CONTRIBUTION ISSUE 2016/07 A PROPOSAL TO -...

ISSUE 2016/07 MARCH 2016 A PROPOSAL TO

REVIVE THEEUROPEAN FISCALFRAMEWORK

GRÉGORY CLAEYS, ZSOLT DARVAS AND ÁLVARO LEANDRO

Highlights

• Pro-cyclical fiscal tightening might be one reason for the anaemic economic recoveryin Europe, raising questions about the effectiveness of the EU’s fiscal framework inachieving its two main objectives: public debt sustainability and fiscal stabilisation.

• In theory, the current EU fiscal rules, with cyclically adjusted targets, flexibilityclauses and the option to enter an excessive deficit procedure, allow for large-scalefiscal stabilisation during a recession. However, implementation of the rules ishindered by the badly-measured structural balance indicator and incorrect forecasts,leading to erroneous policy recommendations. The large number of flexibility clausesmakes the system opaque.

• The current inefficient European fiscal framework should be replaced with a systembased on rules that are more conducive to the two objectives, more transparent,easier to implement and which have a higher potential to be complied with.

• The best option, re-designing the fiscal framework from scratch, is currentlyunrealistic. Therefore we propose to eliminate the structural balance rules and tointroduce a new public expenditure rule with debt-correction feedback, embodiedin a multi-annual framework, which would also support the central bank’s inflationtarget. A European Fiscal Council could oversee the system.

Grégory Claeys ([email protected]) is a Research Fellow at Bruegel. ZsoltDarvas ([email protected]) is a Senior Fellow at Bruegel. Álvaro Leandro([email protected]) is a Research Assistant at Bruegel. The authors aregrateful for the comments of Bruegel colleagues and of the participants and the fourdiscussants, Svend Hougaard Jensen, Kathrin Muehlbronner, Lucio Pench and FoccoVijselaar, at the 3 March 2016 presentation of the paper.

Telephone+32 2 227 4210 [email protected]

www.bruegel.org

BRU EGE LPOLICYCONTRIBUTION

A PROPOSAL TO REVIVE THE EUROPEANFISCAL FRAMEWORK

GRÉGORY CLAEYS, ZSOLT DARVAS AND ÁLVARO LEANDRO

02

BR U EGE LPOLICYCONTRIBUTION

1. See for instance Buiter etal (1993).

2. Reasons for the deficitbias include informational

problems, impatience, elec-toral competition, common-

pool problems andtime-inconsistency (see for

example Portes and Wren-Lewis, 2014).

3. See for example Blan-chard and Leigh (2013),

Holland and Portes (2012),Wren-Lewis (2013) and

Barbiero and Darvas(2014).

4. The first two rules arefrom the EU Treaty, while

the specification of the1/20th debt reduction

requirement is from the Six-Pack. The third rule origi-

nates from the Stability andGrowth Pact (SGP) require-

ment for the budget to be“close to balance or in sur-

plus”, while the MTOappeared in the 2005

reform of the SGP, and theminimum numerical

requirements for the euroarea come from the Fiscal

Compact. The fourth rule isfrom the Six-Pack.

1 INTRODUCTION

The European Union’s fiscal framework, which con-sists of fiscal rules, budget procedures and insti-tutions, has been the subject of majorcontroversies since it was put in place in the1990s1. Member state non-compliance with therules in the early 2000s, and the perceived rigid-ity of the rules, led to reforms in 2005. The globaland European economic and financial crises ledto further major changes to the fiscal frameworkin the form of the so-called Six-Pack (2011), FiscalCompact (2012) and Two-Pack (2014).

Assessments of the current framework varywidely. Marzinotto and Sapir (2012) and Micossiand Peirce (2014) argue that the current rules rep-resent a sophisticated system of surveillance andex-post control that provides sufficient room formanoeuvre under exceptional circumstances. Bycontrast, Manesse (2014) and Ódor and P. Kiss(2015) propose to design fundamentally newfiscal frameworks. Several authors from the Inter-national Monetary Fund (Andrle et al, 2015) sug-gest various options for simplifying and makingthe EU fiscal governance framework more effec-tive, of which the most ambitious would pro-foundly change the current rules.

Revision of the EU’s fiscal rules appears to be offthe table in the short term. The Five Presidents’Report (Juncker et al, 2015) did not make anyproposal to amend the numerical fiscal rules. Thispreference for the status quo is probably rootedin the political difficulty of starting a newdiscussion about European fiscal rules so soonafter the 2011-14 reforms. But it might also berelated to the currently calm government bondmarket situation.

Meanwhile, the European fiscal framework mightnot be effective at achieving its two key objec-tives: (1) to discourage the deficit bias of the gov-

A PROPOSAL TO REVIVE THE EUROPEAN FISCAL FRAMEWORK

ernment in order to ensure the long-term sustain-ability of public debt2 and (2) to leave scope forcounter-cyclical fiscal policy. The latter objectivehas been the main subject of discussion, becausemany researchers3 concluded that the fiscalstance has been too restrictive since 2010, takinginto account the economic situation in most EUcountries and in the euro area as a whole. Pro-cyclical fiscal tightening in a recession impliesthat long-term public debt sustainability isachieved in an ineffective way, because unduefiscal consolidation in a recession can prolongeconomic weaknesses and keep the debt ratiohigher, triggering further fiscal consolidation. Otherkey issues are whether the framework is suffi-ciently implementable, transparent and under-standable to the general public, and whether thereis strong national ownership of the rules.

This Policy Contribution assesses the suitabilityof the current European fiscal framework for ful-filling its two key objectives. We argue thatbecause the status quo would preserve an ineffi-cient system, while the first-best solution for aEuropean fiscal framework is politically unrealis-tic, a change in the Stability and Growth Pact andthe Fiscal Compact and the establishment of aEuropean Fiscal Council are needed.

2 THE EU’S CURRENT FISCAL FRAMEWORK

The fiscal framework includes numerical fiscalrules and requirements for budgetary proceduresand independent fiscal councils.

Numerical fiscal rules

The basic fiscal rules are relatively simple4:

1 The budget deficit must be below 3 percent ofGDP;

2 Gross public debt must be below 60 percent ofGDP:

03

BR U EGE LPOLICYCONTRIBUTIONA PROPOSAL TO REVIVE THE EUROPEAN FISCAL FRAMEWORK

5. “The expenditureaggregate shall exclude

interest expenditure,expenditure on Union

programmes fully matchedby Union funds revenue

and non-discretionarychanges in unemployment

benefit expenditure. Theexcess expenditure growth

over the medium-termreference shall not be

counted as a breach of thebenchmark to the extent

that it is fully offset byrevenue increasesmandated by law.”Regulation (EU) no

1175/2011 of the EuropeanParliament and of the

Council, http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32

011R1175&from=EN.

• If it is higher, it must decline annually by atleast 1/20th of the gap between the actualdebt level and the 60 percent referencevalue;

3 The structural budget balance (that is, thebudget balance which excludes the impact ofthe economic cycle and one-off fiscalmeasures) must be higher than the country-specific medium-term objective (MTO), which,in the case of euro-area countries, has to bechosen at or above -0.5 percent of GDP, or -1percent for countries with a debt-to-GDP ratiobelow 60 percent. • If the structural balance is lower than the

MTO, is must increase by 0.5 percent of GDPper year as a baseline;

4 An adjusted measure of real governmentexpenditures (nominal expenditures deflatedby the GDP deflator forecast)5 cannot growfaster than the medium-term potential eco-nomic growth if the country’s structural balanceis at its MTO or higher; • If the structural balance has not yet reached

its MTO, expenditure growth must be lowerthan potential growth, in order to ensure anappropriate adjustment towards the MTO.

When the first two rules are met, the country is inthe so-called ‘preventive arm’ of the Stability andGrowth Pact (SGP). If one or both of the first tworules are not met, the country is in the ‘correctivearm’ of the SGP and an Excessive Deficit Procedure(EDP) is opened. Breaching the rules can lead tofinancial sanctions in the corrective arm for allcountries, and in the preventive arm for euro-areacountries.

Flexibility and discretion

The numerical rules are rather simple, at least con-ceptually. However, there are so many flexibilityclauses and exceptions that the whole frameworkbecomes opaque. Certain deviations from therules are allowed for an unusual event outside thecontrol of the member state concerned and whichhas a major impact on the financial position of thegeneral government, a severe economic downturnin the member state, a severe economic downturnfor the euro area or the EU as a whole, an unex-pected adverse economic event, when structuralreforms are implemented or planned, when the

government contributes to EU-funded invest-ments, when the government implements pen-sion reforms, or when “relevant factors” emerge.The 3 percent deficit rule can be disregarded whenthe deviation from it is small and temporary, whilethe 1/20th debt reduction rule can be disregardedwhen the country is assessed as doing enoughfiscal consolidation.

