Fiscal Policy, Inequality and the
Poor in the Developing World
Nora Lustig
Tulane University
Nonresident Fellow CGD and IAD
National Treasury and World Bank Workshop
Fiscal Policy and Redistribution in an Unequal Society
Pretoria, South Africa, November 5, 2014
When using material in this ppt please
cite as:
• Lustig, Nora. 2014. “Fiscal Policy, Inequality and the Poor in
the Developing World. Round 1.” CEQ Working Paper No. 23,
Center for Inter-American Policy and Research and
Department of Economics, Tulane University and InterAmerican Dialogue, forthcoming.
www.commitmentoequity.org
2
CEQ Teams
(Year of Survey; C=consumption & I=income)(MWB Version)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
10
11
12
13
14
15
16
17
Argentina (2009, I): Nora Lustig and Carola Pessino (CEQ Web Dec 2013) Public Finance Review, May 2014,
Volume 42, Issue 3
Armenia (2011; I): Stephen Younger and Artsvi Khachatryan (May 31, 2014; paper)
Bolivia (2009; I): Veronica Paz Arauco, George Gray-Molina, Wilson Jimenez and Ernesto Yañez (CEQ Web
Dec 2013) Public Finance Review, May 2014, Volume 42, Issue 3
Brazil (2009; I): Sean Higgins and Claudiney Pereira (CEQ Web Dec 2013) Public Finance Review, May 2014,
Volume 42, Issue 3
Chile (2009, I): Jaime Ruiz-Tagle and Dante Contreras (Oct. 25, 2014)
Colombia (2010, I): Marcela Melendez, Nora Lustig and Valentina Martinez (May 2014)
Costa Rica (2010; I): Pablo Sauma and Juan Diego Trejos (February 2014; paper)
El Salvador (2011; I): Margarita Beneke, Nora Lustig and Jose Andres Oliva (March 11, 2014)
Ethiopia (2010/11; C): Ruth Hill, EyasuTsehaye, Tassew Woldehanna (Sept. 28, 2014)
Guatemala (2011; I): Maynor Cabrera, Nora Lustig and Hilcias E. Moran (August 27, 2014)
Indonesia (2012; C) : Jon Jellema and Matthew Wai-Poi (Sept. 9, 2014)
Jordan (2010; C) : Morad Abdel-Halim, Shamma Adeeb Alam, Yusuf Mansur, Umar Serajuddin, Paolo
Verme (May 16, 2014)
Mexico (2010; I): John Scott (CEQ Web Dec 2013) Public Finance Review, May 2014, Volume 42, Issue 3
Peru (2009; I): Miguel Jaramillo (CEQ Web Dec 2013) Public Finance Review, May 2014, Volume 42, Issue 3
South Africa (2010; I): Ingrid Woolard, Precious Zikhali, Mashekwa Maboshe, Jon Jellema (Aug. 25, 2014)
Sri Lanka (2009/10; C): Nisha Arunatilake, Gabriela Inchauste and Nora Lustig (April 8, 2014; paper)
United States (2011; I): Sean Higgins, Nora Lustig, Whitney Ruble and Timothy Smeeding (paper Oct.
2014)
Uruguay (2009; I): Marisa Bucheli, Nora Lustig, Maximo Rossi and Florencia Amabile (CEQ Web Dec 2013)
3
Public Finance Review, May 2014, Volume 42, Issue 3
Handbook
• Lustig, Nora and Sean Higgins. 2013. Commitment to Equity
Assessment (CEQ): Estimating the Incidence of Social
Spending, Subsidies and Taxes. Handbook. CEQ Working Paper
No. 1, Center for Inter-American Policy and Research and
Department of Economics, Tulane University and InterAmerican Dialogue, September.
4
Commitment to Equity Assessments (CEQ)
• Accounting Approach: no behavioral, no general equilibrium
effects and no intertemporal effects
• Point-in-time
• Mainly average incidence; a few cases with marginal incidence
• Comprehensive standard fiscal incidence analysis of current
systems
• Harmonized definitions and methodological approaches to
facilitate cross-country comparisons
• Uses income/consumption per capita as the welfare indicator
• Tax shifting assumptions are the standard ones
• Allocators vary => full transparency in the method used for
each category, tax shifting assumptions, tax evasion
• Secondary sources are used to a minimum
5
• Handbook (Lustig and Higgins, 2013)
6
Basic elements of standard fiscal incidence
• Before taxes and transfers income of unit h, or Ih
• Taxes Ti
– personal income taxes; contributions to social security
– consumption and production taxes and subsidies
• Transfers Ri
– social spending: cash & near-cash transfers; in-kind transfers
(education and health)
– consumption and production (agriculture) subsidies
• “Allocators” of tax i and transfer j to unit h, or Sih ,
Sjh (the share of tax i borne or transfer j received
by unit h) => Incidence
• Post-taxes and transfers income of unit h (Yh)
7
• Post-taxes and transfers income of unit h
(Yh) is:
Taxes & Transfers
Post-fisc
Income
Yh = Ih - ∑i TiSih + ∑j RjSjh
Pre-fisc
Income
Incidence of Taxes & Transfers
8
MARKET INCOME
PLUS DIRECT TRANSFERS
MINUS DIRECT TAXES
GROSS INCOME
NET MARKET INCOME
MINUS DIRECT TAXES
PLUS DIRECT TRANSFERS
DISPOSABLE INCOME
MINUS NET INDIRECT TAXES
Construction
of Income
Concepts
POST-FISCAL INCOME
PLUS MONETIZED VALUE OF PUBLIC SERVICES: EDUCATION & HEALTH
FINAL INCOME
9
Methods to Construct Income
Concepts
• Direct Identification Method
• Imputation Method
– Direct (Education and Health)
– Simulation (Direct and Indirect Taxes)
• Inference Method
• Alternate Survey
• Secondary Sources Method
10
Reconciling the Two Economies:
Survey Data vs. Administrative Accts.
