International Experience with Fiscal Redistribution · International Experience with Fiscal...
Transcript of International Experience with Fiscal Redistribution · International Experience with Fiscal...
International Experience with Fiscal Redistribution
SANJEEV GUPTADeputy Director
Fiscal Affairs DepartmentInternational Monetary Fund
ColomboJuly 14, 2016
Outline of the presentation
Trends in Inequality and the Redistributive Role of Fiscal PolicyTrends in Inequality and the Redistributive Role of Fiscal Policy
Lessons for Policy Design:
Tax PoliciesTax Policies
Expenditure Policies
Lessons from Case Studies
Fiscal Consolidation and Income Distribution
Conclusions
2
The presentation builds on the findings included in the recent IMF book
http://www.imfbookstore.org/ProdDetails.asp?ID=IRFPEA&PG=1&Type=BL
3
Trends in Inequality and e ds equa ty a d the Redistributive Role of
Fi l P liFiscal Policy
While world income inequality has been going down …
Evolution of World Inequality, 1820-201080
140
160
6
70
75
80
Historical series
Recent period
1990 1995
2000 2005
100
120
55
60
65
Gap
cen
t (G
ini)
Gini coefficient
2000
2010
2000
1990
40
60
80
40
45
50
Per
Mean income gap between top and bottom 10%2010
2005
20001995
20
40
30
35
5
Sources: Bourguignon, The Globalization of Inequality. The indexes of PPP that Angus Maddison used for the historical data referenced the year 1990. The data for the recent period use PPP data based on price statistics that were collected in 2005, which sometimes resulted in significant revisions to the parity indexes. This explains much of the discontinuity between the two series in 1990.
… within countries inequality has been increasing in most d d i d i l i advanced economies and in large emerging ones
Income Inequality in the 1980s and 2000sAdvanced Economies Developing Economies
HondurasJamaica
South AfricaBotswana
C t l Af i
0.6
0.65
Advanced Economies Developing Economies
U it d St t
0.4
0.45
Malaysia
Brazil
Chile
Colombia
Costa RicaDominican Republic
Mexico
Panama
ParaguayCentral African
Republic
Kenya
Lesotho
Mozambique
Nigeria
Rwanda
Swaziland
Zambia
0 45
0.5
0.55
i In
dex
) la
testUnited States
United Kingdom
France
Italy
CanadaJapan
Greece
Portugal
SpainAustraliaNew Zealand
Cyprus
Israel
Taiwan0 3
0.35
nd
ex)
late
st
Cambodia
Sri Lanka
I diIndonesia
Laos
Philippines
Thailand
Vietnam
China
Mongolia TurkeyAlbania
Georgia
Kyrgyzstan
TurkmenistanUzbekistan
ArgentinaBolivia
EcuadorEl Salvador
Mexico
NicaraguaPeru
Uruguay
Venezuela
Iran
MauritaniaMorocco Tunisia
Cameroon
Ghana
Guinea
Cote d`Ivoire Madagascar
Niger
SenegalTanzania
Uganda
Burkina Faso
0.35
0.4
0.45
neq
ual
ity
(Gin
Austria
BelgiumDenmark
GermanyLuxembourg
NorwaySweden
SwitzerlandFinland
IrelandCyprusTaiwan
Czech RepublicSlovak Republic
Slovenia
0.25
0.3
qu
alit
y (G
ini
In
Bangladesh IndiaNepal
Armenia
Azerbaijan
Kazakhstan
Kyrgyzstan
Moldova
Hungary
Poland
Jordan
Egypt
Pakistan
Burundi
Ethiopia
Mali
g
0.25
0.3
0 25 0 35 0 45 0 55 0 65
In
Netherlands
0.15
0.2
0 15 0 2 0 25 0 3 0 35 0 4 0 45
Ineq
0.25 0.35 0.45 0.55 0.65Inequality (Gini Index) 1980s
6
0.15 0.2 0.25 0.3 0.35 0.4 0.45Inequality (Gini Index) 1980s
Wealth inequality is significantly more pronounced than income inequality
Inequality of Wealth and Income0 9
0 6
0.7
0.8
0.9
Wealth Gini; Average = 0.68Disposable Income Gini; Average = 0.36
0.4
0.5
0.6
0.1
0.2
0.3
Sources: Davies and others (2008); Luxembourg Income Study; Organisation for Economic Co-operation and Development; Socio-Economic
0.0
JPN
CH
N
ES
P
KO
R
ITA
AU
S
NLD
TW
N
BG
D
GE
R
IND
VN
M
CA
N
GB
R
PA
K
RU
S
TH
A
TU
R
FR
A
NG
A
AR
G
ME
X
IDN
BR
A
US
A
CH
E
7
( ) g y g p p
Database for Latin America and the Caribbean (CEDLAC and the World Bank).