The European Commission has wide-ranging dis-cretionary power in the assessment of fiscal per-formance and plans. Discretion can be a blessingbut also a curse. In unusual times it can be helpfulto get rid of rigid fiscal rules and calibrate fiscalpolicy to the specific circumstances. But discre-tion might also encourage neglect of the rules inother times. It might also lead to unequal treat-ment of countries. Ódor and P. Kiss (2015) arguethat it is difficult to predict the Commission’s deci-sions on flexibility.

Budgetary processes and fiscal councils

The fiscal framework also includes requirementsfor budgetary processes, such as the establish-ment of an effective and transparent medium-term budgetary framework, based on high-qualityforecasts. Each country is requested to submit aStability Programme (euro-area members) or aConvergence Programme (non-euro area mem-bers) in April and a Draft Budgetary Plan (euro-area members) by October of each year. TheCommission assesses the plans for compliancewith the fiscal framework. The Six-Pack also intro-duced a requirement for each country to set up anindependent body, such as a fiscal council, that isresponsible for monitoring compliance with thefiscal rules.

3 ASSESSMENT OF THE EU FISCAL FRAMEWORK

A fiscal framework has two basic objectives: (1) todiscourage the deficit bias of governments inorder to ensure fiscal discipline and the long-termsustainability of the public debt, and (2) to sup-port countercyclical fiscal policy in both good andbad times. In theory, both objectives can beachieved with the current European framework ifthe rules are implemented, but there are so manyfactors hindering their implementation that theframework is ineffective in practice.

A PROPOSAL TO REVIVE THE EUROPEAN FISCAL FRAMEWORKBR U EGE LPOLICYCONTRIBUTION

04

Long-term sustainability

If European fiscal rules are fully adhered to andthere are no unexpected shocks, the public debtratio should generally decline to low levels,because of the debt and structural balance rules.For example, with a nominal GDP growth of 3 per-cent, respecting an MTO of -1.0 percent of GDP (theminimum MTO for euro-area countries with debtbelow 60 percent) ensures that public debt con-verges to 34 percent of GDP.

Given the probability of negative shocks and thecurrent high levels of debt in some euro-areacountries, however, the debt-ratio will remain highand the 60 percent target will probably not bereached at the euro-area level for a long time, evenif rules are complied with. The recent Fiscal Sus-tainability Report (European Commission, 2016)concluded that there is a high medium-term sus-tainability risk for almost a dozen EU countries,which, in our view, could also increase when theEuropean Central Bank ends its quantitativeeasing programme.

The conduct of counter-cyclical policy has animpact on public debt sustainability too. An insuf-ficient counter-cyclical policy in good times leadsto a higher debt level and the inability to providesufficient fiscal stabilisation in bad times. Aninsufficient counter-cyclical policy in bad timesamplifies economic and social problems, and canaffect negatively potential growth and publicfinances in the long run, if hysteresis effects arepresent6.

Countercyclical policy

The other basic objective of a fiscal framework is tosupport countercyclical fiscal policy both in goodand bad times. Here we focus on options for badtimes.

In theory, the 3 percent headline deficit rule andthe structural deficit rule, if respected, allow auto-matic stabilisers to operate even in reasonably

6. As argued by De Long andSummers (2012),

downturns can havepersistent negative

consequences on futureeconomic activity through

various channels: “reducedlabour force attachment on

the part of the long-termunemployed, scarring

effects on young workerswho have trouble beginning

their careers, reductions ingovernment physical and

human capital investmentsas social insurance

expenditures make priorclaims on limited public

financial resources, reducedinvestment in both in

research and developmentand in physical capital,

reduced experimentationwith business models and

informational spillovers, andchanges in managerial

attitudes”.

7. IMF estimates suggestthat the average fiscal sta-bilisation coefficient is 0.7

for advanced countries(Buti and Gaspar, 2015). For

16 EU countries the esti-mated coefficient does not

exceed 0.7 (for most ofthem it is well below), whilefor 5 EU countries it is larger

than 0.7.

8. These calculations arebased on the European

Commission’s 2016 winterforecast.

deep recessions. For example, a structural bal-ance of -0.5 percent of GDP (which is the minimumMTO for euro-area countries with public debt over60 percent of GDP) makes it possible for auto-matic stabilisation of up to 2.5 percent of GDPwithout breaching the 3 percent deficit rule. If thefiscal stabilisation coefficient (which measuresthe response of the overall fiscal deficit tochanges in the output gap) is 0.7, then a 3.6 per-cent of GDP negative output gap is compatible withthe 3 percent deficit criterion7. A negative outputgap equal to or larger than 3.6 percent is a rela-tively rare event: based on the empirical distribu-tion of estimated output gaps between 1965 and2016, such a negative output gap is expected inevery twenty-second year in the 10 core EU15countries (EU members before 2004 not includ-ing five periphery countries) and in every sixthyear in the five periphery EU15 countries (Spain,Portugal, Greece, Ireland and Italy)8, if the histori-cal distribution of output gaps is a good indicationof their future distribution. For the 13 countriesthat joined the EU in 2004 and after, the 1997-2016 period suggests such an output gap can beexpected every ninth year.

In addition, countries might decide to performmore cyclical stabilisation than what is allowed bythe 3 percent deficit rule and thereby enter anexcessive deficit procedure, as highlighted byMicossi and Peirce (2014). Flexibility clauses alsoallow delayed fiscal consolidation after anincrease in the budget deficit. In 2008, the Euro-pean Commission proposed the European Eco-nomic Recovery Plan (European Commission,2008) and invited EU countries to “agree to animmediate budgetary impulse amounting to €200 billion (1.5 percent of GDP), to boost demandin full respect of the Stability and Growth Pact”.When calling for the stimulus, the Commissionnoted that countries that would breach the 3 per-cent deficit limit would be placed under the exces-sive deficit procedure.

Fiscal policy in the United States from 2008-10 isoften portrayed as a good example of effective

‘We find it unacceptable that the EU’s fiscal framework strongly relies on the change in the

structural balance as an indicator, for which the typical one-year revision of the estimate is

larger than the required policy action.’


05

9. We also note that the UKwas requested to increase

its structural balance by1.75 percent of GDP annu-

ally under the SGP: fiscalconsolidation in the US was

done at exactly the samepace.

10. For example, accordingto the estimated structural

balance indicator, since2013 the UK stopped fiscal

consolidation, yet non-com-pliance with the fiscal

requirements was not sanc-tioned. Instead, in June

2015, the Council issued anew recommendation forthe UK to reach the 3 per-

cent deficit threshold in twoyears, which will be sup-

ported by the expectedimprovement in the UK’s

cyclical situation.

countercyclical policy because it provided a largestimulus in response to the financial crisis. Weconclude that such a stimulus would have been inline with the current EU fiscal rules. With the stim-ulus, the US structural deficit increased to 10 per-cent of GDP, similar to Greece, Ireland, Romania,Spain and the United Kingdom. In the case of theUK, which had been under an excessive deficitprocedure since 8 July 2008, the Council of the EUon 30 November 2009 assessed that the stimu-lus was “an appropriate response”9.

However some countries were constrained bymarket pressure and others decided not to stimu-late their economies as much. In particular, in thelargest EU country, Germany, the structural deficitpeaked at a mere 2.2 percent in 2010.

When the economic cycle started to deteriorateagain in 2012, fiscal consolidation continued inmost EU countries, leading to pro-cyclical fiscalpolicy even in those countries that had amplefiscal space, as argued by Barbiero and Darvas(2014). Barbiero and Darvas also showed thatpublic investment, the expenditure category withthe greatest impact on output growth, suffered themost among the various public expenditure cate-gories throughout the EU.

Germany corrected its excessive deficit in 2011,two years ahead of the deadline set by theCouncil, and fiscal consolidation continued up to2014 when the structural balance increased to asurplus of 0.8 percent of GDP, well above the -0.5percent MTO and also above the requirement setby Germany’s own debt-brake rule. The Germanstructural balance increased much more quicklythan planned in Germany’s Stability Programmesin 2010-13, highlighting the fact that the structuralbalance is an inadequate fiscal target because thegovernment has only limited control over it.Therefore, we conclude that the post-2012 pro-cyclical fiscal tightening in Germany was not theresult of EU fiscal rules, but most likely the resultof domestic political preferences and the relianceon an inappropriate fiscal indicator, the structuralbalance.

To examine how the fiscal rules were interpretedby the Commission and the Council in 2012, welook at the country-specific recommendations

made in summer 2012 and their assessments in2013, for the six largest EU countries.

Among these six countries, the Council requesteda fiscal tightening which seemed pro-cyclical atthe time for only Poland, because in spring 2012the Commission forecast a 0.6 percentage pointdeterioration in the Polish output gap. But for thefive other large countries (France, Germany, Italy,Spain and the UK) the Commission forecast someimprovements in their output gaps in 2012 and2013. Therefore, pro-cyclical fiscal tighteningresulted more from incorrect Commission fore-casts than from a deliberate pro-cyclical fiscalpolicy.

