No 3 2003
ISSN 0379-0991
EUROPEANECONOMY
EUROPEAN COMMISSIONDIRECTORATE-GENERAL FOR ECONOMIC
AND FINANCIAL AFFAIRS
Public finances in EMU2003
EURO
PEAN
ECON
OM
YN
o 3
2003
Price (excluding VAT) in Luxembourg EUR 45
European Economy (6 issues minimum per year) EUR 145
The annual subscription runs from 1 January to 31 December of each year
Payments to be made only to the agents in the countries listed on page 3 of the cover of European Economy
These are surface mail rates for air subscription rates please apply to the sales offices
httpeuropaeuintcommeconomy_finance10
KC
-AR
-03-003-EN
-C
7IJ2I9-efaifaISBN 92-894-5085-1
BELGIQUEBELGIEuml
Jean De LannoyAvenue du Roi 202Koningslaan 202B-1190 BruxellesBrusselTeacutel (32-2) 538 43 08Fax (32-2) 538 08 41E-mail jeandelannoyinfoboardbeURL httpwwwjean-de-lannoybe
La librairie europeacuteenneDe Europese BoekhandelRue de la Loi 244Wetstraat 244B-1040 BruxellesBrusselTeacutel (32-2) 295 26 39Fax (32-2) 735 08 60E-mail maillibeuropbeURL httpwwwlibeuropbe
Moniteur belgeBelgisch StaatsbladRue de Louvain 40-42Leuvenseweg 40-42B-1000 BruxellesBrusselTeacutel (32-2) 552 22 11Fax (32-2) 511 01 84E-mail eusalesjustfgovbe
DANMARK
J H Schultz Information ASHerstedvang 4DK-2620 AlbertslundTlf (45) 43 63 23 00Fax (45) 43 63 19 69E-mail schultzschultzdkURL httpwwwschultzdk
DEUTSCHLAND
Bundesanzeiger Verlag GmbHVertriebsabteilungAmsterdamer Straszlige 192D-50735 KoumllnTel (49-221) 97 66 80Fax (49-221) 97 66 82 78E-Mail vertriebbundesanzeigerdeURL httpwwwbundesanzeigerde
ELLADAGREECE
G C Eleftheroudakis SAInternational BookstorePanepistimiou 17GR-10564 AthinaTel (30) 21 03 25 84 40Fax (30) 21 03 25 84 99E-mail elebooksbooksgrURL wwwbooksgr
ESPANtildeA
Boletiacuten Oficial del EstadoTrafalgar 27E-28071 MadridTel (34) 915 38 21 11 (libros) 913 84 17 15(suscripcioacuten)Fax (34) 915 38 21 21 (libros) 913 84 17 14(suscripcioacuten)E-mail clientescomboeesURL httpwwwboees
Mundi Prensa Libros SACastelloacute 37E-28001 MadridTel (34) 914 36 37 00Fax (34) 915 75 39 98E-mail libreriamundiprensaesURL httpwwwmundiprensacom
FRANCE
Journal officielService des publications des CE26 rue DesaixF-75727 Paris Cedex 15Teacutel (33) 140 58 77 31Fax (33) 140 58 77 00E-mail europublicationsjournal-offi-cielgouvfrURL httpwwwjournal-officielgouvfr
IRELAND
Alan Hannarsquos Bookshop270 Lower Rathmines RoadDublin 6Tel (353-1) 496 73 98Fax (353-1) 496 02 28E-mail hannasiolie
ITALIA
Licosa SpAVia Duca di Calabria 11Casella postale 552I-50125 FirenzeTel (39) 05 56 48 31Fax (39) 055 64 12 57E-mail licosalicosacomURL httpwwwlicosacom
LUXEMBOURG
Messageries du livre SARL5 rue RaiffeisenL-2411 LuxembourgTeacutel (352) 40 10 20Fax (352) 49 06 61E-mail mailmdlluURL httpwwwmdllu
NEDERLAND
SDU Servicecentrum UitgeversChristoffel Plantijnstraat 2Postbus 200142500 EA Den HaagTel (31-70) 378 98 80Fax (31-70) 378 97 83E-mail sdusdunlURL httpwwwsdunl
PORTUGAL
Distribuidora de Livros Bertrand Lda
Grupo Bertrand SARua das Terras dos Vales 4-AApartado 60037P-2700 AmadoraTel (351) 214 95 87 87Fax (351) 214 96 02 55E-mail dlbippt
Imprensa Nacional-Casa da Moeda SASector de Publicaccedilotildees OficiaisRua da Escola Politeacutecnica 135P-1250 -100 Lisboa CodexTel (351) 213 94 57 00Fax (351) 213 94 57 50E-mail spoceincmptURL httpwwwincmpt
SUOMIFINLAND
Akateeminen KirjakauppaAkademiska BokhandelnKeskuskatu 1Centralgatan 1PLPB 128FIN-00101 HelsinkiHelsingforsPtfn (358-9) 121 44 18Ffax (358-9) 121 44 35Saumlhkoumlposti akatilausakateeminencomURL httpwwwakateeminencom
SVERIGE
BTJ ABTraktorvaumlgen 11-13S-221 82 LundTfn (46-46) 18 00 00Fax (46-46) 30 79 47E-post btjeu-pubbtjseURL httpwwwbtjse
UNITED KINGDOM
The Stationery Office LtdCustomer ServicesPO Box 29Norwich NR3 1GNTel (44-870) 60 05-522Fax (44-870) 60 05-533E-mail bookordersthesocoukURL httpwwwtsocouk
IacuteSLAND
Bokabud Larusar BloumlndalEngjateigi 17-19IS-105 ReykjavikTel (354) 552 55 40Fax (354) 552 55 60E-mail bokabudsimnetis
NORGE
Swets Blackwell ASHans Nielsen Hauges gt 39Boks 4901 NydalenN-0423 OsloTel (47) 23 40 00 00Fax (47) 23 40 00 01E-mail infonoswetsblackwellcom
SCHWEIZSUISSESVIZZERA
Euro Info Center Schweizco OSEC Business Network SwitzerlandStampfenbachstraszlige 85PF 492CH-8035 ZuumlrichTel (41-1) 365 53 15Fax (41-1) 365 54 11E-mail eicsosecchURL httpwwwoseccheics
BLGARIJA
Europress Euromedia Ltd59 blvd VitoshaBG-1000 SofiaTel (359-2) 980 37 66Fax (359-2) 980 42 30E-mail MilenamboxcitbgURL httpwwweuropressbg
CYPRUS
Cyprus Chamber of Commerceand IndustryPO Box 21455CY-1509 NicosiaTel (357-22) 88 97 52Fax (357-22) 66 10 44E-mail staloccciorgcy
EESTI
Eesti Kaubandus-Toumloumlstuskoda(Estonian Chamber of Commerce and Industry)Toom-Kooli 17EE-10130 TallinnTel (372) 646 02 44Fax (372) 646 02 45E-mail einfokodaeeURL httpwwwkodaee
HRVATSKA
Mediatrade LtdStrohalov Prilaz 27HR-10000 ZagrebTel (385-1) 660 08 40Fax (385-1) 660 21 65E-mail mediatradehihinethr
MAGYARORSZAacuteG
Euro Info ServiceSzt Istvaacuten krt12IIl emelet 1APO Box 1039H-1137 BudapestTel (36-1) 329 21 70Fax (36-1) 349 20 53E-mail euroinfoeuroinfohuURL httpwwweuroinfohu
MALTA
Miller Distributors LtdMalta International AirportPO Box 25Luqa LQA 05Tel (356) 21 66 44 88Fax (356) 21 67 67 99E-mail infomillermaltacom
POLSKA
Ars PolonaKrakowskie Przedmiescie 7Skr pocztowa 1001PL-00-950 WarszawaTel (48-22) 826 12 01Fax (48-22) 826 62 40E-mail books119arspolonacompl
ROMAcircNIA
EuromediaStrDionisie Lupu nr 65 sector 1RO-70184 BucurestiTel (40-21) 260 28 82Fax (40-21) 260 27 88E-mail euromediamailcitycom
SLOVAKIA
Centrum VTI SRNaacutemestie Slobody 19SK-81223 Bratislava 1Tel (421-2) 54 41 83 64Fax (421-2) 54 41 83 64E-mail europtbb1cvtisrskURL httpwwwcvtisrsk
SLOVENIJA
GV Zalozba dooDunajska cesta 5SI-1000 LjubljanaTel (386) 13 09 1800Fax (386) 13 09 1805E-mail europgvzalozbasiURL httpwwwgvzalozbasi
TUumlRKIYE
Duumlnya Aktuumlel ASGlobus Duumlnya Basinevi100 Yil Mahallessi 34440TR-80050 Bagcilar-IstanbulTel (90-212) 440 22 27Fax (90-212) 440 23 67E-mail aktuelinfodunyacom
ARGENTINA
World Publications SAAv Coacuterdoba 1877C1120 AAA Buenos AiresTel (54-11) 48 15 81 56Fax (54-11) 48 15 81 56E-mail wpbooksinfoviacomarURL httpwwwwpbookscomar
AUSTRALIA
Hunter PublicationsPO Box 404Abbotsford Victoria 3067Tel (61-3) 94 17 53 61Fax (61-3) 94 19 71 54E-mail admintekimagingcomau
BRASIL
Livraria CamotildeesRua Bittencourt da Silva 12 CCEP20043-900 Rio de JaneiroTel (55-21) 262 47 76Fax (55-21) 262 47 76E-mail livrariacamoesincmcombrURL httpwwwincmcombr
CANADA
Les eacuteditions La Liberteacute Inc
3020 chemin Sainte-FoySainte-Foy Queacutebec G1X 3V6Teacutel (1-418) 658 37 63Fax (1-800) 567 54 49E-mail libertemediomqcca
Renouf Publishing Co Ltd
5369 Chemin Canotek Road Unit 1Ottawa Ontario K1J 9J3Tel (1-613) 745 26 65Fax (1-613) 745 76 60E-mail orderdeptrenoufbookscomURL httpwwwrenoufbookscom
EGYPT
The Middle East Observer
41 Sherif Street11111 CairoTel (20-2) 392 69 19Fax (20-2) 393 97 32E-mail meosoficomcomegURL httpwwwmeobservercomeg
MALAYSIA
EBIC Malaysia
Suite 4701 Level 47Bangunan AmFinance (letter box 47)8 Jalan Yap Kwan Seng50450 Kuala LumpurTel (60-3) 21 62 62 98Fax (60-3) 21 62 61 98E-mail ebictmnetmy
MEacuteXICO
Mundi Prensa Meacutexico SA de CV
Riacuteo Paacutenuco 141Colonia CuauhteacutemocMX-06500 Meacutexico DFTel (52-5) 533 56 58Fax (52-5) 514 67 99E-mail 1015452361compuservecom
SOUTH KOREA
The European Union Chamber ofCommerce in Korea
Suite 2004 Kyobo Bldg1 Chongro 1-Ga Chongro-GuSeoul 110-714Tel (82-2) 725-98805Fax (82-2) 725-9886E-mail eucckeucckorgURL httpwwweucckorg
SRI LANKA
EBIC Sri Lanka
Trans Asia Hotel115 Sir ChittampalamA Gardiner MawathaColombo 2Tel (94-1) 074 71 50 78Fax (94-1) 44 87 79E-mail ebicslsltnetlk
TrsquoAI-WAN
Tycoon Information Inc
PO Box 81-466105 TaipeiTel (886-2) 87 12 88 86Fax (886-2) 87 12 47 47E-mail eiutpems21hinetnet
UNITED STATES OF AMERICA
Bernan Associates
4611-F Assembly DriveLanham MD 20706-4391Tel (1-800) 274 44 47 (toll free telephone)Fax (1-800) 865 34 50 (toll free fax)E-mail querybernancomURL httpwwwbernancom
ANDERE LAumlNDEROTHER COUNTRIESAUTRES PAYS
Bitte wenden Sie sich an ein Buumlro IhrerWahlPlease contact the sales office ofyour choiceVeuillez vous adresser aubureau de vente de votre choix
Office for Official Publicationsof the European Communities2 rue MercierL-2985 LuxembourgTel (352) 29 29-42001Fax (352) 29 29-42700E-mail info-info-opocececeuintURL httppublicationseuint
72003
Venta bull Salg bull Verkauf bull Pvlegraveseiw bull Sales bull Vente bull Vendita bull Verkoop bull Venda bull Myynti bull Foumlrsaumlljninghttpeur-opeuintgeneralens-adhtm
European Economy appears six times a year It contains important reportsand communications from the Commission to the Council and theParliament on the economic situation and developments ranging from theBroad economic policy guidelines and its implementation report to theEconomic forecasts the EU Economic review and the Public financereport As a complement Special reports focus on problems concerningeconomic policy
Subscription terms are shown on the back cover and the address of thesales offices are shown on the inside back cover
Unless otherwise indicated the texts are published under the responsibilityof the Directorate-General for Economic and Financial Affairs of theEuropean Commission BU1 B-1049 Brussels to which enquiries otherthan those related to sales and subscriptions should be addressed
European Commission
EUROPEANECONOMY
Directorate-General for Economic and Financial Affairs
2003 Number 3
copy European Communities 2003
Printed in Belgium
Public finances in EMU mdash 2003
Abbreviations and symbols used
Member States
BE BelgiumDK DenmarkDE GermanyEL GreeceES SpainFR FranceIE IrelandIT ItalyLU LuxembourgNL The NetherlandsAT AustriaPT PortugalFI FinlandSE SwedenUK United KingdomEUR-12 European area Member States currently participating in the monetary union
(BE DE EL ES FR IE IT LU NL AT PT FI)EU-15 European Union 15 Member States (EUR-12 plus DK SE and UK)
Candidate countries
BG BulgariaCY CyprusCZ Czech RepublicEE EstoniaHU HungaryLV LatviaLH LithuaniaMT MaltaPL PolandRO RomaniaSK Slovak RepublicSI SloveniaTR TurkeyAC-10 Accession countries (CY CZ EE HU LV LH MT PL SK SI)CC-13 Candidate countries (AC-10 plus BG RO and TR)
iv
Currencies
EUR euroECU European currency unitDKK Danish kroneGBP Pound sterlingSEK Swedish kronaCAD Canadian dollarCHF Swiss francJPY Japanese yenSUR Russian roubleUSD US dollar
Other abbreviations
bn billion 1 000 millionCPI consumer price indexEC European CommissionECB European Central BankECSC European Coal and Steel CommunityEDF European Development FundEIB European Investment BankEMCF European Monetary Cooperation FundEMS European Monetary SystemEMU economic and monetary unionERM exchange rate mechanismEuratom European Atomic Energy CommunityEurostat Statistical Office of the European CommunitiesFDI foreign direct investmentGDP (GNP) gross domestic (national) productGFCF gross fixed capital formationHICP harmonised index of consumer pricesILO International Labour OrganisationIMF International Monetary FundLDCs less developed countriesMio millionMrd 1 000 millionNCI New Community InstrumentOCTs overseas countries and territoriesOECD Organisation for Economic Cooperation and DevelopmentOPEC Organisation of Petroleum Exporting CountriesPEP Pre-accession economic programmesPPS purchasing power standardSCP Stability and convergence programmesqoq quarter-on-quarter percentage changeSMEs small and medium-sized enterprisesVAT value added taxyoy year-on-year percentage change not availablendash none
v
vi
Acknowledgements
This report was prepared in the Directorate-General for Economic and Financial Affairs under the direction ofKlaus Regling Director-General and Servaas Deroose Director for the Economy of the Euro Area and the UnionThe main contributors were Declan Costello Elena Flores Gual Gabriele Giudice Andrea Montanino AlessandroTurrini and Peter Wierts
Specific contributions were prepared by Riccardo Maggi Joatildeo Nogueira Martins and Jan inrsquot Veld The country chap-ters in Part VI were prepared in the Directorate for the Economies of the Member States under the responsibility ofAntonio Joseacute Cabral The contributors were Ronald Albers Georg Busch Joaquim Ayuso Casals Per EckfeldtFrancesco De Castro Heinz Jansen Ulrich Jocheim Harri Kaumlhkonen Martin Larch Karin Abelskov MadeleineMahovsky Laurent Moulin Joatildeo Medeiros Stellios Pantazidis Lucio Pench Elena Reitano Joseacute Luis Robledo FragaMatteo Salto Mirella Tieleman Charlotte Van Hooydonk Keith Vernon Editorial collaboration was provided byIoanna Soulioti Karine Jabet and Marko Mrsnik The statistical annex was prepared by Maarten Van de Stadt (avail-able on the web see at httpeuropaeuintcommeconomy_financepublicationspublicfinance_enhtm)
Comments and suggestions by colleagues in the Directorate-General for Economic and Financial Affairs as well as byother services of the Commission are gratefully acknowledged
Secretarial support was provided by Maria Davi-Pilato and Aliki Drossou
Comments on the report would be gratefully received and should be sent to
Directorate-General for Economic and Financial AffairsPublic Finances Unit with particular reference to the euro zoneEuropean CommissionB-1049 Brussels
or by e-mail to ElenaFloresceceuint
Contents
Summary and main conclusions 1
Part I Current developments and prospects 9
1 Budgetary developments in the euro area and EU Member States 13
11 Short-term developments and prospects for the budget balance and public debt 13
12 Government revenue and expenditure 16
13 The fiscal stance and policy mix 18
2 Overview of the 2002 updates of the stability and convergence programmes 22
21 The medium-term budget targets 22
22 Composition of the budgetary adjustment 26
3 The sustainability of public finances based on the 2002 updates of stability and convergence programmes 31
31 Introduction 31
32 How the sustainability of public finances was assessed 31
321 The quantitative indicators 31
322 The data used 34
323 The results of the quantitative indicators 36
33 Policy conclusions per Member State 39
4 Budgetary developments in candidate countries 44
41 Short-term budgetary developments and prospects in candidate countries 44
42 Overview of the 2002 updates of the pre-accession economic programmes 47
421 Introduction 47
422 Medium-term budgetary developments 48
423 Composition of the adjustment 51
424 Other considerations 52
Part II Evolving budgetary surveillance 55
1 Implementing the Stability and Growth Pact 59
11 Introduction 59
12 The enforcement mechanisms of the SGP 59
121 The preventive part of the Pact 59
122 The dissuasive elements of the Pact 60
13 The use of enforcement mechanisms since spring 2002 62
131 Slippage from budget targets in many Member States 62
132 Portugal 63
133 Germany 65
134 France 67
2 Strengthening the coordination of budgetary policies 69
21 Background to the debate a mandate from the Barcelona European Council 69
vii
22 Commission proposals to strengthen the coordination of budgetary policies 71
221 A diagnosis of the shortcomings of the SGP in the first four years of EMU 71
222 Avoiding pro-cyclical policies and accounting for transitory elements in the assessment 72
223 A minimum annual rate of adjustment for countries still in deficit 75
224 The goals of the Lisbon strategy ensuring that public finances contribute to growth and employment 76
225 Ensuring the sustainability of public finances 78
226 Concrete measures for the enforcement of the Pact 78
23 The agreement of the European Council on strengthening the coordination of budgetary policies 78
3 Public debt and the excessive deficit procedure 80
31 Introduction 80
32 Debt dynamics and compliance with the Treaty requirements 80
33 Debt dynamics in EU countries 81
34 What could constitute a satisfactory pace of debt reduction 82
4 The governance of budgetary statistics in EMU 87
41 Introduction 87
42 The governance of budgetary statistics in the EU 87
421 Main elements 87
422 Other aspects of the governance of budgetary statistics 88
43 Assessing the quality of budgetary statistics 90
431 Reliability 90
432 Transparency and consistency 91
433 Timeliness 91
44 Recent measures to improve the quality of budgetary statistics 92
441 The code of best practice 92
442 Towards quarterly accounts 92
45 Conclusion and challenges for the future 93
Annex A Budgetary surveillance for long-term sustainability in EU Member States 95
Part III Public investment and its interaction with the EUrsquos budgetary rules 99
1 Introduction 103
2 Public investment definition and broad trends 104
21 The definition of public investment 104
22 Broad trends of public investment in industrialised countries 104
3 Public investment its rationale and impact on efficiency 107
31 The rationale for public investment 107
32 Public investment productivity and growth the empirical evidence 108
4 A closer look at public investment in Member States and the interaction with the EU fiscal rules 112
41 The evolution of public and private investment in EU countries 112
411 Trends in recent decades 112
412 Is there a link between changing levels of public and private investment 115
42 Budgetary consolidation in light of EMU and its impact on public investment 115
viii
5 Catering for public investment needs in the Stability and Growth Pact 122
51 How public investment is treated under the existing Treaty and SGP rules 122
52 Public investment and the golden rule 123
521 A rationale for the golden rule 123
522 Limitations and drawbacks 124
523 Practical experiences 125
524 Why a golden rule would not be desirable for EMU 125
53 Public-private partnerships 128
531 Definition taxonomy and recent experiences 128
532 The economics of PPPs 129
533 Publicndashprivate partnerships and budgetary practices in EMU 130
Part IV Can fiscal consolidations in EMU be expansionary 133
1 Introduction 137
2 Can budgetary consolidations be expansionary What the theory says 138
21 Budgetary consolidations the standard view 138
22 Non-Keynesian effects of fiscal consolidation 139
3 Characteristics and effects of fiscal consolidations in the EU evidence from cross-country analysis 143
31 Survey of existing studies 143
32 Were there expansionary fiscal consolidations in the EU A close look at the data 145
321 How to define periods of budgetary consolidation with expansionary effects 145
322 When does a fiscal consolidation occur 147
323 When is a fiscal consolidation expansionary 149
324 Summary of findings 154
4 Assessing ex-ante the effects of fiscal consolidations Simulation results from the QUEST model 158
41 Introduction 158
42 Tax increases 159
43 Expenditure cuts 159
44 Summary of findings 163
Part V Meeting the EUrsquos budgetary requirements national expenditure rules and fiscal relations across levels of government 167
1 Introduction 171
2 Expenditure rules in EU Member States 172
21 The need for expenditure rules as a means to control public finances 172
22 The design and implementation of expenditure rules 174
221 The design of expenditure rules 174
222 The implementation of expenditure rules 176
223 A taxonomy of expenditure rules 177
23 National expenditure rules 177
231 Main features of expenditure rules within EU Member States 177
232 How have national expenditure rules worked in practice 178
ix
24 Conclusions 186
3 Fiscal relations across levels of government 187
31 Fiscal relations across different levels of government in EU Member States 187
32 Fiscal decentralisation and its interaction with the EUrsquos fiscal rules 190
321 Fiscal decentralisation and the goal of sound and sustainable public finances 190
322 Recent measures in several Member States to coordinate budgetary positions across levels of government in light of EU requirements 193
33 Fiscal decentralisation and automatic stabilisation 195
34 Case studies 199
341 Spain 199
342 Germany 202
Part VI Member State developments 205
1 Belgium 207
2 Denmark 210
3 Germany 214
4 Greece 217
5 Spain 221
6 France 225
7 Ireland 228
8 Italy 232
9 Luxembourg 237
10 The Netherlands 240
11 Austria 243
12 Portugal 247
13 Finland 250
14 Sweden 256
15 United Kingdom 260
Part VII Resources 265
1 Code of best practice on the compilation and reporting of data in the context of the excessive deficit procedure 267
11 Compilation of budgetary data by Member States 267
12 Reporting of budgetary data by Member States to the Commission 267
13 Securing the quality of the actual budgetary data 268
14 Publication of the budgetary data by the Commission 269
2 Glossary 270
3 References 275
4 Useful Internet links 282
Statistical annex 285
x
Tables
I1 General government budgetary position mdash euro area 1999ndash2004 ( of GDP) 14
I2 Budget balances in EU Member States 2001ndash04 ( of GDP) 14
I3 Composition of changes in government debt ratio in EU Member States 2001ndash04 ( of GDP) 15
I4 Euro area government resources and expenditures 2000ndash04 ( of GDP) 16
I5 Total revenue and expenditure in EU Member States 2001ndash04 ( of GDP) 17
I6 Euro area mdash Growth projections and macroeconomic developments in the 2002 updates and comparison with the 2001 updates and the Commission forecasts 23
I7 GDP growth projections in the 2002 updates 23
I8 Actual budget balances in the 2002 updates and in the Commission forecasts in of GDP 24
I9 Cyclically-adjusted balances in the 2002 updates and in the Commission forecasts
on the basis of the production function method in of GDP 25
I10 Euro area net lending by sub-sectors in the 2002 updates 26
I11 Euro area mdash Gross debt level and changes in the 2001 updates ( of GDP) 27
I12 Debt levels in the 2002 updates of the stability and convergence programmes ( of GDP) 27
I13 Expenditure and revenue ratios in the 2002 updates 28
I14 Euro area Budget developments for the general government 29
I15 Long-run budgetary projections included in the 2002 updates to stability and convergence programmes ( of GDP) 35
I16 Data used to run the sustainability indicators in the lsquoSGP compliance scenariorsquo ( of GDP) 36
I17 Projected evolution of debt levels up to 2050 37
I18 Results of the tax gap indicator 38
I19 Policy conclusions on the sustainability of public finances 41
I20 General government balances in candidate countries ( of GDP) 46
I21 GDP growth in candidate countries ( pa) 46
I22 Macroeconomic projections in the 2002 PEPs 49
I23 General government balances in the 2002 PEPs ( of GDP) 49
I24 General government debt in the 2002 PEPs ( of GDP) 50
I25 General government revenue and expenditure in the 2002 PEPs ( of GDP) 50
I26 Composition of general government expenditure in the 2002 PEPs ( of GDP) 51
I27 Main measures in the PEPs concerning pension reform 53
II1 Comparison of growth and budgetary developments for 2002 between autumn 2002 Commission forecasts and the 2001 updates of the programmes 62
II2 Breakdown of revision of 2001 budget balance of Portugal 64
II3 Average annual percentage change of public debt to GDP ratios 82
II4 Development in debt levels in several EU high debt countries since the mid-1990s 83
II5 Debt dynamic according to different budget balances and nominal GDP growth rates (initial government debt to GDP ratio 100) 84
II6 The implied rate of debt reduction by a constant primary surplus 84
II7 The implied primary surplus by defining a rate of reduction of the debt ratio 85
xi
III1 The effect of public investment on output productivity and growth 111
III2 Public and private investment Granger causality tests 117
III3 The composition of fiscal consolidations general government (1970ndash2002) 119
III4 The determinants of public investment in the EU regression analysis (EU-15 1970ndash2002) 119
IV1 Some puzzling effects of fiscal policy 140
IV2 Cross-country evidence on fiscal consolidations 144
IV3 Expansionary consolidations description of episodes with alternative definitions of consolidation 148
IV4 Size and composition of expansionary consolidations alternative definitions of consolidation 149
IV5 Macroeconomic environment in expansionary consolidations alternative definitions of consolidation 150
IV6 Expansionary consolidations description of episodes with alternative definitions of expansion 151
IV7 Correlation indexes among alternative indicators of expansionary consolidations 152
IV8 Size and composition of expansionary consolidations alternative definitions of expansion 152
IV9 Macroeconomic scenario in expansionary consolidations alternative definitions of expansion 153
IV10 Permanent increase in labour income tax of 1 of GDP 160
IV11 Permanent increase in corporate tax of of 1 of GDP 160
IV12 Permanent increase in VAT of 1 of GDP 161
IV13 Permanent reduction in government purchases of 1 of GDP 161
IV14 Permanent reduction in government transfers to households of 1 of GDP 162
IV15 Permanent reduction in spending on government employment of 1 of GDP 162
IV16 Temporary expenditure cuts (1 of GDP) 163
IV17 Permanent expenditure cuts (1 of GDP) with accommodating monetary stance 164
V1 Trends in public expenditure items in selected EU countries (variation in percentage points between 1998 and 2001) 174
V2 A taxonomy of expenditure rules 177
V3 The features and implementation of expenditure rules within Member States (general targets) 179
V4 The impact of expenditure rules on spending trends 185
V5 Total expenditure targets and spending rules 185
V6 Expenditure and revenues at State and local government level 189
V7 Sub-national government spending by function as a percentage of total local spending 191
V8 The composition of total revenues at state and local level as a percentage of GDP (year 2000) 191
V9 Aggregate budget balances at state level (A B E D) local level (DK FIN S) and output gaps 196
VI1 Composition and balances of general government Belgium (as of GDP) 207
VI2 Key figures of the Belgian stability programme (2003ndash05) 209
VI3 Composition and balances of general government Denmark (as of GDP) 211
VI4 Key figures of the Danish convergence programme (2001ndash06) 212
VI5 Composition and balances of general government Germany (as of GDP) 215
VI6 Key figures of the German stability programme (2002ndash06) 216
VI7 Composition and balances of general government Greece (as of GDP) 218
xii
VI8 Key figures of the Greek stability programme (2001ndash06) 219
VI9 Composition and balances of general government Spain (as of GDP) 222
VI10 Key figures of the Spanish stability programme (2002ndash06) 222
VI11 Main features of recent tax reforms 224
VI12 Composition and balances of general government France (as of GDP) 226
VI13 Key figures of the French stability programme (2002ndash06) 226
VI14 Composition and balances of general government Ireland (as of GDP) 229
VI15 Key figures of the Irish stability programme (2003ndash05) 229
VI16 Composition and balances of general government Italy (as of GDP) 233
VI17 Key figures of the Italian stability programme (2002ndash06) 233
VI18 Composition and balances of general government Luxembourg (as of GDP) 238
VI19 Key figures of the Luxembourg stability programme (2001ndash05) 238
VI20 Composition and balances of general government Netherlands (as of GDP) 241
VI21 Expenditure in the Netherlands mdash relevant ceilings and outcome 242
VI22 Composition and balances of general government Austria (as of GDP) 244
VI23 Key figures of the Austrian stability programme (2002ndash06) 244
VI24 Composition and balances of general government Portugal (as of GDP) 248
VI25 Key figures of the Portuguese stability programme (2003ndash06) 248
VI26 Composition and balances of general government Finland (as of GDP) 251
VI27 Key figures of the Finnish stability programme (2002ndash06) 251
VI28 Difference of budgetary outcomes and expenditure ceilings (excluding interest payment) 253
VI29 Central government expenditure frames 1994ndash2003 253
VI30 Central government expenditure frames 2004ndash07 by the ministeries 255
VI31 Composition and balances of general government Sweden (as of GDP) 258
VI32 Key figures of the Swedish convergence programme (2001ndash04) 258
VI33 Central government expenditure in of GDP 259
VI34 Composition and balances of general government United Kingdom (as of GDP) 261
VI35 Key figures of the United Kingdomrsquos stability programme (2001ndash02 to 2006ndash07) 261
Boxes
I1 The impact of ageing populations on tax revenues and social contributions 33
I2 Candidate countriesrsquo budgetary data and EU standards 45
I3 The pre-accession fiscal surveillance procedure for candidate countries 47
II1 What constitutes an rsquoexceptional circumstancersquo under the excessive deficit procedure 61
II2 The Convention on the Future of Europe the debate on the coordination of budgetary policies 70
II3 Transitory elements affecting the budgetary position 74
III1 Public investment in national account statistics 105
III2 Empirical evidence on the effects of public investment methodologies and results 109
xiii
III3 Public and private investment in EU countries crowding in or crowding out 116
III4 The determinants of public investment in the EU an empirical analysis 120
III5 How would the introduction of a golden rule modify the fiscal architecture of EMU 126
IV1 Expansionary fiscal consolidations and confidence indicators 155
IV2 Fiscal consolidation as catalyst for structural reforms in Germany 165
V1 Key arguments of the theory on fiscal federalism 188
Graphs
I1 Euro area fiscal stance and cyclical conditions 2000ndash04 17
I2 Policy-mix in the euro area 1999ndash02 19
I3 Fiscal stance and cyclical conditions in EU Member States in 2002 20
I4 Policy-mix in EU Member States in 2002 20
I5 Fiscal stance and cyclical conditions in EU Member States in 2003 21
I6 Potential and real GDP growth rate and output gaps for the euro area derived from the 2002 updates 22
I7 Contributions to change in budgetary position 2002ndash05 (in points of GDP) 30
I8 A comparison of debt projections for the EU-15 based on the lsquoSGP compliance scenariorsquo and the lsquo2002 starting positionrsquo scenario 38
I9 A comparison of debt projections for four Member States based on the lsquoSGP compliance scenariorsquo and the lsquo2002 positionrsquo scenario 39
I10 Contributions to change in budgetary position 2001ndash05 according to the 2002 PEPs (in points of GDP) 52
II1 Budgetary divergence from target in Portugal 63
II2 Budgetary divergence from target in Germany 66
II3 Budgetary divergence from target in France 68
II4 Compliance with the lsquoclose to balance or in surplus requirement for countries that completed the transition process 73
II5 The budgetary adjustment path of Member States still in transition to the lsquoclose to balance or in surplusrsquo objective 76
II6 Illustration of a small temporary deviation to cater for the intertemporal budgetary effect of a large structural reform 77
II7 Pace of adjustment for the debt ratio 86
III1 Gross fixed capital formation general government of GDP at current market prices 105
III2 Gross public private and total investment of GDP average values for the period 1970ndash2002 113
III3 Net public private and total investment of GDP average values over the period 1974ndash2001 113
III4 Average annual changes in investment shares (1970ndash2002) 114
III5 Cross-country relations between growth rates in public and private investment (average yearly growth rates over the period 1970ndash2002) 114
III6 Interest expenditure and public investment EU-15 1970ndash2002 118
xiv
III7 Public invest changes in the 1990s (average yearly growth rate in gross fixed capital formation general government of GDP at current market prices) 118
III8 Nominal budgets structural budgets the SGP and the golden rule 127
IV1 Economic sentiment indicators during expansionary consolidations 155
IV2 Expenditure-based fiscal consolidations and structural reforms effects on GDP ( change from baseline) 165
V1 Trends in different items of public expenditure at EU level 173
V2 Fiscal decentralisation and budgetary situation 192
V3 The contribution of lower levels of government to general government (net lending (ndash) or borrowing (+) in 2002) 193
V4 Germany output gap and changes in the level of real revenues and expenditure at State level 197
V5 Spain output gap and changes in the level of real revenues and expenditure at State level 198
V6 Denmark output gap and changes in the level of real revenues and expenditure at local level 198
xv
Summary and main conclusions (1)
The most difficult period for budgetary policies since the launch of the euro1
The year 2002 and the early part of 2003 has been a dif-ficult period both in terms of actual budgetary develop-ments and as regards the implementation of the EUframework for fiscal surveillance The nominal deficitfor the euro area as a whole increased from 16 ofGDP in 2001 to 22 in 2002 and according to the lat-est Commission forecast it is projected to rise to 25 of GDP in 2003 This aggregate outcome is the result ofstriking contrasts in the performance across Member StatesBy the end of 2002 only six EU countries including foureuro area countries (accounting for some 18 of euroarea output) had achieved budget positions (both in nom-inal and cyclically-adjusted terms) that met the lsquoclose tobalance or in surplusrsquo requirement of the Stability andGrowth Pact whereas two euro area countries (account-ing for half of the euro area output) had deficits abovethe 3 of GDP reference value
The Portuguese authorities succeeded in reducing thenominal deficit from 41 of GDP in 2001 to 28 in2002 although very significant challenges remain if thedeficit is to remain below 3 of GDP in 2003 as muchof this improvement is due to one-off measures whichhave only led to a transitory improvement in the budgetbalance A deficit of 36 of GDP in 2002 has resultedin Germany being placed in an excessive deficit posi-tion while the authorities are taking measures aimed atreducing the cyclically-adjusted budget deficit only avery limited improvement in nominal terms is expectedin 2003 as growth conditions deteriorate Despite clearevidence of budgetary slippage emerging in early 2002the French authorities did not take corrective measuresand a deficit of 31 of GDP occurred in 2002 resultingin the excessive deficit procedure being activated An
even higher deficit of 37 of GDP is forecast by theCommission services for 2003 on the basis of currentpolicies Large deficits remain in Italy (23 of GDPin 2002 and in 2003) and by 2004 are projected to riseabove the 3 of GDP reference value (2) budgetaryconsolidation efforts in Italy continue to rely on one-offmeasures rather than on reforms of a structural natureneeded to ensure a permanent improvement in thebudget balance Deficits have also re-emerged in 2002 incountries that had already reached balanced budget posi-tions notably Austria (06 of GDP) the Netherlands(11 ) and the UK (13 )
Higher nominal deficits are only partly due to the economic cycle
At first sight these developments compare relativelyfavourably with previous economic downturns when def-icits reached much higher levels and debt ratios enteredrapidly increasing trajectories In addition governmentshave not pursued fine-tuning policies and while fiscalpolicies were slightly looser monetary conditions haveeased thanks mainly to low real interest rates
However a closer consideration of underlying budgetarytrends reveals that the deterioration in nominal deficitsalso results from high and rising cyclically-adjusted def-icits in several countries This indicates a discretionaryloosening of the fiscal stance by some Member Statesover the past two years brought about by a combinationof unfunded tax cuts discretionary expenditure increasesand failures as regards budgetary execution While theoutcome of the euro area in 2002 was unchanged com-pared to 2001 it should be noted that the cyclically-adjusted budget balance for 2001 has recently beenrevised upwards to 21 of GDP from 15 of GDPimplying that the deterioration in the underlying budgetbalance in that year was considerably worse than earlier
yen1part The summary and main conclusions of this report have been adopted bythe College of Commissioners in the form of a communication from theCommission to the Council and the European Parliament lsquoPublic financesin EMU mdash 2003rsquo COM(2003) 283 adopted on 21 May 2003
yen2part European Commission spring 2003 forecast 2004 figures are based on theassumption of no policy change
1
P u b l i c f i n a n c e s i n E M U 2 0 0 3
estimates showed moreover the cyclically-adjustedbudget balance includes the impact of one-off budgetarymeasures which only have a transitory effect on budgetpositions The deterioration has been particularly pro-nounced in Germany (where the CAB increased to 32 of GDP in 2002) and France (to 33 ) In Italy it remainshigh at 21 of GDP
In a medium-term perspective the latest updates of thestability and convergence programmes contain a targetby most Member States to reach budget positions oflsquoclose to balance or in surplusrsquo by 2005 or 2006 How-ever it should be noted that the medium-term targets ofMember States are based on growth assumptions whichin light of developments in recent months now appearto be optimistic In countries where large cyclically-adjusted deficits remain the time frame for reaching thelsquoclose to balance or in surplusrsquo objective has beenpushed back to 2006 or 2007 even this date will only bemet if additional consolidation measures are undertaken
Commission proposals to strengthen the coordination of budgetary policies
The deterioration in budget positions has placed consid-erable stress on the EUrsquos framework for fiscal surveil-lance and three Member States have been placed inexcessive deficit positions In response to these develop-ments and in line with a mandate from the BarcelonaEuropean Council conclusions the Commission adopteda communication on strengthening the coordination ofbudgetary policies (1) It identified a number of short-comings with the implementation of the SGP in the firstfour years of EMU and outlined a strategy based onMember States reassuming political ownership of thePact Inter alia it called for more account to be taken ofunderlying economic conditions when assessing budget-ary positions an interpretation of compliance with SGPrequirements that would (depending on country-specificcircumstances) cater for the budgetary impact of reformsthat enhance growth and employment increasing theemphasis placed on the sustainability of public financesand outstanding debt positions and improving theimplementation of the SGP including stricter and moretimely recourse to the existing enforcement instrumentsAt the same time the Commission adopted proposals toimprove the governance of budgetary statistics whichprovide the foundations for effective surveillance
The European Council of March 2003 endorsed key conclusions of the Ecofin Council
The spring European Council of March 2003 endorseda report of the (Ecofin) Council which shared many ofthe Commissionrsquos proposals on strengthening the coor-dination of budgetary policies It confirmed that theachievement of a budget position of lsquoclose to balance orin surplusrsquo is in the economic self-interest of MemberStates both individually and collectively In the shortrun it provides room for the automatic stabilisers tooperate freely and cushion the effect of economicshocks in the medium run it creates room for budgetarymanoeuvre to either cut taxes or divert expenditures tomore productive items such as investment and RampD inthe long run compliance will help Member States meetthe budgetary costs of ageing population while securingadequate and accessible pensions and healthcare
In addition to re-stating their commitment to the goal ofthe SGP the Council agreed that compliance with thelsquoclose to balance or in surplusrsquo requirement should beassessed in cyclically-adjusted terms with due accounttaken of one-off budgetary measures which only have atransitory impact on budget positions For euro-areacountries agreement was reached that Member Stateswith deficits should achieve an annual improvement inthe cyclically-adjusted budget deficit of at least 05 ofGDP until the lsquoclose to balance or in surplusrsquo require-ment is reached It underlined the need for automatic sta-bilisers to operate symmetrically over the economiccycle and the particular importance of avoiding a pro-cyclical loosening of fiscal policies in good times TheCouncil also confirmed the importance of running downpublic debt at a satisfactory pace towards the 60 ofGDP reference value and that the existing provisions ofthe Treaty (ie the debt criterion of the excessive deficitprocedure) can contribute to achieving this goal
An opportunity to ensure consistent and transparent budgetary strategies
To ensure that the agreement of the European Councilrepresents a real progress towards a consistent and trans-parent implementation of SGP it is essential that the pol-icy guidelines endorsed by the European Council andthe specific budgetary commitments given by MemberStates in their updated stability and convergence pro-gramme are respected
To this end policies adopted at national level need torespect the budgetary goals agreed at EU level Indoing so budgetary consolidation strategies need to be
yen1part Communication from the Commission lsquoStrengthening the coordination ofbudgetary policiesrsquo COM(2002) 668 final of 27 November 2002
2
S u m m a r y a n d m a i n c o n c l u s i o n s
designed in a way that tackle and do not exacerbatestructural weaknesses leading to slow growth and missedemployment opportunities This requires careful designas regards the balance between measures on the revenueand expenditure side and choices on the composition ofpublic expenditures Contrary to what is often arguedthe existing framework for budgetary surveillance cansimultaneously achieve a consistent approach that bal-ances the need for budgetary consolidation re-ignitingthe recovery and strengthening growth potential
Significant advances have been made in the framework for budgetary surveillance
This yearrsquos report on Public finances in EMU mdash 2003highlights three areas where substantial progress hasbeen made in the framework for budgetary surveillanceover the past year (i) the integration of candidate coun-tries into the EUrsquos fiscal surveillance framework (ii) anincreased focus on the sustainability of public financesand (iii) an improvement in the governance of budgetarystatistics These advances show that tangible progresscan be made to the benefit of Member States and the EUas whole when there is a political will to do so It alsoshows that the framework for budgetary surveillance iscapable of evolving in the light of growing experienceand new policy challenges
Integrating acceding and candidate countries into the EUrsquos fiscal surveillance framework
With 10 countries set to join the EU in 2004 a majorpolicy challenge is to prepare for their integration into theEU economic policy framework in particular for budget-ary surveillance A key requirement has been to developreliable government accounts and economic forecasts ona par with existing EU countries At the same time theEU surveillance of budgetary developments needs todevelop so that appropriate account is taken of the impor-tant structural and institutional changes underway inaccession countries These are partly due to the comple-tion of the transition from a command to a market econ-omy and partly due to the additional effects which EUmembership will entail (associated with the need toupgrade public infrastructure and the commitment toimplement the acquis communautaire)
Clear strides have been taken in recent years althoughbudgetary data are still neither fully comparable acrosscountries nor completely in line with EU definitionsData reported by the candidate countries and forecastsprepared by the Commission services indicate that budg-etary developments are closely mirroring those in the
EU with nominal and cyclically-adjusted budget deficitsin 2002 rising in most countries Looking ahead to 2003and 2004 the Commission forecast of spring 2003 envis-ages an improvement in the budgetary balances of ninecountries with marked deficit reductions forecasted inHungary Slovakia and Turkey and to a more limitedextent in Malta However very limited improvements inbudget balances are projected in the Czech RepublicPoland and Cyprus
An important step to integrate the candidate countriesinto the existing surveillance process was completed inNovember 2002 when the second set of pre-accessioneconomic programmes (PEPs) submitted by candidatecountries were examined The annual programmes out-line the medium-term policy framework including pub-lic finance objectives and structural reform prioritiesand moreover provide an opportunity for candidatecountries to develop their institutional and analyticalcapacity The 2002 updates revealed an improved effortto develop a consistent and credible medium-termmacroeconomic framework although further analyticalcapacity building is called for
The sustainability of public finances received increased prominence in the assessment of sustainability and convergence programmes
Progress has also been made as regards placingincreased emphasis on the sustainability of publicfinances in the SGP as requested by the 2001 StockholmEuropean Council For the second time an assessment ofthe sustainability of public finances was carried out onthe basis of budgetary targets and measures announcedin the 2002 updates to stability and convergence pro-grammes leading to firm policy conclusions by theCouncil The policy conclusions which are based onquantitative indicators and long-run budgetary projec-tions prepared by the Economic Policy Committee andnational authorities are worrying
Even assuming that all Member States achieve thebudget targets for 2006 set down in their stability or con-vergence programmes there is a risk of unsustainablepublic finances emerging in about half the EU MemberStates especially Belgium Germany Greece SpainFrance Italy Austria and Portugal To ensure sustaina-ble public finances Member States with deficits firstneed to achieve and sustain the SGP goal of budget posi-tions of lsquoclose to balance or in surplusrsquo Furthermorepreliminary estimates by the Commission show that anadditional permanent budgetary adjustment of between
3
P u b l i c f i n a n c e s i n E M U 2 0 0 3
1 and 2 percentage points of GDP is needed in MemberStates where the sustainability of public finances is aconcern To close this financing gap governmentsshould try to avoid raising taxes (especially on labour)and concentrate efforts on reducing (in terms of ratio toGDP) age-related expenditure by reforming of pensionand healthcare systems andor reducing non-age-relatedprimary spending while increasing employment ratesand fostering growth
Progress has been made on the governance of budgetary statistics
The quality of economic statistics is crucial to ensure anadequate understanding of the economic situation andeffective policy making Budgetary statistics are thefoundation of the EU fiscal surveillance tools and theirquality has improved considerably over the last decadeGovernment accounts are now more reliable completetransparent and detailed and are published in a muchmore timely fashion than when the excessive deficit pro-cedure was set up However some weaknesses remainin several countries data on government deficit and debtratios are not yet as reliable as they should be and aresubject to large revisions Furthermore the governmentaccounts of several Member States are not fully transpar-ent and there have been problems in terms of theirtimely submission These concerns are clearly amplifiedwith the perspective of enlargement
To address outstanding challenges the (Ecofin) Councilrecently agreed to implement a code of best practice (1)From the Member Statesrsquo side this involves increasingthe transparency of government accounts in particularfor the lower government subsectors the strict respect ofdeadlines an overall increase in the data quality but alsoa clarification of the independence statute of the nationalstatistical offices as the main compilers of governmentdata The Commission (Eurostat) is aiming at reinforc-ing its ability to scrutinise the Member Statesrsquo govern-ment accounts in more detail and accelerating thedecision-making process for deciding upon the record-ing of government transactions The new steps to com-pile quarterly budgetary statistics is a major challengefor statisticians but also for economists policy-makersand budgetary policy analysts that will need to interpretquarterly data with due care since these will necessarily
be more volatile and perhaps less transparent than annualdata
The Commission role in upgrading the analysis of economic and budgetary policies
In its communication on strengthening the coordinationof budgetary policies the Commission committed itselfto upgrading the analysis of economic and budgetarypolicies To this end a number of detailed studies arecontained in the report Public finances in EMU mdash 2003as follows
bull Firstly the report examines the impact of budgetaryconsolidation on growth It considers whether theassertion that budgetary consolidation has a nega-tive impact on output is always valid or whether fis-cal consolidations in EMU under certain conditionscan have a positive effect on output
bull Secondly and as part of the effort to focus on thequality of public finances the report analyses publicinvestment It examines the reasons why publicinvestment as a share of GDP has fallen in recentdecades and whether this is in part due to the processof budgetary consolidation and the development offiscal rules at EU level It also analyses the linkbetween public investment and productivity andconsiders the merits and feasibility of developingspecific provisions for public investment within theEUrsquos framework for budgetary surveillance
bull A third chapter examines various aspects of the chal-lenge facing national authorities in ensuring soundpublic finances It reviews the experience of Mem-ber States in using expenditure rules as an instru-ment to better manage public finances and improvetheir quality In addition the chapter examines howthe allocation of public finance functions across dif-ferent levels of governments influences the capacityof Member States to fulfil their budgetary commit-ments at EU level This analysis is a good exampleof the role of the Commission in undertaking com-parative cross-country analyses that enable MemberStates to learn from the experiences and best prac-tices of other countries
Is fiscal consolidation always contractionary
While there is a broad consensus among both academicsand policy-makers on the need for fiscal discipline toensure the smooth functioning of EMU and provide
yen1part Conclusions of the 2 485th Council meeting Economic and FinancialAffairs Brussels 18 February 2003
4
S u m m a r y a n d m a i n c o n c l u s i o n s
conditions conducive to growth and employment crea-tion concerns have been expressed that budgetaryconsolidation could have a negative effect on output inthe short run This issue is relevant given the need forseveral Member States to reduce large cyclically-adjusted budget deficits especially against the currentbackground of slow economic growth
An empirical analysis of the experiences of EU MemberStates however demonstrates that roughly half of theepisodes of fiscal consolidation undertaken in the pastthree decades have been accompanied by an accelerationin economic growth These findings appear to be consist-ent with theories that identify a positive impact of budg-etary consolidation on consumer expectations of lowertaxes in the future inducing them to raise their consump-tion plans andor on business expectations of higherprofitability enabling them to raise investment Confi-dence factors may play a more prominent role in thefuture in the light of large unfunded pension liabilities
Simulations using the QUEST model confirm that ifappropriately designed budgetary consolidation cancontribute significantly to the goal of the Lisbon strategyin terms of raising output and employment in themedium term Budgetary consolidation has a slightcontractionary effect on output in the short run depend-ing on the composition of the budgetary adjustmentHowever budgetary consolidation has a positive impacton output in the medium run if it takes place in the formof expenditure retrenchment rather than tax increasesMoreover the effect of budgetary consolidation on out-put could be reinforced and even positive in the shortrun if fiscal consolidation is combined with structuralreform of factor and product markets and accompaniedwith an accommodating monetary stance Indeed budg-etary consolidation often acts as a catalyst for structuralreforms
Public investment
Public investment as a share of GDP has fallen in mostindustrialised countries in recent decades It has beenclaimed that the budgetary requirements of the Treatyand SGP result in public investment expenditures beingat excessively low levels and that a sustained growth inpublic investment expenditures would improve the EUrsquosgrowth potential However an analysis shows that thedecline in public investment rates is a long-run tendencythat started already in the 1970s and affected all indus-trialised countries and not just EU Member StatesDeclining levels of public investment as a share of GDP
have been attributed to factors such as increased levels ofeconomic development (with developed countries alreadyhaving a high stock of physical capital and the emphasisswitching towards investment in human capital (1)) andthe changing boundaries between public and privateinvestment (in part linked to the process of privatisa-tion) Some of the decline in public investment levelsappears to be related to efforts to consolidate publicfinances which was necessary irrespective of EMU Acareful analysis of the data however fails to show anyclear-cut link between change in investment ratios andthe provisions of the EUrsquos framework for fiscal surveil-lance Indeed public investment expenditures in manyMember States have stopped falling after the beginningof monetary union
Public investment can make an important contribution tomeet the output and employment goals of the Lisbonstrategy However in considering the links between pub-lic investment and growth it is important to focus on netas opposed to gross investment levels (that is takingaccount of the depreciation of the existing capital stock)and also the interaction between trends in public and pri-vate investment level Existing studies reveal that publicinvestment has a positive impact on output and produc-tivity although the results are not very strong This isexplained by the fact that only a fraction of public invest-ment expenditures are devoted to projects which aim atdirectly raising productivity (for example investment intransport infrastructure) whereas a significant propor-tion of public investment is devoted to projects that pur-sue other objectives such as environmental protection orredistribution across regions which have an indirectcontribution to productivity
The important role of public investment is recognised inthe existing framework for budgetary surveillance forexample Member States are required to specify plannedpublic investment levels in their annual updates to stabil-ity and convergence programmes and the BEPGs fre-quently recommend that an increased share of publicexpenditures be devoted to productive items In brief thebudget balance requirements of Treaty and SGP arecompatible with a high share of public spending beingdevoted to public investment The recent Commissioncommunication on strengthening the coordination ofbudgetary polices sought to cater for the budgetary
yen1part Communication from the Commission lsquoInvesting efficiently in educationand training an imperative for Europersquo COM(2002) 779
5
P u b l i c f i n a n c e s i n E M U 2 0 0 3
impact of large investment projects while at the sametime respecting the commitment to sound and sustaina-ble public finances (1)
Several calls have been made to introduce a so-calledgolden rule into the SGP which would allow govern-ments to borrow to finance investment However thereare strong theoretical and practical arguments against itsintroduction especially in a framework of multilateralsurveillance such as the SGP First a golden rule basedon a national accounts system could lead to a bias inexpenditure decisions in favour of physical capital andagainst spending on human capital (education training)or other productive items (healthcare RampD) which alsocontribute to growth and employment Secondly ifapplied to gross investment depending on the specificdesign and implementation of the reform the adoption ofa golden rule into the SGP framework may imply sub-stantially higher deficits thus compromising the objec-tive of sustainability of public finances Finally the rel-evant concept for the application of the golden rulewould be net investment However it is not always pos-sible to compute reliable comparable and timely data onthis type of investment
There is a growing practice of financing public purposeinvestment projects through publicndashprivate partnerships(PPPs) A large share of the PPPs in the EU financeinfrastructure and supplement public investment (2) Themain implication for public finances of choosing PPPs asopposed to traditional public investment is in fact thatof converting up-front fixed expenditures into a streamof future obligations This practice has a sound micro-economic rationale in that it can lead to increased effi-ciency without compromising public objectives It isimportant however to avoid recourse to PPPs wherethis is solely motivated by a desire to bypass budgetaryconstraints by putting capital spending outside govern-ment budgets This could lead to PPP projects whichentail higher overall costs which would not be in linewith the objective of sustainable public finances Effortsare also required to ensure transparency in nationalaccounts
Efforts at national level to meet EU budgetary requirements expenditure rules and fiscal relations across different levels of governments
Many Member States in recent years have introducedexpenditure rules as a means to improve the manage-ment of their public finances mostly in the form ofex ante targets rather than binding legal obligationsNational expenditure rules can enable Member States tomeet the budget balance requirements of the Treaty andSGP by helping them to better control expenditure itemsthat are subject to overruns The specific design and thestrength of the enforcement mechanisms are key to theireffectiveness Depending on their design they can alsocontribute to other policy objectives such as avoiding apro-cyclical loosening of fiscal policy in good times andimproving the quality of the composition of publicspending
There is a great deal of variety in the design of expendi-ture rules across EU Member States as regards the typesof expenditure covered by a rule the time frameinvolved and the robustness of surveillance and enforce-ment mechanisms Preliminary empirical analysis indi-cates that the existing expenditure rules have not had avisible impact on trends in public spending Howeverjudging compliance with expenditure rules is difficult asin many cases they cover several years and are subject torevisions In some countries expenditure rules are notambitious enough and adherence with them is easilyreached in other cases the rule has been adjusted orabandoned if it is perceived as being too ambitiousOverall even a relatively weak expenditure rule can pro-vide useful guidance and signals to actors involved in thebudgetary process
The Treaty and SGP requirements are defined in terms ofthe budget balance of the general government (that iscentral and localstate governments and social security)although the specific budget targets in stability andconvergence programmes are set by the central govern-ment The challenge in meeting EU budgetary require-ments is therefore affected by the way in which Mem-ber States allocate fiscal functions (both revenues andexpenditures) across different levels of governmentThis is especially the case in federal countries and theMember States where local authorities have considera-ble budgetary autonomy The contribution of sub-centralauthorities to the overall budget position is changing in anumber of countries in light of efforts to devolve certainpublic functions to regionallocal authorities
yen1part The Council has shown some flexibility in interpreting compliance withthe lsquoclose to balance or in surplusrsquo requirement to reflect significantplanned increases in public investment programmes
yen2part See also communication from the Commission lsquoDeveloping the trans-European transport network innovative funding solutions interoperatibil-ity of electronic toll collection systemsrsquo COM(2003) 132 of 24 April 2003
6
S u m m a r y a n d m a i n c o n c l u s i o n s
The direct contribution of lower levels of government tothe general government deficit is generally limited sinceall Member States apply restrictions to local governmentborrowing the exception is Germany where net borrow-ing by local and state governments accounts for nearlyhalf of the general government budget deficit in 2002However it should be borne in mind that de facto centralgovernments often have to bear the cost of financing dif-ficulties that emerge at sub-central level To help complywith the EUrsquos fiscal rules the federal Member States andItaly and Spain have recently introduced arrangementsthat aim at coordinating the budgetary position acrosslevels of government (usually referred to as nationalstability pacts) More experience with the implementa-tion of these arrangements is needed before conclusionscan be drawn on their effectiveness in contributing to theobjectives of the EU fiscal framework A priori a strong
legal base and enforcement mechanism would beexpected to contribute to the credibility and effective-ness of the arrangements
The process of decentralising responsibility for somepolicies raises a second issue in the context of EMUnamely the operation of automatic stabilisers Experi-ence shows that in general systems are designed toshield sub-national governments from cyclical varia-tions However empirical evidence for the US and Ger-many suggests some degree of pro-cyclical behaviour atthe level of the states Further research would be usefulto analyse the possible interaction between fiscal decen-tralisation and automatic stabilisation and to identify thebest practices to reconcile the process of decentralisationwith ensuring sound and sustainable public finances
7
Part I
Current developments and prospects
Summary
Against a background of a prolonged period of lowgrowth 2002 and the early part of 2003 has been a diffi-cult period in terms of actual budgetary developmentsThe nominal deficit for the euro area increased from16 of GDP in 2001 to 22 in 2002 and is forecast torise to 25 of GDP in 2003 according to the latestCommission forecast However this aggregate outcomeis the result of striking contrasts in the performanceacross Member States By the end of 2002 only six EUcountries including four euro area countries (accountingfor under 18 of euro area output) had achieved budgetpositions in both nominal and cyclically-adjusted termsthat respected the lsquoclose to balance or in surplusrsquo require-ment of the Stability and Growth Pact (SGP) in contrasttwo euro area countries accounting for half of the euroarea output had nominal deficits above 3 of GDP
Among the countries recording high deficits Portugalsucceeded in reducing the nominal deficit from 41 ofGDP in 2001 to 27 in 2002 although very significantchallenges remain concerning 2003 as much of thisimprovement is due to one-off measures such as a taxamnesty A deficit of 36 of GDP in 2002 has resultedin Germany being placed in an excessive deficit positionand while the authorities are taking measures aimed atreducing the cyclically-adjusted budget deficit only avery limited improvement is expected in 2003 as growthconditions deteriorate Despite clear evidence of budget-ary slippage emerging in early 2002 the failure ofFrench authorities to take corrective measures resulted ina deficit of 31 of GDP in 2002 recent forecasts showan even higher deficit for 2003 at 37 of GDP and thatthe deficit in 2004 would be 35 in 2004 that is stillabove the reference value of the Treaty Large deficitsremain in Italy (23 of GDP in 2002) and the deficitlevel is projected to remain unchanged in 2003 and beabove the 3 of GDP reference value by 2004 budget-ary consolidation efforts in Italy continue to rely on one-off measures rather than reforms of a structural natureneeded to ensure a permanent improvement in thebudget balance Deficits have also re-emerged in coun-
tries that already had reached balanced budget positionsnotably Austria (06 of GDP in 2002) the Netherlands(11 ) and also in the UK (13 ) These three coun-tries are forecast to record an important deterioration ofthe deficit in 2003
At first sight these developments compare relativelyfavourably with previous economic downturns whendeficits reached much higher levels and debt ratiosentered rapidly increasing trajectories In addition gov-ernments have not pursued fine-tuning policies andwhile fiscal policies were slightly looser monetaryconditions have eased thanks mainly to low real interestrates
However a closer consideration of underlying budget-ary trends reveals that the deterioration in nominal def-icits results from high and rising cyclically-adjusteddeficits in several countries This indicates a discre-tionary loosening of the fiscal stance by some MemberStates brought about by a combination of unfunded taxcuts discretionary expenditure increases and slippagesas regards budgetary execution While the outcome ofthe euro area in 2002 was unchanged compared to 2001it should be noted that the cyclically-adjusted budgetbalance for 2001 has recently been revised upwards to21 from 15 of GDP implying that the deteriorationin the underlying budget balance in that year was consid-erably worse than earlier estimates showed moreoverthe cyclically-adjusted budget balance includes the impactof one-off budgetary measures which only have a transi-tory effect on budget positions The deterioration has beenparticularly pronounced in Germany (where the CABincreased to 32 of GDP in 2002) and France (to 33 )In Italy it has improved but remained high (at 21 )
In a medium-term perspective the latest updates of thestability and convergence programmes contain a com-mitment to reach the target of lsquoclose to balance or in sur-plusrsquo both in actual and structural terms by 2005 or2006 although this is not explicitly stated by all Member
11
P u b l i c f i n a n c e s i n E M U 2 0 0 3
States However it should be noted that the medium-term targets of Member States are based on growthassumptions which in light of developments in recentmonths now appear to be optimistic For countries wherelarge underlying deficits remain the date for reachingthe lsquoclose to balance or in surplusrsquo objective has beenpushed back to 2006 or 2007 and even this deadline willonly be met if additional consolidation measures areundertaken It is vital therefore that all efforts are madeto achieve these goals and maintain sound positions overthe medium term This requires that budgetary consoli-dation resumes vigorously as soon as growth picks up inorder to achieve the agreed objectives by the deadlines inthe programmes Meeting these targets will allow allMember States to let automatic stabilisers operate freelyduring future cyclical downturns thereby mitigating thepolicy dilemma that countries in deficit faced in 2002and 2003
EU budgetary surveillance for the second time includesa systematic assessment of the sustainability of publicfinances on the basis of the updated stability and conver-gence programmes submitted in late 2002 The analysisshows that there is a risk of unsustainable public financesin some half of EU countries notably Belgium Ger-many Greece Spain France Italy Austria and Portu-gal With a fast-closing window of opportunity prior tothe budgetary impact of ageing populations taking holdthe risk of unsustainable public finances will increasesubstantially higher if Member States with large deficitsdo not achieve and sustain the budgetary consolidationplans outlined in their stability and convergence pro-grammes In Spain and Greece a substantial share of therisk of emerging budgetary imbalances is due to a verylarge projected increase in pension expenditure In sev-eral Member States (notably Germany France Austriaand Portugal) the risk of emerging budgetary imbalancesis a combination of factors including a projected increasein public spending on pensions and healthcare a slowingin the pace of debt reduction and relatively low labour-force participation rates of older workers High-debtcountries (Belgium Greece and Italy) face a particularset of challenges because they must be able to sustainlarge primary surpluses over several decades SeveralMember States appear to have sustainable public financesincluding Denmark Luxembourg the Netherlands Fin-
land Sweden and the UK but they nonetheless facebudgetary challenges as a result of ageing populationsfor example the maintenance of high tax ratios at over50 of GDP raises concern about competitiveness inthe long run and in some countries the financial sustain-ability of the pension system depends on the perform-ance of private pensions
The framework for budgetary surveillance at EU level isbeing prepared for the accession of 10 countries to theEU in May 2004 The aggregate general governmentdeficit of these 10 countries widened but is projected toimprove in 2003 and 2004 Despite a significant acceler-ation in growth however the projected reduction in theaggregate deficit of the 10 acceding countries is notsufficient to reverse the deterioration recorded in 2002This suggests that structural rather than cyclical factorsunderlie current budgetary imbalances Concerning the13 candidate countries as a whole the aggregate budgetposition is influenced to a large extent by the exceptionaladvance recorded in 2002 and forecast for the comingyears in Turkey
Looking at the pre-accession economic programmessubmitted by candidate countries an improvement by2005 is envisaged in the large majority of cases Ninecountries plan to reduce their budget deficits by 2005leading to a fall in the average deficit Among the fourremaining countries Bulgaria and Estonia plan to movefrom a small surplus to a balanced budget leaving onlyLatvia and the Czech Republic with a projected increasein the general government deficit over the programmeperiod In 2005 projected budget outcomes would varyfrom a balanced budget in Bulgaria and Estonia to a def-icit of 55 of GDP in the Czech Republic Among thecandidate countries only the Czech Republic Malta andthe Slovak Republic refrained from targeting a deficitbelow 3 of GDP in 2005 According to the programmesgeneral government debt-to-GDP ratios would fall orremain virtually stable in all countries with the excep-tion of the Czech Republic and Poland where the debt-to-GDP ratio is projected to rise considerably by the endof the programme period By 2005 however all candi-date countries with the exception of Malta and Turkeywould have a debt-to-GDP ratio below 60
12
1 Budgetary developments in the euro area and EU Member States
11 Short-term developments and prospects for the budget balance and public debt
In 2002 the euro-area budget position deteriorated again(see Table I1) The actual deficit reached 22 of GDP06 of GDP higher than the outcome in 2001 a devel-opment which is largely explained by the working of theautomatic stabilisers in a period of slowing growth Theeuro-area cyclically-adjusted budget deficit in 2002remained high at 22 of GDP almost unchanged from2001
At first sight this outcome does not appear to be undulynegative against a background of slow growth Howeverit should be noted that the cyclically-adjusted budget bal-ance figure for 2001 has recently been revised upwards to21 from 15 of GDP implying that the deterioration inthe underlying budget balance in that year was consider-ably worse than earlier estimates showed Moreover thecyclically-adjusted budget balance includes the impact ofone-off budgetary measures which only have a transitoryeffect on budget positions Overall this points to anunderlying budget position of the euro area which is lessfavourable than in 1999ndash2000
The aggregate outcome for the euro area as a whole is theresult of striking contrasts in budgetary performanceacross Member States As shown on Table I2 thebudget positions of Germany France Portugal and Italyremained weak with deficits ranging from 23 of GDPin Italy to 36 of GDP in Germany As a result of thedevelopments in the course of 2002 Germany and Por-tugal have already been placed in an excessive deficitposition (1) and the procedure has been launched against
France (see Part II1 of this report) In contrast six EUMember States and four in the euro area had actualbudget positions in balance or in surplus in 2002 In spiteof the continued slowdown in growth actual budget bal-ances in 2002 did not deteriorate (or did so only margin-ally) compared to the previous year in Belgium GreeceSpain Finland Italy (although this is because of a largeupward revision in the recorded deficit level for 2001)and Portugal (partially as the result of one-off measures)
Looking ahead to 2003 and 2004 the Commission fore-cast of spring 2003 projects that economic growth in 2003will remain below potential The budget balance for theeuro area as a whole is expected to deteriorate further to25 of GDP and to remain at a similar level in 2004
Developments in Member States show that BelgiumSpain Ireland and Luxembourg are expected to moveinto small budget deficit positions in 2003 Under a no-policy change assumption Belgium and Spain are pro-jected to move back towards a position of balance in2004 while in Ireland and Luxembourg the deficitwould deteriorate further to around 1 of GDP
On the basis of current policies the Commission fore-casts that Germany France Italy and Portugal willhave deficit levels above the 3 of GDP referencevalue in 2003 andor in 2004 The budget deficit inGermany is forecast to remain above 3 of GDP in2003 and to move only slightly below the referencevalue in 2004 The situation in France is more worry-ing since the deficit is forecast to increase further in2003 and remain well above 3 of GDP in 2004 incontradiction with the requirements of the excessivedeficit procedure After the large reduction in the Por-tuguese deficit in 2002 the balance is expected to dete-riorate in 2003 and remain above 3 of GDP in 2004The deficit in Italy is projected to breach the 3 ofGDP reference value in 2004 yen1part The latter for the 2001 deficit discovered only late in 2002
13
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Table I1
General government budgetary position mdash euro area 1999ndash2004( of GDP)
1999 2000 (1) 2001 (1) 2002 2003 2004
Total receipts (1) 475 472 465 462 461 459
Total expenditure (2) 489 471 481 484 486 483
Actual balance (3) = (1) ndash (2) ndash 13 01 ndash 16 ndash 22 ndash 25 ndash 24
Interest (4) 42 40 39 37 36 35
Primary balance (5) = (3) + (4) 29 41 23 15 11 11
UTMS proceeds 11 00 00
Cyclically-adjusted balance (6) ndash 17 ndash 18 ndash 21 ndash 22 ndash 20 ndash 20
Cyclically-adj prim balance = (6) + (4) 26 22 18 15 16 15
Change in actual balance 10 14 ndash 17 ndash 06 ndash 03 01
Due to mdash Cycle 03 05 ndash 04 ndash 06 ndash 04 01
mdash UMTS 11 ndash 11
mdash Interest 06 02 01 02 01 01
mdash Cyclically-adjusted primary balance 01 ndash 04 ndash 04 ndash 03 01 ndash 01
(1) Including UMTS receipts UMTS receipts as a of GDP would be equal in 2000 to 25 for DE 01 for ES 12 for IT 07 for NL 04 for AT 03 for PT24 for UK 11 for the euro area and 12 for the EU-15 In 2001 they would be equal to 02 for BE 02 for DK 05 for EL 01 for FR and 0 for the euro area andthe EU-15 In 2002 they would be equal to 0 for FR 02 for IE and 0 for the euro area and EU-15
NB differences are due to rounding
Source Commission spring 2003 economic forecasts
Table I2
Budget balances in EU Member States 2001ndash04 ( of GDP)
Budget balance excluding UMTS
Cyclically-adjusted budget balance
Cyclically-adjusted primary balance
2001 2002 2003 2004 2001 2002 2003 2004 2001 2002 2003 2004
BE 03 01 ndash 02 ndash 01 ndash 04 01 02 00 62 61 57 49
DE ndash 28 ndash 36 ndash 34 ndash 29 ndash 30 ndash 33 ndash 26 ndash 24 03 ndash 01 06 08
EL ndash 19 ndash 12 ndash 11 ndash 10 ndash 23 ndash 18 ndash 18 ndash 19 40 37 34 30
ES ndash 01 ndash 01 ndash 04 ndash 01 ndash 08 ndash 04 ndash 04 ndash 01 23 25 23 24
FR ndash 16 ndash 31 ndash 37 ndash 35 ndash 22 ndash 33 ndash 35 ndash 33 09 ndash 03 ndash 03 01
IE 12 ndash 03 ndash 06 ndash 09 00 ndash 09 ndash 03 01 15 04 12 16
IT ndash 26 ndash 23 ndash 23 ndash 31 ndash 31 ndash 21 ndash 18 ndash 27 33 36 35 24
LU 64 26 ndash 02 ndash 12 41 20 05 ndash 03 44 23 07 ndash 02
NL 01 ndash 11 ndash 16 ndash 24 ndash 10 ndash 10 ndash 04 ndash 11 25 22 26 18
AT 03 ndash 06 ndash 11 ndash 04 00 ndash 06 ndash 10 ndash 04 35 29 25 30
PT ndash 42 ndash 27 ndash 35 ndash 32 ndash 46 ndash 25 ndash 26 ndash 21 ndash 15 05 05 09
FI 51 47 33 30 42 48 37 33 70 70 58 54
EUR-12 ndash 16 ndash 22 ndash 25 ndash 24 ndash 21 ndash 22 ndash 20 ndash 20 18 15 16 15
DK 28 20 18 21 23 19 20 22 63 55 53 54
SE 45 13 08 12 36 09 11 15 68 38 39 42
UK 08 ndash 13 ndash 25 ndash 25 07 ndash 10 ndash 20 ndash 20 31 11 00 00
EU-15 ndash 09 ndash 19 ndash 23 ndash 22 ndash 14 ndash 18 ndash 18 ndash 18 23 16 15 14
NB Concerning UMTS receipts see footnote to Table I1 Cyclically-adjusted figures are computed with the production function method except for Germany SpainLuxembourg and Austria where the HP filter method has been used
Source Commission spring 2003 economic forecasts
14
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
High deficits are forecast in other countries as wellin the UK it is projected to deteriorate to 25 of GDPin 2003 and 2004 while in the Netherlands the deterio-ration would be progressive to reach 24 of GDPin 2004
In cyclically-adjusted terms the deficit of the euro areawould decrease slightly in 2003 to 20 of GDP andremain unchanged in 2004 which underlines the factthat the budgetary consolidation process has stalled inrecent years At national level the cyclically-adjusted
deficit is projected to remain above 3 of GDP inFrance while in Italy it is expected to move close to thatlevel by 2004 Germany and Portugal are forecast tomove below 3 of GDP by that year Six euro-areacountries and eight EU Member States are expected tocomply in cyclically-adjusted terms with the lsquoclose tobalance or in surplusrsquo requirement of the SGP by 2004The negative effect of the cycle on the nominal balancesis expected to diminish progressively in 2004 (seeTable I1) so that by that year nominal budgets in manycountries would be close to balance as well
After stabilising in 2002 the general government grossdebt level of the euro area is expected to increase slightlyin 2003 to just below 70 of GDP (see Table I3) Debtreduction should resume in 2004 but at a very slow pacedue to the large negative contribution of the interest rate-growth rate differential and an insufficiently high pri-mary surplus Stock-flow operations mdash although modestmdash would increase debt ratios
This overall picture conceals very different situationsacross Member States Italy Belgium and Greece
continue to have debt ratios above the 100 of GDP By2004 only Italy should have a debt level above 100 ofGDP In Greece debt increasing financial operations ofthe government as reflected in the large stock-flow com-ponent would offset to a large extent the positive contri-butions of the primary balance and GDP growth A highdeficit and the poor growth performance will impact thedebt developments in Germany where the debt ratiowent above 60 of GDP in 2002 as well as in Franceand in Portugal where the reference value is projected tobe breached in 2003 and 2004 respectively In Austria
Table I3
Composition of changes in government debt ratio in EU Member States 2001ndash04 ( of GDP)
Gross debt
Change in gross debt 2002ndash04
Change in 2002ndash04 due to
2001 2002 2003 2004Primary balance
Interest and growth contribution
stock flow adjustment
BE 1085 1053 1027 989 ndash 63 ndash 101 34 04
DE 595 608 627 630 21 ndash 01 38 ndash 15
EL 1070 1049 1010 970 ndash 79 ndash 80 ndash 43 44
ES 569 540 525 505 ndash 35 ndash 46 ndash 08 19
FR 568 591 618 631 40 07 28 05
IE 368 333 333 333 00 ndash 15 ndash 18 31
IT 1095 1067 1060 1047 ndash 20 ndash 50 23 06
LU 56 53 41 34 ndash 19 11 00 ndash 30
NL 528 526 524 528 02 ndash 20 25 ndash 03
AT 673 687 685 668 ndash 19 ndash 54 30 05
PT 556 581 594 602 21 07 13 01
FI 438 427 423 414 ndash 13 ndash 105 12 80
EUR-12 692 692 699 696 04 ndash 23 24 03
DK 454 452 427 399 ndash 53 ndash 104 32 20
SE 544 526 509 495 ndash 31 ndash 74 12 30
UK 389 384 390 398 13 09 05 00
EU-15 629 627 635 632 06 ndash 20 27 ndash 01
Source Commission spring 2003 economic forecasts
15
P u b l i c f i n a n c e s i n E M U 2 0 0 3
after the continuous increase in the debt level until 2002to almost 69 of GDP debt should move onto a slowdownward path in 2003
12 Government revenue and expenditure
The deterioration in the cyclically-adjusted budgetarybalance in the past two years (resulting in the euro areamoving further way from the SGP goal of lsquoclose to bal-ance or in surplusrsquo) is the result of diverging trends asregards expenditures and revenue ratios As shown inTable I4 the expenditure ratio for the euro area in cycli-cally-adjusted terms remains static over the 2000ndash04period In contrast cyclically-adjusted revenues for theeuro area fell from 465 in 2000 to 461 of GDP in2001 (which contributed to increasing the deficit incyclically-adjusted terms) but started to rise to 464 and466 of GDP in 2002 and 2003 (which contributes tolowering the deficit)
At Member State level the patterns are generally similarto that of the euro area (Table I5) Only in Germany andPortugal are revenue ratios expected to increase over the2002ndash04 period (although this is to a large extent due toan improvement in the cyclical position) Strongdeclines are set to take place in the Netherlands Luxem-
bourg Austria and Finland Outside the euro area reve-nues in Sweden and the UK are set to increase over thenext two years while in Denmark revenues will diminishover the whole period Expenditure ratios over 2002ndash04are set to increase in France Luxembourg the Nether-lands Portugal and in particular the UK where discre-tionary spending measures are planned to improvepublic services and address infrastructure needsBy contrast a marked decrease is expected in GreecePortugal and Denmark
A number of lessons can be drawn from these develop-ments Firstly tax reforms were introduced before Mem-ber States had completed the transition to the lsquoclose tobalance or in surplusrsquo objective of the SGP and therewas insufficient room for the automatic stabilisers tooperate when growth slowed down resulting in deficitsin several Member States breaching the 3 of GDP ref-erence value To prevent deficits from rising furtherseveral countries have had to take measures to raiserevenue ratios either by raising tax rates (such as Portu-gal) or extending tax bases (such as Germany) therebyreversing the effects of earlier reforms Secondly thereis some evidence that the relatively high growth ratesin 1999 and 2000) resulted in a degree of fiscal illusionwhereby authorities in some countries overestimated the
Table I4
Euro area government resources and expenditures 2000ndash04 ( of GDP)
2000 2001 2002 2003 2004
Total resources 472 465 462 461 459
mdash Cyclically-adjusted 465 461 464 466 463
Taxes on imports and production 136 133 134 134 134
Current taxes on income and wealth 130 125 123 120 120
Social contributions 162 161 160 161 160
of which actual social contributions 151 149 149 150 149
Other resources 45 46 46 46 44
Total expenditure 471 481 484 486 483
mdash Cyclically-adjusted 483 482 484 485 482
Collective consumption 82 82 82 83 82
Social benefits in kind 117 117 118 119 118
Social transfers other than in kind 167 166 170 173 172
Interest 40 39 37 36 35
Subsidies 14 14 13 13 12
Gross fixed capital formation 25 25 24 25 25
Other expenditures 25 38 40 39 38
NB Including UMTS receipts see footnote to Table I1
Source Commission 2003 spring forecast
16
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
Table I5
Total revenue and expenditure in EU Member States 2001ndash04 ( of GDP)
Revenue Expenditure
2001 2002 2003 2004 2001 2002 2003 2004
BE 498 502 495 492 494 501 497 493
DE 455 450 454 455 483 486 489 484
EL 456 465 460 452 470 477 471 462
ES 392 396 393 395 393 396 398 396
FR 510 505 503 503 525 537 541 538
IE 352 337 335 328 341 337 340 336
IT 458 452 451 443 485 475 474 475
LU 466 481 460 451 402 455 463 464
NL 465 461 459 453 464 472 475 477
AT 523 514 510 507 520 520 521 511
PT 421 435 435 436 463 463 470 469
FI 542 539 528 520 490 492 495 490
EUR-12 465 462 461 459 481 484 486 483
DK 581 570 562 561 550 549 544 540
SE 617 595 599 597 572 582 591 585
UK 407 395 395 397 399 407 419 422
EU-15 461 456 456 454 470 474 478 476
NB Including UMTS receipts see footnote to Table I1
Source Commission spring 2003 economic forecasts
Graph I1 Euro area fiscal stance and cyclical conditions 2000ndash04
2004
2003
2002
2001 2000
ndash 10
ndash 05
00
05
10
ndash 20 ndash 15 ndash 10 ndash 05 00 05 10 15 20
Pro-cyclical fiscal tightening
Pro-cyclical fiscal loosening
Counter-cyclical fiscal tightening
Counter-cyclical fiscal loosening
Cyclical condition (Output gap)
Fis
cal s
tanc
e (C
hang
es in
CA
PB
)
17
P u b l i c f i n a n c e s i n E M U 2 0 0 3
level of structural revenues andor the benefits thatwould result from reforms of the tax system Thirdly taxcuts in 1999 and 2000 were not matched by expendituresavings and indeed expenditure cuts made little or nocontribution to reaching the goal of budget positions oflsquoclose to balance or in surplusrsquo
13 The fiscal stance and policy mix
The fiscal stance and policy mix in the euro area
An appropriate policy mix can be defined as a combina-tion of monetary and fiscal policies that ensures pricestability and keeps economic activity close to its poten-tial level In EMU the policy mix results from a mone-tary policy that is centralised and from fiscal policieswhich are decentralised In the euro area nationalauthorities set fiscal policy at Member State level In sodoing national budgetary policies determine implicitlythe fiscal stance for the euro area as a whole The aggre-gate fiscal stance deserves special attention since itaffects the policy mix at the euro-area level and there-fore is one of the elements taken into account by the ECBin setting monetary policy In turn the policy mix for theeuro area will have a feedback effect on the national pol-icy mix via the common interest rate This implies thatthe policy mix needs to be assessed both from the per-spective of the euro area as a whole and from the per-spective of each Member State
Graph I1 examines the fiscal stance (proxied by thechanges in the cyclically-adjusted primary balanceCAPB) in relation to cyclical conditions (that is the sizeof the output gap (1)) for the euro area In this graph fis-cal behaviour in accordance with the general philosophyof the SGP would be represented by a line parallel to thehorizontal axis In other words countries would achieveand sustain broadly balanced budgets over the economiccycle and run a neutral fiscal policy (lsquotax smoothingrsquo)Hence changes in the output gap would not result inmovements in the CAPB Actual budget balances wouldchange reflecting the working of automatic stabilisers Inthe transition period to the extent that a country has yetto reach the medium-term target of the SGP a restrictivefiscal stance mdash that is a rise in CAPB mdash would beneeded (2)
According to the Commission spring 2003 forecasts thefiscal stance loosened again slightly in 2002 This devel-opment follows two years of a looser-than-expected fis-cal policy (given the revision of budgetary positions con-cerning 2001) Such a stance in the past three yearscoupled with the failure to improve cyclically-adjustedbudget balances when growth conditions were favoura-ble has resulted in the current economic slowdown innominal deficits of some Member States approaching orbreaching the 3 of GDP reference value Despite thelonger-than-expected economic slowdown which ledto the appearance of negative output gaps Graph I1illustrates that Member States are not implementingsizeable counter-cyclical measures This is welcomeas the medium-term losses of relaxing fiscal policywould probably outweigh the uncertain short-termgains (see Part IV on this issue) A broadly neutral fis-cal policy stance is projected for 2003 and 2004
Turning to the policy-mix in the euro area Graph I2plots the fiscal stance on the vertical axis and on the hor-izontal axis the monetary stance proxied by the changein the short-term real interest rates Against a back-ground of a prolonged slowdown of the global economythe monetary stance was loosened in 2001 and to a morepronounced degree in 2002 Overall the policy mix inthe early years of EMU has therefore been broadlyappropriate to provide conditions for economic growthand macroeconomic stability
The fiscal stance and policy mix at the national level
The aggregate fiscal stance for the euro area concealsquite disparate national responses to the economic slow-down Graph I3 shows that most EU countries had anegative output gap in 2002 as a result of growth belowpotential in the 2001ndash02 period
France and Ireland loosened their stance in 2002 despitehaving positive output gaps Given the estimated level ofthe output gap the fiscal stance (in particular in Ireland)appears to have been pro-cyclical however the judge-ment on pro-cyclicality has to take into account theuncertainty of the measure of output gap as well as thepoor economic conditions in 2002 Outside the euroarea Sweden substantially eased the fiscal stance inspite of a slightly positive output gap but in view of itsquick deterioration
Several EU countries loosened their fiscal policies in acontext of negative output gaps However the fiscal
yen1part In line with the Council agreement the output gap used in this section iscomputed with the production function method
yen2part However part of the adjustment towards balanced budgets may be origi-nated by reducing interest payments
18
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
stimulus was modest in most of these countries with theexception of the UK where the policy was clearly coun-ter-cyclical Portugal stands out for a clearly pro-cyclicalpolicy in 2002 as it enacted a strong consolidation inorder to bring the deficit below 3 from the level of41 recorded in 2001
As pointed out above for the euro area as a whole thepolicy mix in 2002 has been slightly accommodativewith most Member States experiencing a simultaneousloosening of the fiscal stance accompanied by decliningreal interest rates the real interest rate fell in all countriesexcept Finland and the Netherlands
While Graph I4 shows the changes in the real short-term interest rate its level is also important in assessingthe policy mix After the reductions in the nominalinterest rate decided by the ECB during 2002 the realinterest rate in the euro area (that is the short-terminterest rate corrected by private consumption infla-tion) was around a very low 1 in 2002 However thisaggregate figure for the euro area conceals significantdifferences across countries due to differences in infla-
tion rates across countries In spite of the reduction inshort-term real interest rates in 2002 real interests ratesin Germany France Austria and Finland were justbelow 2 whereas in a number of countries (GreeceSpain Ireland and Portugal) the real interest ratebecame slightly negative
Regarding 2003 the fiscal stance is forecast to bebroadly neutral in most members of the euro area (seeGraph I5) Ireland Germany and the Netherlands areexpected to enact a tightening of the fiscal stance Instark contrast France Portugal and Italy mdash countrieswhich still have high budget deficits mdash are not expectedto make any sizeable progress towards improving theirbudgetary positions in 2003 Finland which is benefitingfrom the past consolidation efforts and consequentlyenjoys a large safety margin is expected to ease the fis-cal stance Some pro-cyclical policy is projected for2003 in Greece Fiscal policy in the three countriesoutside the euro area is expected to be neutral with thenotable exception of the UK where the fiscal stanceagain is set to be loosened
Graph I2 Policy-mix in the euro area 1999ndash2002
1999
20002001
2002
ndash 10
ndash 05
00
05
10
ndash 1 0 1
Fiscalmonetary tightening
Fiscalmonetary loosening
Fiscal loosening monetary tightening
Fiscal tightening monetary loosening
Monetary stance (Change in real short-term interest rates)
Fis
cal s
tanc
e (C
hang
e in
CA
PB
)
19
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Graph I3 Fiscal stance and cyclical conditions in EU Member States in 2002
Graph I4 Policy-mix in EU Member States in 2002
PT
IT
DE
EUR-12
UK
FI
NL BE DKFR
EL
IE
SE
AT
ndash 4
ndash 3
ndash 2
ndash 1
0
1
2
3
ndash 1 0 1 2 3
Pro-cyclical fiscal easing
Counter-cyclical fiscal tightening
Pro-cyclical fiscal tightening
Counter-cyclical fiscal easing
Cyclical conditions (Output gap)
Fis
cal s
tanc
e (C
hang
es in
CA
PB
)
ES
ndash 3
ndash 2
ndash 1
0
1
2
3
ndash 15 ndash 10 ndash 05 00 05 10 15
Fiscal easing monetary tightening
Fiscalmonetary tighteningFiscal tightening monetary easing
Fiscalmonetary easing
FI
NL
BE
DEAT
FRIE
EL
ES
IT
EUR-12
PT
Monetary stance (Changes in real short-term interest rates)
Fis
cal s
tanc
e (C
hang
es in
CA
PB
)
20
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
Graph I5 Fiscal stance and cyclical conditions in EU Member States in 2003
UK
SE
DK
EUR-12
FI
PT
AT
NL
IT
IE
FR
ESEL
DE
BE
ndash 2
ndash 1
0
1
2
ndash 3 ndash 2 ndash 1 0 1 2
Cyclical conditions (Output gap)
Fis
cal s
tanc
e
(Cha
nge
in C
AP
B)
Pro-cyclical fiscal easing
Counter-cyclical fiscal tightening
Pro-cyclical fiscal tightening
Counter-cyclical fiscal easing
21
2 Overview of the 2002 updates of the stability and convergence programmes
21 The medium-term budget targets
The examination of the latest round of updates of the sta-bility and convergence programmes covering the periodto 200506 was particularly prolonged While for most ofthe countries the assessment was completed betweenJanuary and March the Austrian programme was onlyexamined in May and the Dutch programme was consid-ered as provisional pending the submission of a newprogramme after the formation of the new governmentIt should be underlined that the budgetary obligations of
the Treaty and SGP remain in force during periods whennew governments are being formed
To assess the reliability and ambition of budget targetsset by Member States in stability and convergence pro-grammes it is necessary to examine the underlyinggrowth assumptions on which the budgetary commit-ment is given The updated programmes projected a sus-tained economic recovery in the euro area GDP growthwould resume to 21 in 2003 reach 26 in 2004 andstay at 27 in the following years (see Table I6)
Graph I6 Potential and real GDP growth rate and output gaps for the euro area derived from the 2002 updates (1)
(1) Potential GDP growth and output gaps calculated by the Commission
ndash 10
ndash 05
00
05
10
15
20
25
30
2002 2003 2004 2005 2006
Output gapReal GDP growth Potential GDP growth
22
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
Table I6
Euro area mdash Growth projections and macroeconomic developments in the 2002 updates and comparison with the 2001 updates and the Commission forecasts
Macroeconomic developments 2001 2002 2003 2004 2005 2006
2002 updates of the stability programmes
Real GDP growth pc from previous year 16 10 21 26 27 27
GDP deflator 24 22 19 18 18 18
HICP change 26 22 20 16 15 15
Employment growth 12 03 06 12 12 13
Labour productivity growth 03 07 16 16 16 16
2001 updates of the stability programmes 18 18 27 27
Difference ndash 03 ndash 08 ndash 06 ndash 01
Commission autumn 2002 forecast 14 08 18 26
Difference 02 02 03 00
Commission spring 2003 forecast 15 09 10 23
Difference 01 01 11 03
NB Discrepancies are due to rounding For the 2001 updates GDP growth rates used for Germany and the Netherlands are on the basis of the cautious scenario and ofthe revised scenario respectively Since figures for the HICP were not available in the German programme the Commission forecasts have been used to have a rep-resentative aggregate
Source Commission services
Table I7
GDP growth projections in the 2002 updates
2001 2002 2003 2004 2005 2006 Revision (1)
BE 08 07 21 25 25 ndash 05
DE 06 05 15 225 225 225 ndash 04
EL 41 38 38 40 37 36 ndash 01
ES 27 22 30 30 30 ndash 01
FR 18 12 25 25 25 25 ndash 04
IE 57 45 35 41 50 ndash 10
IT 18 06 23 29 30 30 ndash 08
LU 10 05 12 24 31 ndash 42
NL 13 025 075 275 275 275 ndash 09
AT 07 09 14 20 25 25 ndash 05
PT 16 07 13 27 31 35 ndash 08
FI 07 16 28 26 25 24 ndash 01
EUR-12 16 10 21 26 27 27 ndash 05
DK (2) 10 15 18 21 18 17 ndash 01
SE 12 21 25 25 23 ndash 01
UK (3) 20 15 275 325 30 ndash 01
EU-15 16 11 22 27 27 ndash 04
Standard deviation EUR-12 16 14 09 06 07
(1) Difference with respect to the 2001 updates in average growth over 2002ndash04 (2) Taking account of revised information provided by Denmark For 2006 data provided for 2010 has been used(3) Mid-point of the range provided in the programme
23
P u b l i c f i n a n c e s i n E M U 2 0 0 3
The implied euro-area potential growth would be stableat 24 Based on these assumptions the output gap in2002 would be negative at 09 of potential GDP andwiden further in 2003 to 11 thereafter it would closeand disappear by 2006 (see Graph I6 and Table I7)
This projection contrasts markedly with the oneexpected in the 2001 updates which foresaw a lessmarked slowdown and a less sizeable negative gap sothat the output gap would already close in 2003 thanksto the expected rebound
The growth projections were more favourable than theCommissionrsquos autumn 2002 ones which provided thebasis for the assessment of the Commission and CouncilIn view of the Commission spring 2003 forecasts theGDP growth assumptions in the programmes now seemoverly optimistic especially for 2003 (for a comparisonconcerning the euro area see last row of Table I6)
Based on these growth assumption the programmes pro-jected that the budget balance for the euro area wouldimprove from a projected level of 22 of GDP to below1 of GDP by 2005 and should reach zero in the euroarea by 2006 (see Table I8) The overall improvement inthe budget balance relies strongly on the sizeable budget-ary consolidation projected in the largest Member Statessuch as Germany (a consolidation of 38 of GDP overthe next four years in the actual balance) Italy (22 ) andFrance (18 ) Also Portugal (23 ) and Greece (17 )foresee large improvements in the actual budget balanceThe other euro-area countries also project to improve theirbudgetary position over the next four years the only coun-tries forecasting a deterioration were Finland which wouldstill post significant actual surpluses and Ireland wherethe deficit would increase to above 1 of GDP in 2005Outside the euro zone Denmark and Sweden project tomaintain or slightly improve their surpluses over the pro-jection period while in the UK the deficit would remainhigher than 15 in the financial year 2005ndash06
Table I8
Actual budget balances in the 2002 updates and in the Commission forecasts(in of GDP)
2002 updates of stability and convergence programmes (1)
Commission autumn 2002 forecasts (1) (2)
Commission spring 2003 forecasts (1)
2001 2002 2003 2004 2005 2006 2002 2003 2004 2002 2003 2004
BE 02 00 00 03 05 ndash 01 00 03 01 ndash 02 ndash 01
DE ndash 28 ndash 375 ndash 275 ndash 15 ndash 10 00 ndash 38 ndash 31 ndash 23 ndash 36 ndash 34 ndash 29
EL ndash 12 ndash 11 ndash 09 ndash 04 02 06 ndash 13 ndash 11 ndash 11 ndash 12 ndash 11 ndash 10
ES ndash 01 ndash 02 00 00 01 02 00 ndash 03 01 ndash 01 ndash 04 ndash 01
FR ndash 14 ndash 28 ndash 26 ndash 21 ndash 16 ndash 10 ndash 27 ndash 29 ndash 25 ndash 31 ndash 37 ndash 35
IE (3) 16 ndash 05 ndash 07 ndash 12 ndash 12 ndash 12 ndash 12 ndash 10 ndash 03 ndash 06 ndash 09
IT (4) ndash 22 ndash 21 ndash 15 ndash 06 ndash 02 01 ndash 24 ndash 22 ndash 29 ndash 23 ndash 23 ndash 31
LU 61 ndash 03 ndash 03 ndash 07 ndash 01 05 ndash 18 ndash 19 26 ndash 02 ndash 12
NL 01 ndash 07 ndash 10 ndash 07 ndash 04 01 ndash 08 ndash 12 ndash 09 ndash 11 ndash 16 ndash 24
AT 03 ndash 06 ndash 13 ndash 07 ndash 15 ndash 11 ndash 18 ndash 16 ndash 15 ndash 06 ndash 11 ndash 04
PT ndash 28 ndash 24 ndash 19 ndash 11 ndash 05 ndash 34 ndash 29 ndash 26 ndash 27 ndash 35 ndash 32
FIN 49 38 27 21 26 28 36 31 35 47 33 30
EUR-12 ndash 15 ndash 22 ndash 18 ndash 11 ndash 07 ndash 01 ndash 23 ndash 21 ndash 18 ndash 22 ndash 25 ndash 24
DK (5) 28 16 19 24 24 22 20 20 25 20 18 21
SE 48 17 15 16 20 14 12 15 13 08 12
UK (6) ndash 02 ndash 18 ndash 22 ndash 17 ndash 16 ndash 11 ndash 13 ndash 14 ndash 13 ndash 25 ndash 25
EU-15 ndash 11 ndash 20 ndash 17 ndash 10 ndash 07 ndash 19 ndash 18 ndash 16 ndash 19 ndash 23 ndash 22
(1) Excluding UMTS proceeds amounting in of GDP in 2001 to 01 in Belgium 02 in Denmark 05 in Greece 01 in France in 2002 004 in France and 02 inIreland
(2) Based on pre-budget figures for Ireland and the UK For 2004 on the assumption of unchanged policies(3) The targets for the final two years incorporate lsquocontingency provisions against unforeseen developmentsrsquo mdash their size is 04 of GDP in 2004 and 08 in 2005(4) Including lsquofuture measuresrsquo amounting to 16 of GDP in 2004 14 of GDP in 2005 and 08 of GDP in 2006(5) Including revised information provided by Denmark in the supplementary note For 2006 used data relative to 2010 (6) Financial years for data in the convergence programme Figures based on assumptions for output growth which are more prudent than those presented in Table I7
24
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
The comparison between the projections made by theMember States (left panel of Table I8) and by theCommission forecast for 2003 and 2004 made in bothautumn 2002 and spring 2003 (1) (right panels) showsthat in most cases the projections in the programmes forthe budget balance are more favourable than the Com-mission ones This is mostly due to the more optimisticgrowth assumptions presented in the national pro-grammes The only exceptions are Finland and to amuch smaller extent Ireland reflecting their more cau-tious growth assumptions The differences in projectedbudget balances already noticeable for 2003 wouldincrease considerably in 2004 for the euro-area averageThis partly appears to be due to some Member Statesincorporating planned though not-yet-enacted policymeasures in their projections
Most countries provided figures for the cyclically-adjusted budget balance (CAB) in their programmes(see the left panel of Table I9) The central panel ofTable I9 shows the cyclically-adjusted balances com-puted by the Commission and used in the individualassessment of the programmes According to these fig-ures the cyclically-adjusted balance of the euro areawhich deteriorated to 19 of GDP in 2002 is projectedto increase by roughly of GDP per year over thecoming years This is clearly more optimistic than whatwas forecast by the Commission in autumn 2002 (2)
According to the Commission calculations of the ninecountries showing a cyclically-adjusted budget deficit in2002 in the euro area four are projecting to be in deficitin 2006 (Germany France Austria and Portugal) The
yen1part For 2004 based on the assumption of unchanged policies yen2part For 2004 on the assumption of unchanged policies
Table I9