Note: Data labels in the figure use International Organization for Standardization (ISO) country codes.
Inequality and inequality of opportunities tend
ITAGBR
USA
0.5
→)
to be related
FRAESP
CHEUSA
0.4
ss m
ob
ility
→
GERJPN
NZL0.3
s el
asti
city
(le
s
AUS
CANFIN
SWE
0.2
tio
nal
ear
nin
gs
DNK
FINNOR
y = 0.0251x - 0.3709
0.120 25 30 35
Gen
erat
20 25 30 35
Gini (around 1985; higher inequality→)
8
Distributive concerns—to a varying degree–are addressed in IMF advice
► Distributive concerns in IMF policy advice:p y
► Surveillance of economic policies
► Lending
C it b ildi► Capacity building
► First began to crop up in IMF-supported programs in the late 1980s and work intensified in the wake of the global financial crisis
► Growing recognition that inequality may have social consequences, potentially undermining g g y y y gmacroeconomic stability and sustainable growth
► Recent examples where these issues have been addressed in Fund Surveillance: Ethiopia, Bolivia and ► ece t e a p es e e t ese ssues a e bee add essed u d Su e a ce t op a, o a a dKyrgyz Republic
9
Fiscal redistribution reduced inequality by one third in d d i tl th h di
0.30
Average market income Gini: 0.43
advanced economies, mostly through spending
0 20
0.25
on
From taxesFrom transfers
Average market income Gini: 0.43Average disposable income Gini: 0.29
0.15
0.20
e G
ini r
edu
ctio
Total redistribution: 0.14
0.10
Ab
solu
te
From transfers: 0.09
0.00
0.05
NK
ZE EL
VN
OR
BR IN UT
WE
UX
EU
RA
US
RL
LD SR
AN
SA
ST
TA RC
SP
OR
WN
DN
CZ
BE
SV
NO
GB F AU
SW LU DE
FR
AU IR NL IS CA
US
ES IT
GR ES
KO
TW
10
The contribution of fiscal policy to reduce inequality is smaller in developing countriesdeveloping countries
Level and Composition of Tax Revenues and Social Spending , 2010
35
40
45 EducationHealthSocial protectionTax revenuesTotal revenues
20
25
30
ent
of
GD
P
5
10
15
20
Per
ce
0
5
Advanced{30; 29}
Emerging Europe{14; 16}
Latin America and the
Caribbean
Middle East and North
Africa
Asia and Pacific{23; 19}
Sub-Saharan Africa
{34; 35}
11
Source: IMF database.
Note: Numbers in brackets refer to number of countries in the {tax; spending} country samples.