In 2013, when the Commission revised its outputgap estimates to indicate deterioration, seven ofthe eight countries (France, Malta, Netherlands,Poland, Portugal, Slovenia and Spain) which weremissing targets and the deadline under the cor-rective arm were given a delayed deadline for thecorrection of the excessive deficit, citing “unex-pected adverse economic developments”. Theeighth country, Belgium, received a ‘notice’ to stepup fiscal consolidation efforts by 0.25 percent ofGDP to meet the initially planned structural fiscaleffort. Since then, other countries have beengranted similar deadline extensions10. However,while these countries were given more time to cor-rect their deficits, the Commission and the Councilcertainly did not call for a stimulus (even in coun-tries that had ample fiscal space) at a time whenthe cyclical situation deteriorated.

Therefore, concerning counter-cyclical fiscalpolicy in an economic downturn, we concludethat:

• When all rules are met, sizeable fiscal stabili-sation is possible even without entering theexcessive deficit procedure;

• In a deep recession, even a 2008-10 US-stylestimulus is possible by entering an excessivedeficit procedure.

The key problems from the perspective of fiscalstabilisation in a downturn are:

• Estimates and forecasts of the output gap canprove to be incorrect and can therefore mis-

11. The average from 2003-14 was 0.71 percent of GDP

for core EU15 countries,1.84 percent for periphery

EU15 countries, while in2006-14 it was 1.24 percent

for newer member states.IMF and OECD estimates

were characterised by simi-larly large revisions.

12. For example, a 0.3 per-centage point downwardrevision in medium-term

potential growth estimatewould imply that if in spring

2016 a country is allowed toincrease expenditures by

1.5 percent in 2017, inspring 2017 the allowed

growth rate of expendituresis revised downward to 1.2

percent per year. Given thatpublic expenditure amounts

to about half of GDP, a 0.3percent revision in expendi-tures implies a 0.15 percentof GDP impact on the budget

balance, which is muchsmaller than the average

revision in the change in thestructural balance.

13. In some cases therewere major revisions in the

latter two factors. For exam-ple, the 2014 French Stabil-

ity Programme reported thatcyclical unemployment

expenditures amounted to0.2 percent of GDP, while the

2015 French Stability Pro-gramme revised the esti-

mate to 1.3 percent.EU-funded programmes

were indicated at 0.0 per-cent (after rounding) in the

2014 Austrian Stability Pro-gramme, while they was pro-

jected at 0.5 percent in the2013 programme and

reported at 0.4 percent inthe 2015 programme.


06

(left panel of Figure 2 on the next page). However,the estimates using real-time data from the modelof Darvas and Simon (2015) were not subject tosuch large revisions during the crisis years (rightpanel of Figure 2). With the exception of the Com-mission’s 2008 estimates, the typical one-yearrevision for different EU country groups was about0.1-0.5 percentage points per year12. We thereforeconclude that the medium-term potential growthrate estimate was a more suitable indicator thanthe annual change in the structural balance, espe-cially when using a more robust technique thanthe Commission’s current model.

On the other hand, the real-time measurement ofthe expenditure rule is hindered by its depend-ence on GDP deflator forecasts (since the ruleapplies to the real growth of expenditures), theinclusion of EU funding and the non-discretionaryunemployment spending13. Furthermore, inde-

guide fiscal policy and the European Commis-sion’s recommendations;

• The structural balance estimates are subject tomajor revisions and can lead to misguidedpolicy advice;

• When a recession lingers for several years,fiscal rules at best allow the postponement offiscal consolidation instead of suggesting anecessary repeated stimulus;

• In recent years, most EU countries were farfrom their MTO and therefore could not availthemselves of the options offered by the fiscalrules to support the economy with counter-cyclical fiscal policy.

Real-time measurement

While using cyclically-adjusted targets seemsstraightforward and sensible in theory, it is notvery helpful and can even be harmful in practice.Compliance with at least one of the four numericalrules is very badly measured in real time. Thestructural balance and potential output areunobservable variables and their real-timeestimates are extremely imprecise and subject tomajor revisions.

The typical yearly revision both in the level and inthe change in the structural balance is larger than0.5 percent of GDP, ie larger than the requiredbaseline annual adjustment (Figure 1)11. That is,if the Commission forecasts in spring 2016 thatthe structural balance will remain unchanged from2015 to 2016, it is likely that in spring 2017 the2015-16 change in the structural balance will beestimated as half percent or larger (either anincrease or decrease). We find it unacceptablethat EU’s fiscal framework strongly relies on anindicator (the change in the structural balance) forwhich the typical one-year revision in the estimateis larger than the required policy action, especiallysince the revisions are much larger in more uncer-tain times, as indicated by Figure 1.

The revisions of the real-time estimates of themedium-term average potential growth rate(which is used for the expenditure rule) weresmaller than the revisions of the change in thestructural balance estimates, though Commissionestimates were revised substantially during thecrisis, exactly when good guidance was needed

0

0.5

1

1.5

2

2.5

2002

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

EU25Old EU15 coreOld EU15peripheryNew EU10

Figure 1: Average one-year revision in the real-time European Commission estimate of thechange in the structural budget balance (% GDP)

Source: Bruegel. Note: Average absolute revision of the real-time estimate made in spring of the subsequent year. Forexample, the last observation shows the difference betweenthe May 2015 and May 2014 estimates for the 2013-14change in the structural balance (absolute values of thedifferences averaged for the country group indicated in thelegend). We could not find real-time structural balanceestimates made before 2006, but we found real-timecyclically adjusted budget balance estimates made in 2003,2004 and 2005. Therefore, for the first three years shown wereport the revision in the change to the cyclically adjustedbudget balance. EU15 Periphery: Greece, Ireland, Italy,Portugal and Spain; EU15 Core: other 10 countries which weremembers of the EU before 2004. New EU10: member statesjoined in 2004. Bulgaria, Croatia and Romania are notincluded because of data limitations.

07


states which voted for the sanction and againstthe EU as a whole, undermining the cohesion ofthe EU and its peoples. Backlash would be espe-cially harsh if the perception in the sanctionedcountry is that the Commission’s recommenda-

pendent verification of the relevant expenditureaggregate based on publicly available data isimpossible.

Implementation

European fiscal rules are barely implemented. The1/20th debt reduction rule will not be met by Bel-gium, Croatia, Finland, France, Greece, Italy, Por-tugal, Slovenia and Spain in the next three years,according to the IMF’s October 2015 forecasts14.Even the European Commission’s own assess-ment is that only a fraction of the EuropeanSemester recommendations related to the Stabil-ity and Growth Pact are implemented (Figure 3).

Credibility of sanctions

Finally, we note that the threat of sanctions is notcredible. In a time of economic hardship, sanc-tions would make the economic situation worse(Andrle et al, 2015), though when the budgetdeficit is, for example, about 10 percent of GDP, a0.2 percent of GDP sanction would be insignificantcompared to the scale of fiscal problems.

In our view, the political dimensions of a sanctionare more important. Imposition of a financial sanc-tion may lead to backlash against the member

14. The Commission pub-lishes forecasts only one

year ahead, which cannot beused to assess the forecast

change in the debt ratio ofthe next three years, as the

debt rule requires

0

0.5

1

1.5

2

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2003

2004

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

0

0.5

1

1.5

2A. European Commission estimates B. Darvas and Simon estimates

EU25

Old EU15 core

Old EU15 periphery

New EU10 Old EU15 core

Old EU15 periphery

Figure 2: Average one-year revision in the real-time estimate of the medium-term average potentialgrowth rate (%)

Source: Bruegel. Note: Average absolute revision of the real-time estimate made in spring of a year one year later. For example,the last observation on the left panel shows the difference between the May 2015 and May 2014 Commission estimates forthe 2009-18 average potential growth rate, while the right panel shows the estimates for the 2009-14 period using spring2014 and spring 2015 data on the basis of the model of Darvas and Simon (2015) (absolute values of the differences averagedfor the country-group indicated in the legend). The Darvas and Simon (2015) estimates are not available for longer-termforecasts.

0.00

0.20

0.40

0.60

0.80

1.00

2012 2013 2014

Euro area

Non-euro area

Figure 3: Implementation rates of the Stabilityand Growth Pact

Source: Bruegel. Note: We consider recommendations relatedto the SGP made in the context of the European Semester andthe European Commission’s assessments regarding theprogress with the implementation of the recommendations,which is graded on a 5-step scale. We gave a score of 1 to ‘fullimplementation’, a score of 0.75 to ‘substantial progress’, ascore of 0.5 to ‘some progress’, a score of 0.25 to ‘limitedprogress’ and a score of zero to ‘no progress’; we report anunweighted average of those countries for which data isavailable for all years. The horizontal axis indicates the dateof the European Semester recommendations. See Box 1 ofDarvas and Leandro (2015) for further details.