Cyclically-adjusted balances in the 2002 updates and in the Commission forecasts on the basis of the production function method
(in of GDP)
2002 updates of the programmes (1)
Commission calculations based on the 2002 updates (2)
COM autumn 2002 forecasts (2)
COM spring 2003 forecasts (2)
2002 2003 2004 2005 2006 2002 2003 2004 2005 2006 2002 2003 2004 2002 2003 2004
BE 05 06 08 08 03 03 02 02 02 02 01 01 02 00
DE ndash 30 ndash 20 ndash 10 ndash 10 ndash 05 ndash 31 ndash 20 ndash 09 ndash 07 00 ndash 33 ndash 24 ndash 19 ndash 33 ndash 26 ndash 24
EL ndash 15 ndash 16 ndash 08 00 00 ndash 16 ndash 15 ndash 12 ndash 08 ndash 06 ndash 17 ndash 18 ndash 20 ndash 18 ndash 18 ndash 19
ES ndash 03 ndash 01 00 01 03 ndash 01 ndash 02 00 ndash 04 ndash 04 ndash 01
FR ndash 21 ndash 19 ndash 14 ndash 09 ndash 05 ndash 28 ndash 26 ndash 21 ndash 16 ndash 10 ndash 27 ndash 28 ndash 24 ndash 33 ndash 35 ndash 33
IE ndash 10 ndash 04 ndash 02 01 ndash 10 ndash 06 ndash 06 ndash 04 ndash 14 ndash 08 ndash 02 ndash 09 ndash 03 01
IT ndash 12 ndash 05 00 00 00 ndash 14 ndash 09 ndash 02 00 01 ndash 18 ndash 16 ndash 25 ndash 21 ndash 18 ndash 27
LU 21 12 19 20 05 ndash 03
NL ndash 05 03 03 02 04 ndash 06 00 03 ndash 10 ndash 04 ndash 11
AT ndash 04 ndash 09 ndash 04 ndash 13 ndash 11 ndash 04 ndash 09 ndash 04 ndash 13 ndash 11 ndash 17 ndash 15 ndash 14 ndash 06 ndash 10 ndash 04
PT ndash 28 ndash 18 ndash 13 ndash 07 ndash 03 ndash 24 ndash 16 ndash 11 ndash 05 ndash 01 ndash 30 ndash 19 ndash 15 ndash 25 ndash 26 ndash 21
FI 35 25 20 26 29 36 27 24 31 35 37 33 36 48 37 33
EUR-12 ndash 17 ndash 11 ndash 06 ndash 04 ndash 01 ndash 19 ndash 13 ndash 08 ndash 05 ndash 01 ndash 20 ndash 17 ndash 15 ndash 22 ndash 20 ndash 20
DK (3) 21 22 22 22 22 16 19 22 25 21 21 25 19 20 22
SE 23 19 17 11 09 10 13 13 15 09 11 15
UK ndash 12 ndash 15 ndash 13 ndash 15 ndash 15 ndash 12 ndash 14 ndash 14 ndash 15 ndash 16 ndash 06 ndash 09 ndash 10 ndash 10 ndash 20 ndash 20
EU-15 ndash 15 ndash 11 ndash 07 ndash 05 ndash 03 ndash 16 ndash 12 ndash 08 ndash 06 ndash 03 ndash 16 ndash 14 ndash 12 ndash 18 ndash 18 ndash 18
NB Footnotes to Table I8 apply here (1) Germany Austria and Portugal provided figures based on the HP filter method (2) On the basis of the PF method except in the case of Germany Spain and Austria where the HP filter method has been used (3) Commission calculations based on data from programme and information provided in the supplementary note The latter did not provide revised cyclically-
adjusted balances
25
P u b l i c f i n a n c e s i n E M U 2 0 0 3
data show that the deficit countries plan to have anadjustment in cyclically-adjusted terms of at least 05 of GDP per year over the next years (1) France andGreece however would only start the adjustment in2004 Outside the euro area the UK is expected to recorda cyclically-adjusted deficit of 14 of GDP in 2003 setto increase slightly until 2006
The developments in the general government balancecan be decomposed by sectors of government (seeTable I10) (2) For the euro area as a whole the budgetdeficit of the general government in 2002 is the resultof a large deficit of the central government sector (over2 of GDP) and a smaller deficit in the local govern-ment roughly compensated by the small surplus in thesocial security sector The local sector is projected toeliminate its deficit by 2004 and in 2006 should con-tribute with the social security sector to broadly bal-ance the deficit of the central government projected toremain at 07 of GDP in that year
The gross debt-to-GDP ratio in the euro area after theincrease recorded in 2002 is set to resume its gradualdecline in 2003 to arrive to just above 65 in 2006(see Table I11) This is again slower than projected inprevious updates and is due to smaller primary sur-pluses and nominal GDP growth contributions espe-cially for 2002
Table I11 also shows that the estimated stock-flow com-ponent contributes to increase the debt ratio on average
over the period (3) This could either stem from plans tobuild up financial assets (for example in public pensionreserve funds which are invested in non-governmentalassets) or simply indicate that a certain degree of cautionhas been used when setting the targets for debt
Table I12 shows that all Member States will be belowthe 60 of GDP ceiling in 2005 with the exception ofBelgium and Greece where the debt ratio should fallbelow 90 of GDP in 2006 (4) of Italy where it shouldstill be above 95 of GDP in 2006 and of Austriawhere it should decrease very slowly and remain slightlyabove 60 in 2006 In the EU the debt level in 2006 islikely to be below 50 in seven Member States (SpainIreland Luxembourg the Netherlands Denmark Swe-den and the United Kingdom) of which four (IrelandLuxembourg Denmark and the United Kingdom) willhave debt ratios below 40
22 Composition of the budgetary adjustment
The updated programmes show that both revenue andexpenditure ratios are expected to decline over theprojection period (see Table I13) The euro-area totalreceipts are projected to fall by almost 1 of GDPbetween 2002 and 2005 to slightly below 46 of GDPin 2005 This is more than compensated by reductions
Table I10
Euro area net lending by sub-sectors in the 2002 updates (1)
of GDP 2002 2003 2004 2005 2006
General government ndash 22 ndash 18 ndash 11 ndash 07 ndash 01
Central government ndash 20 ndash 18 ndash 16 ndash 13 ndash 07
State plus local governments ndash 04 ndash 03 00 00 02
Social security funds 02 03 03 03 03
(1) Discrepancies are due to rounding and to the non-attribution to a specific sector of future measures in the Italian (for the years 2004ndash06) and German (for 2004)programmes
yen1part In accordance to the pace of adjustment endorsed by the European CouncilSee Part II2)
yen2part To simplify the presentation Table I10 presents the two sectors of stateand local government in one row given that the state government sector isrelevant only for four countries
yen3part As in the previous round of updates very large positive contributions of thestock-flow over the period are identified for Greece (but this time the yearlyaverage is around 2 of GDP rather than 5 of GDP implied in the previ-ous update) Finland Sweden and Ireland (on average around 2 ) andSpain (on average around 1 ) In other countries the stock-flow operationsseem to compensate over the period In Italy they are over 2 of GDP inthe last two years of the programme
yen4part For Belgium assuming that in 2006 nominal GDP growth and the budgetbalance are the same as in 2005
26
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
Table I11
Euro area mdash Gross debt level and changes in the 2001 updates (1)( of GDP)
2001 2002 2003 2004 2005 2006
Gross debt level 695 697 687 668 654 635
Change in gross debt 03 ndash 11 ndash 18 ndash 14 ndash 19
Previous updates of the programmes 691 674 657 636
Difference 04 24 31 34
Contributions to change in gross debt
Primary balance ndash 15 ndash 19 ndash 25 ndash 28 ndash 33
Interest payments 37 37 35 35 34
Nominal GDP growth ndash 22 ndash 28 ndash 31 ndash 30 ndash 29
Other factors influencing the debt ratio (2) 03 ndash 01 03 08 08
(1) Discrepancies are due to rounding(2) The programmes do not always contain enough information to identify directly the contribution from different factors to the development of the euro-area debt ratio
Therefore it has been necessary in some cases to identify the contribution from nominal GDP growth (GDP deflator plus real GDP growth multiplied by the debtratio) In this way the stock-flow adjustment is derived as a residual
Source 2002 updates of the stability and convergence programmes
Table I12
Debt levels in the 2002 updates of the stability and convergence programmes ( of GDP)
2001 2002 2003 2004 2005 2006
BE 1086 1061 1023 979 936
DE 595 610 615 605 595 575
EL 1070 1053 1002 961 921 879
ES 571 552 531 510 490 469
FR 573 587 591 589 583 570
IE 367 341 340 345 349
IT 1099 1094 1050 1004 984 964
LU 53 51 41 38 29
NL 528 519 512 490 474 453
AT 673 678 670 651 638 621
PT 554 588 587 575 553 527
FI (1) 434 425 419 419 414 407
EUR-12 695 697 687 668 654 635
DK (2) 447 439 421 392 367
SE (1) 566 536 509 493 480
UK (3) 382 379 388 389 389 391
EU-15 628 629 621 606 595
(1) Revised national accounts data for 2001 refer to a debt ratio of 435 of GDP for Finland and of 543 of GDP for Sweden(2) Figures for 2002ndash04 may not be consistent with those in the tables for GDP growth and budget balances as they have not been revised by the supplementary note(3) Financial years
Source 2002 updates of the stability and convergence programmes
27
P u b l i c f i n a n c e s i n E M U 2 0 0 3
in the expenditure ratio which over the same periodwill amount to 20 of GDP Both revenue andexpenditure ratios are reduced in most countriesStrong reductions in revenue are projected in FinlandIreland Greece Italy (1) Belgium the Netherlands andLuxembourg and outside the euro area SwedenFrance Denmark and Sweden still have revenue ratiosabove 50 of GDP in 2005 (2) The UK is set toincrease revenues by 2 of GDP between 2002 and2005 Such an increase should finance the almostequivalent increase in total expenditure which remainsamong the lowest in the EU All the other countries areset to decrease total expenditure In several countries(that is Germany Greece Italy France and Portugal)this ratio is expected to be reduced by around 2 percent-age points of GDP or more
Although the information provided in the programmeson the budget components is not always complete (3) itwould seem that the reduction in taxes which has takenplace in most countries in the euro area in 2001 and in2002 (on average ndash 08 of GDP see Table I14) (4) isnot expected to continue thereafter as the ratio wouldremain constant around 251 until 2006 Sizeablereductions are expected in Finland and to a smallerextent Italy while Germany would increase the ratio byone point by 2004 Outside the euro area the tax to GDPratio in Sweden would remain constant after the largereduction in 2002 mdash due to the reduced revenues fromhigh capital income and corporate taxes mdash while it
Table I13
Expenditure and revenue ratios in the 2002 updates
Total revenues Total expenditures
2002 2005 2002ndash05 2002 2005 2002ndash05
BE 495 484 ndash 11 495 480 ndash 15
DE 450 445 ndash 05 485 455 ndash 30
EL 457 444 ndash 13 468 442 ndash 26
ES 398 398 00 401 397 ndash 04
FR 512 506 ndash 06 540 522 ndash 18
IE (1) 348 329 ndash 19 351 341 ndash 10
IT (2) 460 448 ndash 12 481 463 ndash 18
LU 466 456 ndash 10 470 456 ndash 14
NL 464 453 ndash 11 471 457 ndash 14
AT 515 495 ndash 20 521 510 ndash 11
PT 437 436 ndash 01 466 447 ndash 19
FI 513 490 ndash 23 475 464 ndash 11
EUR-12 462 455 ndash 08 484 464 ndash 20
DK (3) 551 542 ndash 09 529 518 ndash 11
SE (4) 565 554 ndash 11 548 538 ndash 10
UK (5) 380 400 20 389 408 19
EU-15 452 449 ndash 03 470 457 ndash 13
NB Discrepancies are due to rounding The improvement in net lending implied by this table may be different from the one resulting from other tables This is due toinconsistencies across tables in the programmes
(1) 2002 figures reflect corrected treatment of UMTS proceeds (2) Not including for 2005 future unspecified measures amounting to 14 of GDP(3) Figures for 2002ndash04 may not be consistent with those in the tables for GDP growth and budget balances as they have not been revised by the supplementary note(4) 2004 and 2002ndash04(5) Financial years
Source 2002 updates of the stability and convergence programmes
yen1part In the case of Italy future unspecified measures amounting to 14 ofGDP in 2005 have not been distributed across budgetary items
yen2part However as no adjustment is made for differences in institutional rules thecomparability of tax ratios is limited across countries
yen3part No information was given by France and Luxembourg only partial infor-mation by Spain and complete data but only up to 2004 by Sweden Insome cases erroneous classifications in the figures provided have beenidentified
yen4part With the notable exceptions of Austria where taxes increased by 18 ofGDP and in more limited measure the Netherlands where the progressiveincrease in taxes is due to the tax reform which shifts revenue from socialcontributions to taxes and reduces social contributions over time
28
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
should decrease between 2002 and 2006 by about 1 in Denmark and increase by the same amount in the UKSocial contributions would be reduced further in thecoming years by around of GDP in the euro area(see Table I14) At national level Germany would com-pensate the increase in taxes by a reduction of a similarsize in social contributions although in different yearsItaly and to a smaller extent Belgium are also expectedto reduce the ratios somewhat Other revenues areexpected to decrease slightly over the period
Graph I7 presents the contribution to the change in thebudgetary position of four budget components primarycurrent expenditures interest payments gross fixed cap-ital formation and total revenues A number of generalconclusions can be drawn
Firstly the development of expenditure components overthe time frame of the programmes appears to be influ-enced by the initial budgetary and cyclical position Mostcountries showing deficits in 2002 plan to reduce substan-tially the expenditure ratios while most countries showingsubstantial surpluses expect lower revenue Germany andPortugal which plan to improve the balance substantiallyover the period expect to do so essentially via cuts in cur-rent primary expenditure However Portugal would alsoreduce public investment while Germany also plans to
implement further tax cuts Italy Greece and Francewhich plan to improve the balance by around 1 to 2 ofGDP would use part of the large reductions in primarycurrent expenditure and in interest payments to finance taxcuts and increased investment (1) Secondly after a slightreduction in 2002 and 2003 gross fixed capital formationis set to increase at the euro area level to 24 of GDPThis would reflect the large increase expected in publicinvestment in Spain and to a smaller extent in Greecewhich would more than offset the reduction expected inFinland and in Portugal (2) Germany would maintain theinvestment ratio constant although at 15 of GDP alevel almost 1 percentage point lower than the euro-areaaverage The UK projects to increase public investment by07 of GDP between 2002 and 2005 to 21 of GDPstill below the EU average In Ireland the reduction inrevenues is compensated by cuts in public investment andreduction in primary current expenditure (3)
yen1part The increase of public investment in Italy between 2002 and 2005 is to alarge extent due to an accounting effect (see also footnote 17)
yen2part The level of public investment in 2002 and 2003 has also been affected bythe accounting treatment of the sales of real assets by the Italian govern-ment in those years sales which were recorded as a reduction in invest-ment This effect should cease by 2004
yen3part However contingency provisions are made in the Irish programme whichare not included in these calculations
Table I14
Euro area Budget developments for the general government
of GDP 2001 2002 2003 2004 2005 2006
Components of revenues
Taxes 255 251 251 252 251 251
Social contributions 154 153 153 150 149 149
Interest income
Other 41 42 41 39 38 38
Total receipts 466 462 460 458 455 455
Components of expenditures
Collective consumption
Social transfers in kind 147 148 145 141 138
Social transfers other than in kind 169 173 171 167 164
Interest payments 40 37 37 35 35 34
Subsidies 14 13 13 13 13
Gross fixed capital formation 24 23 22 24 24 24
Other 31 32 32 31 31
Total expenditures 481 484 479 470 464 458
NB Totals might not correspond to the sum of the components while for totals information is available for all countries several countries are not included in the aggre-gation concerning budgetary components which affects the ratio of the components
Source 2002 updates of the stability and convergence programmes
29
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Thirdly there are a number of countries which whileimproving the balance marginally expect to reduce thesize of the public sector This is most notably the case of
Luxembourg the Netherlands and in a smaller measureSpain and Belgium Outside the euro area this is the caseof Sweden and Denmark
Graph I7 Contributions to change in budgetary position 2002ndash05 (in points of GDP)
Source 2002 updates of the stability and convergence programmes A positive value indicates a positive contribution to net lending A positive value in total variation of budgetary position (value is presented on top of columns) implies an improvement of the balance For SE data refer to 2002ndash04 For FR and LU values of primary current expenditures refer to primary expenditure Net lending for Italy includes unspecified measures totalling 14 of GDP in 2005 Revised figures for net lending for Denmark do not specify the impact of the revision on budget items
ndash 35
ndash 25
ndash 15
ndash 05
05
15
25
35
BE DK DE EL ES FR IE IT LU NL AT PT FI SE UK
28 12 ndash 09 03 17 ndash 12 ndash 01
Variation of budgetary
position (in points of GDP)
13 02 030805 02 19 ndash 09
Revenues
Gross fixed capital formation
Interest payments
Primary current expenditure
30
3 The sustainability of public finances based on the 2002 updates of stability and convergence programmes
31 Introduction
In recent years growing attention has been paid to theneed to extend EU budgetary surveillance beyond thethree or four year time horizon of stability and conver-gence and to consider whether public finances aresustainable in the long run This largely stems fromconcerns about the potential impact of ageing popula-tions on public finances The importance of securingsustainable public finances is not unique to EMU butthere are additional implications in a monetary unionAn unsustainable public finance position in a partici-pating Member State may complicate the implementa-tion of the single monetary policy and possibly resultin interest rates being higher than they would other-wise be
Since the launch of the euro in 1999 the Commissionhas addressed the issue of the sustainability of publicfinances along a number of lines (1) In particular theCommission has sought to integrate an examination ofthe sustainability of public finances into the existing EUframework for the surveillance of Member Statesrsquo eco-nomic and budgetary policies in line with the conclu-
sions of the Stockholm (March 2001) and Barcelona(March 2002) European Council meetings
The chapter presents the second assessment of long-termsustainability carried out by the Commission and theCouncil on the basis of the 2002 updated stability andconvergence programmes which followed a similarapproach to that followed in the first exercise (see Euro-pean Commission 2002a)
32 How the sustainability of public finances was assessed
321 The quantitative indicators
In the absence of an agreed definition a pragmatic defi-nition of what constitutes a sustainable public financeposition was used namely whether on the basis of cur-rent policies Member States will continue to complywith the budgetary requirements of EMU and in partic-ular the Treaty requirement to keep debt levels belowthe 60 of GDP reference value (2) At the same timehowever it was recognised that sustainability of publicfinances is a multifaceted policy challenge Aside fromavoiding deficits and debt accumulation sustainabilityin addition requires that tax burdens remain at reasona-ble levels and that other non-age-related expenditures(infrastructure RampD) are not squeezed out In recogni-
yen1part Firstly projections for age-related expenditures were published for eachMember State up to 2050 additional projections covering the impact ofageing on public spending on education and unemployment transfers willbe published in mid-2003 Secondly on 4 and 5 March 2003 the Directo-rate-General for Economic and Financial Affairs coorganised a conferencewith the Centre for Strategic International Studies (CSIS) on lsquothe economicand budgetary implications of global ageingrsquo The conference papers areavailable on the web site of DG ECFIN at httpeuropaeuintcommeconomy_financeevents2003events_brussels_0303_enhtm Thirdly the impactof demographic changes on growth has been analysed (see Chapter 4 inEuropean Commission (2002b)) Fourthly the need to ensure the financialsustainability of pension systems has been addressed as part of the open-method of coordination on pensions
yen2part This definition based on compliance with pre-determined and arbitrarybudgetary aggregates can be justified on the grounds that continued com-pliance with the SGP and in particular the lsquoclose to balance or in surplusrsquorequirement would de facto lead to the virtual disappearance of public debtin the long run under reasonable assumptions on growth and interest ratesBalassone and Franco (2000a) also review the various approaches to defin-ing the sustainability of public finances
31
P u b l i c f i n a n c e s i n E M U 2 0 0 3
tion of this the Commissionrsquos assessment examinedboth quantitative and qualitative information
On the basis of the work of the Economic Policy Com-mittee (2001) two groups of indicators were used toquantify the sustainability of public finances
The first indicator consisted of extrapolating debt devel-opments up to 2050 so as to verify whether continuedcompliance with the debt requirements of the Treaty canbe expected on the basis of current policies Under anlsquoSGP compliancersquo scenario the starting position interms of the current budget balance level of debt pri-mary spending and tax revenues are the figures reportedby the Member States for the final year of their 2002updated stability or convergence programme for mostMember States this is 2005 or 2006 The Commissionthen extrapolated the evolution of the budget balanceand debt levels up to 2050 assuming that (i) the tax bur-den and non-age-related primary expenditures remainconstant as a share of GDP at the 200506 level over theprojection period (ii) the interest-growth rate differen-tial converge towards an EU average level of around 2 in 2010 (1) and (iii) age-related expenditures evolve inline with the projections of the EPC or alternativenational projections It is then possible to verify whetherthe projected level of debt respects the requirement tostay below 60 of the GDP reference value for publicdebt at all times (2) Failure to do so would a priori indi-cate that there may be a risk of budgetary imbalancesemerging in light of ageing populations and that meas-ures may be required to place public finances on a moresustainable footing
It should be noted that the lsquoSGP compliancersquo scenarioassumes that Member States actually achieve thebudget targets set down in their programmes whichfor several Member States implies a successful process
of budgetary consolidation to the lsquoclose to balance orin surplusrsquo requirement However such an outcomeis by no means assured since several Member Statesstill have to complete the consolidation A lsquo2002positionrsquo scenario is therefore run in the same way asthe lsquoSGP compliancersquo scenario excepting that thestarting budget position is different Debt levels areextrapolated from 200506 to 2050 assuming that nobudgetary changes occur during the programme periodthat is the primary balance in 200506 is the same asthe 2002 level The purpose of this scenario is todemonstrate the long-term impact on debt develop-ments and consequently on the sustainability of publicfinances of a failure to achieve the lsquoclose to balance orin surplusrsquo requirement of the Pact for those countriesstill in deficit in accordance with the timetable setdown in the Member Statesrsquo stability or convergenceprogrammes
For both scenarios the tax gap has been measured It pro-vides a gauge of the scale of budgetary adjustment whichwould be required for a Member State to reach a sustain-able public finance position It measures the differencebetween the current tax ratio and the constant tax ratioover the projection period necessary to achieve a pre-determined debt level at some date in the future A posi-tive tax gap indicates that there is a financing gap toreach such an objective
The choice of both the targeted debt ratio and the lengthof the projection period is arbitrary and the Commissiontherefore calculated three different tax gaps as follows
bull T-1 measures the difference between the currentand constant tax ratio required to reach the samedebt level in 2050 that would result from running abalanced budget position over the entire projectionperiod By definition the debt ratio would convergetowards zero but the level reached in 2050 will dif-fer across countries depending on the starting debtlevel This approach has the advantage that the debttarget to be achieved is consistent with the budget-ary framework of the SGP and the fact that the EPCprojections for age-related expenditures cover theperiod up to 2050
bull T-2 recognises that a requirement for debt levels toconverge towards zero is an overly strict definitionto ensure the sustainability of public finances Ittherefore measures the difference between the cur-
yen1part Real growth is based on the projections included in the report of theEPC(2001) that is GDP growth convergence to some 175 by 2030 inmost Member States reflecting the assumption on labour force participationrates and in particular a prudent assumption on the rate of productivitygrowth An identical nominal interest rate was assumed for all countriesThe interest rate is defined as the sum of the inflation target of the ECB(2 ) the real growth rate of the EU (converging to 175 by 2030) plusthe differential of two between the nominal interest rate and nominal GDPgrowth This leads to assume a nominal interest rate close to 6 To avoida discrete jump in the debt projections it is assumed that the implicit inter-est rate on debt in the final year of the stabilityconvergence programmeconverges towards the common nominal interest rate over 10 years in a lin-ear fashion
yen2part For countries with debt ratios still above 60 it must convergetowards the reference value and stays below it for the remaining periodof projection
32
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
rent and constant tax ratio required to reach a debtlevel of 40 of GDP in 2050 (1)
bull T-3 is a measure which is close to tax gap measuresfound in the economic literature based on the presentvalue budget constraint It indicates the change intax revenues as a share of GDP that would guaranteethe respect of the intertemporal budget constraint ofthe government that is that equates the actualised
Box I1 The impact of ageing populations on tax revenues and social contributions
Government revenues can be decomposed into four main categories according to the tax base labour capital consumptionand social contributions Revenues for each of these categories are simply the product of the respective effective tax rateby the tax base In the case of social contributions it is the product of wages (the tax base) by the contribution rate Anageing population can have a direct effect on tax revenues through a modification of a tax base
Few studies analyse the consequences of ageing populations on government revenues Goudswaard and Ven de Kar (1994)show that income tax revenues in the Netherlands would increase because of a rising share of older workers in the labourforce as these are the highest paid group for seniority reasons However in the long run wages are driven by labour pro-ductivity more than by seniority and the impact of ageing on labour productivity could therefore be negative Auerbech etal (1989) argue that the capacity to adapt to new technologies is lower for older workers and that technological innovationcan render their human capital obsolete Older workers are also less mobile (both geographically and within sectors andlabour tasks) and this implies a lower capacity of economic systems to adjust to structural changes However CambridgeEconometrics (1997) argues that the evidence that at any given time older workers are paid more implies that they shouldalso be more productive Alternatively it would determine lsquoa shift from profits to wages in national income without anyobvious reason that justifies thisrsquo
Ageing also affects consumption and thus savings If consumption increases as a consequence of a higher propensity toconsume amongst elderly people savings will decrease according to the life-cycle model and this will negatively affectlong-term economic growth and revenues Rosevaere et al (1996) argue that national savings both governmental and pri-vate will decline In particular it is estimated that an increase of the old-age dependency ratio in OECD countries of 20 in the next 30 years will reduce private savings by 6
Martinez-Mongay (2000) shows that the evolution of revenues has been driven mainly by the need to finance increasedlevels of public expenditure In particular revenues have adjusted to the evolution of social transfers He shows thatbetween 1970 and 1998 implicit tax rates on labour increased while the tax base (total wages as a percentage of GDP)decreased In contrast tax rates on consumption did not change sharply According to this study demographic changeswould affect tax revenues only to the extent that they lead to additional expenditure
In any case it is rather difficult to isolate the direct effect of ageing on revenues without taking into account the indirecteffect through changes in income levels and distribution There is an endogeneity problem as economic growth is affectedby ageing and this will determine tax bases and revenues But taxation together with social contribution rates affectsemployment and its structure with relevant consequences on participation rates and on the general level of income as wellas its distribution This makes it difficult to carry out any projection on the impact of ageing on tax revenues
To summarise there is a great uncertainty over the effect of ageing populations on revenues Several factors can lead to anincrease in government revenues for example a better-paid workforce (due to seniority effects) an increase in consump-tion and participation rates However several factors could lead to a decline in tax revenues for example a fall in labourproductivity due to an older workforce and a decline in aggregate savings Therefore in making any long-run projectiona very detailed knowledge of income distribution and its evolution is required (since this can change the tax bases for directand indirect taxes) and account needs to be taken of the indirect effect of taxation on labour participation and on incomelevels However past experience already shows that the level of public spending is the main determinant of tax revenuesas a share of GDP
yen1part Interestingly the UKrsquos sustainable debt rule requires that net debt does notexceed 40 of GDP
33
P u b l i c f i n a n c e s i n E M U 2 0 0 3
flow of revenues and expenses over an infinitehorizon (1) As such there is no target for the debtratio what happens is that this will convergence to arelatively low level Moreover there is no cut-offdate in 2050 and this requires the assumption thatage-related expenditures remain constant as a shareof GDP at the projected level in 2050
It is important to interpret the results of these quanti-tative indicators with caution The projected evolutionof debt levels are not a forecast of possible or evenlikely outcomes Instead they are a tool to facilitatepolicy debate and at best provide rough indication ofthe timing and scale of emerging budgetary chal-lenges that could occur on the basis of lsquono policychangersquo In practice it is likely that governmentswould respond to either explosive debt trajectories orthe implosion of debt leading to the accumulation oflarge net assets
A further limitation of both sets of indicators is thatthey provide little guidance on what is the appropriatebudget target which Member States should aim for inthe light of the expected costs of an ageing populationand indeed other contingencies which may affect pub-lic finances in the future Moreover a positive tax gapdoes not imply that tax rates should be raised but ratherthat a financing gap exists which needs to be closed bya variety of means including raising tax revenues cut-ting non-age-related expenditures andor introducingreforms to curb the growth in age-related expendituregrowth The results are also sensitive to underlyingassumptions on parameters such as interest rates andgrowth rates as well as the starting budget position Tosome extent account can be taken of this by running avariety of sensitivity tests but these provide no esti-mate of the risk or probability of various budgetary sce-narios emerging
Finally the utility of the exercise depends heavily on thequality and comparability of the long-run budgetary pro-
jections If greater weight is to be attached to the sustain-ability of public finances in the EU surveillance processand in particular if the Commission and Council wish toprovide clearer recommendations on policy responsesthen considerable efforts should be made to upgrade theprojections
322 The data used
The code of conduct on the content and presentation ofstability and convergence programmes requires MemberStates to address the issue of sustainability and on a vol-untary basis include long-run budgetary projections AllMember States included a specific section on the sus-tainability in their 2002 programmes and there was amarked improvement in the terms of the quality and cov-erage of information compared with the 2001 pro-grammes
Table I15 summarises the budgetary projectionsincluded in the programmes of Member States Twelveof the 15 programmes included budgetary projectionsfrom national sources whereas three Member Statesreferred to the EPC projections A trade-off exists asregards the choice of which projections to use TheEPC projections were made using common demo-graphic scenarios and agreed assumptions on keylabour market and macroeconomic parameters andwere subject to a peer review exercise by the Commis-sion and Member States However national projectionsmay encompass the impact of recent reforms they mayalso capture in more detail the institutional complexityof national tax and benefit systems
Table I16 presents the projections used by the Commis-sion in running its quantitative indicators A number ofimportant choices taken when doing the projections areworth highlighting
bull The Commission as a general rule used the nationalprojections when they consisted of updates based onthe EPC approach For the most part the differencesbetween the EPC and national projections weremodest and would not influence policy conclusionsHowever Spain Germany and Austria submittedrevised projections for spending on public pensionswhich indicated a much smaller increase in spendingover the projection period The revised projectionfor Spain indicated that spending on pensions willincrease by some 5 percentage points of GDP by2050 compared with 8 percentage points in the EPCprojections and the difference is due to a revised
yen1part The applied formula is the following
where τ is the actual share of revenues on GDP (assumed to remain constant)pg is the share of primary expenditures on GDP (assumed to stay constantafter 2050) is the stock of gross debt on GDP at time t while r and nare respectively the discount rate and the nominal growth rate of the econ-omy (assumed to be constant)
τr nndash----------- 1 TG
τ--------+
bt
pgi 1 n+( )i
i 1=
45
sum1 r+( )i
------------------------------------pg
r nndash( )----------------
1 n+1 r+------------
45+ +=
bt
34
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
demographic scenario (1) The revised projectionmade by Germany takes account of the 2001 reformof the pension system and indicates that spendingon pensions would be 2 percentage points of GDPlower than in the EPC exercise The Austrian projec-tion indicates that age-related spending by 2040 willincrease by 18 percentage points of GDP less thanwhat was projected by the EPC and is due to the useof an alternative demographic scenario Of this dif-ference 1 percentage point relates to public spend-ing on pensions and 07 percentage points to lowerspending on acute healthcare It should be noted that
none of these national projections have been subjectto peer review at EU level and their use results in aconsiderably more favourable profile for debt devel-opment compared to what would have occurred onthe basis of EPC projections
bull EPC projections for spending on healthcare and long-term care were included in the calculations even ifthey were not mentioned in the stability or conver-gence programme Also to ensure consistency theCommission excluded projections for non-age-relatedprimary expenditures indicated by some MemberStates (for example Sweden and the UK) Finallyprojections for changes in the tax ratio were includedfor three Member States (Denmark Netherlands UK)as these can largely be attributed to the deferred taxrevenue contributions to funded pension systems aswell as accumulated earnings prior to disbursement
Table I15
Long-run budgetary projections included in the 2002 updates to stability and convergence programmes ( of GDP)
Source
PensionsHealth
and long-term careOther age-related
expenditureTax revenues
Net impact
2005Change by 2050
2005Change by 2050
2005Change by 2050
2005Change by 2050
BE national 87 27 62 20 73 ndash 16 31
DK national 47 25 74 19 541 22 22
DE national 111 38 60 11 177 13 36
EL national 124 102 50 16 444 46 72
ES national 79 51 na na 51
FR EPC 121 37 69 20 57
IE EPC 38 39 61 17 56
IT national 139 02 59 17 19
LU EPC 74 19 na na 19
NL national 83 53 73 31 23 29 55
AT national 146 18 58 21 39
PT national 133 20 na na 20
FI national 107 37 62 29 537 ndash 21 87
SE national 91 18 98 46 319 ndash 16 531 25 23
UK national 50 ndash 02 70 28 61 08 399 ndash 15 49
NB BE the starting data refers to 2000 Other expenditures include family allowances unemployment and early retirement transfers work-related accidents and sick-ness and residual regimes DK of the change in tax revenues the net tax on pension payouts increased by 24 pp of GDP from 2005 and 2050 Also pensionassets are projected to increase from 119 of GDP in 2005 to 206 of GDP in 2040 DE the starting data refers to 2010 Pension projections were made by theBMGS (statutory pension insurance and public service workers pension) Healthcare projections only cover acute healthcare and were made by the EPC Tax reve-nues only concern taxation of payments to private households and was made by the German Institute for Economic Research EL Healthcare only concerns acutehealthcare FR starting date is 2000 and change refers to the period 2000 to 2040 IE data in programme was reported as a of GNP It was converted to GDPassuming a constant differential of 17 over the projection period NL revenues projections refer to income tax revenues on pensions PT starting data refers to2001 FI starting year is 2000 SE expenditure projections include a breakdown covering childcare primary and secondary education adult education other trans-fer payments (ill health childrenstudies labour market transfer payments to firms transfer payments abroad) and public investment
Source 2002 updates of stability and convergence programmes
yen1part This is based on the recent census which indicates that the existing popula-tion size is considerably higher than estimated by Eurostat and also impor-tant differences as regards inward migration The upshot is that thepopulation of working age is considerably higher in the revised projectionthan assumed by the EPC
35
P u b l i c f i n a n c e s i n E M U 2 0 0 3
bull The concept used in the EU surveillance of MemberStatesrsquo budgetary positions is general governmentgross debt It measures the amount of existing finan-cial debt the government will have to service andreimburse The only asset taken into account is gov-ernment debt held within the government sectorother financial assets such as holdings of shares andequity and real assets do not contribute to lower therecorded level of debt It has been argued that whenassessing long-term sustainability there is also a casefor looking at net debt figures However this wouldentail a number of practical measurement problems asa large part of government assets are of a non-finan-cial nature Real assets are typically not easy to valueand moreover it is questionable to what extent theseassets can be used to redeem outstanding debt or sub-stitute for other revenues In running the quantitativeindicators for Finland and Sweden (1) however the
Commission took on board information on financialassets (other than government bonds) in designatedpension funds as well as information on financialassets specifically designated for privatisation andthus available for future debt reduction It wasassumed that the yield on assets is the same as ondebt
323 The results of the quantitative indicators
The results of the quantitative indicators (both theextrapolation of debt and the tax gap indicators) are pre-sented in Table I17 and Table I18 The need to interpretthe results with caution is again underlined and in par-ticular to avoid drawing mechanical policy conclusionsNotwithstanding the caveats the indicators clearly illus-trate that ageing populations pose a very significantbudgetary challenge and the following broad conclu-sions can be drawn
First even assuming that all Member States achieve theirbudget targets for 2006 (SGP compliance scenario)which in most cases represents a position of lsquoclose to bal-ance or in surplusrsquo there is a risk of unsustainable public
Table I16
Data used to run the sustainability indicators in the lsquoSGP compliance scenariorsquo ( of GDP)
Level in 20056 (1) Change by 2050
Net borrowing
DebtTotal
revenues
Total non-age-related spending
PensionHealth-
care
Other age-related
expenditures
Tax revenues
Net change
BE 05 94 484 210 29 16 ndash 11 00 34
DK 22 26 536 266 20 20 02 00 42
DE 00 58 445 243 35 13 00 00 48
EL 06 88 443 218 102 16 00 00 118
ES 02 47 398 235 51 16 00 00 67
FR ndash 10 57 505 290 34 11 04 00 49
IE ndash 12 35 329 227 39 17 00 00 56
IT 01 89 446 194 02 17 00 00 19
LU ndash 01 3 456 380 19 00 00 00 19
NL 01 45 453 274 35 34 05 37 37
AT ndash 11 61 494 271 18 13 08 00 39
PT ndash 04 53 430 242 21 08 00 00 29
FI 28 ndash 17 489 260 30 10 15 00 55
SE 17 18 554 149 18 14 17 ndash 09 58
UK ndash 16 39 399 214 ndash 02 28 08 15 49 (2)
(1) Denmarkrsquos levels are for 2010(2) The net change for UK includes the change in the total non-age-related spending of ndash 14 of GDP
Source Commission services
yen1part According to the last set of stability and convergence programmes grossgovernment debt in Sweden in 2001 was 523 of GDP while the net debttaking pension fund financial assets into account was ndash 31 of GDP Inthe Finnish programme gross interest payments in 2001 was 28 of GDPwhile net interest payments was 07 of GDP
36
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
finances (measured against the 60 of GDP referencevalue) emerging in some half of EU Member States andindeed for the EU as a whole (see Graph I8) Hence cur-rent policies are not sustainable and further policy meas-ures are needed
Secondly the risk of unsustainable public financesincreases considerably if all Member States do notachieve the SGP goal of budget positions of lsquoclose to bal-ance or in surplusrsquo An indication of this can be seen bycomparing the projected debt levels under the lsquoSGPcompliance scenariorsquo with the lsquo2002 positionrsquo scenariofor the EU-15 the failure to reduce the deficit for its2002 levels of some 2 of GDP would result in debtbeing some 100 of GDP higher in 2050 In particularGraph I9 compares debt developments under both sce-narios for the four euro area countries with highest defi-cits in 2002 ie Germany France Italy and Portugal
Thirdly debt developments for most Member Statesfollow a U-shaped pattern In the coming decade or20 years debt levels are projected to decrease thanks tothe running ofa balanced budget position however thistrend would start to reverse once the budgetary impact ofageing starts to take hold with the largest increase inmost countries expected between 2020 and 2030 There
is therefore a limited but fast closing window of oppor-tunity to reduce debt levels
Fourthly the tax gap indicators provide some order ofmagnitude to the budgetary adjustment needed to ensuresustainable public finances In addition to consolidationefforts to correct the 2002 aggregate underlying deficitof some 2 of GDP the tax gap under the lsquoSGP com-pliance scenariorsquo indicates that an additional permanentbudgetary adjustment of between 1 and 2 percentagepoints of GDP is needed in Member States where thesustainability of public finances is a concern A budget-ary adjustment of this magnitude would be between onethird and one half the size of consolidation achieved aspart of the Maastricht process since 1995 However thescale of budgetary adjustment efforts could be evengreater if age-related spending increases faster than inthe baseline EPCnational projections andor if accountis taken of the stated budgetary objectives of some Mem-ber States such as a reduction in the tax ratio Also andas stated above this does not suggest that taxes should beincreased but rather that an appropriate combination isneeded of tax increases reducing the level of non-age-related primary spending andor reform of pension andhealthcare systems to curtail the impact of ageing onexpenditure growth The scale of such a budgetary chal-lenge is presented in Table I18
Table I17
Projected evolution of debt levels up to 2050
SGP compliance scenario 2002 budget position scenario (1)
2010 2030 2050 2010 2030 2050
BE 70 ndash 21 ndash 108 66 ndash 41 ndash 154
DK 26 ndash 23 ndash 51 9 ndash 79 ndash 172
DE 49 56 89 75 186 384
EL 70 48 160 70 64 201
ES 38 17 89 33 4 59
FR 54 107 248 62 144 335
IE 33 85 220 22 52 153
IT 77 17 ndash 38 88 72 91
LU 2 16 51 4 18 52
NL 39 48 99 37 43 91
AT 59 88 123 61 39 19
PT 46 51 107 61 120 281
FI (2) ndash 25 ndash 48 ndash 39 ndash 42 ndash 135 ndash 225
SE (2) 3 2 ndash 35 3 2 ndash 57
UK 38 43 78 39 49 90
(1) As calculated assuming primary balance constant at the level of 2002(2) Government debt net of financial assets
Source Commission services
37
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Table I18
Results of the tax gap indicator
SGP compliance scenario 2002 budget position scenario
T1 T2 T3 T1 T2 T3
BE ndash 17 ndash 20 01 ndash 24 ndash 27 ndash 05
DK ndash 09 ndash 13 01 ndash 27 ndash 31 ndash 20
DE 10 06 32 48 45 69
EL 20 17 48 26 23 55
ES 11 07 24 10 06 24
FR 37 35 46 48 45 57
IE 34 29 51 24 18 40
IT ndash 08 ndash 10 03 09 06 19
LU 12 02 25 10 03 26
NL 12 07 46 10 06 45
AT 26 21 34 10 06 19
PT 13 10 20 37 33 43
FI ndash 06 ndash 11 ndash 05 ndash 32 ndash 37 ndash 08
SE ndash 06 ndash 11 02 ndash 09 ndash 14 ndash 01
UK 19 14 12 13 08 14
NB T1 indicates the constant difference between projected revenues and the revenues required to reach in 2050 the same debt-to-GDP ratio as the close to balance posi-tion holds for the whole projection period T2 indicates the constant difference between projected revenues and the revenues required to reach in 2050 a debt-to-GDP ratio equal to 40 T3 indicates the change in tax revenues as a share of GDP that guarantees the respect of the intertemporal budget constraint of thegovernment that is that equates the actualised flow of revenues and expenses over an infinite horizon
Source Commission services
Graph I8 A comparison of debt projections for the EU-15 based on the lsquoSGP compliance scenariorsquo and the lsquo2002 starting positionrsquo scenario
EU ndash non-compliance SCPs 2002 EU baseline
debtGDP
60 reference value
2018
2020
2002
2004
2006
2008
2010
2012
2014
2016
2022
2024
2026
2028
2030
2032
2034
2036
2038
2040
2042
2044
2046
2048
2050
20
40
60
80
100
120
140
160
180
38
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
33 Policy conclusions per Member State
The policy conclusions in the Commissionrsquos recommen-dations for Council opinions on updated stability and con-vergence programmes were drawn on the basis of qualita-
tive as well as quantitative analysis They addressed threepolicy questions as follows
bull In the light of projected budgetary implications ofageing populations is it likely that the SGP require-ments will continue to be respected on the basis ofcurrent policies
Graph I9 A comparison of debt projections for four Member States based on the lsquoSGP compliancersquo scenario and the lsquo2002 positionrsquo scenario
ndash 50
0
50
100
150
200
250
300
Italy
France
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
2024
2026
2028
2030
2032
2034
2036
2038
2040
2042
2044
2046
2048
2050
30
60
90
120
150
180
210
240
270
300
Germany Portugal
2002
2004
2006
2008
2010
2012
2014
2016
2018
2020
2022
2024
2026
2028
2030
2032
2034
2036
2038
2040
2042
2044
2046
2048
2050
39
P u b l i c f i n a n c e s i n E M U 2 0 0 3
bull Are the medium-term budget target and other policymeasures outlined in the updates compatible withimproving the sustainability of public finances (1)
bull What is the main policy challenge facing MemberStates and what reform measures should be envisaged
Table I19 below summarises the conclusions on each ofthese questions based on the Commissionrsquos assessmentof the 2002 updated programmes and the respectiveCouncil opinions The risk of unsustainable publicfinances is evident in some half of EU countries notablyGermany Greece Spain France Italy Austria andPortugal There are also particular circumstances forBelgium and Ireland which influence the quantitativeindicators of the sustainability of public finances andunderline the need to avoid a mechanical interpretationof results It is possible to group countries according tothe source of potential budget imbalances and the seri-ousness of the risk as follows
bull In two Member States (Spain and Greece) a largeshare of the risk of emerging budgetary imbalancesis due to the very large projected increase in pensionexpenditure According to the EPC public spendingon pensions alone is projected to grow by 8 of GDPbetween 2000 and 2040 in Spain and 12 of GDP inGreece the highest projected increase of all EU coun-tries although both countries have submitted revisedprojections showing substantially lower increasesSpain has already achieved a budget position of lsquocloseto balance or in surplusrsquo and Greece aims at doing soin the coming two years To ensure sustainability themain challenge is to reform the public pension systemso as to contain any increase in spending as a result ofageing populations
bull In several Member States (notably Germany FranceAustria and Portugal) the risk of emerging budgetaryimbalances is a combination of factors First public
spending on pensions and healthcare in these countriesis projected to grow at or above the average rate of theEU in coming decades Secondly the pace of debtreduction is slow due to persistent and large underly-ing deficits Finally they have a relatively poor labourmarket performance and in particular low employ-ment rates of older workers and a low effective retire-ment age Addressing sustainability therefore requiresa more ambitious and comprehensive approach tack-ling all these challenges rather than the unambiguousand piecemeal approaches evident today
bull High debt countries (Belgium Greece and Italy) facea particular set of challenges in ensuring the fastreduction of debt levels At first sight the quantitativeindicators suggest that these countries appear to be rel-atively well placed to meet the costs of ageing popula-tions But the favourable development in debt levels(and consequently on interest payments) hides adegree of fiscal illusion based on an implicit assump-tion that high debt countries are able to sustain largeprimary surpluses over a long period Running theactual budget surpluses implied by such assumptionsover time may be difficult to ensure for the govern-ment as they will be faced with competing budgetarypressures for tax cuts andor increased public expendi-tures (2) In addition the debt may evolve more slowlythan planned because of stock-flow adjustments Onthis aspect the Council expressed concern about theslow pace of debt reduction in Greece and Italy since1999 due to large and persistent financial operationsbesides the unfavourable growth conditions and slip-page from budget balance targets (3)
bull Several Member States appear to have sustainablepublic finances (Denmark Luxembourg the Neth-erlands Finland Sweden and the UK) They have
yen1part The conclusion of the Stockholm European Council did not alter the goalor purpose of the SGP that is to ensure that Member States have medium-term budget positions that are lsquoclose to balance or in surplusrsquo The Com-mission and Council did not attempt to quantify what constitutes an appro-priate budget position for a Member State in light of the budgetary costs ofageing population Whether countries should set more ambitious budgettargets (including surpluses) in the coming years prior to the budgetaryimpact of ageing populations taking hold is clearly a policy issue which theECOFIN Council must address in the future Indeed several MemberStates already go beyond budget positions of lsquoclose to balance or in sur-plusrsquo and are running large surpluses with the explicit purpose of preparingfor the budgetary costs of ageing populations However the obligation onMember States under the SGP remains unchanged
yen2part An indication of this additional budgetary effort can be gauged by lookingat the required primary surplus needed to sustain a balanced budget posi-tion The Commission has calculated this using the same projected increasein age-related spending and assuming that countries achieve the budget tar-get set down in their stability and convergence programme On averageBelgium is estimated to require an average primary surplus of 3 of GDPover the 2010 to 2020 period whereas Greece and Italy would require pri-mary surpluses of 36 and 37 of GDP respectively This compares withan estimated required primary surplus of between 1 and 2 of GDP inmost other Member States with debt levels below the 60 of GDP refer-ence value
yen3part Moreover for Italy the Commission and Council noted that the relativelysmall projected increase in spending on public pensions is based upon anassumption that the reforms enacted in the 1990s are implemented in full(especially the indexation of the entitlement to prices and the adjustment ofbenefits to increases of life expectancy) and on the basis of the assumptionof a significant increase in labour force participation rates in coming dec-ades
40
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
sound budget positions and in most cases pastreform of their pension systems have strengthenedthe link between contributions and entitlementsNotwithstanding the favourable conclusion ageingpopulations will pose budgetary challenges for thesecountries The maintenance of high tax ratios at over50 of GDP in a number of Nordic countriesrequires continued public support and raises concernabout competitiveness there is also a risk that taxbases may become more mobile in the future whichmay make it more difficult for countries to raise rev-enues For the Netherlands the Council consideredthat some additional measures may be needed if theDutch authorities are to achieve the stated aim ofeliminating public debt within one generation Lux-embourg has to provide pensions to a large numberof non-residents financial sustainability will beinfluenced on the number of cross-border workersRegarding the UK the Council concluded that muchof the financial sustainability of the pension systemdepends on the performance of private pension pro-viders If private provision produces significantlyless than the anticipated coverage or level of pen-
sions future governments may face increasedclaims of means-tested benefits
bull In Ireland the indicators point a policy challengethat sooner or later needs to be addressed despitethe improvement in public finances in recent yearsA financing gap may emerge if public spending onpensions and healthcare in Ireland converge towardslevels in other EU countries and if the tax ratio as ashare of GDP remains unchanged (1)
Table I19
Policy conclusions on the sustainability of public finances
yen1part A number of important qualifications need to be made First and as recog-nised in the Commissionrsquos assessment of Irelandrsquos stability programme themedium-term budget position may be substantially better than indicated bythe programmesrsquo targets as it includes an annual transfer of 1 of GNP tothe National Pensions Reserve Fund and a contingency reserve of some06 of GNP The projected evolution of debt levels would be different ifan adjustment was made for these items Secondly there is considerableuncertainty as to what constitutes the potential growth rate of Ireland andthe time frame over which it could be expected to converge to levels seen inother EU countries The growth assumptions used in the sustainability indi-cator are prudent based on recent experience in Ireland Thirdly it shouldbe borne in mind that the tax ratio in Ireland is the lowest of all EU coun-tries and thus there is greater scope to raise taxes if necessary
Are public finances sustainable
Do the budgetary measures in the programme improve sustainability
What are the key policy measures required
BE Appears to be sustainablebut conditional upon sustain-ing large primary surpluses inthe coming decade or more
Policy of sustaining high primary surpluses shouldlead to a fast pace of debt reduction But thisneeds to be complemented with measures toraise employment rates especially amongst olderworkers as the effective retirement age is oneamongst the lowest of all EU countriesSome progress made as regards draft legislationfor setting up the framework for supplementarypensions
Sustaining high primary surplus over the long runwill be a challenge At the same time it is impor-tant that the budgetary cost of structuralreforms notably those involving tax and non-taxburden reduction be kept consistent with thetargeted budgetary adjustment and the reduc-tion of the government debt ratio be ensured
DK Appears to be sustainable Yes Comprehensive approach benefiting fromthe running of budget surpluses and a projectedaccumulation of large net assets in both pensionfunds and the government sector
The tax ratio will remain high compared to otherindustrialised countries and consideration couldbe given to further reductions in a framework ofsound public finances
DE Clear risk of emerging budg-etary imbalances
If achieved a balanced budget position by 2006would help reduce debt at a faster pace Pensionreform of 2001 has helped improve sustainabilitybut the need for further reforms cannot be ruledout
To ensure sustainability compliance with SGP assoon as possible is essential This needs to beaccompanied with far-reaching reforms to raiseGermanyrsquos very low growth potential Urgentreforms are needed not only in the labour mar-ket but also in social security and benefit systemsin general and for a reduction in the regulatoryburden of the economy
(Continued on the next page)
41
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Table I19 (continued)
Are public finances sustainable
Do the budgetary measures in the programme improve sustainability
What are the key policy measures required
EL Clear risk of emerging budg-etary imbalances
Projected move towards a position of budget bal-ance is welcome But programme does notaddress the core issue of pension reform
Further reforms are required to the pension sys-tem to avoid an unsustainable increase in publicspending The Greek authorities are encouragedto promote supplementary privately-funded pen-sion schemes and to take measures to raise partic-ipation rates and to control the evolution of age-related expenditures
ES Clear risk of emerging budg-etary imbalances
Programme contains commitment to sustain abalanced budget position and provides informa-tion on measures to increase employment ratesMeasures to improve incentives for active ageingand private pension schemes were taken
Risk of unsustainable public finances largelystems from the projected increase in spending onpensions (despite the recent downward revisionon estimate) Reform of the pension systemplanned for in 2004 needs to address the issue offinancial sustainability
FR Clear risk of emerging budg-etary imbalances
Overall approach and in particular a failure toreach a position of lsquoclose to balance or in surplusrsquoby the end of the programme is not consistentwith a commitment to sustainable public financesSome progress however has been made asregards structural measures designed to curbexpenditures in the health sector and the actionsaiming at improving the control of budgetaryexecution in the State sector Also the Frenchauthorities announced their intention to reformpension and healthcare systems
To ensure sustainability compliance with SGP assoon as possible is essential Need to pursue theplanned reform of pension system
IE Outlying country Some riskof emerging budgetary im-balances given projected in-creases in spending on pen-sions and healthcare butthere should be scope tomeet financing challengegiven low tax rates and lowlevels of government debt
Some concern as regards projected move to deficitin coming years However when assessing sustain-ability due account should be taken of a contin-gency provision of 08 of GDP in the deficit ofthe final year of the programme and of the even-tual completion of a large programme of publicinvestment Also the gradual build up of assets inthe National Pension Reserve Fund (annual contri-bution of 1 of GNP) will help bear the budget-ary costs of an ageing population
In a good position to meet the costs of ageingpopulations given high degree of funding of pen-sions and the relatively low tax burden Howevera long-term financing challenge may arise asspending on pensions and healthcare as a shareof GDP approach levels in other EU countries
IT Clear risk of emerging budg-etary imbalances
Strategy to prepare for ageing populations givescause for concern There is a need to implement asustained path of budgetary consolidation withone-off measures replaced with structural oneson the expenditure side Council is especially con-cerned that the risks to the programme deficittargets might imply too slow a pace of reductionin the debt ratio The slowdown in the rate ofdebt reduction projected toward the end of theprogramme period also in connection with somelsquobelow the linersquo operations Italyrsquos ability to copewith the budgetary consequences of ageing isbased on implementation of the major pensionreforms adopted in the 1990s and a large increasein the participation rate
To ensure sustainability compliance with SGP assoon as possible is essential It will be necessarygiven Italyrsquos high debt to sustain primary sur-pluses in the order of 5 of GDP for many yearsAlso the goal of reducing the tax burden canonly be safely and effectively achieved within acomprehensive reform plan on both the expendi-ture and the revenue side Italian authorities areencouraged to adopt further measures to pro-mote supplementary privately-funded pensionschemes and to address the outstanding criticalissue in the public pension system namely thelong transition period to the new contributions-based system This should be coupled with themeasures necessary to raise participation ratesand to control the evolution of age-relatedexpenditures
LU Appears to be sustainable Yes comprehensive approach outlined withmeasures announced to improve the attractive-ness of third pillar private pensions
Sustainability is sensitive to the number of cross-border workers
(Continued on the next page)
42
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
Table I19 (continued)
Are public finances sustainable
Do the budgetary measures in the programme improve sustainability
What are the key policy measures required
NL Appears to be sustainable Yes comprehensive approach outlined althoughadditional measures may be needed if the Dutchauthorities are to achieve the stated aim of elimi-nating public debt within one generation Theconclusion on sustainability relies on projectedincreases in the tax ratio although in part this isdue to increases in the deferred taxes on pensionincome
The strategy hinges upon achieving a large andsustained reduction in the debt ratio which mayprove challenging during economic downturnsand in the face of competing pressures to pursueother budgetary objectives While the stabilityprogramme envisages additional savings beingmade so as to absorb the projected increase inage-related expenditures there is a lack of clarityon the precise measures which will be taken toachieve this goal
AT Clear risk of emerging budg-etary imbalances
The Council welcomes the intentions of the Aus-trian authorities to reform pension and health-care systems in light of ageing populations How-ever a greater degree of budgetary ambition isrequired and Austria should complete the transi-tion to a position of budget balance in line withSGP requirements without delay
Need to sustain sound public finances and possi-bly consider further reform of pensions It is vitalto put into operation the planned pensionreform since the measures outlined in the updateaddress many of the key problems
PT Clear risk of emerging budg-etary imbalances
The programme sets down an ambitious pro-gramme for budgetary consolidation which ifsuccessful would make a significant improve-ment to the sustainability of public finances
To ensure sustainability compliance with SGP assoon as possible is essential Also essential to pro-ceed with reforms to achieve a better control ofpublic expenditures at all levels of governmentand in particular in the healthcare system
FI Appears to be sustainable Yes comprehensive approach outlined benefitingfrom the sustained running of budget surplusesand a reformed pension system that has a highdegree of pre-funding Programme also containsinformation of reforms both planned and under-way which aim at raising employment rates ofolder workers
The tax ratio in Finland is high compared withother industrialised countries A major challengewill be to carry out the planned tax reformswhile safeguarding the achievements of the pastdecade of placing public finances on a sustainablefooting
SE Appears to be sustainable Yes comprehensive approach outlined benefitingfrom the sustained running of budget surplusesof 2 of GDP up to 2015 and a reformed pen-sion system that automatically limits futureexpenditure growth
Policy aim of running large surpluses may provedifficult over a long time period A challenge willbe to complete the tax reform while safeguard-ing the achievements of the past decade of plac-ing public finances on a sustainable path
UK Appears to be sustainable The deficit targets in the programme raise someconcern as regards the sustainability of publicfinances A budgetary position of a limited deficitin the medium term would help avoid any risk ofemerging budget imbalances in the context ofageing populations and give greater assurance tothe programme view that lsquothe public financesbased on current policies are sustainable in thelong-termrsquo
Much of the financial sustainability of the pen-sion system depends on the performance of pri-vate pension providers If private provision pro-duces significantly less than the anticipatedcoverage or level of pensions future govern-ments may face increased claims of means-testedbenefits
Source Based on the policy conclusions in the Commissionrsquos assessment of the 2002 updates to stability and convergence programmes and the respective opinions of the Council
43
4 Budgetary developments in candidate countries
41 Short-term budgetary developments and prospects in candidate countries
In 2002 the aggregate budget position of the 13 candi-date countries (CC-13) (1) improved but only due to theexceptional advance recorded in Turkey (see Table I20) (2)The aggregate general government deficit of the 10 coun-tries set to become EU members in May 2004 (AC-10)widened This deterioration occurred despite the fact thataggregate growth for the AC-10 continued at roughly thesame pace as in 2001 (3)
Aggregate budget positions are projected to improve forall country groupings in 2003 and 2004 Despite a signif-icant acceleration in growth however the projectedreduction in the aggregate deficit of the AC-10 is not suf-ficient to reverse the deterioration recorded in 2002 Thissuggests that structural rather than cyclical factorsunderlie current budgetary imbalances
Due caution however should be taken when interpret-ing budgetary trends for the CC-13 Despite significantprogress budgetary data for these countries are still notfully comparable across countries nor completely in linewith EU definitions (see Box I2) Significant revisionsin the budget positions of these countries are still possi-
ble and from a methodological point of view aggregat-ing country figures is only possible to a limited degree
Aggregate figures tend to hide the differences amongindividual countries Outcomes for 2002 range from adeficit of 137 of GDP in Turkey to a surplus of 13 of GDP in Estonia (see Table I20) Relative to 2001 thebudgetary position worsened in seven countries mdash andby more than 1 of GDP in the majority of cases Themost noticeable improvement was recorded in the caseof Turkey followed by Romania and Estonia
Among the seven countries undershooting the budgetarytargets for 2002 set out in their pre-accession economicprogrammes (PEPs) of 2002 Cyprus Malta and aboveall Hungary missed their objectives by a rather largeamount (see Table I20) (4) Five countries on the otherhand overachieved their targets most notably Estoniawhich further increased its surplus position despite hav-ing originally planned to run to a small deficit
In most cases country-specific factors rather than gen-eral macroeconomic trends seem to lie behind countriesrsquobudgetary performance relative to targets Electoraldynamics for instance appear to have played a relevantrole in the case of some of the countries missing theirPEP targets such as Hungary Latvia and the SlovakRepublic Statistical reclassifications and one-off meas-ures also played a part most notably in the case ofHungary (5)
yen1part The CC-13 are Bulgaria Cyprus the Czech Republic Estonia HungaryLatvia Lithuania Malta Poland Romania the Slovak Republic Sloveniaand Turkey The AC-10 exclude Bulgaria Romania and Turkey
yen2part Accounting factors underpin this improvement At 137 of GDP Tur-keyrsquos general government deficit remained very high in 2002 but nearlyhalved relative to 2001 when expenditures were boosted on a one-off basisby the inclusion (in a single year) of the large transfers to the agriculturalsector that had been channelled through the banking system in previousyears
yen3part The sources for all figures used in this section are the 2002 pre-accessioneconomic programmes the 2002 fiscal notification and the Commissionforecast of spring 2003 Given the cut-off data for the preparation of thisreport new and revised budgetary data reported to the Commission in thecontext of the 2003 fiscal notification exercise could not be taken intoaccount
yen4part Following elections the new government of the Slovak Republic formallyrevised upwards its PEP deficit targets prior to the finalisation of the Com-missionrsquos assessements The Slovak Republic is projected to have compliedwith its revised deficit target for 2002 which was 32 of GDP higherthan in the original PEP submission
yen5part Reclassifications also contributed to the upward revision of the PEP deficittarget of the Slovak Republic
44
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
Growth on the other hand did not influence budgetaryperformances uniformly across countries For instanceof the eight countries for which growth in 2002 turnedout higher than envisaged in the PEP framework (seeTable I21) only five achieved a better than targetedbudgetary balance (1)
Looking ahead to 2003 and 2004 the Commissionspring 2003 forecast envisages an improvement in thebudgetary balances of eight countries with particularlymarked deficit reductions in the cases of TurkeyHungary the Slovak Republic and to a more limitedextent Malta (see Table I20) With Estonia projected tomove from a surplus to a small deficit position a rela-tively small deterioration is also expected in the case ofLithuania and Latvia Relative to 2002 country positionswould become less diverse with deficits ranging from06 of GDP in the case of Estonia to 69 in Turkey
Box I2 Candidate countriesrsquo budgetary data and EU standards
The data utilised in this section approximate ESA95 definitions for the general government statistics as much as possibleHowever due to methodological and data availability problems this is only partially possible As the harmonisation ofstatistics progresses significant revisions of general government deficits may be needed Problems of comparability alsoaffect data on the level of total expenditure and revenue and their components
In estimating the data used in this section the Commission services relied upon the government deficit and debt figuresreported in the 2002 fiscal notifications Candidate countries have been formally notifying fiscal statistics to the Commis-sion since 2001 using the same format and aiming at producing the same data as the notifications provided by the MemberStates in the framework of the excessive deficit procedure By completing this exercise candidate countries are becomingfamiliar with the technical and quantitative requirements they will have to apply as soon as they become Member States
The April 2002 fiscal notifications showed that a majority of countries were well advanced in the application of the EUmethodology However further work was still required in all cases and progress remained uneven In particular Estoniashowed a degree of good practice the Czech Republic Hungary Latvia Malta Poland the Slovak Republic and Sloveniawere well advanced in the application of the EU methodology and significant work was still necessary in Bulgaria CyprusLithuania Romania and Turkey
The following issues raised particular concerns
bull While the exhaustiveness of general government statistics had been improved by integrating the activities of privatisationagencies and debt consolidation institutions as well as the quasi-fiscal activities of public enterprises and financial institu-tions further work was still necessary to verify that reclassified revenue and expenditure items were completely and correctlytaken into account
bull Despite a more extensive reliance on accrual figures these often constituted only preliminary estimates In many casesthe correct statistical treatment of large tax and social contribution arrears posed a particular challenge
bull There remained a need to determine with greater precision the component of budgetary support to the enterprise sectorthat constituted a transfer element
bull The classification of compulsory pension funds either within the social security sub-sector of the general governmentor within the insurance sector remained an open question
Candidate countries notified new figures in April 2003 This new set of fiscal notifications is expected to show furtherprogress in the quality and comparability of CC-13 government deficit and debt figures However this report could nottake them into account as their assessment by the Commission services was still ongoing at the time of publication
yen1part These were Bulgaria Estonia Lithuania Romania and the Slovak Repub-lic A diverse picture also emerges if one compares budgetary performancewith the difference between actual and projected nominal GDP growth
45
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Table I20
General government balances in candidate countries ( of GDP)
Actual Forecast PEP target
2001 2002 2003 2004 2002 2004
CY ndash 30 ndash 35 ndash 40 ndash 35 ndash 26 ndash 06
CZ ndash 55 ndash 65 ndash 63 ndash 59 ndash 64 ndash 57
EE 05 13 ndash 05 ndash 06 ndash 02 00
HU ndash 42 ndash 91 ndash 49 ndash 37 ndash 57 ndash 30
LV ndash 19 ndash 25 ndash 29 ndash 26 ndash 18 ndash 22
LH ndash 23 ndash 18 ndash 19 ndash 20 ndash 19 ndash 16
MT ndash 70 ndash 61 ndash 52 ndash 41 ndash 52 ndash 39
PL ndash 31 ndash 42 ndash 42 ndash 40 ndash 41 ndash 33
SK ndash 54 ndash 77 ndash 53 ndash 38 ndash 78 ndash 38
SI ndash 25 ndash 18 ndash 15 ndash 12 ndash 18 ndash 10
AC-10 ndash 37 ndash 53 ndash 44 ndash 39 ndash 47 ndash 34
BG 04 ndash 07 ndash 06 ndash 05 ndash 08 ndash 05
RO (1) ndash 33 ndash 24 ndash 24 ndash 24 ndash 27 ndash 24
TR ndash 289 ndash 137 ndash 98 ndash 69 ndash 131 ndash 29
CC-13 ndash 124 ndash 71 ndash 57 ndash 45 ndash 66 ndash 31
(1) For Romania 2003 spring forecast adjusted to estimated ESA95 balance
Source Commission spring 2003 economic forecasts and 2002 PEPs
Table I21
GDP growth in candidate countries ( pa)
2002 Average 2003ndash04
Forecast PEP Difference Forecast PEP Difference
CY 20 28 ndash 08 29 46 ndash 17
CZ 20 30 ndash 10 33 38 ndash 05
EE 56 43 13 50 58 ndash 08
HU 33 40 ndash 07 39 45 ndash 06
LV 61 50 11 57 56 01
LH 59 47 12 47 54 ndash 06
MT 30 33 ndash 03 34 33 01
PL 13 10 03 31 40 ndash 09
SK 44 38 06 41 42 ndash 01
SI 30 36 ndash 06 36 44 ndash 08
AC-10 24 25 00 35 42 ndash 07
BG 43 40 03 47 51 ndash 04
RO 49 45 04 49 54 ndash 04
TR 78 39 39 41 50 ndash 09
CC-13 43 31 12 39 46 ndash 07
Source 2002 PEPs and Commission services
46
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
According to the projections six countries among theAC-10 would have a general government deficit above3 of GDP in 2004
As a result of the trends outlined above the Commissionforecasts that all countries will miss the 2004 targets setout in the 2002 PEPs with the exception of BulgariaRomania and the Slovak Republic (see Table I20)Targets would be missed by a particularly significantamount in the cases of Turkey Cyprus and to a muchmore limited extent Hungary and Poland Various fac-tors drive these developments including the slippagefrom the targeted adjustment path accumulated by sev-eral countries in 2002 and the worsening of economicprospects (1) Moreover in some cases like Latvia gov-ernments have modified their medium-term budgetarytargets following the submission of the 2002 PEPsFinally in the case of Turkey less optimistic assump-tions on the projected decline of interest rates explain alarge part of the divergence between the Commissionforecast and the PEP targets More information on thelatter is provided in the following section
42 Overview of the 2002 updates of the pre-accession economic programmes
421 Introduction
The examination of the second set of pre-accession eco-nomic programmes (PEPs) submitted by candidate coun-tries within the framework of the pre-accession fiscalsurveillance procedure (see Box I3) was completed inNovember 2002
Compared to the PEPs submitted for the first time bycandidate countries in 2001 the 2002 updates revealed agood and improved effort to develop a consistent andcredible medium-term macroeconomic framework Theprogrammesrsquo information content and their comparabil-ity across countries was greatly enhanced by the presen-tation of data through standardised tables based uponthose envisaged under the code of conduct for currentMember States as well as by the provision of detailedestimates of fiscal variables in principle according toESA95 methodology (2)
yen1part For all countries except Latvia and Malta the average growth rate over the2003ndash04 period is projected to fall below that envisaged in the 2002 PEPs(see Table I21)
yen2part The only exception was Turkey which provided budgetary data based uponGFS methodology Apart from Cyprus Lithuania and Romania all countriesupdated the estimates presented in the fiscal notification of April 2002
Box I3 The pre-accession fiscal surveillance procedure for candidate countries
(Continued on the next page)
In line with the call for the establishment of an annual fiscal surveillance for the candidate countries contained in the 1999and 2000 accession partnerships the so-called pre-accession fiscal surveillance procedure was established in 2001 ThePFSP aims at preparing the candidate countries for the participation in the multilateral surveillance and economic policycoordination procedures currently in place in the EU as part of economic and monetary union
As explained in European Commission (2002a) the PFSP has three components
1 The notification of budget positions mdash requires candidate countries to report data on their general government def-icits and debt in the same format as that used by existing Member States Notifications are then evaluated by the Com-mission services in order to monitor the countryrsquos fiscal positions determine compliance with ESA95 standards andassess their quality as a basis for fiscal analysis
2 The pre-accession economic programmes mdash are submitted on an annual basis by each candidate country for theCommissionrsquos evaluation PEPs have two main aims First to outline the medium-term policy framework includingpublic finance objectives and structural reform priorities needed for EU accession Second they offer an opportunityto develop the institutional and analytical capacity necessary to participate in EMU with a derogation from the adop-tion of the euro upon accession particularly in the areas of multilateral surveillance and coordination of economicpolicies
47
P u b l i c f i n a n c e s i n E M U 2 0 0 3
While all programmes reflected the main challengesahead for the acceding countries the degree of detailedanalysis differed across countries and policy areas as didthe specificity and credibility of the medium-term eco-nomic and fiscal scenarios A rather common problemwas that the costs of structural reforms were insuffi-ciently quantified and integrated in the budgetary frame-work More generally further analytical capacity-build-ing still seemed required for all countries More detailedinformation on the sources of fiscal risks the budgetarycosts of on-going reforms the long-term sustainabilityof public finances and cyclically-adjusted budget bal-ances also appeared to be needed
422 Medium-term budgetary developments
The medium-term macroeconomic framework for the2002 PEPs covers the period 2001 to 2005 For mostcountries the framework envisages accelerating growthdeclining inflation and persisting external imbalances(see Table I22) Although growth projections were gen-erally revised downwards relative to the 2001 PEPs inview of the deterioration in the international economicenvironment growth is generally expected to acceleratein the period 2002ndash05 relative to 2001 (1)
Against this background and taking as a starting pointthe 2001 general government balances most budgetaryplans presented in the 2002 PEPs envisage an improve-ment by 2005 with nine countries planning to reducetheir budget deficits by 2005 thus leading to a fall in theaverage deficit for both the CC-13 and the AC-10 (seeTable I23) Among the four remaining countries Bul-
garia and Estonia plan to move from a small surplus to abalanced budget leaving only Latvia and the CzechRepublic with a projected increase in the general govern-ment deficit over the programme period In the case ofthe Czech Republic in particular the budget deficit wasprojected to increase from 5 of GDP in 2001 to 55 of GDP in 2005 after peaking at 64 in 2002 In 2005projected budget outcomes would vary from a balancedbudget in Bulgaria and Estonia to a deficit of 55 ofGDP in the Czech Republic Among the AC-10 only theCzech Republic and Malta refrained from targeting adeficit below 3 of GDP in 2005
Primary balances are also projected to improve over theprogramme period both on average and for the majorityof individual countries After being projected to worsenin eight cases over 2002 in fact by 2005 primary bal-ances are targeted to improve relative to 2001 for eightcountries The Czech Republic and Latvia would then bethe only countries left running a primary deficit
Compared to the 2001 PEPs eight countries presentedless ambitious budgetary paths in the 2002 updates lead-ing to a deterioration in the average deficit target for theAC-10 over the 2002ndash04 period Among the factorsunderpinning these revisions the reassessment of eco-nomic growth prospects does not seem to have played aconsistently relevant role (2) Contrary to what one mayexpect in fact lower growth projections are met byhigher deficit targets in only five cases out of nine (andto a significantly different degree in each of these)
Box I3 (continued)
3 The discussions in a multilateral context mdash allow present and prospective Member States to jointly debate the fiscalnotifications the PEPs and their evaluation by the Commission Discussions take place in two steps First at a high-level meeting between the member of the Economic and Financial Committee and their counterparts from candidatecountries Secondly at a yearly ministerial meeting between Ecofin and their counterparts
In this context on 5 November 2002 ministers concluded inter alia that lsquosound and credible fiscal policy is crucial notonly for coping with difficult economic policy choices but also for enhancing confidence in the stability of the macro-economic policy framework The weak fiscal positions of several acceding countries argue strongly for taking decisivesteps towards sustainable fiscal consolidation in line with the EUrsquos fiscal surveillance procedures inter alia so as to createroom for private investment Effective public expenditure management and efficient tax collection should be central ele-ments of any consolidation programme Long-term challenges due to ageing populations have also to be factored inrsquo
yen1part Even when a deceleration is expected as in the cases of Latvia Lithuaniaand Romania the average rate of growth is projected to remain above 5
yen2part Other potential factors include a worse starting position than originally tar-geted in the 2001 PEPs methodological changes in the statistics for thegeneral government sector and of course changes in the political willing-ness to pursue the budgetary targets originally set out in the 2001 PEPs
48
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
Table I22
Macroeconomic projections in the 2002 PEPs
Real GDP growth(Annual percentage change)
Consumer price inflation (Annual percentage change)
Current account balance(percentage of GDP)
2001 2002ndash05 (1) Revision (2) 2001 2005 2001 2005
CY 40 42 ndash 06 20 20 ndash 43 ndash 14
CZ 33 37 ndash 03 47 34 ndash 46 ndash 35
EE 50 55 ndash 06 58 35 ndash 61 ndash 64
HU 38 46 ndash 12 92 30 ndash 22 ndash 33
LV 77 55 ndash 03 25 30 ndash 97 ndash 67
LH 59 53 03 13 41 ndash 48 ndash 70
MT ndash 08 34 01 29 24 ndash 50 ndash 24
PL 10 36 ndash 03 55 31 ndash 41 ndash 57
SK 30 43 00 71 45 ndash 86 ndash 42
SI 33 44 ndash 05 84 46 ndash 04 02
AC-10 25 40 ndash 04 59 33 ndash 43 ndash 46
BG 40 49 ndash 14 74 35 ndash 60 ndash 52
RO 53 51 ndash 02 345 80 ndash 59 ndash 35
TR ndash 74 47 na 544 98 23 ndash 08
CC-13 ndash 01 43 na 228 56 ndash 26 ndash 34
(1) Annual average over the period 2002ndash05 (2) Difference between the average rate of growth over the period 2002ndash04 in the 2001 and 2002 PEPs
Source 2001 and 2002 PEPs Commission services
Table I23
General government balances in the 2002 PEPs ( of GDP)
Nominal balance Primary balance Cyclically-adjusted balance (3)
2001 2005 Change Revision (2) 2001 2005 Change 2001 2005 Change
CY ndash 30 ndash 03 27 ndash 09 26 48 22 na na na
CZ ndash 50 ndash 55 ndash 05 ndash 15 ndash 38 ndash 36 02 ndash 53 ndash 56 ndash 03
EE 02 00 ndash 02 06 05 03 ndash 02 na na na
HU ndash 41 ndash 25 16 17 02 07 05 na na na
LV ndash 16 ndash 20 ndash 04 ndash 13 ndash 10 ndash 11 ndash 01 ndash 19 na na
LH ndash 19 ndash 15 04 ndash 02 ndash 02 00 02 na na na
MT ndash 70 ndash 31 39 ndash 01 ndash 34 02 36 ndash 68 ndash 27 41
PL ndash 35 ndash 22 13 ndash 08 ndash 06 15 21 ndash 36 ndash 26 10
SK (1) ndash 54 ndash 20 34 ndash 04 ndash 20 04 24 ndash 39 ndash 27 12
SI ndash 25 ndash 08 17 ndash 07 ndash 05 09 14 ndash 18 09 27
AC-10 ndash 38 ndash 27 11 ndash 05 ndash 11 03 14 na na na
BU 04 00 ndash 04 08 41 32 ndash 09 na na na
RO ndash 34 ndash 24 10 05 07 02 ndash 05 ndash 31 ndash 23 08
TR ndash 151 ndash 05 146 na 86 78 ndash 08 na na na
CC-13 ndash 66 ndash 19 47 na 19 24 05 na na na
(1) Figures for the Slovak Republic refer to the first draft of its 2002 PEP because its officially revised draft did not include a full set of figures The revised deficit targetfor 2005 equals 33 of GDP
(2) Difference between annual averages over the 2002ndash04 period in the 2001 and the 2002 PEPs(3) Countriesrsquo own estimates as presented in the 2002 PEPs
Source 2002 PEPs and Commission services
49
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Shedding more light on this issue would require relia-ble indicators of cyclically-adjusted balances but onlya few countries provided some preliminary estimates inthis regard in their 2002 PEPs Moreover these figuresstill need to be interpreted with considerable cautionThe institutional capacity to estimate cyclically-adjusted balances in fact is still being developed whileshort time series and strong structural changes make itdifficult to isolate structural relationships With thesole exception of the Slovak Republic however theestimates provided would indicate that the budget defi-cits recorded in 2001 were generally equal to the struc-tural deficits With the cyclical component of thebudget playing a relatively small role in the plannedadjustment structural changes in revenue and expendi-ture would be required to achieve the targets set out inthe 2002 updates
In line with decreasing deficits and high nominal GDPgrowth (see Table I24) most countries expect theirgeneral government debt-to-GDP ratios to fall andsharply so in the cases of Turkey and Bulgaria Theonly significant exceptions are the Czech Republic andPoland where the debt-to-GDP ratio is projected to riseconsiderably by the end of the programme period Nev-ertheless according to the projections presented in thePEPs by 2005 all countries would have a debt-to-GDP
ratio below 60 except Malta and Turkey In both ofthese cases however the ratio would be on a decliningtrend
Table I24
General government debt in the 2002 PEPs ( of GDP)
2001 2005 Change
CY 546 512 ndash 34
CZ 236 347 111
EE 48 37 ndash 11
HU 530 500 ndash 30
LV 159 180 21
LH 231 231 00
MT 653 611 ndash 42
PL 387 456 69
SK 430 381 ndash 49
SI 275 244 ndash 31
AC-10 369 409 ndash 41
BU 663 463 ndash 200
RO 233 260 27
TU 1228 730 ndash 498
CC-13 599 483 ndash 116
Source 2002 PEPs and Commission services For the Slovak Republic first ver-sion of the 2002 PEP
Table I25
General government revenue and expenditure in the 2002 PEPs ( of GDP)
Revenue Expenditure
2001 2005 Change 2001 2005 Change
CY 405 422 17 435 425 ndash 10
CZ 421 413 ndash 07 471 468 ndash 03
EE 386 384 ndash 02 384 384 00
HU 461 425 ndash 36 502 450 ndash 52
LV 414 386 ndash 28 430 406 ndash 24
LH 342 361 19 361 376 15
MT 374 358 ndash 17 444 388 ndash 56
PL 418 422 04 453 445 ndash 08
SK 412 398 ndash 14 466 418 ndash 48
SI 431 425 ndash 06 456 433 ndash 23
AC-10 421 415 ndash 06 459 442 ndash 17
BU 406 350 ndash 56 403 350 ndash 53
RO 367 346 ndash 21 401 370 ndash 31
TU 421 401 ndash 20 572 406 ndash 166
CC-13 415 402 ndash 13 482 422 ndash 60
Source 2002 PEPs and Commission services For the Slovak Republic first PEP version
50
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
423 Composition of the adjustment
The 2002 PEPs show that most countries plan to reducethe size of the general government sector in terms of bothrevenue and expenditure ratios (see Table I25) Averag-ing some 13 of GDP the planned reduction in reve-nues appears particularly sharp in the cases of BulgariaHungary and Latvia Only Cyprus Lithuania and Polandforesee an increase in revenues over the programmeperiod As regards the composition of these changes anumber of countries including Bulgaria HungaryLatvia and Lithuania expect a significant cut in taxreceipts often due to sizeable reductions in direct taxesand company taxation in particular Being bound by spe-cific acquis provisions the changes in indirect taxes aremore limited (1)
With the planned reduction in the revenue ratio tending toincrease budget deficits the targeted improvements inbudgetary balances generally hinge upon a relatively
sharper reduction in the expenditure ratio (see Table I25)All countries programme a reduction in their expenditureratio with the exception of the two with the lowest ratio in2001 that is Estonia and Lithuania Reduction targetsappear particularly ambitious in the case of BulgariaHungary Malta and the Slovak Republic mdash each aimingto cut outlays by some 5 of GDP mdash and Turkey
Turkey however constitutes a special case as theplanned cut in the expenditure ratio by nearly 17 percent-age points of GDP is almost fully accounted for by thedramatic fall in interest payments expected to follow thenormalisation of its macroeconomic situation In theother countries instead planned budgetary retrenchmentis driven in most cases by cuts in current expendituresand collective consumption in particular (see Table I26and Graph I10)
Most of the PEPs also envisage a gradual reduction insubsidies but only marginally in the Czech RepublicHungary and Poland that is three of the five countrieswhere in 2001 subsidies amounted to more than 2 ofGDP With the sole exception of Poland and to a lesser
yen1part Most of the adjustments required by the acquis in the area of VAT havealready been made whereas in many countries there is still further need foradjustment in the area of certain excise duties (such as for tobacco)
Table I26
Composition of general government expenditure in the 2002 PEPs ( of GDP)
Collective consumption
Social transfers SubsidiesGross fixed capital
formationOthers including
interest
2001 2005 Change 2001 2005 Change 2001 2005 Change 2001 2005 Change 2001 2005 Change
CY 97 78 ndash 19 149 157 08 16 06 ndash 10 37 37 00 136 147 11
CZ 81 81 00 245 245 01 28 26 ndash 02 50 47 ndash 03 68 70 02
EE 202 193 ndash 09 111 109 ndash 02 08 10 02 34 43 09 29 29 00
HU 101 82 ndash 19 214 234 20 29 24 ndash 05 51 39 ndash 12 107 71 ndash 36
LV 84 na na 229 na na 11 na na 41 na na 65 na na
LH 72 81 09 227 233 06 07 07 00 22 28 06 33 27 ndash 06
MT na na na na na na na na na na na na na na na
PL 72 66 ndash 06 251 235 ndash 16 25 24 ndash 01 26 25 ndash 01 79 95 16
SK 100 80 ndash 20 186 194 08 12 10 ndash 03 27 22 ndash 05 140 118 ndash 23
SI 81 74 ndash 06 180 173 ndash 07 14 13 ndash 01 25 24 00 157 149 ndash 09
AC-10 (1) 83 76 ndash 08 231 227 ndash 04 24 22 ndash 02 34 31 ndash 03 88 88 00
BU 98 82 ndash 16 142 142 00 24 13 ndash 11 35 29 ndash 06 104 84 ndash 20
RO 63 53 ndash 10 99 99 00 21 16 ndash 05 32 40 08 186 162 ndash 24
TU 175 162 ndash 13 91 90 ndash 01 09 10 01 42 43 01 ndash 317 ndash 305 12
CC-13 (1) 107 97 ndash 10 176 173 ndash 03 19 17 ndash 02 36 35 ndash 01 ndash 13 ndash 13 00
(1) Weighted averages excluding Latvia and Malta
Source 2002 PEPs and Commission services For the Slovak Republic first PEP version
51
P u b l i c f i n a n c e s i n E M U 2 0 0 3
extent Slovenia social transfers (1) would not be cut sig-nificantly in any country and would actually increasequite markedly in Hungary Apart from the latter publicinvestment would be mostly shielded from expenditurecuts with the (unweighted) average public investmentratio remaining around 35 of GDP (2) Over the2002ndash05 period the average ratio of government grossfixed capital formation to GDP would be above 4 inthe case of the Czech Republic Estonia Hungary andTurkey
424 Other considerations
When viewed against the rigidity of primary expenditurerecorded in the past (3) the fall in expenditure ratiosrequired to achieve the deficit and revenue reduction tar-gets of the 2002 PEPs highlight the difficult task facedby the authorities in implementing their fiscal plansCompounding this challenge and arguably weakening
the programmesrsquo credibility expenditure cuts are back-loaded in a large majority of the PEPs In only threecases expenditures are expected to be already cut in2002 and only Hungary Malta and Romania aim toachieve a large share of total expenditure adjustment in2003 In seven cases half or more of the total expendi-ture cuts would have to be implemented in 2004 and2005 In addition as shown in Section 12 the Commis-sion 2003 spring forecast indicates that 11 countries outof 13 would fail to meet the PEP budgetary goals for2004 An even larger adjustment than planned in the2002 PEPs would therefore have to be implemented in2005 to achieve the programmesrsquo end-targets
Quite apart from the possibility of partial implementa-tion a wide range of risks to countriesrsquo budgetary plansare identified in the 2002 PEPs most of which stress thedanger posed by governmentrsquos off-budget liabilities Interms of overall stocks of guarantees these seem to berelatively high in Malta and Romania moderate inCyprus the Czech Republic Hungary Lithuania andSlovenia and relatively low in the other countries forwhich information is provided Of course the assess-ment of underlying fiscal risks cannot rely solely on thelevel of the existing stock of guarantees Yet only a few
Graph I10 Contributions to change in budgetary position 2001ndash05 (1) (in points of GDP)
(1) Source 2002 updates of the pre-accession economic programmes A positive value indicates a positive contribution to the change in budgetary positionA positive value for the total variation of the budgetary position (figure presented in bold) implies an improvement of the balance For LV primary cur-rent expenditure refers to total expenditure as no data were provided for interest payment and gross fixed capital formation in 2005 For MA no dataavailable on gross fixed capital formation
yen1part In kind and other than in kindyen2part In Hungary cuts in gross fixed capital formation (from 51 of GDP in
2001 to 39 of GDP in 2005) would account for nearly a quarter of theplanned reduction in the expenditure ratio However Hungaryrsquos publicinvestment ratio was the highest among accession candidates in 2001 andwould remain above the (unweighted) average for the group in 2005
yen3part See World Bank (2002)
52
P a r t IC u r r e n t d e v e l o p m e n t s a n d p r o s p e c t s
countries provided information on the estimated annualbudgetary impact stemming from these contingent liabil-ities making this a key area for the provision of furtherinformation in the next PEP updates (1)
This also applies to the assessment of the long-term sus-tainability of acceding countriesrsquo public finances For thefirst time in 2002 countries were asked on a voluntarybasis to provide data in this regard along the format pro-vided for Member Statesrsquo convergence and stability pro-grammes Only very few countries however provided(incomplete) data
Almost all countries however identified the reform ofthe pension system as one of the key domestic policyareas linked to medium-term fiscal sustainability Inmost cases in fact long-term demographic projectionssuggest that the first (compulsory non-funded) pillar ofthe pension system would become overburdened thusconstituting a source of medium-term budgetary risks Inview of this trend all countries either intend to reformtheir first pillar or have recently done so notably bymatching individual benefits more closely to individualpast contributions and in same cases by raising theretirement age or by adjusting pension indexation rulesAbout half of the candidate countries have introduced amulti-pillar pension system with several others planningto do so (see Table I27) (2)yen1part As for the other identified risks these appeared to be more country-spe-
cific Lithuania for instance highlighted additional risks linked to hard-to-predict remaining costs of the transition process such as those stemmingfrom the restitution of savings and real estate ownership rights the debts ofState-owned enterprises and the decommissioning of the Ignalina nuclearpower plant Restitution issues could also represent a significant share ofGDP under the most pessimistic scenarios presented in the Polish PEPRomania identified policy failures linked to the non-elimination of quasi-fiscal deficits or to additional bank bailouts as the main fiscal risk
yen2part Healthcare reform is another area high on the agenda in most countrieswith considerable implications for the long term sustainability of publicfinances Planned and on-going reforms often include the introduction of amixed publicndashprivate model for insurance and health services provision
Table I27
Main measures in the PEPs concerning pension reform
Introduction of mandatory funded-pillar
Planned reforms
Bulgaria radic Balance 1st pillar by 2007Increase contribution compliance
Cyprus radic Increase contribution to 1st pillar in framework of tax reform
Czech Republic times Reform planned but no details in PEP
Estonia radic None new lower estimates of costs of introduction of 2nd pillar scheme on 1st pillar scheme
Hungary radic Increase contribution rate to 2nd pillarMake 2nd pillar compulsory for new entrants
Latvia radic Increase retirement age further
Lithuania times Introduction of 3 pillar system planned in 2004
Malta times Reform 1st pillar planned
Poland radic Administrative and legal changes to 1st pillar and interaction with 2nd pillar
Romania times Introduction of 3 pillar system is being discussed
Slovak Republic times Parametric reforms of 1st pillarIntroduction of 3 pillar system planned privatisation revenue planned to fund transition cost
Slovenia times None
Turkey times None
53
Part II
Evolving budgetary surveillance
Summary
The year 2002 and the early part of 2003 has been a dif-ficult period as regards the implementation of the EUframework for fiscal surveillance With nominal deficitsbreaching the 3 of GDP reference value Germany andPortugal have been placed in excessive deficit positionsAn early-warning was issued to France in January 2002but subsequent data revealed that a nominal deficit of31 of GDP was recorded in 2002 and the Commis-sion has consequently recommended that France beplaced in an excessive deficit position
A number of lessons can be drawn from these first expe-riences with the enforcement mechanisms of the Treatyand SGP Firstly the credibility in the rules-basedframework was not aided by the Councilrsquos failure toissue an early-warning in February 2002 to Germany andPortugal the recent experience with France furtherunderlined the need for early-warnings to be sent wellbefore nominal deficits are close to 3 of GDP Sec-ondly the repeated upward revisions of deficits under-lined the importance of strengthening the process of col-lection and verification of budgetary statistics Thirdlyand on a positive note surveillance at EU level (with itsbinding deadlines for reporting data and the role of theCommission in providing a neutral assessment of com-pliance with agreed budgetary targets) has prompteddebates at Member State level on the need to face up todifficult budgetary policy challenges In the case of Por-tugal and Germany action at EU level has arguablyfacilitated the introduction of painful reforms necessaryto prevent public finances from entering unsustainablepaths the French authorities however have to datefailed to take measures to address the growing budgetaryimbalances despite these becoming apparent already inmid-2002
In response to these developments and in line with amandate from the Barcelona European Council conclu-sions the Commission adopted a communication onstrengthening coordination of budgetary policies Itidentified a number of shortcomings with the implemen-