• What to do when totals in Survey do not
match administrative accounts?
• Should imputed values be scaled-down or the
rest of the concepts scaled up?
11
Contributory Pensions
• Are they a government transfer or deferred
consumption and hence part of market
income?
– No consensus
– Results, especially for poverty, are extremely
sensitive
=> Do it both ways
12
Importance of Comprehensive
Analysis
• Obvious reason: to capture the full effect of
the net fiscal system
• More subtle reason: partial assessments of
progressivity and regressivity can be
misleading
=> a regressive tax can be equalizing and
re-inforce the equalizing impact of
transfers
13
Lambert’s Conundrum
1
2
3
4
Original income x
10
20
30
40
Tax Liability t(x)
6
9
12
15
Benefit level b(x)
21
14
7
0
Post-benefit income
31
34
37
40
Final income
25
25
25
25
Source: Lambert, 2001, Table 11.1, P. 278
Total
100
42
42
142
100
14
Lambert’s Conundrum
• The Reynolds-Smolensky (R-S) index for taxes in
this example is equal to -0.0517, highlighting
their regressivity.
• Yet, the R-S for the net fiscal system is 0.25,
higher than the R-S for benefits equal to 0.1972.
• If taxes are regressive vis-à-vis the original
income but progressive with respect to the less
unequally distributed post-transfers (and
subsidies) income,
=> regressive taxes exert an equalizing effect over an
above the effect of progressive transfers.
• Two renowned studies found this type of result in
the US and the UK.
15
Results
Redistribution and Inequality
Reduction
16
Redistribution in the rich and
developing countries
Change in Gini: Disposable vs. Market
(in GINI points)
Ireland
Belgium
United Kingdom
Finland
Luxembourg
France
Germany
Netherlands
Austria
Portugal
Denmark
Slovenia
Sweden
Spain
South Africa(2010)
Czech Republic
Malta
US(2011)
Hungary
Italy
Slovakia
Latvia
Romania
Estonia
Cyprus
Lithuania
Greece
Brazil(2009)
Chile(2009)
Uruguay(2009)
Bulgaria
Poland
Mexico(2010)
Costa Rica(2010)
Peru(2009)
Bolivia(2009)
El Salvador (2011)
Colombia(2010)
-0.05
Guatemala(2010)
0.00
-0.10
-0.15
-0.20
-0.25
-0.30
Sources: EUROMOD for EU,Higgins et al. (2014) for US and for CEQ countries see Lustig (2014) and references at the end.
Note: in these calculations contributory pensions are part of market income and NOT treated as a government transfer.