08

BR U EGE LPOLICYCONTRIBUTION A PROPOSAL TO REVIVE THE EUROPEAN FISCAL FRAMEWORK

15. See footnote 2.

16. The EU Treaty explicitlyincludes a ‘no-bailout’

clause and a prohibition ofmonetary financing by the

ECB (respectively in Articles125 and 123 of the TFEU).

However, the authors of theSGP might have believedthat these clauses, com-

bined with market discipline,were not enough to prevent

the free-riding problem orthat they do not represent

credible commitments intimes of crisis, and thereforea European fiscal framework

is needed.

17. Seehttp://ec.europa.eu/econ-

omy_finance/eu_borrower/balance_of_payments/index_en.

htm.

tions were misguided, partly because of the fore-cast and measurement errors described above.

Moreover, game theory also suggests that sanc-tions are unlikely. The final decision on sanctionslies ultimately with the Council, which makesmember states both judges and defendants(Tirole, 2012). Countries might therefore have littleincentive to take an adversarial stance towardsanother member state.

Many countries outside the EU have fiscal rulesbut do not impose financial sanctions on them-selves or on sub-national public entities. In theend, the perception that the rules and the fiscalframework provide economically sound guidancecould be a much more important factor than thefear of sanctions to provide an incentive tomember states to respect the rules.

4 HOW TO REFORM THE EUROPEAN FISCALFRAMEWORK

Do we need a fiscal framework at thesupranational level?

The deficit bias originates from country-specificproblems15. Therefore, it would make sense todesign and implement the fiscal framework at thenational level, which would also increase thedomestic ownership of the framework.

However, there are also good reasons for a certaindegree of European involvement in the design,monitoring and enforcement of fiscal rules,because of cross-border spillovers of fiscalpolicies.

1 Some governments might be tempted to freeride by implementing unsustainable fiscal poli-cies and expecting either a bailout from othergovernments or a monetisation of their debt bythe common central bank (Buiter et al, 1993).This could have a negative impact on their part-ners through increased taxation or inflation. Inthe absence of a bail-out or monetisation, thecountry that runs an unsustainable fiscal policymight face more adverse economic and finan-cial developments, which would impact part-ners through trade and financial links.European involvement in the design, monitor-

ing and enforcement of fiscal rules might limitthe likelihood that any member state will rununsustainable fiscal policies16.

2 Inflationary (deflationary) fiscal policy in oneeuro-area country could impact the averageeuro-area inflation targeted by the EuropeanCentral Bank and trigger a monetary tightening(easing) for everyone (Bénassy-Quéré, Ragotand Wolff, 2016).

3 The differences in fiscal policy in different euro-area countries (such as a competitive and low-debt core and an uncompetitive and indebtedperiphery) provide a major tool to addressprice/wage divergences in a non-optimal mon-etary union, in which factor movements andpurely market-based relative price adjust-ments across countries cannot efficiently com-pensate for price and wage divergences.European involvement in the fiscal frameworkis justified because fiscal policy has a role bothin the build-up and the correction of such diver-gences (Merler and Pisani-Ferry, 2012).

4 Purely national fiscal policies might lead to asuboptimal area-wide aggregate fiscal stancein the absence of proper fiscal policy coordina-tion, and to a suboptimal macroeconomicpolicy mix in the absence of coordinationbetween monetary policy and the aggregatefiscal stance (Buiter, 2006).

Most of these arguments are pertinent for theeuro-area, but trade and financial linkages tend tobe strong between all EU member states and thereis also a bail-out option for non-euro countries (theso called balance of payments facility17). Wetherefore prefer an EU-wide approach and con-clude that given the current institutional setup ofthe euro area and the EU, some EU involvement inthe design of the fiscal framework is justified.

Should the fiscal framework be changed?

We believe that the ‘no change’ vision (no changeto fiscal rules, appointment of an advisory fiscalboard) of the Five Presidents’ Report (Juncker etal, 2015) would be suboptimal. While the frame-work strongly focuses on long-term sustainabilityand allows counter-cyclical policy in a downturnwhen rules are met, several member states per-sistently and even openly disregard the rules, thelarge number of flexibility options makes the

09


18. Barbiero and Darvas(2014) proposed an asym-

metric golden rule for publicinvestment, while Bénassy-

Quéré, Ragot and Wolff(2016) proposed an incre-mental public investment

rule.

19. The potential outputmethod of Darvas and Simon(2015) is conceptually intu-itive and led to more reliablereal-time estimates than the

method of the EuropeanCommission.

20. If the Treaty is notchanged, the 3 percent

deficit rule would continue toexist, but it should not be

given much attention in ourrenewed framework. It wouldcontinue to trigger the open-

ing of an excessive deficitprocedure (EDP), which

should focus on the properimplementation of the

expenditure rule.

whole system opaque and the real-time imple-mentation of the rules is burdened with significanterrors related to the estimation and forecasting ofthe structural budget balance, which can lead tomisguided policy recommendations. While someimprovements can be made to the current frame-work, such as better protection of public invest-ment during an economic downturn18, improvedmeasurement of potential output and therebycyclically-adjusted fiscal indicators19, and clearerprovisions on flexibility options, it would be betterto adopt a framework that is not burdened withsuch problems.

In our view, the best option would be to re-designthe fiscal framework from scratch, which wouldrequire a major overhaul of the EU Treaty. One wayto do that would be to remove completely thebailout option, establish conditions for market dis-cipline to work effectively, allow a large degree offiscal independence to member states and designa cyclical stabilisation mechanism at the Euro-pean level.

However, in our view such an overall of the EU’sand the euro area’s fiscal system is unrealistictoday and therefore we do not develop this sce-nario in this paper. Instead, we make a proposal torevise fiscal rules so that they are more conduciveto long-term debt sustainability and fiscal stabili-sation, more transparent, easier to implement andmore likely to be respected. We also propose theestablishment of a European Fiscal Council tooversee the new framework. Our proposal requiresa change to the Stability and Growth Pact and theFiscal Compact, while the EU Treaty need not bechanged20.

The proposed fiscal rule

We propose to drop the structural balance as anintermediate target of fiscal policy. Instead, wepropose an expenditure rule with a debt-feedbackmechanism, which would make the 1/20th debtreduction rule redundant.

The intuition behind such a proposal is not new.For example, Pisani-Ferry (2002) proposed thatthe emphasis of fiscal discipline should be shiftedaway from the year-by-year monitoring of thedeficit to a more medium-term approach thatfocuses on the long-run sustainability of publicfinances. Anderson and Minarik (2006) arguedthat steering on the expenditure side rather thanon a cyclically adjusted deficit constraint is moretransparent and less susceptible to manipulation.Turrini (2008) found that pro-cyclical bias in goodtimes is an entirely expenditure-driven phenom-enon in the euro area and expenditure rules canbe helpful to curb the expansionary bias of fiscalpolicy. Holm-Hadulla, Hauptmeier and Rother(2012) confirmed that expenditure rules reducepro-cyclical bias. Based on literature surveys, Fab-rizio and Mody (2008) and Darvas and Kostyleva(2011) ranked expenditure rules the best amongthe various fiscal rules when designing fiscal insti-tution quality indices. Ayuso-i-Casals (2012) sum-marised many positive features of expenditurerules. Model simulations for Germany led Brückand Zwiener (2006) to propose the replacementof the SGP deficit rule with an expenditure ruleaugmented by medium-term debt targets. Morerecently, Andrle et al (2015) proposed a similarsetup, supported by literature review and modelsimulations.

Our proposed expenditure rule is similar in spiritto rules suggested in some of the above-men-tioned works, but has certain specific features thatwe regard as important. The rule would put a limiton the growth rate of an adjusted measure of gov-ernment expenditure. Table 1 on the next pagecompares our proposed new rule to the existingEU expenditure rule.

1 The adjusted expenditure aggregate: nominalexpenditure excluding interest expenditure,labour-market related expenditure and one-offexpenditure, while public investment expendi-ture should be smoothed over several yearsand accounted for in the same way that corpo-rate investment is accounted for.

‘The best option, re-designing the fiscal framework from scratch, is unrealistic today and

therefore we propose a better fiscal rule and the establishment of a European Fiscal Council to

oversee the new framework.’

10


Motivation:

• The current EU expenditure rule is based on theforecast GDP deflator, which is burdened withforecast uncertainty. Nominal expenditures areunder the direct control of the government butthe GDP deflator is not.

• The current EU expenditure rule disregards“non-discretionary changes in unemploymentbenefit expenditure”, which is problematicgiven estimation problems21. Excluding allunemployment-related expenditures would notlead to major moral hazard issues, because itis unlikely that a government will adopt meas-ures to increase unemployment just to be ableto spend more on unemployment benefits22.