tation of the SGP in the first four years of EMU and out-lined a strategy that called for more account to be takenof underlying economic conditions when assessingbudgetary positions an interpretation of compliancewith SGP requirements which would (depending oncountry-specific circumstances) cater for the budgetaryimpact of reforms that enhance growth and employmentincreasing the emphasis placed on the sustainability ofpublic finances and outstanding debt positions andimproving the implementation of the SGP includingstricter and more timely recourse to the existing enforce-ment instruments At the same time the Commissionadopted proposals to improve the governance of budget-ary statistics
The spring European Council of March 2003 endorseda report of the (Ecofin) Council which shared many ofthe Commissionrsquos proposals on strengthening the coor-dination of budgetary policies It confirmed that theachievement of a budget position of lsquoclose to balance orin surplusrsquo is in the economic self-interest of MemberStates both individually and collectively In addition theCouncil agreed that compliance with the lsquoclose to bal-ance or in surplusrsquo requirement should be assessed incyclically-adjusted terms with due account taken of one-off budgetary measures which only have a transitoryimpact on budget positions For euro-area countriesagreement was reached that Member States with deficitsshould achieve an annual improvement in the cyclically-adjusted budget deficit of at least 05 of GDP until thelsquoclose to balance or in surplusrsquo requirement is reached Itunderlined the need for automatic stabilisers to operatesymmetrically over the economic cycle and the particu-lar importance of avoiding a pro-cyclical loosening offiscal policies in good times The Council also confirmedthe importance of running down public debt at a satisfac-tory pace towards the 60 of GDP reference value andthat the existing provisions of the Treaty (that is the debtcriterion of the excessive deficit procedure) can contrib-ute to achieving this goal
57
P u b l i c f i n a n c e s i n E M U 2 0 0 3
A debt-to-GDP ratio below 60 (or on a decreasingpath) is warranted to ensure that public finances are on asustainable footing in the light of the projected budget-ary impact of ageing populations In addition the reduc-tion of government debt will create room to pursue othereconomic and social goals in particular to enhance eco-nomic growth High debt levels also leave the creditstanding of the country vulnerable to unfavourable eco-nomic circumstances The challenge now is to operation-alise the debt criterion of the EDP When assessing debtdevelopments careful attention should be devoted to allthe factors which determine its dynamics so as to evalu-ate to what extent debt developments are due to factorsoutside the immediate control of governments It isindeed essential to avoid a too mechanistic approach toassess compliance with the debt criterion
Budgetary statistics are the foundation of the EU fiscalsurveillance tools and their quality has improved consid-erably over the last decade Government accounts are nowmore reliable complete transparent and detailed and arepublished in a much more timely fashion than when theexcessive deficit procedure was set up However someweaknesses remain in several countries data on govern-
ment deficit and debt ratios are not yet as reliable as theyshould be and are subject to large revisions Furthermorethe government accounts of several Member States are notfully transparent and there have been problems in termsof their timely submission These concerns are clearlyamplified with the perspective of enlargement To addressoutstanding challenges the (Ecofin) Council recentlyagreed to implement a code of best practice From theMember Statesrsquo side this involves increasing the transpar-ency of government accounts in particular for the lowergovernment subsectors the strict respect of deadlines andan overall increase in data quality but also a clarificationof the independence statute of the national statisticaloffices as the main compilers of government data TheCommission (Eurostat) is aiming at reinforcing its abilityto scrutinise the Member Statesrsquo government accounts inmore detail and accelerating the decision-making processfor deciding upon the recording of government transac-tions The new steps to compile quarterly budgetary statis-tics is a major challenge for statisticians but also for econ-omists policy-makers and budgetary policy analysts whowill need to interpret quarterly data with due care sincethese will necessarily be more volatile and perhaps lesstransparent than annual data
58
1 Implementing the Stability and Growth Pact
11 Introduction
The fiscal framework of EMU aims at combiningbudgetary discipline with flexibility through two mainrequirements These are the Treaty requirement to avoidexcessive deficit positions (measured against referencevalues for deficits and debt of 3 and 60 of GDPrespectively) and the requirement of the SGP to achieveand maintain a budgetary position lsquoclose to balance or insurplusrsquo over the cycle Compliance with the lsquoclose tobalance or in surplusrsquo requirement secures fiscal disci-pline and the sustainability of public finances and thuscontributes to maintaining an economic environment inwhich monetary policy can effectively pursue price sta-bility It also provides the necessary room for manoeuvreto allow the automatic stabilisers to play freely Therules-based framework of the Treaty and SGP consists ofboth preventive and dissuasive elements both of whichare backed up with enforcement procedures
The deterioration in the budget positions has required theCommission and Council to apply the various enforcementmechanisms of the Stability and Growth Pact (SGP)against several Member States during 2002 and the earlypart of 2003 Against a background of slow economicgrowth this has led to considerable tension in the CouncilThe discussions on the implementation of the SGP gener-ated negative reactions in the press and markets and in partmotivated the Commission proposals to strengthen thecoordination of budgetary policies in November 2002 (1)
The remainder of this chapter summarises the debate onthe implementation of the SGP since spring 2002 (2)
Section 2 describes the enforcement mechanisms pro-vided for in the Treaty and the SGP regulations (3)Section 3 examines the specific cases of the three Mem-ber States (Portugal Germany and France) where theCouncil has already taken action in the framework of theexcessive deficit procedure
12 The enforcement mechanisms of the SGP
121 The preventive part of the Pact
Under the preventive arrangements of the Pact Mem-ber States submit annual stability or convergence pro-grammes in which they set down their short andmedium-term budgetary strategies to reach and sustainbudget positions that are lsquoclose to balance or in sur-plusrsquo The programmes are subject to peer review andmonitoring by the Commission and Council with aview to identifying any lsquosignificant divergencersquo eitherfrom the medium-term budget target or the adjustmentpath towards it This surveillance not only consists ofverifying whether nominal budgetary targets are met italso involves a close examination of the underlyingbudget position taking account of cyclical economicconditions
If the Council identifies such a significant divergencefrom a budget target it shall address a recommendationto the Member State concerned with a view to give anearly warning in order to prevent the occurrence of anexcessive deficit The Council recommendation isadopted by qualified majority on the basis of a Commis-sion recommendation following the procedure outlinedin Article 99(4) of the Treaty and Articles 6 and 10 of
yen1part COM(2003) 668 final See Chapter II for a discussion of the communicationyen2part Part II2 in last yearrsquos report summarises the debate on the Commissionrsquos
recommendation of February 2002 for lsquoearly-warningsrsquo to be sent toGermany and Portugal
yen3part For a more detailed description see Cabral (2001) Costello (2001) andFischer and Giudice (2001)
59
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Council Regulation (EC) No 146697 (1) A second rec-ommendation to take prompt corrective measures can beaddressed to the Member States concerned if the Counciljudges that the divergence is persisting or worsening
Council Regulation (EC) No 146697 does not definewhat constitutes a lsquosignificant divergencersquo from budget-ary targets or the conditions under which the early-warn-ing mechanism is to be activated To ensure consistencythe Commission has developed and used the followingthree factors in deciding whether to activate the early-warning mechanism
bull the size of the budgetary slippage that is the extentto which budget positions diverge from the targetsset down in stability or convergence programmes
bull the reason for the budgetary slippage that iswhether the divergence of actual balances from thetarget can be explained by cyclical or discretionaryfactors
bull the risk of an excessive deficit position that iswhether there is a risk of breaching the 3 of GDPreference value
These criteria distinguish between slippage from budget-ary targets in nominal and cyclically-adjusted terms andreflect whether or not a country has reached the medium-term target of the SGP In brief more leeway is affordedto countries with sound budget positions An early-warn-ing however can be issued even if the nominal deficit issome way below the 3 of GDP reference value Todate recourse has only been made to the early-warningmechanism when deficits were well above 2 of GDPand experience with Portugal Germany and France hasshown that this is likely to be too late to prevent deficitsfrom going above 3 of GDP
122 The dissuasive elements of the Pact
The dissuasive elements of the SGP are set down in Arti-cle 104 which requires all Member States to avoid exces-sive government deficits (2) Under the excessive deficitprocedure (EDP) the Commission monitors budgetary
developments and examines compliance with budgetarydiscipline on the basis of two criteria that is lsquowhetherthe ratio of the actual or planned government deficit togross domestic product exceeds a reference value [3 of GDP]rsquo and lsquowhether the ratio of government debt togross domestic product exceeds a reference value [60 of GDP] unless the ratio is sufficiently diminishing andapproaching the reference value at a satisfactory pacersquo
The EDP is a complicated procedure involving severalsteps Article 104(3) states lsquoIf a Member State breachesone or both of the these criteria the Commission shallprepare a reportrsquo This report shall lsquotake into accountwhether the government deficit exceeds governmentinvestment expenditure and take into account all otherrelevant factors including the medium-term economicand budgetary position of the Member Statersquo After theCommission adopts such a report the EFC must give itsopinion thereon within two weeks As of this point threepossible courses of action are possible
bull the Commission could decide that there is neither arisk nor existence of an excessive deficit positionand the procedure would then stop
bull the Commission could address an opinion on the riskof an excessive deficit position in accordance withArticle 104(5) The Treaty does not specify the pre-cise conditions as to what constitutes a risk of anexcessive deficit but the most clear-cut scenario is aforecast (either Commission or of the nationalauthorities) projecting a deficit level above 3 ofGDP reference value (3) The Council is not requiredto vote on the Commissionrsquos opinion and the proce-dure comes to a halt at this stage It would only bereactivated if subsequent outcome data confirmsthat the 3 of GDP reference value has indeed beenbreached
bull the Commission could adopt an opinion in accord-ance with Article 104(5) on the existence of anexcessive deficit position The Council is thenrequired to vote by qualified majority on whether anexcessive deficit position exists in accordance withArticle 104(6) To be placed in an excessive deficit
yen1part OJ L 209 281997 In addition to these legal obligations on the early-warning mechanism the Commission Member States and Council gave astrong political commitment to the lsquostrict and timelyrsquo implementation ofthe SGP in the resolution of the Amsterdam European Council on theStability and Growth Pact (OJ C 236 281997)
yen2part Under the provisions of its opt-out protocol the UK is not required to avoidexcessive deficit positions but rather must endeavour to do so
yen3part However a forecast deficit above the 3 of GDP reference is not a prereq-uisite requirement for the activation of the EDP Article 104(3) states thatlsquoThe Commission may also prepare a report if notwithstanding the fulfil-ment of the requirements under the [deficit and debt] it is of the opinionthat there is a risk of an excessive deficit positionrsquo
60
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
position outcome data must show that the referencevalues have indeed been breached The rationalebehind waiting for the outcome data is that beingplaced in an excessive deficit position has poten-tially serious consequences for a Member State forexample the possibility of negative reactions bymarkets resulting in a higher risk premium on debtit could prevent a country from joining the eurozone it could eventually lead to the imposition of
financial sanctions on euro-area countries in breachof its obligations
At the same time as it decides upon the existence of anexcessive deficit position the Council must also adopt arecommendation to the Member State concerned (inaccordance with Article 104(7)) with a view to bringingthe situation to an end within a given period Article 3(4)of Council Regulation (EC) No 146797 specifies that
Box II1 What constitutes an lsquoexceptional circumstancersquo under the excessive deficit procedure
A nominal deficit above 3 of GDP does not imply a country is automatically placed in an excessive deficit positionas the Treaty and SGP regulations provide some room for interpretation to take account of lsquoexceptional circumstancesrsquoArticle 104(2) of the Treaty states that a Member State with a deficit to GDP ratio over 3 is in an excessive deficitposition unless lsquohellip the excess over the reference value is only exceptional and temporary and the ratio remains close tothe reference valuersquo Against the background of the current economic slowdown and the effects of war in the Gulf thequestion has been raised as to whether countries could make recourse to this exceptionality clause to avoid being placedin an excess deficit position
Before addressing this question directly it should be noted that any breach must be at the same time exceptional and tem-porary and close to the reference value that is the conditions are cumulative and thus recourse to this Treaty provision isrestricted to a very limited number of cases Moreover the issue of exceptional circumstances only arises when the Com-mission and Council are deciding upon the existence of an excessive deficit position in accordance with Article 104(6) ofthe Treaty There is no scope for the Council to give an ex ante exemption to any Member State allowing to breach the 3 of GDP reference value for deficits Neither could it be applied retroactively to countries such as Portugal and Germanywhich are already in excessive deficit positions
Article 2(1) of Council Regulation (EC) No 146797 provides some further clarification on what constitutes an exceptionalcircumstance The excess of a government deficit over the reference value shall be considered exceptional when lsquoresultingfrom an unusual event outside the control of the Member State concerned and which has a major impact on the financialposition of the general government or when resulting from a severe economic downturnrsquo
A priori the direct costs of participation in a military conflict could be regarded as an lsquounusual event outside the control ofthe Member State concernedrsquo together with costs of additional security measures However it would need to be backed upwith evidence that these have had a lsquoa major impact on the financial position of the general governmentrsquo and thus are amajor contributory factor to the deficit level rising above 3 of GDP Clearly this argument would not apply to deficitsgoing above 3 of GDP in 2002
A more pertinent issue is whether the economic situation constitutes a lsquosevere economic downturnrsquo Article 2(1) ofCouncil Regulation (EC) No 146797 establishes a general rule whereby a severe downturn is considered exceptional iflsquothere is an annual fall of real GDP of at least 2 rsquo The Member State concerned can demonstrate that even a fall ofannual real GDP of less than 2 is lsquonevertheless exceptional in the light of further supportive evidence in particularon the abruptness of the downturn or on the accumulated loss of output relative to past trendsrsquo In the resolution of theEuropean Council on growth and employment Member States committed not to invoke the exceptional clause if GDPfall is less than ndash 075
Based on the spring 2003 Commission forecast a loss in output of 075 is not projected in any Member State and there-fore there is currently no case for considering recourse to the exceptionality clause However it could become relevant ifgrowth in some countries turns out to be substantially lower than is currently forecasted
61
P u b l i c f i n a n c e s i n E M U 2 0 0 3
this recommendation must contain two deadlinesFirstly a deadline of four months at the most must beestablished for the Member State to take effective actionIn addition a deadline must be established for the correc-tion of the excessive deficit position which lsquohellip shouldbe completed in the year following its identificationunless there are special circumstancesrsquo It is worth high-lighting the fact that the initial requirement on the Mem-ber State concerned is to take corrective action ratherthan in achieving immediate results As such the will-ingness of the Member States to respond to the repri-mand of the Council is of critical importance The failureto take corrective actions would trigger the next stage ofthe EDP and move the Member State closer to the stagewhen it may receive sanctions
13 The use of enforcement mechanisms since spring 2002 (1)
131 Slippage from budget targets in many Member States
Throughout 2002 concern grew about the deteriorationin budget positions in several Member States participat-ing in the euro area Table II1 compares the budget out-comes for 2002 projected by the Commission in autumn
yen1part For documents concerning these procedures see the section on fiscal sur-veillance on the web site of the Directorate-General for Economic andFinancial Affairs httpeuropaeuintcommeconomy_financeaboutactiv-itiessgpprocedures_enhtm
Table II1
Comparison of growth and budgetary developments for 2002 between autumn 2002 Commission forecasts and the 2001 updates of the programmes
GDP growth in 2002 ( pa)
Budget balance in 2002 (excluding UMTS) ( of GDP)
Difference from SPCP budget target ( of GDP) due to
of GDP SPCP COM
forecastSPCP target
EDP notification
COM forecast
Difference COM
ndash SPCP
Impact of cyclical conditions
in 2002
Non-cyclical factorsin 2002
pmcyclical and non-cyclical
factors in 2001
1 2 3 4 5 6 = 5 ndash 3 7 8 = 6 ndash 7 9
BE 13 07 00 00 ndash 01 ndash 01 ndash 01 01 02
DE 08 04 ndash 25 ndash 29 ndash 38 ndash 13 ndash 01 ndash 12 ndash 03
EL 38 35 08 08 ndash 13 ndash 21 00 ndash 21 ndash 18
ES 24 19 00 00 00 00 ndash 02 02 ndash 01
FR 15 10 ndash 18 ndash 26 ndash 27 ndash 09 ndash 01 ndash 08 ndash 01
IE 39 33 07 ndash 01 ndash 12 ndash 19 ndash 02 ndash 17 02
IT 23 04 ndash 05 ndash 11 ndash 24 ndash 19 ndash 07 ndash 12 ndash 11
LU 53 01 28 13 05 ndash 23 ndash 19 ndash 03 20
NL 13 02 04 ndash 05 ndash 08 ndash 12 ndash 06 ndash 05 ndash 06
AT 13 07 00 ndash 13 ndash 18 ndash 18 00 ndash 17 02
PT 18 07 ndash 18 ndash 28 ndash 34 ndash 16 ndash 03 ndash 13 ndash 20
FI 16 14 26 36 36 10 ndash 02 12 01
EUR-12 18 08 ndash 11 ndash 17 ndash 23 ndash 11 ndash 05 ndash 06 ndash 04
DK 17 17 19 21 20 01 00 02 02
SE 24 16 21 18 14 ndash 07 ndash 03 ndash 04 02
UK 23 16 ndash 11 ndash 10 ndash 11 00 ndash 04 04 09
EU-15 18 09 ndash 10 ndash 14 ndash 19 ndash 09 ndash 04 ndash 05 ndash 02
NB SPCP = stabilityconvergence programmes submitted in Autumn 2001 EDP notification = September 2002 COM = autumn 2002 Commission forecasts Impactof cyclical conditions shortfall = ndash bonus =
Source European Commission
62
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
2002 with the targets set down in the 2001 updates ofstability and convergence programmes that is the infor-mation which was available to ministers in late 2002when key decisions on the implementation of the SGPhad to be taken Significant slippage from budget targetswas evident in a large number of countries although theconcern was focused on countries where deficits emergedEventually the Council took action against three coun-tries (Germany France and Portugal) although the defi-cits in Greece Italy and Austria also gave cause forconcern
As shown on column 8 of Table II2 approximately halfof the deterioration in budget positions projected for2002 was due to the automatic stabilisers in response toeconomic cycle Non-cyclical factors such as unfundedtax cuts discretionary expenditure increases and spend-ing overruns also contributed to the slippage This indi-cates a reversal in some Member States of budgetaryconsolidation efforts In several Member States how-ever most of the deviation from the 2002 target resultedfrom the slippage that had already occurred by the end of2001 (see column 9)
132 Portugal
On 5 November 2002 the Council decided that an exces-sive deficit existed in Portugal the first time the EDPwas applied since the launch of the euro in 1999 (1)Budget difficulties in Portugal had been apparent forsome time (2) and in January 2002 the Commissionadopted a recommendation that an early-warning be sentto Portugal for having missed its budget target for 2001by a wide margin At that time the Commission (on thebasis of its autumn 2001 forecast) was projecting a defi-cit of 22 of GDP for 2001 compared with a target of11 of GDP set down its stability programme seeGraph II1 The Ecofin Council at its meeting of 12 Feb-ruary 2002 however decided not to endorse the Com-mission recommendation for an early-warning Thisfollowed commitments given by the Portuguese authori-ties to endeavour to prevent the deficit from going abovethe 3 of GDP reference value in 2002
yen1part Council Decision 2002923EC OJ L 32230yen2part See Part II2 in European Commission (2002a)
Graph II1 Budgetary divergence from target in Portugal
ndash 5
ndash 4
ndash 3
ndash 2
ndash 1
0
1
2000 2001 2002 2003 2004 2005 2006
Per
cent
age
of G
DP
Reference value
2001 programme 2002 programme
Spring 2002 forecastAutumn 2002 forecast
63
P u b l i c f i n a n c e s i n E M U 2 0 0 3
On 25 July 2002 the Commission received officialconfirmation from the Portuguese authorities that thegeneral government deficit in 2001 was to be revisedupwards from 22 of GDP (reported in February 2002EDP notification) to 41 of GDP an upward revisionof 19 percentage points of GDP (1) This revision fol-lowed the submission of a report by a special task forcecalled the Commission for the Analysis of PublicAccounts established by the Portuguese Governmentunder the direct responsibility of the Governor of theBank of Portugal This task force set up following thenon-acceptance by Eurostat of budgetary data notified inMarch 2002 was made up of representatives from theMinistry of Finance Bank of Portugal and the NationalInstitutes of Statistics
The size of this ex-post revision and the delay in it com-ing to light underlined serious deficiencies in the collec-tion and processing of general government statisticaldata in Portugal A breakdown of the revised outcomefor data for 2001 shows that the difference of 19 per-centage points of GDP was due in almost equal parts tothe reclassification of certain items in governmentaccounts (2) to bring them in line with the Eurostat defi-nitions and due to a slippage in budgetary execution
A deficit of 41 of GDP in 2001 was confirmed in thePortugalrsquos submission by 1 September 2002 under thesemi-annual reporting of government deficits and debtlevels and the Commission activated the EDP by prepar-ing on 24 September 2002 a report in accordance withArticle 104(3) of the Treaty In this report the Commis-sion drew attention to the failure on the part of Portugalto achieve budgetary consolidation since the mid-1990sand that the deterioration in the budget balance could notbe explained by the cycle as the cyclically-adjustedbudget deficit had risen from 3 of GDP in 1999 to45 of GDP in 2001 (using the HP filter method) Onthe revenue side the shortfall derives from the lossesimplied by the reform of direct taxes implemented in2001 and lower-than-projected efficiency gains in tax
collection and administration At the same time currentprimary expenditures continued to grow faster than nom-inal GDP with the public sector wage bill and socialtransfers repeatedly surpassing targets set by the govern-ment The Commission also concluded that the breach ofthe 3 of GDP reference value could not be attributedto a severe economic downturn (that is the exceptional-ity clause could not apply) Moreover the increase in thedeficit in 2001 could not be attributed to public invest-ment as this remained constant at some 4 of GDP overthe 1999 to 2001 period
The Economic and Financial Committee confirmedthese findings and on the basis of an opinion and a rec-ommendation proposed by the Commission adopted on16 October 2002 the Council decided upon the existenceof an excessive deficit It also adopted a recommenda-tion with a view to bringing the situation to an end (3) Asrequired two deadlines were set down in this recommen-dation (i) a deadline of 31 December 2002 was set forthe Portuguese authorities to take measures to correct theexcessive deficit position (ii) a deadline for the correc-tion of the excessive deficit position which should becompleted in the year following its identification this isunderstood as being the end of 2003
The response of the Portuguese authorities began beforethe Council had decided upon the existence of an exces-sive deficit position (for more details see Part VI12 onPortugal) The newly elected government enacted arectifying budget which became law in June 2002 It
yen1part The impact of this upward revision for 2001 on the budgetary position for2002 is evident on column 9 of Table II1 above
yen2part Regulation EC25162000 requires that taxes and social contributionsrecorded in the accounts may be derived from two sources amounts evi-denced by assessment and declarations or cash receipts If assessments anddeclarations are used the amounts shall be adjusted by a co-efficientreflecting assessed and declared amounts never collected If cash receiptsare used they must be time-adjusted so that the cash is attributed when theactivity took place to generate a liability The Portuguese authorities optedfor the cash method with slight time adjustments notably as regards the col-lection of VAT taxes Portugal was granted a derogation from this provisionup to 30 June 2002
Table II2
Breakdown of revision of 2001 budget balance of Portugal
Reclassification of some operations as subsidies instead of capital injections 02
Recording of expenditure arrears from commitments made in 2001 03
Application of regulation EC25162000 06
Recording of receipts associated with EC structural funds ndash 01
New information on budgetary execution 09
Total 19
Source Portuguese Commission for the Analysis of public Accounts
yen3part See the Directorate-General for Economic and Financial Affairs web sitefor the relevant documents httpeuropa euintcommeconomy_financeaboutactivitiessgpprocedures_enhtm
64
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
included consolidation measures equivalent to 06 ofGDP notably via an increase in the standard VAT ratefrom 17 to 19 It also included measures such as afreeze on the hiring of civil servants and the end of inter-est rate subsidies on new mortgage loans
In addition a firm commitment was given to reduce thedeficit to 28 of GDP in 2002 that is below the 3 of GDP reference value already in the same year thusahead of the formal deadline required under the EDPregulations Additional measures have been taken in anattempt to meet this goal a task made more difficultby deteriorating growth conditions According to theMarch 2003 semi-annual notification the deficit in 2002fell to 27 of GDP an outcome which relied heavilyon one-off measures especially a tax amnesty
Against a background of slow growth and the termina-tion of one-off measures Portugal will face a considera-ble challenge in keeping the nominal deficit below the3 of GDP reference value The Council will shortlyhave to decide whether in accordance with Article 104(11)to abrogate the decision on the existence of an excessivedeficit
Both negative and positive conclusions can be drawnfrom this first experience with the EDP in Stage III ofEMU It underlined the importance of strengthening theprocess of collection and verification of budgetary statis-tics that underline the fiscal rules of EMU On the posi-tive side however the discrepancies in the statisticalreporting framework were picked up albeit with anunsatisfactory delay and the resulting peer pressure hasfacilitated the introduction of painful but necessaryreforms to prevent public finances continuing on whatwas an unsustainable path
133 Germany
On 21 January 2003 the Council decided that an exces-sive deficit exists in Germany (1) Significant divergenceof the budgetary position from targets had becomeapparent already in late 2001 and in January 2002 theCommission adopted a recommendation for an early-warning to be sent to Germany At that time the Com-mission (on the basis of its autumn 2001 forecast) wasprojecting a deficit of 26 of GDP for 2001 comparedwith a target of 15 of GDP set down its stability pro-gramme see Graph II2 The Council decided the Com-
mission recommendation would not be put to vote and toclose the early-warning procedure This followed com-mitments from the German authorities to endeavour toensure that the 3 of GDP reference value for the gen-eral government deficit would not be breached in 2002and to reach a close to balance position by 2004 in linewith previous commitments
Following general elections on 22 September 2002the re-elected federal government on 24 Septemberbelatedly submitted the autumn notification of budg-etary data showing a deficit of 29 of GDP andconfirming a debt ratio of 606 for 2002 Subse-quently on 16 October 2002 the Minister for Financepublicly stated that the deficit for 2002 was likely toexceed the Treatyrsquos reference value On the basis of itsautumn 2002 forecast projecting a deficit of 38 ofGDP for 2002 the Commission activated the EDP bypreparing a report in accordance with Article 104(3) ofthe Treaty
The report drew attention to the very weak growth per-formance of Germany over the past decade Howeverthe deterioration in the budget balance can only in part beattributed to the effects of the economic cycle as thecyclically-adjusted budget deficit which had fallen con-tinuously since 1995 started to increase as of 2000 andgrew to some 32 in 2002 The origins of this budget-ary slippage can be found in the 1998ndash2000 periodinsufficient efforts were made to strengthen the underly-ing budgetary position when growth conditions werefavourable Indeed the cyclically-adjusted deficit startedto rise again as from 2000 not least due to strongerexpenditure growth at the regional level Based on anassumption of continued strong economic growth and aso-called lsquodividendrsquo for public revenues the governmentopted for the carrying-forward to 2001 of the 2002 stageof the tax reform and for a back-loading of the necessarybudgetary consolidation efforts Thus with the advent ofthe business cycle slowdown there was insufficient lee-way for the operation of automatic stabilisers while atthe same time preventing the deficit from rising abovethe 3 of GDP reference value
Although dramatic for the people involved the floodswhich occurred in August 2002 are not expected to haveconstituted a serious drag on public finances in 2002Commission calculations show that the 2002 overall def-icit-raising effect should not be higher than one tenth ofa percentage point of GDP (that is around EUR 2 bil-lion) given that the bulk of repair works would start onlyyen1part Council Decision 200389EC OJ L 3416
65
P u b l i c f i n a n c e s i n E M U 2 0 0 3
in 2003 this was implicitly recognised by the fact thatthe special fund set up by the federal government offi-cially began its operations on 1 January 2003 As in thecase of Portugal the general government deficit hadbeen clearly higher than public investment althoughhigher public investment induced by the flood damagesand the projected decline in the general government def-icit should narrow the gap in 2003
Outcome data for 2002 confirmed a deficit of 36 ofGDP and the Council on 21 January 2003 decided uponthe existence of an excessive deficit position andadopted a recommendation with a view to bringing thesituation to an end It should also be noted that the debtlevel in 2002 reached 608 of GDP which is in excessof the Treaty reference value and on the basis of currentgrowth forecasts it is expected to increase further in2003 Two deadlines were set in the Council recommen-dation (i) a deadline of 21 May 2003 was set for the Ger-man authorities to take measures to correct the excessivedeficit positions (ii) a deadline for the correction of theexcessive deficit position which should be completed inthe year following its identification this is understood asbeing the end of 2004 Germany however was invitedto bring the deficit below 3 of GDP already in 2003
as planned in the updated stability programme if thegrowth conditions projected in the update (GDP growthof 1 ) would materialise The Council also recom-mended that the German authorities ensure that the risein the debt ratio is brought to a halt in 2003 and reversedthereafter
Based on the latest growth prospects a correction of theexcessive deficit situation in 2003 appears uncertainConcerning 2004 the full implementation of the coali-tion agreement and the achievement of the targets setdown in the updated stability programme (see Part VI3on Germany) would ensure a substantial decline in theactual and cyclically-adjusted deficit provided GDPgrowth turns out as expected Even in the event ofgrowth picking up further into 2004 the budgetary roomfor manoeuvre is set to remain limited in view of the fur-ther steps of income tax cuts envisaged A sustainedimprovement in the budgetary position will thus requiregovernment expenditure to remain under firm control
Important lessons can be drawn from the application ofEDP to Germany the largest economy in the euro areaand a leading proponent of the SGP The credibility inthe rules-based framework was not aided by the Coun-
Graph II2 Budgetary divergence from target in Germany
ndash 5
ndash 4
ndash 3
ndash 2
ndash 1
0
1
2000 2001 2002 2003 2004 2005 2006
Per
cent
age
of G
DP
2001 programme 2002 programme
Spring 2002 forecastAutumn 2002 forecast
66
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
cilrsquos failure to issue an early-warning in February 2002nor the subsequent ratcheting up of projections for thedeficit level throughout 2002 This called into questionthe reliability of budgetary statistics and forecasts under-lying the EU surveillance process and indicated a lack ofcapacity and willingness on the part of Member States todeal with growing budgetary imbalances However italso indicated that a debate on difficult budgetary policychallenges could not be avoided on account of bindingdeadlines in the SGP even for large countries duringelectoral campaigns Arguably the debate on the early-warning ensured that the issue of sound public financesplayed a prominent role in the electoral campaign andhas been facilitating discussions on difficult policychoices and trade-offs
Ultimately however the debate on the SGP has shownthat the rising budget deficit is the symptom but not thecause of Germanyrsquos economic problems The key policychallenge is the growth performance during the last dec-ade with an average annual GDP growth rate of 13 between 1992 and 2002 Unless the causes of slowgrowth are tackled at source deficits in Germany willremain high posing continuous stress on the SGP
134 France
On 21 January 2003 the Council adopted a recommen-dation giving an early-warning to France in order to pre-vent the occurrence of an excessive deficit This is thefirst time that an early-warning has been issued by theCouncil and occurred because there was a significantdivergence from the budget target set down in its stabil-ity programme (see Graph II3)
In its 2001 update of the stability programme France pro-jected a general government deficit at 14 and 13 ofGDP in 2002 and 2003 respectively under the assump-tion of an increase in real GDP by 25 in both years (1)The Commission in its autumn 2002 forecast projected adeficit of 27 and 29 of GDP for 2002 and 2003 respec-tively An early warning was merited on account of
bull the size of the slippage from target some 13 per-centage points of GDP for 2002
bull the source of the budgetary slippage According toCommission services calculations at most one half
of the total slippage can be attributed to cyclical fac-tors The cyclically-adjusted government deficitstable at around 2 of GDP between 1999 and2001 has increased in 2002 to slightly above 2
bull the risk of a breach of the 3 of GDP referencevalue given the perilously close margins that wereprojected at that time
In its March 2003 reporting of data the French authori-ties indicated that the deficit in 2002 was 31 (2) clearlyin excess of the reference value considering also its fore-cast for 2003 of a deficit still above 3 of GDP Itshould be noted that this further deterioration in thebudget balance compared with the autumn 2002 forecastcannot be attributed to effects of deteriorating growthconditions and instead is the result of a disappointingbudgetary execution The Commission therefore acti-vated the EDP and on 7 May 2003 recommended to theCouncil to decide on the existence of an excessive deficitin France and to address a recommendation to France toput an end to the present excessive deficit situation asrapidly as possible and by 2004 at the latest
The experience with the early-warning mechanism toFrance has been far from smooth for three reasonsFirstly the fact that the deficit level in 2002 turned out tobe above 3 of GDP and that the EDP was activatedsome eight weeks after the Council had issued an early-warning forcefully illustrates that the mechanism is notproviding an advance signal to Member States on theneed for corrective action Early-warnings to be effectivewould need to be sent well before deficit levels are veryclose to 3 of GDP a point made in the Commissioncommunication of November 2002 on strengthening thecoordination of budgetary policies (see Chapter II2)
Secondly the debate on the early-warning was charac-terised by repeated revisions in budget projections for2002 coupled with strong but unfulfilled commitmentsto avoid excessive deficits position In February 2002the French authorities adjusted their objective for the2002 general government deficit upwards from 14 to18 of GDP reflecting the impact of deterioratingcyclical factors This revision took place very shortlybefore the discussion of the French update in the EcofinCouncil which created inconveniences with respect to
yen1part France subsequently revised its deficit target for 2002 to 18 of GDP asreported in Table II1
yen2part The government deficit for 2002 has been revised from 30 of GDP (asnotified by the French authorities) to 31 of GDP as a consequence of theinclusion in the deficit of the capital injection by the French State to Reacuteseauferreacute de France (RFF) See Press Release STAT0330 of 17 March 2003
67
P u b l i c f i n a n c e s i n E M U 2 0 0 3
the preparatory work made by the Commission and theEFC In May 2002 after the presidential elections thenew government launched an audit on public financeswhich estimated the general government deficit in 2002within a range 23ndash26 of GDP the revision broughtabout by the audit was due to the consideration of thecyclical effect on tax revenues and unemploymentexpenditures following the deceleration in economicactivity estimated at 03ndash04 of GDP and also due toan overrun in expenditures particularly in the State andthe health sectors estimated at 06ndash07 of GDP InJuly 2002 the French authorities presented a correctivebudget bill for 2002 aimed at adjusting the governmentbudgetary forecasts in line with the results of the audit onpublic finances and at implementing a cut in the incometax by 5 In this corrective budget bill the Frenchauthorities decided to target a general government deficitof 26 of GDP in 2002 which is the highest value ofthe range of the auditorsrsquo projections thus not correctingthe observed slippage in the budgetary situation Asnoted above the autumn 2002 forecast and subsequentreporting of data under the EDP has led to a further sub-stantial upward revision
Thirdly and unlike the Portuguese and German authori-ties which did not contest the application of the SGP the
French authorities have to date failed to take any measuresto address the growing budgetary imbalances despitethese already becoming apparent in mid-2002 Moreoverthey have failed to engage in a constructive dialogue at EUlevel on the pace of budgetary consolidation towards thelsquoclose to balance or in surplusrsquo requirement (see the nextchapter for a discussion on these issues) In particularFrance was the only euro area country which did notaccept to pursue a continuous adjustment of the underly-ing balance by at least 05 of GDP per year startingalready in 2003 as agreed by all other ministers at theEurogroup meeting of 7 October 2002 (see Section II21)
The French authorities continue to fail to start taking cor-rective measures in 2003 This was demonstrated in thebudget targets of their 2002 stability programme whichprovided for an improvement of only 02 percentagepoints of GDP in its cyclically-adjusted budget balanceThe Council in its opinion urged lsquohellip the French author-ities to seek an improvement in the underlying budgetposition in each yearhelliprsquo The start of the process ofbudgetary consolidation cannot be postponed indefi-nitely as the Council recommendation (in accordancewith Article 104(7)) on measures to correct an excessivedeficit position includes a deadline of no more than fourmonths for taking corrective actions
Graph II3 Budgetary divergence from target in France
2000 2001 2002 2003 2004 2005 2006
Per
cent
age
of G
DP
ndash 5
ndash 4
ndash 3
ndash 2
ndash 1
0
1
2001 programme 2002 programme
Spring 2002 forecastAutumn 2002 forecast
Reference
68
2 Strengthening the coordination of budgetary policies
21 Background to the debate a mandate from the Barcelona European Council
The Treaty supplemented by secondary legislation hasbestowed on the Union a unique institutional architec-ture for the conduct of economic and monetary policiesThe uniqueness of the framework resides in the fact thata single monetary policy is entrusted to an independentEuropean Central Bank whilst the responsibility foreconomic policies (budgetary and structural policies)remains decentralised in the hands of national (or sub-national) authorities but subject to some common rulesIn particular Member States remain fully responsible fortheir tax and expenditure policies but within a frame-work at EU level to monitor and where necessaryensure that countries pursue the common goal of soundand sustainable public finances
The appropriate degree and instruments of economic pol-icy coordination cannot remain static or be subject to anoverly rigid literal interpretation of rules and proceduresTo remain effective it must evolve over time so as to takeaccount of changing economic circumstances andor theconvergencedivergence of political preferences It isespecially important in the aftermath of a major regimechange such as the launch of EMU that a learning-by-doing approach be followed so that shortcomings arecorrected and the lessons of experience are drawn
On the basis of the experience accumulated in the earlyyears of EMU the Commission in February 2001adopted a communication on strengthening economic pol-icy coordination within the euro area (1) This led to sev-eral positive developments including better and moretimely statistics covering the euro area a quarterly report
on the euro area prepared by the Commission the estab-lishment of a Eurogroup working party attached to theEconomic and Financial Committee (EFC) to help pre-pare debates and regular communiqueacutes (so-called terms ofreference) from the Eurogroup on important policy issuesIn addition an important agreement had been reached onthe streamlining of policy measures in the BEPGs
Subsequently the Barcelona European Council ofMarch 2002 concluded that the euro area needed to makefurther progress with policy coordination and invitedthe Commission to present proposals to reinforce eco-nomic policy coordination in time for the 2003 springEuropean Council
The initial response of the Commission to this mandatewas to suggest that all euro-area countries adhere to com-mon standards for the conduct of economic policies in theeuro area The objective of common standards would beto clarify the respective role of economic policies in threedomains (1) preserving macroeconomic stability (2)enhancing the economic growth potential of the euro areaand (3) responding to economic shocks that affect indi-vidual Member States or the euro area as a whole
Concerning their format and status the intention was forcommon standards to complement the existing Treatyprovisions and Stability and Growth Pact regulationswith non-binding guidelines on the policy stanceexpected of authorities in various circumstances that isa so-called lsquoreaction functionrsquo The aim was to facilitatediscussions amongst ministers on policy challenges asthey emerged and thereby contribute to a more consist-ent policy stance over time and across Member StatesMoreover the Commission argued that setting downbroad ex ante guidelines on the conduct of economic andbudget policies would help demonstrate that the EU andthe euro area have a well-defined economic strategy withmedium-term orientationyen1part COM(2001) 82 final of 7 February 2001
69
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Based on analytical work of the Commission servicesthe Eurogroup in July and September 2002 discussedpossible elements to be included in common standardson the conduct of economic policies in the euro areaHowever these discussions became overshadowed bythe deterioration in budget positions in several Member
States described in Chapter II1 and the challenges in theimplementation of the SGP This forced a major re-con-sideration on the part of the Commission on how torespond to the mandate of the Barcelona EuropeanCouncil The intended approach of adopting commonstandards on the conduct of economic and budgetary pol-
Box II2 The Convention on the Future of Europe the debate on the coordination of budgetary policies
In the two communications to the Convention adopted in the course of 2002 (1) the Commission has put forward specificsuggestions in the area of economic and notably budgetary policy coordination
First the Commission proposes to reinforce the Community dimension of the policy-coordination process To this effectthe instruments of economic policy coordination should be drafted on the basis of proposals from the Commission ratherthan mere recommendations from which the Council may depart by qualified majority As far as Article 99 of the Treatyis concerned this change would notably have an impact on the adoption procedure of the BEPGs on the adoption by Coun-cil of its opinions on the stability and convergence programmes and on the Council recommendations to a Member Statewhich is pursuing economic policies which are not consistent with the BEPG or with the Stability and Growth Pact
Moreover when the economic policies pursued by a specific Member State are not consistent with the broad guidelines orrisk jeopardising the proper functioning of EMU the Commission should be able to issue warnings directly to the MemberState concerned For example the Commission could decide to issue a lsquodirectrsquo early warning to any Member State with abudgetary position which is significantly diverging from the budgetary target set out in its stability or convergence pro-gramme At the same time it would preserve the possibility under Article 99(4) to invite the Council to make the necessaryrecommendations to the Member State concerned As already explained above the Councilrsquos decision would be based ona Commission proposal as opposed to a recommendation These different measures will reinforce the Community dimen-sion of the economic policy coordination framework by allowing the Commission to play its role as a representative of thecommon interest and as the lsquorefereersquo who ensures that the rules of the game are being observed
The Commission furthermore proposes to facilitate decision-making within the euro area While the informal Eurogroupwould continue to exist a lsquoeuro arearsquo Ecofin Council would also be established in order to allow the Member States belong-ing to the euro area to take certain decisions which are mainly or exclusively relevant for participating countries This insti-tutional change would have important consequences for a number of decisions taken in the framework of the excessivedeficit procedure and the SGP (for example early warnings adopted by the Council) particularly when participating Mem-ber States are concerned
The Convention has closely examined the functioning of the EMU framework A Working Group on Economic Govern-ance chaired by Mr K Haumlnsch was established in order to examine a list of different issues falling under three headingsmonetary policy economic policy and institutional issues As far as the Stability and Growth Pact is concerned a majorityof the Group agreed that the Commission should be allowed to issue first warnings on excessive deficits directly to theMember State concerned Some members also agreed with the need to transform Commission recommendations into pro-posals and supported the exclusion from the vote of the Member State concerned for example in the case of an early warn-ings issued by the Council or in relation to decisions on the existence of an excessive deficit The Working Groupconsidered that the Stability and Growth Pact is a political instrument to implement the Treaty provisions and that it shouldtherefore not be integrated into the Constitution The results of the Working Group were discussed by the plenary on7 November 2002 which largely confirmed the main views expressed by the Group The Praesidium has indicated that afirst draft of Part II of the Constitution which will describe the different policy areas and the Conventionrsquos proposals inrelation to each of them will be made available in the course of May 2003
(1) lsquoA project for the European Unionrsquo (COM(2002) 247 of 22 May 2002) and lsquoFor the European Union peace freedom solidarityrsquo (COM(2002) 728 of4 December 2002)
70
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
icies in the euro area was dropped in favour of muchfocused efforts to improve the functioning of the SGP (1)
On 24 September 2002 Commissioner Solbes with theagreement of President Prodi issued a communicationsuggesting a strategy for dealing with pressing budgetarychallenges in the euro area (2) It underlined the impor-tance of the SGP but recognised the need to avoidsetting budget targets that would require very largeimprovements in underlying budget positions in econo-mies suffering from cyclical weakness To this end theysuggested that the medium-term objective of the SGPshould incorporate explicit references to cyclical consid-erations Also countries which have not yet reached the(cyclically-adjusted) lsquoclose to balance or in surplusrsquoobjective should be required to undertake every year aminimum adjustment of 05 of GDP of their cycli-cally-adjusted deficit
The Eurogroup meeting of 7 October 2002 producedlsquoTerms of references on the budgetary developments inthe euro arearsquo very close to the approach of the Commis-sion and which marked an important policy shift asregards the implementation of the Pact In particularlsquoministers re-affirmed their commitment to the Treatyobligation to avoid excessive deficits and to the Stabilityand Growth Pact objective to achieve and maintainbudgetary positions close to balance or in surplus overthe economic cycle Ministers and the ECB concurredtherefore with the Commission that those countrieswhich have not yet reached that objective need to pursuecontinuous adjustment of the underlying balance by atleast 05 of GDP per year All ministers but one[France] accept this to start no later than in next yearrsquosbudgetrsquo
The efforts described in this chapter on measures tostrengthen the coordination of budgetary policies shouldnot be confused with the broader debate underway in theConvention on the Future of Europe The need for abroader and deeper debate on the future of the Unionbecame apparent at the European Council in Nice(December 2000) see Declaration 23 to the Treaty ofNice One year later the European Council meeting inLaeken decided to convene a Convention to examine thefundamental questions raised by the future developmentof the Union The different questions put forward in the
Laeken Declaration mainly relate to the definition of thepowers of the Union the simplification of the Unionrsquosinstruments (legislative instruments implementationmeasures etc) the enhancement of democracy transpar-ency and effectiveness (for example appointment proce-dures for Commissioners and for the Commission Presi-dent EP powers and elections role of the Council roleof the national parliaments etc) and the preparation of aEuropean Constitution The Convention is chaired byMr Giscard drsquoEstaing and is composed of 105 memberswhich represent the different Heads of State or Govern-ment the national parliaments the European Parliamentand the Commission Its work is prepared by the Praesid-ium which is composed of 12 members The Conven-tion started its work in February 2002 and will presentthe results of its work in mid-2003 An intergovernmen-tal conference will be convened either in 2003 or in2004 in order to formally amend the Treaty and proposeit for ratification to the different Member States Box II2provides details on the proposals of the Commissionrelated to the coordination of budgetary policies and thesubsequent debate within the Convention
22 Commission proposals to strengthen the coordination of budgetary policies
221 A diagnosis of the shortcomings of the SGP in the first four years of EMU
The Commission adopted on 27 November 2002 acommunication on strengthening the coordination ofbudgetary policies (3) While arguing that the coordina-tion of budgetary policies is essential for the smoothfunctioning of EMU and that the SGP goal of budgetpositions of lsquoclose to balance or in surplusrsquo remains aneconomically valid objective (4) it provided a candiddiagnosis of significant shortcomings in its implementa-tion as follows
bull political ownership of the SGP by Member Stateshas diminished with a divergence between budget-ary commitments and concrete actions to achievestated targets and unwillingness to acknowledge theimplication of EMU on the conduct of fiscal policyat national level More generally Member Statesfailed to play their role in exerting peer pressure oncountries that miss budgetary targets by a wide mar-gin via the enforcement mechanisms of the SGP
yen1part For a review of problems and challenges concerning the SGP see Giudiceand Montanino (2002)
yen2part SEC(2003) 10096 of 25 September 2002
yen3part COM(2002) 668 finalyen4part For an assessment of Maastrichtrsquos fiscal rules see Buti and Giudice (2002)
71
P u b l i c f i n a n c e s i n E M U 2 0 0 3
bull it has been difficult to establish clear and verifiablebudget objectives which take account of underlyingeconomic conditions While the targets for budgetbalances are set down in stability and convergenceprogrammes in nominal terms the effect of the eco-nomic cycle on the budget position has to be takeninto account when assessing compliance with budg-etary commitments and in particular the adjustmentpath to lsquoclose to balance or in surplusrsquo This proveddifficult in the absence of an agreed method to cal-culate cyclically-adjusted budget balances and alsobecause the nominal deficit targets in the pro-grammes of Member States were sometimes basedon optimistic growth assumptions and with budget-ary adjustment efforts back-loaded towards the endof the time horizon of programmes Measuring com-pliance with budgetary commitments set down inprogrammes has therefore not been straightforwardand this in turn weakened the enforcement mecha-nisms of the SGP
bull the framework for the collection and assessment ofbudgetary statistics has experienced a number ofdifficulties Of greatest concern are the reportinganomalies detected in some Member States whichin the case of Portugal led to a very large upwardrevision of deficit levels Concern was expressedabout the fact that ex post revisions of budgetarydata are getting larger and the discrepancy betweendeficits recorded on accrual basis and debt issuancein cash terms in some Member States Finally thedecision making processes of Eurostat on the classi-fication of certain budgetary operations could bespeeded up
bull some Member States did not run sound budgetarypolicies in good times A failure to pursue budgetaryconsolidation in 1999 and 2000 when growth condi-tions were favourable led to a deterioration in under-lying budget positions and inadequate room for theautomatic stabilisers to operate in the subsequenteconomic slowdown This failure to allow the auto-matic stabilisers to operate symmetrically over theeconomic cycle illustrates inadequate surveillanceand enforcement mechanisms to deal with unwar-ranted pro-cyclical loosening of the fiscal stance
bull the enforcement procedures of the SGP have beenfound wanting at critical junctures In particularthe early-warning mechanism was not effective indealing with significant slippage from budget tar-
gets set down by Member States in their stability andconvergence programmes
bull the SGP has struggled to develop into an effectivecoordination framework for dealing with country-specific circumstances in a consistent manner assur-ing the long-term sustainability of public financeswhile supporting structural reforms that are designedto enhance employment and growth potential
bull it has been difficult to communicate effectivelywith the press markets and the public on thebenefits of achieving and sustaining sound publicfinance positions and also how the SGP worksThis is partly due to the fact that it takes time foreconomic agents to adjust to the new policy frame-work in place since the launch of the euro and alsobecause the institutional procedures of the SGP arecomplex In addition effective communication hasbeen hampered by conflicting statements on theappropriate conduct of budgetary policies
The communication then set out a number of proposalsto tackle these shortcomings It should be noted that theyimplied no change whatsoever to the existing Treaty pro-visions or SGP regulations that is the existing frame-work would be unchanged and no additional procedureswere envisaged On the one hand they consisted of pro-posals to clarify the interpretation of key SGP provisionsso as to strengthen the economic rationale underpinningthe policy decisions On the other hand there were pro-posals to strengthen the implementation of SGP includ-ing the enforcement procedures The main elements ofthe Commission proposals are described below
222 Avoiding pro-cyclical policies and accounting for transitory elements in the assessment
The Commission proposed that in establishing budget-ary objectives at EU level and in carrying out the surveil-lance of Member States budgetary positions dueaccount should be taken of the economic cycle In partic-ular the Commission suggested that the lsquoclose to bal-ance or in surplusrsquo requirement of the SGP would bedefined in underlying terms throughout the economiccycle To this end it is necessary to isolate the impact oftransitory factors on the budget position and in particu-lar the effects of the economic cycle
The underlying budget balance is the actual balance netof transitory elements The main transitory elementtaken into account is the cyclical component However
72
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
other transitory elements beyond the cyclical componentalso have impact on the budget positions (both positivelyand negatively) and thus need to be considered whenassessing the underlying position so as to avoid wrongpolicy conclusions this issue is examined in detail inBox II3 In other words the economic cycle is one butnot the only transitory element that has an importantbudgetary impact Consequently the cyclically-adjustedbudget balance (CAB) is not the same concept as theunderlying budgetary position
To illustrate how the Commissionrsquos proposal wouldwork in practice (and in particular the relevance of thecyclically-adjusted budget balances) Graph II4 illus-trates the budgetary position expected of MemberStates in order to be in compliance with the lsquoclose tobalance or in surplusrsquo requirement of the SGP over theeconomic cycle (1) It refers to a country that has com-pleted the transition to the medium-term goal of thePact and assumes that there are no other transitory
effects on the budget balance other than the effect of thecycle mdash that is the CAB corresponds to the underlyingbudget balance
The underlying budget balance is represented in Graph II4by the bold line which remains unchanged over theeconomic cycle However the nominal budget balance(blue line) fluctuates according to the output gap (darkline) The degree of the fluctuation depends on thecyclical sensitivity of the budget on average an outputlevel that goes 1 below the potential implies anincrease in the nominal deficit of 05 of GDP How-ever automatic stabilisation should show its effects tooduring upturns as automatic stabilisers should operatesymmetrically over the economic cycle this impliesrunning nominal budget surpluses when growth condi-tions are favourable A degree of caution must be usedwhen interpreting changes in cyclically-adjustedbudget balances especially on an annual basis (2)
yen1part Buti and Giudice (2002) illustrate the benefits of focusing on cyclically-adjusted balances for output stabilisation
yen2part See Part II3 of European Commission (2002a)
Graph II4 Compliance with the lsquoclose to balance or in surplusrsquo requirement for countries that completed the transition process
CAB (underlying
budget balance)
of GDP
surp
lus
defi
cit
Nominal balance
Output gap
time
3 refence value
73
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Box II3 Transitory elements affecting the budgetary position
(Continued on the next page)
The surveillance of budgetary positions aims at the maintenance of sound public finances and at ensuring their long-termsustainability In this respect what is important to understand is what the underlying budgetary positions are beyond theimpact of the economic cycle and other transitory effects In the context of the EU rule-based fiscal framework the sur-veillance carried out by the Commission and the Council should take into account the role of measures with only transitoryeffect on the budget As concluded by the Brussels European Council (21 March 2003) lsquoin making an assessment [of theimprovement of the cyclically-adjusted budgetary position] one-off measures will be considered on their own merits on acase-by-case basisrsquo
What could qualify as a one-off measure According to Milesi-Ferretti (2001) lsquoa measure implying an improvement in thefiscal balances is considered to be creative accounting if it does not imply an improvement in the intertemporal budgetarypositionrsquo lsquoCreative accountingrsquo is used in the economic literature as meaning measures with temporary effect or one-offmeasures It is difficult to identify clearly what is transitory or permanent as this depends on what is the reference pointand the degree of country- and situation-specificity is large In the context of EU surveillance the Commission and theCouncil have inevitably a margin of discretion to decide what measures to take into account in order to make the bestpossible assessment However it is important that there is consistency to the degree possible across countries in the dis-tinction between purely transitory elements and other more permanent trends Some examples of transitory elements thatcould be explicitly taken into account are as follows
On the expenditure side large individual sales of real assets such as real estate and the UMTS receipts provide good exam-ples On the revenue side a possible candidate is tax amnesties Here of course what is lsquonormalrsquo in the country concernedis an important reference point since some measures can be exceptional in one country while taking place regularly inanother Other elements may be linked to lsquounusualrsquo events Here size is clearly important as each year there are lsquounusualeventsrsquo Possibly the short-term emergency costs from flooding could be an example Large revenues or expenditures dueto specific court rulings could be another
Along the same line the Congressional Budget Office of the United States produces an estimation for the so-called stand-ardised budget that nets the actual budgetary position from the cyclical component and other temporary factors (see lsquoACBO report the standardised and cyclically-adjusted budgetsrsquo March 2003) It includes in these temporary measures thefollowing unusually large discrepancies between tax payments and liabilities swings in collection of capital gains taxeschanges in the inflation component of the governmentrsquos net interest payments temporary legislative changes in the timingof revenues and outlays asset sales and receipts from auctions of licences to use portions of the electromagnetic spectrum
However the availability of fiscal data on measures with a transitory effect is limited given the difficulties of measurementand the degree of arbitrary Some countries as done by Danish and Swedish authorities in their updated convergence pro-gramme (2002) use of a refined cyclically-adjusted budget balance More specifically by correcting the budget balancefor the deviation of several special factors (that are by definition country-specific) to their calculated trend Large clearlyidentifiable transitory items can be taken into account when assessing underlying budget developments However furtherwork in this area is necessary to upgrade the quality of the analysis
The question of measures that have only a transitory effect (one-off measures lsquocreative accountingrsquo) on the budget positionis also relevant in terms of compliance with the fiscal rules The economic literature proves that the imposition of numericalbudget rules by an outside agent encourages the use of lsquocreative accountingrsquo (see for example Easterly (1999) Eichen-green and Wyplosz (1998) Kopits and Craig (1998)) Policy makers can be induced to explore ways to fulfil budgetarytargets through creative accounting even when the rule results from an agreed commitment and not from an externalconstraint The simple reason to recourse to creative accounting is to avoid the implicit (reputational) or explicit (pecuni-ary) sanctions that occur when the rule is breached In the context of the EU rule-based fiscal framework creative account-ing may contribute to limit reputational sanctions that appear with the lsquoearly-warningrsquo andor with the start of the excessivetends to disappear in the long run due to its temporary nature it is less likely that it can be helpful in avoiding eventual
74
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
223 A minimum annual rate of adjustment for countries still in deficit
The Commission communication built upon the agree-ment of the Eurogroup that countries with underlyingdeficits would be required to achieve an annual improve-ment in the underlying budget position of at least 05 of GDP each year until the lsquoclose-to-balance or surplusrsquorequirement of the SGP has been reached This proposalmakes clear that Member States with underlying deficitsmust make continuous progress towards the medium-term goal of the Pact and thus tackles the problemwhereby targets are being rolled over indefinitely in suc-cessive updates of stability or convergence programmesMoreover it recognises that account must be taken ofeconomic conditions when setting the pace of budgetaryconsolidation
An example of what this proposal implies in practice isillustrated in Grapg II5 The starting position shows thatthe Member State has not completed the transition to thelsquoclose to balance or in surplus requirementrsquo of the SGPNote that there is an assumption of no other transitoryeffects on the budget balance other than the effect of thecycle that is the cyclically-adjusted budget correspondsto the underlying budget balance at all times
The country is required to achieve an annual improve-ment in its underlying budget position of at least 05 of GDP until the medium-term target of the Pact hasbeen reached this minimum rate of underlying budget-ary consolidation should be achieved irrespective ofgrowth conditions (see adjustment path illustrated by thebold line) However this does not imply that the nominal
budget balance must improve every year by an equiva-lent amount There may be some scope to allow the auto-matic stabilisers to operate as illustrated by the deterio-ration in the nominal budget balance during thedownturn when growth falls below its potential rate(between t0 and t1) however a safety margin must beprovided at all times so as to ensure that the nominalbudget deficit does not risk breaching the 3 of GDPreference value
The communication also states that the lsquohellip rate ofimprovement in the underlying budget position shouldbe higher in countries with high deficits or debt Also amore ambitious annual improvement in underlyingbudget positions should be envisaged if growth condi-tions are favourablersquo The latter requirement is illus-trated by a kink in the line representing the cyclically-adjusted budget balance when the output gap starts toimprove As shown between t1 and t2 the output gapstarts to increase and it closes in t2 The requested rate ofadjustment is higher than in the previous period and thenominal budget balance improves at a faster rate than thecyclically-adjusted budget position reflecting the sym-metric operation of the automatic stabilisers As illus-trated in Graph II5 reaching a position of balance innominal terms would not necessarily represent compli-ance with the lsquoclose to balance or in surplusrsquo require-ment The consolidation continues between t2 and t3when the nominal budget becomes positive and theclose to balance position in underlying terms is reachedFrom t3 onwards the transitional period is finished andthe nominal and underlying budget balance are expectedto behave as in Graph II4
Box II3 (continued)
pecuniary sanctions implied by the EDP If nominal budget unbalances is only temporary (due for example to aneconomic shock) the recourse to one-off measures avoids overemphasising the imbalance rightly correcting this temporarysituation as a temporary budgetary measure
But using creative accounting also has costs First fiscal adjustment can be illusory because it temporarily lowers thebudget deficit or the public debt but it does not improve the public sectorrsquos net worth This can imply future measures tocompensate the insufficient structural adjustment that becomes necessary once transitory measures end their effect on thebudgetary position Second the use of creative accounting entails a lack of transparency that could lead to a loss of confi-dence by public opinion in respect of government actions Loss of confidence could also affect financial markets and there-fore the country concerned could face higher risk premium Third these transitory measures can cause distortions in themarkets For example a huge sale of real estate concentrated in a short period of time to reducing the deficit level can havea destabilising impact on prices in the housing market
75
P u b l i c f i n a n c e s i n E M U 2 0 0 3
224 The goals of the Lisbon strategy ensuring that public finances contribute to growth and employment
Perhaps the most innovative elements of the communi-cation concern the proposal to introduce a more flexibleapplication of the lsquoclose to balance or in surplusrsquo require-ment in light of the achievement of the goals of the Lis-bon strategy In particular it is argued that there is a needto lsquohellip cater for the intertemporal budgetary impact oflarge structural reforms (such as productive investmentor tax reforms) that raise employment or growth poten-tial in line with the Lisbon strategy andor which in thelong term improve the underlying public finance posi-tionsrsquo The Commission did not consider it appropriateto develop a list or catalogue of reforms which justify ormerit an exemption This should be judged on a case-by-case basis but it referred to major structural reformsidentified in the BEPGs or as part of the Lisbon strategythat have a clearly identifiable negative impact on thebudget in the short run (for example a reform of the taxsystem pension reform substantial increase in net pub-lic investment) but a positive return in the medium tolong term on growth and the budgetary position
In making this proposal the Commission was aware thatthis initiative could easily be interpreted as a weakeningof the commitment to sound public finances or the corebudgetary goals of the SGP To avoid the impression thatprovisions of this nature would weaken the Pact numer-ous safeguards were outlined in the communication Adistinction was drawn between deviations from thelsquoclose to balance or in surplusrsquo requirements of a lsquotem-poraryrsquo and lsquomore permanent naturersquo
Regarding the former the communication stated that lsquoasmall temporary deterioration in the underlying budgetposition could be envisaged only if the Member Stateconcerned has already made substantial progresstowards the lsquoclose to balance or in surplusrsquo requirementand if general government debt is below the 60 ofGDP reference valuersquo The Commission did not specifya numerical rule as to what would constitute lsquosubstantialprogressrsquo the key issue is to ensure that an adequatesafety margin exists to limit the risk of the nominal def-icit breaching the 3 of GDP reference value and thiswould imply that the underlying budget deficit should bewell below 1 of GDP
Graph II5 The budgetary adjustment path of Member States still in transition to the lsquoclose to balance or in surplusrsquo objective
defi
cit
surp
lus
CAB (underlying BB)
Nominal balance
of GDP
3 reference value
time
Output gap
t 1 t 2 t 3
t 0
76
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
The communication added additional safeguards as fol-lows lsquoIn assessing the programme the Commissionmust ascertain that there is a clear and realistic deadlinefor returning to a position of ldquoclose to balance or in sur-plusrdquo within the time horizon of the stability or conver-gence programme Budgetary projections must be basedon a sound and prudent macroeconomic scenario to beverified against those of the Commission with dueaccount taken of the need to avoid inappropriate pro-cyclical policies An adequate safety margin must beprovided at all times to prevent nominal deficits frombreaching the 3 of GDP reference value Finally theMember State concerned should pre-announce correc-tive measures that would be introduced in the event of afailure to stick to the adjustment path for returning to abudget position of lsquoclose to balance or in surplusrsquo
An example of what this implies in practice is illustratedin Graph II6 The starting position shows a MemberState with an identical nominal (continuous line) andcyclically-adjusted budget deficit (bold line) again it isassumed that there are no other transitory effects on thebudget balance other than the effect of the cycle that is
the cyclically-adjusted budget corresponds to the under-lying budget balance at all times
From that starting position in t1 the Member State imple-ments a major structural reform that initially has a nega-tive impact on the cyclically-adjusted budget balancethis is evident from the downward slope in the CAB lineThere may be some scope to allow the automatic stabi-lisers to operate in the event of a slowdown in growthan even larger increase occurs in the nominal deficit(continuous line) However an adequate safety marginmust be provided at all times so as to ensure that thenominal budget deficit does not risk breaching the 3 of GDP reference value The nominal and the CAB areequal when the output gap is zero (t2) and the MemberState concerned must return to a position of lsquoclose to bal-ance or in surplusrsquo within the time horizon of the pro-gramme (say in t3)
The communication also sought to reflect differencesbetween the sustainability of public finances acrossMember States It therefore proposed that a lsquosmall devi-ation from the ldquoclose to balance or in surplusrdquo require-
Graph II6 Illustration of a small temporary deviation to cater for the inter-temporal budgetary effect of a large structural reform
defi
cit
surp
lus
CAB (underlying BB)
Nominal balance
3 reference value
time
Start of the reform
Output gap
of GDP
t 1t 2 t 3
77
P u b l i c f i n a n c e s i n E M U 2 0 0 3
ment of a longer-term nature could be envisaged forMember States where debt levels are well below the60 of GDP reference value and when public financesare on a sustainable footing This will require a carefulassessment to be made of outstanding public debt con-tingent liabilities (such as implicit pension obligations)and other costs associated with ageing populations Anadequate safety margin must be provided at all times toprevent nominal deficits from breaching the 3 of GDPreference valuersquo
225 Ensuring the sustainability of public finances
The communication also proposed that the sustainabilityof public finances should become a core policy objectiveat EU level and this requires that greater weight isattached to government debt ratios in the budgetarysurveillance process Countries with high debt levelswould be required to set ambitious long-term debt-reduction strategies in their stability and convergenceprogrammes Also the Commission suggested that thehigh-debt countries should be required to achieve a sat-isfactory pace of debt reduction towards the 60 ofGDP reference value and that a failure to do so shouldresult in the activation of the debt criterion of the exces-sive deficit procedure Overall these proposals wereconsidered necessary as the sustainability of publicfinances cannot be assured simply by looking at a three-or four-year time horizon of programmes Chapter II3considers how in practice the debt criterion of theexcessive deficit procedure could be made operational
226 Concrete measures for the enforcement of the Pact
In addition to suggestions on how to interpret certainprovisions of the SGP the communication set downdetailed proposals to improve its practical implementa-tion of how Member States needed to reaffirm theirpolitical commitment to the Pact
Firstly to ensure that Member States assume politicalownership of the SGP the communication called for thespring 2003 European Council to adopt a resolution onstrengthening the coordination of budgetary policiesThe reason for seeking support at the highest politicallevel is that achieving and sustaining the goal of budgetpositions of lsquoclose to balance or in surplusrsquo is extremelychallenging and requires full commitment of all govern-ment departments and all levels of government from thefederal authorities to local councils Substantive conclu-sions of the European Council were deemed helpful for
finance ministers in their difficult task of negotiatingwith spending ministries and representatives of sub-cen-tral governments
Secondly the communication recognised the need toimprove the quality of budgetary statistics and to thisend proposed that all parties mdash Member States and theCommission itself mdash commit themselves to a code ofbest practice on the compilation and reporting of budget-ary statistics (see Part II4 of this volume)
Finally the communication underlined the fact that fis-cal rules need to be backed up with effective and credibleenforcement procedures To this end the Commissionproposed to clarify the criteria to be used when decidingwhether to activate the early-warning mechanism TheCommission also proposed that the interpretation of thedebt criterion of the excessive deficit procedure shouldbe clarified in particular what would constitute a lsquosatis-factory pacersquo of debt reduction towards the 60 of GDPreference value
23 The agreement of the European Council on strengthening the coordination of budgetary policies
The Ecofin Council on 7 March 2003 (1) adopted a reporton strengthening the coordination of budgetary policeswhich was fully endorsed by the European Council of21 and 22 March 2003 The Council agreed that therewas no need to change the current fiscal rules of the EUand that improvements could be made to ensure an effec-tive application of the Stability and Growth Pact
In its report the Ecofin Council endorsed most of theproposals of the Commission It considered that compli-ance with the close to balance or in surplus requirementof the Stability and Growth Pact should be assessed incyclically-adjusted terms and that countries with deficitsmust improve their cyclically-adjusted budget positionand in the case of euro-area countries by a minimumannual reduction of 05 of GDP
The Council also called for automatic stabilisers to oper-ate symmetrically over the cycle and to this end Mem-ber States should avoid pro-cyclical policies especiallywhen growth conditions are favourable
yen1part Ecofin Council report on lsquoStrengthening the coordination of budgetary pol-iciesrsquo 7 March 2003 687703 (Press 61)
78
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
The suggestion of the Commission to allow for devia-tions from the lsquoclose to balance or in surplus requirementrsquoof the SGP was subject to intense debate Concerns wereraised about the practical feasibility of making such aproposal operational while at the same time safeguardingthe commitment to sound public finances In the end theEcofin Council agreed lsquohellip to pay particular attention tocountry-specific circumstances in particular to (i) thelong-term sustainability of public finances (ii) sufficientsafety margins at all times including an allowance forautomatic stabilisers to operate fully without breachingthe 3 of GDP reference value and (iii) the coherencebetween the evolution and quality of the public finances
in the stability and convergence programmes and theclose to balance or in surplus requirementrsquo
Finally the Ecofin Council agreed to pay greater atten-tion to the longer-term sustainability and the quality ofpublic finances with a view to increasing the growthpotential of the EU economies in conformity with theLisbon agenda It recognised that the pace of decline inpublic debt plays an important role in budgetary sur-veillance especially in highly indebted countries Inconformity with the Treaty provisions the excessivedeficit procedure should contribute to ensuring a satis-factory pace of debt reduction
79
3 Public debt and the excessive deficit procedure
31 Introduction
As part of the recent debate on strengthening the coordi-nation of budgetary policies a consensus was reached onthe need to pay increased attention to debt developmentsand the sustainability of public finances One step to thisend is to enhance the assessment of the sustainability ofpublic finances on the basis of stability and convergenceprogrammes (see Part I3 of this report)
The European Council of March 2003 also concludedthat lsquoThe pace of decline in public debt plays an impor-tant role in budgetary surveillance especially in highlyindebted countries In conformity with the Treaty provi-sions the excessive deficit procedure should contributeto ensuring a satisfactory pace of debt reductionrsquo Bothcriteria defined in the Maastricht Treaty (the deficit crite-rion of the 3 reference value and the debt criterion) arerelevant to ensure sound public finances A nominal def-icit-to-GDP ratio below 3 allows automatic stabilisersto smooth (at least partially) the cycle without compro-mising long-term budgetary positions It also helps mon-etary policy to keep inflation under control and to sustainthe economy during slowdowns A debt-to-GDP ratiobelow 60 (or on a decreasing path) is warranted toensure that public finances are on a sustainable footing inthe light of the projected budgetary impact of ageing pop-ulations In addition the reduction of government debtwill create room to pursue other economic and socialgoals in particular to enhance economic growth Highdebt levels also leave the credit standing of the countryvulnerable to unfavourable economic circumstances (1)
So far neither the excessive deficit procedure nor therisk of excessive deficit have been launched for breach-
ing the debt criterion alone The challenge is now toensure that the commitment of reducing debt levelsbelow 60 of GDP is implemented
32 Compliance with the Treaty requirements
Member States have a Treaty obligation to avoid exces-sive deficit positions To this end Article 104(2) of theTreaty states that lsquoThe Commission shall monitor thedevelopment of the budgetary situation and of the stockof debt in Member States with a view to identifying grosserrors In particular it shall examine compliance withbudgetary discipline on the basis of the following twocriteria
(a) whether the ratio of the actual or planned governmentdeficit [hellip]
(b) whether the ratio of government debt to gross domes-tic product exceeds a reference value [60 of GDP]unless the ratio is sufficiently diminishing and approach-ing the reference value at a satisfactory pacersquo
Article 104(3) states that lsquoIf a Member State does not fulfilthe requirements under one or both of the these criteriathe Commission shall prepare a reportrsquo This report is thefirst step in the process that eventually could lead to aCouncil decision on the existence of an excessive deficitposition
To make the debt criterion of the EDP operationalrequires clarifying the conditions under which a debtratio above 60 of GDP lsquohellipis sufficiently diminishingand approaching the reference value at a satisfactorypacersquo yen1part See Bank of America Corporation Economic Research 7 February 2003
80
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
A key question to consider is whether a Member Statecould be in an excessive deficit position for not respect-ing the debt criterion even if the nominal deficit levelremains below 3 of GDP A priori the answer is yessince the Treaty gives the same relevance to both criteria
The focus on government debt in the EUrsquos budgetary sur-veillance process is not new In its decision on MemberStates to adopt the euro (Council Decision of 3 May1998) the Council stated that several countries with agovernment debt-to-GDP ratio still above 60 respectedthe convergence criteria on both the deficit and the debtsince the latter was diminishing at a satisfactory pace
Furthermore the lsquoDeclaration of 1 May 1998 by theEcofin Council accompanying the Councilrsquos recommen-dation on Member States adopting the EMUrsquo stated thatlsquoThe higher the debt-to-GDP ratios of participatingMember States the greater must be their efforts toreduce them rapidly To this end in addition to maintain-ing appropriate levels of primary surpluses in compli-ance with the commitments and the objectives of theStability and Growth Pact other measures to reducegross debt should be put in placersquo As a result high debtcountries remained committed to reduce their govern-ment debt-to-GDP ratios towards the reference valueFor instance Ireland committed to reduce its govern-ment debt-to-GDP ratio to 70 by 1999 (60 deemedachievable early in the 21st century) Italy stated that thegovernment debt-to-GDP ratio would fall below 100 in 2003 and thanks to a constant primary surplus itwould continue to fall in the following years Similarcommitments were taken by the Belgian authorities
33 Debt dynamics in EU countries (1)
Table II3 shows the average annual percentage changeof public debt-to-GDP ratios over the past 10 years intwo sub-periods 1992ndash97 (the so-called period of lsquofiscalconsolidationrsquo) and the years of the Stability and GrowthPact (1998ndash2002) Over the whole period the rate ofvariation has been negative (on average) in only onethird of EU members and among those countries with agovernment debt-to-GDP ratio still above 60 onlyBelgium showed a declining path (ndash 22 on averageeach year) Thanks to the reduced deficit levels and the
further implementation of the Stability and Growth Pactrequirements the debt-to-GDP ratio has had a moreaccentuated declining path during recent years How-ever in those countries that did not comply with the SGPrequirement andor with very high levels of debt thespeed of debt reduction has been slower than in othercountries Six out of 15 countries have had a rate ofreduction of less than 3 each year between 1998 and2002 and among these countries there are the three big-gest EU economies mdash Italy France and Germany mdash thatrepresent more than 60 of total EU public debt in 2002
The pace of debt reduction depends upon both factorsthat can be shaped by government policies (primary bal-ance privatisation) and factors which lie outside theirimmediate control (interest rate changes growth andinflation rates exchange rate movements) Factors out-side the immediate control of the government whosecombined effect is commonly known as the lsquosnowballeffectrsquo are as follows
The interest rates on government debt They includeexpected inflation and a (diversificationdefault) riskpremium A lower interest rate decreases the amount ofinterest payments making the reduction of the debt ratioeasier Ceteris paribus the market interest rate is likelyto decrease the more credible the economic policy is
Real GDP growth A faster rate of real GDP growthincreases the denominator of the debt-to-GDP ratio Italso affects revenues and therefore improves the budget-ary position
Inflation rate As the denominator of the debt-to-GDPratio is expressed in nominal terms a faster inflation ratereduces the value of the stock of debt The inflation ratehas also an impact on government revenues and expen-ditures and in general tends to improve the nominalbudgetary position Contrary to the past given the clearmandate of the ECB to maintain price stability and itsindependence this factor can no longer be expected tocontribute substantially to debt reduction However dif-ferences in inflation across Member States could ceterisparibus be reflected in the pace of reduction of the stockof the debt
The factors more under governmental control are asfollows
The primary balance This factor is determined bygovernment policies (apart from cyclical components)
yen1part The definition of government debt is the one contained in the Protocolannexed to the Maastricht Treaty lsquodebt means gross debt at nominal valueoutstanding at the end of the year and consolidated between and within thesectors of general governmentrsquo
81
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Other things being equal a primary surplus improves thegovernment debt-to-GDP ratio (or limit the deterioration)
Stock-flow adjustments These result primarily fromfinancial operations for example debt issuance policy tomanage public debt privatisation receipts impact ofexchange rate changes on foreign denominated debt (1)In general these should tend to cancel out over timeHowever large and persistent stock-flows (especially ifthey always have a negative impact on debt develop-ments) should give cause for concern as they may be theresult of the inappropriate recording of budgetary opera-tions and can lead to large ex post upward revisions ofdeficit levels Also the debt ratio may fluctuate consid-erably because of changes in the governmentrsquos portfolioof financial assets For instance if the social securitysector decides to shift its reserves from governmentpaper into private securities the government debt asdefined in the Protocol annexed to the Maastricht Treatyincreases
Table II4 shows how the above-mentioned factorsaffected debt development in high debt countries sincethe mid-1990s The impact of interest rates and nominalGDP growth is represented by the so-called lsquosnowballrsquoeffect measured as the difference between the twoSince 1998 beside lsquopurersquo public finance variables thebehaviour of the stock of debt has been negativelyaffected by stock-flow adjustments in all three high-debtcountries
34 What could constitute a satisfactory pace of debt reduction
Table II5 shows the expected debt dynamic for a countrywith a starting government debt-to-GDP ratio of 100 under different nominal GDP growth conditions (therange is between 3 and 5 ) and when the lsquoclose to bal-ance or in surplusrsquo requirement is always respected (2) Asshown the government debt-to-GDP ratio is expected toreach the reference value in maximum 17 years unlessgrowth conditions remain very adverse over the wholeperiod (that is below 3 in nominal terms)
Respect of the lsquoclose to balance or in surplusrsquo require-ment will clearly ensure a fast pace of debt reductionHowever for the purpose of operationalising the debtcriterion of the EDP a minimal requirement of whatconstitutes a lsquosatisfactory pacersquo of debt reduction couldbe defined to be used as a reference in the assessment ofdebt developments This operational indicator should berelated to the level of the debt ratio with a faster pace ofreduction required in countries where debt levels arewell above the 60 of GDP reference value It shouldalso be consistent with the overall policy frameworkThe indicator should be strict enough to allow debtreduction below the reference value in a reasonablenumber of years but not be over-demanding
A number of different methods can be used to measure asatisfactory pace of debt reduction Depending on howparameters of the rule are fixed the speed of debt reduc-tion towards the reference value can be very different Afirst set of operational indicators can refer to the budgetbalance position either in terms of required primary sur-plus or required budget balance An example of how thisindicator could work in practice for a stylised countrywith initial government debt at 100 of GDP is shown
Table II3
Average annual percentage change of public debt-to-GDP ratios
1992ndash2002 1992ndash97 1998ndash2002
Countries with debt ratio above 60 in 2002
BE ndash 22 ndash 11 ndash 32
DE 37 74 ndash 01
EL 21 47 ndash 06
IT 00 23 ndash 23
AT 18 26 09
Countries with debt ratio below 60 in 2002
DK ndash 35 ndash 12 ndash 58
ES 18 77 ndash 41
FR 42 85 ndash 01
IE ndash 103 ndash 82 ndash 124
LU 03 58 ndash 52
NL ndash 38 ndash 21 ndash 55
PT 08 18 ndash 02
FI 12 69 ndash 45
SE ndash 40 ndash 15 ndash 56
UK 00 55 ndash 54
EUR-12 12 41 ndash 17
EU-15 ndash 08 21 ndash 25
Source Commission services
yen1part Exchange rate developments may affect the flow of interest payments andhence the implicit interest rate paid on debt when part of the latter isdenominated in a foreign currency yen2part Nominal implicit interest rates are set up at 6
82
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
in Table II6 for different combinations of nominalgrowth and primary surpluses (1)
The exact minimum primary surplus required woulddepend on the pace of debt reduction which would beconsidered necessary and feasible A main conclusion tobe drawn from the table is the critical influence of thenominal GDP growth rate If nominal growth rates are
low then the pace of debt reduction slackens considera-bly for a given primary surplus For example if the nom-inal growth rate would be 3 instead of 4 it wouldtake 26 as opposed to 17 years for debt to fall below thereference value with a primary surplus of 4 of GDPWhile primary surplus is the policy variable that drivesdebt reduction over which the government has mostcontrol the budgetary effort becomes higher the lowerthe debt-to-GDP ratio is In fact the implied rate ofreduction of the debt-to-GDP ratio increases the lower isthe debt-to-GDP ratio
Table II4
Development in debt levels in several EU high-debt countries since the mid-1990s
Belgium 1995 1996 1997 1998 1999 2000 2001 2002
Debt level ( GDP) 1340 1302 1248 1196 1149 1096 1085 1053
Change in debt level ndash 19 ndash 38 ndash 54 ndash 52 ndash 47 ndash 53 ndash 11 ndash 32
Due to Primary deficit (1) ndash 49 ndash 50 ndash 60 ndash 68 ndash 65 ndash 69 ndash 70 ndash 61
Snowball effect 45 57 19 31 17 13 36 29
Stock-flow adjustment ndash 15 ndash 45 ndash 12 ndash 15 01 03 23 00
pm
Implicit interest rate on debt 71 68 64 63 61 62 62 58
Real GDP growth (pa ) 24 12 36 20 32 37 08 07
GDP deflator (pa ) 13 12 13 17 14 13 20 23
Greece 1995 1996 1997 1998 1999 2000 2001 2002
Debt level ( GDP) 1087 1113 1082 1058 1051 1062 1070 1049
Change in debt level 08 26 ndash 31 ndash 24 ndash 07 11 08 ndash 21
Due to Primary deficit (1) ndash 10 ndash 31 ndash 42 ndash 53 ndash 54 ndash 51 ndash 49 ndash 43
Snowball effect ndash 05 07 ndash 25 ndash 09 06 ndash 05 ndash 12 ndash 22
Stock-flow adjustment 23 50 36 39 41 68 69 44
pm
Implicit interest rate on debt 116 107 82 78 73 71 63 56
Real GDP growth (pa ) 21 24 36 34 36 42 41 40
GDP deflator (pa ) 98 74 68 52 30 34 34 37
Italy 1995 1996 1997 1998 1999 2000 2001 2002
Debt level ( GDP) 1232 1221 1202 1163 1149 1106 1095 1067
Change in debt level ndash 06 ndash 11 ndash 19 ndash 39 ndash 14 ndash 43 ndash 11 ndash 28
Due to Primary deficit (1) ndash 39 ndash 44 ndash 67 ndash 52 ndash 50 ndash 58 ndash 38 ndash 34
Snowball effect 23 40 41 31 31 07 15 23
Stock-flow adjustment 11 ndash 07 06 ndash 18 06 08 11 ndash 18
pm
Implicit interest rate on debt 101 99 80 70 60 59 60 55
Real GDP growth (pa ) 29 11 20 18 17 31 18 04
GDP deflator (pa ) 50 53 24 27 16 21 27 27
(1) The primary surplus include UMTS proceeds which amounted to 12 of GDP in Italy in 2000 02 of GDP in Belgium and 05 of GDPin Greece in 2000
Source Commission services
yen1part Nominal implicit interest rates are set up at 6
83
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Alternatively the lsquosatisfactoryrsquo pace of debt reductioncan be defined looking directly at the rate of reduction ofthe debt ratio For instance this can fall by a fixed per-centage of the debt ratio each year (Table II7a) or as afixed percentage of the distance between the actual debt-to-GDP ratio and the 60 reference value (Table II7b)Note this approach is defined in terms of a specified per-centage of reduction in the debt-to-GDP ratio each yearand not in terms of a fixed reduction of debt as a share ofGDP
Table II7a shows the required primary surplus at thebeginning and at the end of the adjustment period (forexample first three years and last three years beforereaching 60 ) according to different annual rate ofreduction and nominal growth assumptions when a fixed
rate of reduction is set up (1) For instance a reduction inthe debt ratio of 3 each year would bring the debt levelfrom 100 to 60 of GDP within 17 years If nominalGDP growth is assumed constant at 5 this wouldrequire an average primary surplus of 38 of GDP inthe first three years of the consolidation process As debtlevels fall over time a lower primary surplus would beneeded to achieve a constant reduction in the debt ratioof 3 each year in the last three years of the consolida-tion process an average primary surplus of 33 ofGDP would be sufficient
Table II7b shows the debt development when the rate ofreduction of the debt ratio is based on the distance of the
debt ratio from 60 As the debt ratio declines towards60 the further reduction that is required becomessmaller and approaches zero the closer it gets to 60 (2)
Throughout a fixed rate of debt reduction the MemberState reaches the reference value of 60 in a reasonablenumber of years without the rule being over-demandingat the beginning of the adjustment path However itcould be too stringent for countries with a governmentdebt-to-GDP ratio below 65 but still above 60 Conversely a percentage of debt reduction that decreaseas the debt approaches 60 of GDP makes a clear dis-tinction between very high and high debt countries
Table II5
Debt dynamic according to different budget balances and nominal GDP growth rates (initial government debt-to-GDP ratio 100 )
Nominal GDP growth rate
5 4 3
Budget balance Years to reach 60
Years to reach 60
Years to reach 60
0 10 13 17
ndash 05 12 15 22
ndash 10 14 19 30
ndash 15 17 25 53
Source Commission services
Table II6
The implied rate of debt reduction by a constant primary surplus(starting point 100 of government debt-to-GDP ratio)
Nominal GDP growth
3 4 5
Average primarysurplus
Annual rate of reduction
Years to reach 60
Annual rate of reduction
Years to reach 60
Annual rate of reduction
Years to reach 60
3 GDP 02 gt30 11 29 21 19
4 GDP 12 26 22 17 32 13
5 GDP 23 16 33 13 44 10
NB The table shows the average annual reduction in debt levels as pp of GDP in the first five years of a budgetary consolidation programme for different combinationsof a constant primary surplus and interest-growth rate differential It also shows the number of years required to bring debt levels from 100 to 60 of GDP
yen1part Nominal implicit interest rates are set up at 6
yen2part The formula to be applied is the following bt = bt ndash 1 ndash x (bt ndash 1 ndash 60) where btis government debt-to-GDP ratio at time t bt ndash 1 is government debt-to-GDPratio at time t ndash 1 x is the fixed percentage of reduction ie 0 lt x le 1
84
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
However the debt ratio would approach 60 at adecreasing speed without ever reaching it (1) In addi-tion to achieve the reference value within a reasonableperiod of time the required adjustment at the beginningof the period could become unsustainably high in termsof the required primary surplus
Graph II7 compares the implied debt dynamic of thethree described approaches with the expected path if acountry complies with the Stability and Growth Pactrequirement of a budget balance lsquoclose to balance or insurplusrsquo The three approaches are set in order to deliverthe same rate debt reduction in percentage points of GDP
during the first year for a stylised country with initialgovernment to GDP ratio of 100 (2) Once the param-eters are fixed the approach is then maintained over theyears
To summarise the pace of debt reduction depends uponboth factors that can be shaped by government policies(primary balance privatisation) and factors which lieoutside their immediate control (interest rate changesgrowth and inflation rates exchange rate movements)When assessing debt developments careful attentionshould be devoted to each of these factors so as to eval-uate to what extent unfavourable debt developments are