{14; 16} Caribbean{23; 21}
Africa{19; 18}
{23; 19} {34; 35}
Lessons for Policy Design
Designing efficient redistributive fiscal policy
► Redistributive fiscal policy should be consistent with macroeconomic objectives► Redistributive fiscal policy should be consistent with macroeconomic objectives
► Policies need to be carefully designed taking into account also indirect and medium-long term effectsg
► Design should take into account administrative capacity
► The impact of tax and spending policies should be evaluated jointly► The impact of tax and spending policies should be evaluated jointly
13
Efficient design requires looking at tax and spending together
India
Si l t d N t I t b D il f C bi d T d T f R fSimulated Net Impact by Decile of Combined Tax and Transfer Reforms
2.5
pti
on
0 5
0.5
1.5
ho
ld C
on
sum
p
-2 5
-1.5
-0.5
cen
t o
f H
ou
seh
-3.5
2.5
1 2 3 4 5 6 7 8 9 10
Per
c
Deciles
14
Source: IMF staff estimates based on the 2011/12 Indian National Sample Survey and the Employment Unemployment Survey
Deciles
Tax policy: reform options to achieve more efficient redistribution of taxation
Direct taxes Advanced Developing
✓ ✓Implement progressive personal income tax ✓
✓
✓(PIT) rate structures
Relieve low-wage earners from tax or social contributions
✓ ✓
✓Expand coverage of PIT
More effectively tax multinational corporations
Utilize better opportunities for recurrent property taxation
✓ ✓
15
Tax policy: reform options to achieve more efficient redistribution of taxation
Indirect taxes Advanced DevelopingIndirect taxes Advanced Developing
✓ ✓Minimize VAT exemptions and special VAT rates
✓Use specific excises mainly for purposes other than redistribution
✓
16
Expenditure policy: reform options to achieve more efficient redistribution of spending
Education Advanced Developing
✓ ✓Improve access to education of low-income ✓ ✓Improve access to education of low-income families
Health Advanced Developing
✓Expand coverage of publicly financed basic health package
✓
health package
Ensure or maintain access of low-income groups to essential health services
✓
17
Expenditure policy: reform options to achieve more efficient redistribution of spending
Social transfers Advanced Developing
✓
✓
Intensify the use of active labor market programs and in-work benefits
Expand conditional cash transfer (CCT) ✓
✓
programs as administrative capacity improves
Increase effective pension retirement age while protecting low-income pensioners
✓
✓Expand noncontributory means-tested social pensions
18
Other lessons
Heterogeneity among countries must be taken into account when designingHeterogeneity among countries must be taken into account when designing fiscal policies: income levels and pre-tax & transfers levels of inequality differ significantly among countries
administrative capacity may constrain policy options (for instance for ensuring tax compliance or effective targeting)
institutional design varies among countries and may have an effect on distributional outcomes institutional design varies among countries and may have an effect on distributional outcomes (for instance decentralization/centralization of certain functions)
19
Fiscal consolidation and redistributionUnemployment Rates and Gini Coefficients During Fiscal Adjustment
(percent)1030.0
929.5
Disposable income Gini coefficient
Unemployment rate (right scale)
829.0
728.5
Sources: Solt (2014); Eurostat; and IMF World Economic Outlook
628.0Before consolidation During consolidation
20
Sources: Solt (2014); Eurostat; and IMF, World Economic Outlook .
Note: Fiscal adjustment episodes are defined as in Escolano and others 2014 based on changes in cyclically adjusted primary balances in countries
with positive primary gaps. The sample covers 91 episodes across 49 advanced and developing economies between 1945 and 2012.
Fiscal consolidation and redistributionChanges in Market- and Disposable-Income Gini Coefficients, 2007-13
0.08
0.02
0.04
0.06
-0.02
0.00
-0.06
-0.04
PR
TLV
AR
OM
GR
CN
LDG
BR
FIN
BE
LLU
XIT
AS
VN
IRL
CZ
EB
GR
AU
TD
NK
ES
TC
YP
PO
LS
WE
DE
UM
LTE
SP
SV
KF
RA
HU
NLT
U
Disposable Market
Source: EUROMOD statistics on Distribution and Decomposition of Disposable Income, accessed at http://www.iser.essex.ac.uk/euromod/statistics/ using EUROMOD version No. G2.0.
Note: An increase in the Gini coefficient indicates an increase in inequality. The Gini coefficient for market income is estimated based on disposable-income micro data by adding back (in the case of taxes) or deducting (in the case of benefits) each income component, using the EUROMOD microsimulation model. Estimates for market-income Gini in 2007 and 2013 are based on European Union Statistics on Income and Living Conditions 2008 (income reference period: 2007) and European Union Statistics on Income and Living Conditions 2010 (income reference period: 2009), respectively. For the latter, market-income updates from the income reference period to later years are based on a combination of updating factors For more information on the exact updating factors used for each country please refer to the Country
P R G N G B L S C B A D E C P S D M E S F H L
21
updates from the income reference period to later years are based on a combination of updating factors. For more information on the exact updating factors used for each country, please refer to the Country Reports (https://www.iser.essex.ac.uk/euromod/resources-for-euromod-users/country-reports). Changes between years and tax-benefit components are not necessarily statistically significant. Data labels in the figure use International Organization for Standardization (ISO) country codes.
Conclusions
Conclusions
Fiscal policy is a powerful and adaptable tool for achieving distributionalFiscal policy is a powerful and adaptable tool for achieving distributional objectives
Improving both distributional outcomes and efficiency is possible
Considering taxes and spending programs together enhances the effectiveness of fiscal redistribution
23