17
Redistribution in Middle and Low
Income Countries: CEQ 16
Change in Gini: Disposable vs. Market
(in GINI points)
South Africa(2010)
Chile(2009)
Brazil(2009)
Uruguay(2009)
Mexico(2010)
Costa Rica(2010)
Peru(2009)
Bolivia(2009)
-0.03
El Salvador (2011)
-0.02
Colombia(2010)
-0.01
Guatemala(2010)
0.00
-0.04
-0.05
-0.06
-0.07
-0.08
-0.09
18
The impact of direct taxes and transfers
on inequality (Gini coefficient): CEQ 16
Changes in Gini: Disposable VS Market Income
0.8
16000
14000
0.7
12000
10000
0.6
8000
0.5
6000
4000
0.4
2000
0.3
0
Mexico(2010)
Chile(2009)
Colombia(2010)
Brazil(2009)
South Africa(2010)
Gini of Disposable Income
Gini of Market Income
GNI per capita (2005 PPP)
19
The impact of net indirect taxes
on inequality (Gini coefficient): CEQ 16
Changes in Gini Coef icients
0.9
16000
0.8
14000
0.7
12000
0.6
10000
0.5
8000
0.4
6000
0.3
4000
0.2
2000
0.1
0.0
0
SA
Bra
Col
Chi
(2010) (2009) (2010) (2009)
Gini of Market Income
Mex
(2010)
Gini of Disposable Income
Gini of Post- iscal
GNI per capita (2005 PPP)
20
Lindert’s (2006) historical result is also found in cross
section: Higher GDP/capita, more redistribution
Change in Gini: Post-fiscal vs. Market
(decline in Gini points shown in posi ve quadrant)
0.09
South Africa
0.08
0.07
0.06
0.05
CHL
0.04
BRA
0.03
MEX
CRI
0.02
URY
PER
SLV
0.01
COL
0.00
0
2000
BOL
GTM
4000
6000
8000
10000
GNI/Capita (2005 PPP)
12000
14000
16000
21
However, no Robin Hood Paradox
And results do not depend on South Africa
Change in Gini points: Post-fiscal vs. Market
(decline in Gini points shown in posi ve quadrant)
0.09
0.08
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0.00
Sputh Africa
URY
SLV
0.30
0.35
0.40
0.45
MEX
CHL
BRA
CRI
PER
BOL
COL
GTM
0.50
0.55
0.60
Mkt Income Gini
0.65
0.70
0.75
0.80
22
Results
Redistribution and Poverty
Reduction
23
Direct Transfers (net of direct taxes)
reduce poverty (except in Ethiopia): CEQ 16
Change in Headcount Ratio ($2.5 PPP/Day):
Disposable vs. Market Income
(in percentage points)
2%
0)
)
01
ri
ca
(2
20
So
ut
h
Af
az
i l(
20
Br
Ur
ug
ua
y(
(2
ia
09
)
09
9)
00
)
10
20
Bo
ic
ex
M
li v
o(
(2
or
ad
lv
Sa
El
01
1)
9)
00
(2
0)
Ch
ile
01
)
(2
10
ca
20
Co
st
a
Ri
a(
al
m
ua
te
Pe
ru
(2
00
10
G
-6%
Co
lo
m
bi
a(
20
-2%
-4%
9)
)
0%
-8%
-10%
-12%
-14%
24
Indirect Taxes increase poverty over and
above market income poverty in six out
of the CEQ 16 countries
25
Changes in Headcount Ratio: Post- iscal vs. Market Income
(Poverty Line: US$2.50ppp/day)
0.9
16000
0.8
14000
0.7
0.6
0.5
12000
South
Africa
10000
8000
0.4
0.3
6000
Brazil
4000
0.2
Headcount of Market Income
Headcount of Disposable Income
Headcount of Post- iscal
Chi(2009)
Mex(2010)
Bra (2009)
0
Peru (2009)
0.0
Col (2010)
2000
SA (2010)
0.1
GNI per capita (2005 PPP)
26
Note that Net Indirect Taxes can be equalizing
and yet poverty increasing: Ethiopia
Change in Gini: Marginal Contribu on of Net Indirect
Taxes
Change in Headcount Ra o ($2.5 PPP/Day): Marginal Contribu on from Net Indirect Taxes
(in percentage points)
Gu
at
Ind
Eth
iop
ia
em
al
Bo
liv a(20
ia(
10
20
)
In
09
do
)
ne
Sr sia(2
iL
an 012
k
)
Co a(2
00
lo
m
9)
bi
a
El
Sa (201
lv a
0)
Pe d or
ru
(2
(2
01
00
1)
Et
9)
hi
op
i
Co a(20
st
a R 11)
ic
Jo
rd a(20
an
10
(2
01 )
M
ex
0
ico )
(
2
Ar
m 010
en
)
i
Ur a(20
ug
11
ua
)
y
(
Br
az 200
9)
il(
2
Ch 009
)
ile
(2
00
So
ut 9)
h
A
(20
11)
one
sia
(20
12)
Sr i
Lan
ka(
200
9)
Per
u(2
009
)
Col
om
b ia
(20
10)
Gu
ate
ma
la(2
010
Cos
)
ta R
ica
(20
10)
Chi
le(2
009
)
El S
alv
ad o
r (2
011
Jor
)
dan
(20
10)
Me
xico
(20
10)
Bol
ivia
(20
09)
Arm
eni
a( 2
011
)
Uru
gua
y(2
009
)
Bra
zil(
200
9)
Sou
th
Afr
ica
(20
10)
(in GINI points)
8%
0.01
0.00
-0.01
-0.02
-0.03
-0.04
-0.05
-0.06
-0.07
-0.08
-0.09
6%
4%
2%
0.045
0.026
0.036
0%
-2%
0.006
-0.007
-0.015
0.012
0.006
0.006
-0.002
-0.012
-0.016
-0.024
-4%
-0.019
-0.028
-6%
-8%
-0.072
Contribu on from Net Indirect Taxes
Change in Headcount: Post-Fiscal vs Mkt
Contribu on from Net Indirect Tax
Gini Change: Post-fiscal vs Mkt
27
Thank you!
28