Table II7
The implied primary surplus by defining a rate of reduction of the debt ratio
(a) Implied primary surplus by a constant rate of debt reduction(starting point 100 of government debt-to-GDP ratio)
Nominal GDP growth
3 4 5
Annual rate of reduction
Years to reach 60
First 3 years Last 3 years First 3 years Last 3 years First 3 years Last 3 years
3 17 56 48 47 41 38 33
4 13 65 43 56 37 46 31
5 10 72 50 63 44 55 38
NB The table shows the implied primary surplus in the first and last three years of a budgetary consolidation process necessary to achieve a constant annual reductionin debt levels as a of GDP Implicit interest rates constant at 6
(b) Implied primary surplus by a fixed percentage of debt reduction based on distance from 60 reference value(starting point 100 of government debt-to-GDP ratio)
Nominal GDP growth
3 4 5
Fixed percentage of debt reduction
Years to reach 60 (1)
First 3 years Last 3 years First 3 years Last 3 years First 3 years Last 3 years
75 39 56 21 47 15 38 09
10 29 64 22 55 16 46 10
15 19 77 24 69 18 60 12
(1) Since the rule is asymptotic to 60 it never reaches the reference value Therefore the table shows the number of years to approach the reference value (ie toreach 62 of government debt-to-GDP ratio) The table shows the implied primary surplus in the first and last three years of a budgetary consolidation process bya fixed percentage of debt reduction Implicit interest rate is constant at 6
yen1part To avoid the asymptotic problem at 60 it could be proposed to move thetarget to a value lower than 60 say 40 when the country has alreadyreduced its debt-to-GDP ratio to a value well below 100 but still far from60 eg 80
yen2part The implied reduction in the first year is 3 percentage points of GDP iegovernment debt-to-GDP ratio falls from 100 to 97 Primary surplus at4 constant rate of reduction at 3 fixed percentage of reduction at75 Implicit interest rate at 6 Nominal GDP growth at 5
85
P u b l i c f i n a n c e s i n E M U 2 0 0 3
due to factors outside the immediate control of govern-ments Also the year-on-year development of the debt-to-GDP ratio can be influenced by the volatility of somevariables and for this reason the dynamic of the debt
should also take into account government debt develop-ments in previous years It is indeed essential to avoid atoo mechanistic approach to assess compliance with thedebt criterion
Graph II7 Pace of adjustment for the debt ratio
40
50
60
70
80
90
100
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 24years
Deb
t le
vel i
n
of
GD
P
reference value
(CTB)
(2)
(1)
(3)
(1) Fixed primary surplus CTB
(2) Fixed rate of reduction(3) Reduction based on distance from 60
86
4 The governance of budgetary statistics in EMU
41 Introduction
The quality of economic statistics is crucial to ensure anadequate understanding of the economic situation and tocontribute to effective policy making Low quality statis-tics may lead to poor economic analysis mistaken con-clusions about the behaviour of economic agents andeven to inappropriate policy decisions The quality of thebudgetary statistics of Member States is particularlyimportant given that these statistics are the foundation ofthe budgetary surveillance framework
The quality of budgetary statistics is used here as a verygeneric term It includes the appropriateness of theaccounting rules compliance of data with the account-ing rules the reliability credibility completeness time-liness across-time and across-country comparabilityconsistency and transparency of data
The quality of the statistics depends primarily on theirgovernance Governance includes the accounting princi-ples rules procedures and behaviour of institutions onthe compilation and publication of figures on the distri-bution of responsibilities among different institutionsand on the mechanisms to resolve technical difficultiesor even to mediate conflicts
Throughout the last decade since the Maastricht Treatycame into force there has been considerable progressin the budgetary statistics in the EU Governmentaccounts are now more reliable complete transparentand detailed and are published in a much more timelyfashion than when the excessive deficit procedure (EDP)was set up Moreover the governance of statistics has alsoimproved with the respective roles of the Member Statesand of the Commission being progressively clarified
However some weaknesses can be still identified in thecompilation and publication of government accounts bythe Member States In several countries the governmentdeficit and debt ratios are not yet as reliable as theyshould be and are subject to large revisions Further-more the government accounts of several countries arenot fully transparent and there have been some problemsin terms of timeliness and of inappropriate politicalpressure on the national statistical institutes All theseconcerns are clearly amplified with the perspective ofenlargement since most acceding countries have statis-tical systems that are less developed than in currentMember States and some of them have serious budgetaryimbalances (see Part I2)
The next section of this chapter describes the main ele-ments of the governance of budgetary statistics in EMUSection 3 assesses the quality of the main budgetaryindicators the government deficit and debt in terms ofreliability transparency and timeliness Section 4 is onrecent progress to improve the quality of budgetary sta-tistics the first steps towards the compilation of govern-ment accounts with a quarterly frequency and the code ofbest practice recently endorsed by the Ecofin CouncilSection 5 concludes and describes the challenges for thefuture
42 The governance of budgetary statistics in the EU
421 Main elements
The main elements of the governance of budgetary sta-tistics in EMU were established already in 1992 in theprotocol on the excessive deficit procedure annexed tothe Maastricht Treaty The authors of the MaastrichtTreaty were already mindful that an effective implemen-tation of the budgetary surveillance in the EU depended
87
P u b l i c f i n a n c e s i n E M U 2 0 0 3
on the quality of statistics and that the latter should besupported by good governance
ESA as the accounting reference The protocol statesthat the data for the budgetary surveillance should becompiled according to the objective and well-definedaccounting rules of the European system of integratedeconomic accounts (ESA) A main advantage of an eco-nomic accounting system like ESA (1) is that transac-tions and policy measures are recorded in a meaningfuland suitable way for economic analysis forecasting andpolicy making In addition the ESA accounts try toreflect the economic reality irrespective of the legal andadministrative arrangements and therefore lead to com-parable results even if the Member States have quitedifferent institutional settings
There is a wide agreement that ESA is an appropriatetool to assess economic developments The usefulness ofESA for budgetary surveillance is also widely acceptedalthough the accounting system was not developed spe-cifically for budgetary surveillance purposes
The Commission authority The protocol also helps toensure sound governance by stating that the statisticaldata to be used for the implementation of the excessivedeficit procedure are to be provided by the CommissionThis implies that the Commission is the statisticalauthority in this domain This principle is understanda-ble and logic Since the budgetary data will be used bythe European institutions to check whether MemberStates adhere to fiscal discipline it is sensible that thesedata are officially provided by an impartial institutionand not by the Member States themselves The provisionof the budgetary data by the Commission ensures thatsuch statistics are properly checked their quality is per-manently monitored and that they are comparable amongMember States
However this does not mean that the budgetary data arecompiled directly from basic sources by the Commissionservices That would clearly be an inefficient option Thecompilation of government accounts involves collecting
data on millions of transactions by thousands of govern-ment units by the central government including theState and several other public units such as publicautonomous funds and services public hospitals univer-sities and other education units by the regional and localgovernments and by the social security Clearly theCommission does not have the means to compile thegovernment accounts of each Member State Accordingto the principle of subsidiarity this task belongs to eachMember State However the statistics compiled byMember States are then reported to the Commissionwhich validates them after a thorough examination
422 Other aspects of the governance of budgetary statistics
Besides the basic elements of governance of budgetarystatistics contained in the protocol there are some otherimportant aspects that were developed in secondarylegislation or that evolved over the last decade Theseinclude the rules on the reporting of EDP-related data tothe Commission the rules on the transmission to theCommission of more complete budgetary statistics andthe role of Eurostat as the Commission service that exer-cises the Commissionrsquos role as statistical authority
EDP reporting Given the Commission task of officiallyproviding the statistical data for the excessive deficitprocedure there was a need to organise the transmissionor reporting of data by Member States This was done ina Council regulation of 1993 (2) Member States reporttheir deficit and debt figures twice a year for 1 Marchand 1 September
This twice a year reporting is adequate The first report-ing allows the Commission to get a first estimate of theoutcome of the budgetary implementation in the previ-ous year so that the formal implementation of the exces-sive deficit procedure can be put in motion shortly afterthe end of the year The second reporting confirms orrevises the estimate with data that are much more stableand reliable
The reporting tables contain important information tocheck whether the deficit and debt data comply with theaccounting rules Namely Member States should reportinformation that explains the adjustments made to thecash-basis deficit to transform it into the ESA definition
yen1part The version of ESA that was in force in 1992 was ESA79 This system wasreplaced in 2000 with the European system of national and regionalaccounts or ESA95 The adoption of ESA95 as the accounting frameworkfor the budgetary surveillance in Europe in 2000 was a major step in thecompilation of national accounts and in particular of governmentaccounts ESA95 is a modern system of national accounts which has astrong legal basis in the form of a legally binding regulation while the pre-vious accounting system was simply an administrative document
yen2part Council Regulation (EC) No 360593 This regulation was slightly revisedin 2000 and 2002
88
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
of government deficit Member States should also trans-mit information on the contribution of the governmentdeficit and the other relevant factors to the variation inthe government debt level that is the so-called stock-flow adjustment In practice this consists of transmittinginformation on the government financial transactions(such as privatisation loans etc) that affect the govern-ment debt but are eligible to be excluded from the gov-ernment deficit
Transmission of other budgetary statistics The EDPreporting covers the data that are strictly indispensablefor the surveillance of the budgetary situation in the EUand that are specifically mentioned in the Treaty asconvergence criteria That is the government deficit anddebt However there are plenty of other elements thatare relevant when analysing budgetary policy and thedevelopments in the fiscal position of Member States
In fact Member States transmit many other budgetarystatistics to the Commission These other statistics aretransmitted according to the transmission programme ofnational accounts and include
bull the complete government account which is transmit-ted thrice a year at the end of March end of Augustand end of December This is detailed informationon tax revenue and on all other government receiptson salaries paid on purchases of goods and serviceson investment and on all other government expend-iture categories The data transmission of Decemberis even broken down by sub-sector (central Stateand local government and social security)
bull the government financial account which is transmit-ted at the end of September This is information ontransactions on financial assets such as the sales andpurchase of enterprisesrsquo shares loans granted by thegovernment and on all government liabilities
bull the government financial balance sheets which aretransmitted at the end of September This is informa-tion on the stocks of assets and liabilities owned orowed by government
bull details on taxes and social contributions collected bythe general government and each of its sub-sectorswhich is transmitted in December for the previousyear
bull the breakdown of government expenditure by func-tion which is also transmitted in December for theprevious year
All this information is disseminated by Eurostat
Although these other statistics are compiled under alegal context other than EDP they may be used for eco-nomic analysis in the context of the budgetary surveil-lance and for cross checking the deficit and debt figuresreported for 1 March and September
The role of Eurostat In the internal organisation of theCommission the statistical authority role is exercised byEurostat The aim of this delegation of powers was thatthe accounting and statistical issues are treated inde-pendently by an impartial and technically competentbody that guarantees the quality of data and lends credi-bility to the whole process
The tasks of Eurostat in this field have developed alongtwo lines The first has been checking and validating thedata reported by Member States This work has beendone on the basis of the reporting tables on other infor-mation transmitted by Member States when reportingtheir EDP data and on regular technical meetings withthe national authorities in charge of compiling the deficitand debt figures
In practice Eurostat has become progressively moreactive and stricter when checking the data transmittedby the Member States Several times notably duringthe last two years the control of data by Eurostat led theMember States to amend the reported figures MoreoverEurostat has itself amended the reported government sta-tistics figures and publicly expressed reservations aboutthe quality of data reported by a few Member States thuscontributing to the transparency and credibility of budg-etary surveillance
The second part of the Eurostat task has been in clarify-ing the application of the accounting rules wheneverthere were doubts over how specific measures and trans-actions should be recorded In fact despite the high levelof detail of the ESA accounting rules there are govern-ment transactions for which the accounting treatment isnot straightforward This owes to the specificity of eachcountry as the same accounting system is applied bycountries with fairly different institutional arrangementsto the diversity and multitude of operations performedby government each year and also to the increasing
89
P u b l i c f i n a n c e s i n E M U 2 0 0 3
sophistication of government transactions To guaranteethat data reported by each country are comparable thereis a need to interpret the accounting rules in these cir-cumstances
The accounting issues with relevance for the governmentdeficit and debt that had so far to be considered by Euro-stat can be classified in four broad groups
bull issues about the delimitation of general governmentthat is whether a specific publicly owned or control-led unit is government or whether it should beclassified outside general government as a publicenterprise in the corporate sector
bull issues about the nature of specific transactions thatis to know whether a specific government transac-tion has any direct impact on the government deficitIn more technical terms this means that one shoulddecide whether a transactions has a financial or anon-financial nature In the former case the transac-tion has no direct impact on the deficit while in thelatter case the deficit improves or deteriorates
bull issues about the time of recording of transactionsThis issue is particularly relevant since in ESAtransactions are recorded on an accruals basis Theaccruals basis imply that transactions are recordedwhen economic value is created transformed orextinguished or when claims and obligations ariseare transformed or extinguished which does notnecessarily coincide with a cash disbursement
bull issues about the calculation of the government debtEurostat had to decide about the inclusion in thegovernment debt of unusual financing instrumentssuch as share-convertible and share-exchangeablebonds of bonds issued by the government specifi-cally for the financing of public enterprises and ofbonds issued by special purpose vehicles in the con-text of securitisation
The Eurostat decisions have been very important toensure comparable results In some cases they have hadsubstantial impact on the accounts of some MemberStates
Multilateral discussion and accountability Given thatthe accounting decisions on specific transactions mayhave significant consequences on the government deficitand debt ratios of Member States Eurostat has taken its
decisions as openly as possible after discussion with thestatistical authorities of all Member States and the con-sultation of the CMFB (1) Although the CMFB opinionis not binding Eurostat always takes the utmost accountof the opinions expressed by the CMFB In practice inmost cases Eurostat follows the opinion expressed bythe majority of CMFB members whenever it was aquestion of deciding on the accounting treatment of gov-ernment transactions Furthermore the Eurostat deci-sions and the CMFB opinions on the recording of gov-ernment transactions and the respective rationale aremade public thus ensuring accountability
43 Assessing the quality of budgetary statistics
The section above described governance of budgetarystatistics In particular the distinction between theCommission and the Member Statesrsquo role is widely rec-ognised as adequate and contributing to the quality ofbudgetary statistics However the quality of statisticsmust be assessed directly that is whether the budgetaryfigures in particular the deficit and the debt ratiosreported by Member States are reliable transparentconsistent and timely
431 Reliability
The reliability of statistics is difficult to measure andeven to define The concept of reliability that is used hererefers to the successive revisions in data Are the deficitand debt ratios reported in March each year reliable inthe sense that they are only slightly revised after sixmonths or later or are deficit and debt figures subject tolarge revisions after the publication of the first estimate
Over the last three and half years (that is since 2000when ESA95 replaced ESA79 as the accounting frame-work for the compilation of government accounts) theaverage absolute revision in the deficit ratios of MemberStates has been 015 of GDP after six months 022 after one year and 026 after 18 months (2) This is avery small figure if one considers that the EU average of
yen1part The CMFB or Committee on Monetary Financial and Balance of PaymentStatistics gathers senior statisticians and national accountants from thenational statistical institutes and national central banks of all MemberStates as well as Commission and ECB representatives
yen2part This indicator is the GDP-weighed average of the absolute differencebetween the deficit (or debt) ratio for year t reported in March t+1 and thedeficit (or debt) ratio for the same year reported in September t+1 andMarch t+2
90
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
government total revenue and expenditure that lead tothe deficit is around 47 of GDP
However in some cases the revisions in the governmentdeficit ratios were unacceptably high For example thedeficit to GDP ratio for 2001 as reported by Portugal wasrevised upwards by 2 points from spring 2002 tospring 2003 by Greece by 15 and by Italy by 12 The government surplus of Luxembourg for 2001 wasalso revised upwards by 14 of GDP
Concerning the debt ratios the average absolute revisionin data has been 031 038 and 041 of GDP aftersix 12 and 18 months The largest revisions in the debtratio in recent years took place in Greece and Austria (1)
In most cases the revisions in the deficit and debt ratiosare because the national statistical institutes receivedbetter data from their basic sources However in otheroccasions the revisions were because Eurostat requestedcountries to amend their data since the accounting ruleshad not been fully respected or following a clarificationof such rules In some cases the revision in the GDP fig-ures also played a role in the revision of deficit and debtratios
Therefore while the deficit and debt ratios reported byMember States have been generally reliable there werevery large revisions in a few countries Although allcountries may still improve the reliability of their datathis issue is particular relevant for the countries forwhich the deficit and debt data were recently signifi-cantly revised
432 Transparency and consistency
All Member States publish complete governmentaccounts that is they publish not only the governmentdeficit figures but also details about their expenditureand revenue even if in most cases such informationappears around one month after the transmission of thedeficit data In this sense government accounts are
transparent as one may understand what is behind anymovement in the deficit ratio (in terms of increase ordecrease in specific revenue and expenditure categories)from one year to the other
For the sake of consistency it is also important that a linkis established between the ESA government deficit andthe cash-based public accounts deficits This is impor-tant because the cash-based balances are easier to com-pile and to monitor as they are directly observable Inaddition the public accounts deficits are scrutinised bythe national institutions like the national parliaments andcourts of auditors Therefore if one is able to explain thelink between the two deficit concepts the ESA govern-ment accounts profit from the scrutiny made at the levelof the public accounts
All countries transmit to the Commission data on the linkbetween the cash basis figures and the ESA governmentdeficit for central government However for severalcountries this information is relatively confusing or notcomplete or there are important statistical discrepanciesMoreover only one Member State (Spain) has transmit-ted detailed information on the link between the cash fig-ures and the ESA accounts for the lower subsectors(regional and local authorities and social security) Thisis clearly an area where there is still much progress to bemade
433 Timeliness
Most countries always transmit their data to the Com-mission within the reporting deadlines However somecountries consistently report their data to the Commis-sion several weeks after the established deadlines Inmost cases these delays are because of technical diffi-culties in compiling the government accounts in time forthe reporting deadline However in a few occasionsMember States have also postponed the transmission ofdata on purpose for political reasons such as the proxim-ity of elections
These delays may hinder an effective and expeditedimplementation of the budgetary surveillance mecha-nisms both for the concerned countries but even for allother countries Moreover delays in the transmission ofdata by Member States lead to delays in the publicationof the EU aggregates and impede a proper validation ofdata by Eurostat
yen1part From spring 2002 to spring 2003 the Greek government debt ratio for 2001was revised upwards by 73 of GDP mainly because of the inclusion inthe debt of bonds issued in the context of securitisation of share-exchange-able bonds and of share-convertible bonds In Austria the debt ratio wasrevised upwards by 41 of GDP mainly because of the inclusion in thegovernment debt of bonds issued by the federal government for the financ-ing of public enterprises (Rechtstraumlgerfinanzierung)
91
P u b l i c f i n a n c e s i n E M U 2 0 0 3
44 Recent measures to improve the quality of budgetary statistics
441 The code of best practice
The Ecofin Council of 18 February 2003 endorsed acode of best practice on the compilation and reporting ofEDP data The aim of the code which follows the Com-mission communication on the need and the means toupgrade the quality of budgetary statistics of 27 Novem-ber 2002 (1) is to streamline procedures both at MemberStates and Commission level that may contribute toimproving the quality of budgetary statistics
The main elements of the code of best practice (the fulltext of the code of best practice as endorsed by the Coun-cil can be found in the annex) are the following
bull the authority of the Commission (and of Eurostat onbehalf of the Commission) in assessing the qualityof reported data and in interpreting the accountingrules is clarified and reinforced
bull the Member Statesrsquo responsibility to compile andreport data to the Commission and their commit-ment to strictly respect the accounting rules and thereporting deadlines
bull the need to ensure transparency and consistency inbudgetary statistics and to report figures that are asupdated as possible
bull the reporting tables will be revised as experiencehas shown that more precise and detailed informa-tion is needed (2) while each Member State willprovide an inventory of methods procedures andsources (3)
bull Member States are encouraged to address account-ing issues at the earliest stage when there are doubtson the correct accounting treatment of a government
measure Eurostat should be formally consulted onthe recording of specific transactions
bull the procedure leading to the Eurostat decisions onaccounting issues is streamlined and accelerated Asa rule no accounting issue should be left pending atthe time of the EDP reporting of 1 March and 1 Sep-tember Moreover as a general rule the Eurostatdecisions should be taken within six weeks (4) aftera formal request has been received
bull Eurostat is entitled to examine in depth the ESAgovernment accounts of each Member State tocheck compliance with the accounting rules toexpress reservations to the reported figure and toamend such figures if need be
In the above-referred communication of 27 November2002 the Commission concluded that lsquoan improvementin the quality of budgetary statistics requires effort andstrong commitment from all parties The Commissionbelieves that the reliability of budgetary statistics wouldprofit from a clarification and streamlining of proceduresfollowed both by the Member States and by the Commis-sion This clarification and streamlining should take theform of a code of best practice that all concerned partiescommit themselves to implementrsquo
442 Towards quarterly accounts
EU budgetary surveillance is based on annual data Thismeans that data that are relevant for deciding whether acountry is complying with the SGP requirement ofbudget positions of lsquoclose to balance or in surplusrsquo orwhether such a country is in an excessive deficit positionare the deficit and debt ratios for each year Given thatthe government budgets are adopted by the politicalinstitutions of each country and implemented in a yearlyfrequency it would not make any sense to implement theEDP and SGP on a basis other than annual
However quarterly accounts for general government canbe very important for budgetary surveillance for severalreasons First quarterly government data allow thebudgetary policy analysts to better understand the inter-action between the fiscal positions of countries and theeconomic activity Second quarterly data allows policymakers to better calibrate their measures within each
yen1part COM(2002) 670 finalyen2part This concerns in particular the lower government subsectors given that the
central government is already relatively well covered The new reportingtables will be prepared by the Commission in cooperation with the CMFBand will be implemented from March 2004
yen3part Such an inventory is a kind of document that the national statistical insti-tutes have already prepared in other circumstances It is an important toolto check that deficit and debt figures are compiled according to the account-ing rules and that the data sources and estimation methods are appropriateThe inventory requested by the Council in the code of best practice shouldbe ready for each Member State by the end of 2004
yen4part Please note that this deadline of six weeks does not appear specifically inthe code as there is a cross reference to the CMFB rules of procedure
92
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
year whenever any deviation from plans becomes evi-dent Third the budgetary efforts made by any countrycan be better and more quickly appreciated by the Coun-cil and the Commission Moreover experience fromother statistics shows that the compilation of data with ahigher frequency (quarterly or monthly figures) has afavourable impact on the quality of statistics with alower frequency (annual data)
The compilation of quarterly statistics for general gov-ernment is still at an early stage and should be under-stood as a medium-term project The quarterly govern-ment accounts are governed by three legal acts
First according to Regulation (EC) No 2642000 allMember States are required to transmit to the Commissionquarterly data on taxes and social contributions and onsocial benefits other than in kind since mid-2000 Thesedata are transmitted with a three-month lag after the end ofthe respective quarter Such data have not yet entered theusual rhythm of regular publication as their quality is stillbeing assessed by both the Commission and the MemberStates However one expects that the publication of thesefigures would start later in 2003 Although the variablescovered by Regulation (EC) No 2642000 represent a rel-atively small part of the complete government accountand do not allow the compilation of a quarterly govern-ment deficit they have the potential of becoming very rel-evant indicators as they are the government account itemsthat are most sensitive to economic activity
Second according to Regulation (EC) No 12212002Member States should compile and transmit quarterlydata for all other items of the government account lead-ing to the compilation of a quarterly government deficitMost countries are already compiling these figures andall of them will do so by mid-2004 However as in thecase with the data on taxes social contribution and socialbenefits such figures will be subject to a quality assess-ment period and the publication of data per country is notexpected before the end of 2005
Third the compilation of quarterly statistics on the gov-ernment financial transactions and of the governmentfinancial balance sheets is being envisaged The relevantlegal acts still have to be adopted by the European Par-liament and the Ecofin Council but the plans are thatthese data will be compiled from 2003 or 2004 on
For the time being there is no legal act on the compilationof the government debt with an infra-annual frequencyalthough a few Member States do compile such figures
45 Conclusion and challenges for the future
This chapter described the main elements of the govern-ance of budgetary statistics in Europe The main ele-ments of the governance mdash well-defined accountingrules and a clear distinction of roles between the Com-mission and the Member States mdash have shown to be nec-essary adequate and have contributed to the increase inthe quality of budgetary statistics in the EU
However there is still scope to improve the reliabilitythe transparency and timeliness of budgetary statistics inmany countries A strict implementation of the recentlyagreed code of best practice will also give a majorcontribution to the quality of budgetary statistics Fromthe Member Statesrsquo side this requires increasing thetransparency of government accounts in particular withrespect to the government subsectors a stricter respect ofdeadlines an overall increase in the data quality as wellas a reinforcement of the independent role of the nationalstatistical institutes as the main compilers of governmentdata From its side the Commission needs to reinforceits ability to scrutinise the Member Statesrsquo governmentaccounts in more detail Moreover it should acceleratethe process to decide whenever there are doubts howspecific government transactions are recorded in theaccounts
ESA has performed well as the accounting reference andits usefulness as a budgetary surveillance tool has notbeen challenged However one should acknowledge thatit is an extremely complex system which is not alwaysproperly understood by policy makers and that the com-pilation of the ESA government deficit and debt is noto-riously difficult lengthy and costly This is partiallybecause the foundations of the accounting system weredeveloped in a context other than budgetary surveillanceand before EDP and SGP were set up
Moreover in a context of evolving surveillance theaccounting rules need to be further developed to take dueaccount of innovative transactions or the changingnature of government units (1) The accounting systemshould remain consistent provide policy makers with
yen1part For example the reform of the public pension schemes the development ofthe securitisation of government assets or of the partnerships between thepublic and private sectors for the construction of public infrastructures andthe provision of public services etc
93
P u b l i c f i n a n c e s i n E M U 2 0 0 3
reliable data and the adequate set of incentives to remainthe appropriate tool for budgetary surveillance
The compilation of quarterly budgetary statistics is amajor challenge for the next years The challenge ismainly for the statisticians who will compile the datasince the quarterly data are notoriously more difficult tocompile than annual figures However it is also a chal-
lenge for economists policy-makers and budgetary pol-icy analysts who will need to learn how to read quarterlydata since these will necessarily be more volatile sub-ject to more revisions and perhaps less transparent thanannual data Anyhow whilst quarterly data will give asignificant contribution for the public finance analysisthe formal budgetary surveillance mechanisms willremain on a yearly basis
94
Annex A Budgetary surveillance for long-term sustainability in EU Member States
Part I3 of this report described how the sustainabilityof public finances is assessed on the basis of annualupdates to stability and convergence programmes andexplained that the Economic Policy Committee is con-tinuously working on the production of more compara-ble long-run projections on the budgetary impact ofageing populations on public expenditures As part ofits work on the sustainability of public finances theworking group on ageing populations attached to theEconomic Policy Committee (EPC) recently carriedout a questionnaire survey on whether and how thesustainability of public finances is systematicallyaddressed as a part of the budgetary-setting process inMember States
This annex presents a short summary of the results Thequestionnaire was divided into two main parts A firstsection examined how Member States carried out thelong-run budget projections A second part of the surveyexamined how such projections are used in the budget-ary-setting process and in particular whether considera-tions on the sustainability public finances are taken onboard in the setting of short- and medium-term budget-ary priorities
Long-term projections coverage and updating
All Member States currently produce long-term projec-tions for at least some expenditure or revenue items Pen-sions and healthcare represent the most relevant publicexpenditures affected by ageing and they are generallyfully covered in the projections exercise mainly thanks
to the common projections carried out by the EPC at theend of 2001 see Table II9) (1)
The coverage of revenue projections is more limited dueto methodological difficulties Any projection of taxrevenues should make assumptions on development oftax rates as they tend to adjust to the level of publicexpenditures (2) It also requires a detailed knowledge ofincome distribution and its evolution since this canchange the tax bases for direct and indirect taxes More-over the indirect effect of taxation on labour participa-tion and on income levels should be assessed to projectthe likely impact of ageing on revenues
A key issue is the demographic scenarios used toperform the projections of age-related expenditures andrevenues All Member States run several projections totake account of different possible scenarios to take intoaccount uncertainty over long-term demographic devel-opments The demographic scenarios are not fully con-sistent across countries since in many cases they arebased on national projections and not on Eurostat dataHowever the use of national scenarios makes it easier totake on board the latest demographic projections whichtake account of fast-changing variables such as migra-tion flows
In most Member States long-term projections are regu-larly updated to take into account at least changes in theeconomic environment andor the demographic scenarioIn Denmark the UK and Sweden they are updated more
yen1part The information provided below comes from a survey across MemberStates carried out by the Economic Policy Committee of the EuropeanUnion
yen2part See Martinez-Mongay C (2000)
95
P u b l i c f i n a n c e s i n E M U 2 0 0 3
often than once a year in Belgium Germany and Italyprojections are updated annually (1) Longer time spansare considered in Ireland (two years) Austria (threeyears) and the Netherlands (four years) Irregular updat-ing is being done in France Finland and Portugal InGreece a lsquoNational Actuary Authorityrsquo has just beenestablished and it will produce long-term projections ona regular basis in the coming years
The process of producing long-term budgetary projectionsgenerally involves several actors In most cases the finalresponsibility for producing the projections is within a gov-ernmental body mainly the TreasuryFinance Ministry orthe LabourSocial Affairs Ministry Social partners inde-pendent experts and social security institutions are fre-quently involved at some stage in the preparation of techni-cal assumptions and in the feedback of the first wave ofresults In many Member States there are ad hoc public bod-ies (committees and working groups) composed of officialsfrom the public administration and external experts socialpartners and representatives of the national Parliament
For instance in Germany consultation is a regular fea-ture of each annual update a workshop on methodologyand the main assumptions that involve the Pension Insur-ance Institutions (VDR) and the Federal Ministry ofHealth and Social Affairs is organised Other institutionsare also consulted and at the end of the process a special
advisory board assesses the results and forwards theassessment to the Federal Parliament In Austria a con-sultant body to the federal government composed of min-istry representatives social partners and researchers dis-cusses projections and it presents subsequently a report tothe government In Portugal there is an interministerialworking group on ageing that discusses technical aspectsof the projections An ad hoc group is also established inthe Irish Finance Ministry (the Long-Term Issues Group)and in Belgium (Comiteacute drsquoetudes sur le vieillissement)In France a body attached to the Prime Ministerrsquos Officecoordinates the consultation with many different actors(social partners Parliament Ministry of Finance etc)
The use of projections in budgetary procedures
All Member States use long-term projections at somestage of the budgetary process reflecting a shift in recentyears from budgetary procedures that only focused onshort-term targets to procedures that incorporate morelonger-term considerations
Currently long-term projections are used in SwedenFinland the Netherlands Belgium and Denmark as atool in setting the medium-term budgetary targets of thegovernment
Long-term projections are also used in the majority ofcountries at the design stage of major reforms in partic-
yen1part In the case of Germany this applies to projections performed for the gen-eral statutory pension scheme
Table A ndash Long-term public expenditures development covered by national projections
BE DE EL ES FR IE IT LU NL AT PT FI DK SE UK
Pensions of public employees X X X X X X na X X X X X X
Pensions of private employees X X X X X X X na X X X X X X
Pensions of employers X X X X X X X na X X X X X
Second pillar pensions X na X X
Third pillar pensions na X X
Healthcare X X X X na X X X X X X
Education X X na X X X X
Others (1) X X X na X X X X X X X
(1) IE Other areas of social welfare such as child benefit and unemployment benefit payments NL All other expenditure items (for example defence general gov-ernment transfers abroad) FI Services long-term care child day care Benefits family allowances unemployment benefits sickness insurance allowances hous-ing allowances living allowances etc DK unemployment benefits labour market- and maternity leave cash benefits early retirements benefits pension benefitspayable between early retirement and normal retirement (efterloslashn) child care and residential support for elderly SE All public sector expenditures UK All spend-ing for example long-term care non-pension social benefits (for example child benefit incapacity benefit housing benefit) net transfers abroad etc IT age-related lump sums other than pensions will be projected in the coming years AT contributions and federal transfers PT long-term care projections available inDecember 2003 BE all social security expenditures are included sickness and disability Family allowances unemployment early retirements
96
P a r t I IE v o l v i n g b u d g e t a r y s u r v e i l l a n c e
ular those related to pensions or tax systems Projectionsare generally used as additional information for prepar-ing specific provisions of legislation In some countriesthere is a legal obligation for each new law or amend-ment to be accompanied by a technical report on thelong-term budgetary effects for which the use of projec-tions is indispensable For instance in Italy the long-term (10 year) impact of a pension reform must beassessed and annexed to the law proposal In the UKindividual reforms are generally assessed for their long-term fiscal sustainability before policies are imple-mented In Germany such projections were used whenthe 2001 pension reform was devised
The assessment of long-term sustainability of public finances
The assessment of long-term sustainability of publicfinances is conducted primary by ministries of financeeconomy but there are cases where the SocialLabourMinistries or other public institutions are involved(Table B)
A key issue is the definition of long-term sustainabilityof public finances It can refer to debt dynamics or to abudget balance position In the Netherlands and Den-mark public finances are considered sustainable if debtis not on an lsquoexplosive pathrsquo implying a constant debt-to-GDP ratio over the long term Other countries refer tothe Treaty requirement of 60 in the debt-to-GDP ratioas in Sweden where a sustainable debt path is one whichnever exceeds the Treaty reference value In Italy thereare currently two ways to assess long-term sustainabilityof public finances One has been developed in the 2002updated stability programme for the first year and refers
to a debt reduction towards 60 of GDP A second def-inition of sustainability refers to the impact of differentdebt structure scenarios on the cost of debt and on realGDP growth rates This analysis is then used to projectthe evolution of the debt-to-GDP ratio in a long-termperspective
Belgium and Austria refer more explicitly to the defini-tion given by the EPC that is each year to maintain abudget position which is balanced or in surplus A ratherdifferent definition is the one used in the UK where sus-tainability is defined as meeting the governmentrsquos sus-tainable investment rule which says that net debt shouldremain below 40 of GDP over the economic cycle
On the basis of the above-mentioned definitions of long-term sustainability countries use a number of indicators
bull budget balance the country is not sustainable if thebudget balance cannot be maintained for the wholeperiod covered by the projections
bull fiscal gaps tax ratios (whether the current tax ratiois sustainable)
bull increase of expenditure and revenue which are sen-sitive to changes in the composition of the popula-tion (mainly pension expenditures)
bull economic dependency ratios
bull a measure of generational fairness where benefitsfrom government expenditure enjoyed by a genera-tion minus taxes paid by this generation should besimilar across generations
Table B ndash Who makes the assessment of long-term sustainability
BE DE EL ES FR IE IT LU NL AT PT FI DK SE UK
Ministry of FinanceTreasuryEconomy X X na X na X X X X X X X
Ministry of Health X na X na
Ministry of Social AffairsLabour X na na X
Others (1) X X na na X X
(1) BE Conseil Supeacuterieur des Finance - public research institute NL Netherlands Bureau of economic policy analysis DK independent institutions DE Ministry ofthe Interior (if the civil servantsrsquo pension scheme is assessed)
97
Part III
Public investment and its interaction with the EUrsquos budgetary rules
Summary
Public investment as a share of GDP has fallen in the EUin recent decades and currently public investmentexpenditures are relatively low compared with otherindustrialised areas There is a widespread perceptionthat the process of budgetary consolidation (both beforeand after the launch of the euro) and the application ofthe EUrsquos fiscal rules has contributed to excessively lowlevels of public investment it is claimed that a sustainedgrowth in spending would improve the EUrsquos growthpotential in accordance with the Lisbon strategy
However data analysis shows that the decline in publicinvestment rates is a long-run tendency that had alreadystarted in the 1970s and affected all industrialised coun-tries and not just EU Member States Declining levels ofpublic investment as a share of GDP have been attributedto factors such as economic development and structuralchange (with developed countries already havingacquired a high stock of physical capital) and the chang-ing boundaries between public and private investment(in part linked to the process of privatisation) Some ofthe decline in public investment levels appears to berelated to efforts to consolidate public finances whichwas necessary irrespective of EMU A careful analysisof the data taking account of other explanatory variableshowever fails to show any clear-cut link betweenchanges in investment ratios and the provisions of theEUrsquos framework for fiscal surveillance Indeed publicinvestment expenditures in many Member States havestopped falling since the beginning of monetary union
Public investment can make an important contribution tomeet the output and employment goals of the Lisbonstrategy However in considering the links between pub-lic investment and growth it is important to focus on netas opposed to gross investment levels (that is takingaccount of the depreciation of the existing capital stock)and also the interaction between trends in public and pri-vate investment levels Existing studies reveal that pub-lic investment has a positive impact on output and pro-ductivity although the results are not very strong and
depend quite crucially on the analytical methodologiesemployed This is explained by the fact that only a frac-tion of public investment expenditures are devoted toprojects which aim directly at improving the allocationof resources and raising productivity (for exampleinvestment in transport infrastructure) a significant pro-portion of public investment is devoted to projects thatpursue other objectives such as environmental protectionor redistribution across regions which only indirectlycontribute to output
Understanding and measuring the links between publicand private investment is also crucial to assessing theoverall impact of public investment on the economy andits growth potential A priori both a complementarity ora substitution relationship can be expected between pub-lic and private investment depending on whether crowd-ing-out effects via reduced savings and increased interestrates are compensated by higher productivity of privatecapital associated with enhanced public infrastructure Inrecent decades both public and private investment rateshave declined in the EU as a whole although there aresignificant differences across countries In some coun-tries such as Greece Ireland Luxembourg and Portugalboth public and private investment have been risingConversely both type of investments have been fallingin other countries such as France Germany Italy andthe Netherlands Finally in other countries such as Aus-tria Denmark and the UK the fall in public investmenthas been coupled with a moderate increase in privateinvestment The analysis of the data shows that publicinvestment has a poor explanatory power on the dynam-ics of private investment the effect is generally not sig-nificant with the exception of the UK where there issome evidence of crowding-out and that of Portugal andSpain where instead the evidence indicates a crowding-in effect In summary the hypothesis that a generalisedincrease in public investment expenditures in the EUwould contribute to growth via higher private investmentreceives little empirical support
101
P u b l i c f i n a n c e s i n E M U 2 0 0 3
The important role of public investment is recognised inthe existing framework for budgetary surveillance forexample Member States are required to specify plannedpublic investment levels in their annual updates to stabil-ity and convergence programmes and the BEPGs fre-quently recommend that an increased share of total pub-lic expenditures be devoted to productive items such asinvestment In brief the budget balance requirements ofthe Treaty and SGP are compatible with a high share ofpublic spending being devoted to public investment Therecent Commission communication on strengthening thecoordination of budgetary policies sought to cater for thebudgetary impact of large investment projects while atthe same time respecting the commitment to sound andsustainable public finances
Several calls have been made to introduce a so-calledgolden rule into the SGP which would allow govern-ments to borrow to finance investment However thereare strong theoretical and practical arguments against itsintroduction especially in a framework of multilateralsurveillance such as the SGP First a golden rule basedon a national accounts system could lead to a bias inexpenditure decisions in favour of physical capital andagainst spending on human capital (education and train-
ing) or other productive items (healthcare and R amp D)which also contribute to growth and employment Sec-ond if applied to gross investment the adoption of agolden rule into the SGP framework may imply substan-tially higher deficits thus compromising the objective ofsustainability of public finances Finally to be effectiveit would need to apply to net investment however dataon net investment is neither reliable nor timely
There is a growing practice of financing public purposeinvestment projects through publicndashprivate partnerships(PPPs) The main implication for public finances ofchoosing PPPs as opposed to traditional public invest-ment is in fact that of converting up-front fixed expen-ditures into a stream of future obligations While thispractice has a sound microeconomic rationale (increasedefficiency without compromising public objectives)there is the risk that the recourse to PPPs is increasinglymotivated instead by the purpose of putting capitalspending outside government budgets in order to bypassbudgetary constraints If this is the case then it may hap-pen that PPPs are carried out even when they are morecostly than purely public investment Efforts are alsorequired to ensure a transparent recording of PPP trans-actions in national accounts
102
1 Introduction
Public investment as a share of GDP has fallen in mostindustrialised countries in recent decades promptingmany commentators to argue that this is having negativeconsequences on productivity In the EU context it hasbeen claimed that the deficit targets of the Treaty andSGP may contribute to keeping public investmentexpenditures at excessively low levels and that consid-eration should be given to allowing for a special budget-ary treatment for public investment
This part of the report analyses and discusses the issue ofpublic investment in the framework of the EUrsquos fiscalrules Public investment is analysed from a long-runmacroeconomic perspective Issues related to sectoralpatterns or microeconomic efficiency (for example cost-benefit analysis) are therefore left aside and the focus ison the aggregate trends in public investment and theirdeterminants and on the impact of public investment onoutput growth and private investment
While the effects of public investment on output andgrowth have been extensively studied empirically in thepast decade there is little work investigating systemati-cally how public investment relates to private investmentin EU countries New empirical analysis is thus carriedout to investigate this issue Original analysis is alsoundertaken to study the relationship between public andprivate investment in EU countries and the impact of theadvent of EMU on the evolution of public investment
Chapter 2 provides a definition of public investment anddescribes the broad trends in public investment level indeveloped economies in recent decades
Chapter 3 examines the economic rationale for publicinvestment and its potential impact on productivity Inparticular it surveys the main empirical findings on thismatter
Chapter 4 takes a closer look at developments as regardspublic investment in EU Member States It focuses onthe relationships between public and private investmentlevels in EU countries and also considers whether publicinvestment levels have been affected by the Treaty andSGP budgetary requirements both before and after thelaunch of the euro While this section focuses on the linkbetween budgetary consolidation and investment itshould also be borne in mind that a reverse causationcould exist as transparent public procurement proce-dures can contribute to budgetary savings (1)
Chapter 5 is forward looking and examines the pros andcons of proposals to modify the existing EU fiscal rulesto include a golden rule for public investment It alsopresents the main features and the budgetary implica-tions of publicndashprivate partnership agreements forundertaking public investments
yen1part OECD (2003a)
103
2 Public investment definition and broad trends
21 The definition of public investment
Through public investment governments increase andimprove the stock of capital employed in the productionof the goods and services they provide It is important tonote that the term lsquopublic investmentrsquo used in this chap-ter refers to a rather unique definition used in nationalaccount statistics and thus excludes certain expenditureswhich typically might be considered as constitutinginvestment (Box III1) It includes the relevant transac-tions that lead to changes in the stock of physical capitalbut excludes a large amount of expenditures related to theaccumulation of human capital For example the construc-tion of research laboratories or the purchase of computersoftware is included in the definition of public invest-ment but wages paid to researchers and scientists arenot in national account statistics this type of spending isclassified as current expenditures of the public sector inspite of the fact that the labour services provided by theseprofessional categories contribute to the accumulation ofhuman capital Equally investment in knowledge (edu-cation training or R amp D) also enhances productivityperformance in the long run by favouring more knowl-edge-intensive higher value-added job creation but thisis not captured by the national account definition (1)
With regard to the contribution of the stock of publiccapital a distinction should also be drawn between grossand net investment by the public sector Only the con-cept of net investment takes into account depreciation(that is the loss of economic value of the current capitalstock due to usage or obsolescence) and as such is thecorrect measure of the actual change in value of the stockof public capital However the available statistics on netinvestment are the result of estimation methods and are
of limited reliability It is therefore common to refer tothe statistical aggregate lsquogross fixed capital formation ofthe general governmentrsquo to obtain country-level infor-mation on public investment
22 Broad trends of public investment in industrialised countries
In most OECD countries (gross) public investment hason average been below 5 of GDP in the past 30 yearsa fraction about five times lower than private investmentFrom the 1970s onwards public investment rates havebeen falling significantly in a number of OECD coun-tries although the picture is quite differentiated acrosscountries (see for example Roubini and Sachs 1989Oxley and Martin 1991) (2)
Focusing on the EU US and Japan Graph III1 showsthat gross public investment as a share of GDP fell visi-bly in the US and in the EU during the 1970s and the firsthalf of the 1980s whereas in Japan the trend was broadlypositive (3)
yen1part European Commission (2002d)
yen2part A downward trend in public investment as a share of GDP is quite substan-tial in non-EU OECD countries such as Norway Canada Australia Icelandand New Zealand In Switzerland the share of public investment on GDPhas instead remained quite stable The main exceptions among OECDcountries are Japan and South Korea where on average the role of publicinvestment has been growing
yen3part In the whole analysis data to Germany in the years before unification referto West Germany only Moreover in this part of the report ESA95 grosspublic investment data are linked with those referring to the previous clas-sification systems system according to the following criterion
where the subscript lsquoFormerrsquo refers to the classification used before ESA95This linking methodology assumes that the growth rates in the variables arethe same irrespective of the accounting system employed and has the advan-tage of avoiding lsquojumpsrsquo in time series in correspondence with the year inwhich the accounting system changes
Xt Xt 1ndash 1Xt Xt 1ndashndash
Xt 1ndash----------------------
FORMER
+ =
104
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
Graph III1 Gross fixed capital formation general government of GDP at current market prices
Box III1 Public investment in national account statistics
(Continued on the next page)
0
1
2
3
4
5
6
7
US Japan EU-15
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
In national account statistics investment is defined as expenditures in fixed assets for example in items that last for morethan one year So while for instance teachersrsquo wages are classified as current expenditures buildings or furniture used inthe education sector enter the definition of investment The most common statistical definition of public investment is thegross fixed capital formation of the general government Since the general government is the relevant institutional unitthis definition includes investments carried out by the central government and by local authorities but excludes invest-ments by public enterprises classified as market units
In the ESA95 system of accounts (see Council Regulation (EC) No 222396) gross fixed capital formation consists of lsquores-ident producersrsquo acquisitions less disposals of fixed assets during a given period plus certain additions to the value of non-produced assets realised by the productive activity of producer or institutional units Fixed assets are tangible or intangibleassets produced as outputs from processes of production that are themselves used repeatedly or continuously in processesof production for more than one yearrsquo
Some remarks concerning the above definition are warranted First gross fixed capital formation does not take necessarilypositive values Negative values may be recorded if the public capital stock is reduced through sales of assets Secondchanges in inventories are excluded meaning that the stock of items other than fixed assets that can be cumulated and car-ried over (for example materials and supplies used as intermediate inputs in production) are not part of gross fixed capitalformation Third fixed assets are not necessarily physical Intangible assets like patents or software enter in fact the def-inition of gross fixed capital formation Finally it should be noted that some types of military expenditures such as thelsquopurchase of military weapons and their supporting systemsrsquo are not included in the category of gross fixed capital forma-tion whereas all military expenditures with a possible civilian use (for example hospitals) are included This is a majordifference with respect to the accounting system previous to ESA95
105
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Public investment as a share of GDP in the EU continuedto fall throughout most of the 1990s but started to rise inlater years In contrast public investment in the US hadstarted to rise already by the mid-1980s and by the endof the 1990s it had surpassed the EU
A large body of studies has identified several factors thatcould explain this downward trend in public investmentlevels (1) First there are reasons linked to economicstructural development The supply of public capital(public infrastructure especially) depends upon the levelof economic development of a country At very low lev-els of development the supply of public infrastructuresis limited by the availability of financial (savings) andtechnical resources At intermediate levels of develop-ment the limiting role of these factors weakens and thecontribution of public infrastructure to the economybecomes more important At high levels of developmentthe marginal productivity of public physical capital startsdecreasing while the role of knowledge and human cap-ital becomes more important In brief public investmentlevels are likely to be highest in countries at intermediatelevels of economic development
Second there are reasons related to the changingboundaries between the public and private sector asregards the provision of overall investment in the econ-omy In recent decades the private sector has increas-ingly replaced the government in the realisation of riskylong-term projects due to the development of more effi-cient capital markets and better possibilities of hedgingrisk via market instruments Also many industrial andindustrialising countries in the 1980s and 1990s have
been characterised by privatisation practices throughwhich activities owned and managed by the public sectorhave been transferred totally or partially to the privatesector (2) Moreover in a number of countries a growingshare of investments in public interest have been carriedout through the operation of publicndashprivate partnershipagreements (PPPs) Frequently investments carried outin this way are not registered as government investmentin national account statistics (see Section 53 of thischapter)
Finally there are reasons related to the need to consoli-date public finance positions From the 1980s onwardsmany industrial countries especially in Europe werefaced with rising public deficits and debts In manyinstances governments found it easier to achieve a partof the consolidation of public finances by reducing pub-lic investment
Overall the trend towards falling levels of public invest-ment has led to an extensive debate as to whether this isin part responsible for lower productivity and growthrates This issue is examined in the next section of thischapter There has been an added dimension to this pol-icy debate in the EU namely whether the need to respectthe budgetary requirement of the Treaty and SGP hasaffected the level of public investment an issue which istaken up in Section 4
Box III1 (continued)
The concept of net fixed capital formation takes into account the flow of resources that are used up during the year in main-tenance operations (repairing or substituting capital goods) and the depreciation of existing fixed assets of the public sectorThe quantification of net fixed capital formation is obtained by subtracting capital consumption from gross fixed capitalformation In available national account statistics capital consumption figures are the result of an estimation method Inthe ESA95 system of classification the suggested estimation method is based on the value of the stock of fixed assets(obtained through the perpetual inventory method) and the probable average economic life of the different capital items
yen1part For empirical evidence on this issue see for instance de Haan Sturm andSikken (1996)
yen2part Privatisation practices may result in falling public investment figuresbecause of two reasons The first is that the sales of non-financial assetsowned by the general government enters with a negative sign in the defini-tion of government investment statistics The second is that after privatisa-tion the investments related to the transferred activities (for example toimprove or expand their services) stop being undertaken by the governmentand exits from public investment statistics
106
3 Public investment its rationale and impact on efficiency
31 The rationale for public investment
Public sector economics identifies a number of reasonswhy governments should undertake public invest-ment (1) In many instances the promotion of economicgrowth is not the main (or even minor) rationale for agovernment to undertake a particular public investmentand therefore the link between public investment andefficiency (productivity) is very often only of an indirectnature
A first reason for public investment is the supply of pub-lic goods that is goods for which there is no rivalry inconsumption and that would be under-supplied by theprivate sector alone A typical example would be publicinvestment in transport infrastructures such as roadsharbours or railways In general these are intermediatepublic goods that is they produce their benefits asinputs in the production process rather than as finalgoods and have an important impact on the efficiency ofthe private sector investments However not all govern-ment investment on public goods is likely to have adirect impact on productivity For example investmentin infrastructures to ensure clean air and water whileessential for the general welfare of citizens may onlyindirectly feed through to efficiency
A second rationale for public investment comes from thepresence of various sources of market failures Invest-ments in infrastructures with environmental purposesserve to deal with pollution or other types of environ-ment-related externalities Investment in the educationsector can be justified on the ground of human capitalexternalities and knowledge spillovers Due to such phe-
nomena the social marginal productivity of educationwould exceed the private one In the absence of publicintervention under-investment in schooling and educa-tion-related activities would arise (2)
Another category of market failures that justifies publicintervention in the provision of infrastructures comesfrom the presence of increasing returns and naturalmonopoly-type arguments The provision of networkinfrastructures (in energy distribution or telecommunica-tion for instance) could be subject to increasing returnsassociated with so-called network externalities resultingin a natural tendency towards monopolisation In suchindustries public intervention through the direct supplyof services or the regulation of the sector is desirable toovercome the inefficiencies associated with the under-supply by the private sector Since public utilities pro-vide important intermediate inputs in private sector pro-duction their efficient provision has an impact on over-all productivity However it should be pointed out thatdue to technological and institutional innovation (forexample international liberalisation of air transport andpublic utilities) in recent decades the role of naturalmonopolies has been shrinking thereby enabling gov-ernments to leave the provision of such goods and serv-ices to the private sector
A third argument in favour of public investment is thatof missing markets for capital or insurance that resultfrom asymmetric information problems In the absenceof properly functioning capital and insurance marketsprivate firms may not be willing to undertake riskyprojects or projects that can be recovered only over avery long time horizon In these cases the only alterna-
yen1part For a general treatment of the rationale for public sector activity see forexample Atkinson and Stiglitz (1990) See also European Commission(2002a)
yen2part It should be noted however that only spending on education infrastruc-tures (such as school buildings etc) is recorded as public investment innational account statistics whereas spending on teachersrsquo salaries isrecorded as current expenditures
107
P u b l i c f i n a n c e s i n E M U 2 0 0 3
tive to have such type of projects carried out is throughthe public sector
Summarising there are several reasons that justify thedesirability of public investment in terms of a more effi-cient allocation of resources It should be noted thoughthat in many cases the principal rationale for a particularpublic investment is not to increase efficiency in the sup-ply of goods and services that enter production statistics(GDP) but rather to pursue some other policy objectivethat raises overall welfare for example protection of theenvironment or a fair distribution of resources This isalso the case for investment related to the provision ofseveral types of welfare state services (for example hos-pitals public housing hellip) (1) Hence a priori a stronglink between government investment productivity andgrowth should not be expected
In principle public investments are desirable until thesocial marginal benefit of public capital exceeds itssocial marginal cost Social marginal benefits exceedingsocial marginal costs indicate that public capital is inshort supply and that higher public investment wouldimprove social welfare In practice however the supplyof public capital can be far from the welfare maximisinglevel for several reasons
A basic reason has to do with the lack of information ofthe policy-makers about the costs and benefits of publicinvestment The outcome of the actual economic evalu-ations of policy-makers concerning public investment(for example though cost-benefits analysis) is subject topotentially large errors related to limited information onthe technical characteristics of projects and on citizensrsquopreferences (free-riding problem) The potential dis-crepancy between the outcome of actual cost-benefitanalyses and the lsquotruersquo social marginal costs and benefitsbecome evident by considering that an appropriate esti-mate of social costs should refer to the concept of oppor-tunity cost (which requires an estimation of the benefitsfrom alternative uses of public funds) and should takeinto account the cost of alternative means of financingpublic investment including an assessment of the impactof distortionary taxation
Political economy considerations may also lead toinvestments which are not welfare increasing for thesociety as a whole A basic reason is that public invest-ments such as infrastructures tend to concentrate thebenefits among a clearly identifiable and relatively smallsubset of the population while the costs tend to spreadamong a larger and more diffused group Such types oflsquopork-barrelrsquo projects may end up being over-providedby the public sector (see for example Drazen 2000 onthis subject) (2)
In sum for a number of reasons public capital may eitherbe in short or in excess supply Understanding whetherpublic investment is socially desirable in a particularcountry or region is most often an empirical matter
32 Public investment productivity and growth the empirical evidence
In the 1990s a large amount of research was carried outwith the aim of measuring the contribution of publiccapital in terms of increased production possibilitiesreduced costs for the private sector or enhanced growthprospects In spite of the different approaches and meth-odologies followed and different measures of public cap-ital employed (for example total public investment fromnational account statistics estimates of the net publiccapital stock estimates of the stock of public infrastruc-tures or estimates of transport infrastructure only) allthese analyses assume that public capital is a productionfactor of a particular type
Aschauer (1989a) found a significant and strong positiveimpact of public investment on aggregate output for theUS case whereby a 1 percentage point increase in thepublic capital stock would raise aggregate output byalmost 04 percentage points This result generated avivid debate in academic and policy circles Empiricalwork proliferated investigating alternative datasets (dif-ferent periods or countries) and following new method-ologies In these subsequent analyses not only is theestimated impact of public investment on output smallerbut quite often the results are insignificant or even nega-tive (see Box III2 and Table III1)
yen1part By definition such investments will not necessarily have a direct positiveimpact on overall efficiency However by contributing to social cohesionthey may improve a countryrsquos lsquosocial capitalrsquo and to its long-run productivepotential an efficient allocation of resources
yen2part This does not mean that political economy factors lead to a bias of publicexpenditure in favour of investment expenditure Political economy reasons(existence of political clienteles and pressure groups) may equally explainsa bias towards excessive current public expenditure
108
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
While results do not seem to depend crucially upon theparticular country or period considered the level ofaggregation of the dataset and the way dynamic rela-tions among the variables are modelled seem to matterSome work (for example Bernd and Hansson 1991Conrad and Seitz 1994 La Ferrara and Marcellino2000) compare estimates of the social marginal benefits(proxied by shadow prices) with estimates of the socialmarginal costs of public capital with the aim of deter-mining whether public capital is in short or in excess
supply (1) Quite often the results are not supportive ofthe view that public capital is under-supplied
Box III2 Empirical evidence on the effects of public investment methodologies and results
In recent empirical analyses different methodologies have been followed to analyse the impact of public investment oneconomic activity A first strand of studies follows the so-called lsquoproduction function approachrsquo The aim is that of esti-mating the parameters of an aggregate production function in which public capital enters as a separate productive factorThe obtained estimate of the marginal productivity of public capital is thus chosen as a measure for the benefits of publicinvestment This approach has been followed for the first time in the seminal work of Aschauer (1989a) The analysis fol-lowing this approach generally finds quite ambiguous results (see Table III1) Results appear to depend quite crucially onthe level of aggregation of the dataset and the way dynamic relations among the variables are modelled In general studiesusing panel datasets disaggregated at the state or regional level find a weaker or insignificant impact of public investmentConcerning dynamics once proper techniques are used to obtain stationary series (thus avoid estimating possible spuriousrelations between public capital and output) results tend to become ambiguous
In other studies a different approach has been followed Instead of production functions cost or profit function of privatesector firms have been estimated The idea is that public capital affects the costs and profits of firms as an unpaid fixedinput This approach has the advantage of imposing less restrictions on the equations to be estimated and allowing for theestimation of the shadow price of public capital In most of the cases public capital is found to reduce the costs of privatesector firms However in several studies (for example Berndt and Hansson 1991 La Ferrara and Marcellino 2000) it isfound that public capital is in excess supply since its social marginal productivity (proxied by its shadow price) is lowerthan its social marginal cost
Some analyses followed an atheoretical approach Instead of deriving measures of the contribution of public capitalfrom the estimation of production or cost function equations these studies investigate the dynamic relationship betweenpublic investment and other aggregate variables (output private investment etc) through vector auto regressions (VAR)analysis Under this approach no a priori assumptions are made concerning causal relations all variables are jointlydetermined In most of this work measures of public investment are found to increase aggregate output but there areexceptions (see Table III1) (1)
A different strand of studies analyses the impact of public capital on the growth potential of countries or regions The ideais that public capital (transport or communication infrastructure for instance) has an impact on the accumulation possibil-ities of the economy rather than on the level of output The empirical methodology to test this hypothesis is that of cross-section growth regressions Growth rates in per-capita income over a given time period for a collection of countries orregions are regressed on initial conditions and a list of conditional variables (for example measures of human capitalstock) including the stock of public capital Results from these studies appear to be very fragile Depending on the set ofcountries and regions considered the impact of public capital may or may not be significant
(1) Such results are obtained by means of Granger causality tests
yen1part The shadow price of public capital measures the impact on private sectorfirmsrsquo costs of a unitary increase in the stock of public capital This meas-ure is thus an adequate proxy of the social marginal productivity of publiccapital under the assumption that the main role of public capital is as anintermediate input Estimates of public capital shadow prices are com-monly used in cost-benefit analysis and project evaluation Measures forthe social cost of public capital are based on estimates of the public invest-ment deflator rates of return and depreciation rates
109
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Overall a majority of studies indicate that public capitalhas a positive impact on output productivity or growthHowever results appear to be quite weak and fragileWhen positive the estimated impact in most of the stud-ies is not a strong one and there are cases in which theimpact is insignificant or even negative A certain con-sensus is emerging that public investment is not asimportant for growth as other factors such as invest-ments in human capital (see for example Barro andSala-i-Martin 1998)
These results are mainly explained by the fact that thepurpose of a non-negligible share of public investmentexpenditures is not that of (static or dynamic) efficiencybut rather that of supporting the provision of welfareservices and affecting the distribution of income Dataon the sectoral distribution of public investment in EUcountries indicate that the investment projects directlyaffecting overall productivity and growth potential are
hardly the majority (1) Even if the most important cate-gory is transport infrastructure (roads and bridges in par-ticular) which accounts by itself for almost one third ofthe gross fixed capital formation of the general govern-ment in the EU the rest is devoted to purposes not nec-essarily related to productivity and growth A sharebetween 10 and 15 of public investment is absorbedby fixed expenditures for education and health (forexample construction and maintenance of school build-ings and hospitals) while the provision of public hous-ing and community amenities (for example water andsewers) accounts for roughly 10 of public investmentThe remaining share is mainly devoted to general publicservices (for example administration) defence andsecurity
yen1part Matha et al (2000)
110
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
Table III1
The effect of public investment on output productivity and growth
Study Data Results
1 Production function approach
Aschauer (1989a) US time series 1949ndash85 Positive effect of public capital on output
Sturm and De Haan (1995) US time series 1949ndash85 Positive effect of public capital on output insignificant effects using time differences
Evans and Karras (1994) US panel data on 48 states 1970ndash86 Insignificant effect of public capital on output
Baltagi and Pinnoi (1995) US panel data on 48 states 1970ndash86 Insignificant effect of public capital on output
Garcia Milagrave et al (1996) US panel data on 48 states 1970ndash83 Insignificant effect of public capital on output
Aschauer (1989c) G-7 panel data 1966ndash85 Positive effect of public capital on output
Ford and Poret (1991) 11 OECD countries time series 1960ndash89 Significant positive effect in Belgium Canada and Germany
Merriman (1990) Japan panel data on 9 regions 1954ndash63 Positive effect of public capital on output
Bajo-Rubio and Sosvilla-Rivero (1993) Spain time series 1964ndash88 Positive effect of public capital on output
Dalamagas (1995) Greece time series 1950ndash92 Ambiguous effects
Kavanagh (1997) Ireland time series 1958ndash90 Insignificant effect of public capital on output
Ligthart (2000) Portugal time series 1965ndash95 Positive effect of public capital on output
La Ferrara and Marcellino (2000) Italy regional panel 1970ndash94 Negative effect of public capital on output
2 Cost or profit function approach
Berndt and Hansson (1991) Sweden time series 1960ndash88 Reduction in costs Public capital in excess supply
Conrad and Seitz (1994) Germany panel on three sectors 1961ndash88 Reduction in costs Public capital in short supply during 1961ndash79 in excess supply during 1980ndash88
Dalamagas (1995) Greece time series 1950ndash92 Reduction in costs
Lynde and Richmond (1993a) UK time series 1966ndash90 Reduction in costs
Lynde and Richmond (1993b) US time series 1958ndash89 Increase in output
Morrison and Schwartz (1996a) US panel on 48 states 1970ndash87 Infrastructures have a negative impact on costs
Morrison and Schwartz (1996b) US panel six New England states 1970ndash78 Public infrastructure reduces costs but less than private investment
Seitz and Licht (1995) Germany panel on 11 states 1971ndash88 Reduction in costs
La Ferrara and Marcellino (2000) Italy regional panel 1970ndash94 Insignificant effect on costs Public capital in excess supply for Italy as a whole
3 VAR studies
Clarida (1993) US France Germany UK time series 1964ndash89
TFP and public capital are cointegrated but direction of causality is unclear
Sturm et al (1999) Netherlands time series 1853ndash1913 Public infrastructure Granger-causes output
Otto and Voss (1996) Australia time series 1959ndash82 No significant relation between public capital and output
Ligthart (2000) Portugal time series 1965ndash95 Public investment Granger-causes output
4 Cross-section growth regressions
Barro (1991) 76 countries 1960ndash85 No effect of public investment on per capita GDP growth
Easterly and Rebelo (1993) 100 countries 1970ndash88 Insignificant effect of public investment on per capita GDP growth significant effect of transport and communication spending
Crinfield and Panggabean (1995) 282 US metropolitan areas 1960ndash77 Ambiguous or insignificant effects of local and federal public capital on per capita GDP growth
Host-Eakin and Schwartz (1994) 48 US states 1971ndash86 Insignificant effects of public capital on per capita GDP growth
Mas et al (1994) 17 Spanish regions 1955ndash91 Not always significant effects of public capital on per capita GDP growth
Matha et al (2001) EU countries 1960ndash97 Positive effect of public investment on per capita GDP levels negative on output growth
La Ferrara and Marcellino (2000) Italian regions 1970ndash94 (panel structure) Positive effect of public infrastructure investment on TFP growth
111
4 A closer look at public investment in Member States and the interaction with the EU fiscal rules
41 The evolution of public and private investment in EU countries
411 Trends in recent decades
The EU has been characterised by a prolonged down-ward trend in public investment rates in recent decadesThere is a quite widespread view that such a tendencymay have contributed to reducing the productive poten-tial of EU countries However since what matters foroutput and growth is the accumulation of overall capitalrather than that of public capital only to support thisargument one needs to assess how the decline in publicinvestment shares relates with trends in private invest-ment in EU countries
On average gross public investment in the EU in the1970ndash2002 period has been slightly above 3 of GDPOver the same period private investment averaged about19 of GDP in the 1970ndash2002 period The differencebetween public and private investment is less markedwhen using net fixed capital formation figures The shareof net public investment in GDP was about 14 ofGDP over the same period while that of net privateinvestment is just above 6 This smaller difference ismainly explained by the fact that as on average the stockof private capital is higher than that of public capital alarge part of the investment is devoted to maintenance
Graph III2 provides a breakdown of average gross pub-lic private and total investment-GDP shares over the1970ndash2002 period for each Member State Regardingpublic investment the lowest shares are recorded forItaly Germany and the UK while the highest are thoseof Ireland Luxembourg and Sweden
Evidence concerning net investment is reported inGraph III3 (1) Net investment shares are generally lessthan one half of gross investment shares In Denmarkaverage net public investment during the 1974ndash2001period has been particularly low compared with grossinvestment being slightly negative At the opposite endin Ireland Spain and Portugal net public investment hasbeen relatively high in comparison with gross figuresThese differences across countries between gross and netinvestment figures reflect primarily differences in thesize and composition of the capital stock but may alsobe related to non-uniform practices for imputing depre-ciation
Graph III4 reports average annual changes in the shareof gross private public and total gross fixed capital for-mation during the period 1970ndash2002 For the EU-15 areduction is observed in both the public and private com-ponent of investment resulting in a reduction of the totalinvestment share of about half a percentage point peryear The average annual reduction is stronger for thepublic component which is above 16 percentage pointsper year
Overall the evidence shows that investment sharesdiffer quite widely across countries Differences ininvestment shares seem mainly to reflect differences inper-capita income and levels of economic develop-ment Cohesion countries (Greece Ireland Portugaland Spain) registered relatively high overall investmentshares and both public and private investment rates
yen1part Data are reported for the 1974ndash2001 instead of 1970ndash2002 due to missingvalues
112
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
Graph III2 Gross public private and total investment of GDP average values for the period 1970ndash2002 (1)
(1) Gross fixed capital formation
Graph III3 Net public private and total investment of GDP average values over the period 1974ndash2001 (1)
(1) Net fixed capital formation(2) Excluding Greece and Luxembourg
0
5
10
15
20
25
30
EU-15 BE DK DE EL ES FR IE IT LU NL AT PT FI SE UK
Gross fixed capital formation general governmentGross fixed capital formation private sector Gross fixed capital formation total economy
ndash 2
0
2
4
6
8
10
12
14
EU-13 (2) BE DK DE ES FR IE IT NL AT PT FI SE UK
Net fixed capital formation general government Net fixed capital formation private sectorNet fixed capital formation total economy
113
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Graph III4 Average annual changes in investment shares (1970ndash2002) (1)
(1) Gross fixed capital formation
Graph III5 Cross-country relations between growth rates in public and private investment (average annual changes in shares 1970ndash2002)
ndash 4
ndash 3
ndash 2
ndash 1
0
1
2
3
EUndash 15BE DK DE
EL ES FRIE
ITLU
NLAT PT
FI SEUK
Gross fixed capital formation general governmentGross fixed capital formation private sectorGross fixed capital formation total economy
BE
DK
DE
EL
ES
FR
IE
IT
LU
NL
ATPT
FI
SE
UKy = 00534x ndash 00828
R2 = 00626
ndash 10
ndash 08
ndash 06
ndash 04
ndash 02
00
02
04
06
08
ndash 40 ndash 30 ndash 20 ndash 10 00 10 20 30
Gross fixed capital formation general government
Gro
ss f
ixed
cap
ital
for
mat
ion
pri
vate
sec
tor
114
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
have generally been growing in these countries (1)Conversely investment rates have been generally rela-tively low and falling in countries with relatively highper-capita income This is particularly evident by look-ing at changes in public investment rates with strongnegative values observed for countries such as AustriaBelgium Germany and Sweden characterised by per-capita income higher than the EU average
Another factor that helps to explain cross-country differ-ences in the evolution of public investment rates is theoccurrence of changes in the ownership structure of pro-ductive assets The reduction in the investment activity ofthe public sector is partly the result of privatisation initi-atives especially in the UK Austria and Germany (2)
412 Is there a link between changing levels of public and private investment
A relevant question is the following how does the fall inpublic investment relate with changes in private invest-ment A clear a priori effect of public investment on pri-vate investment is not evident On the one hand as withother types of public expenditure public investmenttends to crowd out private investment via reduced avail-able savings and higher interest rates Public investmentmay also crowd out private investment if the publicsector engages in activities that are strictly substitutedwith those normally carried out by the private sector(for example productive investment by publicly ownedenterprises) On the other hand public investment mayexert a positive effect on private investment (crowdingin) via increased productivity of private sector firmshigher expected profits and better investment opportuni-ties This is typically the case of public infrastructuresthat are used as common inputs in private sector firmsrsquoactivities (for example transport and communicationfacilities)
In Graph III5 growth rates in private investment areregressed against growth rates in public investment
across countries The relationship appears to be positivealthough weak indicating that the countries experienc-ing bigger reductions in public investment are morelikely to also experience bigger reductions in privateinvestment Such an analysis however does not provideany information on the direction of causality so that it isnot possible to say if it is public investment causing pri-vate investment if it is the opposite or if there is a thirdfactor that is simultaneously affecting both public andprivate investment To investigate this issue further timeseries analyses have been performed separately for eachcountry with the aim of assessing the effect of changesin public investment on future developments in privateinvestment (see Box III3 and Table III2) Results areweak and vary considerably across countries In mostcountries public investment did not play a significantrole Crowding-in effects are found for Spain and Portu-gal while for the UK there is evidence of crowding-out
In sum there is no evidence that changes in publicinvestment had a relevant or systematic impact on pri-vate investment developments in EU countries
42 Budgetary consolidation in light of EMU and its impact on public investment
Among the factors that may have contributed to explainthe downward trend in public investment has been theefforts by Member States especially during the mid-1990s to consolidate public finances in light of mount-ing public debt which was accompanied by a consequentincrease in interest expenditure While budgetary con-solidation was necessary in any event the prospect ofstage III of EMU and the entry into force of EU budget-ary rules may also have played a role With the entry intoforce of the Maastricht Treaty Member States commit-ted to avoid excessive deficits and high debt levels (anentry condition for joining the euro area) An additionalbudgetary requirement came into force with the launchof the euro in 1999 namely the objective of the Stabilityand Growth Pact to achieve budget positions of lsquoclose tobalance or in surplusrsquo
The purpose of this section is to examine the impact ofbudgetary consolidation on public investment rates inEU countries It should be stressed that this analysisexamines the relationship between budgetary consolida-tion in terms of deficit levels and changes in public
yen1part An additional reason why public investment shares may have been in gen-eral higher and growing in cohesion countries is the availability of Commu-nity structural funds However this should not be considered as a structuraldeterminant and the size and direction of structural funds will change afterthe accession of new Member States
yen2part The shift of ownership concerned mainly energy and telecommunicationinfrastructure As a result of privatisation public investment in these coun-tries became even more concentrated into fewer sectors such as transportinfrastructure health and education (OECD 1998) In the UK case afterthe privatisation of telecom and energy companies and of airports and rail-ways about 15 of UK gross fixed capital formation was transferred out-side the general government sector (Pollitt 2000)
115
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Box III3 Public and private investment in EU countries crowding in or crowding out
The purpose of this analysis is to assess which impact public investment had on private investment in EU countries Acommon methodology followed in time-series analyses to test whether one variable has a significant impact on anothervariable (or set of variables) is through Granger causality tests This test permits the understanding of whether the pastvalues of the variable to be tested (public investment in this case) adds explanatory power to an existing relationshipbetween one variable (private investment) and its lags (1)
Granger causality tests are performed for all 15 current EU countries Yearly data are used ranging from 1970 to 2002 Thechosen specification to perform Granger causality tests is the simplest possible and it is the same for all countries Privateinvestment at time t is assumed to depend upon its own value at time t-1 and upon public investment at time t-1 This formu-lation permits to save degrees of freedom given the limited number of time series observations Variables are expressed asfirst differences of their logarithm This transformation permits to obtain stationary time series so that ordinary least squaresestimation methods can be used The logarithmic transformation permits the interpretion of the variables employed in theregressions as growth rates of the underlying variables Formally the equations to be estimated are as follows
where (resp ) is the difference between the log of private (resp public) investment at time t and time t-1 while is a random term
Testing whether public investment has an impact on private investment (Granger causes) in the above specification simplyamounts to test whether the parameter γ is significantly different from zero A significantly negative value for γ indicatescrowding out a positive value would be associated with crowding in
Results are reported in Table III2 for all EU-15 countries The coefficient of public investment normally turns out to benot significant with the exception of three countries Spain Portugal and the UK For Spain and Portugal the estimatedimpact of public on private investment is positive conversely for the UK it is negative A possible interpretation of theresults for Spain and Portugal can be related with decreasing returns in public capital Spain and Portugal are character-ised by a relatively low publicprivate investment ratio during the period considered (see Graph III2) For these coun-tries since the stock of public capital is relatively low (and thus its marginal productivity relatively high) an increase inpublic capital results in higher productivity for the private sector and then in enhanced profits and better investmentopportunities for private firms (2) In the case of the UK a possible explanation comes from the process of privatisationof the 1980s and 1990s and the growing involvement of private sector firms in the realisation of projects of public interestDue to changing ownership of assets from the public to the private sector falling public investment in the UK may havecoincided to a certain extent with a corresponding increase in investment by the private sector Something else to noteis the negative and almost significant coefficient for public investment in the case of Sweden In this country the publicprivate investment ratio is higher compared with the rest of EU countries This may indicate a relatively low marginalproductivity of public capital so that an increase in public investment would mainly crowd out private investmentthrough reduced available savings and higher interest rates (3)
(1) In existing studies on the relation between public investment and private investment using Granger causality tests results depend on the particular coun-tries and periods analysed and on the specific methodology followed (for example Aschauer 1989c Eremburg 1993) Flores de Frutos et al 1998 findevidence of crowding-in while Monadjemi et al 1998 Lightart 2000 and Voss 2001 find support of the crowing-out hypothesis) Among the existingstudies there is none analysing systematically all EU countries
(2) The result for Spain is consistent with those found in previous studies (for example Flores de Frutos et al 1998) while Lightart (2000) finds noevidence of crowding-in in the case of Portugal
(3) This interpretation for the Swedish case is consistent with existing work estimating that the stock of public capital in Sweden is above the optimal one(Berndt and Hanson 1991)
∆itp α β∆it 1ndash
p γ∆it 1ndashG εt+ + +=
∆itp ∆it
G
εt
116
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
investment rather than a detailed examinations of spe-cific provisions of the EU framework for budgetary sur-veillance (1)
Graph III6 shows that in EU countries public investmentand interest expenditure followed quite opposite tenden-cies during the past decades It shows that the share ofinterest expenditure reached its maximum in the mid-1990s and declined in subsequent years Public invest-ment reached its minimum level around 1997 and stayedbroadly constant afterwards Table III3 presents evi-dence consistent with the hypothesis that fiscalconsolidations induced by high debt levels and the needto satisfy the Maastricht criteria coincided with rela-tively larger cuts in public investment For each Euro-pean country the average annual change occurred ingovernment revenues total primary expenditures andpublic investment (shares on GDP) during consolidationperiods is reported With the exception of Greece andPortugal public investment in all countries dropped dur-ing phases of consolidation and in general did so moremarkedly than total primary expenditures
Table III3 also reports the average annual change ingovernment revenues total primary expenditures andpublic investment for the EU-14 aggregate separately forthe overall period for consolidations periods only andfor consolidation periods occuring after 1985 only Pub-lic investment cuts during consolidations occurredthroughout the whole period but were on average deeperduring consolidations that took place after 1985 whichwere concentrated on the expenditure side
Graph III7 reports the average annual change in publicinvestment shares in each EU country and in the EUaggregate during the 1990s distinguishing several sub-periods The first sub-period (1991ndash93) coincides withphase I of EMU The second sub-period (1994ndash98) cor-responds to phase II of EMU It is in those years that theMaastricht calendar for monetary unification exercisedthe strongest pressure on governments urging them tokeep their budget deficits below 3 of GDP as a condi-tion for entering EMU Between 1994 and 1998 publicinvestment in the EU registered the largest drop How-ever it can be noted that during this period publicinvestment also fell in all the countries that chose not tojoin the single currency (2) The third sub-period (1999ndash2002) coincides with the years of operation of the euroIn spite of the fact that in this period the Maastricht
Table III2
Public and private investment Granger causality tests
Dep Variable N obs
Adj R squared
BE 0213(0189)
ndash 0114(0137)
31 00385
DK 0194(0183)
0101(0161)
30 ndash 00118
DE 0473 (2)(01749
ndash 0024(0117)
31 0159
EL 0063(0204)
006(0173)
31 ndash 00577
ES 0491 (3)(0152)
0158 (1)(0081)
31 0271
FR 0449 (2)(0165)
0012(0141)
31 0155
IE 025(0191)
ndash 0084(0155)
29 ndash 0063
IT 0287(0179)
ndash 0076(0106)
31 0036
LU ndash 0153(0199)
0122(0264)
31 ndash 0047
NL 0301(0177)
ndash 0137(0149)
31 0066
AT ndash 0002(0178)
ndash 0026(0121)
31 ndash 0069
PT 039 (2)(016)
0262 (1)(0138)
31 0227
FI 06 (3)(0154)
005(0182)
31 032
SE 0438 (2)(0163)
ndash 0246(0176)
31 0223
UK 0377 (2)(0161)
ndash 0144 (2)(007)
31 0234
NB Estimation method OLS constant term included (3) (2) (1) denoterespectively significance at 1 5 10 level Coefficient standard devia-tions are reported in parentheses is the difference between the (log of)real gross fixed capital formation of the private sector at time t and at timet-1 is the difference between the (log of) real gross fixed capital for-mation of the general government at time t and at time t-1 The deflatorused is that of gross fixed capital formation total economy
Data source AMECO database
yen1part The EU framework for budgetary surveillance is presented in previousissues of this report See European Commission (2000 2001 and 2002a)
∆itp
∆it 1ndashp ∆it 1ndash
G
∆itp
∆itG
yen2part While in Denmark and Sweden this reduction was not particularly strongthe UK is the European country registering the largest drop in publicinvestment in this period The reduction of UK public investment in thisperiod concerned mostly central government investment in health educa-tion and defence (Clarke Elsby and Love 2001)
117
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Graph III6 Interest expenditure and public investment EU-15 1970ndash2002
Graph III7 Public invest changes in the 1990s (average annual changes in GDP shares) (1)
(1) Gross fixed capital formation general govenment
0
1
2
3
4
5
6