• Sometimes large one-off expenditure becomesdesirable, such as a bank bailout, repair ofpublic infrastructure after a natural disaster ora one-off discretionary fiscal stimulus in a deeprecession. The decision on what can be quali-fied as a one-time expense should be dele-gated to an appropriate fiscal council – an issuewe discuss later.

• The current EU expenditure rule allows publicinvestment expenditure to be averaged overfour years, in order to reduce the impact of apossibly large investment in a given year onother expenditure. While this is helpful, our pro-posal goes further and suggests treating publicinvestment as corporate investment is treatedin corporate accounting; that is, the cost of aninvestment is distributed over future yearsduring the service life of investment. Forimproved transparency and increased effi-

Table 1: Comparison of the current EU expenditure rule with our proposed expenditure rule

Current EU expenditure rule Our proposed expenditure ruleExpenditure aggregate Real (depends on GDP deflator forecast) Nominal

Items excluded from theexpenditure aggregate

Interest, EU funded programmes,non-discretionary changes in

unemployment benefit expenditure

Interest, all labour-market relatedexpenditure, one-offs

Treatment of public expenditureFour-year backward-looking moving

averageAs in corporate accounting; separate

current and investment budgets

Expenditure growth benchmarkReal medium-term potential GDP

growthReal medium-term potential GDP

growth + 2% for inflation targetRevenue correction Yes YesDebt correction No YesExpenditure-overrun correction No YesSource: Bruegel.

ciency of public investment, it would be impor-tant to separate the investment budget fromthe current budget and to manage publicassets in a transparent holding company.

• The current EU expenditure rule excludes“expenditure on Union programmes fullymatched by Union funds revenue”, but suchspecial treatment of EU-funded investments isnot needed when all public investment isaccounted for as we propose.

2 The benchmark for expenditure growth:medium-term potential growth rate plus thecentral bank’s inflation target (2 percent in theeuro area and those other EU countries thathave this target, and 2.5 or 3 percent in thecase of some central European memberstates23). Euro-area countries subject to theBalassa-Samuelson effect may add a higherinflation rate.

Motivation:

• In some non-EU countries expenditure rulesdefine the ratio of expenditure to GDP, but thatintroduces some pro-cyclicality, because itallows more spending when output is abovepotential and less spending when it is below.This pro-cyclicality can be avoided by the useof potential output, at the cost of estimationerror related to medium-term potential growth.This estimation error, however, is not so large(Figure 2).

• The current EU expenditure rule (nominalexpenditure deflated by the forecast GDP defla-

21. We note that Bénassy-Quéré, Ragot and Wolff

(2016) suggested toexclude the incremental

increase in unemploymentpayments from the budget

deficit, when in a recessionthe European Commission’s

advisory Fiscal Board con-cludes that the situation

justifies such an exclusion.

22. It is true that the gen-erosity of unemployment

insurance can be increasedtoo (eg before an election to

gain popularity amongunemployed), yet in our

view the possible extraexpenditure under such

behaviour tends to be smallcompared to overall unem-

ployment expenditures.

23. The inflation target is2.5 percent in Poland andRomania and 3 percent in

Hungary.

11


tor) is subject to measurement and forecasterrors and takes expected inflation as given.Instead, we propose to add the central bankinflation target to the growth rate of potentialoutput24, which also helps the central bank toachieve its inflation target, while providing fur-ther cyclical stabilisation: when inflation ishigh, fiscal policy helps to reduce it, and wheninflation is low, fiscal policy supports the econ-omy and the return of inflation to its target.

• The convergence of the price level concurrentwith the convergence in productivity (the Bal-assa-Samuelson effect) is an equilibrium phe-nomenon, which is relevant for convergingcountries. Non-euro area countries are able toset their own inflation targets to reflect theimportance of this effect, but this is not thecase for euro-area countries. We therefore pro-pose the European Fiscal Council (see later) toallow a higher than 2 percent inflation rate forthose euro-area countries which are subject tothe Balassa-Samuelsson effect. Since thesecountries are small, their impact on the averageeuro-area inflation rate is minor.

• The Commission’s methodology for estimatingmedium-term potential output growth was sub-ject to major revisions at the height of the crisisand therefore should be improved by incorpo-rating open-economy considerations, as sug-gested in Darvas and Simon (2015).

3 Debt correction: the allowed maximum expen-diture growth is reduced by 0.02 times the dif-ference between the debt level in the previousyear and the 60 percent of GDP debt criterion.

Motivation:

• While gross public debt is not the best indica-tor of public sector sustainability risks, it is stilla useful and widely-used indicator, which canserve as a long-term anchor of fiscal policy.

• The 60 percent of GDP criterion for public debtis included in the Protocol of the EU Treaty.While this number does not have an academicunderpinning, the academic literature on theoptimal level of debt is inconclusive. There aresome advanced countries with debt levelsbelow 60 percent (eg Switzerland, Australia,New Zealand), where fiscal policy seems tooperate relatively well. We therefore accept the

60 percent criterion as the political choice of EUleaders.

• The exact value of the parameter of the debtcorrection term should be open to discussion.However, a back-of-the-envelope calculationsuggests that a 0.02 coefficient is reasonable.For example, for a country with a 110 percentof GDP debt ratio, this coefficient implies thatexpenditure growth should be 1.0 percentagepoints lower than the sum of potential growthand the inflation target. Since public expendi-ture typically amounts to about half of GDP, a 1percentage point slower expenditure growthimplies a fiscal tightening of about half percentof GDP, which is similar to the benchmark struc-tural balance adjustment requirement in thecurrent EU framework. We find this magnitudereasonable for a country with a 110 percentdebt ratio.

• Debt correction is included in our proposal in asymmetric way: in accordance with the EUTreaty, governments with public debt below 60percent of GDP should be allowed to increasetheir debt towards that level. However, theexpenditure growth limit resulting from our pro-posed rule represents a limit and not a target.Any government can opt for lower expendituregrowth if it prefers to have a public debt ratiobelow the 60 percent criterion.

4 Overrun correction: the difference betweenactual expenditure growth and the expendituregrowth limit should be corrected in subsequentyears if the gap was positive, while it can be cor-rected if the gap was negative.

Motivation:

• Even though nominal expenditures (excludinginterest and unemployment-related payments)are under the control of the government, over-runs are possible, which necessitates a latercorrection mechanism, while correcting anactually more-restrictive expenditure growthwould not endanger public debt sustainability.

• The debt correction mechanism does not makean overrun correction redundant, because debtcorrection works slowly. For example, a 1 per-centage point excess expenditure growth in thelast year would imply an approximately halfpercentage of GDP higher public debt (when

24. Brück and Zwiener(2006) also proposed to

consider the ECB’s inflationtarget in the definition of the

expenditure limit.

12


expenditure amounts to half of GDP), whichwould necessitate only a 0.01 percentage pointslower expenditure growth in the next year.

5 Consideration of revenues: a permanentincrease in the level of spending is allowed onlyif appropriate revenue measures are intro-duced; conversely, a cut in taxes is allowedonly if the expenditure level is cut too.

Motivation:

• A government might prefer to spend more,especially when a new government is formedafter an election, given the mandate the gov-ernment received. Yet long-term sustainabilityrequires that a permanent increase in expendi-tures should be compensated by increasedrevenues.

• Conversely, we propose to allow tax cuts only ifthey are matched by an appropriate reductionin expenditure growth.

Thereby, our proposed rule would be conducive tofiscal stabilisation through both expenditures (viathe inflation target, unemployment payments andpublic expenditures) and revenues (revenue-based automatic stabilisers are allowed to workfully)25. It would also be conducive to public debtsustainability, because of the incorporation ofexplicit debt correction and the elimination of thepro-cyclical bias in expenditure during good times,while limiting hysteresis effects in bad times.Implementation of our proposed rule would bemuch easier than the implementation of the cur-rent web of EU fiscal rules with all flexibilityclauses, given that nominal expenditure is underthe control of the government and the real-timeestimation and measurement errors in the expen-diture limit is much smaller than in the case of thestructural balance indicator. The simplicity andincreased transparency of the rule would alloweasier surveillance and enforcement and muchbetter communication with the general public.

Given the benefits of medium-term budgeting26

(such as better allocation of expenditures, avoid-ance of the negative effects of current expendituredecisions on future expenditure27, greater pre-dictability and transparency, increased account-ability of policymakers and higher effectiveness

in stabilisation terms), our rule should also be setin a multiannual budgeting framework.

As an illustration, we simulated the real-time work-ing of our proposed rule for some EU countries in2004-15. We cannot fully mimic our rule, becausewe do not have data on discretionary revenuechanges and also do not have sufficient informa-tion to smooth public investment. We thereforecalculated the growth rate of nominal publicexpenditures excluding interest expenditure,labour-market related expenditure, and one-offexpenditure, but make no correction for revenuesand public investment. We compare expendituregrowth to the real-time estimate of potentialoutput growth using the Darvas and Simon (2015)model. For simplicity, we do not consider theexpenditure-overrun correction.