Interest expenditure general government ( of GDP at current market prices)Gross fixed capital formation general government ( of GDP at current market prices)
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
ndash 15
ndash 10
ndash 5
0
5
10
15
EUndash 15
BE DK DE
EL
ES
FR
IE
IT
LU
NL
AT
PT FI
SE UK
1991ndash93 1994ndash98 1999ndash2002
118
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
requirements for fiscal discipline (integrated with the pro-visions contained in the Stability and Growth Pact) contin-ued to operate the share of public investment in GDPstopped falling in the EU aggregate (1) In several countriespublic investment shares actually rose (Ireland especially)
On the basis of the data it appears that public investmentin EU Member States has been cut especially during theperiods of fiscal consolidation occurring in the late1980s and in the 1990s These dynamics may be partlyexplained by the fact that in periods of financial distresspublic investment is often more likely to be cut than cur-
rent public expenditure since the former is made of fixedexpenditures which can be delayed or moved to futureperiods with relatively low political costs The evidencealso seems to suggest that the effects of the fiscal disci-pline provisions of EMU were quite different before andafter the introduction of the euro The years precedingthe introduction of the euro coincided with a particularlystrong reduction in public investment shares in mostcountries Conversely and also thanks to the progressmade in reducing interest expenditure the introductionof the euro coincided with a halt in the downward trendin public investment that characterised the EU since theearly 1970s Interestingly the trends are similar in coun-tries that do not form part of the euro zone
Results of regression analysis presented in Table III4(see also Box III4) suggest that the requirements of fis-cal discipline associated with EMU have produced botha direct and an indirect effect on public investment withopposite signs
Table III3
The composition of fiscal consolidations general government (1970ndash2002)
Total revenues
Total primary
expenditure
Gross fixed capital
formation
EU-14 average
Overall period 09 0977 ndash 098
Consolidation periods
152 ndash 086 ndash 413
Consolidation periods after 1985
059 ndash 148 ndash 462
Individual countries during consolidation periods
BE 059 ndash 075 ndash 600
DK 078 ndash 214 ndash 234
DE 089 ndash 006 ndash 683
EL 260 ndash 045 373
ES 050 ndash 100 ndash 630
FR 194 091 170
IE 291 ndash 217 ndash 532
IT 244 ndash 031 ndash 478
NL 246 ndash 140 ndash 235
AT 161 050 ndash 800
PT 053 112 145
FI 014 ndash 243 ndash 428
SE 173 ndash 290 ndash 446
UK 210 ndash 095 ndash 1070
NB Figures refer to average annual changes in shares on GDP Cross-coun-try averages are unweighted The years of fiscal consolidation in each EUcountry are those reported in European Commission (2000) p 20 for the1990s while for the remaining period are those reported in IMF (1996)p 57
Source Commission services
yen1part For the EU aggregate public investment shares are constant at 23 GDPpercentage points for the 1999ndash2001 period and equal to 22 percentagepoints in 2002
Table III4
The determinants of public investment in the EU Regression analysis (EU-15 1970ndash2002)
RPCGDP(t-1) ndash 0112 (3)(0011)
ndash 0145 (3)(0013)
ndash 0172 (3)(0014)
RLIR(t-1) ndash 0036 (2)(0014)
ndash 0022(0014)
ndash 0021(0014)
CAB(t-1) ndash 0038 (3)(0013)
ndash 0057 (3)(0013)
ndash 0064 (3)(0013)
DEBT(t-1) ndash 0021 (3)(0002)
0025 (3)(0002)
ndash 0024 (3)(0002)
TOTGREV(t-1) 0051(0012)
0051 (3)(0012)
0055 (3)(0012)
EMU 0566 (3)(0122)
0881 (3)(0149)
EMUCAB(t-1) 0115 (3)(0032)
R squared within groups
044 047 048
NB Dependent variable Gross fixed capital formation general government( of GDP)
Estimation method fixed effects panel regressionCountry effects coefficients are not reported Hausman tests rejected random effects in linear panel regressions(3) (2) (1) denote respectively significance at 1 5 and 10 confidence Coeffi-cient standard deviations are reported in parenthesesRPCGDP(t-1) Real per capita GDP lagged one year RLIR(t-1) Real interest rate on 10 year government bonds lagged one yearCAB(t-1) Cyclically-adjusted budget balance of GDP) lagged one yearDEBT(t-1)Gross nominal public debt of GDP) lagged one yearTOTGREV(t-1) Total revenue general government of GDP) lagged one yearEMU dummy variable equal to 1 for years following 1993 and for EMU countries
Source Commission services
119
P u b l i c f i n a n c e s i n E M U 2 0 0 3
bull on the one hand EMU is associated with higher pub-lic investment shares keeping other factors constantThis direct effect may be associated to changed gov-ernmentsrsquo expectations concerning the state of theirpublic finances induced by the framework for fiscalstability The expectation of lower future deficits anddebts may have induced governments to increaseexpenditures devoted to public investment
bull on the other hand EMU appears to have reduced pub-lic investment indirectly by inducing a negative effectof budget deficits on public investment This mayindicate that in order to qualify for the adoption of theeuro countries running relatively large budget deficitshad to reduce their public investment expenditures torespect the EMU requirements of fiscal discipline
The overall effect of monetary unification on publicinvestment expenditures in EU countries is therefore notclear-cut and may be different depending on the countryconsidered While the net effect on countries running rel-atively large budget deficits in the 1990s may have beennegative public investments in countries with relativelylow deficits and debt levels may have instead received astimulus (1)
yen1part Gali and Perotti (2003) in their empirical analysis report evidence consist-ent with these findings They similarly do not find support for the hypothe-sis that the advent of EMU reduced public investment rates in EUcountries Their analysis however shows that with EMU public investmenthas become more pro-cyclical
Box III4 The determinants of public investment in the EU an empirical analysis
(Continued on the next page)
The aim of this analysis is to investigate the main factors affecting the evolution of public investment across EU countriesin the past decades with a special focus on the impact of the process of monetary unification To that end panel data regres-sions have been performed The dataset includes all Member States and covers the period 1970ndash2002 (1) The data sourceis the AMECO database
The dependent variable is the share of public investment (gross fixed capital formation general government) on GDP atcurrent market prices As for explanatory variables the real GDP per capita (RPCGDP) captures the different role that pub-lic investment has in different stages of countriesrsquo development Public investment is likely to exert a more prominent rolein countries at intermediate stages of development Since in the European context all countries are at an advanced or inter-mediate stage of development in the period considered the expected effect of RPCGDP on public investment is negative
To take into account the opportunity cost of funds used up in public investment real long-term interest rates (RLIR) areadded in the equation to be estimated This variable also hase an effect on public investment through the current andexpected cost of public debt The expected sign for the coefficient of RLIR is thus negative The fiscal stance is capturedby the cyclically-adjusted budget balance (CAB) It has been found in previous empirical analysis (for example Sturm etal 1996) that budget deficits can be negatively associated with public investment expenditure so that the expected signfor the coefficient of CAB is thus positive (in CAB deficits are negative entries while surpluses are positive entries) Afurther variable is the stock of gross public debt as a share of GDP (DEBT) Other things being equal the larger the stockof accumulated debt the higher the flow of interest payments to be paid by governments Hence the expected sign for thecoefficient of DEBT is negative
To account for the cross-country variation in the scope of government intervention and for its evolution in time the shareof government revenues on GDP (TOTGREV) is included in the equation A higher value for TOTGREV is an indicationof a greater role of the public sector in the economy The expected sign for the coefficient of TOTGREV is positive sincea higher value for TOTGREV is likely to be associated with a higher share of resources devoted to public expenditureincluding public investment To overcome endogeneity (reverse causation) problems all the above-mentioned variables
(1) Due to missing observations for some variable in particular years and countries the total number of observations used in regressions is somewhat lowerthan the maximum of 15 times 33 = 495
120
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
Box III4 (continued)
have been used with a one-year lag Finally a dummy variable (EMU) equal to one in all EMU countries in all years fol-lowing the start of phase II of EMU (1994) is included to have a measure of the effect played by the fiscal constraints ofmonetary unification on public investment All the remaining idiosyncratic factors that may explain differences in publicinvestment expenditures across countries (for example publicprivate ownership of infrastructures etc) are captured bycountry effects (1)
Results are displayed in Table III4 The first specification tested excludes the EMU dummy All variables have theexpected sign When the EMU dummy is included as a constant term its coefficient is significantly positive and close to05 The interpretation is the following other things being equal EMU is associated with larger public investment by abouthalf a percentage point of GDP Keeping unchanged per capita GDP public debt deficit government revenues and interestrates a country would devote a larger fraction of resources to public investment Under this specification the assumptionis that EMU only has a direct effect on public investment It may be argued however that EMU also produces an indirecteffect on public investment by changing the impact of budget deficits To test for this hypothesis the EMU dummy hasbeen interacted (multiplied) with the CAB variable (EMUCAB) Under this specification the effect on EMU is both directand indirect The impact of CAB outside EMU is captured by the coefficient of the CAB variable when not interacted theone on EMU is given by the sum of this coefficient and that of the CAB variable interacted with the EMU dummy Resultsshow that while outside EMU the impact of CAB is negative in EMU it is significantly positive In fact summing up thecoefficient of CAB and that of EMUCAB yields a positive value It is also to note that EMU still plays a significant directeffect on public investment represented by a significantly positive coefficient for the constant EMU dummy
The overall results can be interpreted as follows The requirements of macroeconomic convergence and fiscal disciplineaccompanying the process of monetary unification appear to have produced both a direct and an indirect effect on publicinvestment On the one hand EMU is associated with a shift of resources towards public investment keeping other factorsconstant This direct effect may be due to reduced interest expenditure but also to changed government expectations con-cerning the state of their public finances induced by the EMU fiscal framework The expectation of lower future deficitsand debts may have induced governments to devote a higher amount of resources to public investment On the other handmonetary unification induced a negative effect of budget deficits on public investment Starting with phase II of EMU therequirement of fiscal discipline was strengthened by specific time deadlines and started to be perceived as binding thistranslated into countries running larger budget deficits making bigger cuts in public investment
(1) The regressions results presented in Table III4 hold qualitatively unchanged under alternative specifications The exclusion of the variable TOGREV(which due to its correlation with CAB leads to multicollinearity problems) does not alter significantly the coefficients of the remaining variables A listof additional explanatory variables affecting the expected benefits of public investment (the net stock of capital over GDP private investment as a shareof GDP) and representing cyclical factors (inflation rate growth rate of real GDP) have also been considered but their coefficients resulted in beinginsignificantly different from zero in all specifications Specifications including a time trend have also been tested In such specifications the time trendturns out to have a significant negative effect on public investment while real per capita GDP and the debt variable result in being insignificant
121
5 Catering for public investment needs in the Stability and Growth Pact
51 How public investment is treated under the existing Treaty and SGP rules
The Treaty obliges countries to avoid excessive deficitpositions (defined as general government deficit below areference value of 3 of GDP) and the SGP requirescountries to achieve budget positions lsquoclose to balance orin surplusrsquo These requirements imply that most publicexpenditure including those in investment projectshave to be funded from current revenues
While the existing framework provides for no specialtreatment of public investment as regards the definitionof the budget balance (and consequently in terms ofthe budgetary objectives which Member States mustrespect) the framework for budgetary surveillance doeshowever take account of public investment as part of theassessment of Member Statesrsquo fiscal position For exam-ple Member States are required to report public invest-ment levels and plans in their annual updates to stabilityand convergence programmes The Council has shownsome flexibility in interpreting compliance with thelsquoclose to balance or in surplusrsquo requirement to reflectsignificant planned increases in public investment pro-grammes (for example see recent Council opinions onthe stability programme of Ireland and on the conver-gence programme of the UK)
Moreover public investment levels are taken intoaccount in the excessive deficit procedure As describedin Part II2 the Commission activates the EDP by pre-paring a report if the actual or planned deficit goes above3 of GDP Article 104(3) states that when preparingits report the Commission lsquohellipshall also take intoaccount whether the government deficit exceeds govern-ment investment expenditurehelliprsquo
In brief public investment does feature in the existingframework for budgetary surveillance and in particularconcerning the assessment of the budgetary position ofMember States This chapter considers whether there isscope for a more specific treatment of public investmentexpenditures in the EUrsquos framework for budgetary sur-veillance Two specific issues are examined
First it has been suggested by several scholars and pol-icy makers to amend or reinterpret the EU legislation insuch a way as to exclude investment expenditures fromthe deficit ceilings relevant to the EDP that is to intro-duce a lsquogolden rulersquo (1) Section 52 considers the meritsand feasibility of applying a golden rule for publicinvestment in the EUrsquos budgetary rules
Second private sector corporations are increasinglyinvolved in the building and operating of public projectsin EU countries (2) Section 53 examines the rationalefor publicndashprivate partnerships (PPPs) and how these arehandled within the existing framework for budgetarysurveillance
The issue of a more specific and flexible treatment ofpublic investment within the EU framework for budget-ary surveillance is timely In the communicationlsquoStrengthening the coordination of budgetary policiesrsquoadopted on November 2002 (3) the Commission pro-posed to introduce a more flexible application of thelsquoclose to balance or in surplusrsquo requirement to betterachieve the goals of the Lisbon strategy Point 5 (iv) ofthe communication states that there is a need to lsquohellipcater
yen1part For the academic debate on this point see for instance Balassone andFranco (2000b) Blanchard and Giavazzi (2002) Buiter and Grafe (2002)and Buti Effijnger and Franco (2002)
yen2part See for instance European Commission (2003) on alternative proposals tofinance trans-European transport networks including partnerships betweengovernments and private operators
yen3part European Commission (2002c) See also Part II2 of this report
122
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
for the intertemporal budgetary impact of large structuralreforms (such as productive investment or tax reforms)that raise employment or growth potential in line withthe Lisbon strategy helliprsquo The Commission made clearthat this should not put in jeopardy the core budgetarycommitment to sound public finances and thereforestated that lsquosmall temporary deteriorationsrsquo in underly-ing budgets position can only apply to countries alreadyhaving made substantial progress towards the lsquoclose tobalance or in surplusrsquo requirement and whose debt isbelow the 60 of GDP In other words the Commissiondid not propose a golden rule per se but rather that on atemporary basis a planned increase in public investmentcould provide grounds for a flexible interpretation of thelsquoclose to balance or in surplusrsquo requirement providedthere was an adequate safety margin ensuring respect ofthe 3 of GDP reference value for deficits
52 Public investment and the golden rule
521 A rationale for the golden rule
The golden rule consists of excluding investment spend-ing from the computation of the deficit measures whichare considered for the definition of fiscal discipline tar-gets This is not a new idea and was debated already inthe 1930s (1) A number of countries (for example Bel-gium the Netherlands and Sweden) adopted this ruleduring the 1950s and 1960s but subsequently aban-doned it The golden rule debate has been revivedrecently partly as a consequence of decisions taken bysome governments (UK Australia and New Zealand) toallow for public borrowing to finance public investment
The idea behind the golden rule is relatively simple Aswith private companies a government should notattribute entirely the full cost of a project that is likely togenerate gains for a long time period to a single yearrsquosaccounts Since public investments normally implyreturns over several years (and in some cases over a verylong time horizon) the cost should be distributed overseveral years as the returns materialise
A proper working of golden rule provisions requiresadopting a dual public budget one budget should onlyinclude current operations a separate budget should bedevoted to capital operations (2) Gross investments would
enter only in the asset side of the capital budget while inthe liabilities side of the capital budget would be regis-tered the cumulative amortisation of the public capitalstock and the deficit of the current budget As for thecurrent budget it would be affected only by the amorti-sation of the capital stock which would be recorded onthe expenditure side (3) Since the balance of the currentaccount equals general government net lendingborrow-ing after subtracting net public investment for countriesadopting a dual-budget system targets for the balance ofthe current budget are equivalent to standard budgetarytargets amended by the golden rule A dual budgetsystem would have the added advantage of improvinginformation on the contribution of public investment tothe net worth of the public sector (see for exampleFottinger 2000)
Several arguments have been advanced in favour ofadopting a golden rule First in the presence of deficitlimits socially desirable public investment projects maynot be undertaken This may happen for several reasons
bull Financing investment from current revenues mayclash with consumption-smoothing objectives ofpolicy authorities If policy-makers are inclined toavoid large variations of consumption possibilitiesover time they may decide not to carry out poten-tially profitable investment projects if this implies asubstantial reduction in current disposable incomeWhen growth prospects and public investmentreturns are high while public borrowing is not toocostly constraints that impose financing all publicexpenses through current revenues may be counter-productive In such conditions in fact profitableinvestments may be rejected under a balancedbudget rule because the additional gains generatedby public investment projects will only materialisein the future mdash when income is expected to be highmdash while current consumption would be furtherreduced by higher taxation Under such conditionsamending the balanced budget constraint by agolden rule may increase profitable investmentssince deficit finance permits current consumptionnot to be compressed
bull A more subtle motive for under-investment arisingfrom deficit ceilings builds on the analysis of
yen1part See for example Musgrave (1939)yen2part As it is currently done in the UK and in the past by Belgium the Nether-
lands and Sweden
yen3part So by construction the balance of the capital budget equals net investmentminus the balance of the current budget
123
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Tabellini and Alesina (1990) who show that govern-ments may have a tendency to run large deficits forstrategic purposes (1) In this setting Peletier Durand Swank (1999) analyse public investment expen-ditures and show that in the presence of deficit ceil-ings governments may be induced to under-investfor strategic reasons The reason is that by reducinginvestment current policy-makers can assure them-selves of a high level of current expenditure of thepreferred type reducing at the same time the amountof resources that will accrue to future governmentsfrom the returns on investment Under such a frame-work a golden rule that excludes investment expend-iture from the deficit ceiling could help to avoid thetendency towards strategic underinvestment
A second reason why potentially desirable investmentprojects may not be carried out in the presence of deficitlimits is the existence of institutional or political constraintsIt has been argued (for example Oxley and Martin (1991)Lane 2002) that cutting public investment is often polit-ically easier to do than to achieve reductions in currentexpenditure or raising taxes Under such circumstancesinvestment may not take place simply as a result offinance constraints of institutional and political origin
A third rationale in favour of a golden rule concernsintergenerational equity As emphasised for instance inBalassone and Franco (2001) the adoption of deficitceilings that do not distinguish between current andinvestment expenditure may redistribute income awayfrom current generations due to the creation of a lsquodoubleburdenrsquo Current generations continue to pay back thedebt accumulated to finance investment undertaken bypast generations (in the form of taxes levied on theirincomes) However budget rules prohibiting deficitfinancing would require them to also pay entirely fornew investments carried out by themselves without thepossibility of deferring their cost to future generationsthrough debt The double burden issue is a transitory onethat continues until all the debt of previous generationsis repaid Once achieved all future generations will onlyhave to pay for their current investment without inherit-ing debt used to finance past investment However thetransition may be very long penalise the generations
alive during the shift in the financing regime and lead toa bias towards excessively low investment levels for aprolonged period
522 Limitations and drawbacks
In spite of the potential benefits of a golden rule thereare also considerable drawbacks and implementationproblems
A first set of basic problems with a golden rule relatesto its desirability effectiveness and relevance As illus-trated in Section 121 there are no strong theoretical orempirical arguments in favour of the view that govern-ments undertake too few public investments If the proc-ess of public decision-making produces a bias towardsexcessive rather than insufficient public investmentthen the adoption of a golden rule may prove counter-productive (2)
A further substantial drawback of the golden rule has todo with possible distortions in resource allocation Theidea of the golden rule is that of distributing over timethe costs of public projects that are likely to generateincome streams across several years This is a principlethat is normally followed in private sector accountingHowever the analogy is very limited since there aremajor differences between the concepts of economicreturns for the public and the private sector While forprivate firms economic returns of investment projectsmust translate into financial returns at least in the longrun this is not necessarily the case for the public sector(such as for instance concerning investment projectswith environmental purposes) Moreover the adoptionof a golden rule is likely to produce an effect on the com-position of productive public expenditure Finance con-straints would be released on investment in fixed assetswith a physical component (normally covered by thedefinitions of public investment from national accountstatistics) while investments in human capital mayremain constrained by deficit ceilings This may lead toa distortion in the allocation of resources in favour of thephysical
A sound application of the golden rule would require thatit should be investment net of amortisation which isexcluded from the computation of the deficit Howeverimplementation problems arise especially with the deter-mination of net investment The calculation of amortisa-yen1part When current governments have a preference over a certain type of current
expenditure but are uncertain about the preferences of future governmentsa bias towards too-high deficits may emerge since by running deficits pol-icy authorities will influence the composition of current expenditure andlimit the spending possibilities of their successors yen2part See for example Fottinger (2001) for a formal development of this argument
124
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
tion is subject to technical difficulties and ambiguitiesMoreover the difficulties with the computation of netinvestment may induce opportunistic accounting prac-tices with the consequence of an overestimation ofamortisation rates
523 Practical experiences
Though not a very common practice some form ofgolden rule has been operational in some countries orsub-national jurisdictions In the European context thecountries currently operating some form of a golden ruleare Germany and the UK In both cases the rule isdesigned in such a way that budget deficits should not behigher than some definition of public investment but thecharacteristics of the German and the UK golden rule arequite different (1)
In the German legislation Article 115 of the Constitutionstates that the annual budget deficit of the general govern-ment cannot be higher than gross fixed capital formationin the federal budget Exceptions are permitted to avoidlsquodisturbances to the overall economic equilibriumrsquo Acrucial feature of the German golden rule is that the targetis defined in terms of gross public investment not netinvestment as would be preferable in principle
In the UK since the institution of the Code for FiscalStability in 1997 the general government and thebroader public sector are allowed to borrow only to fundinvestment while current spending must be fullyfinanced from current revenues The compilation of sep-arate current and capital budgets facilitates the distinc-tion between gross and net investment in nationalaccounts Consistently the UK golden rule applies to netinvestment It can also be noted that the UK golden ruleis applied over the budget cycle so that a transitorydecline in revenues would not affect medium-termexpenditure targets Finally it is to be remarked that thegolden rule in the UK is complemented by a rule aimedat guaranteeing that leaving net investment out of defi-cits is not incompatible with sustainable public financesThis is the so-called lsquosustainable investment rulersquo whichrequires the debt-to-GDP ratio to be maintained at belowthe prudential 40 ceiling
Is the golden rule effective in stimulating public invest-ment expenditures by reducing finance constraints Inspite of the fact that a number of countries have experi-enced alternative forms of the golden rule very few sys-tematic analysis of the effects of such rules on publicinvestment exist One notable exception is the analysisby Poterba (1995) who studies the impact of the differentbudgetary rules across states in the US The analysisallows to identify the states that make a budgetary dis-tinction between capital and current expenditures andthose that use pay-as-you-go constraints to finance pub-lic projects The results show that on average separatecapital budgets are associated with higher capital expen-ditures
The cross-section dimension used in the analysis by Pot-erba (1995) for US states is lost when analysing EU coun-tries since only Germany and the UK adopted a goldenrule in recent years By simply looking at the evolution ofpublic investment figures one notes that in spite of thepresence of a golden rule Germany is among the EUcountries in which public investment has been fallingmore markedly in past decades (see Graph III7) As far asthe UK is concerned the evolution of net public invest-ment after the introduction of the golden rule does notseem so far very different from that before its introduction(see the section on the UK in part VI15 of this report)
524 Why a golden rule would not be desirable for EMU
Various proposals have been made to introduce someform of a golden rule into the EUrsquos fiscal rules that isexclude investment expenditures from the measure ofbudget balance This would imply shifting from budget-ary targets and ceilings common to all countries andfixed ex ante in numerical terms to country-specificceilings and targets related to some form of investmentexpenditure planned by national governments The pre-cise effect of such a move would depend on the way thea golden rule is designed and implemented For examplea golden rule could concern the upper ceiling for nomi-nal deficits in the EDP (as in the German golden rule)andor the medium-term target of lsquoclose to balance or insurplusrsquo (as in the UK golden rule) Box III5 examineshow the EMUrsquos fiscal architecture would be affected if agolden rule was introduced along the lines of Germanand UK approaches
Overall and building upon the drawbacks and limita-tions identified in Section 522 above there are severalarguments which suggest that the adoption of the golden
yen1part Note that the working of a golden rule in Germany and the UK is not incon-sistent with the respect of the budgetary requirements of the Treaty and theSGP In both countries deficits are required to be below ceilings defined interms of investment expenditures These ceilings will be binding only ifmore stringent than the Maastricht 3 Moreover these ceilings are notinconsistent with the medium-term goal of lsquoclose to balance or in surplusrsquo
125
P u b l i c f i n a n c e s i n E M U 2 0 0 3
rule in the EMU framework would largely outweigh thepossible benefits as follows
First the likely impact of a golden rule on actual levelsof public investment and its share in total public spend-ing is questionable For example a golden rule whichallows deducting net investments from medium-termbudgetary targets (as in the UK) would probably onlyhave a limited impact An indirect indication of the orderof magnitude can be inferred from past values of netinvestment in European countries During the 1980s and1990s average annual net public investment rates in theEU-15 area were well below 2 of GDP with valuesaround 1 of GDP for countries like Belgium the Neth-erlands the UK and Sweden while the average rate wasnegative for Denmark (see Graph III5)
Second the introduction of a golden rule could undermineefforts to improve the sustainability of public finances if(either because of the way the rule is designed or imple-
mented) the medium-term target for deficits ends upbeing increased by an amount equal to planned grossinvestment rates Simulations show that if governmentsrun constant deficit levels of 2 of GDP over the period2005ndash50 then debt levels would be some 45 percentagepoints of GDP higher in 2050 than what would resultfrom running a balanced budget position over the pro-jection period The difference would amount to some90 percentage points were governments to run constantdeficits of 4 of GDP (equivalent to the gross invest-ment ratio in some Member States) The impact of defi-cits (either temporary or permanent) on the sustainabilityof public finances depends on many factors not least theprojected increase in age-related expenditures in comingdecades As pointed out in Part I4 debt reduction has akey role to play in the strategies of many countries tomeet the costs of an ageing population and the risks ofan unsustainable public finance position is greatlyincreased by a failure to respect the lsquoclose to balance orin surplusrsquo requirement of the SGP
Box III5 How would the introduction of a golden rule modify the fiscal architecture of EMU
(Continued on the next page)
The impact of the introduction of a golden rule on the EMU fiscal architecture depends crucially on how the golden ruleis designed The current requirements of the Treaty and the Stability Pact are (i) nominal budget balances below 3 ofGDP at each year t (ii) a budget position lsquoclose to balance or in surplusrsquo Graphically the present state of the EMU fiscalarchitecture can be described as in Graph III8a) The nominal budget balance and the cyclically-adjusted budgets are plot-ted as functions of the output gap Assuming a constant sensitivity of the budget deficit with respect to the output gap thenominal budget balance can be represented by a linear function of the output gap The lsquoclose to balancersquo requirement con-strains the CAB to be non-negative This constraint is represented by the continuous horizontal line in correspondence witha value of zero Deficits must not breach the 3 reference value set by the Treaty this ceiling is represented by the dottedhorizontal line
The application of a golden rule of the German type would prescribe nominal deficits in each year to be lower than theprogrammed gross investmentGDP share Compared with the current situation the only change that would be a revisionof the value for the nominal deficit ceiling that is the SGP objective of lsquoclose to balance or in surplusrsquo would beunchanged As illustrated in Graph III8b) the upper ceiling for a deficit would change from 3 to the planned grossinvestment share at time t denoted by It
Amending the EMU fiscal architecture with a UK-type golden rule would instead change the medium-term target but leavethe 3 of GDP reference value unchanged As illustrated in Graph III8c) the revised lsquoclose to balance or in surplusrsquorequirement would require the CAB to be at most equal to the average net investmentGDP rates planned over the cycle(denoted by NIAV )
It has also been proposed (for example by Modigliani et al 1998 Blanchard and Giavazzi 2002) to modify the EMUfiscal framework by excluding net investments from the definition of deficits both nominal and CAB Such a reform wouldresult in a revision of both the upper ceiling for deficits and the medium-term targets Both would increase by the amountof planned net investment rates In other words the medium-term objective would be the same as in the UK proposal (NIAV)whereas the upper ceiling would be equal to 3 +NIAV
126
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
Box III5 (continued)
Graph III8 Nominal budgets structural budgets the SGP and the golden rule
CAB (underlying budget balance)
Positive outputgap
Negative outputgap
Positive outputgap
Negative outputgap
Positive outputgap
Negative outputgap
deficit
surplus
Maastricht 3 upperceiling
0
(SGP medium-run target)
deficit
surplus
0
(SGP medium-run target)
deficit
surplus
0
(revised medium-run target)
3
NIAV
It
Output gap
Nominal balance
Gross investmentGDPprogrammed for year t(revised upper ceiling)
Maastricht 3 upper ceiling
0
0
0
Nominal balance
Nominal balance
CAB (underlying budget balance)
CAB (underlying budget balance)
(a) Current SGP
(b) SGP amended by a German-type golden rule
(c) SGP amended by a UK-type golden rule
Output gap
Output gap
3
127
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Third there would be specific implementation problemsin implementing a golden role in a multilateral settingThe crucial distinction is the one between net and grossinvestment As illustrated in Section 321 an applica-tion of the golden rule in accordance with its economicrationale would require using the concept of net invest-ment However EU countries normally do not dispose ofa dual-budget accounting system which would instead berequired for an efficient application of the golden ruleapplied to net investment As stressed previously the cal-culation of amortisation is a complex process whichrequires estimating the economic value of each publiccapital item and its expected life period These difficul-ties would become particularly relevant in a multilateralframework Amortisation rates should be evaluated by allcountries following common methodologies but oppor-tunistic accounting practices may be difficult to avoidwith governments attempting to underestimate amortisa-tion rates Finally a golden rule applied to net investmentin a multilateral framework would discriminate againstthe countries with a larger stock of public capital For agiven amount of gross investment net investment forthese countries will normally be lower since amortisa-tion applies to a larger public capital stock Howevercross-country differences in the magnitude of the publiccapital stock may simply relate to a different allocation ofownership of facilities and infrastructures between thepublic and the private sector so that a larger public capi-tal stock does not necessarily imply a weaker need forpublic investment
53 Publicndashprivate partnerships
531 Definition taxonomy and recent experiences
The involvement of private sector corporations to buildand operate public projects has become an increasinglywidespread practice in EU countries Following theexperience of the UK private finance initiative the con-struction and operations of infrastructures such as roadsbridges or airports are made jointly in a number of coun-tries by the government and private sector enterprisesthat finance the projects through so-called publicndashpri-vate partnership (PPPs) Currently PPPs cover about15 of the finance provided yearly to publicly spon-sored investment projects in the UK (Spackman 2002)In other European countries such as Germany SpainFrance the Netherlands Portugal Austria and FinlandPPP projects have been recently carried out mainly inthe field of transport infrastructure Almost all the otherEU Member States have planned PPP projects
There is no unambiguous definition of what constitutes aPPP Broadly speaking PPPs concern the transfer to theprivate sector of investment projects that traditionallyhave been executed or financed by the public sector (seefor example Grout 1997) Four elements howeverseem required to qualify PPPs
bull the project should concern the construction or theoperation of physical assets in areas characterised bya strong public function (for example transporturban development security etc) and involve thepublic sector (general government) as the principalpurchaser Although PPPs are especially relevant intransport infrastructure examples of publicndashprivatepartnerships can be found in the provision ofdefence health education and cultural services thebuilding and operation of prisons or the area ofwater and waste management
bull the PPP must involve a corporation outside the gen-eral government (normally a private corporation) asthe principal operator that is the agent that carriesout the project
bull the principal finance of the project should not comefrom public debt but from other sources such as pri-vate bonds
bull by way of the partnership the way the project is exe-cuted must change compared with the alternative ofpure public supply This means that in PPPs the pri-vate operator provides significant inputs in thedesign and conception of the project and bears a rel-evant amount of risk
The main distinction between PPPs and alternative pri-vatisation schemes is that the public sector plays a keyrole as purchaser of services While in the case of pureprivatisation (for example of public utilities) the clientsof the private operator are private users in the case ofinfrastructure building realised though PPPs the govern-ment normally pays for the services to be supplied or hasan influence in their specification What instead distin-guishes PPPs from the traditional public procurementmodel is the origin of the funds to accomplish theproject Instead of relying on government borrowingmost PPPs are financed through bonds issued by the pri-vate operator
128
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
For accounting and performance evaluation purposes itis useful to classify PPP schemes according to the typeof financial operations involved as follows (1)
Sale of services After having funded and executed theproject what the private operator sells to the purchasinggovernment is the flow of services from a capital asset (forexample a road a bridge a prison) In addition to theservices emanating from the use of the assets additionalservices can be provided by the private counterpart for theregular operation of the asset (for example maintenance)The contracts specify on which conditions the governmentcan access these services This is the most frequent case ofPPP and has been extensively used in the financing build-ing and operation of infrastructures such as prisons rail-ways or roads In a sense PPPs can be assimilated to aform of leasing rather than a case of asset purchase
Financial free standing The private operator designsbuilds finances and operates the asset and recovers thecosts through direct charges to users without direct pay-ments from the government The involvement of thepublic sector is in the provision of licenses in securingconformity of the project with public purposes and inregulating the private operator This scheme has beenused especially in projects concerning transport infra-structures such as bridges and highways Compared withthe classic privatisation schemes the government playsa greater role in contributing to the definition of the char-acteristics of the services to be provided by the asset
Joint ventures In this case the finance to build theproject does not come fully from the private operator butis partially provided by the government
Also relevant for accounting and evaluation purposes arethe characteristics of the private operator involved in thePPP and how the contract is designed The private oper-ator can be either an existing firm or a new firm createdon purpose Its activities can either be multiple anddiversified or confined to those of the PPP contractMoreover the operator may be fully private or partici-pated in by the public sector In particular a number ofrecent PPPs are dealt through operators that are publicenterprises not belonging to the general government sec-tor (so called lsquoproject vehiclesrsquo)
Regarding the design of the contract a crucial aspect isthe specification of the modalities with which paymentsare made to the private operator by the government Pay-ments may be in fixed yearly amounts proportional tosome measure of the cost of the asset provided by theoperator (for example in the case of road building pro-portional to the length of the road) or proportional to theeffective flow of services provided by the asset (forexample still in the case of roads proportional to thenumber of vehicles using the road) The way governmentpayments are specified in the contract are crucial indetermining how risks are shared between the govern-ment and the private operator Also key in determiningthe sharing of risk between the public and the privatecounterparts is the possible presence of guarantees bywhich the government backs the bonds issued by the pri-vate operator to finance the project
Another characterising feature of PPP contracts are theirlong-term nature (due to the fact that the revenues for theprivate operator must be distributed over sufficientlylong time horizons to cover up-front costs) and the pos-sible inclusion of clauses by which the government com-mits to buy back the asset after a given number of years
532 The economics of PPPs
Which is the rationale for using PPP schemes to financeand operate public purpose investment
In the policy debate it is often emphasised that PPPs havethe desirable property of putting capital spending outsidegovernment budgets thus easing the effects of externalbudgetary constraints on public investment Though verypopular this argument has little substance First it doesnot address why PPPs should be preferred to alternativeschemes to finance capital formation with public purposesthat do not imply an increase in government borrowing(for example classical privatisation) Second even if theimpact on current budget balances of PPP schemes is mostlikely to be smaller compared with the alternative of purepublic procurement the long-term impact of PPPs on pub-lic finances is to be assessed carefully
The main implication for public finances of choosingPPPs as opposed to traditional public investment is thatof converting up-front fixed expenditures into a streamof future claims In computing the actuarial value of thegovernment commitments of PFI schemes one has toestimate not only the size and distribution of the regularpayments specified in the contract but also the cost ofthe possible buy-back of the asset and the possibility that
yen1part This taxonomy has been proposed by Pollitt (2000) to classify UK privatefinance initiative projects Note that the taxonomy is not fully exclusivesince PPP cases may have characteristics common to more than one of thecases identified
129
P u b l i c f i n a n c e s i n E M U 2 0 0 3
debt guarantees are exercised Comparing the impact ofPPPs schemes on the actuarial value of public financeswith that which would arise from traditional publicinvestment is thus a complex issue which requires a greatdeal of information The argument that PPP schemes arepreferable to publicly funded investment expendituresfrom the viewpoint of long-term public finance sustain-ability is thus not well grounded The distinguishing fea-ture of PPPs is rather that of permitting to smooth out thecost of public investment This in turn may be effectivein releasing finance constraints on public investment inthe presence of formal ceilings on budget deficits (1)
The rationale for the use of PPP schemes is rather that ofmicroeconomic efficiency Even assuming that competi-tive tenders for the selection of private counterparts arefeasible and efficient pure privatisation schemes may notbe optimal when there are reasons that justify a form ofcontrol on the design of the project by the public sectorThis is the case when the project concerns the delivery ofpure public goods (for example a prison) when external-ities are particularly relevant (for example when projectshave a considerable environmental impact) or when thedistributive consequences of the project are a major con-cern (for example the provision of health facilities) Inthose cases regulation mechanisms may not be sufficientto ensure that public objectives are satisfactorily met Thestandard alternatives are direct public provision or publicprocurement through competitive tenders
In many instances public procurements (contracting out)guarantees higher cost-efficiency than direct public provi-sions (2) In both alternatives however it is the public sec-tor that provides the financial funds to carry out the projectand that exercise the control on the design of the assetPPP schemes offer a third alternative In such a case thefinance of the project is provided by the private sector asin privatisation schemes but the public sector plays a rel-evant role as client of the services provided by the assetIn particular PPP contracts may specify that the privateoperator will be remunerated only if the actual supply ofservices is judged to be successful The fact that the objectof PPP contracts is the supply of services rather than theprovision of the asset can make a major difference with
respect to public procurement schemes Specifying andmonitoring the desired characteristics of services is nor-mally easier than specifying and monitoring those ofassets Thus contracts that have as their object the flow ofservices rather than the building of assets help to reducethe incentives that the private supplier may have to cut onquality while preserving the incentives to contain costs(Grout 1997) (3) The microeconomic rationale of PPPschemes is thus that of shaping incentives in such a way asto achieve cost efficiency without compromising publicobjectives relating to the quality and characteristics of theservices provided by the asset (4)
533 Publicndashprivate partnerships and budgetary practices in EMU
Although there are microeconomic reasons that may jus-tify the use of PPP schemes there is the risk that PPPsare increasingly used by EU governments to evade SGPconstraints on public deficits As already pointed out theimpact of PPPs on long-run public finance sustainabilityas an alternative to traditional public investment dependsupon a complex set of factors and should be assessedcase by case In general when resorting to PPP schemesgovernments should conform to the Eurostat guidelineson accounting practices and to a series of transparencyprinciples
Concerning the treatment of PPP schemes in nationalaccounting Eurostat fixes a set of guidelines that nationalstatistical institutes should respect A crucial issue is thatof evaluating the effective sharing of risk and rewardsbetween the general government and the project operatorassociated with the building and operation of the assetAccording to the Eurostat guidelines whenever there areregular payments made by the government to the opera-tor the asset should be recorded in the balance sheets ofthe contracting party that effectively bears most part ofthe risks and rewards form the project
yen1part The conditions under which external constraints on budget deficits caneffectively reduce public investment have been discussed in Section 521
yen2part The reasons are well-known (see for example Domberger and Jensen(1997) for a survey) In particular bureaucracy theories suggest that gov-ernment officials tend to focus on objectives different from that of costminimisation (eg maximising the size of their budget)
yen3part Hart Shleifer and Vishny (1997) develop an incomplete-contracts model ofpublic procurement and show that compared with direct public provisionsprivate operators will in general have higher incentives to keep costs lowbut lower incentives to keep quality high They provide supporting evi-dence in the context of prisons in the US
yen4part The resort to PPPs to finance trans-European transport infrastructure is alsoconsidered by the Commission in its communication COM(2003) 132 final)There the view is expressed that lsquoUse of publicndashprivate partnerships(PPPs) to supplement public financing may be envisaged for some types ofproject However there are still too many unknowns regarding the projectsto be carried out mdash particularly railway and cross-border projects mdash andregarding transport policy choices The private sector has insufficient confi-dence to commit to financing them Moreover PPPs almost always requiremajor public financial support in the form of subsidies and guaranteesrsquo
130
P a r t I I IP u b l i c i n v e s t m e n t a n d i t s i n t e r a c t i o n
w i t h t h e E U rsquo s b u d g e t a r y r u l e s
Evaluating the sharing of risks in PPP schemes has thusmajor implications for the computation of public deficitsand debts If it is the operator that bears most of the risksthen the budget balance of the government will beaffected only by the regular payments made by the gov-ernment If instead most of the risk lies with the govern-ment then public debts and deficits will be affected bythe full cost of the project Given the uncertainties sur-rounding the appropriate evaluation of risk-sharing inPPP schemes it is desirable that national statisticaloffices exchange (among themselves and with Eurostat)detailed information on the criteria used to make suchevaluations It may also be desirable to have the defini-tion of further operational guidelines concerning riskevaluation on the part of Eurostat
It is also relevant that national statistical institutes con-form to transparency principles concerning the recordingof operations giving origin to so-called contingent liabil-
ities Contingent liabilities normally arise when in PPPcontracts governments offer a guarantee to the debtissued by the private operator to finance the project Pub-lic guarantees do not constitute effective government lia-bilities because there is no certainty that they will trans-late into increased debt in the future However this maybe the case if certain contingencies occur that is in thecase of default of the private counterpart Since withpublic guarantees there is no certainty concerning theimpact on public debt they are recognised only undercash accounting if and when the contingent event (thePPP counterpart default) actually occurs and payment ismade However given the possible relevant debt impactof contingent liabilities the inclusion of information(also quantitative when possible) on each provision giv-ing raise to contingent liabilities in supplementary budg-etary documents is recommended in international codesof fiscal transparency (for example the OECD best prac-tices for fiscal transparency)
131
Part IV
Can fiscal consolidations in EMU be expansionary
Summary
While there is a broad consensus among both academicsand policy-makers on the need for fiscal discipline toensure the smooth functioning of EMU and to provideconditions conducive to growth and employment creationconcerns have been expressed that budgetary consolida-tion could have a negative effect on output in the short runThis issue is relevant given the need for several MemberStates to reduce large cyclically-adjusted budget deficitsespecially against the current background of slow eco-nomic growth
According to standard macroeconomic models a restric-tive fiscal stance would result in short-run negativeimpact on aggregate demand and then on output andemployment However the indications of the standardmodels approach have not always been supported by thefacts Growing evidence has been accumulated that thevalue of fiscal multipliers is likely to be quite small andfalling over time Moreover there is evidence that in thecase of fiscal consolidations the effects of fiscal policyon short-run growth may be even opposite to those pre-dicted by traditional macroeconomic models Cases havebeen documented of EU countries in which tax increasesor expenditure cuts have been followed by acceleratedgrowth in the short run Through systematic cross-coun-try analysis new evidence is reported in this partshowing that roughly half of the episodes of fiscalconsolidations undertaken in EU countries in the pastthree decades have been followed by an immediateacceleration in growth
Academic research in the past decade focused on theidentification of the most relevant offsetting factors thatmay explain the emergence of possible expansionaryeffects of fiscal consolidations A number of rationalisa-tions have been provided for what are commonly calledlsquonon-Keynesianrsquo effects of fiscal policy Some of thesefactors concern the impact of fiscal policy on privateconsumption In particular it has been shown that thereduction of budget deficits may lead to an increase inaggregate consumption already in the short run through
wealth and confidence effects In this sense the credibil-ity of consolidations is crucial that is fiscal adjustmentshould be perceived to lead to a permanent increase infuture disposable income streams via reduced taxationConsolidations leading to a substantial improvement ofthe budget balance or starting from situations of highdebt-to-GDP ratios are more likely to affect consumersrsquoexpectations and induce an immediate increase in con-sumption through confidence and wealth effects Withthe awareness of the implications of ageing the effectsof fiscal consolidation on confidence may have becomemore important Fiscal consolidations may also affectaggregate supply via the investment channel They maylead to higher-expected profits and then higher invest-ment by reducing the tax burden on firms and inducingwage moderation In this respect the composition of thefiscal adjustment and the institutional characteristics ofthe labour market may play a major role
Consistently with the predictions of theory the empiricalevidence reported in existing studies shows that the sizeand persistence of the fiscal adjustment (as measured bya sufficient degree of improvement in cyclically-adjustedbudget balances) the composition of adjustment (that isthe extent to which it is achieved through tax increases orexpenditure cuts) and the initial state of public finances(mainly the debt-to-GDP ratio) are relevant for episodesof expansionary consolidations
Interpreting cross-country evidence ex-post is subject toa number of problems above all difficulties in isolatingthe effect of concomitant factors (other than fiscaladjustment) that may have acted on growth Model sim-ulations have therefore also been carried out to investi-gate whether fiscal consolidations can actually produceexpansionary effects The policy experiments performedwith the European Commission QUEST model refer tothe German economy and focus on the composition ofthe adjustment They permit the evaluation of the likelyimpact of fiscal retrenchment obtained either through taxincreases or via cuts in different expenditure items con-
135
P u b l i c f i n a n c e s i n E M U 2 0 0 3
trolling for other factors such as the stance of monetarypolicy
The results of simulations using the QUEST model con-firm that if appropriately designed budgetary consoli-dation can contribute significantly to the goal of Lisbonstrategy in terms of raising output and employment in themedium term Budgetary consolidation have a slightcontractionary effect on output in the short run depend-ing on the composition of the budgetary adjustment
However budgetary consolidation has a positive impacton output in the medium run if it takes place in the formof expenditure retrenchment rather than tax increasesMoreover the effect of budgetary consolidation on out-put could be reinforced and even positive in the shortrun if fiscal consolidation is combined with structuralreform of factor and product markets and accompaniedwith an accommodating monetary stance Indeed budg-etary consolidation often acts as a catalyst for structuralreforms
136
1 Introduction
There is consensus among both academics and policy-makers on the need for fiscal discipline to ensure thesmooth functioning of EMU and provide conditions thatare conducive to growth and employment creation Thisconsensus is reflected in the Treaty requirement to avoidexcessive deficit positions and the goal of the Stabilityand Growth Pact for Member States to achieve andmaintain budget positions of lsquoclose to balance or in sur-plusrsquo With significant cyclically-adjusted budget defi-cits remaining even increasing in several MemberStates (see Part I of this report) the process of budgetaryconsolidations needs to resume if these budgetary goalsare to be achieved
Concerns however have been expressed that budgetaryconsolidation could have a negative effect on output inthe short run and this is particularly relevant against thecurrent background of slow economic growth This sec-tion of the report analyses whether the assertion thatbudgetary consolidation has a negative impact on outputin the short run is always valid or whether it can have apositive effect on output and the conditions under which
this can occur It builds on the work on automatic stabi-lisers and discretionary policy presented in the 2001 and2002 reports on public finances in EMU
Chapter 2 presents a survey of the existing theoreticalexplanations based on consumption-side or investment-side effects through which fiscal consolidation may leadto higher output in the short run
Chapter 3 reviews the empirical evidence from existingstudies on the impact of fiscal consolidations on outputIt then carries out a statistical analysis on the effects ofpast fiscal consolidations in the EU
Chapter 4 presents simulations made using the QUESTmodel investigating the effects on output of varioustypes of fiscal consolidations in a representative EUcountry It examines a variety of consolidation scenarioson both the expenditure and revenue side as well as theimplications of budgetary consolidation through spend-ing cuts being accompanied with an accommodatingmonetary policy response or structural reforms
137
2 Can budgetary consolidations be expansionary What the theory says
21 Budgetary consolidations the standard view
Following the textbook macroeconomics approach a fis-cal consolidation has a negative impact via the multiplieron domestic demand national output and employmentDisposable income and private consumption would benegatively affected by tax increases while a cut in publicspending would directly reduce aggregate demandGiven that the simple form of the multiplier (the standardKahn-Keynes multiplier) depends on the responsivenessof consumption to income its value is by definitionhigher than one (1)
The models generated by the so-called neo-classical syn-thesis (the IS-LM model and its variants) develop theoriginal Keynesian approach to consider also the effectsof various characteristics of the real and money marketson the fiscal multiplier In these models several factorsare likely to interact with the direct effect of fiscal policyon aggregate demand The final impact of the fiscal con-solidation may be therefore smaller implying that thevalue of the multiplier may be below 1 Several factorshave to be considered for the evaluation of fiscal policymultipliers in complex open-economy neo-Keynesianmodels
Real sector substitution effects and investment crowdingout Substitution effects are likely to reduce somewhatthe multiplier some of the goods or services no longerdemanded by the public sector would be demanded by
the private sector or could be directed towards theexport markets The sensitivity of investment spendingto interest rates and income is also relevant A larger sen-sitivity to interest rates would imply a bigger adjustmentof aggregate demand to reduced interest rates (ie wouldflatten the IS curve) leading to a more extensive offset-ting of the initial fiscal contraction By contrast currentincome could affect investment more than proportion-ally (as in the case of multiplier-accelerator models)which may depress investment more markedly in case ofa fiscal consolidation
The functioning of the money market The lower activ-ity implied by the fiscal consolidation would beaccompanied by a reduced demand for money Thiswould lead to a fall in interest rates which would inturn create an incentive for increased investment off-setting part of the effect of the consolidation on outputThis effect crucially depends on the responsiveness ofmoney demand to income and interest rates If thedemand for money is highly sensitive to income and alittle to interest rates (that is the LM function is steep)the reduced activity will have a strong effect on thedemand for money implying a very large adjustmentin interest rates The effect of fiscal policy wouldeventually be small due to the offsetting behaviour ofprivate investment In such a case most of the initialadjustment would be rapidly absorbed via a change ininterest rates
Wealth effects In lsquomodernrsquo neo-Keynesian models con-sumption is determined not only by current disposableincome but also in some measure by current wealth Thelarger the importance of financial wealth in determiningprivate consumption is the more it is likely that a wealtheffect would offset the contraction in public sector
yen1part This value hinges on the typical assumption that output is determined byaggregate demand this results from excess capacity with rigid prices whichdo not adjust (at least in the short run) to the mismatch between demandand supply
138
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
demand (1) The wealth effect could be generated bylower interest rates or in an open economy by fluctua-tions in the exchange rates Lower interest rates increasethe value of nominal fixed-rate debt holdings and ofother assets held by households with the size of its effecton consumption depending on the level of debt holdingsand maturity In an open economy a similar effect couldarise following a depreciation or a devaluation of theexchange rates which would affect the nominal value ofassets denominated in foreign currencies
Openness The degree of openness of the economy andthe exchange rate regime affect the way the external sec-tor responds to the fiscal adjustment (this is evident inthe Mundell-Fleming open economy version of theIS-LM model) The more open an economy is the morethe external sector is likely to react to the change in mon-etary conditions induced by the fiscal adjustment Theextent to which fiscal policy will be crowded-out via thecurrent account crucially depends on the exchange rateregime With flexible exchange rates the currency willtend to depreciate after a fiscal contraction as a result ofcapital outflows The currency depreciation in turn stim-ulates net exports which reduces the effect of the fiscalcontraction on output Conversely with fixed exchangerates an automatic monetary policy response to keep theexchange rate constant increases the effectiveness offiscal policy on output (2)
The interaction between the labour and goods marketsThe degree of price flexibility is a crucial factor in thedetermination of the impact of the fiscal consolidationIn the neo-classical synthesis prices are assumed to berigid in the short run Softening this assumption changesthe effects of a fiscal adjustment The reduction in aggre-gate demand caused by the fiscal contraction will befollowed by an adjustment process whereby pricereductions (or a lower inflation) increase demand in thedirection of a new equilibrium Naturally the magnitudeof this effect will depend on the degree to which pricesare assumed to be less rigid in the short run (3)
In general even if according to the standard models inthe Keynesian tradition fiscal multipliers are expected tobe positive there are several instances that can justifysmall fiscal multipliers also within this approach This isespecially the case for economies with a high degree ofopenness Adjusting the real exchange rate would behelpful if fiscal consolidations occur in countries whoseexchange rate floats the output effects of fiscal policywill be offset by an improvement in the current accountbalance In countries adhering to exchange rate regimesthe negative output effects of fiscal consolidations couldbe offset by accompanying devaluation policies Moreo-ver the value of multipliers will be lower the higher therelevance of wealth in determining consumption asopposed to that of current income which in turn dependsupon the availability of financial instruments to smoothincome and the efficiency of financial marketsIncreased openness and financial wealth may explain thefact that the value of estimated multipliers has been fall-ing in the last decades when the pace of economic inte-gration was accelerating and financial markets devel-oped as a result of liberalisation and institutional andtechnological innovation (4)
22 Non-Keynesian effects of fiscal consolidation
While successive developments of the neo-Keynesianapproach explain why the value of fiscal multipliers isfalling they all assume the multiplier to be positiveThe idea that fiscal policy may have short-run effectsopposite to those predicted by the Keynesian modelwas first suggested by Giavazzi and Pagano (1990)who looking at the fiscal consolidation experiences ofDenmark and Ireland in the mid-1980s documented inboth cases an acceleration in growth just after the gov-ernments put in place measures that drastically reducedbudget deficits
Table IV1 shows deficits and debt ratios as well as GDPgrowth rates in Denmark and Ireland during the cited fis-cal consolidation episodes and in Sweden during theearly 1990s when its deficit and debt rose dramaticallyWhile growth accelerated after the Irish and Danishconsolidations the Swedish fiscal stimulus was followedby an output contraction
yen1part These wealth effects are often referred to also as Pigou effects or real bal-ance effects
yen2part Needless to say for individual euro-area countries the impact of exchangerate movements on the fiscal multiplier is absent or negligible
yen3part While in a closed economy price reductions (or a lower inflation) wouldreduce the output effect of the fiscal consolidation in an open economyprice flexibility softens the effects of the exchange rate regime described inthe previous parafigure in a fixed exchange rate regime there will be morecrowding out (that is a lower fall in output) than with price rigidity and in aflexible exchange rate regime the crowding out is less (that is output wouldfall more) than with sticky prices The price-wage loop will determine thespeed and relevance of this factor
yen4part On this issue see European Commission (2002a) For a recent survey on theestimated value of fiscal multipliers see for instance Hemming Kell andMahfouz (2002)
139
P u b l i c f i n a n c e s i n E M U 2 0 0 3
The possibility that fiscal policy may have non-Keynesianeffects has attracted increasing attention among academ-ics Some studies aimed at further investigating empiri-cally the case of expansionary fiscal consolidations (seeChapter 3 of this part) Some of the research was directedat providing a conceptual framework in which non-Key-nesian effects of fiscal policy could be rationalised
Starting from the 1970s the so-called new classicalmacroeconomics paradigm has challenged many of thefounding assumptions of the standard neo-classical syn-thesis models with major implications for the concep-tual grounds for macroeconomic policy-making Amongthe basic tenets of the so-called new classical macro-economics there is the acknowledgement that agentstake their economic decisions from a forward-lookingperspective and that in so doing they will use rationallyall the information available to them
While according to macroeconomic models in the Keyne-sian tradition consumption is essentially a function of cur-rent income in new classical macroeconomics modelsconsumers are assumed to be forward looking that is tobase their consumption decisions upon the expected futurestreams of income (permanent income) (1) Moreover whileplanning consumption decisions consumers are also in the
position to identify the intertemporal budgetary constraintwhich has to be respected by solvent governments
Concerning the modelling of how expectations areformed macroeconomic models before the neo-classicalsynthesis generally were based upon static or adaptiveexpectations The idea was that agentsrsquo expectationsabout the future could not be too dissimilar from theobserved present This view has been challenged innew classical macroeconomics by the requirement thatexpectations should be rational that is economic agentsshould rationally use available information This meansthat past errors will be considered when formulating newexpectations in a continuous learning process More-over in this context perceptions about the behaviour ofthe government become relevant especially about thenature of the measures taken by the authorities In partic-ular if it is perceived that current policy measures willaffect future variables credibly and permanently thenagents will adapt their behaviour immediately
The emphasis of the new classical macroeconomic para-digm on forward-looking behaviour and expectationshelp to rationalise the apparent puzzle of fiscal consoli-dations with expansionary effects Recent theoreticalmodels belonging to this paradigm show that consump-tion may react to fiscal policy measures in an oppositeway than predicted by standard models in the Keynesiantradition thus leading to effective output expansions(contractions) when fiscal policy is meant to be contrac-tionary (expansionary) The same has been shown in thecase of investment under particular circumstances pol-icy measures aimed at adjusting the budget deficit maylead to a boost in investment with a potentially expan-sionary effect on aggregate output The lsquoconsumptionchannelrsquo and the lsquoinvestment channelrsquo through whichfiscal policies may operate in a non-Keynesian fashionare illustrated below
Non-Keynesian effects of fiscal policy the consumption channel
If agents are forward-looking and rational in forming theirexpectations they will anticipate that a tax cut todayfinanced by government debt will translate into highertaxes at some point in the future If in addition govern-ment intervention is non-distortionary capital markets areperfect and consumers sufficiently long-lived the so-called Ricardian equivalence should hold namely perma-nent income will be unaffected by fiscal policy and soconsumption Under these abstract circumstances fiscalmultipliers will be zero since higher government savings
Table IV1
Some puzzling effects of fiscal policy
Country YearDeficitGDP (1)
Debt GDP
GDPgrowth (2)
Denmark (3) 1982 89 625 30
1986 ndash 33 623 36
Ireland 1986 105 1138 03
1989 17 1001 62
Sweden 1989 ndash 54 453 24
1993 122 758 ndash 22
The table presents the changes in the fiscal stance and its impact on debt andGDP growth Values are shown for the year before the consolidation (stimulus)started and its last year(1) Negative values correspond to a surplus(2) Annual change(3) In Denmark the debt was on a downward path after a peak of 734 in
1984 and real GDP growth accelerated to 44 in 1984 before returning tothe previous level in 1986
Source Giavazzi and Pagano (1996)
yen1part However in this respect it should be mentioned that forward-lookingbehaviour was also incorporated in some Keynesian consumption modelsnotably the life-cycle model In the original formulation of these modelshowever there is no requirement of consumersrsquo expectations to be rational
140
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
obtained through fiscal consolidations will be compen-sated by an equivalent reduction in private savings (1)
However if distortions introduced by taxation are takeninto account a first reason for expecting non-Keynesianeffects of fiscal policy emerges This can be the case forinstance when a current expenditure cut is expected tobe offset in the future by a reduction in future distortion-ary taxes Such a case for non-Keynesian effects of fiscalpolicy was first illustrated by Blanchard (1990) In thismodel it is shown that the effects of fiscal policy onaggregate consumption are likely to be non-linear Thereason is that the dead-weight loss of taxation increasessignificantly with the extent of taxation So if a consol-idation is made starting from a low level of current debta traditional positive fiscal multiplier will result (2) Ifinstead a fiscal consolidation is made starting from ahigh debt level consumption may react positively as aresult of an expected increase in permanent income Thereason is that by consolidating now the government willnot raise taxes too much in the future to pay back thedebt This reduces the dead-weight loss imposed bytaxes thus raising agentsrsquo permanent income (3)
A different motive to expect fiscal policy to have non-lin-ear effects has been proposed by Bertola and Drazen(1993) The assumption here is that when public expend-iture becomes alarmingly high then agents start antici-pating a future major fiscal adjustment to occur This mayoffset any loosening of fiscal policy At the same time aconsolidation occurring when public spending is highmay then change agentsrsquo expectations concerning afuture major retrenchment and the lower expected levelof taxes raises permanent income and consumption (4)
A further rationale for possible non-Keynesian effectsthrough the consumption channel emerges if fiscalconsolidations are assumed to affect the risk of govern-ment insolvency By reducing their budget deficits gov-ernments will signal to markets their willingness toswitch to lsquosound financesrsquo If this signal is taken as cred-ible interest premiums on government bonds will fall
The consequent reduction in interest rates will in turncontribute to raise agentsrsquo permanent income since theywill discount future income streams at a lower rate Thecrucial ingredient of this explanation for the emergenceof non-Keynesian effects is the credibility of governmentaction to make public finances sustainable As empha-sised for instance by Feldstein (1982) the credibility ofthe regime shift can be enhanced by the size of the con-solidation While small adjustments in the budget may bebelieved to be short-lived or not enough to correct theimbalances major fiscal retrenchments may signal thewillingness of the government to face the political costsassociated with the shift to sound public finances Fur-thermore as illustrated for instance by Cotis et al (1998)the introduction of fiscal rules for the maintenance ofbudgetary discipline (like the SGP) may increase the per-ception of the intertemporal budget constraint andthereby the credibility of the fiscal adjustment and thelikelihood of the emergence of non-Keynesian effects
Non-Keynesian effects of fiscal policy the investment channel
Expansionary consolidations working through theconsumption channel act on aggregate demand leavingsupply conditions unaffected (factor supply TFP hellip)Output expansions above potential obtained throughthe consumption channel are therefore inevitably short-lived However recent empirical research has shownthat fiscal consolidations may produce significantshort-run expansionary effects also through the invest-ment channel thus affecting not only demand but alsosupply factors (Alesina and Ardagna 1998 AlesinaPerotti and Tavares 1998 Alesina et al 2002)
The rationale for fiscal policies producing non-Keyne-sian effects through an investment channel has been for-malised in Alesina et al (2002) The highlighted channelgoes beyond possible reductions in real interest ratesassociated with fiscal contractions as predicted by stand-ard macroeconomic models The link between fiscalpolicy and investment behaviour is rather represented bythe labour market
As in models rationalising non-Keynesian effects throughthe consumption channel agents are assumed to be for-ward-looking and to behave on the basis of the actualvalue of future income streams The relevant agents are inthis case firms that decide about their factor service pur-chases by looking at the present value of profits Invest-ment decisions are driven by the expected present value ofthe net marginal product of capital which in turn is anegative function of real wages Fiscal consolidations
yen1part If consumers have short-term horizons or are affected by liquidityconstraints Ricardian equivalence will no longer hold and fiscal policy willaffect consumption according with the predictions of standard models inthe Keynesian tradition (see for example Blanchard 1985)
yen2part In Blanchard (1990) this is due to the fact that agentsrsquo horizons are shortterm since each of them are faced with a constant positive probability ofdeath Hence Ricardian equivalence does not hold in this model even in theabsence of tax distortions
yen3part Results similar to those to Blanchard (1990) are obtained in Perotti (1999)In this model however Ricardian equivalence does not hold on aggregatebecause a fraction of consumers are assumed to be liquidity-constrained
yen4part A similar non-linear effect of fiscal policy is obtained in Sutherland (1997)
141
P u b l i c f i n a n c e s i n E M U 2 0 0 3
obtained through expenditure cuts may increase short-runinvestments via reduced wage pressures for a number ofreasons A reduction in the government wage bill will ingeneral contribute to wage moderation in the private sec-tor as well A similar effect would be obtained by meansof reductions in government transfers The possibility forfiscal consolidations to exhibit non-Keynesian effectsthrough the investment channel will then crucially dependupon the composition of adjustment (expenditure cuts ver-sus tax increases) and on institutional factors above all theworking of the labour market (1)
In sum in view of the latest developments in the theoret-ical paradigm a number of reasons have been identifiedin the theoretical literature that may explain why fiscalconsolidations may have expansionary effects The pos-sibility of non-Keynesian effects working through theconsumption channel is expected to be mainly affectedby factors affecting the credibility of the adjustment andagentsrsquo expectations such as the size of the consolida-tion and the initial state of public finances The likeli-hood of non-Keynesian effects acting via the investmentchannel is instead crucially affected by the compositionof adjustment As illustrated in the next chapter theempirical research on budgetary consolidations hasfocused on the above factors to identify the characteris-tics of expansionary consolidations and the relevantchannels
yen1part Clearly adjustments in the tax structure which mdash within an overall fiscalconsolidation mdash favour a reduction in the tax wedge on labour would alsoimply an increase in the net present value of profits
142
3 Characteristics and effects of fiscal consolidations in the EU evidence from cross-country analysis
31 Survey of existing studies
In existing cross-country studies (see Table IV2) thehypothesis that fiscal consolidation may have expan-sionary effects is analysed empirically in several waysCrucial to this end is the definition of what fiscal consol-idation is Usually it is defined in terms of a givenimprovement in the budget balance as a fraction of GDPachieved over a time period of several years In order toexclude changes in the budget balance associated withthe economic cycle measures of the cyclically-adjustedbudget balance have generally been used Moreover tobetter isolate fiscal policies of a discretionary type inter-est expenditures have been deducted from the structuralbudget balance in most studies that is changes in theprimary cyclically-adjusted budget balance have there-fore been adopted to identify consolidation periods
Depending on the particular study considered theconcept of fiscal consolidation has been focused eitheron the idea of a sufficiently strong fiscal adjustmentachieved in a given period (size criterion) or on the ideaof a sufficiently long time period during which thebudget balance constantly improves (persistence crite-rion) Some studies refer to a further refinement of theconcept of consolidation by defining as successful thoseconsolidations that manage to bring about a sustainedreduction in the debt-to-GDP ratio
The methodologies adopted in the existing studies differquite widely In almost all studies there is a descriptiveanalysis of the sample characteristics of relevant fiscaland macroeconomic variables before during and afterconsolidation periods This allows for the checking ofthe general requirement for the identification of expan-
sionary fiscal consolidations the occurrence of positivegrowth development after the fiscal adjustment Bylooking at sample averages of fiscal variables it is possi-ble to describe the characteristics (in terms of size ofadjustment initial conditions of public finances or com-position of adjustment) of fiscal consolidations and toidentify how these characteristics differ depending onwhether consolidations turned out to be expansionary orcontractionary In some studies ProbitLogit regressionshave also been performed in order to identify economet-rically the main factors affecting the probability for fis-cal consolidation to be successful (Von Hagen Hughes-Hallet and Strauch 2001) or expansionary (Alesina andArdagna 1998) Sample evidence on relevant macroeco-nomic variables (for example interest rates or exchangerates) permits to judge whether fiscal consolidationshave in general been accompanied by active monetarypolicies or devaluations Some studies complementdescriptive sample statistics with country case studiesaimed at better understanding the policy environmentduring consolidation periods (for example wage agree-ment policies exchange rate devaluations etc)
In a number of studies empirical verifications of theoret-ically grounded hypotheses are also provided Giavazziand Pagano (1996) estimate consumption functions totest whether fiscal consolidations may have non-Keynesian effects via the consumption channel due toconsumersrsquo revised expectations and increased expectedlife-time income Giavazzi Jappelli and Pagano (2000)perform a similar test by estimating saving functionsAlesina et al (2002) instead verify empirically thehypothesis that non-Keynesian effects of consolidationsmay come from the investment channel by estimatinginvestment equations
143
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Table IV2
Cross-country evidence on fiscal consolidations
Study and sample Definition of consolidation Aim of the analysis Type of analysis Main findings
McDermott and Westcott (1996) IMF (1996)20 OECD countries 1970ndash95
The primary structural balanceimproves by at least 15 ofGDP over two years and does notdecrease in any year
Analyse the characteristics andeffects of successful consolida-tions that is of consolidationsleading to a 3 of GDP reduc-tion in debt
Descriptive Successful consolidations leads onaverage to increased growth un-successful to reduced growth Sizeand composition both important toidentify successful consolidations
Giavazzi and Pagano (1996)19 OECD countries 1970ndash92
The cumulative change in theprimary structural balance isabove a given threshold as a of GDP (5 4 or 3) over a givennumber of years (resp 4 3 or 2)
Analyse the existence of non-Keynesian effects of fiscal con-solidations via the consumptionchannel
Panel data estima-tion of consump-tion functions
Size of adjustment is relevant toidentify episodes exhibiting non-Keynesian features
OECD (1996)18 OECD countries 1975ndash95
The cumulative change in thestructural budget balance isabove 3 of GDP over a periodof at least two years
Analyse characteristics and ef-fects of fiscal consolidations
Descriptive There were fiscal consolidationsduring which growth was abovepotential Accommodating mone-tary policy seems to matter to limitoutput contractions
Cour et al (1996) 17 OECD countries 1970ndash94
Continuous improvement in theprimary structural budget bal-ance with a period of at mostthree years during which the pri-mary structural budget balanceimproves by at least 3 of GDP
Analyse characteristics and ef-fects of fiscal consolidation epi-sodes with a particular focus onthe consumption channel ofnon-Keynesian effects
Descriptive andestimation of con-sumption func-tions
Size of adjustment is relevant toidentify expansionary episodes
Alesina Perotti and Tavares (1998) 19 OECD countries 1960ndash95
The primary structural balanceimproves by at least 15 ofGDP
Analyse characteristics and ef-fects of fiscal consolidation ex-ploring alternative channels fornon-Keynesian effects
Descriptive Successful consolidations morelikely to lead to expansions Com-position more important than sizeto identify expansionary episodesLabour market structure alsomatters
Alesina and Ardagna (1998) 20 OECD countries 1960ndash94
The primary structural balanceimproves by at least 2 of GDPor by at least 15 of GDP peryear over two years
Analyse characteristics and ef-fects of fiscal consolidation ex-ploring alternative channels fornon-Keynesian effects
Descriptive Probitregressions collec-tion of case studies
Composition more important thansize to identify expansionary epi-sodes Wage agreements and ex-change rate devaluations are alsorelevant accompanying factors
Perotti (1999) 19 OECD countries 1965ndash94
na Analyse whether initial fiscalconditions are relevant for theeffects of fiscal policy
Estimation of dy-namic consump-tion functions
High debt levels are associated witha higher probability for fiscal policyto have non-Keynesian effects
Giavazzi Jappelli and Pagano (2000) 18 OECD countries 1970ndash96
The structural balance improvesby at least 15 of GDP per yearover two years
Analyse the existence of non-Keynesian effects of fiscal con-solidations via the consumptionchannel
Panel data estima-tion of savingfunctions
Size of adjustment is relevant toidentify episodes exhibiting non-Keynesian features Non-Keynesianeffects more likely for tax changesthan expenditure changes and forfiscal consolidations than for fiscalexpansions
Von Hagen Hughes-Hallet and Strauch (2001)20 OECD countries 1960ndash98
The structural balance improvesby at least 125 of GDP peryear over two years or by at least15 of GDP in one year and bya positive amount in a consecu-tive year
Describe characteristics and ef-fects of fiscal consolidations withspecial reference to the EU
Descriptive analy-sis case studiesProbit regressionsestimation of out-put equations andmonetary and fis-cal policy reactionfunctions
Fiscal policies exhibit in generalKeynesian effects but in the EU inthe nineties there is no evidenceneither in favour nor against Key-nesian effects
Alesina et al (2002)18 OECD countries 1960ndash96
The primary structural balanceimproves by at least 2 of GDPor by at least 125 of GDP peryear over two years
Analyse the existence of non-Keynesian effects of fiscal con-solidations via the investmentchannel
Estimation of in-vestment equa-tions descriptiveanalysis
Cuts in public expenditure particu-larly in public employeesrsquo compen-sations boost investmentExpansionary consolidationsassociated with acceleration ininvestment growth
144
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
In spite of the above-mentioned differences in methodol-ogy a number of results are common to almost all stud-ies
bull There is evidence of fiscal consolidations likely toexhibit non-Keynesian features in almost all studies
bull Consolidations leading to a permanent reduction indebt (lsquosuccessfulrsquo) are more likely to be expansion-ary
bull During expansionary consolidations both an accel-eration in private consumption and business invest-ment is observed
bull The policy environment in which fiscal consolida-tions are undertaken matters In particular themonetary exchange rate and wage policies accom-panying consolidations may affect significantly theimpact of fiscal adjustments on growth
Where consensus is missing is on the specific factorsdetermining the expansionary effects of fiscal consolida-tions Some papers find that fiscal adjustments withexpansionary effects are more likely when the size ofconsolidation is large (Giavazzi and Pagano 1996Giavazzi Jappelli and Pagano 2000) In other studiesinstead it is found that what is most significant to char-acterise expansionary consolidations is the compositionof the adjustment Fiscal adjustments based on expendi-ture cuts rather than tax increases have expansionaryeffects with a higher probability especially if expendi-ture cuts are concentrated on public employeesrsquo compen-sations and on government transfers (Alesina Perottiand Tavares 1998 Alesina and Ardagna 1998 Alesinaet al 2002) Finally there are studies that emphasise theinitial state of public finances Consolidations are morelikely to have non-Keynesian effects when they occur incountries and periods where debt-to-GDP ratios are high(Alesina and Ardagna 1998 Perotti 1999)
Overall although cross-country empirical analyses permitto shed light on several features of fiscal consolidationsthe results arising from such analyses need to be inter-preted with caution for a number of reasons First thereare problems in measuring and defining fiscalconsolidation episodes In particular relying on deficit-based measures tends to exclude fiscal reforms with a lim-ited impact on current budget balances but potentiallylarge effects on long-term public finances such as pensionreforms Second existing empirical analyses quite often
fail to take properly into account relevant factors such asdevelopments in monetary and exchange-rate policiesthat contribute to shaping the links between fiscal consol-idations and economic activity (1) Third when interpret-ing the links between fiscal policy and economic activitysimultaneity issues are to be taken into account Not onlyfiscal consolidations affect output growth but actual andexpected growth affect budget balances and policy mak-ersrsquo choices (2) Finally there is the possibility that resultsare driven to some extent by a sample selection bias prob-lem Most of the episodes of fiscal consolidations thatonce started have been aborted early due to very adversegrowth consequences are by definition missing from thesamples used in cross-country analyses
32 Were there expansionary fiscal consolidations in the EU A close look at the data
321 How to define periods of budgetary consolidation with expansionary effects
This section carries out a statistical analysis of the fiscalconsolidations that took place in the EU in the past dec-ades It covers the current EU countries with the excep-tion of Luxembourg during the period 1970ndash2002 (3)The source of the data used in the analysis is theAMECO database developed by the Directorate-Generalfor Economic and Financial Affairs
The main purpose of the analysis is that of identifying anddescribing the characteristics of the fiscal consolidationepisodes that appear to be expansionary An analysis ofthe macroeconomic scenario preceding and following thefiscal consolidation episodes is also provided Comparedwith existing event studies of expansionary fiscal consol-idations the focus here is on testing the robustness of thisconcept with respect to alternative definitions of fiscalconsolidation episodes and of their expansionary status
As shown in the previous section in the existingliterature analysing fiscal consolidation episodes using
yen1part In Von Hagen Hughes-Hallet and Strauch (2001) there is an attempt totake into account the links between fiscal and monetary policies by esti-mating together with output equations fiscal and monetary policyreaction functions
yen2part Some studies (Giavazzi and Pagano 1996 Giavazzi Jappelli and Pagano2000) account for possible simultaneity problems by using 2SLS estima-tion techniques
yen3part The exclusion of Luxembourg is due to missing data As will be clear in thefollowing exposition the very last years of the sample are necessarilydropped when identifying expansionary consolidations since it is not possi-ble to evaluate countriesrsquo growth performances after those years
145
P u b l i c f i n a n c e s i n E M U 2 0 0 3
countryyear panel datasets quite different definitions offiscal consolidation have been proposed so that the com-parison of findings is not always easy and immediate Inthe statistical analysis carried out here in order to cap-ture changes in the government budget balance of discre-tionary nature consolidation periods are generally iden-tified by looking at changes in cyclically-adjustedfigures for budget balances budget balance (possibly netof interest payments to isolate discretionary fiscal adjust-ments from developments in the interest rates)
By fiscal consolidation period it is generally meanteither (see Table IV2)
(i) a period in which a given country experiences a suffi-ciently large improvement in its budget balance due todiscretionary policy or
(ii) a period of continuous improvement of the budgetbalance due to discretionary policies
(iii) or a combination of both the above criteria
Criterion (i) emphasises the size aspect of the adjustmentin a given time period while criterion (ii) focuses on thepersistence aspect that is the fact that fiscal consolida-tions are protracted policy actions which are notreversed in their immediate aftermath For instance thedefinitions provided in Alesina and Ardagna (1998) orAlesina et al (2002) mainly refer to the size criterionwhile those in Cour et al (1996) Giavazzi and Pagano(1996) or OECD (1996) refer especially to the persist-ence criterion (see Table IV2)
As mentioned in the previous section in several analysesthere is reference to a further refinement of the conceptof fiscal consolidation that is that of successful fiscalconsolidation (see for example Alesina and Ardagna1998 or Von Hagen Hughes-Hallet and Strauch 2001)By lsquosuccessfulrsquo consolidation it is meant a consolidationepisode that contributes to improve the budget balanceover a relatively long time period (that is debt levels arepermanently lowered) In the following analysis thenotion of successful consolidations is not used (1)
Concerning the definition of expansionary fiscal consoli-dations the criteria used in existing work differ widely Ingeneral for a fiscal consolidation period to be defined asexpansionary the economy must perform sufficientlywell (for example growth sufficiently fast with respect toprevious years or some benchmark growth rate) after thefiscal adjustment takes place It is to note that the refer-ence period considered to evaluate the growth perform-ance of consolidating countries is generally a relativelyshort-term one (one to three years after consolidation)
The benchmark definition of fiscal consolidation used inthis study is taken from Alesina and Ardagna (1998)According to this definition a year of fiscal consolida-tion is a lsquoyear in which the cyclically-adjusted primarybalance improves by at least 2 of GDP or a period oftwo consecutive years in which the cyclically-adjustedprimary balance improves by at least 15 per year inboth yearsrsquo This notion of fiscal consolidation putsemphasis on the size of the improvement in the primarybudget balance
The benchmark notion of expansionary fiscal contrac-tion used in the present study is the same as that pro-posed by Alesina et al (2002) This criterion classifies asexpansionary an episode of fiscal consolidation if lsquotheaverage real GDP growth in each adjustment year and inthe two years after is greater than the average real GDPgrowth in the two years beforersquo (2)
In order to test the sensitivity of the results to the differ-ent definition of both the fiscal consolidation and of theexpansionary effects in Section 322 while keepingconstant the benchmark definition of expansion expan-sionary consolidation periods will be identified anddescribed according with the benchmark size-based def-inition of consolidation and with an alternative criterionof consolidation based on persistence
In Section 323 instead while keeping constant thesize-based benchmark definition of consolidation
yen1part The focus of the present analysis is in fact on the distinction betweenexpansionary and non-expansionary consolidations Moreover the conceptof successful consolidation tends to overlap with that of fiscal consolida-tion based on a persistence criterion
yen2part The above criterion is different for instance with respect to that employed inAlesina and Ardagna (1998) which specifies that the average real GDPgrowth rate (in difference from the G7 average) in the period of consolidationand in the two years after it must be greater than the average value of thesame variable across all episodes of consolidation Therefore the concept ofexpansion used in Alesina and Ardagna (1998) identifies those consolidationepisodes after which growth has been higher relative to the average consoli-dation periods of the sample In this study the criterion based on growthacceleration proposed by Alesina et al (2002) is chosen as the benchmarkbecause it is better suited to identify fiscal consolidation episodes specific tothe country and potentially exhibiting non-Keynesian features
146
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
expansionary consolidation periods will be identifiedand described according to the benchmark definition ofexpansion based on growth acceleration and to alterna-tive expansion criteria acceleration in trend growth inthe cyclical component of growth and on the growthdifferential with the EU average
In identifying expansionary consolidations a further dis-tinction will be made in order to isolate those expan-sionary consolidation episodes that are unlikely to beattributable to concomitant monetary policy easing orexchange-rate devaluation policies It has been shown infact that fiscal contractions have been quite frequentlyaccompanied by expansionary monetary policies in EUcountries (see for example OECD 1996 Alesina andArdagna 1998) The notion of lsquopurersquo expansionary fis-cal consolidation is thus proposed as one during whichshort-term real interest rates do not fall (1)
Using the different definitions several characteristics ofconsolidation periods are analysed in the next twosections including their size the initial state of publicfinances and how the fiscal adjustment is achieved(tax increases or expenditure cuts) The macroeconomicenvironment before during and after consolidation peri-ods is analysed by reporting average statistics on growthoutput gaps interest rates and on the change in the com-ponents of aggregate demand
322 When does a fiscal consolidation occur
The first exercise is to identify consolidation episodes bycomparing the characteristics of expansionary fiscal con-solidations that arise using different definitions To thisend the results obtained when using the benchmark defini-tion by Alesina and Ardagna (1998) based on the size ofadjustment are compared with an alternative definitionbased on the persistence of adjustment According to thisalternative criterion fiscal consolidations occur when theprimary cyclically-adjusted budget balance improves by atleast 3 percentage points of GDP over three consecutiveyears (the note to Table IV3 provides a formal definition)
Table IV3 reports the number of fiscal consolidationsidentified and describes which countries and in which peri-ods experience expansionary episodes In the sample of
462 observations used (14 EU countries 33 years) 49 fiscalconsolidation episodes have been identified which areconsistent with the definition based on size (2) Using theconcept of fiscal consolidation based on persistence thenumber of consolidation episodes rises to 59
Among the episodes of fiscal consolidation identifiedroughly half of the total number of consolidation experi-ences amount to being expansionary This result does notseem to depend on the definition of fiscal consolidationemployed (size or persistence) Refining further the con-cept of expansionary consolidation to account for themonetary stance or possible devaluations about half ofthem are found to be lsquopurersquo (11 and 16 episodes usingrespectively the size and the persistence concept of con-solidation period)
Concerning the description of the expansionary consoli-dation episodes the evidence of expansionary effectsregistered in Belgium Denmark and Ireland reported inprevious studies is confirmed Sweden also appears to behave experienced expansionary consolidations duringthe mid-1980s and in the late 1990s The identificationof expansionary consolidations in the remaining EUcountries depends quite strongly on the concept used todefine consolidation periods Overall the correlationindex between lsquosizersquo and lsquopersistencersquo expansionaryconsolidation indicators is positive but quite low(033) (3)
Table IV4 reports statistics concerning the characteris-tics (size of adjustment initial state of public financescomposition of adjustment) of the fiscal consolidationsidentified distinguishing whether the consolidationproved to be expansionary or not
Results appear to be very robust with respect to theconcept of fiscal consolidation employed (size or persist-ence) and supportive of findings reported in previousstudies (Alesina and Ardagna 1998 Alesina et al 2002)In particular it is not the simple fact that an adjustment iscarried out that really matters rather it is the compositionof the adjustment which explains its expansionary effectIndeed the amount of the adjustment (measured by the
yen1part Under likely assumptions non-decreasing real interest rates tend to excludeboth monetary expansions under floating exchange rates and devaluationpolicies under fixed exchange rates regimes This is the case for instancein a Mundell-Fleming open economy setting with uncovered interest rateparity (see for example Krugman and Obstfeld 2001)
yen2part The episodes may not coincide with those reported in Alesina and Ardagna(1998) because the method used to obtain cyclically-adjusted figures differ(HP filter in the present study Blanchard-type trend regressions in Alesinaand Ardagna 1998)
yen3part Expansionary consolidation indicators take the value 1 for countryyearcombinations in which an expansionary consolidation occur and zerootherwise
147
P u b l i c f i n a n c e s i n E M U 2 0 0 3
change in the primary cyclically-adjusted budget bal-ance) does not seem to be significantly different betweenexpansionary and non- expansionary consolidation peri-ods using both definitions of fiscal consolidation Whatappears to be relevant to distinguish expansionary fromnon-expansionary periods of fiscal adjustment mdash usingboth definitions of consolidation mdash is its compositionFiscal adjustments based on expenditure cuts are morelikely to be expansionary than consolidation periodsbased on tax increases Looking at overall values forprimary expenditure and for government revenues
(cyclically-adjusted or not) differences are statisticallysignificant irrespective of the concept of consolidationemployed (size or persistence) The definition of consol-idation appears to matter instead as far as the compositionof expenditure is concerned In particular the reductionin the public wage bill found to be relevant to character-ise expansionary fiscal consolidations in other studies issignificantly higher in expansionary than in non-expan-sionary consolidations only when adopting a persistence-type definition of fiscal consolidation
Concerning the initial state of public finances the averagevalue of debtGDP ratios are found to be higher in expan-sionary fiscal consolidation periods by about 10 GDPpercentage points irrespective of the concept of consoli-dation employed (size or persistence) However t testsshow that this difference is not statistically significant (1)
Table IV5 presents data characterising the macro-economic environment preceding during and follow-ing consolidation periods Several results emerge Firstconsolidations are more likely to be expansionary afterperiods characterised by relatively low growth and bynegative output gaps (2)
Second growth appears to accelerate during the consol-idation year and during the following year for expan-sionary episodes while in non-expansionary episodesgrowth is more likely to decelerate Interestingly trendgrowth accelerates in expansionary consolidation peri-ods when consolidation is defined according to persist-ence while trend growth appears to be constant beforeand after consolidation when using a definition based onsize As for unemployment it worsens during non-expansionary consolidations while this is not the casefor expansionary fiscal adjustments
Third both private consumption and business investmentaccelerate during expansionary consolidation periods withinvestment registering a much higher acceleration By con-trast investment decelerates during non-expansionaryperiods and even drops after the consolidation (negativegrowth rates of investment) (3) Moreover an acceleration
Table IV3
Expansionary consolidations description of episodes with alternative definitions of consolidation
Size Persistence
Number of consolidation episodes 49 59
Number of expansionary episodes 24 31
Number of lsquopurersquo expansionary episodes
12 16
Description of expansionary episodes
Size Persistence
BE 1984 1985 1985 1986 1987
DK 1983 1984 1984
DE 1982 1982 1983 1984
EL 19821987 1994 1996 1994 1997 1998
ES 1986
FR 1996 1997
IE 1976 1987 1988 1984 1987 1988 1989
IT 1976 1977 1993 1993
NL 1993 1982 1983
AT 1996 1997
PT 1986
FI 1993 1977
SE 1983 1987 1995 1998 1982 1983 1984 1995 1997 1998
UK 1997 1981 1982 1997
Definitions of fiscal consolidationSize The primary cyclically-adjusted budget balance improves by at least 2 per-
centage points of GDP at time t or by 15 at least points in two consecutiveyears (ie t and t-1 or in t and t+1)
Persistence The primary cyclically-adjusted budget balance improves by at least3 percentage points of GDP over three consecutive years (ie between t-2and t or between t-1 and t+1 or between t and t+2) and in each year thechange in the primary cyclically-adjusted budget balance cannot be belowndash05 percentage points of GDP
Definition of an expansionary fiscal consolidationA fiscal consolidation in which the average real GDP growth between t andt+2 is greater than between t-1 and t-2
Definition of a pure expansionary consolidationAn expansionary fiscal consolidation in which the average change in realshort run interest rates between t-1 and t+1 is non-negative
Source Commission services
yen1part When performing comparisons between variables t tests permit to take intoaccount both measures of position (averages) and of variability (standarddeviations) This helps in understanding when apparently large differencesin averages are mainly driven by the fact that variables are highly volatile
yen2part This effect however may be due in part to the autonomous development ofthe business cycle which resumes from a period of below potential growth
yen3part It should be noted that in spite of the greater variation in investment thecontribution of consumption changes to growth is always higher than thatof investment due to the larger weight on total aggregate demand
148
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
in TFP growth is recorded Concerning the current accountbalance while during non-expansionary consolidations amarked worsening of the current account is observed thisis not the case for expansionary consolidations
Finally during consolidation periods both expansionaryand non-expansionary there is a reduction in nominalinterest rates irrespective of the definition of consolida-tion employed This finding is consistent with the factthat in these periods the fiscal stance is meant to be con-tractionary Moreover falling short-term nominalinterest rates are a possible indication of concomitantmonetary expansions or exchange rate devaluations (1)
323 When is a fiscal consolidation expansionary
The second exercise undertaken is that of analysing therobustness of the characteristics of expansionary fiscalconsolidations with respect to different criteria to iden-tify expansionary episodes
For this purpose after having seen that alternative defi-nitions of consolidation produce similar results the
benchmark size-based definition of consolidation is keptconstant and different definitions of expansion are usedTwo alternatives to the benchmark expansion definitionby Alesina et al (2002) are proposed
The first definition employs the notion of trend outputgrowth as opposed to the of real GDP growth to identifyexpansionary episodes (see note to Table IV6 for the for-mal definition) Resting on the assumption that fiscal pol-icy may affect potential output the idea is that an episodeof fiscal adjustment is meant to be expansionary providedit is associated with an acceleration of trend output Suchdefinition of expansion permits to distinguish consolida-tion periods associated with positive developments in theeconomic cycle from those which are associated withpositive output developments of a more structural natureThe second criterion defines as expansionary those fiscalconsolidations that are associated with an increase in thedifference between the growth rate in countriesrsquo GDP andthe EU average GDP The aim of this criterion is that ofidentifying those expansionary episodes which are associ-ated with a growth acceleration which is not attributable tothe EU-wide economic cycle (2)
Table IV4
Size and composition of expansionary consolidations alternative definitions of consolidation
Criterion of fiscal consolidation Size Persistence
Non exp Exp t test for (1) ne (2)
Non exp Exp t test for(3) ne (4)Variables (change as a of GDP) Average values Average values
(1) (2) (3) (4)
Primary CAB 29 28 08 17 16 02
Debt (level as a of GDP) 654 751 ndash 09 619 760 ndash 15
Primary expenditure 00 ndash 16 29 (2) 01 ndash 11 28 (2)
Government investment ndash 02 ndash 03 16 ndash 01 ndash 03 20 (2)
Public employees compensation 00 ndash 02 14 00 ndash 02 19 (1)
Total government revenues 23 10 41 (2) 14 05 29 (2)
Total cyclically-adjusted government revenues 24 11 33 (2) 13 04 26 (2)
NB t test values labelled by (1) and (2) refer respectively to cases in which the average value of variables during expansionary and non-expansionary consolidations arestatistically different at a 90 and 95 confidence interval
Source Commission services
yen1part Depending upon the evolution of inflation falling nominal interest ratesmay not correspond necessarily to effectively monetary easing Howeverin most of the cases real interest rates also appear to fall between t and t-1both during expansionary and non-expansionary consolidations whilechanges between t and t+1 show a less clear pattern (unreported) Concern-ing nominal exchange rates it is found that during both expansionary andnon-expansionary consolidations the exchange rate of the consolidatingcountry with respect to the US dollar tends to depreciate overall the wholeperiod between t-1 and t+1 (unreported)
yen2part A third definition of expansion complementary to the first one as also beentested According to this definition an episode of fiscal consolidation isexpansionary provided it is associated with an increase in the differencebetween actual and trend output growth The results however are particu-larly close to the baseline scenario (correlation index at 098) given thatmost of the movements in this variable is due to changes in actual growthTherefore the results of such exercise are not shown
149
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Table IV6 reports the number of fiscal consolidationsidentified and describes which countries and in whichperiods experience expansionary episodes An interest-ing result is already that irrespective of the definitionused to identify expansionary episodes the number ofexpansionary consolidations is about 20 that is roughlyhalf the total number of consolidation experienced in EUcountries Most of the country-specific cases are identi-fied under the three definitions
Adopting the narrower definition of lsquopurersquo expansionaryconsolidation that is excluding the expansionary consolida-tion periods likely to be associated with monetary expan-sions or exchange rate devaluations the number of expan-sionary episodes reduces to about 10 Again this result isfairly robust with respect to the definition of expansion used
So irrespective of the definition of expansion usedabout half of the consolidation periods experienced by
Table IV5
Macroeconomic environment in expansionary consolidations alternative definitions of consolidation
Criterion for fiscal consolidation Size Persistence
Variables Non-exp Exp Non-exp Exp
Growth rate of real GDP ()
t-1 26 16 26 12
t 11 21 20 22
t+1 07 34 24 31
Output gap ( of trend output)
t-1 04 ndash 11 03 ndash 18
t 02 ndash 15 02 ndash 16
t+1 ndash 03 ndash 08 07 ndash 10
Trend GDP growth
t-1 26 26 29 25
t 25 26 28 26
t+1 25 26 27 28
Growth rate of real private consumption ()
t-1 24 14 23 13
t 14 18 17 19
t+1 15 30 23 26
Growth rate of real business investment ()
t-1 35 03 54 06
t ndash 06 37 05 35
t+1 ndash 33 67 ndash 15 64
Growth rate in real current account surplus ()
t-1 30 01 31 00
t ndash 01 ndash 02 ndash 41 00
t+1 ndash 02 04 ndash 43 ndash 05
Growth rate in TFP ( change)
t-1 09 10 12 09
t 02 16 08 17
t+1 02 21 13 21
Unemployment rate ( of labour force)
t-1 65 87 7 95
t 69 9 76 94
t+1 89 89 79 91
Short -term nominal interest rates
t-1 113 125 109 97
t 109 115 102 93
t+1 99 102 98 86
Source Commission services
150
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
EU countries appear to be expansionary and about onequarter appear to be lsquopurersquo expansionary (not likely tobe accompanied by expansionary monetary policies ordevaluations)
Turning to the description of the expansionary consoli-dation episodes all criteria used to identify expansion-ary episodes permit to isolate the experiences of Den-mark (1983ndash84) and Ireland (1987ndash88) which sinceGiavazzi and Pagano (1990) are known to be the clas-sical examples of fiscal adjustment exhibiting possiblenon-Keynesian features Expansionary fiscal consoli-dations are found in Spain and Portugal in 1986 as wellas in West Germany in 1982 Non-expansionary epi-sodes are instead found in France Findings concerningGreece Italy Sweden and the UK depend quite cru-cially on the criterion chosen to define an expansionaryadjustment Concerning Finland not surprisingly anexpansionary period in 1993 is found using all criteriaexcept that based on trend growth since results are very
much driven by the strong output contraction experi-enced in 1991
Correlation indexes (reported in Table IV1) amongexpansionary consolidation indicators based on differentdefinitions of expansion help to understand the extent towhich alternative criteria tend to yield overlappingresults The benchmark criterion based on the accelera-tion of real GDP growth is correlated to a certain extentwith the trend growth criterion (063) The resultsobtained using the definition of expansionary consolida-tions using the criterion of actual minus EU growth ismore correlated with the baseline scenario (076) thanwith trend growth scenario (051)
Table IV8 reports average values and t tests for thecharacteristics of the fiscal consolidations identified dis-tinguishing according to the expansionary status of theconsolidation and repeating the analysis for the differentdefinitions of expansion
Table IV6
Expansionary consolidations description of episodes with alternative definitions of expansion
Growth Trend growth Actual minus EU growth
Number of consolidation episodes 49
Number of expansionary episodes 24 22 21
Number of lsquopurersquo expansionary episodes 11 11 11
Description of expansionary episodes
BE 1984 1985 1984 1985 1984 1993
DK 1983 1984 1983 1984 1983 1984
DE 1982 1982 1982
EL 19821987 1994 1996 19861987 1991 1994 1996 19821991 1994 1996
ES 1986 1986 1986
FR
IE 1976 1987 1988 1987 1988 1987 1988
IT 1976 1977 1993 1997 1976 1977 1992 1993
NL 1993 1993
AT 1984
PT 1986 1986 1986
FI 1993 1993
SE 1983 1987 1995 1998 1995 1996 1998 1983 1998
UK 1997 1980 1997 1998
NB All fiscal consolidations are of the size-type (see note to table ) A lsquopurersquo expansionary fiscal consolidation is an expansionary fiscal consolidation in which the average change in real short run interest rates between t-1 and t+1 isnon-negative
Definitions of expansionary fiscal consolidationGrowth average real GDP growth between t and t+2 greater than between t-1 and t-2Trend growth average trend growth between t and t+2 greater than between t-1 and t-2Actual minus EU growth average difference (actual real GDP growth ndash EU average real growth) between t and t+2 greater than between t-1 and t-2
Source Commission services
151
P u b l i c f i n a n c e s i n E M U 2 0 0 3
The results shown in Table IV8 are fairly robust withrespect to alternative definitions of expansion (based ongrowth trend growth difference between growth andtrend growth difference between growth and EU aver-age growth) In general there is support for the view thatthe composition of adjustment is more significant toidentify expansionary episodes rather than the size of theadjustment carried out In particular an expansionaryadjustment would be a combination of tax increases andcuts in primary expenditure where however the lattermeasure would be more sizeable Consistently with theargument proposed by Alesina et al (2002) it is also
found that during expansionary consolidations there is amore marked reduction in expenditure on public employ-eesrsquo wage bill However this difference is significant(and highly so) only when a criterion of expansion basedon trend growth is chosen This finding is consistent withthe idea that the wage bill of public employees affectsreal output through higher profits and then greaterinvestment To the extent that investment affects bothdemand and supply conditions one should expect trendoutput to be affected while this is not the case if wageswould only affect output via the demand channel
Table IV9 shows data illustrating the macroeconomicenvironment before during and after consolidations Itconfirms the result that in general expansionary consoli-dations follow periods of low growth and negative output
gaps Regarding the behaviour of aggregate demand com-ponents again both private consumption and businessinvestment accelerate during expansionary consolidationperiods with investment registering a greater accelerationThe finding that during non-expansionary periods invest-ment decelerates quite significantly is also confirmed androbust with respect to the notion of expansion adoptedAgain the level of short-term nominal interest rates fallsduring both expansionary and non-expansionary consoli-dations irrespective of the definition of expansion used
What seems to differ quite significantly depending on thedefinition of expansion adopted is the behaviour of thecurrent account balance When the expansion is measuredby trend growth the current account balance worsens dur-ing expansionary consolidations while the opposite is true
Table IV7
Correlation indexes among alternative indicators of expansionary consolidations
Growth Trend growthActual minus
EU growth
Growth 1
Trend growth 063 1
Actual minus EU growth
076 051 1
Source Commission services
Table IV8
Size and composition of expansionary consolidations alternative definitions of expansion
Growth Trend growth Actual minus EU growth
Non-exp Exp t test for (1) ne (2)
Non-exp Expt test for (3) ne (4)
Non-exp Exp t test for (7) ne (8)
Average values Average values Average values
(1) (2) (3) (4) (7) (8)
Primary CAB 29 28 06 29 28 02 29 27 07
Debt (level as a of GDP) 654 751 ndash 09 639 777 ndash 16 616 811 ndash 22 (2)
Primary expenditure 00 ndash 16 32 (2) 00 ndash 18 36 (2) ndash 04 ndash 14 17 (1)
Government investment ndash 02 ndash 03 19 ndash 02 ndash 03 ndash 013 ndash 02 ndash 03 00
Public employees compensation 00 ndash 02 16 01 ndash 04 40 (2) 00 ndash 02 12
Total government revenues 23 10 39 (2) 22 09 41 (2) 20 11 28 (2)
Total cyclically-adjusted government revenues
24 11 33 (2) 26 08 49 (2) 21 12 21 (2)
NB t test values labelled by (1) and (2) refer respectively to cases in which the average value of variables during expansionary and non-expansionary consolidations arestatistically different at a 90 and 95 confidence interval
Source Commission services
152
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
when the other criteria for consolidation are used Thisinteresting result highlights that the evolution of the exter-nal sector may be a consequence of the type of fiscaladjustment carried out rather than its determinant There-fore a less unfavourable evolution of the current accountbalance may help to identify fiscal consolidation episodesleading to cyclical improvements (that is when theexchange rate tends to depreciate) but not those associ-
ated with growth improvements of a structural nature (thatis when the exchange rate would tend to appreciate) Inthe latter cases in fact the current account gives on averagea negative contribution to the evolution of growth Theview that expansionary fiscal consolidations should beseen as a phenomenon mainly associated with exchangerate depreciations or devaluations and consequent currentaccount improvements is thus not supported
Table IV9
Macroeconomic scenario in expansionary consolidations alternative definitions of expansion
Definition of expansion Growth Trend growth Actual minus EU growth
Variable Non-exp Exp Non-exp Exp Non-exp Exp
Growth rate of real GDP ()
t-1 26 16 22 20 28 11
t 11 21 12 21 13 20
t+1 07 34 16 27 13 31
Output gap ( of trend output)
t-1 04 ndash 11 09 ndash 19 00 ndash 07
t 02 ndash 15 01 ndash 16 ndash 03 ndash 11
t+1 ndash 03 ndash 08 ndash 02 ndash 10 ndash 02 ndash 10
Trend GDP growth
t-1 26 26 31 20 26 25
t 25 26 29 22 26 25
t+1 25 26 28 23 26 26
Growth rate of real private consumption ()
t-1 24 14 21 17 23 14
t 14 18 12 20 18 13
t+1 15 30 15 31 19 26
Growth rate of real business investment ()
t-1 35 03 04 34 45 ndash 08
t ndash 06 37 ndash 06 41 18 12
t+1 ndash 33 67 ndash 26 68 ndash 07 49
Growth rate in real current account surplus ()
t-1 30 01 03 26 22 03
t ndash 01 ndash 02 01 ndash 04 00 ndash 04
t+1 ndash 02 04 00 02 00 02
Growth rate in TFP ()
t-1 09 10 07 12 13 04
t 02 16 05 14 02 17
t+1 02 21 08 15 04 20
Unemployment rate ( of labour force)
t-1 65 87 62 79 66 88
t 69 90 70 80 69 94
t+1 89 89 75 80 72 94
Short run nominal interest rates
t-1 113 125 124 114 103 132
t 109 115 120 105 95 127
t+1 99 102 108 93 89 112
Source Commission services
153
P u b l i c f i n a n c e s i n E M U 2 0 0 3
324 Summary of findings
The analysis carried out in this chapter relying upon alter-native definitions of fiscal consolidation and on differentcriteria to identify expansionary fiscal adjustments leads toa number of findings that can be summarised as follows
bull Fiscal consolidation episodes exhibiting non-Key-nesian features can be found in Europe growthappears to have accelerated after about half of theconsolidation episodes identified and in roughly onequarter of the cases this happened without a mone-tary stimulus Hence there is an indication thatroughly half of the expansionary fiscal consolida-tions are unlikely to be attributable to concomitantmonetary policy easing or devaluations
bull Expansionary fiscal consolidations are more likelyto be based on expenditure cuts than on tax increasesirrespective of the definition of fiscal consolidationor expansion employed Expansionary fiscal adjust-ment periods also appear to be associated with initialhigh levels of debt while the simple fact that therehas been an adjustment is not enough to guaranteeits positive effect on output Consolidations basedon cuts in wage expenditure seem to be more likelyto spur potential growth (1) Consistent findings arefound in previous studies (Alesina and Ardagna1998 Alesina et al 2001)
bull The macroeconomic environment preceding expan-sionary consolidation periods is characterised byslow growth and negative output gaps comparedwith that characterising non-expansionary consoli-dations This finding appears robust with respect tothe definition of consolidation used and the defini-tion of expansion adopted
bull There is evidence that the acceleration in growthfollowing fiscal consolidations may have either astructural nature (trend growth is affected) or acyclical one or have both a structural and a cyclicalcomponent During expansionary consolidationsboth consumption and investment accelerate Thebehaviour of business investment seems especiallyhelpful in distinguishing between expansionary andnon-expansionary episodes Irrespective of the defi-nition of consolidation and expansion used while innon-expansionary cases investment falls in expan-
sionary periods there is a strong acceleration in thiscomponent of aggregate demand
bull The results presented above highlight the fact thatcredibility and confidence may play a role in deter-mining the effects on output of the fiscal adjustmentBox IV1 provides some examples about the rela-tionship between confidence and fiscal adjustmentsThe indicators of householdsrsquo and businessesrsquo con-fidence are below average before the consolidationstarts and tend to improve in those cases of budgetconsolidations followed by acceleration in growth
Before taking firm policy conclusions however itshould be recalled that the above results are to be inter-preted with caution As mentioned in Section 31 cross-country empirical analysis on fiscal consolidations aresubject to a series of problems and limitations In partic-ular in interpreting results referred to the short-run it isquite difficult to understand to what extent consolida-tions affect growth or if it is actual and expected outputgrowth which affects budget balances and budget poli-cies (2) Moreover it is quite difficult to isolate the effectof external factors (such as monetary and exchange-ratepolicies) that shape the links between fiscal consolida-tions and economic activity
An ideal way to overcome the above difficulties in inter-preting the empirical evidence would be that of creatinga policy experiment in which a fiscal shock occurs in iso-lation from other policies and from other types of shocksto macroeconomic variables Though real-world policyexperiments are not feasible the use of applied macro-economic models helps to understand how such hypo-thetical policy experiments would work in reality Tothis end the next chapter presents simulations on theeffects of alternative types of fiscal consolidations fromthe Directorate-General for Economic and FinancialAffairs QUEST model
yen1part However this finding is quite fragile with respect to the definition of fiscalconsolidation or expansion used
yen2part The possibility of a mistaken interpretation of results is somehow sup-ported by the fact that expansionary consolidations are more likely to occurafter weak growth and when output gaps are negative The growth pick-upobserved after expansionary consolidations may therefore be related tosome extent to independent cyclical developments However even restrict-ing the analysis to relatively homogenous cases from the viewpoint ofcyclical conditions the evidence still seems potentially consistent with thehypothesis of consolidations with non-Keynesian effects In order to avoidextreme cases where this independent cyclical effect is more likely to playa relevant role the sample could be limited to fiscal consolidations epi-sodes (according to the benchmark definition) occurring when output iswithin 2 percentage point from potential (which is the case for about 80 of the cases) In this case average growth is 17 14 and 18 respectively inthe year before during and after consolidation which still shows that theconsolidation would not have recessionary effects
154
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
Box IV1 Expansionary fiscal consolidations and confidence indicators
(Continued on the next page)
The rationale for the emergence of non-Keynesian fiscal consolidations is based on the forward-looking behaviour of con-sumers and investors The improvement of the government budget position may lead to increased consumption or invest-ment if the expectations of economic operators are affected positively Consumers may raise their consumption spendingif the consolidation of public finances credibly signals reduced future taxation thus leading to a higher perceived perma-nent disposable income Businesses may invest more if the fiscal consolidation leads to the perception of improvedexpected future profit opportunities A common indicator to measure the level of confidence of consumers and businessesis the economic sentiment indicator (ESI) developed by the National Bureau of Economic Research (NBER) which sum-marises the attitudes and judgements on the current economic situation through surveys conducted on a large number ofeconomic actors For EU countries the economic sentiment indicator is calculated on a monthly basis by the EuropeanCommission within the framework of the joint harmonised EU programme of business and consumer surveys (1)Table IV1 reports values of the economic sentiment indicator during periods of fiscal consolidations resulted to be expan-sionary (see Table IV3 and Table IV4) The fiscal consolidation episodes chosen are those in Ireland in 1987 and 1988in Italy in 1993 in Greece in 1994 and 1996 and in Sweden in 1998 On the left-hand axis the value of the ESI is reported(changing on a monthly basis) on the right-hand axis the value of the primary cyclically-adjusted CAB (annual data) isreported
(1) The European Commission economic sentiment indicator is the result of monthly surveys collected on a sample of 67 000 firms and 24 000 consumersacross the EU The questions included in the surveys concern the perception of economic actors on the state of the economy in the coming monthsAnswers are ranked according to their degree of lsquooptimismrsquo with higher scores for more optimistic answers Answers from different economic actorsare aggregated using predetermined weights (40 for industrial firms 20 for households 20 for construction firms 20 for retail trade firms)Figures are seasonally adjusted and normalised such that 1995=100
Graph IV1 Economic sentiment indicators during expansionary consolidations
Economic sentiment indicators during expansionary consolidations Ireland 1987 1988
Primary CAB 1999
Primary CAB 1998
Primary CAB 1987
Primary CAB 198695
96
97
98
99
100
101
102
0
1
2
3
4
5
6
7
1986
M01
1986
M03
1986
M05
1986
M07
1986
M09
1986
M11
1987
M01
1987
M03
1987
M05
1987
M07
1987
M09
1987
M11
1988
M01
1988
M03
1988
M05
1988
M07
1988
M09
1988
M11
1989
M01
1989
M03
1989
M05
1989
M07
1989
M09
1989
M11
155
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Box IV1 (continued)
(Continued on the next page)
Economic sentiment indicators during expansionary consolidations Italy 1993
Primary CAB 1994
Primary CAB 1993
Primary CAB 1992
92
93
94
95
96
97
98
99
100
101
0
05
1
15
2
25
3
35
4
1992
M01
1992
M02
1992
M03
1992
M04
1992
M05
1992
M06
1992
M07
1992
M08
1992
M09
1992
M10
1992
M11
1992
M12
1993
M01
1993
M02
1993
M03
1993
M04
1993
M05
1993
M06
1993
M07
1993
M08
1993
M09
1993
M10
1993
M11
1993
M12
1994
M01
1994
M02
1994
M03
1994
M04
1994
M05
1994
M06
1994
M07
1994
M08
1994
M09
1994
M10
1994
M11
1994
M12
Economic sentiment indicators during expansionary consolidations Greece 1994 1996
Primary CAB 1997Primary CAB 1996
Primary CAB 1995
Primary CAB 1994
Primary CAB 1993
98
985
99
995
100
1005
101
1015
102
ndash 3
ndash 2
ndash 1
0
1
2
3
4
5
1993
M01
1993
M03
1993
M05
1993
M07
1993
M09
1993
M11
1994
M01
1994
M03
1994
M05
1994
M07
1994
M09
1994
M11
1995
M03
1995
M05
1995
M07
1995
M09
1995
M11
1995
M01
1996
M03
1996
M05
1996
M07
1996
M09
1996
M11
1996
M01
1997
M03
1997
M05
1997
M07
1997
M09
1997
M11
1997
M01
156
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
Box IV1 (continued)
It is difficult to compare annual budgetary data with monthly surveys and to attribute to specific months the effect of eachbudget one should determine whether the impact on confidence materialises when the budget is announced when it isapproved or when evidence about its implementation emerges Nevertheless some gross information can be obtained bylooking at the above figures In all cases an improvement in the economic sentiment indicator is observed in the monthsduring which the consolidation was taking place In Ireland confidence improved after the consolidation of 1987 tookplace and the ESI continued to improve also after the consolidation occurred in 1988 In Italy the consolidation of 1993coincided with an inversion of the trend in the ESI during 1992 confidence was falling starting from mid-1993 confidencebegan to rise In Greece it is worthwhile to observe that the improvement in the ESI during the consolidation of 1994 wasonly temporary The worsening of the primary cyclically-adjusted balance in 1995 coincided with a fall in confidenceConfidence improved again during the consolidations of 1996 and 1997 Concerning Sweden between 1997 and 1999 therise and fall pattern in the ESI follows quite closely that in the primary cyclically-adjusted budget balance
This evidence is generally consistent with the view that fiscal consolidations may have expansionary effects via improvedeconomic confidence of economic operators
Economic sentiment indicators during expansionary consolidations Sweden 1998
Primary CAB 1999
Primary CAB 1998
Primary CAB 1997
96
965
97
975
98
985
99
995
100
1005
0
1
2
3
4
5
6
7
8
9
1997
M01
1997
M02
1997
M03
1997
M04
1997
M05
1997
M06
1997
M07
1997
M08
1997
M09
1997
M10
1997
M11
1997
M12
1998
M01
1998
M02
1998
M03
1998
M04
1998
M05
1998
M06
1998
M07
1998
M08
1998
M09
1998
M10
1998
M11
1998
M12
1999
M01
1999
M02
1999
M03
1999
M04
1999
M05
1999
M06
1999
M07
1999
M08
1999
M09
1999
M10
1999
M11
1999
M12
157
4 Assessing ex ante the effects of fiscal consolidations simulation results from the QUEST model
41 Introduction (1)
This chapter describes the macroeconomic effects offiscal consolidations based on simulations using theQUEST model QUEST is an applied macroeconomicmodel whose foundations can be characterised as a mod-ern version of the neoclassical-Keynesian synthesisBehavioural equations in the model are based on inter-temporal optimisation of households and firms with for-ward-looking expectations (2) Unemployment is gener-ated by imperfect matching between workers and firmsPrices adjust sluggishly and the nominal wages responseis delayed because of overlapping wage contracts Themodel has Keynesian features in the short run but theeffectiveness of fiscal policy is more limited than in thetraditional econometric models because of the built-inintertemporal budget constraints However since plan-ning horizons are finite there is no complete tax dis-counting and Ricardian equivalence does not holdMoreover total consumption is represented as the aggre-gation of the responses of two groups of households oneforward-looking group that follows the optimal con-sumption rule given by the life cyclepermanent incomehypothesis and a liquidity-constrained group whose con-sumption depends on current disposable income
Taxes are in general distortionary in the model and affectlong-term employment and capital formation as well asconsumption decisions by private agents Consolidationsthrough tax increases have therefore long-term negative
consequences in the model The only exception to this islump-sum taxes which do not create any distortions butthis is of limited practical relevance
A reduction in government expenditure in QUEST affectsconsumption of the liquidity-constrained householdswho see their current disposable income decline if wagesand employment are falling However the non-liquidity-constrained households could increase their consump-tion as interest rates fall and if they anticipate higherdisposable incomes in the future The removal of distor-tions that this entails could boost employment and outputand already affect life-time income in the short runExpansionary effects through the consumption channelmay occur in the medium term but if a sizeable share ofhouseholds is liquidity-constrained it is unlikely that inthe short run the boost to consumersrsquo spending thatmight result from the fiscal consolidation will be strongenough to offset the negative impact of the reduction ingovernment spending Thus the emergence of non-Keynesian effects of fiscal consolidations through consum-ersrsquo spending crucially depends on the severity of creditconstraints and on the degree of distortions associatedwith public intervention (3)
Besides the consumption channel QUEST allows for theworking of non-Keynesian effects through the workingof the investment channel A reduction in public expend-iture in particular public employment will raise unem-ployment and exert downward pressure on wages Thisin turn tends to boost profits and raise investment spend-
yen1part For an extended analysis see Giudice et al (2003)yen2part The model has a richer theoretical structure than most macroeconometric
models Moreover as in standard computable dynamic general equilibriummodels it allows for adjustment costs and nominal rigidities For a presen-tation of QUEST II model see Roeger and inrsquot Veld (1997 2002)
yen3part The fraction of liquidity-constraint households in QUEST is obtained fromavailable estimates
158
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
ing This is the investment channel emphasised forinstance in Alesina et al (2002)
The question arises whether the non-Keynesian channelsdescribed above could prevail over the traditional Keyne-sian channels and lead to expansionary fiscal consolida-tions If this is the case it is then relevant to understandwhich type of fiscal consolidations are more likely toinduce a prevalence of non-Keynesian channels
For comparability all scenarios of consolidation in thesimulations are of equal ex ante size and standardised toconsolidations of 1 percentage point of GDP that ispermanent increases in taxation or reductions in expend-iture of 1 of (baseline) GDP (1) It should be noted thatan annual budgetary adjustment equivalent to 1 ofGDP would be large relative to what Member States areenvisaging in their stability and convergence pro-grammes in order to reach the SGP goal of budget posi-tions which are lsquoclose to balance or in surplus Thisimplies that two of the factors that are generally investi-gated in analysing fiscal consolidations that is the sizeof the adjustment and the initial state of public financesare not directly explored here (2) The policy experimentsare also applied to one country in isolation (Germany)and no attention is paid to possible cross-country spill-over effects (3)
All scenarios assume that the fiscal consolidations arepermanent and credible that is private agents fully andcorrectly anticipate the effects of fiscal consolidationand do not expect the fiscal policies stance to be revertedin the future An exercise is carried out in Section 431to assess the implications of fiscal consolidation whichare perceived as non-credible since they are reversed inthe following years
The default monetary policy assumption in the scenariosdescribed below is based on a forward looking Taylor-
type rule The monetary authorities are assumed to setshort-term interest rates at a level that depends both onthe deviation of the forecast of inflation from the targetinflation rate and on the magnitude of the output gap Toevaluate the impact of the monetary policy stance on theeffects of fiscal consolidation an alternative monetarypolicy rule leading to a looser policy stance is alsoconsidered (see Section 432)
The simulation results are presented as changes in levelsof relevant macroeconomic variables These results areequally interpretable as deviations from baseline steady-state growth
42 Tax increases
With distortionary taxes it should come as no surprisethat a fiscal consolidation through tax increases has anegative impact on output The purpose of this section ismerely to provide a comparison for the simulations ofexpenditure reductions and contrast the potential effectson output Three simulations are carried out below per-manent tax increases of 1 of GDP in labour incometax corporate profit tax and VAT respectively
As expected all these scenarios show negative GDPeffects in the short and medium run the tax rises increasethe distortions in the economy and lower output Labourincome tax and VAT affect consumption more thaninvestment and they both reduce employment Incontrast the consolidation through an increase in thecorporate tax rate has the largest impact on capital for-mation which falls sharply on impact while the increasein unemployment is only of temporary nature On thewhole these negative output effects broadly confirm thefindings in the previous section that consolidationsthrough tax rises are seldom expansionary
43 Expenditure cuts
In the episode study in the previous section it was foundthat fiscal adjustments based on expenditure cuts seem tohave a higher probability to be expansionary than thosebased on tax increases The set of scenarios below are fis-cal consolidations through alternative types of expenditurereductions cuts in government purchases in governmentemployment or in government transfers to households
All the policy experiments considered lead to negativeGDP effects on impact in the short run but are reversedin the medium to long run Permanent cuts in govern-
yen1part The simulated fiscal consolidations have an impact on the size and evolu-tion of public debt The solution of the model requires the debt to be sus-tainable In the simulations the debt is stabilised at a 10 lower level as apercentage of GDP through reductions over time in labour income taxes
yen2part The non-linearities in the model are not substantial enough to analyse theimportance of larger versus smaller fiscal consolidations and the modelresults are close to proportional for larger adjustments than the standard-ised consolidations of 1 percentage point considered here Nor are weexploring here the significance of the initial state of public finances Insteadwe focus our attention on the composition of fiscal adjustments and look atthe effects for different tax and expenditure categories
yen3part Note that as the simulations are performed under an existing EMU frame-work there is also no role for an exchange rate channel a potentiallyimportant element in some of the episodes studied in the previous chapter
159
P u b l i c f i n a n c e s i n E M U 2 0 0 3
ment purchases and government employment can alreadyboost consumption spending in the short run as forward-looking households are anticipating higher disposableincome in the future (1) The simulations show that thiseffect becomes relatively stronger over time strongenough to offset after some years the negative effect onGDP on impact associated with reduced aggregate demandfrom the public sector When instead it is transfer pay-ments to households to be reduced the boost to private
spending in anticipation of lower tax liabilities in the futureappears not to be large enough to offset the negative impactof lower transfer receipts Consumption remains belowbase although investment spending gradually recovers
A reduction in government employment can from thethird year on increase investment spending This scenariodisplays the largest potential gains in terms of highergrowth after the initial decline in the first years The short-term rise in unemployment in this case puts downwardpressure on real wages in the private sector and increasesprofits for firms (Alesina et al 2002) Lower real wagecosts also boost private sector employment again in the
Table IV10
Permanent increase in labour income tax of 1 of GDP
change from baseline 1st year 2nd year 3rd year 4th year 5th year 10th year
GDP ndash 036 ndash 047 ndash 06 ndash 071 ndash 080 ndash 109
Private consumption ndash 090 ndash 110 ndash 119 ndash 125 ndash 131 ndash 142
Private investment ndash 029 ndash 057 ndash 086 ndash 109 ndash 129 ndash 191
Real wage costs 070 094 071 056 058 019
Real effective exchange rate 014 008 ndash 001 ndash 010 ndash 016 ndash 042
Absolute change from baseline
Short-term interest rate ndash 008 ndash 006 ndash 005 ndash 005 ndash 005 001
Real short-term interest rate ndash 004 ndash 009 ndash 009 ndash 007 ndash 007 000
Unemployment rate 028 075 098 107 115 138
Debt ( of GDP) ndash 037 ndash 121 ndash 192 ndash 259 ndash 329 ndash 763
Deficit ( of GDP) ndash 100 ndash 083 ndash 074 ndash 073 ndash 082 ndash 086
Source Commission services
Table IV11
Permanent increase in corporate tax of 1 of GDP
change from baseline 1st year 2nd year 3rd year 4th year 5th year 10th year
GDP ndash 034 ndash 023 ndash 023 ndash 027 ndash 031 ndash 009
Private consumption 085 138 137 130 125 147
Private investment ndash 424 ndash 529 ndash 518 ndash 501 ndash 496 ndash 396
Real wage costs ndash 013 ndash 025 ndash 025 ndash 029 ndash 040 ndash 132
Real effective exchange rate 010 011 007 003 ndash 001 003
Absolute change from baseline
Short-term interest rate ndash 005 ndash 004 ndash 003 ndash 003 ndash 001 003
Real short-term interest rate 001 ndash 003 ndash 004 ndash 005 ndash 004 005
Unemployment rate 008 005 002 001 ndash 001 ndash 081
Debt ( of GDP) ndash 044 ndash 163 ndash 280 ndash 395 ndash 511 ndash 956
Deficit ( of GDP) ndash 112 ndash 116 ndash 118 ndash 122 ndash 121 ndash 066
Source Commission services
yen1part Higher future disposable income is associated with the lower future taxesthat become possible after the consolidation under the assumption that thedebt ratio is stabilised at a 10 of GDP lower level
160
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
medium term and total employment recovers graduallyThe cross-country evidence on expansionary consolida-tions supports this view As found in the previous sectionreductions in the wage bill of the public sector are morelikely to be linked with expansionary consolidations whenthese are defined in terms of trend growth
431 Temporary versus permanent cuts
One of the regularities found in cross-country analysesof fiscal consolidations is that lsquosuccessfulrsquo consolida-tions leading to a permanent reduction in debt are morelikely to produce expansionary effects than adjustmentswhich are reversed in the subsequent years All the
QUEST simulations presented so far concerned fiscalconsolidations resulting from permanent tax increases orexpenditure cuts The hypothesis of a permanent adjust-ment is crucial Since the QUEST model builds upon theassumption of rational expectations agents anticipatethe permanent reduction in government spending and thelower tax liabilities even though these will only materi-alise in the future
To test the importance of the credibility of the adjust-ment (proxied here by its permanence) and how thisfeeds through the channel of agentsrsquo expectations thehypothesis of permanent consolidation is relaxed the
Table IV12
Permanent increase in VAT of 1 of GDP
change from baseline 1st year 2nd year 3rd year 4th year 5th year 10th year
GDP ndash 014 ndash 021 ndash 034 ndash 044 ndash 051 ndash 063
Private consumption ndash 068 ndash 023 ndash 029 ndash 036 ndash 044 ndash 051
Private investment ndash 015 ndash 051 ndash 080 ndash 097 ndash 112 ndash 133
Real wage costs 049 069 050 037 038 ndash 006
Real effective exchange rate ndash 008 ndash 018 ndash 026 ndash 031 ndash 035 ndash 043
Absolute change from baseline
Short-term interest rate ndash 006 ndash 003 ndash 002 ndash 002 ndash 002 003
Real short-term interest rate ndash 009 ndash 008 ndash 006 ndash 004 ndash 004 003
Unemployment rate 016 046 061 068 073 074
Debt ( of GDP) ndash 049 ndash 137 ndash 215 ndash 291 ndash 371 ndash 805
Deficit ( of GDP) ndash 093 ndash 087 ndash 082 ndash 083 ndash 09 ndash 081
Source Commission services
Table IV13
Permanent reduction in government purchases of 1 of GDP
change from baseline 1st year 2nd year 3rd year 4th year 5th year 10th year
GDP ndash 033 ndash 006 ndash 004 ndash 005 ndash 004 041
Private consumption 140 211 214 212 212 255
Private investment ndash 063 ndash 085 ndash 086 ndash 084 ndash 081 015
Real wage costs ndash 007 ndash 010 ndash 005 ndash 005 ndash 010 ndash 079
Real effective exchange rate 002 001 002 004 005 037
Absolute change from baseline
Short-term interest rate 000 002 002 003 004 007
Real short-term interest rate ndash 004 001 002 002 003 010
Unemployment rate 011 005 002 001 ndash 001 ndash 082
Debt ( of GDP) ndash 047 ndash 179 ndash 297 ndash 415 ndash 534 ndash 970
Deficit ( of GDP) ndash 113 ndash 117 ndash 12 ndash 123 ndash 122 ndash 061
Source Commission services
161
P u b l i c f i n a n c e s i n E M U 2 0 0 3
following simulations concern therefore expenditurecuts of the same type and size as those considered previ-ously (Tables IV13ndashIV15) but are temporary that isare reversed in the subsequent year
Results are reported in Table IV16 and show that whenconsolidations are temporary (and perceived by economicagents as likely to be reversed) they tend to produce largercontractionary effects in the short run This is due to thefact that since consolidations are only temporary it willbe unlikely that future tax liabilities will be reduced Con-sequently since agents will not expect any change in their
future income no wealth effects will materialise in thiscase to offset the contractionary fiscal stance
In all the three cases considered (reduction of govern-ment purchases in government transfers and in gov-ernment employment) without any rise in permanentincome there is now a more pronounced reduction inconsumer spending as current income declines Giventhe real interest rates decline following the markedcontraction in GDP investment is likely to increasesomewhat but not enough to offset the negative effectsof the consolidation on output
Table IV14
Permanent reduction in government transfers to households of 1 of GDP
change from baseline 1st year 2nd year 3rd year 4th year 5th year 10th year
GDP ndash 020 ndash 015 ndash 008 ndash 006 ndash 006 019
Private consumption ndash 027 ndash 027 ndash 023 ndash 022 ndash 022 013
Private investment ndash 065 ndash 060 ndash 049 ndash 047 ndash 048 ndash 002
Real wage costs ndash 009 ndash 014 ndash 007 ndash 004 ndash 003 ndash 058
Real effective exchange rate 008 015 018 019 019 034
Absolute change from baseline
Short-term interest rate 001 001 000 000 001 004
Real short-term interest rate 008 004 001 000 000 007
Unemployment rate 004 004 003 003 003 ndash 046
Debt ( of GDP) ndash 047 ndash 148 ndash 252 ndash 354 ndash 458 ndash 900
Deficit ( of GDP) ndash 100 ndash 102 ndash 103 ndash 107 ndash 109 ndash 071
Source Commission services
Table IV15
Permanent reduction in spending on government employment of 1 of GDP
change from baseline 1st year 2nd year 3rd year 4th year 5th year 10th year
GDP ndash 093 ndash 059 ndash 02 002 016 063
Private consumption 087 121 146 159 166 206
Private investment ndash 100 ndash 031 049 093 116 193
Real wage costs ndash 141 ndash 197 ndash 140 ndash 104 ndash 084 ndash 112
Real effective exchange rate 001 029 053 069 079 120
Absolute change from baseline
Short-term interest rate 014 010 007 005 004 007
Real short-term interest rate 028 026 017 012 008 011
Unemployment rate 148 065 023 007 002 ndash 050
Debt ( of GDP) 028 ndash 055 ndash 164 ndash 277 ndash 392 ndash 880
Deficit ( of GDP) ndash 052 ndash 081 ndash 101 ndash 110 ndash 116 ndash 076
Source Commission services
162
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
432 Accommodating monetary stance
One of the findings in the episodes analysis was thathalf the expansionary fiscal consolidations appearedto have been accompanied by a possible effectivemonetary relaxation (that is falling real interestrates) All scenarios described above were character-ised by a rise in real interest rates (lsquopurersquo expansion-ary fiscal consolidations in terms of the episode anal-ysis in the previous section) By contrast the scenariospresented in Table IV17 assume a monetary policy rule(monetary targeting) consistent with a small fall in realinterest rates on impact (1)
This reduction in interest rates reduces the negativeimpact of the fiscal consolidations and helps to boostgrowth in all cases considered Private consumptionregisters a bigger increase in the case of cuts in govern-ment purchases and government employment and a
smaller reduction in the case of cuts in governmenttransfers Moreover investment is boosted further bythe fact that real interest rates fall It is to note that inthe case of cuts in government purchases or transfersfiscal consolidations accompanied by a supportivemonetary stance appear to have expansionary effectsalready on impact
As in previous simulations after some negative effectsin the first years of the adjustment the largest benefits onoutput appear in the case of cuts in public employmentgiven the sizeable effects on investment
44 Summary of findings
The scenarios in this chapter have illustrated the realeffects of fiscal consolidations and shown under whatcircumstances these effects can be expansionary on out-put A number of results emerge
bull The impact on output of budgetary consolidationdepends on whether it takes place on the revenue orexpenditure side Tax increases are likely to have anegative impact on output both in the short and
Table IV16
Temporary expenditure cuts (1 of GDP)
1st year 2nd year 3rd year 4th year 5th year 10th year
Reduction in government purchases
GDP ndash 075 014 007 004 003 001
Private consumption ndash 012 009 007 005 006 017
Private investment 058 033 008 ndash 001 ndash 001 023
Real short-term interest rate ndash 017 ndash 004 ndash 002 ndash 000 000 001
Reduction in government transfers
GDP ndash 021 004 002 001 002 012
Private consumption ndash 056 003 003 003 004 014
Private investment 013 006 ndash 000 ndash 002 ndash 001 022
Real short-term interest rate ndash 004 ndash 001 000 000 001 001
Reduction in spending on government employment
GDP ndash 121 021 023 014 010 012
Private consumption ndash 036 009 017 011 009 013
Private investment 045 063 042 022 011 018
Real short-term interest rate 005 ndash 004 ndash 005 ndash 003 ndash 001 001
NB Figures refer to changes from baseline except in the case of real short-term interest rates where changes are reported in absolute terms
Source Commission services
yen1part See Giudice Turrini and inrsquot Veld (2003) for details on the monetary ruleassumed in the simulations reported in Table IV17 Alternatively the fall ininterest rates could be interpreted as linked to a reduction in risk premiumsFor highly indebted countries a credible fiscal consolidation could lead toa reassessment of the marketsrsquo perceptions of the risks involved and lead toan elimination or at least a reduction of a risk premium on that countriesrsquobonds
163
P u b l i c f i n a n c e s i n E M U 2 0 0 3
medium run By contrast the short-run negativeimpact on output of permanent expenditure cuts islikely to turn positive in the medium to long runThis may be the result of non-Keynesian featuresvia the anticipated effects of higher future disposa-ble income or profitability
bull The short-run immediate negative impact on outputwill be smaller in case of permanent as opposed totemporary expenditure reductions
bull The consumption channel is a major offsettingforce to the standard Keynesian effects but theinvestment channel can also be of great relevancefor consolidations occurring through cuts in thegovernment wage bill
bull The expansionary effects of fiscal consolidationsoccurring both through the consumption or theinvestment channel are likely to be reinforcedwhen the fiscal consolidations are associated witha favourable monetary stance or with structural
reforms improving the efficiency of factor andproduct markets (see Box IV2) Under such cir-cumstances positive effects on output of the fis-cal consolidation may emerge already in the shortrun
Overall given the limited impact on output in the shortrun in spite of a budgetary adjustment equivalent to1 of GDP (which is larger than what currentlyplanned by Member States at this juncture) theseresults are in contrast with the assertion that fiscal con-solidation should be avoided during slowdowns andshow that sizeable positive effects could materialise inthe medium to long run
The QUEST simulations performed consider some butnot all the possible channels through which consolida-tions could have expansionary effects Further work mayattempt to also take into account the effects of consolida-tions on the risk premia on countriesrsquo public debt thatneed in this case to be determined endogenously in themodel as a function of debt levels
Table IV17
Permanent expenditure cuts (1 of GDP) with accommodating monetary stance
1st year 2nd year 3rd year 4th year 5th year 10th year
Reduction in government purchases
GDP 026 040 025 016 009 032
Private consumption 154 236 227 222 218 251
Private investment 100 005 ndash 022 ndash 042 ndash 057 ndash 010
Real short-term interest rate ndash 054 ndash 004 ndash 003 ndash 002 ndash 001 008
Reduction in government transfers
GDP 035 030 021 017 011 017
Private consumption ndash 015 ndash 004 ndash 010 ndash 011 ndash 014 011
Private investment 084 026 017 -001 ndash 016 ndash 014
Real short-term interest rate ndash 037 001 ndash 002 -003 ndash 002 005
Reduction in government employment
GDP ndash 050 ndash 036 ndash 012 007 017 053
Private consumption 095 131 147 158 163 197
Private investment 017 005 068 100 118 166
Real short-term interest rate ndash 020 022 015 010 007 010
NB Figures refer to changes from baseline except in the case of real short-term interest rates where changes are reported in absolute terms
Source Commission services
164
P a r t I VC a n f i s c a l c o n s o l i d a t i o n s i n E M U
b e e x p a n s i o n a r y
Box IV2 Fiscal consolidation as catalyst for structural reforms in Germany
(Continued on the next page)
The evidence suggests that fiscal consolidations based on expenditure cuts are more likely to be expansionary than consol-idations based on tax increases The mechanism described in the existing literature (for example Alesina et al 2002)through which expenditure-based consolidations can produce expansionary effects emphasises the role of increased profitexpectations associated with reduced wage pressure A further reason why expenditure-based fiscal consolidations mayhave non-Keynesian effects is that they accompany and quite often induce the realisation of structural reforms in labourand product markets The positive impact on output of fiscal consolidations may thus also be associated with expectedimproved profit opportunities coming from a better functioning of markets and from productivity gains that are the resultof structural reforms
In order to capture the overall impact of fiscal consolidations that occur together with structural reforms some simulationsare performed in conjunction with shocks to the parameters of the model that reproduce the effect of structural reformsThe simulations are made using the QUEST model concerning Germany and are similar to the simulations shown inTable IV13 to Table IV15 (respectively reduction in government purchases transfers and employment) The shocks rep-resenting structural reforms are based on those presented in European Commission (2002b) they concern reforms in boththe labour and product markets and reforms inducing an increase in overall efficiency Structural reforms in the labour mar-ket are modelled as an lsquoemployment-friendlyrsquo shift of the wage-setting curve encouraging employment participation andreducing the wage mark-ups Reforms in the products markets are modelled instead via a fall in the price mark-ups reflect-ing an improvement in competitive conditions These shocks are further combined with an increase the total factor produc-tivity parameter capturing an improvement in the overall productive efficiency (1)
(1) The size of the shocks to model parameters is one third of that assumed in the simulations presented in the European Commission (2002 b) wherelabour market reforms product market reforms and productivity improvements were analysed separately In the present simulations the assumedreduction of wages along the wage-setting curve is by 13 of a percentage point as it is the supposed increase in TFP As for the reduction in mark-upsit is assumed to be equal to 16 of a percentage point
Graph IV2 Expenditure-based fiscal consolidations and structural reforms effects on GDP ( change from baseline)
ndash 05
0
05
1
15
2
1st year 2nd year 5th year 10th yearWithout structural reforms With structural reforms
165
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Box IV2 (continued)
Graph IV2 reports the percentage change in baseline GDP when expenditure cuts occur with and without such structuralreforms Results show that when expenditure cuts are accompanied by structural reforms the effect of consolidations onoutput becomes generally positive already in the short run In all the simulations performed after two years GDP increasesby about percentage points compared with the baseline In the long run (after 10 years) the effect on output of expendi-ture-based fiscal consolidations is positive irrespective of whether they are accompanied by structural reforms When struc-tural reforms take place however the impact on output is about four times higher instead of being about 05 percentagepoints it amounts to around 2 percentage points
In sum it seems that expenditure-based consolidations accompanied by structural reforms in labour and product marketscould have a very positive impact on output and growth not only in the medium (where it may be very large) but also inthe short run
ndash 05
0
05
1
15
2
1st year 2nd year 5th year 10th year
ndash 1
ndash 05
0
05
1
15
2
25
1st year 2nd year 5th year 10th year
Without structural reforms With structural reforms
Without structural reforms With structural reforms
166
Part V
Meeting the EUrsquos budgetary requirementsnational expenditure rules and fiscal relations across levels of government
Summary
The EUrsquos fiscal rules impose important and challengingbudgetary obligations on Member States Countries arerequired by the Treaty to avoid excessive deficit posi-tions (defined against a reference value for deficits of3 of GDP) and under the Stability and Growth Pactthey are required to achieve and maintain a budget posi-tion of lsquoclose to balance or in surplusrsquo These EU fiscalrules focus on the budget balance that is the differencebetween total revenues and total expenditures and not onthe level or the composition of the two
During the 1990s and in particular since 1997 most EUcountries introduced expenditure rules There is a greatdeal of variety in their design as regards the types ofspending items covered by a rule how the rule is defined(in real or nominal terms as a ceiling or a rate ofgrowth) the time frame involved and the robustness ofsurveillance and enforcement mechanisms In the major-ity of Member States expenditure rules feed into thebudgetary-setting process (rather than representing abinding obligation which must be respected) and ex postcontrol and implementation mechanisms are ratherweak While expenditure rules for the most part wereintroduced with national policy objectives in mind theycan also enable Member States to meet the budget bal-ance requirements of the Treaty and SGP by helpingthem to better control expenditure items that are subjectto overruns Depending on their design they can alsocontribute to other policy objectives such as avoiding apro-cyclical loosening of fiscal policy in good times (viaa discretionary increase in public spending) and improv-ing the quality of the composition of public spending
Preliminary empirical analysis indicates that the existingrules have not had a significant impact on trends in pub-lic spending Judging compliance with expenditurerules however is difficult as in many cases they coverseveral years and are subject to revisions In some coun-tries expenditure rules are not ambitious enough andadherence with them is easily achieved in other casesthe rule has been adjusted or abandoned if perceived as
being too ambitious Nonetheless even a relatively weakexpenditure rule can provide useful guidance and signalsto actors involved in the budgetary process
The Treaty and SGP requirements are defined in terms ofthe budget balance of the general government (that iscentral and localstate governments and social security)although the specific budget targets in stability and con-vergence programmes are set by the central governmentThe challenge in meeting EU budgetary requirements istherefore affected by the way in which Member Statesallocate fiscal functions (both revenues and expendi-tures) across different levels of government This isespecially the case in federal countries and the MemberStates where local authorities have considerable budget-ary autonomy The contribution of sub-central authori-ties to the overall budget position is changing in anumber of countries in light of efforts to devolve certainpublic functions to regionallocal authorities
The direct contribution of lower levels of government tothe general government deficit is generally limited sinceall Member States apply restrictions to local governmentborrowing the exception is Germany where net borrow-ing by local and state governments accounts for nearlyhalf of the general government budget deficit in 2002However it should be borne in mind that de facto centralgovernments often have to bear the cost of financing dif-ficulties that emerge at sub-central level To help complywith the EUrsquos fiscal rules federal Member States andItaly and Spain have recently introduced arrangementsthat aim at coordinating the budgetary position acrosslevels of government (usually referred to as national sta-bility pacts) More experience with the implementationof these arrangements is needed before conclusions canbe drawn on their effectiveness in contributing to theobjectives of the EU fiscal framework A priori a stronglegal base and enforcement mechanism would beexpected to contribute to the credibility and effective-ness of the arrangements
169
P u b l i c f i n a n c e s i n E M U 2 0 0 3
The process of decentralising responsibility for some pol-icies raises a second issue in the context of EMU namelythe operation of automatic stabilisers Experience showsthat in general systems are designed to shield sub-national governments from cyclical variations However
empirical evidence for the US and Germany suggestssome degree of procyclical behaviour at state level Fur-ther research would be needed however before policyconclusions can be drawn on the interaction of fiscaldecentralisation and automatic stabilisation
170
1 Introduction
The EUrsquos fiscal rules impose important and challengingbudgetary obligations on Member States Countries arerequired by the Treaty to avoid excessive deficit posi-tions (defined against a reference value for deficits of3 of GDP) and under the Stability and Growth Pactthey are required to achieve and maintain a budget posi-tion of lsquoclose to balance or in surplusrsquo These EU fiscalrules focus on the budget balance ie the differencebetween total revenues and total expenditures and not onthe level or the composition of the two
A further important feature of the EUrsquos fiscal rules isthat notwithstanding the fact that budgetary commit-ments are given at Community level by the central gov-ernment the requirements in terms of the budget balanceconcern the general government this covers central andlocal (state) governments and social security that is itdoes not distinguish between the allocation of fiscalunbalances across different levels of government butonly looks at the overall budgetary position It is theresponsibility of Member States to organise their fiscalrelations across different levels and sectors of govern-ment so as to ensure that they can meet the budgetaryrequirements set down in the Treaty and SGP
This part examines some of the challenges which Mem-ber States face in complying with the EUrsquos fiscal rulesand also analyses a number of policy instruments that arebeing developed to this end
Section 2 considers the role and effectiveness of nationalexpenditure rules which many Member States haveintroduced in recent years with the purpose of establish-ing a better control of public spending It analyses thedesign of expenditure rules across Member States andconsiders how they relate to the EU framework and cancontribute to the objective of sound and sustainable pub-lic finances Particular attention is paid to a preliminaryevaluation of how rules worked in practice during thefirst years of application
Section 3 considers the issue of fiscal decentralisation Itprovides a brief overview of the allocation of responsi-bility for public expenditure and revenues items acrossdifferent levels of government and then examines howthis interacts with the EU framework for fiscal surveil-lance Firstly attention is paid to the contribution of eachlevel of government to the budget balance of the generalgovernment as a whole Consideration is given to vari-ous institutional arrangements (such as national stabilitypacts) that have been put in place by Member States tocoordinate the budgetary positions across levels of gov-ernment in part to comply with the provisions of theTreaty and SGP Secondly the impact of fiscal decen-tralisation on the potential for automatic stabilisation isexamined Finally the link between the EU fiscal sur-veillance framework and recent institutional reforms isexamined in more detail in case studies on Spain andGermany
171
2 Expenditure rules in EU Member States
21 The need for expenditure rules as a means to control public finances
Since the beginning of 1990s a growing literature hasinvestigated the design of fiscal rules which have beenintroduced in many countries as a response to growingbudgetary imbalances (1) According to Hallerberg et al(2001) lsquoa fiscal rule is a combination of a fiscal targetwith a set of prescriptions of what governments are sup-posed to do to achieve this targetrsquo Any ex ante constraintto budget deliberation could constitute a fiscal rule Eventhe presentation of the budget law or a budget documentwith some political commitments that sets up an ex antetargets for at least the following year could be consideredas a fiscal rule
Kopits and Symanski (1998) define a fiscal rule as lsquoapermanent constraint on fiscal policy expressed in termsof a summary indicator of fiscal performance such as thegovernment budget deficit borrowing debt or a majorcomponent thereofrsquo This definition is narrower becausea fiscal rule should have two specific characteristicsnamely being lsquopermanentrsquo and defined through an lsquoindi-catorrsquo that can be easily monitored
Within the wide spectrum of fiscal rules expenditurerules are of growing relevance in a number of MemberStates (2) Growing recourse to expenditure rules can beexplained by the importance of expenditure control aspart of a successful strategy of budgetary consolida-tion (3) Moreover slippage from agreed budget targetsoften arises on the expenditure side that is more subjectto discretionary actions
The issue is how national budget processes based onexpenditure rules interact with the EU fiscal frameworkNational expenditure rules can act as complementaryinstruments to the EU rules-based framework for severalreasons
bull First their respect does not prevent automatic stabi-lisers to play since the greatest impact of the cycleover the budget is on the revenue side
bull Second appropriate recourse to expenditure rules atnational level could also help contribute to the pol-icy objective of improving the quality of the compo-sition of public spending that is help restructure thecomposition of spending toward so-called produc-tive items such as investment in infrastructures andR amp D Expenditure rules at national level could bedesigned in a way that places a stricter control onspending on items that are considered as being lessconducive to long-term growth and ensuring thatmore productive items receive more favourableconsideration in budgetary consolidation efforts
bull Third expenditure rules could also make an impor-tant contribution to broader economic and budgetarypolicy objectives established at EU level as part ofthe BEPGs and the Lisbon strategy For example inits communication on lsquoStrengthening the coordina-tion of budget policiesrsquo (see Part II2 of this report)the Commission suggested that a temporary deterio-ration in the budget balance could be envisaged ifthis is due to large structural reforms However thisshould not compromise the objective of sound andsustainable public finances A close monitoring ofMember Statesrsquo compliance with their own nationalexpenditure rule could reduce the degree of uncer-tainty on the budget balance intertemporal profile
The contribution of different expenditure items to meet-ing overall budgetary objectives is illustrated inGraph V1 This shows in index form the evolution of
yen1part See for a review of the main features of fiscal rules Kell (2001) Kopits andSymansky (1998) Kopits (2001) Inman (1996)
yen2part See for example Brunila (2002) Hallerberg et al (2001) and Mills andQuinet (2001) for some comparison of expenditure rules across EU Mem-ber States
yen3part The 2001 and 2002 Broad Economic Policy Guidelines recommendedMember States lsquoto introduce or enhance mechanisms that help assess andcontrol spending including budgetary proceduresrsquo
172
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
different components of primary expenditure as a shareof GDP over the last 10 years for EU-15 countries as awhole On average the level of primary expendituredecreased by almost 9 between 1992 and 2000 whileit increased by 5 during the subsequent two yearsWhereas spending on compensation of public employ-ees and gross fixed capital formation fell significantlyover the period spending on intermediate consumption(purchases of goods and services by the public adminis-tration) and social transfers other than in kind actuallyincreased (1) These different trends across expenditureitems illustrate that it is more difficult to curtail spend-ing on particular categories of expenditure
Table V1 examines expenditure trends for several Mem-ber States for the 1998ndash2001 period Public expenditure isclassified according to seven main lsquofunctionsrsquo of the Stateseveral functions (such as defence public order andsafety general public services) refer to the core activitiesof the State for example those functions that are at thebase of the functioning of a modern State (2) Other
expenditure programmes aim at addressing market fail-ures These programmes include mainly education andhealthcare but also economic services Finally part ofpublic expenditure has primary a redistribution roleachieved through social protection programmes
The numbers highlighted in bold show spending itemswhich changed by more than total public spending (lastcolumn of the Table V1) (3) Clearly it should be takenin mind that the expenditure-to-GDP ratios differ mark-edly across different functions social protection repre-sents the greatest share of expenditure (around one thirdof total expenditure) while defence public order andsafety rarely reach more than 2 of GDP
With the exception of Austria spending on healthcare inall countries increased substantially more than totalexpenditure during the last years For instance totalexpenditure in Italy contracted by 3 between 1998
Graph V1 Trends in different items of public expenditure at EU level
Source Commission services
70
80
90
100
110
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Social transfers other than in kind Compensation of employees Intermediate consumptionGross fixed capital formation Primary expenditure
ExpenditureGDP (1992=100)
yen1part Subsidies and lsquoother current and capital expendituresrsquo are not presenteddue to their low share of GDP
yen2part The so-called COFOG classification See European Commission (2002a)for a detailed explanation of the functional classification
yen3part The residual component (lsquoothersrsquo) in Table V1 includes limited expendituresas environmental protection community amenities and religious expenses
173
P u b l i c f i n a n c e s i n E M U 2 0 0 3
and 2001 whereas healthcare expenditure increased by105 In Portugal spending on healthcare grew at twicethe pace of total expenditure while in Belgium despitethe strong decrease of total expenditure (ndash 95 ) health-care increased by 82 However healthcare expendi-
ture is not the only item that presents such a dynamicSpending on education and social protection also tendedto increase faster than total expenditure albeit by a lesseramount and in a smaller group of countries
22 The design and implementation of expenditure rules
221 The design of expenditure rules
According to the standard theory of economic policy(Tinbergen 1956) expenditure targets should be strictlyunder the control of the government since their develop-ment represents an intermediate objective to reach thefinal aim of budgetary control Therefore when design-ing an expenditure rule several choices have to be madeas follows
bull whether the rule should target an outcome of a par-ticular expenditure item (or an aggregate of expend-iture items) or whether it should target the effects ofa particular policy measure
bull whether certain expenditure items should beexcluded from an aggregate expenditure target
bull whether an expenditure target should be defined innominal or real terms
bull whether the target should be defined in terms of alevel of expenditure or as a rate of change
bull the time span of the target
Targeting an expenditure outcome or the impact of apolicy measure A rule that targets the expenditure out-come can be defined as follows
[1]
where is the targeted outcome of expenditure item Eat time t is the projected level of expenditure intime t forecasted at time t ndash 1 and is the policyaction planned in t ndash 1 to correct the trend and thereforeto achieve the target
The outcome Et is defined as
[2]
Table V1
Trends in public expenditure items in selected EU countries (variation in percentage points between 1998 and 2001)
General public
services
Economic affairs
Health EducationSocial
protection
Defence public order and safety
Others Total
BE ndash 238 ndash 170 82 ndash 46 ndash 74 ndash 97 167 ndash 95
DK 00 ndash 125 39 105 ndash 16 00 91 03
DE ndash 60 49 00 ndash 45 ndash 05 ndash 67 43 ndash 10
EL ndash 90 1000 83 27 12 48 00 08
IT ndash 167 ndash 24 105 20 ndash 11 ndash 63 ndash 37 ndash 30
NL ndash 110 188 25 21 ndash 60 33 31 ndash 17
AT ndash 34 78 ndash 275 ndash 33 23 ndash 80 00 ndash 35
PT 15 34 115 30 55 ndash 27 120 50
FI ndash 72 ndash 148 17 ndash 31 ndash 84 ndash 94 ndash 160 ndash 72
NB Figures in the table show the cumulated change of expenditure-to-GDP ratios for each spending category and for total expenditure (last column) For instance the100 increase registered in Greece for economic affairs spending-to-GDP ratio implies that expenditure doubled between 1998 and 2001
Source Commission services
Et
Et1ndash( )
Pt1ndash( )+=
Et
Et1ndash( )
Pt1ndash( )
Et
Et Et1ndash( )
Pt δ+ +=
174
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
where Pt is the policy action as it results ex-post and δ isthe forecast error on Et which is positive when expendi-ture turns out higher than forecasted
Assuming that the outcome Et equals the expectedoutcome it can be shown that
[3]
If the trend expenditure turns out to be lower thanexpected (ie the forecast error is negative such thatδ lt 0) the ex-post policy correction neededto reach the target can be smaller than what wasplanned ex ante On the contrary if expenditure growsfaster than expected (and δ gt 0) then the expendituretarget ( ) will be overshot even if the planned policyaction ( ) is implemented in full
The formulae show that the following steps are neededto define an expenditure rule first the definition of theexpenditure item (or aggregate of expenditures) to tar-get second a forecast of the trend in targeted expendi-ture item third an ex-ante quantification of the policyaction necessary to achieve the target fourth the possi-bility to accurately verify ex-post compliance In partic-ular they highlight the central importance of definingthe appropriate target For instance attention should bepaid as to whether to include high volatile items in anaggregate expenditure rule as a high forecast error δ willaffect the outcome rendering the expenditure rule lesseffective for the purpose of budgetary control (1)
An alternative approach to setting a target in terms of thelevel of expenditure(s) would be to directly target theimpact a particular policy action (P) as follows
[4]
Under this approach the expenditure rule is respected ifthe policy maker implements the announced correctivemeasures irrespective of the actual outcome interms of expenditure The advantage of directly targetingthe specific policy action is that it is not necessary to takeaccount the economic cycle or other exogenous factorsthat affect the outcome However in practice it may bedifficult to quantify ex ante the precise budgetary impact
of the planned policy measures rendering it difficult toassess compliance
Whether to exclude certain items from the aggregateexpenditure target While the main purpose of an aggre-gate expenditure rule is to contribute to sound publicfinance positions certain categories of expendituresmight be excluded from the target so as to ensure consist-ency with other public policy goals At least three cate-gories of expenditure items warrant consideration in thisregard
bull Firstly it may be appropriate to exclude interestpayments from the target and focus on primaryexpenditure which is more under the discretionarycontrol of government The inclusion of interestpayments within the expenditure target increases therole of forecasts errors Other things being equal arule that targets an aggregate that includes interestpayments can be fulfilled with a lower policy effortif interest payments are overestimated (2)
bull Secondly unemployment-related transfers could beexcluded from the target to prevent pro-cyclicalbehaviour For example the rigid adherence to anominal expenditure target that includes unemploy-ment transfers in periods of low growth would cete-ris paribus result in a tightening of the fiscal stanceas de facto it would prevent part of the automatic sta-bilisers from operating
bull Thirdly one may wish to exclude specific categoriesof productive public spending (such as publicinvestment) from an expenditure target this wouldprevent corrective measures needed to achieve theexpenditure target from affecting these desirablepublic expenditure items
yen1part On the importance of forecast errors in the functioning of fiscal rules seealso Auerbach (1994)
Et
Pt Pt1ndash( ) δ+=
Pt Pt1ndash( )ndash( )
Et
Et
Pt1ndash( )
Pt Pt1ndash( )
Pforall=
Pt1ndash( ) yen2part Equation [1] when interest payments are targeted can be rewritten as
[1a]where expenditure E is now the sum between interest payments I andanother primary expenditure item A The target is the sum of thetargeted outcome on item A and the outcome on interest payments Theoutcome at time t and can be rewritten as
[2a]Therefore the outcome depends on the forecasted levels for A and I theex-post policy action and the forecast error This has twocomponents one is primary expenditure and one is interest paymentsIf the outcome equals the targeted expenditure level after some passages itresults that
[3a]
At
It1ndash( )+( ) At
1ndash( )It
1ndash( )+( ) Pt1ndash( )+=
At
It+( )
At It+( ) At1ndash( )
It1ndash( )+( ) Pt ε It It
1ndash( )ndash( )+ + +=
ε It It1ndash( )ndash( )+
Pt1ndash( )
Ptndash ε It It1ndash( )ndash( )+=
175
P u b l i c f i n a n c e s i n E M U 2 0 0 3
The definition of the target in nominal or real termsThe difference between a real or a nominal target is rel-evant in the case of forecast errors in inflation projec-tions (1) In fact price deflators differ across governmentexpenditures items and they also differ from GDP defla-tors A target defined in nominal terms has the advantageof transparency making monitoring easier It can alsohelp to keep expenditure under control through a higher-than-expected adjustment if the inflation outcome ishigher than expected However if the rule has defined anlsquoescapersquo clause so that higher-than-expected inflationshould not force a higher real adjustment in order to fulfilthe nominal target then a nominal rule risks to producea bias such that a lower-than-expected inflation allows alower adjustment effort while a higher-than-expectedinflation does not entail a stronger adjustment On thecontrary if the target is defined in real terms complianceis not affected by inflationary developments but it canresult in it being more difficult to measure the compli-ance
The definition of the target in terms of levels (absolutevalues or as a share of GDP) or as a rate of growthWhen the target is defined as a share of GDP the resultcan depend on GDP developments and in particular onGDP forecast errors Thus the rule might turn out to bepro-cyclical since the expenditure ceiling fluctuates inline with GDP around its trend This problem can beovercome by formulating the target as a fixed rate ofgrowth in the expenditure item(s) or as an absolutelevel
The time span covered (2) A multiannual rule is gener-ally superior to a rule where the target is fixed for onlyone year This is because an annual rule can be more eas-ily circumvented simply by postponing expenditures tothe first day of the following budget year and is suscep-tible to accounting practices When the target is fixed exante for several years the possibility to postpone expen-
ditures or structural adjustment to the future becomesmore difficult
222 The implementation of expenditure rules
The implementation and ex-post assessment mecha-nisms are constituent elements of an expenditure ruleSeveral elements warrant consideration (Kopits andSymanski 1998) the availability of instruments tomonitor and if necessary correct the dynamic of budg-etary position during budget execution provisions todeal with non-compliance including sanctions escapeclauses when failures in respecting the rule are beyondpolicy control
Implementation mechanisms can have different formsand degrees of enforcement from automatic contin-gency measures once the deviation from the targetappears to more flexible (and weak) measures such asnon-binding suggestions for discretionary corrections
The availability of data is essential in order to monitorand control the budget execution Data on budgetaryaggregates (such as budget balances or total expendi-tures) are often easier to collect and subject to less revi-sions than information on specific expenditure itemswhere different spending units of the government areinvolved
To avoid a pro-cyclical or perverse outcome an expend-iture rule should usefully define provisions for cateringfor worse-than-expected economic conditions andorother unexpected events (such as a flood) that requireadditional spending However while flexibility of sucha nature is essential in the case of budget balance rulesthe need for such provisions is less evident in the case ofexpenditure rules as the sensitivity of most expenditureitems is very limited (apart from exceptions such asunemployment transfers)
Sanctions in the case of non-compliance with the targetshould always be defined ex ante to make the rule cred-ible and enforceable (Inman 1996) These could takedifferent forms such as an obligation to amend thebudget law (3) automatic sequesters if there is clearinformation that the target is not going to be fulfilledduring the budget year or pecuniary sanctions imposedby a higher level of government While the existence ofwell-defined sanctions is only a necessary condition to
yen1part See Brunila and Kinnunen (2002)yen2part A clear example of the importance of the time span of spending rules is the
experience of the Budget Enforcement Act (BEA) endorsed in 1990 in theUS In brief a series of annual caps on public spending were set for theperiod 1990ndash95 and the rule specifies that lsquoenacted policies cannot raise thedeficit relative to initial projected levelsrsquo (Poterba 199624) A key differ-ence with the Gramm-Rudman-Hollings (GRH) rule that has been in forceduring the second half of 1980s in the US is that the latter fixed targetsyear-by-year on the basis of the outcome of the previous year and thereforeit was easier to postpone expenditure to future fiscal years Auerbach(1994) presents evidence that during the last years of the GRH there was atendency to reduce deficits of the current year at the expense of the follow-ing year yen3part This is the mechanism implied by the GRH rule see Gramlich (1990)
176
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
make the rule credible it is not a sufficient one To befully credible the sanctions should have a legal or con-stitutional basis and should not be based on politicalcommitment alone
In addition an expenditure rule is more credible if thereis an independent authority that monitors the develop-ment of the budgetary position and that is in charge ofthe enforcement measures including the application ofthe sanctions Without a clear legal basis andor wherethere are no clear defined sanctions the penalty for non-compliance with the rule is only reputational
223 A taxonomy of expenditure rules
It is possible to classify expenditure rules according totheir degree of strictness (1) Following the terminologyof Poterba (1996) expenditure rules are classified asbeing either lsquonarrowrsquo or lsquoweakrsquo see Table V2
An expenditure rule is classified as lsquoweakrsquo if
bull the target includes volatile expenditure items so thatthe actual level of spending is subject to a forecasterror and is only partly influenced by the applicationof spending control mechanisms
bull the target includes interest payments In this casethe target could be respected without the govern-ment taking policy actions simply if interest pay-ments develop favourably (2)
bull there is no ex-post control mechanism to verifywhether ex ante targets have been respected and ifenforcement mechanisms are based on a politicalcommitment rather than legal provisions (3)
In contrast a rule is classified as being lsquonarrowrsquo if
bull it targets a less volatile expenditure item(s) and ittargets the budgetary impact of the policy actionrather than the final outcome
bull there are well-defined mechanisms for carrying outan ex-post verification of compliance with targets
bull enforcement mechanisms and sanctions are definedex ante
bull surveillance is carried out by an independent author-ity that can enforce authorities to respect the rule
23 National expenditure rules
231 Main features of expenditure rules within EU Member States
Almost all EU countries have put in place rules to controlwide aggregates of expenditures Table V3 shows themain features of expenditure rules currently in place in
yen1part As reported by Inman (1996) a fiscal index called an lsquoACIR stringentindexrsquo has been developed to measure the tightness of the statersquos budgetbalance rules constraint in US states with higher values indicating a morestringent constraint The index is a composite measure where several fea-tures of the fiscal rule are considered and in particular it awards points forwhether the rule lsquorequires the governor to submit a balanced budget(1 point) requires the legislature to pass a balanced budget (2 points)allows the state to carry a deficit into the next fiscal year (4 points) doesnot allow the state to carry a deficit into the next fiscal year (6 points if abiennium budget 8 points if an annual budget)rsquo (Inman 19967) In addi-tion additional extra points are awarded if the fiscal rule is based on a leg-islative or constitutional instrument
yen2part However Mills and Quinet (2001) underline that since the main goal of anexpenditure rule is to make the objectives of a decreasing debt and a lowertax burden mutually compatible interest payments should be kept withinthe targeted aggregate
yen3part See Bohn and Inman (1996) and Kopits and Symansky (1998)
Table V2
A taxonomy of expenditure rules
Weak Narrow
Design Aggregate expenditures including interest payments Specific target on non-volatile item(s)
Specific target on volatile items Target on variation in levels
Implementation Only ex ante target Ex ante target ex post control
Internal surveillance Surveillance by an independent authority
Statutory instrument Political commitment Constitutional or legal basis
Enforcement mechanism Reputational or economically insignificant sanctions Economically significant (but not excessive) sanctions
177
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Member States (1) The table describes for each countrythe item targeted the definition of the target (either inreal or nominal terms as a ceiling or as a rate of change)the level of application (general central or local govern-ment) the date of introduction of the rule and the timespan It also underlines whether there are measures spec-ified ex ante in case of non-compliance with the rule andpossible lsquoescape clausesrsquo in case of economic shocksFinally it summarises how the rule worked in practiceduring the first years of application
Although the expenditure rules differ substantiallyacross Member States some common features can beidentified (2) First the expenditure rules of many Mem-ber States either as regards their timing or structurewere influenced by the Stability and Growth Pact Apartfrom Spain (which established a ceiling on nominalexpenditure starting from 2003) and Portugal (where thetarget has been introduced in 2002) all other countriesintroduced expenditure rules in past years in particularsince 1997 Such rules have been generally withinmedium-term frameworks in line with the medium-termfocus of the SGP However there are cases where spend-ing rules anticipated the SGP Germany already hadthese kind of rules since the mid-1980s while the Neth-erlands introduced a rule in 1994
Second the target tends to cover a wide aggregate ofexpenditure items that includes all public expendituresIn all cases except Denmark Ireland Greece and Italythe aggregate includes interest payments Public invest-ment is netted out in the case of Denmark and BelgiumItaly recently introduced an expenditure rule (the so-called lsquoexpenditure freezersquo law) whereby all legislationresulting in new or higher public expenditures shouldexplicitly specify the authorised amount (3)
Third almost all Member States apply their expenditurerules to the central government (coupled with borrowing
and budgeting restrictions for lower levels of govern-ment see Section 32) An exception is Germany whereexpenditure rules also apply to the regional and localgovernments In Italy within the context of a domesticstability pact ceilings are established for primary currentexpenditure of regions
Fourth most of the expenditure rules are based on politi-cal commitments rather than legislation This explainswhy in many cases discretionary adjustments to expend-iture rules have taken place when the original targets havestarted to act as a constraint In some cases this has lim-ited the effectiveness of the rule as an instrument to con-trol public finance developments
Besides these common characteristics rules differ acrosscountries in terms of specific design Targets are formu-lated in levels or rates of growth and both in nominal orreal terms These four possibilities can all be found inpractice which illustrates the diversity of arrangementsthat have been put in place Real growth targets can befound in Belgium Denmark and France Nominalgrowth targets are set up in Germany Ireland Italy andLuxembourg Specific ceilings on absolute values (lev-els) have been put in place in Spain Italy the Nether-lands Finland Sweden and the UK Specific arrange-ments exist in Greece and Portugal for publicemployment Here the target is the number of employ-ees rather than financial expenses
Overall expenditure rules in EU countries belong to thecategory of lsquoweakrsquo rules as described in Section 22since in most cases they are based on political commit-ment solely The outcome depends on policy actions indue course as well as on the development of volatileitems included in the targets or on interest paymentsThe implementation and enforcement mechanisms aregenerally less developed since the rules lack a firm legalbasis Sanctions are generally absent or economicallyinsignificant In sum many expenditure rules put inplace in EU countries lack some necessary features to befully credible since they allow for the option to ignoremiss or abandon the rule when a divergence arisesbetween targeted variables and outcomes
232 How have national expenditure rules worked in practice
This section contains a preliminary empirical assessmenton the implementation of national expenditure rules In
yen1part According to information available to the Directorate-General for Eco-nomic and Financial Affairs of the European Commission Either as a spec-ification of the general expenditure rule or as an additional requirementmost Member States also have defined ceilings for individual ministries orspecific spending categories (Hallerberg et al 2001)
yen2part Case studies for several countries (the Netherlands Italy Finland and Swe-den) are carried out in the correspondent country sections of Part VI of thisreport
yen3part However the list of expenditures that can be frozen if the spending ceiling isbreached excludes important items such as pensions public sector wages andunemployment benefits See the country section on Italy in Chapter VI8 ofthis report
178
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
Tab
le V
3
The
fea
ture
s an
d im
plem
enta
tion
of
expe
ndit
ure
rule
s w
ithi
n M
embe
r St
ates
(ge
nera
l tar
gets
)
Exp
endi
ture
item
Def
init
ion
of t
arge
tL
evel
of a
pplic
atio
nD
ate
of in
trod
ucti
onT
ime
span
Act
ion
in c
ase
of n
on-c
ompl
ianc
e
Exc
epti
ons
to r
ule
in c
ase
of e
cono
mic
sh
ocks
Exp
erie
nce
wit
h th
e ru
le
BEPr
imar
y ex
pend
iture
Ann
ual r
eal g
row
th
rate
to 1
5
in
the
med
ium
term
Orig
inal
ly f
eder
al
gove
rnm
ent a
nd
soci
al se
curit
y (e
ntity
1)
From
200
1 on
war
ds f
eder
al
gove
rnm
ent
Firs
t men
tione
d at
end
of 1
998
as
lsquopoi
nt o
f ref
eren
cersquo
Med
ium
term
(tim
e fr
ame
as c
over
ed b
y st
abili
ty p
rogr
amm
e)
No
mea
sure
s spe
cifi
ed
ex a
nte
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed
ex a
nte
Lim
it w
as re
spec
ted
in
2000
and
200
1 b
ut n
ot
in 1
999
Diffi
cult
to
judg
e ad
here
nce
give
n st
atus
of m
ediu
m te
rm
benc
hmar
k
DK
Publ
ic c
onsu
mpt
ion
Ann
ual r
eal g
row
th
rate
to 1
o
n av
erag
e du
ring
1999
ndash200
5
Cent
ral g
over
nmen
tFi
rst m
entio
ned
in
1997
but
bec
ame
fully
bin
ding
in 1
999
Mul
tiann
ual r
ule
(thr
ee
year
s)N
o m
easu
res s
peci
fied
ex
ant
eN
o au
tom
atic
ex
cept
ions
spec
ified
ex
ante
How
ever
di
scre
tiona
ry re
visio
ns
of ta
rget
hav
e ta
ken
plac
e fo
r exa
mpl
e in
20
01 w
hen
targ
et w
as
raise
d fr
om 1
to 2
2
Diffi
cult
to ju
dge
adhe
renc
e g
iven
sp
ecifi
catio
n of
ave
rage
ta
rget
ove
r sev
eral
ye
ars a
nd re
visio
ns o
f th
e ta
rget
dur
ing
that
pe
riod
New
go
vern
men
t is
impl
emen
ting
syst
em
that
aim
s at
recu
pera
ting
slipp
age
in su
bseq
uent
yea
rs
DE
Ove
rall
expe
nditu
reA
nnua
l nom
inal
gr
owth
rate
to b
e ag
reed
on
year
ly b
asis
by F
inan
zpla
nung
srat
(F
PC)
Cent
ral
regi
onal
and
lo
cal g
over
nmen
tsBe
ginn
ing
of 1
980s
Curr
ent a
nd fo
llow
ing
four
yea
rsFr
om 2
004
onw
ards
th
e FP
C w
ould
disc
uss
devi
atio
ns a
nd c
ould
ag
ree
upon
re
com
men
datio
ns
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed e
x an
te H
owev
er
disc
retio
nary
revi
sions
of
targ
ets h
ave
take
n pl
ace
at l
east
in
dow
nsw
ings
Ceili
ng n
ot re
spec
ted
in
2002
it r
emai
ns to
be
seen
how
pos
sible
re
com
men
datio
ns b
y th
e FP
C on
non
-co
mpl
ianc
e w
ould
af
fect
out
com
es
ELCo
mpe
nsat
ion
of e
mpl
oyee
sRe
crui
tmen
t nor
m 5
1
(one
new
recr
uitm
ent
for e
very
five
civ
il se
rvan
ts le
avin
g se
rvic
e) e
xcep
t for
he
alth
edu
catio
n an
d ar
med
forc
es w
here
the
norm
is 1
1
Cent
ral g
over
nmen
t19
97In
defi
nite
No
mea
sure
s spe
cifi
ed
ex a
nte
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed e
x an
te
Polit
ical
com
mitm
ent
not l
egal
ly b
indi
ng
Diffi
cult
to a
sses
s the
im
plem
enta
tion
of th
e re
crui
tmen
t nor
m
(Con
tinu
ed o
n th
e ne
xt p
age)
179
P u b l i c f i n a n c e s i n E M U mdash 2 0 0 3
Tab
le V
3 (
cont
inue
d)
Exp
endi
ture
item
Def
init
ion
of t
arge
tL
evel
of a
pplic
atio
nD
ate
of in
trod
ucti
onT
ime
span
Act
ion
in c
ase
of n
on-c
ompl
ianc
e
Exc
epti
ons
to r
ule
in c
ase
of e
cono
mic
sh
ocks
Exp
erie
nce
wit
h th
e ru
le
ESN
on-fi
nanc
ial
expe
nditu
reFi
xed
ceili
ng se
t up
annu
ally
in th
e Bu
dget
La
w
Cent
ral g
over
nmen
t20
03A
nnua
llyN
o m
easu
res s
peci
fied
ex
ant
eTh
is lim
it in
clud
es a
co
ntin
genc
y fu
nd s
et
at 2
w
ithin
this
limit
so
as t
o m
eet
unfo
rese
en e
vent
s in
the
budg
et T
here
fore
an
y un
expe
cted
non
-fin
anci
al e
xpen
ditu
re
incr
ease
s hav
e to
be
met
thro
ugho
ut th
is co
ntin
genc
y fu
nd a
nd
or b
y de
crea
sing
othe
r sp
endi
ng it
ems
To b
e as
sess
ed si
nce
2003
is th
e fi
rst y
ear o
f ap
plic
atio
n
FRTo
tal e
xpen
ditu
reCu
mul
ativ
e re
al g
row
th
rate
s as
est
ablis
hed
each
yea
r for
the
next
th
ree
year
s
Mai
nly
cent
ral
gove
rnm
ent
1997
Med
ium
term
rol
ling
N
o m
easu
res s
peci
fied
ex
ant
e T
hese
targ
ets
are
not l
egal
ly b
indi
ng
and
are
usua
lly
adju
sted
in m
ediu
m-
term
pro
gram
mes
of
late
r yea
rs a
nd th
e fi
nal
budg
et fo
r any
pa
rtic
ular
yea
r
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed e
x an
te
The
orig
inal
med
ium
-te
rm o
bjec
tives
hav
e no
t bee
n re
spec
ted
H
owev
er i
n ge
nera
l the
in
crea
ses fi
xed
in th
e ye
arly
bud
get h
ave
been
resp
ecte
d e
xcep
t in
200
2
IETo
tal e
xpen
ditu
reA
nnua
l nom
inal
gr
owth
of 4
o
n av
erag
e du
ring
1998
ndash20
02
Cent
ral g
over
nmen
t 19
97Fi
ve y
ears
of t
he
gove
rnm
entrsquos
term
19
98ndash2
002
No
mea
sure
s spe
cifi
ed
ex a
nte
Tar
get
aban
done
d in
bud
get
for 2
001a
s the
cei
ling
of 4
in
nom
inal
te
rms t
urne
d ou
t to
be
ambi
tious
giv
en h
igh
nom
inal
GD
P gr
owth
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed e
x an
te
Rule
aba
ndon
ed in
bu
dget
for 2
001
rath
er
than
adj
uste
d to
refl
ect
high
er th
an e
xpec
ted
nom
inal
GD
P gr
owth
ITPr
imar
y ex
pend
iture
Nom
inal
cei
lings
or
lsquosafe
guar
d ru
lesrsquo
for a
ll pr
ovisi
ons i
nclu
ded
in
all l
egisl
atio
n in
trod
ucin
g ne
w a
nd
high
er e
xpen
ditu
res
Gen
eral
gov
ernm
ent
End
2002
Inde
fini
teA
pplic
atio
n of
le
gisla
tion
is fr
ozen
un
til n
ew le
gisla
tion
mak
es fu
ndin
g av
aila
ble
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed e
x an
te
Too
early
to a
sses
s H
owev
er s
ome
evid
ence
of a
redu
ctio
n in
gen
eral
gov
ernm
ent
cons
umpt
ion
on
quar
terly
dat
a
(Con
tinu
ed o
n th
e ne
xt p
age)
180
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
Tab
le V
3 (
cont
inue
d)
Exp
endi
ture
item
Def
init
ion
of t
arge
tL
evel
of a
pplic
atio
nD
ate
of in
trod
ucti
onT
ime
span
Act
ion
in c
ase
of n
on-c
ompl
ianc
e
Exc
epti
ons
to r
ule
in c
ase
of e
cono
mic
sh
ocks
Exp
erie
nce
wit
h th
e ru
le
Curr
ent p
rimar
y ex
pend
iture
of
regi
ons
In 2
002
+ 4
5
co
mpa
red
to 2
000
enga
gem
ents
In
2003
20
04 a
nd 2
005
200
2 ab
solu
te v
alue
+ ta
rget
in
flat
ion
of D
PEF
Regi
ons
End
2001
2002
ndash04
Non
e di
rect
Rem
ote
actio
n on
ly in
cas
e of
EU
sanc
tions
follo
win
g a
brea
ch o
f the
M
aast
richt
Tre
aty
3
of
GD
P de
fici
t th
resh
old
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed e
x an
te
Too
early
to a
sses
s
Stat
e fu
ndin
g of
he
alth
care
ex
pend
iture
Ceili
ngs o
n ex
pend
iture
by
regi
ons o
ver a
thre
e-ye
ar p
erio
d R
evise
d in
20
01 c
eilin
g of
EU
R 71
3 b
illio
n in
200
1
with
ann
ual i
ncre
ases
in
2002
ndash04
equa
l to
nom
inal
GD
P gr
owth
as
estim
ated
in th
e m
ediu
m-t
erm
pla
n (D
PEF)
Regi
ons
2000
2000
ndash03
(rev
ised
targ
et
for 2
001ndash
04)
Non
e S
tate
ndashreg
ions
ag
reem
ent
How
ever
an
y ex
tra
defi
cit s
houl
d be
cov
ered
by
regi
ons
thro
ugh
own
reso
urce
s or
by
expe
nditu
re c
uts
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed e
x an
te
The
ceili
ng w
as n
ot
resp
ecte
d an
d a
new
ag
reem
ent b
etw
een
Stat
e an
d re
gion
s was
ne
gotia
ted
in 2
001
A
ccor
ding
to
prov
ision
al fi
gure
s th
e ce
iling
was
bre
ache