In the pre-crisis period, our proposed expenditurerule would have disciplined Spain, Ireland and theUnited Kingdom (Figure 4 on the next page), coun-tries that experienced housing booms and rapidpro-cyclical public expenditure increases. It wouldhave disciplined Italy too, where public debt washigh. On the contrary, Germany and Sweden couldhave spent more in 2004-07. After 2009, our rulewould have allowed much more countercyclicalfiscal policies than those that were actually imple-mented in many EU countries. The growth rate ofpublic expenditure was inferior to our limit in Ger-many, Ireland, Spain and the United Kingdom,while the setback in the public expenditure growthrate in Italy in 2010 was justified, given its lowmedium-term potential growth estimate and theincreased level of public debt (for other countries,see the Annex).

The adoption of our proposed rule would not solvedirectly the problem of the non-credibility ofsanctions that has been present in the Europeanfiscal surveillance framework since the adoptionof the Maastricht Treaty. However, we believe thatour proposed rule, which is simple, easy toimplement in real time and not prone to significanterrors, could lead to sound fiscal policyrecommendations. Thereby, there would bestronger incentives for countries to abide by therules. Ultimately, countries should not – and willnot – observe the rules because they fearsanctions or because of peer pressure, but

25. We share the opinion ofButi and Gaspar (2015) that

budget-neutral automaticstabilisers should play a

large role in fiscalstabilisation.

26. See for instanceTarschys (2003).

27. A typical example of thisissue is the decision to end

a vaccination programmetoday to meet a fiscal target

that will eventually lead tohigher expenditure in termsof health care in the future.

013


-8

-4

0

4

8

12

04 06 08 10 12 14-8

-4

0

4

8

12

04 06 08 10 12 14-8

-4

0

4

8

12

04 06 08 10 12 14

-8

-4

0

4

8

12

04 06 08 10 12 14-8

-4

0

4

8

12

04 06 08 10 12 14-8

-4

0

4

8

12

04 06 08 10 12 14

Nominal expenditure growth (excluding interest, labour market and one-off expenditure)Real-time medium-term potential growth estimate plus 2%Real-time expenditure limit estimate with debt correction

GermanyGermanyGermany IrelandIrelandIreland ItalyItalyItaly

SpainSpainSpain SwedenSwedenSweden United KingdomUnited KingdomUnited Kingdom

Figure 4: Actual expenditure growth and real-time expenditure limit estimate based on our proposedrule, selected countries

Source: Bruegel. Note: Nominal public expenditure excluding interest expenditure, labour-market related expenditure and one-off expenditure, but no correction is made for revenues and public investment. The real-time estimate of potential output growthuses the Darvas and Simon (2015) model. The expenditure limit corrects the real-time potential growth estimate plus 2 percentinflation benchmark with the real-time data on public debt, but for simplicity we do not consider expenditure-overrun correction.

because they all agree that the rule represents thebest guidance for their fiscal policies to be bothsustainable and countercyclical.

Comparison of the existing structural balancerules with our proposed expenditure rule

There are a number of reasons to conclude thatour proposed rule is superior to the existing struc-tural balance rules.

Table 2: Comparison of the current EU structural balance rules with our proposed expenditure rule

Structural balance rule Our proposed expenditure rule

Operational targetStructural balance (not under

government control)Adjusted nominal expenditure

(under government control)Role of forecasts GDP and inflation forecasts matter a lot Forecasts do not matter much

Estimation errorLarge (output gap in a given year, elas-ticity of budget balance to output gap)

Small (multi-year average ofpotential growth)

Quantification of one-offs Yes Yes

Counter-cyclicality Good in theory, bad in practiceGood in theory, good prospect for

practice

Debt sustainability Good in theory, dubious in practiceGood in theory, good prospect for

practice

Source: Bruegel.

The estimated structural balance depends on theestimates (forecasts) of the budget balance andoutput gap, on the elasticity of the cyclicallyadjusted balance to the output gap, and on thequantification of one-off revenue and expendituremeasures. For our expenditure rule, estimates(forecasts) of the adjusted expenditure aggregate,estimates of the medium-term potential outputgrowth and quantification of discretionary revenuemeasures are needed.

014


• For a government it is easier to control theadjusted expenditure aggregate that we pro-posed than the budget balance, since the latterdepends on unemployment expenditure andrevenues too (which in turn strongly depend onthe state of the economy), and on interestexpenditure (which might be subject tochanges in market sentiment).

• Fiscal planning under a structural balance rulevery much depends on forecasts of output andinflation, while such dependence is not soimportant for the implementation of the expen-diture rule.

• Irrespective of which potential output methodis used, the estimation error and the expectedrevision is greater in the output gap estimatefor a given year (which is needed for the struc-tural balance estimate of a given year) than fora medium-term average of potential growthestimates (which is needed to set the limit onexpenditure growth). The medium-term aver-age of potential growth is calculated on thebasis of several years, eg the past five yearsand the current year. Even if the current-yearestimate might be subject to a sizeable revi-sion, experience shows that the past potentialgrowth estimate is only subject to small revi-sion. Figures 1 and 2 demonstrate the past per-formance of real-time estimates and underlinethat the medium-term potential growth esti-

mate is subject to smaller errors.• Estimating the elasticity of the cyclically

adjusted balance to the output gap is neededfor the structural balance rule, but not neededfor the expenditure rule. Thereby, only thestructural balance rule is burdened with thisestimation error.

• The quantification of one-off revenue andexpenditure measures (structural balance rule)and discretionary revenue measures (expen-diture rule) is similarly difficult in our view, andtherefore there is no clear ranking between thetwo rules in this aspect.

In fact, the measurement problems concerningstructural budget balances would have made thecurrent smarter rules useless for Spain in theyears preceding the crisis: real-time data from theEuropean Commission and IMF suggests thatSpain would have been compliant with the struc-tural balance rules (Figure 5).

We also checked a quasi-real-time estimate of thestructural balance using the potential outputmethod of Darvas and Simon (2015). To this end,for each year, we calculated the implied elasticityof the difference between the real-time actual andcyclically adjusted budget balance to the outputgap as estimated by the European Commission,and applied this elasticity to the real-time output

-1 2

-1 0

-8

-6

-4

-2

0

2

2005 2010 2015-1 6

-1 2

-8

-4

0

4

2005 2010 2015

Real-time budget balance

Real-time structural budget balance: ECFIN

Real-time structural budget balance: Darvas-Simon

Real-time structural budget balance: IMF

Spain Ireland

Figure 5: Spain and Ireland, real-time estimates of the actual budget balance and structural budgetbalance made in spring each year (% of GDP)

Source: Bruegel. Note: For each year, the real-time estimate for the given year made in the spring of that year is indicated. Egfor 2010, the spring 2010 estimate for 2010 is included, for 2011 the spring 2011 estimate for 2011 is included, etc. Structuralbalance estimates from the European Commission are available only from 2006 onwards, so for 2003-05, we show thecyclically adjusted balance estimates instead.

015


gap estimate of Darvas and Simon (2015). Wethen corrected the resulting cyclically adjustedbudget balance with the one-off estimates of theEuropean Commission to obtain a quasi-real-timeestimate of the structural balance using the poten-tial output method of Darvas and Simon (2015).The results are also reported in Figure 5, indicat-ing that Spain would have complied with the struc-tural deficit rules even when using the real-timeoutput gap estimates of Darvas and Simon(2015).

The consensus today, as argued for instance byMartin and Philippon (2014), is that Spanish fiscalpolicy was not countercyclical enough before2008 and that Spain should have entered thecrisis with an even lower debt-to-GDP level thanthe 35.5 percent ratio in 2007, which would havehelped the country dampen the unsustainableboom before the crisis and allowed the govern-ment to have more room for manoeuvre when thecrisis hit. Figure 4 shows that our expenditure rulewould have constrained Spain quite significantlyin the pre-crisis period, while we demonstratedabove that the current structural balance rulewould have not constrained Spain in 2000-08.

The conclusion for Ireland is broadly similar,though the real-time structural balance estimatebased on the Darvas and Simon (2015) outputgap model suggests that in 2004-06 the real-timestructural balance estimate was slightly worsethan the Fiscal Compact’s -1.0 percent minimumvalue for euro-area countries with debt below 60percent of GDP.

Transition

An appropriate transition period will be needed tomove from the current system of rules to our pro-posed new rule. Otherwise, the different startingpositions could imply similar expenditure growthlimits for countries that have similar debt levelsand potential growth rates, but very differentbudget deficits even though they have similarcyclical situations. We again would recommend asimple transition rule: for countries with budgetdeficits over a certain threshold (eg 2 percent ofGDP), the expenditure growth limit is reduced by0.5 percentage points per year until the thresholdis reached. The threshold should be country-spe-

cific and should be calibrated, given country-spe-cific medium-term growth and expected interestrates, so that if public debt was at 60 percent ofGDP, it would stay at this level if the expenditurerule is followed. After this transition period is com-pleted, two countries with similar potential growthrates and public debt levels will have significantlydifferent budget balances only if they facemarkedly different economic situations, such asa rapid boom (leading to a budget surplus) andrecession (leading to a deficit), in which case sim-ilar recommendations for two such countrieswould be justified.