d al
so in
200
1
NL
Expe
nditu
re a
s de
fine
d by
the
ceili
ngs
Med
ium
-ter
m re
al
expe
nditu
re c
eilin
gs
tran
slate
d ea
ch y
ear
into
nom
inal
am
ount
s
Gen
eral
gov
ernm
ent
Firs
t int
rodu
ced
in
1994
ada
pted
in
1998
and
200
2
Med
ium
term
cov
erag
e ac
cord
ing
to c
abin
et
perio
d
Com
mitm
ent t
o of
fset
ov
erru
ns o
f ex
pend
iture
cei
lings
by
expe
nditu
re c
uts
Spec
ific
rule
s fo
rmul
ated
for d
ivid
ing
win
dfal
ls be
twee
n lo
wer
ing
the
defi
cit o
r th
e ta
x bu
rden
Gen
eral
exp
endi
ture
ce
iling
has
bee
n ad
here
d to
but
ov
erru
ns h
ave
occu
rred
as
rega
rds t
he sp
ecifi
c ta
rget
s for
subs
ecto
rs
(hea
lthca
re)
It is
gene
rally
ass
umed
that
th
e fr
amew
ork
has h
ad
a re
stra
inin
g im
pact
on
expe
nditu
re
ATA
dmin
istra
tive
expe
nditu
reCu
ts in
per
sonn
el
mos
tly th
roug
h no
t re
plac
ing
civi
l ser
vant
s le
avin
g fo
r ret
irem
ent
Cent
ral g
over
nmen
tPr
evio
us ru
le 2
000
Fort
hcom
ing
rule
20
03
End
of le
gisla
tion
perio
d (p
revi
ous r
ule
20
03 in
theo
ry b
ut
gove
rnm
ent c
olla
psed
in
200
2 fo
r fo
rthc
omin
g ru
le e
nd
of 2
006)
No
mea
sure
s spe
cifi
ed
ex a
nte
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed e
x an
te
The
plan
ned
pers
onne
l cu
ts w
ere
impl
emen
ted
as p
lann
ed fr
om 2
000ndash
02 D
espi
te a
n in
crea
se
in p
ensio
n ex
pend
iture
fo
r pub
lic se
rvan
ts i
t is
assu
med
that
this
rule
ha
s had
a re
stra
inin
g im
pact
on
expe
nditu
re
(Con
tinu
ed o
n th
e ne
xt p
age)
181
P u b l i c f i n a n c e s i n E M U mdash 2 0 0 3
Tab
le V
3 (
cont
inue
d)
Exp
endi
ture
item
Def
init
ion
of t
arge
tL
evel
of a
pplic
atio
nD
ate
of in
trod
ucti
onT
ime
span
Act
ion
in c
ase
of n
on-c
ompl
ianc
e
Exc
epti
ons
to r
ule
in c
ase
of e
cono
mic
sh
ocks
Exp
erie
nce
wit
h th
e ru
le
Tota
l exp
endi
ture
Budg
et b
alan
ce ru
le
How
ever
bud
geta
ry
targ
ets c
an b
e at
tain
ed
via
expe
nditu
re si
de
mea
sure
s onl
y
Regi
onal
and
loca
l go
vern
men
ts20
01En
d of
the
curr
ent
finan
cial
equ
alisa
tion
Fina
ncia
l san
ctio
ns
simila
r to
thos
e of
the
exce
ssiv
e de
fici
t pr
oced
ure
of th
e SG
P
via
reve
nue
dist
ribut
ion
mec
hani
sm b
etw
een
cent
ral a
nd lo
wer
leve
ls of
gov
ernm
ent
The
floo
d di
sast
er in
20
02 le
d to
a te
mpo
rary
su
spen
sion
of th
e ru
le
that
is n
ot ta
king
into
ac
coun
t of fl
ood-
rela
ted
expe
nditu
re in
th
e ye
ars 2
002
and
2003
Ceili
ng n
ot re
spec
ted
in
2001
Not
resp
ecte
d in
20
02 b
ut su
spen
ded
for
that
yea
r In
gen
eral
di
fficu
lt to
mea
sure
st
ruct
ural
savi
ngs o
f re
gion
s
PTCo
mpe
nsat
ion
of
empl
oyee
sN
o ne
w la
bour
co
ntra
cts i
n th
e ce
ntra
l ad
min
istra
tion
are
to
be si
gned
unl
ess
auth
orise
d by
the
Min
ister
of F
inan
ce
Cent
ral g
over
nmen
t20
02Cu
rren
t leg
islat
ure
(200
2ndash05
)N
o m
easu
res s
peci
fied
ex
ant
eTh
e Fi
nanc
e M
inist
er
alon
e ca
n ov
errid
e th
e fr
eezi
ng i
n pa
rtic
ular
fo
r sen
sitiv
ity a
reas
like
he
alth
care
Too
early
to b
e as
sess
ed
FITo
tal e
xpen
ditu
reFr
eezi
ng re
al c
entr
al
gove
rnm
ent s
pend
ing
at th
e le
vel o
f 199
9 ou
tcom
e
Cent
ral g
over
nmen
t on
-bud
get
expe
nditu
re
excl
udin
g ex
tra-
budg
etar
y fu
nds
(pen
sion
etc
)
1999
but
ann
ual
fram
es fo
r cen
tral
go
vern
men
t sp
endi
ng w
ere
desig
ned
alre
ady
at
the
begi
nnin
g of
19
90s
Cabi
net p
erio
d (1
999
to
Mar
ch 2
003)
No
mea
sure
s spe
cifi
ed
ex a
nte
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed e
x an
te H
owev
er
decl
inin
g go
vern
men
t de
bt a
nd fa
lling
un
empl
oym
ent h
ave
crea
ted
leew
ay fo
r ad
ditio
nal e
xpen
ditu
re
Ove
rrun
s occ
urre
d in
20
01 a
nd 2
002
and
acco
rdin
g to
the
2003
sp
endi
ng g
uide
line
cent
ral g
over
nmen
t bu
dget
ary
spen
ding
is
estim
ated
at E
UR
12
billi
on o
ver t
he
outc
ome
of 1
999
It is
ge
nera
lly a
ssum
ed th
at
the
fram
ewor
k ha
s had
a
rest
rain
ing
impa
ct o
n ex
pend
iture
SEPr
imar
y ex
pend
iture
pl
us e
xpen
ditu
re fo
r th
e ol
d-ag
e pe
nsio
n sy
stem
out
side
the
budg
et
Ann
ual c
eilin
g on
no
min
al e
xpen
ditu
re
expe
nditu
re co
vere
d by
th
e ce
iling
shou
ld n
ot
rise
fast
er th
an
(pro
ject
ed) n
omin
al
GD
P
Cent
ral g
over
nmen
t19
97Th
ree
year
s ahe
ad
rolli
ngBi
annu
al m
onito
ring
requ
ired
by th
e Bu
dget
La
w I
f the
re a
re si
gns
of o
verr
uns (
over
all)
the
gove
rnm
ent s
hall
prep
are
a pr
opos
al fo
r co
rrec
tion
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed e
x an
te
The
expe
nditu
re
ceili
ngs h
ave
been
re
spec
ted
in e
ach
year
sin
ce 1
997
whe
n th
ey
wer
e fi
rst i
ntro
duce
d It
is
gene
rally
ass
umed
th
at th
e fr
amew
ork
has
had
a re
stra
inin
g im
pact
on
expe
nditu
re
(Con
tinu
ed o
n th
e ne
xt p
age)
182
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
Tab
le V
3 (
cont
inue
d)
Exp
endi
ture
item
Def
init
ion
of t
arge
tL
evel
of a
pplic
atio
nD
ate
of in
trod
ucti
onT
ime
span
Act
ion
in c
ase
of n
on-c
ompl
ianc
e
Exc
epti
ons
to r
ule
in c
ase
of e
cono
mic
sh
ocks
Exp
erie
nce
wit
h th
e ru
le
UK
Dep
artm
enta
l Ex
pend
iture
Lim
its
(DEL
) (1 )
Gov
ernm
ent
depa
rtm
ents
are
set
spen
ding
pla
ns fo
r the
le
vel o
f nom
inal
ex
pend
iture
for t
hree
ye
ars a
head
in so
-cal
led
Com
preh
ensiv
e Sp
endi
ng R
evie
ws
(CSR
) Pa
rliam
enta
ry
auth
ority
to sp
end
mus
t stil
l be
obta
ined
ea
ch y
ear
Gov
ernm
ent
depa
rtm
ents
Firs
t lau
nche
d un
der
the
1998
CSR
for t
he
perio
d 19
99ndash2
002
A
new
bat
ch o
f thr
ee
year
s was
set i
n th
e 20
00 C
SR a
nd a
gain
in
the
2002
CSR
thre
e ye
ars
The
CSR
take
pla
ce e
very
two
year
s mdash th
e th
ird y
ear
of th
e pr
evio
us e
xerc
ise
beco
mes
the
firs
t yea
r of
the
succ
eedi
ng
exer
cise
The
DEL
pla
ns a
re
bind
ing
but
they
can
be
alte
red
in th
e bu
dget
pro
cess
and
are
su
bjec
t to
appr
oval
by
gove
rnm
ent a
nd
parli
amen
t U
nder
- or
over
spen
ding
in o
ne
year
can
be
offs
et in
an
othe
r yea
r with
in th
e cu
rren
t thr
ee-y
ear
batc
h
No
auto
mat
ic
exce
ptio
ns sp
ecifi
ed e
x an
te
The
gove
rnm
entrsquos
m
ediu
m-t
erm
pla
ns
publ
ished
in th
e Bu
dget
repo
rt a
nd
whi
ch fo
rm th
e fr
amew
ork
for D
EL
prog
ram
mes
are
re
quire
d u
nder
the
term
s of t
he C
ode
for
Fisc
al S
tabi
lity
to m
eet
the
gove
rnm
entrsquos
fisc
al
rule
s Th
ey h
ave
satis
fied
thes
e ru
les s
o fa
r
(1 )T
he tw
o m
ain
part
s of
the
UK
rsquos b
udge
ting
and
cont
rol f
ram
ewor
k ar
e D
EL
(de
part
men
tal e
xpen
ditu
re li
mits
) an
d A
ME
(an
nual
ly m
anag
ed e
xpen
ditu
re)
Gov
ernm
ent d
epar
tmen
ts a
re g
iven
thre
e-ye
ar s
pend
ing
limits
th
e D
EL
s A
ny s
pend
ing
that
can
not r
easo
nabl
y be
sub
ject
to s
uch
mul
ti-ye
ar li
mits
is in
clud
ed in
AM
E (
for
exam
ple
soc
ial s
ecur
ity s
pend
ing
net
pay
men
ts to
the
EC
) A
ll A
ME
pro
ject
ions
for
fut
ure
year
s ar
e es
ti-m
ates
whi
ch a
re u
pdat
ed tw
ice-
year
ly in
the
budg
et a
nd p
re-b
udge
t rep
orts
Tog
ethe
r A
ME
and
DE
L s
um to
tota
l man
aged
exp
endi
ture
(T
ME
) a
nat
iona
l acc
ount
s m
easu
re d
efine
d as
pub
lic s
ecto
r cu
rren
t exp
endi
ture
plus
net
inve
stm
ent p
lus
depr
ecia
tion
In
the
atta
ched
tabl
es o
nly
DE
L s
pend
ing
is in
clud
ed s
ince
this
is th
e on
ly p
art o
f TM
E w
hich
is s
ubje
ct to
mul
ti-ye
ar li
mits
183
P u b l i c f i n a n c e s i n E M U 2 0 0 3
particular it examines spending on various expenditureitems before and after the introduction of an expenditurerule It also examines whether there is a link between theSGP commitments and the non-compliance with Mem-ber Statesrsquo expenditure targets
The results should be interpreted with caution Expendi-tures are affected by many other factors outside directcontrol of policy makers since most national expendi-ture rules target expenditure outcomes and not theimpact of specific policy actions it is difficult to identifythe net effect of the rule on the budgetary position
Table V4 shows the average rate of growth of the itemsfalling under an expenditure rule when data on specificexpenditure categories covered by a rule are not availa-ble a close proxy is used Averages are calculated for thethree years before the introduction of the rule and for theyears following the introduction of the rule up to2002 (1) In four countries (Denmark France Luxem-bourg and Belgium) the rule is defined in terms of realrate of growth of expenditure while in Ireland the targetis in terms of nominal rate of growth Five countries(Sweden Greece the Netherlands Austria and Finland)have a more composite definition that can be proxied toa nominal ceiling However since the final aim is to con-trol expenditure to assess whether the rule worked thereal rate of growth is used as a proxy of the target
In five out of 10 countries the rate of growth after theintroduction of the rule has been lower compared withthe years immediately prior to its introduction In partic-ular the rate of growth of the expenditure targeted fell inall those countries where the target has been defined as aceiling rather than as a rate of change however asshown by t-statistics differences in means are rarely sta-tistically significant
Whenever there is a correction of expenditure trends theeffectiveness of an expenditure rule depends on its levelof ambition In general over-ambitious targets risk notto be fulfilled in some cases Member States havechanged the medium-term targets when it became clearthat these objectives would be difficult to respect A tar-get that is not ambitious enough on the other hand cre-ates the risk of pushing expenditure up to the ceiling Astrict adherence to a credible and realistic framework
would have probably enhanced the credibility of theframework and contributed to the overall compliancewith the EU fiscal rules
In many cases spending rules do not seem to be suffi-ciently ambitious (2) In France the 2000 updated stabil-ity programme fixed a target of 49 in real growth (ona cumulative basis) for the three-year period 2001ndash03During the same period real GDP growth is expected tobe at 41 (3) Therefore expenditure as a share of GDPcan increase without breaching the rule (4) In Belgiumthe rule fixes a real growth of primary expenditure of15 each year (5) Real GDP grew by 08 in 200107 in 2002 and it is expected to grow by 13 in2003 Thus compliance with the rule does not necessar-ily imply a reduction of the expenditure-to-GDP ratioalthough admittedly the increase is at least in part due tothe low growth rates
Table V5 investigates whether there is a link between afailure on the part of a Member State to respect its ownexpenditure rule and respect of the lsquoclose to balance orin surplusrsquo requirement of the SGP It compares the dif-ference between the targeted (in the relevant stability orconvergence programme) and the actual outcome of totalexpenditures as a share of GDP for cases identifiedwhere countries missed their expenditure rule (6) Thefinal column shows the deterioration in the cyclically-adjusted budget balance for the year in question com-pared with the previous year In all cases except Italy in2002 and Finland in 2001 and 2002 non-compliancewith a national expenditure rule coincided with a wors-ening cyclically-adjusted budget position ConcerningFinland the behaviour can be explained by the fact thatthe Finnish expenditure rule stands out as a very ambi-tious one as it aims at freezing real expenditure at thelevel of 1999 (7) In Italy cyclically-adjusted improve-ments rely mainly on one-off measures that loweredexpenditure
yen1part 2000 for Ireland when the rule was abandoned Countries not included inthe Table implemented expenditure rules too recently to be assessed
yen2part An exception is Ireland it decided not to enforce its spending limits for2001ndash02 which were quite tight compared with nominal GDP growth Fin-land also has an ambitious target that is a frozen real central governmentspending at the 1999 level In Sweden the norm is that the real rate ofchange should be zero
yen3part According to the European Commission spring 2003 forecastsyen4part Nevertheless the target has been revised in the following updates of the
stability programme it became 52 in the updated programme 2001 and65 in the updated programme 2002
yen5part The rule regards primary expenditure in entity I (Federal government andsocial security)
yen6part Cases have been identified according to information available by the Euro-pean Commission Directorate-General for Economic and FinancialAffairs
yen7part See also the chapter on Finland in Chapter VI13 of this report
184
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
To sum up the overall picture signals that there are noevident changes in the behaviour of expenditure oncespending rules are introduced Nevertheless the com-pliance with the rule is difficult to judge Targets arein many cases set up over several years and there areoften revisions in due course In some countries tar-gets are not ambitious enough and adherence with
them is easily reached In other cases the rule hasbeen adjusted or abandoned since it was perceived tobe too ambitious Moreover the assessment of howrules worked is limited by a lack of quantitative infor-mation so that expenditure trends are difficult tomonitor This is particularly true in those cases wherethe target concerns detailed expenditure categories
Table V4
The impact of expenditure rules on spending trends
Item controlledDefinition of the rule
Year of introduction
Real rate of growth of the measured item (1)
3 years before the introduction of the rule (A)
After the introduction of the rule (B)
T-testfor A ne B
DK Public consumption Real rate of growth 1999 24 16 10
FR Total expenditure Real rate of growth 1997 17 21 08
LU Total expenditure Real rate of growth 1999 35 62 10
BE Primary expenditure Real rate of growth 1999 20 26 10
SE Primary expenditure Real rate of growth 1997 09 33 19
IE Total expenditure Nominal rate of growth 1997 56 94 16
EL Compensation of employees Nominal ceiling 1997 123 71 11
NL Total expenditure Nominal ceiling 1994 22 10 11
AT Compensation of employees Nominal ceiling 2000 00 ndash 27 07
FI Total expenditure Nominal ceiling 1999 11 10 00
(1) Nominal rate of growth for Ireland as defined in the expenditure rule
Source Commission services
Table V5
Total expenditure targets and spending rules
ExpenditureGDP Changes in cyclically-adjustedbudget balance ( GDP)Target Outcome Difference
BE (1999) 480 501 21 ndash 04
DK (1999) (1) 519 524 05 ndash 13
DE (2002) 480 486 06 ndash 03
FR (2002) 523 537 14 ndash 11
IT (2001) (2) 475 485 10 ndash 07
IT (2002) (2) 467 475 08 10
FI (2001) (3) 240 249 09 01
FI (2002) (3) 243 251 08 06
(1) Outcome before statistical revision(2) Targets recalculated by Commission services according to EU standards to ensure consistency with outcomes Planned sales of real assets have been subtracted
from the expenditure targets (06 percentage points of GDP only in 2002) and additional expenditure items have been included as required under Commis-sion Regulation (EC) No 15002000 (03 percentage points of GDP both in 2001 and 2002) Sales of real assets lowered outcomes by 02 percentage points ofGDP in 2001 and by 09 percentage points of GDP in 2002
(3) Central government total expenditure
Source Commission services on the basis of data provided by Member States in their stability or convergence programmes
185
P u b l i c f i n a n c e s i n E M U 2 0 0 3
andor when the rule covers only part of the generalgovernment
24 Conclusions
Expenditures rules are becoming a common featureamong EU Member States as an additional tool to con-trol budgetary development In the majority of cases theyare lsquoex antersquo rules they fix a target that helps to keepexpenditures under control during the process of budget-ary formation However implementation mechanismsand lsquoex postrsquo control are rather weak As a consequencethe medium-term expenditure target tends to be revisedif it becomes clear it cannot be reached
Thus what counts for an effective expenditure rule is agood design and the existence of control mechanismsthat allow to correct trends in the course of budget imple-mentation Control mechanisms should be accompanied
by enforcement mechanisms to render the rule fullyimplemented
In the EU context national expenditure rules can com-plement the fiscal framework currently in place but can-not be seen as a substitute First because they are notsubject to budgetary surveillance at EU level Secondtheir current designs and implementation mechanismsdo not ensure the achievement and maintenance of soundpublic finances over the long term
However even a lsquoweakrsquo rule can be helpful as a guid-ance of fiscal policy and to signal to the actors involvedin the budgetary process which are the components ofthe budget that create more concern Also the redirec-tion of public expenditure towards those items that aremore conducive to economic growth becomes easierTherefore in all cases in which specific items less undercontrol crowd out other perhaps more productiveexpenditures a rule can increase the efficiency of publicexpenditure
186
3 Fiscal relations across levels of government
31 Fiscal relations across different levels of government in EU Member States
In recent years the management of public finances in EUMember States has not only been affected by the processof European integration it has also been influenced by aprocess of decentralisation whereby the budgetary auton-omy of lower levels of government has been increasedThis reshaping of the division of budgetary competenciesbetween layers of government within Member States hasconsequences for the budgetary requirements at the EUlevel as the Treaty and SGP obligations concern the gen-eral government as a whole that is central state andlocal government plus social security
The process of transferring more budgetary authority tolower levels of government is motivated in part by polit-ical factors namely as a way of reconciling divergenceor tension between communities with national politicalcohesion or has been an expression of the citizensrsquo rightto participate in the conduct of public affairs (Committeeof the Regions 2001) (1) Decentralisation may also bejustified on economic grounds in particular lower lev-els of government may be able to better tailor the provi-sion of public services to local needs and preferencesand to establish a link with the taxes that are needed tofinance them thereby increasing accountability at thelocal level Box V1 provides an overview of the keyarguments of the theory of fiscal federalism
There are large differences between EU Member Statesin the way budgetary responsibilities are divided between
different levels of government This is in part linked tothe system of government and particular whether thecountry is a federal (Austria Belgium and Germany) orunitary State However the distinction is not clear cutSpain and Italy could be classified in both groups sincethey are unitary States with some characteristics of a fed-eral State (2) The Nordic countries (Denmark Finlandand Sweden) also have some special characteristics asthey are unitary States where the principle of lsquoself-gov-ernmentrsquo is grounded in the constitution
A common indicator for assessing the degree of fiscaldecentralisation is to look at sub-national expendituresand revenues both as a percentage of GDP and of totalpublic expenditures Table V6 reports this indicatorbased on the data available in the European system ofaccounts (3) The figures are based on a calculation ofrevenues and expenditures at different levels of govern-ment as the sum of their components since there are noharmonised data ESA95 available for total revenues andexpenditures at lower levels of government (4) Thesefigures must be interpreted with care as they give anapproximate indication of the size of lower levels of gov-ernment but do not measure budgetary autonomy
yen1part See page 48 of Committee of the Regions (2001) for examples of devolu-tion in Europe It should be noted that decentralisation is not a uniformtrend in all Member States It is a long-term process that has taken placeduring the last few decades See the decentralisation web site of the WorldBank (wwwworldbankorgpublicsectordecentralization) for an overviewof the different arguments surrounding the debate on decentralisation
yen2part In Spain the constitution does not directly specify the regions (lsquoautono-mous communitiesrsquo) which account for a large part of public expenditure(Table V6) Italy also has federalist characteristics since regional authori-ties exercise legislative powers comparable with those of regions or statesin federal Member States (Committee of the Regions 2001)
yen3part The most common databases for cross-country comparison in this field arethe Government Finance Statistics of the IMF and the OECD Revenue Statis-tics
yen4part Total expenditure is calculated as the sum of (ESA categories are indi-cated) D3 subsidies D4 Property income D5 Current taxes on incomeand wealth D62 Social benefits other than transfers in kind D7 Other cur-rent transfers P3 Final consumption expenditure D9 Capital transfers P5Gross capital formation K2 Acquisitions of non-produced non-financialassets Total Resources are calculated as the sum of K1 Consumption offixed capital B2 Operating surplus D2 Taxes on production and importsD4 property income D5 Current taxes on income and wealth D61 Socialcontributions D7 Other current transfers D9 Capital transfers
187
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Table V6 shows that in general the federal and the Nor-dic countries are the most decentralised according to theindicator When measured in terms of sub-governmentexpenditure (that is State and local) as a percentage oftotal government spending then Denmark (57 ) Ger-many (43 ) Belgium (41 ) Sweden (40 ) and Spain(38 ) stand out as having a highly decentralised fiscalstructure (1) A second group consists of the Netherlands(35 ) Finland (34 ) Austria (33 ) Italy (30 ) theUK (26 ) and France (19 ) The most centralisedMember States are Luxembourg Portugal (both 14 )and Greece (4 ) With respect to the development oflower levels of government over time the figures gener-
ally show slow changes in the level of decentralisationsince 1995 the first year for which figures are availablefor all Member States Nevertheless a relative increasesince 1995 is recorded in the size of the states in Austriaand Spain and the local level of government in DenmarkSweden and Italy A relative decrease is recorded in thesize of the local government in the Netherlands
Table V7 examines the composition of public spendingby sub-central levels of government in Member Stateswhere data was available According to the theory of fis-cal federalism public spending of sub-central authoritiescould be expected in policy domains where there arelarge differences in preferencesneeds across regionsbut less so in areas whereas economies of scale and spill-over effects prevail
Box V1 Key arguments of the theory on fiscal federalism
The theory of fiscal federalism has developed criteria for the assignment of government activities to different layers of gov-ernment The main benefits of centralisation are the internalisation of externalities and spillovers (that is when market fail-ures have cross-border effects on other jurisdictions) and the exploitation of economies of scale However these need tobe weighed against the benefits of decentralisation which include a capacity to adjust the provision of public goods andservices to local preferences and needs the avoidance of diseconomies of scale more competition and innovation in theprovision of public goods and services and improved accountability and transparency of policy makers by establishing amore direct link between the benefits of public expenditures and the taxes levied to finance them
Fiscal federalism yields no clear-cut policy conclusions on the assignment of public functions whose main objective is toensure an efficient allocation of resources A costbenefit assessment is needed on a case by case basis Some public goodsservices may need to be centralised where there are large spillover effects covering the entire country (national transportinfrastructure) whereas others may be more efficiently provided at local level (local transport infrastructures)
In contrast stronger policy conclusions are drawn as regards the benefits of centralising the public function that aim atredistribution (either across regions or individuals) for several reasons Firstly the demand for redistribution policies maycover an entire country in that citizens may be concerned about the living standards of the entire population and not justin their own locality or State Secondly it may be very difficult to operate redistribution policies efficiently at sub-centrallevel labour mobility may result in the migration of low-income persons to regions providing the most generous benefitswhereas high-skilled persons may move to regions with the lowest taxes
Fiscal federalism in general reaches strong policy conclusions on centralising the stabilisation function This is becauselower levels of government might not have the right incentives to provide an optimal level of stabilisation since a consid-erable part of their stabilisation efforts would leak away to other jurisdictions Furthermore the possibilities of local gov-ernments to run counter-cyclical policies (for example by means of letting the automatic stabilisers work) are often limitedgiven the existence of borrowing restrictions
In general the theory of fiscal federalism provides stronger arguments in favour of centralised revenue collection comparedwith expenditures Centralised revenue collection could lower the costs of collection and compliance due to economics ofscale it could prevent tax evasion induced by mobile tax basis and prevent excessive tax competition This can give riseto a rsquovertical fiscal imbalancersquo whereby central sub-national governments have to rely on the central government to providethem with revenues to finance decentralised public expenditures
yen1part One should keep in mind that this figure does not measure local autonomyin deciding on expenditure
188
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
As expected defence is never decentralised reflecting thepresence of spillover effects economies of scale and polit-ical considerations A considerable percentage of theresources of sub-central authorities are devoted to itemssuch as education housing recreation and culture decen-tralised provision of these items may be justified on theground of tailoring public goods and services to localneeds and preferences The largest differences betweenMember States can be found in the categories of healthand social security and welfare where sub-central author-ities in several countries have an important role to play
It should be noted however that the scale and composi-tion of public spending by sub-central authorities doesnot coincide with the actual degree of budgetary auton-omy of sub-national authorities This is because the cen-tral government can influence to a large degree theexpenditure choices of sub-central authorities for exam-ple by mandating standards of public goods and servicesthat sub-central authorities must provide Local or state
government expenditures for example include expendi-tures that are part of national programmes In the NordicMember States central control is generally confined tosetting a broad policy framework leaving them a highdegree of independence in areas like primary educationsocial and health services Their counterparts in theNetherlands Germany Austria and Italy have a role tooin providing the major welfare services though withmore detailed steering by higher tiers of government(Committee of the Regions 2001)
Sub-central authorities can be financed through taxesgrants service charges and fees (1) Table V8 shows themain categories as according to the ESA95 classifica-tions Taxes that are collected by the central governmentand automatically transferred to the local and state gov-
Table V6
Expenditure and revenues at State and local government level
MS Structure
Total expenditures Total revenues
of GDP of total of GDP of total
1995 2000 2001 1995 2000 2001 1995 2000 2001 1995 2000 2001
BE Federal State 14 13 14 26 27 27 13 14 14 27 27 28
Local 7 7 7 12 14 13 7 7 7 14 13 13
DK Unitary local self-government Local 32 31 31 53 56 57 33 31 31 56 53 54
DE Federal State 13 14 14 27 29 28 12 13 12 26 28 27
Local 8 7 7 15 15 14 8 7 7 16 15 16
EL Unitary Local 2 2 2 3 4 4 2 2 2 5 4 4
ES Unitary federal features State 7 9 9 15 22 23 6 8 8 16 21 21
Local 6 6 6 13 15 15 6 6 6 15 16 16
FR Unitary Local 10 10 10 18 19 19 10 10 10 20 19 19
IT Unitary federal features Local 13 14 14 24 30 30 13 14 15 28 30 32
LU Unitary Local 7 6 15 14 7 6 15 13
NL Unitary Local 23 16 16 45 35 35 23 16 16 49 34 35
AT Federal State 8 10 10 14 18 18 9 10 10 16 20 19
Local 9 8 8 16 15 15 8 8 8 16 16 15
PT (1) Unitary Local 5 7 12 14 5 5 14 12
FI Unitary local self-government Local 19 16 17 31 33 34 20 16 16 36 29 30
SE Unitary local self-government Local 23 22 23 34 39 40 23 23 23 37 38 38
UK Unitary four constituent nations Local 12 10 11 26 28 26 11 10 11 29 25 26
EUR-12 16 16 31 33 16 15 33 32
EU-15 16 15 31 33 16 15 33 32
(1) Figures for PT concern 1999
Source Commission services
yen1part That is in the absence of borrowing See Graph V3 on the contribution oflower levels of government to general government borrowing
189
P u b l i c f i n a n c e s i n E M U 2 0 0 3
ernments (for example as part of a tax-sharing agree-ment) are registered as if they were collected directly bythe local or state government According to ESA95 thecategory of transfers within the general governmentmainly shows block transfers to the local and state gov-ernments that do not correspond to any specific categoryof taxes
There are large differences in the way Member Statesfinance their expenditure at lower levels of governmentIn Belgium the states rely mostly on transfers from thecentral government For the states in Austria and Spaintransfers also account for a large part of their revenuesalthough to a lesser extent than in Belgium In Austriatax sharing represents another important part of incomewhile the states in Spain have increased their tax auton-omy in the second half of the 1990s For the Germanstates the transfers from the central government aremuch smaller and tax income is the most importantsource of revenue This reflects the importance of taxsharing of national taxes with the central government
Transfers to local governments are relatively high inthe UK and the Netherlands which indicates their rela-tively centralised system of financing local govern-ments This contrasts with Italy and France where theautonomy of lower levels of government in raisingtaxes is higher In Italy in particular reforms in the1990s have strongly decreased local governmentsrsquodependence on transfers from the centre and extendedtheir autonomy in raising taxes Finally the data for thecategory of taxes on income and wealth show verylarge differences between Denmark Finland and Swe-den where figures range from 10 to 15 of GDP andother Member States where this figure is usually below2 of GDP in line with the fact that income taxes arethe most important source of income at local level forthe Nordic countries
32 Fiscal decentralisation and its interaction with the EUrsquos fiscal rules
321 Fiscal decentralisation and the goal of sound and sustainable public finances
The data in Section 31 clearly illustrate the importanceof public finances at sub-central level when consideringthe overall budgetary situation of a Member State Aquestion arises whether there is a link between thedegree of fiscal decentralisation and the budgetary per-formance in particular the capacity of Member States to
meet the budget balance and debt requirements for thegeneral government set down in the Treaty and SGP
Graph V2 compares an indicator for fiscal decentrali-sation with indicators for budget balance and debt Itshows that at first glance there is no apparent linkbetween the degree of fiscal decentralisation and budg-etary performance
However a possible link between fiscal decentralisationand budgetary performance may exist depending uponwhether or not a sub-central authority faces a hardbudget constraint (for example Rodden 2000) Theargument is that lower levels of government may nottake adequate account of the spillover effects of theirbudget policies and may face incentives to shift the costsof their expenditure decisions to the central level of gov-ernment The extent to which they might be able to actaccording to these incentives depends on the institutionalset-up of the system of financing of lower levels of gov-ernment (Eichengreen and von Hagen 1996 Rodden2002 Ter-Minassian and Craig 1997)
There may be a tendency for higher levels of publicspending and deficits if there is a vertical fiscal imbal-ance that is when sub-central authorities have importantresponsibilities for public expenditures but limited ownresources and are thus reliant on transfers and grantsfrom central authorities These transfers may create theperception that local public spending is funded by non-residents As a consequence expenditure discipline andcost-awareness might deteriorate the costs of grantsmay not be fully internalised at the local level causing todemand above-optimal levels of public expenditures onitems that are financed by grants for central authorities(for example Rodden and Wibbels 2002) This pressurefor increased transfers to sub-central authorities couldtranslate into higher deficits and debt of the general gov-ernment On a related point sub-central authorities mayengage in excessive levels of borrowing if they considerthat in the event of default they will be bailed out by ahigher level Pressures to bail out sub-central authoritiesmay rise with the degree of vertical imbalance since thesmaller the tax base and the control over it at sub-national level the smaller are the possibilities at thatlevel to raise taxes in the event of financial problems
In response to these pressures governments in recentyears have paid close attention to the incentives embed-ded in the design of grants and revenue sharing arrange-ments with sub-central authorities Many countries have
190
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
Tab
le V
7
Sub-
nati
onal
gov
ernm
ent
spen
ding
by
func
tion
as
a pe
rcen
tage
of
tota
l loc
al s
pend
ing
Gen
eral
pu
blic
se
rvic
esD
efen
ce
Pub
lic
orde
ran
d sa
fety
Edu
cati
onH
ealt
h
Soci
al
secu
rity
an
d w
elfa
re
Hou
sing
an
d co
mm
unit
y am
enit
ies
Rec
rea-
tion
al
cult
ural
an
d re
ligio
us
affa
irs
Fue
l an
d en
ergy
Agr
i-cu
ltur
e
fore
stry
fi
shin
g an
d hu
ntin
g
Min
ing
m
anuf
ac-
turi
ng a
nd
cons
truc
-ti
on (e
xcep
t fu
el a
nd
ener
gy)
Tra
ns-
port
atio
n an
d co
mm
uni-
cati
on
Oth
er
econ
omic
af
fair
s
Oth
er
func
tion
s
ES6
90
42
183
205
51
107
56
01
35
10
72
27
142
DE
64
06
218
310
620
18
63
50
32
20
15
84
313
6
DK
39
00
312
316
157
50
92
80
00
00
12
82
50
8
FR10
60
23
196
23
177
241
77
42
00
00
36
00
78
NL
94
03
417
92
622
620
05
80
50
00
56
70
010
6
UK
40
012
328
70
032
55
43
10
00
10
04
91
08
0
IE2
30
18
113
455
52
149
19
38
02
00
113
07
11
Figu
res
are
for
1997
exc
ept f
or F
I (1
993)
AT
(19
94)
DE
(19
96)
UK
(19
98)
and
DK
(20
00)
Sour
ce I
MF
Gov
ernm
ent F
inan
ce S
tatis
tics
Tab
le V
8
The
com
posi
tion
of
tota
l rev
enue
s at
sta
te a
nd lo
cal l
evel
as
a pe
rcen
tage
of
GD
P (
year
200
0)
AT
BE
ES
DE
DK
FI
SEF
IIT
LU
NL
UK
EL
PT
(1 )
Stat
eLo
cal
Stat
e Lo
cal
Stat
eLo
cal
Stat
e Lo
cal
Taxe
s o
n in
com
e an
d w
ealt
h1
92
00
71
08
55
14
151
103
154
07
16
23
07
15
01
06
Taxe
s o
n p
rod
uct
ion
an
d im
po
rts
13
29
08
12
15
22
41
61
10
36
46
02
07
00
21
6
Cu
rren
t tr
ansf
ers
wit
hin
gen
eral
go
vern
men
t4
81
510
52
74
21
91
52
111
93
73
83
34
82
210
97
11
14
Oth
er2
22
23
19
12
12
22
21
24
21
ndash 1
42
32
81
43
81
50
71
3
Tota
l rev
enu
es10
86
136
65
79
61
132
72
305
161
178
99
138
61
161
102
14
9
(1 )Fi
gure
s fo
r PT
con
cern
199
9
Sour
ce C
omm
issi
on s
ervi
ces
191
P u b l i c f i n a n c e s i n E M U 2 0 0 3
also introduced borrowing restrictions for lower levels ofgovernment (for an overview see Ter-Minasian andCraig 1997) and empirical studies indicate that higherdegrees of vertical imbalance and sub-national borrow-ing restrictions are indeed associated (Eichengreen andVon Hagen 1996)
Borrowing restrictions are usually found to be effectivein restraining fiscal policies at lower levels of govern-ment (for example Bayoumi and Eichengreen 1995)All EU countries apply restrictions to local governmentspending and borrowing but in various forms anddegrees (Dafflon 2002) Their impact within the EU isexamined on Graph V3 which contrasts the general gov-
Graph V2 Fiscal decentralisation and budgetary situation
Source Commission services
ndash 500
ndash 400
ndash 300
ndash 200
ndash 100
000
100
200
300
000 010 020 030 040 050 060
Fiscal decentralisation
CA
B
DKSEFI
UK
EL DEIT
PT
FR ES
NLBE
AT
2000
4000
6000
8000
10000
12000
000 010 020 030 040 050 060
Fiscal decentralisation
Gro
ss d
ebt
( o
f G
DP
)
DKSE
FI
UK
EL
DE
IT
PTFR ES
NL
BE
AT
192
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
ernment budget balance (dark column) and the budgetbalance of local and (where relevant) state governmentsLocal and state governments usually balance their budg-ets or run small deficits or surpluses The only notableexception is Germany where net lending by local andstate government accounts for almost half of the generalgovernment deficit in 2002
Clearly this does not provide for an adequate picture ofthe overall contribution of lower levels of government tothe general government budget balance For example asub-national government that is facing a borrowingrestriction might obtain a higher amount of grants or per-centage of shared taxes from the central governmentincreasing the deficit at that level
322 Recent measures in several Member States to coordinate budgetary positions across levels of government in light of EU requirements
In recent years a number of Member States have re-considered the fiscal relations across different levels ofgovernment which take into account the need to complywith EU budgetary requirements These initiatives alsosought to correct a form of vertical institutional imbal-ance whereby the Treaty and SGP obligations concernthe general government as a whole (ie central state andlocal government plus social security) but commitmentsgiven at European level (notably in the annual updates tostability and convergence programmes) are made by thecentral government Compliance with budgetary com-mitments given at EU levels is dependent upon the budg-etary performance of all levels of government whereas
the costs of non-compliance (either the reputational costor ultimately in the form of a pecuniary sanction) areborne by central government
Apart from borrowing and budgeting restrictions forsub-national authorities (as discussed in the previoussection) the federal Member States and Italy and Spainhave also introduced institutional arrangements atnational level usually referred to as national stabilitypacts These arrangements can be summarised accordingto the formulation and scope of their targets the meas-urement of the targets their legal status the process ofsurveillance and the enforcement including possiblesanctions The usual hypothesis is that a complete designacross all these dimensions will contribute to the effec-tiveness of the arrangements
Graph V3 The contribution of lower levels of government to general government net lending (+) or borrowing (ndash) in 2002
ndash 4
ndash 3
ndash 2
ndash 1
0
1
2
3
4
BE DK DE EL ES FR IE IT LU AT NL PT FI SE UK
Lower levels of government
General government
193
P u b l i c f i n a n c e s i n E M U 2 0 0 3
In January 1999 a domestic stability pact was enacted inAustria In October 2000 it was amended by an agree-ment between the federal government provinces and thelocal authorities that covers the period until 2004 Theagreement covered the joint achievement of a balancedbudget by 2002 as well as financial burden sharingarrangements The provinces undertook to contribute toan average budget surplus over the whole period of notless than 075 of GDP up to 2004 Temporary under-runs of ndash 015 of GDP are allowed if over the wholefinancial burden-sharing period the averaged value of075 GDP is maintained The local authorities under-took to balance their budget up to 2004 Temporaryoverruns of 01 of GDP are allowed if over the wholeperiod the average position of a balanced budget isattained The system of monitoring and enforcementincludes possible fines subject to unanimous decisionfrom all interested parties The flood disaster in 2002 ledto a temporary suspension of the rule that is not takingaccount of flood-related expenditure in the years 2002and 2003
In Belgium the coordination of the budget balance posi-tion of various levels of government is ensured by theagreement concluded initially in 2000 and renewed in2002 between the federal government the communitiesand regions to adhere to the budgetary targets as recom-mended each year by the High Council of Finance (1)The communities and regions draw up internal medium-term stability programmes each year at least equal induration to the Belgian stability programme which areevaluated by the High Council of Finance The agree-ment covers the period of 2001ndash05 The cooperationagreement does not include formal sanctioning proce-dures in case of deviation from the permissible deficitsHowever the federal government can restrict the bor-rowing capacity of communities and regions for a periodof up to two years upon recommendation of the HighCouncil of Finance and after the regions involved havebeen consulted (IMF 2001)
On 21 March 2002 the federal government and theLaumlnder in Germany agreed on a kind of National Sta-bility Pact for the implementation of the SGP (for amore detailed description see the case study on Ger-many in Section 342) The federal government and theLaumlnder (including the local governments falling within
their competence) commit to comply with the budget-ary rules of EMU and lsquoshall strive towards a reductionin net borrowing with the aim of achieving balancedbudgetsrsquo The Financial Planning Council to which theFederal Minister for Finance the Federal Minister forCommerce and Labour the Finance Ministers of theLaumlnder as well as representatives of local authoritiesand local authority associations belong discusses thecompatibility of the budgetary developments of territo-rial authorities with the provisions of the SGP TheFinancial Planning Council will issue recommenda-tions on budgetary policies mdash and in particular on acommon expenditure line mdash taking into account theeconomic and fiscal factors Regarding enforcementthe Financial Planning Council will discuss the reasonsof non-respect of the rules and give recommendationsin order to restore budgetary discipline
In Italy a domestic stability pact came into force throughlegislation adopted in connection with the budget law for1999 It aims at improving the budget balances of localgovernments by fixing targets for the reduction of theirdeficits Healthcare expenditure which accounts for overtwo thirds of regional expenditure is subject to a separateagreement The Treasury is to monitor cash flows duringthe year and report on a quarterly basis to the conferencefor relations between regions and State and the confer-ence for state-municipalities which are expected to indi-cate measures to achieve the targets in case of divergencePossible fines under the budgetary rules of the Treaty andthe SGP are to be levied on the local authorities that havefailed to meet their targets in proportion to the overshootfor which they are responsible
In Spain the General Law of Budgetary Stability enactedin 2001 has taken effect from 2003 (for a more detaileddescription see the case study on Spain in Section 341)The central feature is that all general government sub-sectors should show a surplus or a balanced budgetTemporary deficits are allowed only in exceptional situ-ations where two-to-three year plans will be discussedin Parliament to return to a surplus or a balanced budgetThe central government monitors budgetary executionand assesses the degree of fulfilment of the objectivesAs a part of enforcement the central government will beable to condition any recourse to debt by sub-nationalgovernments Possible fines under the budgetary rules ofthe Treaty and the SGP will be shared by those publicentities responsible for the deficits
yen1part Advisory board on fiscal policy of the government and communities andregions
194
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
This short overview shows differences and similarities inthe way Member States address the challenge of coordi-nating the overall budgetary position across levels ofgovernment The differences reflect historical circum-stances variation in political structure and diversity inbudgetary processes Some Member States have chosento replicate the medium-term objective of the SGP oflsquoclose to balance or in surplusrsquo at the local or regionallevel while others have chosen to define specific budget-ary targets on a yearly basis In some cases the arrange-ments are laid down in national law while in others theyare formulated as an agreement between levels of govern-ment There are also institutional differences with respectto the way the arrangements are implemented and moni-tored Finally some arrangements specify the specificactions to be taken in case of non-compliance such asimposing sanctions while others do not
More experience with the implementation of thearrangements is needed before firm conclusions can bedrawn on their effectiveness in contributing to the over-all fiscal objectives of the SGP A crucial issue is howthe mechanisms are implemented when a divergencearises between targets and expected outcomes that can-not be attributed to exceptional circumstances In thisrespect a strong legal base and enforcement mechanismwould be expected to contribute to the credibility andeffectiveness of the arrangements
33 Fiscal decentralisation and automatic stabilisation
Apart from the issue of how fiscal decentralisationaffects the capacity of Member State to achieve soundand sustainable public finances it may also be relevantas regards the effects of fiscal policy on the stabilisationof economic activity and in particular the operation ofautomatic stabilisers Stabilisation could be beneficialboth to smooth taxes and consumption over time and toavoid excessive output and employment variability andboom-bust fluctuations
EMU raises particular concerns as regards the role ofnational fiscal policies for stabilisation purposes (seeEuropean Commission 2001a) it is widely argued thatgiven the loss of national monetary policy in EMUbudgetary policy may need to play a more significantrole in smoothening the impact of country-specificshocks on real output The philosophy underlying theTreaty and the SGP reflects widespread scepticism onthe use of discretionary fiscal policies for stabilisation
purposes (European Commission 2002a) and the normfor budgetary behaviour in EMU should be to let auto-matic stabilisers operate freely over the economiccycle Adhering to budgetary positions of lsquoclose to bal-ance or in surplusrsquo will provide for an adequate safetymargin to prevent nominal budget deficits from breach-ing the 3 of GDP reference value while letting auto-matic stabilisers play fully
The traditional literature on fiscal federalism providesarguments in favour of centralising the stabilisationfunction Lower levels of government might not have theright incentives to provide an optimal level of stabilisa-tion since a large part of their stabilisation effort wouldleak away to other jurisdictions Likewise local govern-ments could try to free ride on the effort of others Fur-thermore the possibilities of local governments to runcounter-cyclical policies (for example by means of let-ting automatic stabilisers work) are in many cases lim-ited given the existence of borrowing and budgetingrestrictions As a result it is widely believed that theremay be good reason to shield the income of lower levelsof government to some extent from cyclical fluctuationsTer-Minassian (1997) summarises the broad consensusin the literature that the central government should beassigned taxes that among other things have a higherincome elasticity lsquothat is to provide the central govern-ment with stabilisation instruments and also to shelter tothe extent possible the budgets of sub-national govern-ments from cyclical fluctuationsrsquo (1) This kind of sheltercan be achieved either by only assigning tax bases tolower levels of government that are sufficiently stableover the cycle or by devising a system of shared taxes orgrants that correct for cyclical variability in own taxes atlower levels of government
The empirical literature on fiscal federalism and auto-matic stabilisation focuses mainly on the US andsometimes on other large federalist States as well (suchas Canada and Germany) where state budgets are largeenough to potentially influence overall automatic stabi-lisation The results indicate agreement that more strin-gent borrowing controls are associated with less cycli-cal response of the budget balance at the level of thestates within federations (Soslashrensen et al (2001)Alesina and Bayoumi (1996) Bayoumi and Eichen-green (1995)) Moreover Alesina and Bayoumi (1996)
yen1part Ter-Minassian (1997) also provides an overview of which taxes would bemore suitable for assignment to the central government or to lower levels ofgovernment
195
P u b l i c f i n a n c e s i n E M U 2 0 0 3
find that the lower flexibility of the budget balancedoes not affect output variability at state level withinthe US indicating that balanced budget rules for the USstates are effective in enforcing fiscal discipline buthave no costs in terms of increased output variabilityLastly Soslashrensen et al (2001) specifically investigatethe cyclical variability of different components of thebudget for the US states indicating that state revenuesand expenditure are both pro-cyclical but the pro-cyclicality of revenue dominates so that the overallbudget balances improve during upturns and worsenduring downturns These results indicate mdash at least forthe US mdash a gap between practice and the recommenda-tion of shielding sub-national revenues and expenditurefrom cyclical variations No firm conclusions arereached however on the desirability of automatic sta-bilisation at lower levels of government
From the point of view of the EU a relevant question iswhether the trend of fiscal decentralisation might impacton the extent to which budgets of sub-national govern-ments are shielded from cyclical variations Greater taxautonomy at lower levels of government might increasethe cyclical variability of revenue at local level and leadto a degree of procyclical behaviour if borrowingrequirements are in place For example if tax revenuesat lower levels of government decrease in a recessionthen expenditure would have to be cut as a result
To investigate the issue it is necessary to examine thevariability of budget balances at the stateregional levelsof government and the cyclical variability of revenuesand expenditure at state levels of government Table V9presents figures for aggregate net lending (ndash) and bor-rowing (+) at the state-level government for the federal-ist Member States (including Spain) and the local levelfor the Nordic countries It also shows the developmentof the output gaps over time The aggregate budget bal-ances at state or local level generally show little cyclicalvariation over time For Belgium and especially Ger-many the aggregate budget balances of the states couldindicate a small degree of cyclical sensitivity Neverthe-less the limited period for which data are available doesnot allow for a firm conclusion in this respect
The crucial question is whether a low degree of cyclicalresponse of the budget balance at state or local level isdue a low cyclical variability of revenues and expendi-ture According to the proposition that lower levels ofgovernment should be shielded from cyclical variationsin their revenues and expenditures one would expect asteady growth in real revenues and expenditures of lowerlevels of government at the rate of trend GDP (in absenceof any change in the size of lower levels of government)If on the contrary revenue and expenditure at lower lev-els of government would be responsive to the cycle thena lack of cyclical movement of state budget balancescould indicate a degree of pro-cyclical behaviour possi-
Table V9
Aggregate budget balances at state level (AT BE ES DE) local level (DK FI SE) and output gaps
1995 1996 1997 1998 1999 2000 2001
AT B balance 01 03 06 04 03 02
Output gap ndash 08 ndash 09 ndash 15 02 07 19 05
BE B balance ndash 08 ndash 04 ndash 01 03 04 02 08
Output gap ndash 07 ndash 15 ndash 01 00 10 22 10
ES B balance ndash 06 ndash 06 ndash 03 ndash 03 ndash 02 ndash 05 ndash 05
Output gap ndash 30 ndash 31 ndash 18 ndash 03 07 16 09
DE B balance ndash 12 ndash 11 ndash 12 ndash 07 ndash 05 ndash 04 ndash 13
Output gap 02 ndash 08 ndash 10 ndash 07 ndash 02 11 04
DK B balance 05 ndash 03 ndash 04 ndash 04 01 00 ndash 01
Output gap 04 04 08 07 10 14 06
FI B balance 14 09 ndash 04 ndash 01 ndash 01 02 ndash 03
Output gap ndash 25 ndash 11 17 28 23 43 14
SE B balance 00 ndash 02 ndash 49 ndash 02 ndash 54 03 ndash 02
Output gap ndash 04 ndash 12 ndash 10 ndash 02 15 30 13
Source Commission services
196
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
bly influenced by borrowing restrictions (1) In this casethe typical pattern would be that revenues show a degreeof cyclical variability and that expenditures are adjustedin a pro-cyclical manner as a result
The following figures therefore plot the yearly changesin the level of real revenues and expenditures at the stateregional level of government as well as the output gapsGraph V4 shows the results for Germany from 1991onwards Changes in the level of revenues and spendingshow a strong correlation and seem to move to somedegree at least in line with the economic cycle The lastyear shows a stronger drop in revenues than expenditurewhich is in line with the recent increase of the budget
deficit for the states In sum the figures suggest a degreeof cyclical variation in revenues followed by a smallerdegree of pro-cyclicality on the spending side (and hencesome degree of cyclical variation of the deficit ratio atthe level of the states)
Graph V5 shows the available results for the Spanishregions The growth rates of expenditure and revenuewell above at the trend rate of GDP are in line with theincrease in fiscal responsibilities of the Spanish regionsas reported elsewhere (for example Committee of theRegions 2001) The figures seem to provide little or noindication of pro-cyclicality in spending A relevantquestion however is whether the recent budgetaryreform in Spain which combines a higher degree of taxautonomy for lower levels of government and balancedbudget requirements could lead to a greater degree ofpro-cyclicality in the future (see case study on Spain inSection 341)
Graph V6 shows the results for Denmark a unitary Statewith a high degree of decentralisation and local auton-omy A relevant feature of the Danish system is thatblock grants to lower levels of government are adjusted
for changes in the burden of tasks that the central govern-ment assigns to local governments and the effects ofbusiness-cycle fluctuations (OECD 2003b) Thesecyclical variations are covered by the lsquobudget guaran-
yen1part Rodden (2002) presents an index of borrowing autonomy for lower levelsof government where a score of 1 implies no borrowing autonomy and ascore of 5 a high degree of borrowing autonomy The scores for the Mem-ber States as shown here are States Austria 185 States Spain 28 StatesGermany 27
Graph V4 Germany output gap and changes in the level of real revenues and expenditure at State level
Source Commission services
ndash 6
ndash 4
ndash 2
0
2
4
6
8
1992 1993 1994 1995 1996 1997 1998 1999 2000 2001
TE State TR State Output gap
197
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Graph V5 Spain output gap and changes in the level of real revenues and expenditure at State level
Source Commission services
Graph V6 Denmark output gap and changes in the level of real revenues and expenditure at local level
Source Commission services
ndash 4
ndash 2
0
2
4
6
8
10
12
14
16
1996 1997 1998 1999 2000 2001
TE State TR State Output gap
ndash 4
ndash 3
ndash 2
ndash 1
0
1
2
3
4
5
1996 1997 1998 1999 2000 2001
TE TR Output gap
198
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
teersquo which implies that the State adjusts its grants forchanges in expenditure (such as unemployment benefits)that are caused by cyclical variations As expected sincelocal governments are not allowed to run deficits (exceptin short periods) changes in revenues and expendituresshow a large correlation over time In principle thecounter-cyclical adjustments of the block grant need notimply that total expenditures at local level are counter-cyclical Nevertheless a marked difference with thecases described above is that revenues and expendituresat local level show a rise in 2001 against a background oflower growth which may indicate a degree of counter-cyclicality in line with the philosophy of the SGP
In interpreting the results of the examples above oneshould be careful to recognise the preliminary nature ofthe analysis The analysis is a partial one as it onlyinvestigates the cyclical pattern of government financesat the state and local level and is based on aggregate datafor a limited period of time A follow-up study couldinvestigate the total degree of cyclical variability ofbudgets of Member States and investigate the contribu-tion of all levels of government (that is central statelocal and social security) as well as its actual effects onoutput
34 Case studies
341 Spain
Introduction
Spainrsquos budget deficit was gradually reduced during thesecond half of the 1990s in line with EMU fiscal rulesMore recently attention has focused on how to ensurebudgetary stability in a context of increasing fiscal decen-tralisation At present territorial governments representmore than 30 of total general government expenditureand nearly 70 of general government investment
The General Law of Budgetary Stability (GLBS) cameinto force in 2003 Its adoption follows the new financingsystem for regional governments implemented in 2002which implies considerably greater taxation powers forregional authorities and widens joint fiscal responsibilityThe GLBS aims at ensuring that increased decentralisa-tion of public finances should not put at risk overall budg-etary stability It does so by requiring that all the generalgovernment sub-sectors should show a surplus or bal-anced budget and by introducing new budgetary proce-
dures and norms These institutional changes are seen asmore important in ensuring budgetary stability than themere setting of quantitative targets
The principle of budgetary stability implying a surplusor a balanced budget will be applied to all public entitiesThus apart from the central government (State socialsecurity and autonomous entities) and regional and localgovernments public bodies (even if they are notincluded in the general government definition on anational accounts basis) are covered by this new legalframework Therefore the scope of the GLBS is widerthan the general government definition on ESA95 basis
Main principles of the GLBS
The GLBS is based on four basic principles
bull each entity of the public sector must fulfil the crite-rion of budgetary stability which is defined in thelaw as lsquoa situation of balance or in surplus in termsof financing capacity according to ESA95 method-ologyrsquo As far as public entities and enterprises areconcerned the definition of budgetary stability ismore vague For these entities budgetary stabilitymeans lsquoa balanced financial situation which mightimply if necessary the adoption of restructuringstrategies to avoid or lower economic losses andprovide adequate profits for the fulfilment of theirinstitutional purposesrsquo
bull multiannual framework for budgetary setting whichimplies that each public entity must prepare thebudget on the basis of medium-term projections(three years) This framework is in line with the timehorizon of the stability programme
bull transparency meaning that each public entity mustprovide enough information to allow the assessmentof its budgetary situation and the fulfilment of thebudgetary stability criterion
bull efficiency in the use of public funds in order toaccomplish the budgetary stability criterion for eachpublic agent To give practical implementation tothese principles the GLBS introduces new budget-ary procedures both for the public sector as a wholeand for each sub-sector
199
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Budgetary procedure innovations concerning the whole of the public sector
All public sector entities are required to modify theirspecific budgetary procedures to ensure compliance withthe budgetary stability criterion The central governmentis responsible for the assessment of budgetary stability inthe public sector as a whole (1) In particular all publicentities will have to consider the following
bull budgetary deficits have to be justified by the publicsector entity concerned and will require the formula-tion of a medium-term (three years) plan to restore abalanced budget situation
bull in the first quarter of each year the central govern-ment will release the budgetary objectives for thenext three years for the whole general governmentsector and for each sub-sector These objectives areto be discussed in the Parliament together with themacroeconomic scenario for the same period set outin the stability programme
bull before 1 September of each year the General Inter-vention of the State (IGAE henceforth) will submita report on the degree of fulfilment of the stabilityobjectives in the previous year The report will besent to the Fiscal and Financial Policy Council (theofficial body responsible for coordinating fiscal pol-icy between central and regional governments)Imbalances will have to be justified and will requirethe formulation of a plan to correct them
bull the financial penalties due to the non-fulfilment ofthe commitments assumed by Spain within SGP willbe shared by those public entities responsible for thedeficits
Budgetary procedures innovations for central government and social security
The following rules apply to the central government andsocial security system taking into account that the budg-etary stability objectives for these two sub-sectors willbe considered jointly until the process of separation ofsocial security financing is completed in 2012
bull Before releasing the Budget Law the Ministry ofthe Economy prepares a multiannual programme ofexpenditure and revenues for each year specifyingthe spending commitments for lsquoevery budgetarypolicyrsquo
bull Along with the budgetary stability objectivesannounced in the first quarter of each year for thenext three years the government sets a maximumexpenditure limit for the State on an annual basis forthe same period
bull Based on the maximum limit set for the non-financialexpenditure for the State a contingency fund is cre-ated up to a maximum of 2 of the such limit Thisfund is to be used for changes in spending commit-ments to meet unforeseen circumstances The Minis-try of Public Finances has to inform parliament aboutthe use of such funds on a quarterly basis Thus newspending commitments have to be financed throughthis fund or by reducing other expenditures It is notpossible to carry over unspent amounts in the fundfrom one year to the following one
bull Surpluses recorded by the State are allocated to debtreduction while those registered by the social securityare allocated with priority to the pension reserve fund
bull Public entities and enterprises not included in the gen-eral government sector on a national accounts basisbut dependent on the central government have to con-tribute to budgetary stability Thus in case of lossesaffecting negatively the central government budget-ary objective a medium-term plan containing appro-priate measures to remedy this situation has to beadopted The legal procedures contents and deadlinesfor these plans will be set out in a specific regulation
Budgetary procedure innovations concerning regional governments
In addition to the rules that apply to the whole of the pub-lic sector regional governments are required to respectthe following criteria
bull The budgetary stability objective for the wholeregional government sub-sector released by the cen-tral government in the first quarter of each year willhave to take into account a previous report by theFiscal and Financial Policy Council Once the budg-etary stability objective for the whole regional gov-
yen1part This responsibility will be to some extent shared when assessing thebudgetary stability for regional and local governments The so-called lsquoFis-cal and Financial Policy Councilrsquo made up of central and regional author-ity representatives will carry out the assessment for regional governmentsIn turn the lsquoNational Committee for Local Entitiesrsquo co-assesses the finan-cial situation of local governments
200
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
ernment sub-sector is released by the centralgovernment the Fiscal and Financial Policy Councilhas one month to translate it into budgetary objec-tives for each regional government If in this periodthe Council does not reach an agreement eachregional government has to prepare its budgetrespecting at least a balanced budget objective
bull If a regional government approves its budget fore-seeing a deficit or fails to achieve the budgetarystability objective set in the previous budget amedium-term plan aiming at rectifying this situationis required This plan must include all the necessaryrevenue and expenditure measures to restore budg-etary stability within three years and requires theapproval by the Fiscal and Financial Policy CouncilThe Ministry of Public Finances is responsible forthe monitoring of this plan
bull The State can condition the issuance of new debt andthe recourse to bank credit by the regional govern-ments to the fulfilment of their budgetary stabilityobjectives In addition the Ministry of PublicFinances is allowed to request information fromregional governments and to set up a public informa-tion agency for loan operations debt issuance andassumption of risks by regional authorities
bull Regional governments are able to take measures soas to reach budgetary stability in relation to publicentities and enterprises not included in the generalgovernment sector on a national accounts basis butdependent on regional authorities
Budgetary procedure innovations concerning local governments
The procedures that apply to local governments arebroadly the same as to those addressed to regionalauthorities In this case the National Committee forLocal Entities plays a role analogous to that of the FiscalFinancial Policy Council
A tentative assessment
The GLBS can be seen as a means to keep publicfinances on a sound basis so as to respect the fiscal com-mitments undertaken at European level while allowingto share fiscal responsibility among all general govern-ment tiers It redresses the potential asymmetry regard-ing budgetary stability between the central government
which is the only responsible for fiscal commitmentsvis-agrave-vis EU authorities and territorial governmentswith an increasing role in public expenditure
The law aims at entrenching the dynamics of fiscal con-solidation based on the expenditure side through theannual limit on expenditure at the State level and the so-called contingency fund An additional positive featureof the GLBS is the multiannual stability objectivesannounced by the government for budgetary setting Inorder to keep credibility the budgetary target set in thefirst quarter of each year should be only subject to lim-ited changes in the Budget Law and the USP
However despite the validity of the central goal ofensuring that fiscal decentralisation remains compatiblewith the budgetary stability as defined by EMU fiscalrules some questions arise as to the means foreseen toachieve it A principal criticism is that the objective ofbudgetary stability defined in nominal terms mighthamper the stabilisation function of fiscal policy sincethe effect of the cycle on the budget is not taken intoaccount Nevertheless the difficulty of estimating cycli-cally-adjusted balances at sub-national level can explainthe choice of applying objectives in nominal terms
The adoption of the GLBS complements that of the newsystem for financing of regional governments By mak-ing regions finances increasingly dependent on own taxrevenues the new system of financing increases the sen-sitivity of regional budgets to the economic cycle aggra-vating the risk of pro-cyclical policies Should deficitsoccur the responsible public entities have three years torestore the balanced budget In case of a severe reces-sion this period might prove to be insufficient Somemargin of flexibility however can be expected in theimplementation of the three-year plans to restore a bal-anced budget situation The issue raised by GLBS con-cerning its compatibility with a proper functioning ofautomatic stabilisers will need some time to be assessed
It is also claimed that the GLBS could reduce publicinvestment causing serious problems for a catching-upcountry such as Spain The objective of a balancedbudget or a surplus would mean that public investmentcould only be financed through current revenues How-ever recent research suggests that public investment isnot constrained by a ban on deficits as no significantdirection of causality has been detected between publicinvestment and deficits
201
P u b l i c f i n a n c e s i n E M U 2 0 0 3
The law could have been more ambitious in someaspects The joint objective of budgetary stability for theState and social security system (up to 2010) will allowthe State to record deficits without having to presentconsolidation plans at all A stricter formulation couldhave required that the social security surpluses beentirely allocated to the reserve fund together withrequirements for the State balances
The transparency at each level of government is essentialfor the effectiveness of the GLBS The law includesadvances in this direction particularly concerning sub-national governments for which budgetary informationhas traditionally been poor The GLBS may pave theway for additional information requirements recognis-ing that stability and transparency are complementary
Finally this law will only produce the awaited results interms of long-term sustainability only if the GeneralBudgetary Law which contains the main budgetary pro-cedures and the social security system are properlyreformed The latter is especial relevant for Spain giventhe expected budgetary impact due to ageing which can-not be tackled by the simple implementation of the GLBS
342 Germany
The constitutional framework
In constitutional terms the Federal Republic of Germanywas actually founded by the Laumlnder (see Praumlambel) As aconsequence Article 70 (1) of the German constitutionclearly states that lsquothe Laumlnder have the right to legislationas long as this constitution does not defer this right to thefederal levelrsquo The following articles then define whichare the responsibilities of the federation and in which areasboth government levels may intervene In line with theseregulations Article 104a states that lsquothe federation and theLaumlnder are mdash in clear separation mdash in charge of theexpenditures which result from the responsibilities whichthis Constitution confers upon themrsquo
Article 106 then specifies which taxes are collected bywhich level of government Article 105 (3) states thatfederal laws on taxation which regard taxes at least par-tially collected by lower level of government are subjectto approval by the Bundesrat (Upper chamber of Parlia-ment composed of Laumlnder representatives) (1) Further-more Article 106 (3) states that the uniformity of livingconditions on the federal territory has to be guaranteed
Expenditure and revenue by levels of government
On the basis of these constitutional provisions theregional and local governments play a substantial rolefor the development of public finances in Germany asindicated already in Table V7
On average expenditure by regional governments is of asimilar magnitude as federal spending If one incorpo-rates spending by local entities federal expenditureaccounts for only 40 of total expenditure by all levelsof government (that is excluding social security sys-tems) As shown in Graph V3 the share of lower levelsof government account for almost half of the generalgovernment budget deficit
Given the importance of lower levels of government forpublic spending and revenue collection the coordinationof budgetary policies is obviously very relevant for therespect of the SGP Furthermore given that only the fed-eral government is at the European level responsible forthe respect of the deficit and debt criteria the FederalMinister has a strong interest in having his policies sup-ported by the other levels of government In the run-upto EMU however attempts to agree upon a national sta-bility pact failed due to constitutional problems andpolitical considerations Recently however the respon-sibilities of the Finanzplanungsrat (Financial PlanningCouncil) have been clearly reinforced
The role of the Financial Planning Council
The Finanzplanungsrat itself consists of the FederalMinister for Finance the Laumlnder Ministers of Financethe Federal Minister for Economics and representa-tive(s) of local authorities It meets twice a year (nor-mally in June and November) following the presenta-tion of the economic forecast and the respective meetingof the working group Steuerschaumltzung (lsquotax revenuesestimatersquo) (2) Given that the April economic forecasthas a medium-term time horizon (that is normally thefollowing three years) the June meeting of the Finanzpla-
yen1part Following regional elections in the Laumlnder of Lower Saxony and HesseChristian Democratic-led Laumlnder governments currently lsquocontrolrsquo 41 votesout of a total of 69 votes conferring on the opposition parties an importantrole in economic policy-making
yen2part At the end of January the Federal Government normally presents itsJahreswirtschaftsbericht (annual economic review) which also contains itseconomic projections for the spring meeting of the Finanzplanungsratthese projections can be revised The projection of the Federal Governmentserves as a basis for discussion in the Finanzplanungsrat however currentprojections of economic research institutes and of international organisa-tions including the EU are also presented The medium-term projections ofthe federal government are only published once a year in April
202
P a r t VM e e t i n g t h e E U rsquo s b u d g e t a r y r e q u i r e m e n t s
n a t i o n a l e x p e n d i t u r e r u l e s a n d f i s c a l r e l a t i o n sa c r o s s l e v e l s o f g o v e r n m e n t
nungsrat also discusses medium-term budgetary projec-tions while the November meeting mdash based on theshort-term forecast published in October mdash will onlydiscuss prospects for the current and the following year
Automatic stabilisation
When the Finanzplanungsrat met for the first time inMarch 1968 it was to contribute to the fine-tuning of thebusiness cycle in line with Article 109 (2) of the Germanconstitution whereby the federal government and theLaumlnder in their budgetary management have to respectthe requirements of the overall macroeconomic equilib-rium (1) Since the early 1980s the focus has shifted tofiscal consolidation in the mid-1990s it was decidedthat the rise in (nominal) expenditure of all levels of gov-ernment should not exceed 2 per year Regarding theissue of automatic stabilisation past experience showsthat lower levels of government tend to follow a pro-cyclical pattern if in one year tax revenues turn out to behigher than originally budgeted nominal deficits firsttend to decline In the following year the growth rate ofexpenditure would clearly accelerate (see Graph V4)
The Law on Budgetary Principles
Following the adoption of the SGP and the introductionof the euro the Ecofin Council in its opinion on theupdated German stability programme had repeatedlyrecommended to the German authorities to agree upon akind of lsquonational stability pactrsquo in order to make theattainment of the budgetary targets of the updated pro-grammes more credible and in order to avoid pro-cyclicalpolicies In line with this recommendation a modifica-tion to Article 51a Haushaltsgrundsaumltzegesetz (lsquoLaw onBudgetary Principlesrsquo) was decided upon on 20 Decem-ber 2001
Article 51a (1) of the new law now contains a clear ref-erence to the responsibilities of all levels of governmentto respect Article 104 of the EU Treaty and proclaimsthe overall aim of bringing the deficit down in orderto reach a balanced budget Article 51a (2) states thatlsquothe Finanzplanungsrat will issue recommendations on
budgetary policies in particular on a common expendi-ture line taking into account the economic and fiscalfactors The Finanzplanungsrathellip shall discuss the con-sistency of budgetary developments in particular of thedevelopment of expenditures and deficits by the federa-tion and by the Laumlnder (including the local authorities)with the regulations of Article 104 of the EU Treaty andwith the European Stability and Growth Pactrsquo FinallyArticle 51a (3) now stipulates that in case of non-respectof the principles described in Article 51a (1) and (2) theFinanzplanungsrat will discuss the reasons thereof andlsquogive recommendation in order to restore budgetary dis-ciplinersquo
Following the Commission recommendation to the Coun-cil of 30 January 2002 to give an early warning to Ger-many the date of implementation of the new Article 51aHaushaltsgrundsaumltzegesetz was carried forward from1 January 2005 to 1 July 2002
Furthermore in its special meeting of March 2002 theFinanzplanungsrat agreed upon ambitious expendituretargets for 2003 and 2004 Federal expenditure was pro-jected to decrease by 05 per year and Laumlnder expend-iture was to rise by 1 per year in nominal terms onlyIn the November 2002 meeting it was decided that theexpenditure line for 2005 and 2006 should be discussedin the first meeting in 2003 Furthermore not least due tothe costs implied by reconstruction from the floods ofsummer 2002 the expenditure pattern was changed butthe targeted overall rise in expenditure remained almostunchanged
In its opinion on the updated stability programmeadopted in January 2003 the Council urged the Germanauthorities to respect the agreed expenditure targets for2003 and 2004 and to reach an agreement on ambitiousexpenditure targets for 2005 and 2006
While recent legal developments clearly constitute animprovement compared with the preceding rules itremains to be seen how effective the new law turns outto be in practice The lack of threat of sanctions goingabove recommendations could imply less compliancewith mutually agreed targets
yen1part lsquoBund und Laumlnder haben bei ihrer Haushaltsfuumlhrung des Erfordernissen desgesamtwirtschaftlichen Gleichgewichts Rechnung zu tragenrsquo
203
Part VI
Member State developments
1 Belgium
Recent developments
Despite an unfavourable macroeconomic context a 01 of GDP general government surplus was achieved in2002 Initially a government surplus of 03 of GDPwas planned in the 2002 budget under a 13 real GDPgrowth assumption however in the course of the yearactivity proved to be more subdued than expected andreal GDP growth reached 07 only
In March and July 2002 budgetary control exerciseswere organised associating all levels of governmentExpenditures were contained particularly at federallevel of government applying the lsquoanchor principlersquowhich consists in holding the utilisation rate of credits
under or at the rate of 2001 in part of the year was instru-mental in controlling spending In the social securitysector health spending was better controlled Thus thegovernment primary surplus in 2002 was maintained at ahigh level 61 of GDP
The government debt ratio which reached 1085 ofGDP in 2001 was lowered to 1053 of GDP in 2002During the period 2000ndash02 the pace of debt reductionslowed primarily due to low GDP growth in real andnominal terms Moreover in 2001 and to a lesser extentin 2002 financial operations included in the stock-flowadjustment decelerated the reduction process theseoperations consisted of the assumption by the State ofdebt in a number of public entities
Table VI1
Composition and balances of general government Belgium (1) (as of GDP)
2000 2001 2002 2003 2004
Government balance (2) 01 04 01 ndash 02 ndash 01
mdash Total revenue 495 498 502 495 492
Of which mdash current taxes 304 302 304 302 298
mdash social contributions 161 164 165 164 164
mdash Total expenditure (2) 494 494 501 497 493
Of which mdash collective consumption 78 79 81 81 81
mdash social transfers (3) 287 293 300 305 305
mdash interest expenditure 68 65 60 55 49
mdash gross fixed capital formation 18 15 16 14 15
Primary balance (2) 69 70 61 53 48
Pm Tax burden 459 460 463 459 455
Government debt 1096 1085 1053 1027 989
Pm Cyclically-adjusted balance ndash 12 ndash 04 01 02 00
Pm Cyclically-adjusted primary balance 56 62 61 57 49
(1) Commission spring 2003 economic forecasts(2) Data for 2001 (except cyclically-adjusted) include UMTS receipts of 02 of GDP(3) In kind and other than in kind
Source Commission services
207
P u b l i c f i n a n c e s i n E M U 2 0 0 3
The 2003 budget expected the general governmentaccounts to be in balance on the basis of a 21 realGDP growth assumption Attaining a government sur-plus of this order of magnitude was also recommendedin the 2002 BEPGs The balanced budget objective isexpected to be reached while implementing further taxcuts that are to be compensated by equal decline in inter-est payments
The federal government real primary expenditure is pro-jected to increase by 13 in 2003 The general govern-ment primary surplus was expected to decline somewhatin 2003 as a result of a reduction in the primary surplusof Entity I (Federal government and social security) Thegovernment debt ratio was projected to be lowered to anestimated 1023 of GDP and no ad hoc financial oper-ations are planned for 2003
However due to further deterioration in the externalenvironment the real GDP growth assumption underly-ing the budget estimates was revised downwards to14 in February 2003 while the government balanceobjective was maintained Such an adjustment is possi-ble by further containment of federal expenditure apply-ing the lsquoanchor principlersquo higher-than-expected taxrevenues registered at the end of 2002 and a morefavourable projection for interest payments (made possi-ble by interest rate developments)
In the 2003 Commission spring forecasts real GDPgrowth is further adjusted downward to 12 in 2003consequently under the assumption of no further adjust-ment measures a government deficit of 02 of GDP isforecasted At the same time the government debt isexpected to decline to 1027 of GDP in 2003
Reducing the debt ratio within a global budgetary strategy
In recent years Belgium has succeeded in reducing itsgovernment debt ratio by means of achieving highgovernment primary surpluses The 2002 update of thestability programme projects a reduction in the govern-ment debt ratio to 94 of GDP in 2005 this is to beachieved by sustaining primary surpluses in the order of55 of GDP The programme also states that no finan-cial operations are foreseen in the time period covered bythe stability programme (2002ndash05) although a decisionto assume a part of the debt of the national railway com-pany (SNCB) may have an impact in coming years
The need to run down debt levels is motivated in part byconcerns about the projected budgetary impact of ageingpopulations (see Part I3) of this report The strategy ofthe Belgian authorities to meet the budgetary costs ofageing populations is based on achieving a steadydecline in the government debt ratio by sustaining highprimary surpluses even beyond the time horizon of thestability programme However sustaining a high pri-mary surplus over the very long run poses a substantialbudgetary challenge especially if at the same timeother budgetary objectives such as a reduction in the taxratio are being pursued in parallel
The Belgian authorities are currently making a muchneeded effort to alleviate the tax burden weighing partic-ularly on labour Since 1999 social securitycontributions paid by employers have been lowered anda programme of tax cuts is currently being implementedcovering the period 2002ndash05 In order to foster labourmarket participation which is low in Belgium by inter-national standards particularly among older workersboth demand and supply of labour are to be encouragedFiscal measures have been decided aiming at a phasedreduction in personal income taxes over the period2002ndash05 also reducing the lsquomarriage penaltyrsquo betterproviding for children charges and sustaining lsquogreenrsquoinitiatives The global impact of the reform is estimatedin the 2002 update of the stability programme at 13 ofGDP in 2006
In 2003 personal income taxes have been cut by 04 of GDP including in particular lowering the top rateimplementing the final stage of phasing out of the crisiscontribution and introducing an income tax credit forlow-wage earners At the same time as from April 2002further reductions in social security contributions havebeen introduced concerning both contributions paid byemployees in order to improve low-paid categoriesrsquo dis-posable income and contributions paid by employersparticularly for older workers Moreover in 2003 ratesfor enterprises income taxes have been lowered from4017 to 3399 while the more favourable fiscal regimefor SME has been maintained the tax rates applied tothem being reduced from 2884 to 2498 These meas-ures in favour of the enterprises are meant to be budget-ary neutral being compensated by offsetting measures
As far as pensions reform is concerned draft legislationhas been tabled to Parliament providing a regulatoryframework for the lsquosecond pillarrsquo of supplementary pen-sions This covers about one third of the employees in
208
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
the private sector Reform of the healthcare sector is alsoan objective of the government
A measure specific to Belgium is the creation of aSilver Fund (September 2001) financed by budgetarysurplus social security surpluses and non-tax rev-enues In 2003 EUR 625 million should be allocated
to the Fund from non-tax revenues At the end of theyear the capital of the Fund is expected to reach about07 of GDP (excluding interests on investments)Use of the reserves of the Fund will be allowed from2010 only under the condition that the level reachedby the government debt ratio would be below 60 of GDP
Table VI2
Key figures of the Belgian stability programme (1) (2003ndash05)
2001 2002 2003 2004 2005
Real GDP growth (annual change) 08 07 21 25 25
General government budget balance ( of GDP) 03 00 00 03 05
Primary surplus ( of GDP) 69 61 55 56 56
Government debt ( of GDP) 1086 1061 1023 979 936
(1) UMTS receipts excluded (02 of GDP in 2001)
Source 2002 update of the stability programme of Belgium
209
2 Denmark
Recent developments and medium-term prospects
In 2002 a surplus of 20 of GDP was recorded thesixth consecutive year with a surplus Swaps amountedto 013 of GDP leaving the surplus in nationalaccounts definition at 19 of GDP (1) This compareswith a target surplus of 16 of GDP in the most recentupdate of the convergence programme and the differ-ence is mainly due to higher-than-expected tax revenues
Compared to 2001 the surplus fell by 08 percentagepoints of GDP In addition to the effects of slowergrowth this resulted from the disappearance of theone-off effect of the UMTS-sale in 2001 amounting to02 percentage points of GDP (2) and secondly theimpact of lsquothe special pension contributionrsquo beingchanged from a public pension scheme to a private one(as a result of the redistributive element being removedfrom the scheme) which has led to a permanent reduc-tion of the budget balance of percentage point
Revenues are still affected by the volatile pension fundyield tax The taxation on pension fund yields waschanged in 2000 (3) The change has resulted in revenuesbeing far more volatile It is estimated that revenues canfluctuate by slightly more than 1 of GDP on averageleading to increased volatility of the surplus on publicfinances of the same amount and changes are very diffi-cult to predict Furthermore the prolonged downturn inthe stock market has resulted in this tax generating
hardly any revenues in 2001 or 2002 and only smallrevenues are expected in 2003 as well In a year withlsquonormalrsquo stock market developments the tax is expectedto generate revenues of around 1 of GDP
The tax burden fell markedly in 2002 due to the afore-mentioned changes to lsquothe special pension contributionrsquoIn the forecast period the tax burden is set to declineonly marginally despite the inclusion of an announcedtax reform in 2004 amounting to 04 of GDP This isa result of the revenue lost by the tax reform being morethan outweighed by the positive effects of increasedGDP growth on public finances and by the fact that thepension fund yield tax as mentioned should begin to gen-erate revenues again
The ratio of primary expenditure to GDP was largelyunchanged in 2002 Over the forecast horizon a fall ofaround 1 of GDP in total government expenditure is pro-jected mostly as a result of declining interest payments
Government consumption rose in real terms by 1 in2002 This is in line with the multiannual target of 1 increase in consumption growth but slightly lower thanthe target actually set for the year The target for the yearwas set at 13 with a subsequent reduction of the targetin 2003 to 07 averaging 1 over the two years
Over the forecast horizon continued surpluses on gen-eral government finances are expected at around 2 ofGDP every year This should result in the debt-to-GDPratio being reduced from the current level of 45 to 40 by the end of 2004 In cyclically-adjusted terms the pri-mary balance remains largely unchanged over the fore-cast horizon thereby indicating that the fiscal policiesare neutral with respect to the cycle
The governmentrsquos medium-term public finance strategycontinues to be focused on reducing the debt by keepingsurpluses of 1 ndash2 of GDP on average every yeartowards 2010 and adherence to the tax freeze The debt-
yen1part Compared to the figures used in the autumn forecast Statistics Denmarkhas implemented methodological changes to the public finance statisticsamounting to a downward revision of the surplus of 013 of GDP
yen2part It should still be noted that Statistics Denmark has decided to treat theUMTS proceeds as an annuity over the next 20 years which is not in linewith Eurostatrsquos recommendation
yen3part The tax rate on yields on equities was increased and the tax rate on yieldson bonds was reduced to ensure the same tax rate on yields from the twotypes of assets As the development in prices on equities is far more volatilethan on bonds the volatility of the revenues from this tax has increasedmarkedly Given the poor performance of the stock market in 2001 thisresulted in lower revenues
210
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
reduction strategy has mainly been implemented in orderto prepare public finances for the budgetary impact of anageing population Sustainability calculations show thatthe public finances are in a good position to handle theimpact of rising expenditure due to the ageing popula-tion However in order to make room for the targeted
average annual growth in real public consump-tion (1) increases in labour force participation rates areneeded to ensure the continued high surpluses
Tax freeze and tax reform
When the current government took office late in Novem-ber 2001 a novelty was introduced into Danish publicfinances as a tax freeze was implemented This sectiondeals with the implications of the tax freeze for expend-iture control and tax reforms
The tax freeze means that no direct or indirect tax mdashwhether legislated in kroner or as a percentage mdash can beincreased Furthermore a nominal ceiling has been puton the property value tax However if there are compel-ling reasons for introducing or raising a tax rate another
tax rate has to be reduced by an amount which leavestotal tax revenues unchanged (2)
The introduction of the tax freeze has several implica-tions First the wording of the tax freeze is actuallyslightly misleading as the implication of the tax freeze isa trend-wise reduction of the tax burden The reductionof the tax burden stems from the fact that excise dutiesexpressed in kroner and the tax base for property valuetax are no longer adjusted in parallel with priceincreases thereby eroding the effective value of rev-enues from these sources The erosion is estimated to bearound of GDP between 2002 and 2010
Second given the overall target for public finances ofa surplus of 1 ndash2 of GDP every year the taxfreeze also has implications for the need for expenditurecontrol If the tax freeze is strictly implemented itimplies that the marginal budget improvement in orderto secure the surplus will have to be taken on the expend-iture side as the tax side has to be taken as given Thiswould mark a break compared to the historic tradition
Table VI3
Composition and balances of general government Denmark (as of GDP)
2000 2001 2002 2003 2004
Government balance (1) 26 28 20 18 21
mdash Total Revenue 573 581 570 562 561
Of which mdash current taxes 468 472 472 469 469
mdash social contributions 33 32 27 26 26
mdash Total expenditure (2) 547 550 549 544 540
Of which mdash collective consumption 77 77 80 80 80
mdash social transfers (2) 349 354 356 357 354
mdash interest expenditure 42 39 36 33 32
mdash gross fixed capital formation 17 19 17 17 17
Primary balance (2) 68 70 56 51 53
Pm Tax burden 496 499 493 490 490
Government debt 474 454 452 427 399
Pm Cyclically-adjusted balance 15 23 19 20 22
Pm Cyclically-adjusted primary balance 57 63 55 53 54
(1) Data exclude UMTS receipts amounting to 02 of GDP in 2001(2) In kind and other than in kind
Source Commission spring 2003 economic forecasts
yen1part The multiannual target for public consumption growth has been 1 for theyears up to 2004 in 2004 and 2005 the target is and from 2006 the tar-get is reduced to
yen2part The definition of the tax freeze also covers user charges and fees therebycovering all sources of income for the public sector
211
P u b l i c f i n a n c e s i n E M U 2 0 0 3
where marginal budget improvements often have beenfound by increasing taxes
A very tentative assessment might suggest that the taxfreeze actually has acted as a mechanism of discipline ofexpenditure control for 2003 and thereby supported theachievement of a growth rate of real public consumptionlower than 1
The strict implementation of the tax freeze is howevernot given as it implies adherence by all levels of govern-ment In Denmark counties and municipalities havecompletely autonomous taxing powers and they governaround two thirds of public consumption Each year anagreement is made between central government and theassociations of counties and municipalities concerningexpenditures in counties and municipalities and theblock grant from the State to these These agreementsare however not legally binding for individual countiesand municipalities
In order to ensure the implementation of the tax freezethe government therefore announced that any breach ofthe tax freeze would be penalised by the central govern-ment This penalising mechanism consists of centralgovernment recuperating the extra revenues earned fromtax increases and using it for reducing State taxes leav-ing the counties and municipalities with a higher tax bur-den but no extra revenues
Despite the announcement of a sanction mechanismbreaches of the tax freeze did occur in late 2002 whencounties and municipalities made their budgets for 2003This could perhaps be viewed as a test by counties andmunicipalities of whether the government would be will-ing and able to implement sanctions However countiesand municipalities have so far only agreed to respect thetax freeze for 2003 No agreements have been made forsubsequent years and a risk therefore exists that the gov-
ernment might find it more and more difficult to enforcethe tax freeze
As mentioned previously the tax freeze in itself implies areduced tax burden over time The reduction is achieved viareducing the tax burden on property and goods submitted toexcise duties It is not as a result of an explicit political pri-oritisation that the reduction in taxes falls on these itemsbut only a result of the fact that these tax rates happened tobe expressed in terms of kroner and not in percentage termswhen the tax freeze was implemented
Increasing the labour supply is considered to be one ofthe main challenges for the Danish economy but the cur-rent prioritisation of where tax reductions are takingplace as a result of the tax freeze does not seem toaddress this problem to any large degree
Judging by comments made by leading ministers a verystrict interpretation has been chosen for the interpreta-tion of when lsquocompelling reasons arriversquo for changing atax rate (1) thereby effectively blocking for revenue-neutral tax reforms which might reduce further the taxeson earned income while increasing other tax rates
The government has announced a tax reform from 2004to 2007 especially intended to help increase the laboursupply The proposal includes an increase in the thresh-old for the intermediate bracket of the State tax and theintroduction of an earned income tax credit When fullyimplemented the tax reductions are estimated to be of GDP by 2007
The main objective of the reform is stated to be anincrease in the labour supply by reducing marginal taxes
Table VI4
Key figures of the Danish convergence programme (2001ndash06)
2001 2002 2003 2004 2005 2006
Real GDP growth (annual change) 10 15 22 18 17 na
General government budget balance ( of GDP)
28 21 22 25 24 na
Primary surplus ( of GDP) 43 34 33 35 33 na
Government debt ( of GDP) 447 439 421 392 367 na
Source 2002 update of the convergence programme of Denmark
yen1part An example given of a reason for changing a tax rate is tax changesimposed by a decision within the EU
212
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
However the way the reform is constructed is probablyone of the more expensive ways to increase the laboursupply so if this was the only target more could havebeen achieved for the same costs by reducing the topState tax or further reductions to the intermediate Statetax It therefore appears that also there has been a strongeye to distributional effects in the proposal for the taxreform This focus does not seem to be an obvious policychoice at the expense of increasing labour supply furthergiven the very equal income distribution in Denmark
The overall reduction of of GDP is larger than whathad been announced as tax cuts within the governments
medium-term projection However if the induced laboursupply effect is too small to achieve the targets set in thepublic finance strategy and further additional fiscal lee-way cannot be found consideration could be given toimplementing a less rigid interpretation of when lsquocom-pelling reasons arriversquo for changing a tax rate in order toincrease the scope for reducing taxes on earned incomeby a reprioritisation within the current tax frameworkThis would also be in line with the Council opinion onthe Danish convergence programme 2001 in which thetax freeze was first presented where it was noted that thetax freeze should not be an impediment to reductions ofmarginal taxes on labour
213
3 Germany
Recent developments
Following a revised 2001 outcome of 28 of GDP thegeneral government deficit is currently estimated to havereached 36 of GDP in 2002 This rise in the deficit ismostly due to a marked shortfall in tax revenues com-pared with the official May 2002 tax estimate andanother widening of the deficit in social security sectorsNot surprisingly the debt ratio has breached the respec-tive Treaty criterion by rising to close to 61 of GDP byend-2002 With the 2002 general government deficitclearly above the respective reference value of theTreaty the Council on 21 January 2003 decided that anexcessive deficit existed in Germany The Council interalia recommended to the German government to imple-ment up to 21 May 2003 the measures announced in thebudget for 2003 amounting to 1 of GDP and to bringthe deficit below the 3 of GDP reference value of theTreaty by 2004 at the latest
The lsquoannual economic reportrsquo by the federal governmentimplies mdash in line with the updated German stability pro-gramme of December 2002 mdash a decline in the nominaldeficit to 28 of GDP in the current year However theCommissionrsquos spring forecast for the 2003 deficit isclearly more pessimistic (34 of GDP) this is mostlydue to the Commission forecast for nominal GDP growthin 2003 being more than 1 percentage point of GDP lowerthan projected by the federal government in its annualeconomic report of January 2003 Furthermore the gov-ernmentrsquos deficit projection still incorporates a consider-able rise in tax revenues from the tax amnesty and theSteuerverguumlnstigungsabbaugesetz which was howeverrejected by the Bundesrat on 14 March (1) Finally theCommission forecast does not incorporate all of theexpenditure savings projected at the level of the Federal
Labour Office by the federal government According toinformation available in early May 2003 the cyclically-adjusted balance would improve in 2003 by around08 percentage points of GDP
In 2004 accelerating growth conducive to employmentcreation should result in a more important fall in thenominal deficit in spite of the implementation of thenext step of income tax reform with a volume of 03 per-centage points of GDP The debt ratio however is pro-jected to rise further to 63 of GDP Based on thestandard no-policy-change assumption (2) and due notleast to the planned tax cuts the improvement in thecyclically-adjusted deficit in 2004 would be minor
A low nominal rise in expenditures and strict expendi-ture control are indispensable to create the margin for thetax cuts planned for 2004 (03 of GDP) and for 2005(around 1 of GDP) To make the achievement of theambitious expenditure targets of the December 2002update of the German stability programme more proba-ble a profound reform of social security systems is anabsolute necessity
Budgetary impact of current reforms
Following the general elections in September 2002 thereconfirmed federal government has embarked on amore ambitious course in economic policy mattersMore courageous structural reforms are to be imple-mented with the overall objective of increasing thegrowth potential of the German economy and reducingunemployment At the same time the federal govern-ment claims that budgetary consolidation will
yen1part In early April this law was still under discussion in the lsquoMediation Com-mitteersquo (Vermittlungsaus-schuss) between the Bundestag and Bundesrat
yen2part The spring 2003 deficit forecast for 2004 is based on the assumption ofunchanged social security contribution rates Furthermore none of themeasures announced by the Chancellor on 14 March could be taken intoaccount as some of them are not yet outlined in detail or have not yet beenintroduced into the legislative procedure
214
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
continue in spite of some expenditure-raising measuresannounced in the meantime (for example the investmentprogramme for local public investment and the pro-gramme for the promotion of private construction mdash bothannounced by Chancellor Schroumlder on 14 March 2003)
Regarding the labour market the first proposals pre-sented in August 2002 by the Hartz Commission havealready been implemented in January and in April 2003(incentive schemes for older workers creation of per-sonal service agencies gradual benefit phase-out for lowincome earners) Furthermore the government hasannounced that further proposals in line with thoseadvanced by the Hartz Commission would enter intoforce in 2004 In addition to the more structural meas-ures the rules for entitlement to Arbeitslosenhilf (assist-ance for the long-term unemployed) have been tight-ened as the income of the unemployed personrsquos partneris now being taken into account
For public finances the overall impact of these measuresis clearly positive but in general difficult to quantifyThe projections of the federal government in this regardappear very optimistic Based on a real GDP growthforecast for 2003 of 1 the average number of unem-ployed is projected to reach 421 million a rise by
150 000 on the preceding year At the same time the fed-eral budget assumes that the Bundesanstalt fuumlr Arbeit(Federal Labour Office) mdash in stark contrast to 2002 mdashwill need no federal transfers in the current year In spiteof a projected rise by 150 000 in the overall number ofunemployed in 2003 the number of unemployed peopleentitled to Arbeitslosengeld is to decrease by 203 000 toa large part due to the implementation of the above-men-tioned measures
On 14 March Chancellor Schroumlder in an official lsquodecla-ration of the governmentrsquo to the Bundestag announcedadditional reform measures regarding Arbeitslosenhilfethe maximum period of entitlement shall be reduced to12 to 18 months from presently up to 32 months Thecurrent benefit withdrawal for long-term unemployedtaking up work will be reduced and the sanctions forthose who refuse a job offer will be tightened FinallyArbeitslosenhilfe and Sozialhilf (social assistance) are tobe merged with the benefit level of Arbeitslosenhilfegoing down to the level of Sozialhilfe
Without any doubt these measures will increase incen-tives to take up a job instead of relying on social benefitpayments which in the past often discouraged unem-ployed from accepting a job offer Furthermore they are
Table VI5
Composition and balances of general government Germany (1) (as of GDP)
2000 2001 2002 2003 2004
Government balance (2) 11 ndash 28 ndash 36 ndash 34 ndash 29
mdash Total revenue 470 455 450 454 455
Of which mdash current taxes 246 230 226 230 232
mdash social contributions 186 185 184 186 185
mdash Total expenditure 459 483 486 489 484
Of which mdash collective consumption 80 79 78 79 78
mdash social transfers (3) 299 300 307 310 307
mdash interest expenditure 34 33 32 32 33
mdash gross fixed capital formation 18 17 16 16 16
Primary balance 45 05 ndash 04 ndash 02 03
Pm Tax burden 432 415 407 413 414
Government debt 602 595 608 627 630
Pm Cyclically-adjusted balance ndash 21 ndash 30 ndash 33 ndash 26 ndash 24
Pm Cyclically-adjusted primary balance 13 03 ndash 01 06 08
(1) Commission spring 2003 economic forecasts(2) Data for 2000 (except cyclically-adjusted) include UMTS receipts of 25 of GDP(3) In kind and other than in kind
Source Commission services
215
P u b l i c f i n a n c e s i n E M U 2 0 0 3
fully in line with the projections of the updated stabilityprogramme submitted by Germany in December 2002which projected a clear decrease of the share of lsquosocialtransfers other than in kindrsquo from a (rounded) 19 ofGDP in 2002 to 18 in 2004 mdash the year when most ofthese measures are to be implemented
Regarding healthcare the government declaration of14 March announced a whole catalogue of differentmeasures insurers are to gain stronger bargaining power
and shall be allowed to negotiate costs directly with doc-tors and the cost sharing by patients shall be increased toraise their cost awareness The Chancellor also proposedthat in the future employees should seek a private insur-ance for the continued salary payment in case of sick-ness which is currently paid by the public health system
According to the Chancellor the overall aim of thereform of the public healthcare system is to bring thecontribution rate down below 13 of gross income
(from a currently estimated 144 ) While this target ishighly welcome given the negative effect of currentcontribution rates on employment the tax financing ofthe so-called versicherungsfremden Leistungen (1) alsoannounced by the Chancellor may constitute a risk to thevery ambitious consolidation programme laid out in themost recent update of the German stability programmeof December 2002 Furthermore Mr Schroumlder remainedelusive on some aspects indicating only that parts of thenecessary measures were being prepared by the respec-tive ministries while important questions regarding thefinancing part should be presented by the Ruumlrup Com-mission in May
Regarding the pension system the government declara-tion conceded that the projections underlying the recentreform (Riester-Rente) have already proved too optimis-tic and underlined that he expected detailed proposalsfrom the Ruumlrup Commission on how to further adopt thepension system
In his speech of 14 March the Chancellor also announcedmeasures aimed at liberalising regulations for crafts andat reducing bureaucracy especially for small and
medium-sized companies As had been frequentlypointed out by the Commission in the past the implemen-tation of such measures appears very important to raisethe currently very low growth potential of the Germaneconomy However their impact on public finances willin the short term probably be negligible In the mediumterm however stronger average growth appears as thebest way to put public finances on a sustainable basis
All in all the reform measures already implemented orcurrently discussed go into the right direction Howeverat the current juncture many proposals are not yetelaborated in such a way as to allow a final judgementon whether their implementation would allow theachievement of the very ambitious budgetary targets ofthe updated German stability programme of December2002 In particular it remains to be seen how courageousthe proposals for the health and pension system turn outand whether the currently discussed reforms will actu-ally be implemented
In this regard a reform of the social security systemsaimed at cutting expenditures and at bringing contribu-tion rates down to acceptable levels appears as the bestmeans to raise the growth potential of the German econ-omy and to put public finances again on a sustainablebasis
Table VI6
Key figures of the German stability programme (2002ndash06)
2002 2003 2004 2005 2006
Real GDP growth (annual change) 05 15 23 23 23
General government budget balance ( of GDP) ndash 38 ndash 28 ndash 15 ndash 10 00
Primary surplus ( of GDP) ndash 05 05 20 20 30
Government debt ( of GDP) 610 615 605 595 575
Source 2002 update of the stability programme of Germany
yen1part Benefits currently paid by health insurance systems deemed not to be cov-ered by contributions
216
4 Greece
Recent developments
Despite strong economic growth in 2002 the generalgovernment deficit was only slightly reduced to 12 ofGDP from 14 in 2001 The government gross debtfell after two consecutive years of increase but remainedat a high level equal to 1049 of GDP in 2002
The favourable domestic economic conditions in com-bination with measures of further containment of taxevasion and enhancement of the tax base contributed toan increase in total revenues at the targeted rate How-ever the general government surplus of 08 of GDPanticipated in the 2002 State budget and in the 2001 sta-bility programme did not materialise mainly due to thedisappearance of the one-off effects of the UMTS-salein 2001 and the reclassification of a number of opera-tions which primarily affected the expenditure side andto a lesser extent government revenues In additionoverruns in almost all categories of primary expendi-tures increased the deficit of the State budget to 35 of GDP compared to the estimated 29 The primarysurplus in 2002 declined to 43 of GDP from 49 achieved in 2001
The 2003 State budget projects a central government def-icit of 43 of GDP corresponding to a general govern-ment deficit of 09 of GDP Notwithstanding the signif-icant increase in the deficit of ordinary budget as well asin the deficit of public investment programme the Statebudget deficit in percent of GDP is estimated to be thesame as in 2002 as a result of the projected high GDPgrowth On the other hand the surplus of social securityfunds and local authorities which is estimated to slightlyincrease in 2003 appears to be the main factor shaping theforeseen reduction in the general government deficit (1)
According to the Commission forecasts the general gov-ernment deficit for 2003 is projected to stand at 11 ofGDP compared to an estimated 12 of GDP in 2002The deceleration in government consumption expendi-tures projected in the State budget could be consideredrather optimistic (2) In fact the uncertainty characteris-ing some categories of expenditures mainly wages mayresult in an overshooting of the projected primary expen-ditures The general government consolidated grossdebt-to-GDP ratio is projected to stand at 101 at theend of 2003 compared to 1002 estimated in the Statebudget and in the 2002 stability programme
In the 2002 stability programme a further improvementin the budgetary position is projected throughout theperiod covered by the programme The central govern-ment deficit is expected to stand at 37 in 2004 and todecline to 23 in 2006 The general government deficitis projected to turn into a surplus from 2005 onwards andthe government debt ratio is expected to decline by123 percentage points in the period 2004ndash06 reaching879 of GDP at the end of 2006
The projected progressive deceleration in primary spend-ing is subject to the risks mentioned above Consideringthe foreseen slowdown in budget revenues an overshoot-ing of the projected spending would lead to a ratherslower than the expected reduction in the general govern-ment debt-to-GDP ratio
Revisions of budgetary data and their impact on the general government deficit and debt
In 2002 the bilateral discussions between Eurostat andthe Greek statistical authorities led to a significant revi-
yen1part According to the last notification (March 2003) the surplus of social secu-rity funds and local government amounted to 27 of GDP in 2001 and isestimated to have been equal to 35 of GDP in 2002
yen2part In the 2002 update of the stability programme government final consump-tion expenditure is estimated to increase by 25 in 2003 compared to67 in 2002 However according to revised data in 2002 the generalgovernment consumption expenditure appears to have been increased by115 in nominal terms
217
P u b l i c f i n a n c e s i n E M U 2 0 0 3
sion of the Greek budgetary data for 2000 and 2001 Infact the agreed changes resulted in an upward revisionof the government debt ratio while the anticipated gov-ernment surpluses for 2001 and 2002 turned into deficitsThe slow decline in the debt ratio from 1997 onwardshas been reversed and in fact the debt ratio increasedboth in 2000 and 2001
The reclassifications required by Eurostat were relatedto large debt management financial operations that havebeen reflected in the unusually high stock-flow adjust-ments as well as to government expenditures previouslyexcluded from the budget (1) Specifically the increasein general government deficit for the years 2000 and2001 resulted from treating debt assumptions as well ascapital injections to public enterprises as capital trans-fers Moreover before the adjustments the debt man-agement related financial operations were either notincluded or incorrectly treated in government books cre-ating a picture of government balances not correspond-ing to the actual evolution of the debt-to-GDP ratioThus the treatment of debt creating financial operations
according to generally accepted practices is expected toimprove the transparency and credibility of the Greekbudgetary accounts
The reclassified operations can be grouped into the fol-lowing categories
bull Capital injections The treatment of capital injections(that is of increases in the share capital of publicenterprises) in the ESA government accounts is notstraightforward These transactions are recorded asfinancial transactions without any direct impact onthe government deficit when the national account-ants consider that the government is acting as a share-holder which provides funds and expects to receiveproperty income in future However when the capitalinjections are no more than a grant that is paid tocover losses or to compensate the company for publicservices the transaction should be recorded as gov-ernment expenditure and included in the governmentdeficit In the case of Greece Eurostat considered thatthe distinction between these two cases was not strictenough and asked the Greek statistical service torecord capital injections to a group of enterprises inthe deficit
Table VI7
Composition and balances of general government Greece (1) (as of GDP)
2000 2001 2002 2003 2004
Government balance (2) ndash 19 ndash 14 ndash 12 ndash 11 ndash 10
mdash Total revenue 470 456 465 460 452
Of which mdash current taxes 108 96 94 93 90
mdash social contributions 140 139 140 139 139
mdash Total expenditure (2) 489 470 477 471 462
Of which mdash collective consumption 97 93 97 95 94
mdash social transfers (3) 226 222 225 224 222
mdash interest expenditure 70 63 55 52 49
mdash gross fixed capital formation 40 39 38 40 39
Primary balance (2) 51 49 43 41 39
Pm Tax burden 388 366 364 362 357
Government debt 1062 1070 1049 1010 970
Pm Cyclically-adjusted balance ndash 19 ndash 23 ndash 18 ndash 18 ndash 19
Pm Cyclically-adjusted primary balance 51 40 37 34 30
(1) Commission spring 2003 economic forecasts(2) Data include UMTS receipts amounting to 05 of GDP in 2001(3) In kind and other than in kind
Source Commission services
yen1part The large financial operations which affected the evolution of the debt ratiowere mirrored in the stock-flow adjustments which reached 71 of GDPin 2001 and stood at 37 of GDP in 2002
218
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
bull Debt assumptions According to the ESA95 rulesthe assumption of State guaranteed debt of public orprivate companies is recorded as capital transferthat is a government expenditure However Greeceused to register debt assumptions as financial trans-actions and not as government expenditure
bull Coinage In Greece as in most EU Member Statescoins are issued by the Treasury and are a compo-nent of government debt This implies that the pro-ceeds collected by government at the issuance ofcoins have a financial nature and therefore do nothave any direct favorable effect into the deficit (1)However in Greece the coinage proceeds used to berecorded as government revenue and improved thedeficit The correction of this accounting inconsist-ency had a negligible impact on the years before2002 since the regular issuance of coins is verysmall However for 2002 the impact was quiteconsiderable since it was the year that the wholestock of coins in circulation was replaced because ofthe euro cash changeover
bull Securitisation The Greek government has been rel-atively active in securitising expected future reve-nue The Greek statistical authorities used to recordthe securitisation proceeds outside the governmentdeficit and therefore the Eurostat decision of July2002 did not have any impact on the Greek govern-ment deficit However Eurostat also decided thatthe proceeds collected through securitisation shouldbe treated as loans and therefore included in the gov-ernment debt
bull Finally Eurostat considered that according to theESA95 rules share exchangeable and share convert-
ible bonds should be treated as the normal govern-ment bonds and should be included in the stock ofgovernment debt
The agreed total revisions of budgetary accountsamounted to 13 of GDP in 2001 and 19 of GDPin 2002 Correspondingly the debt ratio was revised by73 percentage points in 2001 and 80 percentage pointsin 2002
Specifically in 2001 more than half of the revision of thegeneral government deficit was accounted for by thecapital injections to public enterprises The rest of theadjustments resulted almost entirely from debt assump-tions whereas the impact on interest payments was lessthan 01 of GDP
Overall the estimated surplus of 01 of GDP reportedin the September 2002 notification was turned into a def-icit equal to 12 of GDP in the revised notificationsubmitted in November 2002 whereas in the March 2003notification the general government deficit for 2001 wasfurther raised to 14 of GDP due mainly to an increasein public investment and interest payments
Concerning government debt an amount equal to 54 of GDP related to proceeds from securitisation of futurerevenues and share convertible bonds was alreadyincluded in the debt in the September 2002 notificationHowever the proceeds from share exchangeable bondsadded an amount equal to 19 of GDP raising the gov-ernment debt to 107 of GDP from 1051 of GDPreported in the September 2002 notification
In 2002 the reclassification of capital injections to pub-lic enterprises raised government deficit by 07 ofGDP while debt assumptions added another 02 ofGDP to the deficit In the same year an amount equal to04 of GDP representing proceeds from coinage was
Table VI8
Key figures of the Greek stability programme (2001ndash06)
2001 2002 2003 2004 2005 2006
Real GDP growth (annual change) 41 38 38 40 37 36
General government budget balance ( of GDP) ndash 12 ndash 11 ndash 09 ndash 04 02 06
Primary surplus ( of GDP) 51 44 44 46 50 52
Government debt ( of GDP) 1070 1053 1002 961 921 879
Source 2002 update of the stability programme of Greece
yen1part In fact there is even a small detrimental impact into the government deficitas the cost of minting the coins is recorded as government expenditure
219
P u b l i c f i n a n c e s i n E M U 2 0 0 3
excluded from government revenues Moreover thegeneral government deficit for 2002 was furtherincreased by 06 of GDP due mostly to the budgetaryoutcome of the year reaching 19 of GDP Thus theestimated surplus of 08 of GDP reported in the 2001update of the stability programme turned into a deficit of11 of GDP The same deficit is reported in the 2002update while the March 2003 notification reports ahigher deficit equal to 12 of GDP due to higher publicinvestment spending
Government debt was similarly revised upwards by 8 of GDP in 2002 the main contributors to this increasebeing the same as in 2001 Thus the estimated govern-ment debt of 973 of GDP reported in the 2001 updateof the stability programme was revised up to 1053 of
GDP in the 2002 update However the March 2003 noti-fication reports a slightly lower debt ratio equal to1049 of GDP due to a higher than the previously esti-mated increase in GDP
The correction of budgetary data while improving thecredibility of the Greek public accounts points at thesame time to the imperative need for further and moredecisive consolidation of the Greek public finances Allthe adjustments in the budgetary data were linked to cur-rent expenditures indicating the persisting pressures forincreases in public spending even under the favourableconditions of strong growth Additional measures aretherefore required to limit current spending and bringdown the debt ratio at a faster pace in view of theexpected budgetary costs from an ageing population
220
5 Spain
Recent developments
Since the mid-1990s Spainrsquos fiscal strategy based onexpenditure restraint has been successful in reducing thegeneral government deficit from 66 of GDP in 1995to 01 in 2002 In the same period the debt-to-GDPratio declined from 639 in 1995 to 540 in 2002While fiscal consolidation was supported by the strongexpansion of the economy until mid-2001 the reductionof the deficit continued in 2001ndash02 The balanced budgetin 2002 was obtained thanks to revenue buoyancysupported by relatively strong domestic demand andemployment creation coupled with savings on publicconsumption and interest payments In particular thecorporate income tax posted a record increase (24 ona cash basis) due to a surge in declared capital gainsfollowing the introduction of a favourable regime fortheir reinvestment Additionally civil service pay wasincreased below the CPI inflation rate helping to moder-ate public consumption Finally the interest paymentsfall reflected both the decreasing debt burden and lowerinterest rates In cyclically-adjusted terms the fiscalstance in 2002 can be regarded as slightly restrictive
As a result the GDP share of current resources rose to395 in 2002 (390 in 2001) while total currentexpenditure remained broadly stable at 353 of GDP(351 in 2001) The slight increase in public saving asa percentage of GDP was accompanied by the marginalincrease in the share of gross capital formation reaching33 of GDP in 2002
In the baseline scenario of the 2002ndash06 updated stabilityprogramme a target of a general government balancedbudget is envisaged for 2003 and 2004 (maintaining thetarget for 2003 and revising only marginally the targetfor 2004 with respect to the previous update) Targets for2005 and 2006 are small surpluses of 01 and 02 ofGDP respectively The primary surplus is set to remainbroadly unchanged at close to 27 of GDP throughoutthe programme period
The debt-to-GDP ratio is expected to continue to declineby around 2 percentage points every year throughout theprogramme period falling below 47 of GDP by 2006These objectives assume the economy grows at a rateclose to potential (3 ) between 2003 and 2006 withhigher inflation compared with the previous update(average annual growth of the GDP deflator of 28 and25 respectively) The fiscal strategy continues to relyon restraint of primary current expenditure supported bylower interest payments while allowing for a strength-ening of public investment The programme incorporatesthe effects of the recently implemented reform of per-sonal income tax with an estimated direct cost of 04 of GDP in 2003 and 2004 (see next section for details)
Finally the programme incorporates the full effect ofthe new financial system for regional governmentswhich has involved further decentralisation of tax andspending powers The parallel implementation of theGeneral Law of Budgetary Stability which prescribesthat all public entities have to present their accounts inbalance or in surplus (1) is meant to provide furthersupport to the consolidation strategy in a context ofhigher fiscal decentralisation
For 2003 the target of a general government balancedbudget is being tested given the worsening in the macr-oeconomic scenario and the implementation of the per-sonal income tax reform By comparison the Commis-sion forecasts a deficit of 04 of GDP based on GDPgrowth of 20 reflecting also a less optimistic assess-ment of the effects of the tax reform on domesticdemand However for the rest of the projection periodfiscal objectives appear to be based on cautious growthassumptions and seem attainable (for comparison theCommission services forecast a GDP growth of 3 anda slight deficit of 01 in 2004 on an unchanged policy
yen1part Deficit budgets can be presented in certain circumstances if justified andreturning to balance or surplus within the three-year planning period (see thecase study on Spainrsquos General Law of Budgetary Stability in Chapter V5)
221
P u b l i c f i n a n c e s i n E M U 2 0 0 3
basis) All in all the underlying budgetary positionimproves by over a percentage point over the pro-gramme period to a surplus of 03 of GDP in 2006
The new income tax reform in Spain main features and implications
Main features of the new tax reform
The new personal income tax was approved at the end of2002 The 2003 income tax declaration on incomesearned in 2002 will still be based on the previous systemand the first tax declaration according to the new taxrules will be submitted in June 2004 However theeffects of this reform are already evident in 2003 as the
new tax rates are currently being applied to income sub-ject to withdrawal at the source
The reform introduces fewer and smaller changes thanthe previous reform adopted in 1998 of which the newreform is considered to be a continuation Accordinglythe measures aim at simplifying further the tax systemby lowering the number of tax brackets and increasingthe income threshold below which an individual is nolonger required to file a tax return The major changescan be summed up as follows
Minimum and maximum marginal rates are loweredfrom 18 and 48 to 15 and 45 respectively (20 and56 in the income tax prevailing up to 1998) In turn
Table VI9
Composition and balances of general government Spain(as of GDP)
2000 2001 2002 2003 2004
Government balance (1) ndash 09 ndash 01 ndash 01 ndash 04 ndash 01
mdash Total revenue 390 392 396 393 395
Of which mdash current taxes 222 218 226 223 224
mdash social contributions 133 136 135 135 135
mdash Total expenditure (2) 399 393 397 398 396
Of which mdash collective consumption 75 74 75 76 76
mdash social transfers (2) 224 223 226 228 227
mdash interest expenditure 33 31 29 27 25
mdash gross fixed capital formation 31 32 33 34 34
Primary balance (2) 25 30 28 22 24
Pm Tax burden 357 356 361 359 358
Government debt 605 569 540 525 505
Pm Cyclically-adjusted balance (1) ndash 16 ndash 08 ndash 04 ndash 04 ndash 01
Pm Cyclically-adjusted primary balance (1) 17 23 25 23 24
(1) Data exclude UMTS receipts amounting to 01 of GDP in 2000(2) In kind and other than in kind
Source Commission spring 2003 economic forecasts
Table VI10
Key figures of the Spanish stability programme (2001ndash06)
2001 2002 2003 2004 2005 2006
Real GDP growth (annual change) 27 22 30 30 30 30
General government budget balance ( of GDP) ndash 01 ndash 02 00 00 01 02
Primary surplus ( of GDP) 30 27 28 27 27 27
Government debt ( of GDP) 571 552 531 510 490 469
Source 2002 update of the stability programme of Spain
222
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
the number of tax brackets is reduced from six to five(see Table 1)
The minimum level of earned income which impliesthe obligation of filling a tax return is increased toEUR 22 000 when this income stems from a singlewage earner compared to EUR 21 035 in the previoussystem (1) The new tax keeps the concept of taxableincome introduced in the 1998 reform taxable incomeis obtained after deducting a tax-free allowance theso-called lsquoexempted minimum living standardrsquo whichreplaced a vast set of tax reliefs As in the previousscheme this tax-exempt minimum living standardvaries depending on personal and family circum-stances However these thresholds have now beenraised For instance the basic individual exemptedincome increases by 28 to EUR 3 400
The reform introduces a more favourable fiscal treat-ment for some disadvantaged groups to promote theirinsertion in the labour market and raise employmentSpecifically tax reliefs for working women with chil-dren (EUR 1 200 for every child aged below three)higher tax reliefs for unemployed workers accepting ajob implying geographical mobility tax advantages fordisabled workers and workers aged above 65
With a view to remedying low geographic mobility ofworkers partially attributable to the poor functioning ofthe housing market the new income tax establishesfiscal advantages for letting of accommodation Thisincludes a relief of 25 of net rental income an addi-tional relief of 25 for five years for housing at presentunoccupied and let before the end of 2004
A more favourable fiscal treatment of income from cap-ital in the reform is targeted at promoting saving In par-ticular realised capital gains (net worth gains) on assetsheld for more than one year are now taxed at 15 (18 previously) while tax relief for capital incomes fromassets held for more than two years is raised to 40 from30 In addition capital gains obtained from invest-ment funds will not be taxable as long as they are rein-vested in other funds
Private pension plans are promoted by raising the thresh-olds to be deducted in the tax base In addition pension
insurance contracts are to have a similar fiscal treatmentas pension funds when the main purpose of the insurancepolicy is long-term saving for retirement
Finally the withholding tax rate applied to interest fromdeposits capital gains stemming from share sales andother non-labour incomes is lowered from 18 to 15 inline with the minimum marginal income tax
A tentative assessment of the new income tax
The new income tax regime does not change dramati-cally the existing structure put in place by the 1998 taxreform However the reform is likely to reinforce thereduction in the share of taxation directly paid by house-holds observed since the mid-1990s According to offi-cial estimates the current reform will involve a reduc-tion of the effective average income tax rate of 11 while by level of income the impact is officially esti-mated as being progressive However estimates carriedout by independent researchers highlight an unevenimpact of the reform with a significant reduction in theeffective average tax rate occurs only for the lowest andthe highest income brackets (2) marginal tax rates formiddle-range incomes are broadly unchanged
The reform can be expected to boost the labour supplygiven the reduction in the marginal tax rate at the bottomof the income scale The tax relief introduced for work-ing women with children aims at increasing the femaleemployment rate which at present is very low It mightalso have some positive impact on the fertility rate whichis one of the lowest among developed countries The fis-cal rebates on labour incomes for workers aged above65 could also have positive effects on the labour partici-pation of older workers
The more favourable taxation of unemployed workersaccepting a job offer implying a move to another placeof residence is a positive measure given the low labourmobility in Spain and the wide disparities in unemploy-ment between regions Nevertheless mobility is unlikelyto increase significantly in the absence of a reform of thecurrent benefit system and a better functioning of thehousing market In this respect the new fiscal treatmentof letting of accommodation might imply an increase inthe current rather limited supply of this kind in SpainHowever some doubts arise on the final effect of thismeasure given the absence of reduction in the current
yen1part According to the Spanish Statistical Institute the average gross wage(including non-labour costs) in annual terms is estimated at approximatelyEUR 23 450 yen2part See for example Pampillon and Raymond (2002)
223
P u b l i c f i n a n c e s i n E M U 2 0 0 3
fiscal relief for housing purchase which is more favour-able than that envisaged for letting
In general the new income tax contains positive meas-ures to encourage labour participation and improve thefunctioning of labour market However a fuller assess-ment will only be possible after a certain period of time
As for the fiscal treatment of saving a positive aspect ofthe new income tax refers to long-term savings related toretirement The rise in the annual thresholds to qualifyfor tax relief for private pension contributions is a posi-tive measure in the light of the future financial pressureson the public pension system stemming from ageingLikewise the tax exemption for capital gains realised onan investment fund and reinvested in another fundshould promote saving Additionally this measure isexpected to increase competition not only among differ-ent types of funds but also among financial institutions
Regarding the macroeconomic and budgetary impactaccording to government estimates the income taxreform will add 05 percentage points to GDP growth(although whether this impact takes place over one ortwo years is not specified) The estimated loss of reve-nues is set at 03 of GDP in 2003 including the sec-ond-round effects caused by the reform This estimatemight prove optimistic given the average tax reductionof 11 the cost of the reform excluding second-roundeffects could be roughly estimated at around 06 ofGDP As to the effects on growth both in terms ofimmediate impact and second-round effects an econo-metric simulation performed with the CommissionQUEST model suggests that the impact of the reformon GDP growth is an increase by 01 in 2003 In themedium term this impact is estimated as adding around02 percentage points per year to GDP between 2004and 2006
Table VI11
Main features of recent tax reforms
1998 income tax 2002 income tax
Net tax base up to EUR
Taxable income
Rest up to EUR
Marginal tax rate
Net tax base up to EUR
Taxable income
Rest up to EUR
Marginal tax rate
000 000 3 67819 180 000 0 4 000 150
3 67819 66207 9 19548 240 4 00000 600 9 800 240
12 87370 2 86899 12 26064 283 13 80000 2 952 12 000 280
25 13432 6 33875 15 32580 372 25 80000 6 312 19 200 370
40 46030 12 03995 26 97342 450 45 00000 13 416 mdash 450
67 43355 24 17799 mdash 480 mdash mdash mdash mdash
224
6 France
Recent developments
In 2002 the situation of French public finances deterio-rated markedly The general government deficit reached31 of GDP thus breaching the 3 of GDP referencevalue of the Treaty This outcome is to be compared witha general government deficit of 14 of GDP plannedby the French authorities in the Finance Law for 2002For the first time since 1998 the general governmentdebt ratio increased in 2002 reaching 591 of GDP upfrom 568 of GDP in 2001 The larger part of the slip-page in the 2002 general government deficit is due to adeterioration in the cyclically-adjusted budget positionaccording to Commission calculations the cyclically-adjusted balance worsened by 11 percentage points ofGDP in 2002 to reach 33 of GDP This results from(1) an overrun in expenditures (2) the implementation oftax cuts worth percentage point of GDP and (3) a rel-atively low tax-to-GDP elasticity The remaining reflectsa base effect due to the incorporation in the deficit of acapital injection in the firm RFF (Reacuteseau Ferreacute deFrance) worth 01 percentage point of GDP and theimpact of adverse cyclical developments Indeed realGDP grew by 12 last year as against 25 projectedin the budget for 2002 According to Commission calcu-lations the cyclical component of the deficit worsenedby 05 percentage points of GDP in 2002
The developments in public finances in 2002 provideevidence of the existence of an excessive deficit positionin France The Commission has therefore decided inMarch 2003 to initiate an excessive deficit procedure(EDP) in the case of France This procedure follows theearly warning issued by the Council in January 2003and recommending that France lsquoensures that the 3 ofGDP reference value for the general government deficitwill not be breached in 2003rsquo A recommendation to theCouncil to recommend to France measures to correct theexcessive deficit has been adopted by the Commissionon 7 May 2003 and is expected to be endorsed by theEcofin Council on 3 June
However under current policies French public financesare expected to deteriorate further in 2003 the budget for2003 does not contain measures reducing sizeably theunderlying deficit and unfavourable cyclical develop-ments will continue weighing on fiscal revenues The gen-eral government deficit is projected by the French author-ities to reach 34 of GDP and by the Commission toreach 37 of GDP In the same year the gross govern-ment debt is projected by the Commission and the Frenchauthorities to breach the 60 Treaty reference value
These projections are subject to downside risks In par-ticular after a year of freeze in real wages wage claimsby unions in the central government sector could triggerstronger than projected current expenditures This effectcould overlap with the financing of the priorities of thenew government in the field of national security On theother side the French authorities decided to cancelexpenditures worth 01 of GDP and to freeze creditsfor an amount of 015 of GDP However as long asthey are not cancelled the credits frozen can still bespent in the current budgetary exercise
In 2004 despite the acceleration in economic activity toa close to potential rate the general government deficitis projected by the Commission under a no policychange assumption to decrease only marginally to 35 of GDP The decline in the deficit in 2004 will be bur-dened by the lagged impact of the low 2003 GDP growthon fiscal revenues The French authorities expect astronger decline in the government deficit in 2004 to29 This forecast is based on the assumption thatmeasures ensuring an improvement in the cyclically-adjusted balance by 05 percentage point of GDP in 2004will be implemented in the budget for that year
The urgency of reforming the French pension system
As in many other European countries public finances inFrance will face problems arising from ageing popula-
225
P u b l i c f i n a n c e s i n E M U 2 0 0 3
tion over the next decades Existing projections showthat the old-age dependency ratio will increase rapidlyto almost double between 2000 and 2050 In the absenceof further reforms the share of pension expenditures toGDP will increase by around 4 percentage points of GDPover the next 40 years The pension system todaybroadly balanced could then post a deficit as high as38 of GDP in 2040
While the magnitude of ageing comparable in France tothat expected in the other EU countries the timing willbe somewhat different as the larger part of the demo-graphic change will occur earlier in France than in most
other EU countries Pension expenditures will startincreasing faster than GDP as from 2005 and the growthdifferential will widen markedly as from 2010 Pensionexpenditures are projected to increase by 3 percentagepoints of GDP in France between 2000 and 2020 and byonly 1 percentage point of GDP on average in the EU
Therefore reforming the pension system in view toensure its financial sustainability is urgent The Frenchauthorities committed themselves to implement a com-prehensive reform of the pension system before thesummer 2003 This commitment follows a long period ofbroad consultations with the social partners Since the
Table VI12
Composition and balances of general government France (1) (as of GDP)
2000 2001 2002 2003 2004
Government balance (2) ndash 14 ndash 15 ndash 31 ndash 37 ndash 35
mdash Total revenue 512 510 505 503 503
Of which mdash current taxes (3) 277 275 255 255 255
mdash social contributions 163 163 163 163 163
mdash Total expenditure (4) 526 525 537 541 538
Of which mdash collective consumption 93 92 93 93 92
mdash social transfers (4) 317 317 328 332 332
mdash interest expenditure 31 31 32 32 33
mdash gross fixed capital formation 32 31 31 31 30
Primary balance (4) 17 16 00 ndash 05 ndash 02
Pm Tax burden 450 447 440 438 439
Government debt 572 568 591 618 631
Pm Cyclically-adjusted balance ndash 23 ndash 22 ndash 33 ndash 35 ndash 33
Pm Cyclically-adjusted primary balance 09 09 ndash 03 ndash 03 01
(1) Commission spring 2003 economic forecasts(2) Data include UMTS receipts amounting to 01 of GDP in 2001(3) Taxes on production and imports and current taxes on income and wealth(4) In kind and other than in kind
Source Commission services
Table VI13
Key figures of the French stability programme (2002ndash06)
2001 2002 2003 2004 2005 2006
Real GDP growth (annual change) 18 12 25 25 25 25
General government budget balance ( of GDP) ndash 14 ndash 28 ndash 26 ndash 21 ndash 16 ndash 10
Primary balance ( of GDP) 17 04 06 10 15 20
Government debt ( of GDP) 568 587 591 589 583 570
Source 2002 update of the stability programme of France The figures presented from 2004 are those of the lsquocautiousrsquo scenario of the stability programme which projects real GDP growth at 25 as from 2004 and was considered by the Commission as the most plausible one
226
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
implementation of the reforms of 1993 and 1995 whichincreased the number of contribution years giving rightto a full pension the main new measure has been thecreation of a Pension Reserve Fund in 1999 This Fundis projected to have accumulated assets for a totalamount of 10 of GDP in 2020 and is only intended tosmooth the financial shock during the retirement of theearly baby boom cohorts Indeed the assets of the fundwill be largely insufficient to cover increased spendingon pensions
Limited information has been released up to now on themain lines along which the reform will be based (seeBox VII) However it appears likely that a driving prin-ciple is to safeguard the compulsory schemes financedon a pay-as-you-go basis which the national strategyreport regards as an essential condition for inter- andintra-generational solidarity The reinforcement of thethird pillar based on encouraging private saving seemsalso to be among the intentions of the current govern-ment as well as the willingness of putting on a moreequal footing conditions for retirement in the public andin the private sectors
In this context any possible reform of the pension systemwill necessarily be organised around the modification ofthe three essential parameters of the system (1) thereplacement ratio (2) the rate of social contributionsand (3) the retirement age The reform will probably relyon a change of all these parameters Indeed according to
the estimates presented by the French authorities in thenational strategy report on pensions achieving the sus-tainability of the pension system changing only one ofthese parameters would require too large an effort Cur-rent estimates show that solving the problem via a reduc-tion in the replacement ratio alone requires a reduction ofthis parameter by more than 30 percentage points In thesame vein changing only the average retirement agewould necessitate an increase by at least six years of thecontribution period and relying only on increasingsocial contributions would imply an increase in thelabour cost by more than 6
If this latter option was to be followed the resultingincrease in the labour cost would be huge and couldhave strong negative effects on labour demand It isindeed generally accepted that a rise in the labour costby 1 percentage point reduces the demand for labour by03ndash05 percentage points This option would not befully compatible with the Lisbon strategy neither sat-isfy the objective of fairness across generations
Finally in France still exists a large scope for increasingthe effective retirement age by encouraging labour sup-ply of older people as the employment rate of olderworkers is among the lowest in the EU This could beachieved notably by strengthening the link betweencontribution and benefits in order to increase the incen-tives to remain at work when entering in more maturecohorts
Box VI1 Main features of the French pension system
The French pension system is based on compulsory pay-as-you-go schemes which cover 98 of total pension expenditureand are financed by social security contributions and taxes The larger scheme is the general scheme which covers themajority of workers of the private sector Civil servants and State enterprise employees are covered by a variety of specialschemes which are generally more generous than those for private sector employees The basic schemes contain solidarityelements and give pension on the basis of the number of contribution years Alongside the basic schemes compulsorycomplementary schemes exist which cover workers by category (AGIRC for managers and for ARRCO other employers)The benefit formula of these supplementary schemes is based on a point system and ensures a close link between contri-butions and benefits
On its current form the system is relatively generous compared to other countries Indeed the legal retirement age set at60 and the effective average retirement age currently at 587 years according to Eurostat are among the lowest in the EUThe replacement ratio is relatively high in France compared to other EU countries and retired households are at no higherrisk of poverty than other households as a minimum level is guaranteed The current system provides weak incentives tocontribute beyond the age at which a full pension entitlement is acquired
227
7 Ireland
Recent developments
After five years of surpluses with a peak of 43 ofGDP in 2000 the out-turn for the general governmentbalance in 2002 is estimated to have been a negligibledeficit of 01 of GDP (or 02 of GDP excludingUMTS receipts) around percentage point below tar-get (1) As in 2001 the deviation from target is due to alarge tax undershoot and some expenditure overruns (ona general government basis) While cyclical develop-ments are undoubtedly in part to blame for the large rev-enue shortfalls of the past two years difficulties seem tohave arisen in accurately costing tax packages and fore-casting revenues Regarding expenditure in 2002 sav-ings on interest payments and public investment partlycompensated for overruns on current primary spendingGiven high nominal growth the debt-to-GDP ratiowhich has been the second-lowest in the EU since 2000fell by three percentage points to one third of GDP at theend of 2002
The cyclically-adjusted balance is estimated to havedeteriorated by almost 1 percentage point of GDP in2002 Although calculations of the output gap are subjectto a particularly large margin of error in Ireland thispoints to a discretionary easing of budgetary policy in2002 rather than broadly neutral as planned in the previ-ous update of the stability programme 2002ndash04 and asrecommended by the broad economic policy guidelinesfor 2002 By the same measure fiscal policy was loos-ened by more than 3 percentage points of GDP over theperiod 2000ndash02
By contrast the budget plans for 2003 unveiled on4 December 2002 together with the new stability pro-gramme 2003ndash05 implement a tightening of fiscal
policy by some of GDP The budget for 2003 (2)increases the tax take by 09 of GDP compared to ano-policy change scenario two thirds of which comesfrom increases in indirect taxes and stamp duties On theexpenditure side a 1 cut in nominal capital spending(which translates into a very significant real cut in view ofhigh construction inflation) makes room for increased cur-rent spending Even so the budget plans a marked reduc-tion in the growth rate of current discretionary expendi-ture (3) to 8 for 2003 from 15 in 2002 and 22 in2001 The main measures regarding current spending area relatively modest social welfare package and a rise in thepublic sector pay and pensions bill by 11 (4)
The original budget-day target for the general govern-ment balance in 2003 a deficit of 07 of GDP wasrevised to 08 in March (5) The Commission spring2003 economic forecasts project a slightly better out-come with a deficit of 06 of GDP for 2003 corre-sponding to a restrictive fiscal stance In 2004 the deficitis expected to widen further to 09 of GDP (on a no-policy change basis)
According to the updated stability programme the gen-eral government deficit is projected to reach 12 ofGDP in 2004 and to remain at that level in 2005 Thesetargets incorporate technical provisions for unspecifiedfuture budget measures with a full-year cost of 07 ofGDP in each year which is subject to review lsquoin light ofemerging economic conditionsrsquo They also include
yen1part For this assessment the original budget-day target (+ 07 of GDP)has been adjusted to (i) include UMTS receipts of 02 of GDP and(ii) exclude a transfer from the Central Bank of 05 of GDP which had tobe reclassified below the line
yen2part The rest of this parafigure is based on the budget plans in terms of theExchequer cash accounts
yen3part This refers to the concept of lsquovotedrsquo current spending for which annualapproval by Parliament is needed and which excludes inter alia the serv-ice of the national debt and the contribution to the EU budget
yen4part This includes a provision of 04 of GDP for payment of the first quarterof the lsquobenchmarkingrsquo awards (backdated to December 2001) The bench-marking process was initiated in mid-2000 to adjust pay rates in the publicsector by reference to rates in the private sector for comparable jobs Thebenchmarking bodyrsquos report of mid-2002 recommended pay increases dif-ferentiated by grade leading to an 89 rise in public sector pay costswith an estimated full-year cost of 08 of GDP
yen5part From the March 2003 reporting of government deficits and debt levels inaccordance with Council Regulation (EC) No 360593 as amended byCouncil Regulation (EC) No 4752000
228
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
contingency provisions (against unforeseen develop-ments) of 04 of GDP in 2004 and 08 in 2005 Incyclically-adjusted terms the targets for 2004ndash05 implya broadly neutral stance but excluding the contingencyprovisions would have a tightening bias The debt ratiois projected to rise by less than 1 percentage point overthe programme period to just below 35 by 2005 With-out the build-up of non-general government assets in theNational Pensions Reserve Fund (NPRF) (1) howeverthe (gross) debt ratio would continue to fall to 2005
Recent initiatives on expenditure management
Driven by the need to improve infrastructure and publicservices discretionary spending almost doubled between1997 and 2002 (2) The occurrence of overruns andconcerns about securing lsquovalue for moneyrsquo raise theissue of whether control and management systems areadequate This section reviews progress on implement-ing the recommendations on expenditure management in
Table VI14
Composition and balances of general government Ireland (as of GDP)
2000 2001 2002 2003 2004
Government balance (1) 43 11 ndash 02 ndash 06 ndash 09
mdash Total revenue 364 352 337 335 328
Of which mdash current taxes 268 252 239 239 234
mdash social contributions 56 58 57 56 56
mdash Total expenditure (1) 320 341 339 341 337
Of which mdash collective consumption 52 55 57 58 58
mdash social transfers (2) 167 178 182 187 184
mdash interest expenditure 21 16 14 15 15
mdash gross fixed capital formation 37 46 44 39 38
Primary balance (1) 65 27 11 09 06
Pm Tax burden 321 307 293 292 287
Government debt 393 368 333 334 333
Pm Cyclically-adjusted balance (1) 26 00 ndash 09 ndash 03 01
Pm Cyclically-adjusted primary balance (1) 46 15 04 12 16
(1) Data exclude UMTS receipts amounting to 02 of GDP in 2002(2) In kind and other than in kind
Source Commission spring 2003 economic forecasts
Table VI15
Key figures of the Irish stability programme (2003ndash05)
2001 2002 2003 2004 2005
Real GDP growth (annual change) 57 45 35 41 50
General government budget balance ( of GDP) (1) 16 ndash 05 ndash 07 ndash 12 ndash 12
Primary surplus ( of GDP) (1) 31 10 09 03 04
Government debt ( of GDP) 367 341 340 345 349
(1) Data exclude UMTS receipts amounting to 02 of GDP in 2002 and include contingency provisions (against unforeseen developments) of 04 of GDP in 2004and 08 of GDP in 2005
Source 2002 update of the stability programme of Ireland
yen1part The NPRF receives 1 of GNP annually from general governmentresources At the end of 2002 it was worth over 7 of GNP yen2part Total lsquovotedrsquo spending (see above) rose by 924
229
P u b l i c f i n a n c e s i n E M U 2 0 0 3
the 1996 report lsquoDelivering better governmentrsquo (DBG)thereby putting the range of measures taken or announcedsince 2002 in context (1)
In 1994 the strategic management initiative (SMI) waslaunched to enhance the quality of public services As afollow-up the DBG report proposed to change manage-ment structures human resource management and finan-cial management systems (2) Concerning the latter itmade the following recommendations
Firstly DBG favoured moving to multiannual budgetingThe budget for 1997 marked a first step in this directionproducing aggregate budgetary projections on a no-pol-icy change basis for 1998 and 1999 Subsequent budgetshave provided more detailed projections over a three-yearhorizon thus covering the same period as the stabilityprogrammes that have accompanied each budget In 2000it was decided to defer the original plan to introducethree-year financial envelopes for each departmentbecause existing multi-year programmes (such as thethree-year social partnership agreements (3) and theseven-year investment programme) were considered toprovide sufficient medium-term guidance In November2002 however the Minister for Finance announced thathe was considering extending working with five-yearfinancial envelopes (as is currently done for public trans-port) to other large capital spending areas (4)
Secondly DBG sought to delegate financial authority tothe maximum extent possible This would be accompa-nied by the requirements that each department report atyear-end on outcome versus plan mdash not only financiallybut also in output terms mdash and that regular expenditurereviews be carried out (see next point) (5)
Thirdly the government launched the expenditurereview initiative (ERI) in May 1997 which was origi-nally intended to review all expenditure from a results
perspective over a three-year period This turned out tobe over-ambitious and by end-2000 only 62 reviewshad been completed with 21 more underway togetherrepresenting at most 37 of government spendingDrawing on an assessment of the ERI by the Comptrollerand Auditor General (6) the government decided in mid-2001 to adopt somewhat revised arrangements (i) theselection of topics for review should ensure that the ERIfocuses on significant areas of expenditure and criticalareas of government policy (ii) the Department ofFinance is to enhance its provision of central supportand (iii) departments are encouraged to publish thereviews they carry out The topics for review in thesecond round (2002ndash04) were approved in May 2002and include two pilot cross-departmental reviews
Related to this the government set up an IndependentEstimates Review Committee (IERC) in the course of2002 to help prepare the spending plans for 2003 (7) Theobjective of this group of three former civil servants wasto find EUR 900 million (07 of GDP) worth of sav-ings on the estimated cost of maintaining the lsquoexistinglevel of servicersquo in 2003 (8) This was to be achieved byidentifying programmes (i) which were no longer justi-fied because of changed circumstances or (ii) whichcould be deferred or spread over a longer period and bysuggesting new delivery or user-charging mechanismsThe ensuing budget was based on a reduction of theexisting level of service cost in 2003 by 06 of GDPA similar review is to be carried out in 2003 regardingthe spending targets for 2004ndash05
Fourthly DBG recommended to strengthen the admin-istrative budget system introduced in 1991 Adminis-trative budgets cover the administrative (or running)costs of departments of which pay constitutes by farthe largest part Each department has to agree with theDepartment of Finance on a level of administrativespending for three years and can manage these fundsrather flexibly (rights to switch to carry-over and toappoint staff all subject to certain limits) The systemhas been enhanced and extended to other departmentsand offices The IERCrsquos report recommended to intro-duce an efficiency dividend of 2
yen1part See also National Economic and Social Council (2002)yen2part See Coordinating Group of Secretaries (1996) A working group set up to
oversee implementation of DBGrsquos financial management recommenda-tions produced detailed proposals in July 1999
yen3part These agreements not only contain a wage clause but also cover tax andspending measures
yen4part In addition revised arrangements for capital spending are being put inplace providing detailed information of the size and time-scale of spendingcommitments
yen5part In this context DBG recommended that departments be rewarded for sug-gesting savings In November 2002 the Minister for Finance announcedthat savings would be earmarked to the same department for use in high-priority areas
yen6part See Comptroller and Auditor General (2001)yen7part The IERCrsquos report to the Minister for Finance is available at www
budgetgovieyen8part This concept has recently replaced the no-policy change basis and allows a
clearer distinction between technical and policy adjustments
230
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
Fifthly as recommended by DBG a new managementinformation system is to be implemented by the end of2005 In addition to cash data it will provide data on anaccruals basis In addition to financial data it will providenon-financial information By focusing on outputs it willenable measurement of performance thus providing valu-able information for value for money studies and expend-iture reviews The development of performance measure-ment systems has received much attention under the socialpartnership process because recent national agreementshave included a conditionality clause for public sector payrises The final pay rise under the Programme for prosper-ity and fairness (expiring mid-2003) was conditional ondeveloping and agreeing performance indicators and mak-ing sufficient progress towards modernisation targets Allpay rises in the proposed new agreement Sustainingprogress are similarly dependent on verifiable progresson modernisation and flexibility (1)
Apart from reforms prompted by DBG several measureshave recently been taken to address the need to remainwithin spending allocations which has become morepressing in view of the tighter budgetary situation InNovember 2002 the Minister for Finance announcedthat the assessment of expenditure overrun risks wouldbe improved and that contingencies to cater for unfore-seen pressures would be introduced Special attentionwould be given to demand-led schemes In January2003 a report on accountability was endorsed contain-ing recommendations to strengthen internal financialcontrol and audit systems and to adopt formal risk man-agement strategies
An area of particular concern regarding budgetarycontrol and value for money is the health sector (2) In2002 an Independent Commission on Financial Man-
agement and Control Systems was set up to make recom-mendations on how to enhance the timeliness and qualityof the available financial information in the health serv-ice While not yet published its report is expected tohighlight serious deficiencies in existing financial man-agement procedures and to recommend the establish-ment of a new national health services executive as wellas the urgent implementation of new budgetary andaccountability systems and performance managementprogrammes (3)
Finally two recent measures are worth mentioning eventhough they are not directly concerned with improvingthe management of expenditure In November 2002 thegovernment decided to publish intra-year spending pro-files by department which can be compared with theevolution of expenditure as published in the monthlycash Exchequer accounts The profiles for 2003 (also fortax revenues) were published at the end of January In afurther move to enhance the transparency of spendingdevelopments the Department of Transport decided toissue monthly progress reports on major transport andinfrastructure projects The first such fiche was pub-lished in early March for the Luas (Dublin light-rail)project it provides the original and updated estimates of(i) total cost and (ii) completion dates
In conclusion the 1996 DBG report constituted an ambi-tious programme to improve public expenditure man-agement While progress has been made on severalfronts implementation of the recommendations remainsincomplete The measures announced during 2002 willfurther improve expenditure management but someissues require ongoing attention such as the roll-out ofthe management information system mdash and its contribu-tion to securing value for money mdash as well as themedium-term planning of spending
yen1part The new agreement not only awards general pay rises (a cumulative 7 )but also allows for a gradual and full implementation of the benchmarkingawards (89 on average mdash see footnote above) The first tranche ofbenchmarking (25 ) is unconditional and is to be paid in 2003 backdatedto December 2001
yen2part The involvement of the private sector in the provision of healthcare as in thedelivery of public investment falls outside the scope of this contribution yen3part Irish Times 30 and 31 January 2003
231
8 Italy
Recent developments
The general government accounts recorded a deficit of23 of GDP in 2002 against an initial target of 05 of GDP The considerable shortfall with respect to plansis in part explained by economic growth assumptionswhich from the outset did not sufficiently reflect theobserved deterioration in the global economic out-look (1) It is also due to the emergence of stronger thaninitially assessed expenditure trends also as a result ofthe marked revision of the 2001 budgetary out-turn (2)Almost half of this revision (around 05 percentagepoints of GDP) was due to an underestimation ofexpenditure of the central and the decentralised adminis-trations In the course of 2002 the government tooksteps to address the serious deficiencies that hadappeared in the budgetary process by passing a lawtightening expenditure controls (reviewed in the nextsection) and adopting provisions aimed at improving theinformation base on disbursements of the central govern-ment and the decentralised administrations
The 2002 budgetary out-turn benefited from a lower debtservicing costs than projected for that year in theNovember 2001 stability programme update (3) Legisla-tion enforced in the second half of the year blockingadditional tax credits for employment creation and
investment increasing tax receipts curbing healthcareexpenditure and as recalled above improving expendi-ture controls played a role in keeping the government def-icit in check In addition measures of a temporary naturecontributed over 1 percentage point of GDP to the 2002result sales of public real assets largely through securiti-sation amounted to 09 of GDP and receipts from a taxamnesty on assets held abroad were 01 of GDP
According to Commission calculations after the deteri-oration recorded in 2001 the cyclically-adjusted budgetbalance improved markedly in 2002 although the cycli-cally-adjusted primary balance posted a much smallerrecovery If the impact of sales of real estate is netted outin both years however the cyclically-adjusted overallbudget balance shows only very slender improvementbetween 2001 and 2002 while the cyclically-adjustedprimary balance deteriorates by 04 percentage points
The general government debt ratio decreased by almost3 percentage points of GDP to 1067 in 2002 well belowthe governmentrsquos own revised estimate of 1094 mainly thanks to operations carried out in the final monthsof the year most notably a debt conversion whichreduced the face value of government debt by 19 ofGDP (4)
The update of the stability programme covering theperiod 2002ndash06 targets a reduction in the actual deficitratio to 15 of GDP in 2003 with the budget approach-ing a balance in 2005 the year in which the debt ratio isto fall below 100 of GDP The budgetary target for2003 relies heavily as in 2002 on one-off measuresincluding the sale of publicly-owned real estate assetsthrough securitisation operations an accelerated tax liti-
yen1part Moreover since 2001 official projections of tax receipts have tended to beconsistently higher than would be warranted by already overoptimisticgrowth forecasts
yen2part In March 2002 the Italian statistical office reported a deficit of 14 ofGDP for 2001 The 2001 deficit estimate was then revised upwards threetimes to 16 in June 2002 mainly because of a higher estimate forhealthcare expenditure to 22 in July 2002 as a result of Eurostatrsquos deci-sion on the treatment of securitisation in national accounts and to 26 inFebruary 2003 following a further correction of healthcare expenditureand State sector expenditure (chiefly purchases of goods and services) andrevenue (tax receipts) After the latest revision the discrepancy betweendeficit reported in the government accounts in cash terms and theMaastricht deficit (in accrual terms) is significantly reduced in 2001
yen3part The sensitivity of Italian government debt to changes in interest rates is stillrelatively high given that over a third of it consists of short-term or variablerate instruments In 2002 interest expenditure was also reduced byEUR 19 billion through swap operations
yen4part Other factors influencing the result were financial operations whichreduced the cash deficit including the disposal of loans and other financialassets by the Cassa Depositi e Prestiti (the public savings and loans bank)in December 2002 the effect of the re-evaluation of the euro on debtdenominated in foreign currencies and a reduction compared to end-2001in the assets held by the Treasury with the Bank of Italy at end-2002
232
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
gation settlement scheme and new tax amnesties The2004 budgetary objective depends on replacing the mainone-off measures implemented in 2003 and on an addi-tional significant budgetary correction
The Commissionrsquos spring forecasts show a deficit innominal terms of 23 of GDP in 2003 on the back ofone-off measures amounting to 12 of GDP In 2004the lsquono policy changersquo projection of 31 principally aresult of the expiry of one-off measures implies that avery substantial fiscal correction would have to be car-ried out that year The debt ratio also remains distantfrom the targeted values although it decreases over theforecast period The difference between the Commis-sionrsquos forecasts and the targets in the stability pro-
gramme are in part due to a markedly lower assumptionfor real GDP growth in 2003 in part to a more cautiousevaluation of the fiscal policy measures (although theevaluation of the temporary measures mdash programmedsales of public real assets and tax amnesties mdash is alignedwith the official estimates) Interest rates at historicallylow levels are expected to continue to exert a dampeningeffect on the interest burden The cyclically-adjusteddeficit would improve in 2003 by 03 percentage pointsof GDP over the previous year while the cyclically-adjusted primary balance would remain largely stable
On 18 April 2003 the Ministry for the Economy andFinance released a new projection for the general gov-ernment deficit in 2003 which now stands at 23 of
Table VI16
Composition and balances of general government Italy (as of GDP)
2000 2001 2002 2003 2004
Government balance (1) ndash 18 ndash 26 ndash 23 ndash 23 ndash 31
mdash Total revenue 462 458 452 451 443
Of which mdash current taxes 298 296 288 282 280
mdash social contributions 127 126 127 128 128
mdash Total expenditure (1) 480 485 475 474 475
Of which mdash collective consumption 71 71 70 72 72
mdash social transfers (2) 279 283 289 289 288
mdash interest expenditure 65 64 57 53 51
mdash gross fixed capital formation 24 25 18 21 26
Primary balance (1) 46 38 34 30 20
Pm Tax burden 427 425 418 417 411
Government debt 1106 1095 1067 1060 1047
Pm Cyclically-adjusted balance (1) ndash 24 ndash 31 ndash 21 ndash 18 ndash 27
Pm Cyclically-adjusted primary balance (1) 41 33 36 35 24
(1) Data exclude UMTS receipts amounting to 12 of GDP in 2000(2) In kind and other than in kind
Source Commission spring 2003 economic forecasts
Table VI17
Key figures of the Italian stability programme (2002ndash06)
2001 2002 2003 2004 2005 2006
Real GDP growth (annual change) 18 06 23 29 30 30
General government budget balance ( of GDP) ndash 22 ndash 21 ndash 15 ndash 06 ndash 02 01
Primary surplus ( of GDP) 44 38 45 50 53 55
Government debt ( of GDP) 1099 1094 1050 1004 984 964
Source 2002 update of the stability programme of Italy
233
P u b l i c f i n a n c e s i n E M U 2 0 0 3
GDP based on an economic growth forecast of 11 (asagainst 23 in the November 2002 stability pro-gramme update) This is likely to have negative reper-cussions on the planned medium-term adjustment pathmaking the policy dilemma concerning the necessaryadjustment in 2004 more acute
Monitoring and controlling public expenditure the lsquoexpenditure freezersquo law
Amongst the provisions adopted in 2002 to correct fiscalimbalances Law No 2462002 (the so-called leggeblocca spese lsquoexpenditure freezersquo law) represents in theintentions of the Italian authorities an important instru-ment to improve the control of expenditure and to allowtimely correction of any deviation from the publicfinance objectives established in governmentrsquos medium-term economic and financial plan (DPEF) The law is anexample of how the need to fulfil obligations under theStability and Growth Pact provides an incentive tochange national rules and procedures emphasisingbudgetary discipline and fiscal sustainability
The overriding principle of ensuring the consistency offiscal trends with budgetary objectives is contained in arti-cle 81 of the Italian Constitution which states that lsquoWiththe adoption of the budget no new taxes and no newexpenditures can be establishedrsquo (indent 3) and lsquoAny otherlaw [beyond the budget] that establishes new expendituresmust indicate the means through which they will befinancedrsquo (indent 4) The implementation of the constitu-tional principle has proved problematic In 1978 a seriesof instruments (including the Financial Law) were intro-duced in order to rationalise the budgetary process andcontrol public finances (1) Since the late 1980s legisla-tion introducing newhigher current or capital expendi-tures or lower revenues above and beyond the financingmeans provided by the Financial Law in the special capitaland current account funds can be financed only within thelimits of newhigher receipts andor reductions of othercurrent expenditure with an additional constraint that cap-ital revenues cannot be used to finance current expendi-ture (2) All legislation introducing higher expenditure orlowering revenues and any amendment to existing legisla-tion must be complemented by a technical reportquantifying the budgetary impact of each provision andsupplying indications on means of financing
Inevitably there is a risk of underestimating the budget-ary impact of proposed policy measures or overestimat-ing the impact of provisions ensuring their financingThis would not represent an insurmountable problem inthe presence of effective expenditure control mecha-nisms However the experience is that budgetary moni-toring and control have suffered from the complexityand opaqueness of budgetary procedures delays inavailability of information loose practices and institu-tional weaknesses
The lsquoexpenditure freezersquo law aims at making existingcontrol mechanisms more effective To do so it intro-duces new stringent procedures on the one hand facili-tating the monitoring and control of the implementationof legislation and on the other hand allowing emergencyaction to deal with significant slippages from the overallpublic finance objectives of the governmentrsquos medium-term economic and financial plan (DPEF) (3)
In practice all legislation introducing new or higherexpenditure must indicate for every year and for everyprovision the lsquoauthorised expenditurersquo that is dependingon the type of law either the maximum ceiling of dis-bursements or the expenditure forecast attached to the pro-vision In the first case the General Accounting Office(Ragioneria Generale dello Stato RGS) which is the armof the Ministry for the Economy and Finance entrustedwith surveillance and control powers determines whetherthe ceilings are reached and in this event blocks furtherexpenditure appropriations (4) The Minister for Economyand Finance reports to Parliament and the application oflegislation is suspended unless additional financing can beprovided through a new legislative provision In the sec-ond case legislation must define a lsquospecific safeguardclausersquo to compensate for expenditure in excess of fore-cast Surveillance is entrusted to the RGS
The provisions for lsquoemergency actionrsquo empower theMinister for the Economy and Finance to intervene in thegeneral case in which the RGS detects a lsquosignificantdivergencersquo from the overall budgetary objectives of theDPEF (5) In this case the Minister for the Economy and
yen1part Law No 4681978yen2part Law No 3621988
yen3part This may be considered the more structural aspect of the law as suggestedby Grilli (2003)
yen4part An appropriation is a legal obligation to carry out a payment and as such itis distinct from a disbursement which is the moment in which the paymentis effectively carried out
yen5part The size of the budgetary slippage which gives rise to a lsquosignificant diver-gencersquo is not defined in the law An estimated divergence of 03ndash05 per-centage points from the revised budgetary objective of 21 of GDP wasconsidered grounds in late 2002 for applying the law (Grilli 2003)
234
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
Finance proposes the necessary corrective measures tothe Council of Ministers The government decides on themeasures and the Prime Minister adopts a decree (atto diindirizzo) containing the general guidelines for thegovernment action to secure more effective expendituremonitoring and control The relevant ParliamentaryCommittees have 15 days to express an opinion afterwhich the Minister for the Economy and Finance can inany case adopt the necessary measures by decree Cru-cially the Minister for the Economy and Finance canlimit State budget appropriations by fixing proportionalacross-the-board cuts on State expenditure excludingwages of public sector employees pensions and otherfixed or obligatory expenditure (1)
Finally the new law also bars the widespread practice ofadministrations charging new expenditure appropriationsto a financial year after it has ended (that is 31 Decem-ber) In addition it reduces the number of years duringwhich capital expenditure arrears can be carried over inthe budget Both these measures are likely to exert adampening effect on expenditures
Given the short period elapsed since the adoption oflsquoexpenditure freezersquo law a full evaluation of its impactmust necessarily be left for future discussion (2) At thisstage it is only possible to examine some of its featureswhich may shed light on its potential to ensure an effec-tive control of expenditures
In some respects the law is less innovative than it mightappear budget appropriations for some non-obligatoryexpenditures (for example purchases of goods and serv-ices) already constituted absolute expenditure ceilingsYet in view of the prevailing practices the reinstating ofprocedures should not be underrated Moreover the mon-itoring of expenditures is now carried out more stringentlyon each single provision instead of in terms of balance-sheet items as in the past
A shortcoming of the law is the limited field of applica-tion in case of detection of a significant budgetary slip-page de facto the expenditure cuts concern essentiallyexpenditures for goods and services and some capitalexpenditures Moreover while the lsquosignificant diver-gencersquo is identified for the general government thedomain of application of the law is restricted to expend-iture of the central government and of non-territorialpublic institutions and organisations such as universitiesor social security funds (3) leaving out expenditure ofthe local administrations (4) This represents a clear lim-itation in the light of past experience (the significantoverruns due to regional expenditure for healthcare in2000 and 2001) and of the budgetary control challengesgenerated by the decentralisation process Anotherweakness of the law could be the fact that since expend-iture cuts in case of significant divergence are uniformlyapplied in principle this does not allow selective actionFinally the lsquoemergencyrsquo instrument of the law hasintrinsically temporary effects in the absence of morefundamental action in the following fiscal year pastexpenditure levels can again be reinstated
The innovative features of the lsquoexpenditure freezersquo laware undoubtedly the wide-ranging powers granted to theMinistry for Economy and Finance in particular theobligation for the RGS to suspend expenditure appropri-ations when the ceilings are reached and the possibility(not the obligation) for the minister to act in the event ofsignificant divergences (even if necessary by supersed-ing the role of parliament) The effective implementationof the emergency procedures depends to a great extenton the relative powers of the Ministry for Economy vis-agrave-vis the rest of the government In any case the powerto impose across the board expenditure limitations is jus-tified by the seriousness of the event a significant diver-gence This confers to the instrument an exceptionalnature whose effectiveness may well depend on itsimplicit threat content The threat of ex post action mayenhance ex ante the role of the Ministry for Economy inthe phase in which the budget is drawn giving addedsubstance to the pivotal role it has long been assigned inthe budgetary process Moreover the credibility of amore stringent application of surveillance procedurestogether with the risk of across the board expenditure
yen1part Interest on public debt accrued liabilities for loan reimbursements obliga-tions ensuing from Community law and international agreements obliga-tions ensuing from contracts etc The law does not contain an exhaustivelist of expenditures In practical terms the RGS has chosen to identify aslsquoobligatory expenditurersquo also lsquojuridically complete obligations vis-agrave-visthird partiesrsquo (Grilli 2003)
yen2part The ministerial decree of end-November 2002 established a cut in budget-ary appropriations of 15 The RGS estimates this resulted in a reductionof EUR 18 billion (02 of GDP) in the State sector borrowing require-ment and EUR 21 billion (02 of GDP) in the general government deficit(Maastricht definition)
yen3part The Minister for Economy can reduce working expenses of non-territorialpublic institutions Surveillance is carried out by the internal audit bodies
yen4part In 2002 the law was applied also to the local health institutions (aziendesanitarie locali ASL) However the action spurred strong reaction from theregions and the ASL As a result the Minister for the Economy has ruledthat in future the law will not be applied to regional funding for healthcare
235
P u b l i c f i n a n c e s i n E M U 2 0 0 3
limitations will likely discourage unrealistic ex anteexpenditure projections
The law should increase reliability of budgetary forecastand transparency enhance surveillance strengthen con-trols on expenditures and avoid overruns and its applica-
tion should allow quick action to address budgetaryimbalances However it remains to be seen whether itsprovisions in particular the freezing of laws reachingtheir authorised expenditure limits and the across theboard limitations on all State non-obligatory expendi-ture leads to an effective control of expenditures
236
9 Luxembourg
Recent developments
Luxembourg has enjoyed many consecutive years of highgrowth and substantial fiscal surpluses In 2000 and2001 the general government surplus even increased tothe unprecedented levels of 61 and 64 of GDP respec-tively In 2001 the general government balance increasedby an estimated 2 percentage points of GDP due to atransaction concerning the satellite company ASTRAThis was recorded as the sale of a non-produced non-financial asset that is as a negative capital expenditureIn 2002 the fiscal accounts remained in surplus net lend-ing of general government amounted to 26 of GDPThe outcome for 2002 was more favourable than previ-ously expected in the light of the economic slowdownlargely owing to strong corporate tax revenues and lower-than-expected expenditure from special funds
In 2002 tax revenues remained strong in spite of thesharp economic slowdown This is to a large extent dueto high corporate tax revenues partly reflecting thelagged impact of strong earnings in previous years withhigh economic growth Some acceleration in the collec-tion of corporate tax arrears played a decisive role aswell while income tax receipts also remained relativelyfavourable In 2002 taxes on income and wealthincreased by 75 while the receipts of indirect taxesrose by a modest 22 By contrast the rate of growthof social security premium revenue slowed down to66 compared to an increase of around 13 in 2001Total current resources of general government increasedby around 5 However total current expenditure rosemuch faster at around 10 which is higher than theincrease by around 9 in 2001 Strong increases in thepublic wage bill (by approximately 10 ) and in socialtransfers other than in kind (by some 14 ) were mainlyresponsible for the acceleration in expenditure The lat-ter largely reflect the substantial increases in pension andsocial security payments agreed among social partners atthe so-called Rentendeumlsch In addition total publicinvestment also increased rapidly in 2002 by 116
compared to close to 10 in 2001 As a consequencethe ratio of total harmonised government expenditure toGDP (disregarding the satellite transaction mentionedabove) rose by some 1 percentage points to more than45 the highest level since the mid-1990s
Relatively strong revenue from corporate taxes can beexpected to continue into 2003 and possibly 2004despite adverse developments in corporate profits overthe last few years This is the case because the receiptscan be levied up to five years back Hence payment of aconsiderable part of claims accrued in the favourableyears before 2001 is still due while the advances to bepaid are still relatively high as they are being computedon the basis of the still-favourable company accounts ofa few years back As regards the development of taxablepersonal income the effect of decelerating employmentand compensation on revenue will be mainly felt from2003 onwards only
According to the Commission spring 2003 forecasts thegeneral government balance is expected to deteriorate in2003 and 2004 to a deficit of 02 and 12 of GDPrespectively This reflects the lagged impact of theeconomic slowdown on tax revenue and social securitycontributions in combination with continued high ratesof growth of public current and investment expenditureAs regards tax revenues inevitably the impact of thecollection of tax arrears will fade after some years Thiswill also influence revenues of local government that aredirectly linked to corporate taxes Adverse developmentsin the central government accounts would account formost of the deterioration of the general government bal-ance in 2003 and 2004 while local government revenueswould be affected negatively by the effects of the earliertax reforms from 2004 onwards By contrast the balanceof social security funds would still be positive albeit toa lesser extent than in recent years The already lowgross government debt ratio is forecast to decline some-what further to 34 of GDP in 2004
237
P u b l i c f i n a n c e s i n E M U 2 0 0 3
Fiscal prospects in a period of weaker growth
Luxembourg is a very open economy sensitive to cycli-cal swings which have a sizeable impact on publicfinances The sensitivity of government finances to thebusiness cycle is accentuated by the large cross-borderflows of workers (which affect the payment of socialsecurity contributions) and the large exposure to thehighly volatile financial sector However reliable esti-mates of the impact of the current economic slowdownon government finances in Luxembourg are very diffi-cult to make for a number of reasons
First the higher volatility of key macroeconomic aggre-gates and the importance of cross-border labour flowsleads to a higher margin of uncertainty for fiscal projec-tions compared to larger EU economies (the net pay-ments of excise taxes from the customs union with Bel-gium is a notably difficult item) Second the structure ofthe tax system is such that a sizeable part of taxes are col-lected with substantial and variable lags This also makesit harder to assess the impact of the tax reforms of 2000and 2001 on revenue Finally the impact of operations ofspecial funds that accumulated from the fiscal surplusesachieved in past years may cloud forecasts of changes inthe government balance based on past trends
Table VI18
Composition and balances of general government Luxembourg (1) (as of GDP)
2000 2001 2002 2003 2004
Government balance (2) 61 64 26 ndash 02 ndash 12
mdash Total revenue 457 466 481 460 451
Of which mdash current taxes 302 297 306 300 291
mdash social contributions 114 124 129 129 126
mdash Total expenditure (2) 396 402 455 463 464
Of which mdash collective consumption 67 71 77 80 82
mdash social transfers (3) 233 249 273 289 297
mdash interest expenditure 03 03 04 02 02
mdash gross fixed capital formation 41 43 47 52 55
Primary balance (2) 64 67 30 00 ndash 11
Pm Tax burden 415 418 432 426 413
Government debt 56 56 53 41 34
Pm Cyclically-adjusted balance na na na na na
Pm Cyclically-adjusted primary balance na na na na na
(1) Commission spring 2003 economic forecasts(2) UMTS receipts excluded(3) In kind and other than in kind
Source Commission services
Table VI19
Key figures of the Luxembourg stability programme (1) (2001ndash05)
2001 2002 2003 2004 2005
Real GDP growth (annual change) 10 05 12 24 31
General government budget balance ( of GDP) 61 ndash 03 ndash 03 ndash 07 ndash 01
Primary balance ( of GDP) 64 02 00 ndash 05 01
Government debt ( of GDP) 53 51 41 38 29
(1) UMTS receipts excluded
Source 2002 update of the stability programme of Luxembourg
238
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
Clearly the starting position of government finances ofLuxembourg is very favourable given the healthy surplusreached before the current economic slowdown started thevery low level of government debt (which stood at 53 of GDP in 2002) and the sizeable net asset position stem-ming from cumulated surpluses However some potentialrisk factors should not be underestimated Most impor-tantly the rise in government expenditure in Luxembourghas been very strong in the recent years Total governmentexpenditure increased by a cumulative 40 between 1998and 2002 the average yearly increase during that periodwas around 8 Public investment also increased veryrapidly in the past few years by close to 10 on averagein nominal terms between 1998 and 2002
Luxembourg has accumulated some reserves that areavailable to cushion the negative budgetary impact of atransitory economic slowdown However such reservescould not cater for the long-term impact on the budgetposition of a prolonged slowdown in economic growth
Existing reserves should be used to fund the budgetarycosts of ageing that will increase markedly in the longerterm Thus the existence of substantial reserves in thesocial security system (estimated at around 22 of GDPin 2001) is fully justified by the budgetary costs of age-
ing and is necessary to guarantee the payment of pensionclaims also for non-resident workers in the event ofreduced employment growth In addition the existenceof substantial surpluses in social security funds is not afeature that can be for certain in the medium term asthey depend on continuing rapid growth of employmentof both residents and cross-border workers
While it is clearly important to address the budgetaryconsequences stemming from a cyclical slowdown aperhaps even more daunting challenge for fiscal policyin the period ahead stems from the possibility that eco-nomic growth may slow down for a prolonged period torates well below the ones enjoyed for many years in thepast decades At the present juncture estimates of therate of potential real GDP growth of the Luxembourgeconomy are surrounded by large margins of uncer-tainty However it seems likely that in the years aheadLuxembourg will experience growth rates that on aver-age will be lower than in the decade up to and including2000 Hence policy measures might be needed to adjustthe growth of public expenditure to a rate consistent withthe revenue base This would help ensure that a budget-ary position close to balance or in surplus would beachieved and maintained in the medium term
239
10 The Netherlands
Recent developments
The fiscal accounts in the Netherlands deteriorated mark-edly in 2002 against the background of a sharp economicslowdown After having reached a surplus of 01 ofGDP in 2001 the general government balance turned intoa deficit of 11 of GDP in 2002 (a worse outcome thanthe 05 of GDP deficit expected in the 2003 budget)This deterioration mainly reflects the impact of the eco-nomic slowdown as well as the lagged impact of taxreforms Revenue shortfalls related mainly to corporatetaxes and increases in tax-exempt pension premiums paidinto private schemes According to Commission calcula-tions the cyclically-adjusted general government deficitremained broadly constant in 2002 at around 1 of GDP
The fiscal outlook is highly uncertain at the time of writ-ing due to the political situation Following elections inMay 2002 a new cabinet was formed in July The coali-tion parties formulated their policy proposals in a coalitionagreement known as the lsquostrategic accordrsquo which set outthe fiscal policy objectives for the whole 2003ndash06 cabinetterm of office However the cabinet fell on 16 Octo-ber 2002 which led to general elections on 22 January2003 At the time of writing of this report negotiations toform a new coalition government following the electionsare still ongoing The 2003 budget was approved by Par-liament with some relatively minor changes despite thefall of the government and contains a package consistingof reallocations expenditure increases in some areas andtax revenue raising measures The net combined impact ofthese measures on the government balance is an estimatedimprovement of around EUR 5 billion or approximately1 percentage point of GDP in 2003 According to the 2003budget this would result in a stabilisation of the deficit at05 of GDP in 2003 with real GDP growth of 1
However in the absence of additional measures the deficitin 2003 will increase further than forecast in the budget dueto the less favourable starting position of public finances in2002 according to the latest data available and the worsened
economic outlook Under the technical assumption of nopolicy changes and assuming real GDP growth of 05 the Commission spring 2003 forecasts expect the generalgovernment balance to deteriorate in 2003 to a deficit of16 of GDP as unfavourable cyclical developments willcontinue weighing on fiscal revenues The deficit wouldincrease further to 24 of GDP in 2004 under the no pol-icy change assumption reflecting the weakness of theexpected upturn as well as the lagged impact of slowgrowth in 2003 According to the estimates in the springforecast the cyclically-adjusted general government bal-ance would improve slightly in 2003 and deteriorate againto a deficit of around 1 of GDP in 2004 (1) As a result ofprojected deficits and low economic growth the gross gov-ernment debt ratio would fall only slightly in 2003 andincrease again somewhat in 2004 to 528 of GDP
An incoming government will most likely announce con-solidation measures even though a full assessment canonly be given on the basis of a coalition agreement Hencefor 2004 the current projected deficit may be interpreted asan upper bound to expected deficits An important chal-lenge for the incoming government is to leave some scopefor automatic stabilisers to work while achieving the nec-essary improvement of the underlying balance
The sensitivity of public finances to the business cycle and the importance of expenditure rules
As in other open EU economies Dutch governmentfinances are sensitive to cyclical swings in activity Thevery marked deterioration of government finances to alarge extent due to the impact of the economic slowdown
yen1part Note that the forecast horizon here covers the period up to 2004 Owingto the characteristics of the estimation method the outcome for the cycli-cally-adjusted balance in any year will change when the forecast horizonis altered Thus forecasts covering a longer time horizon such as forinstance used in stability programmes will typically yield a somewhatdifferent result even if the same input data would be used for the periodup to and including 2004
240
P a r t V IM e m b e r S t a t e d e v e l o p m e n t s
that started in 2001 testifies to this According to Com-mission calculations based on the spring 2003 forecaststhe cyclically-adjusted general government balanceremained stable in 2002 (showing an estimated deficit of10 of GDP) Thus the deterioration of the nominalgeneral government balance by 12 percentage points in2002 seems to be wholly accounted for by cyclical factorswith a shortfall in revenues being the main determinant
The precise impact of cyclical conditions on publicfinances of course also depends on the composition ofGDP growth Typically a change in the growth rate ofdomestic demand will have a more profound impact onthe fiscal accounts than a shock to external demand Inaddition the structure of revenue and expenditure alsodetermines the sensitivity of public finances to the busi-ness cycle For instance the share of corporate taxeswhich are highly sensitive to cyclical conditions in totaltax receipts has increased markedly in the Netherlands inthe past decade or two In 1987 this share was approxi-mately 13 of total tax revenue It increased to more than17 in 1997 and 1998 and stood at around 16 in 2001
This mainly reflects three factors First an increase incorporate tax receipts as a percentage of GDP from36 in 1987 to 41 in 2001 Second the fact that thesubstantial reduction in the overall tax burden since the
late 1980s was largely achieved through a reduction inincome taxes thus raising the relative share of corporatetaxes as a source of revenue Third the economic boomof the second half of the 1990s which strongly boostedcorporate tax receipts The latter effect is likely to be atleast partially reversed as the impact on tax revenue oflower corporate profits in the wake of the cyclical down-turn feeds through Thus seen from a longer-term per-spective the increasing share of corporate taxes as asource of revenue has arguably increased the sensitivityof Dutch government finances to cyclical fluctuations
Expenditure restraint is important in the current situationto avoid a further deterioration of the general govern-ment balance in response to prolonged weak economicactivity A continuation of the basic tenets of the currentbudgetary framework which uses expenditure rules as acornerstone for fiscal policy may be a useful instrumentto help maintain stable government finances in themedium term and contain the growth of public expendi-ture In this respect the Dutch experience may be instruc-tive (see part V of this report for an analysis of the roleof expenditure rules in the EU context)
It suggests that it helped reduce the incidence of ad-hocmeasures in response to unexpected changes in the vola-tile nominal government balance Note however that
Table VI20
Composition and balances of general government the Netherlands (1) (as of GDP)
2000 2001 2002 2003 2004
Government balance (2) 22 01 ndash 11 ndash 16 ndash 24
mdash Total revenue 474 465 461 459 453
Of which mdash current taxes 242 245 244 239 235
mdash social contributions 171 153 148 156 153
mdash Total expenditure (2) 453 464 472 475 477
Of which mdash collective consumption 106 109 113 112 113
mdash social transfers (3) 238 239 248 252 254
mdash interest expenditure 39 35 32 30 29
mdash gross fixed capital formation 32 34 35 36 35
Primary balance (2) 61 36 21 15 05
Pm Tax burden 416 400 394 396 389
Government debt 558 528 526 524 528
Pm Cyclically-adjusted balance ndash 06 ndash 10 ndash 10 ndash 04 ndash 11
Pm Cyclically-adjusted primary balance 33 25 22 26 18
(1) Commission spring 2003 economic forecasts(2) Data for 2000 and 2001 (except cyclically-adjusted) include UMTS receipts of 07 and 02 of GDP respectively(3) In kind and other than in kind
Source Commission services
241
P u b l i c f i n a n c e s i n E M U 2 0 0 3
while the use of real expenditure ceilings should in prin-ciple facilitate the working of automatic stabilisers iteffectively mitigates somewhat their working on theexpenditure side That said this need not be a majordrawback since in the Netherlands automatic stabilisersmainly work via the income side
Rules to limit the increase of a large part of total publicexpenditure under pre-determined ceilings defined in realterms were introduced in the Netherlands by the first pur-ple cabinet in 1994 to be applied from 1995 onwardsSince then relevant government expenditure was success-fully kept below these ceilings (see Table VI21 for theoutcome during the second purple cabinet) although thiswas facilitated by strong economic growth and conse-quent lower expenditure (mainly on interest paymentsand unemployment benefits) for most of the period 1996ndash2002 covered This enabled a redistribution of windfallsto sub-sectors of general government with higher-than-projected expenditure growth as well as additionalexpenditure increases in areas such as healthcare educa-tion infrastructure investment and public safety
The overall experience with the expenditure rule sincethe mid-1990s has been positive as it is widely believedthat it helped contain expenditure growth Furthermoreclearly defining fixed expenditure margins for the wholecabinet period helped anchor expectations of economicactors However the fact that the budgetary frameworkwas respected was not enough to prevent public financesfrom deteriorating in the wake of the economic slow-
down This was partly because higher revenue and lowerexpenditure during the years of higher-than-expectedeconomic growth had been used to intensify spendingrather than to use it as an additional buffer But moreimportantly one has to bear in mind that the parametersof any budgetary rule need to be adjusted to changingeconomic prospects in order to ensure that fiscal policyremains consistent with a budgetary position close tobalance or in surplus in the medium term
In the strategic accord the outgoing government retainedmany of the basic characteristics of the previous budget-ary expenditure rule with some modifications (seeBox VI2) It should be noted that the future of the budg-etary framework is uncertain at the present juncture withnegotiations to form a new government still ongoing Inany case the budgetary rules put in place by an incominggovernment will be put to a genuine test given the sever-ity of the economic slowdown
Table VI21
Expenditure in the Netherlands mdash relevant ceilings and outcome
1999 2000 2001 2002
Targeted (bn EUR ) 1507 1573 1666 1740
Outcome 1492 1563 1666 1738
Overrun (+) underachievement (ndash) ndash 15 ndash 10 00 ndash 02
Source Ministry of Finance
Box VI2 The Dutch framework for expenditure ceilings
The budgetary framework adopted by the outgoing government resembles the one embedded in the coalition agreement ofthe two previous governments The use of expenditure ceilings in real terms for a large part of total expenditure is the pivotalmechanism of the framework (1) In particular each of the three main sectors of the general government (central govern-ment social security and healthcare) will have to respect separate expenditure ceilings for the relevant expenditure itemsidentified (irrespective of revenues) In case overruns occur they should be compensated within each sector
bull Fiscal projections are based on cautious macro-economic assumptions
bull The automatic stabilisers will be allowed to work freely on the revenue side as long as the government balance will bebetween 0 of GDP and a surplus of 25 of GDP There are also provisions in case a (nominal) surplus of more than1 of GDP would emerge mdash a scenario that at present appears to be not very relevant
yen1part A technical change to the framework adopted by the previous government is that in the deflator for gross domestic expenditure will be used to calculateexpenditure ceilings in nominal terms The previous budgets used the GDP deflator which is more sensitive to shocks to the terms of trade
242
11 Austria
Recent developments and medium-term prospects
In 2002 general government finances in Austria weak-ened markedly From a surplus of 03 of GDP in 2001the budgetary position deteriorated by almost 1 percent-age point to a deficit of 06 of GDP despite the factthat output growth accelerated slightly to 10 from07 in 2001 This outcome compares with an initialobjective of a balanced budgetary position set in theNovember 2001 stability programme based on a realgrowth assumption of 13 and also exceeds the deficittarget of ndash 02 of GDP retained in the low-growth sce-nario assuming a real GDP expansion of 09
While in 2001 a strong rise in tax revenues helped toimprove the cyclically-adjusted position despite low out-put growth the decline in domestic demand in 2002depressed tax revenues The gross tax intake accountingfor 97 of the revenues in the budget 2002 fell short ofthe budgeted amounts by 32 or 08 of GDP How-ever the marginal item lsquoother revenuesrsquo increased to theextent that total revenues came in only slightly below thebudget (by ndash 01 or 003 of GDP)
Although expenditure exceeded the budgeted figure by23 or 06 of GDP this increase was lower thananticipated against the background of rising unemploy-ment and almost stagnating employment entailinglower pension contributions and thus higher federal out-lays for public pensions In addition the flood disasterin summer 2002 and the emergency package adopted inits aftermath was expected to increase spending but hadcontrary to expectation virtually no budgetary impactin 2002
These factors had led the Austrian fiscal authorities to apreliminary deficit estimate of 10 of GDP in 2002Due to statistical reasons and data revisions the actualdeficit of 06 of GDP turned out clearly lower despitea decline in the surplus at the Laumlnder level
The March 2003 update of the stability programmeprojects that in the near term the general governmentfinancial position will deteriorate markedly both innominal and in cyclically-adjusted terms before improv-ing again as late as in 2007 The deficit will remain onaverage at 1 of GDP until 2007 which is in sharpcontrast to the objective of the previous programme aim-ing at a balanced budget position in 2003 and a smallsurplus in 2004 and 2005
Specifically according to the stability programmes thedeficit is forecast to deteriorate from 06 in 2002 to13 of GDP in 2003 despite higher output growth Whilea temporary improvement is expected in 2004 the plannedincome tax reform will take its toll as of 2005 when thedeficit is estimated to increase to 15 of GDP and toremain above 1 of GDP in 2006 A sizeable improve-ment to a deficit of 04 of GDP is forecast only for 2007
The budgetary strategy has changed significantly com-pared with the previous programme The highlights ofthe new strategy are twofold a fundamental reform ofthe public pension system tackling many of its key prob-lems and a sizeable income tax reform
On the revenue side the tax reform is intended to reducethe tax burden to 43 of GDP by the year 2006 It is esti-mated to cost EUR 3 billion or 13 of GDP and to takeeffect in two steps a smaller one in 2004 (EUR billion)and a more substantial one of EUR 1 billion or almost1 of GDP in 2005 The reform aims at lowering taxesfor low and middle incomes as well as on retained profitsstrengthening work incentives for lower incomes reduc-ing non-wage labour costs and rendering the tax systemmore environment-friendly Conversely an increase inenergy taxes effective as of 2004 should entail additionalrevenues of EUR 450 million or some 03 of GDP
On the expenditure side a comprehensive pensionreform which is announced to start in 2004 and will bephased in until 2009 is without doubt the most remark-able feature of the updated stability programme
243