Surveillance

To increase ownership of the rule by governmentsand parliaments, our proposed European ruleshould be transposed into national law and mon-itored at the national level by independentnational fiscal councils. These councils should beresponsible for validating the potential growthestimates used in the rule and for monitoring theconsistency of the government policies with therule during the drafting of the budget, during thebudget implementation and also after the fiscalyear is closed and the final numbers on the exe-cution of the budget are available.

Still, every possible rule, including our proposedrule, has limitations and we believe that discre-tionary decisions are needed to face special cir-cumstances. For example, in an exceptionallydeep recession, further fiscal stimulus beyondwhat is allowed by our proposed rule might be jus-tified, or a natural disaster might necessitateunusually large public investment. We proposethat such decisions be taken at the Europeanlevel, because of the potential cross-bordersexternalities. We see two options for the European-level involvement:

• The current setup involving the European Com-mission and the Council,

• Creation of a new European Fiscal Council.

Currently, the perception of some stakeholders isthat Commission does not always give unbiasedrecommendations to the Council. Moreover, Mody(2014) argues that the political process alwaysundermines the proper application of any fiscal

16


rule. Such perceptions and political difficultieswould likely be reduced if the EU’s fiscal frame-work were to eliminate the current opaque systemof exceptions and opt for the simple fiscal rule wepropose. However, in order to avoid any possibilityof political mismanagement of the discretionarypowers at the European level, a new EuropeanFiscal Council (EFC) should be set up, similar tothe EMU Stability Council proposal of von Hagen(2007)28. The mandate of the EFC should be tosafeguard the proper implementation of the fiscalrule with the ultimate objectives of long-termpublic debt sustainability and countercyclicalfiscal policy. In particular, the EFC should beentrusted with taking the discretionary decisionsconcerning the implementation of the Europeanexpenditure rule, such as:

• The occasions when the rule can be suspendedeither in a particular country or in the EU as awhole;

• Acceptable one-off measures;• The way investment expenditures should be

smoothed over several years;• Allowing higher inflation in countries charac-

terised by the Balassa-Samuelson effect.

The Commission’s 2015 decision29 about theestablishment of an independent advisory Euro-pean Fiscal Board (EFB) is not sufficient for thetask, given that the EFB was created to be an inter-nal advisory body for the Commission: the missionstatement specifies an advisory role, the nomina-tion of members depends almost entirely on theCommission, members are not accountable andthe transparency regulation requires only onepublic annual report30 by the EFB.

Instead, similar to the ECB Governing Council31, aEuropean Fiscal Council should be established,consisting of an executive board (six members)and the chairs of each EU member state fiscalcouncil. Given the importance of the EFC’s deci-sions, the required qualifications and appointmentprocedures of the executive board membersshould be as strict as those for ECB ExecutiveBoard members, in order to guarantee profes-sionalism and avoid political appointments andthe representation of national preferences32. TheEuropean Fiscal Council should be accountable toEuropean citizens.

Accountability of unelected officials should betwo-dimensional, as explained by Schedler(1999). First, an accountable board should beobliged to inform citizens and its representativesabout its decisions and should be able to justifythem. This could take the form of press confer-ences and hearings at the European Parliament ona regular basis, accompanied by the publicationof reports justifying its decisions. Second, theEuropean Parliament should be able to imposesanctions on the body in case it fails to fulfil itsmandate.

The aggregate fiscal stance of the euro area andthe EU

There is a debate on whether the aggregate fiscalstance of the euro area and the EU makes senseand if the fiscal framework should force countrieswith ample fiscal space to have larger budgetdeficits when other countries are forced toimplement pro-cyclical fiscal tightening. Forexample, Blanchard et al (2014) suggested afiscal expansion in countries with significant fiscalspace might be desirable from a euro-areaperspective in case of under-use of productivecapacities and too-low inflation, especially in aliquidity trap. But if these countries have a lessslack than the euro-area average and they believethat cross-country spillovers from fiscal policy aresmall, they will most likely provide less stimulusthan what would be needed to align the aggregatefiscal stance of the euro area with the aggregateeconomic situation.

Our proposed framework would ensure that eachmember state runs responsible fiscal policies ingood times and thereby have the fiscal space toprovide adequate fiscal stabilisation during badtimes. Thereby, the likelihood that some countrieswill be forced to implement pro-cyclical fiscaltightening during a downturn will be less likely too.For this reason, the aggregate fiscal policy of euro-area and EU member states will more aligned withthe aggregate economic situation of the euro areaand the EU, even though our new expenditure rulewould not try to tackle directly the problem of theaggregate fiscal stance.

We do not propose a rule which can force a coun-try to have a higher budget deficit than what is

28. Our proposal is quitesimilar to the proposals of

von Hagen (2007), with twokey differences: we suggest

including the chairs ofnational fiscal councils in

the EFC in addition to anexecutive board, while von

Hagen proposed only aprofessional executive

board, and we propose anEU-level council, while von

Hagen proposed a euro-area council.

29. See http://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32

015D1937&from=EN.

30. “Article 6. Transparency.The Board shall publish an

annual report of itsactivities, which shall

include summaries of itsadvice and evaluations

rendered to theCommission”.

31. The ‘Eurosystem ofFiscal Policy’ (EFP) proposal

by Sapir and Wolff (2016)was also motivated by the

ECB Governing Council. Akey difference between

their EFP and our EFCproposal is that financeministers would be the

members of the EFP, whilethe chairs of the national

fiscal councils would be themembers of the EFC. Thereare also differences in theproposed mandates of the

EFP and EFC.

32. As suggested inCalmfors and Wren-Lewis

(2011) fiscal councils couldbe composed of a mix of

academics, public-financeexperts, financial-sectoranalysts and even well-

known former politicians, aslong as they are not

influenced by the prospectsof future career

opportunities in thegovernment administration.

17


deemed appropriate domestically. In our view, itis unrealistic to expect that some countries willrun larger budget deficits (and consequently taxtheir citizens more) just because some othercountries do not have fiscal space and are forcedto implement pro-cyclical fiscal tightening.National policymakers are accountable to theirnational parliaments and focus on national inter-ests. If the euro-area or EU aggregate fiscal stanceis to be managed when some countries face fiscalconstraints in a recession, a centralised instru-ment, such as a European unemployment insur-ance scheme (ie an automatic mechanism) or aspecific investment facility (ie a discretionarymechanism), should be developed.

5 CONCLUDING REMARKS

The EU’s current fiscal framework is rather ineffi-cient. In theory, the new fiscal rules, with cyclicallyadjusted targets, flexibility clauses and the optionto enter into an excessive deficit procedure, allowfor large fiscal stabilisation during a recession,while they can also support the sustainability ofpublic debt. However, in practice, the implemen-tation of the rules is hindered by badly-measuredindicators and incorrect forecasts, which can leadto misleading policy recommendations. The largenumber of flexibility clauses makes the frame-work opaque and leads to never-ending bargain-ing between the countries that do not comply withthe rules and the European Commission, whichundermines trust in the rules. Compliance with thefiscal rules is low. Several politicians in countriesthat breach the rules regard the rules as inappro-priate, while other politicians in countries thatcomply with the rules worry that the rules are notenforced on their partners. Preserving this ineffi-cient fiscal framework would be suboptimal.

We recommend changing the EU fiscal framework.The first-best option, in our view, would requireredesigning the whole framework from scratch,which is unrealistic. We therefore make a proposalthat might be realistic even in the near term, bychanging the Stability and Growth Pact and theFiscal Compact. Our proposal would maintain anEU-wide fiscal rule with supranational surveil-lance. We propose to drop all rules related to thebadly-measured structural balance indicator and

adopt an expenditure rule with a debt-correctionmechanism, embodied in a multi-annual fiscalframework.

The expenditure rule should set a limit on thegrowth rate of nominal public expenditure exclud-ing interest, labour-market related and one-offexpenditure, while public investment expenditureshould be smoothed over several years andaccounted for in the same way as corporateinvestment. The limit should be specified as the(appropriately-measured) medium-term potentialgrowth rate of GDP plus the central bank’s inflationtarget, and should be corrected for deviations ofpublic debt from the 60 percent of GDP Maastrichtdebt criterion, discretionary revenue measuresand possible expenditure-overruns in previousyears. This European rule should be transposedinto national laws and monitored by national fiscalcouncils. We also propose to get rid of the opaqueweb of flexibility clauses in current fiscal rules.Instead, an independent European Fiscal Councilshould be set up with an appropriate mandate,appointment procedures and accountability, tooversee the system and exercise the necessarydiscretion in unusual times.

This overhauled framework would be simple,transparent, easy to monitor, easy to explain andwould involve a fiscal indicator that is under thedirect control of the government. It would be moreconducive than the current system to public debtsustainability and fiscal stabilisation, the two keyobjectives of a fiscal framework. The delegation ofthe discretionary power to an independent Euro-pean Fiscal Council would eliminate the percep-tion of a possibly improper or politically-motivatedapplication of the rule.

Enforcement of the rules at the European levelshould move away from the threat of financialsanctions, which is anyway not credible in the cur-rent framework. The political consequences of aneventual financial sanction could be highly nega-tive. The perception that the fiscal framework pro-vides economically-sound guidance would be amuch more important factor than the fear of sanc-tions, to give an incentive to countries to respectthe rules.

18

BR U EGE LPOLICYCONTRIBUTION A PROPOSAL TO REVIVE THE EUROPEAN FISCAL FRAMEWORKA PROPOSAL TO REVIVE THE EUROPEAN FISCAL FRAMEWORK

REFERENCES

Anderson, B. and J.J. Minarik (2006) ‘Design Choices for Fiscal Policy Rules’, OECD Journal onBudgeting 5(4):159-208.

Andrle, M., M.J.C. Bluedorn, L. Eyraud, M.T. Kinda, P.K. Brooks, M.G. Schwartz and A. Weber (2015)‘Reforming Fiscal Governance in the European Union’, Staff Discussion Notes No. 15/9, InternationalMonetary Fund

Ayuso-i-Casals, J. (2012) ‘National Expenditure Rules – Why, How and When’, European EconomicPapers 473, DG ECFIN, European Commission

Barbiero, F. and, Z. Darvas (2014) ‘In sickness and in health: protecting and supporting publicinvestment in Europe’, Policy Contribution 2014/02, Bruegel

Bénassy-Quéré, A., X. Ragot and G.B. Wolff (2016) ‘Which fiscal union for the euro area?’ PolicyContribution 2016/05, Bruegel

Blanchard, O.J., C.J. Erceg and J. Lindé (2015) ‘Jump-Starting the Euro Area Recovery: Would a RiseIn Core Fiscal Spending Help The Periphery?’ NBER Working Paper No. 21426

Blanchard, O.J. and D. Leigh (2013) ‘Growth forecast errors and fiscal multipliers’, No. w18779,National Bureau of Economic Research

Brück, T. and R. Zwiener (2006) ‘Fiscal policy rules for stabilisation and growth: A simulation analysisof deficit and expenditure targets in a monetary union’, Journal of Policy Modeling 28(4): 357-369

Buiter, W., G. Corsetti and N. Roubini (1993) ‘Excessive deficits: sense and nonsense in the Treaty ofMaastricht’, Economic Policy 8(16): 57-100

Buiter, W.H. (2006) ‘The ‘Sense and Nonsense of Maastricht’ Revisited: What Have we Learnt aboutStabilization in EMU?’ Journal of Common Market Studies 44(4): 687-710

Buti, M. and V. Gaspar (2015) ‘Designing Fiscal Policy for Steady, Enduring Growth’, VOXEU, availableat http://www.voxeu.org/article/fiscal-policy-enduring-growth

Calmfors, L. and S. Wren-Lewis (2011) ‘What should fiscal councils do?’ Economic Policy 26(68):649-695

Darvas, Z. and A. Simon (2015) ‘Filling the gap: open economy considerations for more reliablepotential output estimates’, Working Paper 2015/11, Bruegel

Darvas, Z. and V. Kostyleva (2011) ‘The fiscal and monetary institutions in Central, Eastern andSouth-Eastern European countries’, OECD Journal on Budgeting 11(1): 147–185

Darvas, Z. and Á. Leandro (2015) ‘The limitations of policy coordination in the euro area under theEuropean Semester’, Policy Contribution 2015/19, Bruegel

European Commission (2008) ‘A European Economic Recovery Plan’, Communication from theCommission to the European Council, 11 November

European Commission (2016) Fiscal Sustainability Report 2015

Fabrizio, S. and A. Mody (2008) ‘Breaking the impediments to budgetary reforms: evidence fromEurope’, Working Paper No. 08/82, International Monetary Fund, Washington DC

Holland, D. and J. Portes (2012) ‘Self-defeating austerity?’ National Institute Economic Review,222(1), F4

Holm-Hadulla, F., S. Hauptmeier and P. Rother (2012) ‘The impact of expenditure rules on budgetarydiscipline over the cycle’, Applied Economics44(25): 3287-3296

Juncker J.C., D. Tusk, J. Dijsselbloem, M. Draghi, M. Schulz (2015) The Five President's Report:Completing Europe’s Economic and Monetary Union, 22 June

Manesse, P. (2014) ‘Time to scrap the Stability and Growth Pact’, VOXEU, available athttp://www.voxeu.org/article/time-scrap-stability-and-growth-pact

19


Martin, P. and T. Philippon (2014) ‘Inspecting the mechanism: leverage and the Great Recession inthe Eurozone’, NBER Working Paper No. 20572

Marzinotto, B. and A. Sapir (2012) ‘Fiscal rules: timing is everything’, Policy Brief 2012/03, Bruegel

Merler, S. and J. Pisani-Ferry (2012) ‘The simple macroeconomics of North and South in EMU’,Working Paper 2012/12, Bruegel

Micossi, S. and F. Peirce (2014) ‘Flexibility clauses in the Stability and Growth Pact: No need forrevision’, CEPS Policy Briefs 319

Mody, A. (2014) ‘Are the Eurozone’s fiscal rules dying?’ Bruegel Blog, 29 October, available athttp://bruegel.org/2014/10/are-the-eurozones-fiscal-rules-dying/

Ódor, L. and G. P. Kiss (2015) ‘Lost in complexity: building a new fiscal framework for Europe’, mimeo

Pisani-Ferry, J. (2002) Fiscal Discipline and Policy Coordination in the Eurozone: Assessment andProposals, report prepared for the European Commission President’s Group of Economic Analysis,available at http://ec.europa.eu/dgs/policy_advisers/archives/experts_groups/docs/pisani_ferry.pdf

Portes, J. and S. Wren-Lewis (2014) ‘Issues in the design of fiscal policy rules’, University of OxfordDepartment of Economics Discussion Paper Series No. 704

Schedler, A. (1999) ‘Conceptualizing accountability’, chapter in Schedler, A., L. Diamond and M. F.Plattner, The self-restraining state: Power and accountability in new democracies, Lynne RiennerPublishers

Summers, L. (2012) ‘Fiscal policy in a depressed economy’, Brookings Papers on Economic Activity,44(1): 233-297

Tarschys, D. (2003) ‘Time horizons in budgeting’, OECD Journal on Budgeting, 2(2): 77-103

Tirole, J. (2012) ‘The euro crisis: some reflexions on institutional reform’, Financial Stability Review16: 225-242

Turrini, A. (2008) ‘Fiscal policy and the cycle in the euro area: the role of government revenue andexpenditure’, European Economic Papers 323, DG ECFIN, European Commission

Von Hagen, J. (2007) ‘From the Stability and Growth Pact to a Sustainability Council for EMU’, inBerger, H. and T. Moutos (eds) Designing the New European Union, Amsterdam: Elsevier

Wren‐Lewis, S. (2013) ‘Macroeconomic Stabilisation in the Eurozone: Lessons from Failure’, GlobalPolicy 4(s1): 66-73

20


ANNEX

Figure A1: Actual expenditure growth and real-time expenditure limit estimate based on ourproposed rule, selected countries

Source: Bruegel. Note: see explanations in the note to Figure 4. Euro area 11 is the aggregate of the first twelve memberstates of the euro area excluding Luxembourg (because of data limitations).

0

1

2

3

4

5

6

7

04 06 08 10 12 140

1

2

3

4

5

6

7

8

9

04 06 08 10 12 140

1

2

3

4

5

6

7

04 06 08 10 12 14

1

2

3

4

5

6

7

8

04 06 08 10 12 141.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

04 06 08 10 12 14-12

-8

-4

0

4

8

12

04 06 08 10 12 14

-2

0

2

4

6

8

10

04 06 08 10 12 14-12

-8

-4

0

4

8

04 06 08 10 12 140

1

2

3

4

5

6

04 06 08 10 12 14

Austria Belgium Denmark

Finland France Greece

Netherlands Portugal Euro area 11

Nominal expenditure growth (excluding interest, labour market and one-off expenditure)Real-time medium-term potential growth estimate plus 2%Real-time expenditure limit estimate with debt correction

CONTRIBUTION ISSUE 2016/07 A PROPOSAL TO -...

Documents

Transcript of CONTRIBUTION ISSUE 2016/07 A PROPOSAL TO -...