Income Inequality and Fiscal Policy in Low -Income Countries

Sanjeev GuptaFiscal Affairs Department

International Monetary FundMarch 15, 2013

Low-Income Countries Seminar

Presentation is based on:

F. Bastagli, D. Coady, and S. Gupta, 2012, “Income Inequality and Fiscal Policy”, IMF Staff Discussion Note, SDN/12/08(Revised).

Available at: http://www.imf.org/external/pubs/cat/longres.aspx?s k=

40024

2

Background

� Income inequality has increased in most advanced and many developing economies over recent decades

� Emphasis on inclusive growth has led to a growing concern about income inequality in developing countries (e.g., China and India)

� So how can fiscal policy contribute to lowering income inequality?

3

Plan of Presentation

I. Role of fiscal policy

II. Trends in income inequality

III. How has fiscal policy affected income inequality in advanced economies?

IV. How effective has fiscal policy been at reducing inequality in developing countries?

V. Lessons for the design of fiscal policy in developing countries

4

I. Role of Fiscal Policy

5

I. Role of Fiscal Policy

� Fiscal policy can affect income distribution

�Directly. By reducing inequality of disposable incomes compared to inequality of market incomes

� Indirectly. Through impact on future earnings of individuals and inequality of market incomes

� Role likely to vary across countries reflecting ran ge of policy instruments available but also social preferences towards equity and efficiency

� But taxes and transfers may distort allocation of resources (equity-efficiency trade-off)

6

II. Trends in income inequality

7

Trends in Disposable Income Inequality, 1980–2010

Income inequality is substantially higher in low -income economies…..

0.20

0.25

0.30

0.35

0.40

0.45

0.50

0.55

1980 1985 1990 1995 2000 2005 2010

Gin

i coe

ffici

ent

Latin America and Caribbean Sub-Saharan Africa Middle East and North Africa

Asia and Pacific Emerging Europe Advanced

8

0

2

4

6

8

10

12

14

16

18

Nep

al

Chi

na

Sri

Lank

a

Indo

nesi

a

Lao

PD

R

Ban

glad

esh

Cam

bodi

a

Indi

a

Vie

tnam

Mon

golia

Phi

lippi

nes

Par

agua

y

Hon

dura

s

Col

ombi

a

Ven

ezue

la, R

B

Bol

ivia

Uru

guay

Cos

ta R

ica

Arg

entin

a

Dom

inic

an R

epub

lic

Jam

aica

Gua

tem

ala

Tur

kmen

ista

n

Kyr

gyz

Rep

ublic

Uzb

ekis

tan

Taj

ikis

tan

Djib

outi

Mor

occo

Tun

isia

Mau

ritan

ia

Egy

pt, A

rab

Rep

.

Rw

anda

Sou

th A

fric

a

Nig

er

Gha

na

Côt

e d'

Ivoi

re

Tan

zani

a

Moz

ambi

que

Zam

bia

Mad

agas

car

Bur

undi

Changes in Disposable Income Inequality Across Regi ons, 1990–2005 (Percentage-point change in Gini coefficient)

…..and has been increasing in many of these

Latin America

Middle East and North

Africa

Sub-Saharan Africa

Asia and Pacific

9

0

5

10

15

20

25

Per

cent

United States United KingdomAustralia CanadaSouth Africa India

0

5

10

15

20

25

Per

cent

France GermanyJapan NetherlandsSweden

Gross Income Share of Top One-Percent in Selected A dvancedand Developing Economies, 1925–2010

More recently, the focus has been on the rising income share of the top income groups

Source: World Top Incomes Database

10

III. How has fiscal policy affected income inequality in advanced economies?

11

0

0.1

0.2

0.3

0.4

0.5

0.6

0

0.1

0.2

0.3

0.4

0.5

0.6

Disposable income Gross income

In advanced economies, fiscal policy has reduced income inequality by one-third ….

Source: OECD, 2008.

Redistributive impact in OECD countries, 2008

0

0.1

0.2

0.3

0.4

0.5

0.6

0

0.1

0.2

0.3

0.4

0.5

0.6

Gross income Disposable income Gross income

� Gross income average = 0.45

� Disposable income average = 0.30

12

…..with about two-thirds of this impact is achieved on the expenditure side

Redistributive impact of tax and spending

Source: Paulus, 2009.

0.00

0.05

0.10

0.15

0.20

0.25

0.30

Cha

nge

in G

ini c

oeffi

cien

t

non means-testedbenefits

means-testedbenefits

personaltaxes

social insurancecontributions

Transfers = 0.10

Transfers and taxes = 0.15

13

Indirect taxes and in-kind transfers also influence the redistributive impact of fiscal policy

� Indirect taxes . Studies find that the value-added tax (VAT) and excise duties are regressive in European countries (O’Donoghue et al., 2004; Warren, 2008)

� In-kind transfers. Spending on education, health care and housing benefits decreased the Gini coefficient by 5.8 percentage points on average in 5 European economies (Paulus et al., 2009)

14

Corporate income taxes may not be as progressive as often assumed

� The incidence of corporate taxes will tend to fall on wages as capital is more mobile

� However, taxation of “rents” (above normal profits) is likely to fall on owners of capital

15

Diminishing Redistributive Impact of Fiscal Policy Since Mid-1990s

However, the redistributive impact of fiscal policy has decreased since the mid-1990s

36.2

26.7

39.2

27.4

39.8

28.3

0

5

10

15

20

25

30

35

40

45

Market Income(left)

Disposable Income(left)

Gin

i coe

ffici

ent

mid-1980s mid-1990s mid-2000s

73.2

52.5

0

10

20

30

40

50

60

70

80

Fiscal Redistribution(right)

Per

cent

Source: Immervoll and Richardson, 2011; OECD, 2011.

16

IV. How effective has fiscal policy been at reducing inequality in developing countries?

17

Levels and Composition of Tax Revenues and Social S pending

Impact of fiscal policy in developing economies is limited by low tax-spending levels…

0

5

10

15

20

25

30

35

40

Advanced EmergingEurope

LatinAmerica

MiddleEast and

NorthAfrica

Asia andPacific

Sub-Saharan

Africa

Per

cent

of G

DP

Tax,2010 or latest

Property

Corporate

Income

Indirect

0

5

10

15

20

25

30

Advanced EmergingEurope

MiddleEast and

NorthAfrica

LatinAmerica

Sub-Saharan

Africa

Asia andPacific

Per

cent

of G

DP

Social Spending,2010 or latest

Education

Health

Transfers

18

…..as well as less progressive taxes and transfer programs

� Greater reliance on indirect taxes and narrower tax bases

� Progressivity of direct taxation is weakened by tax noncompliance and narrow tax bases

� On the spending side, poor targeting limits the redistributive capacity of transfer programs

19

8.6

1.7 1.60.9 0.6

0.0

4.6

3.0

1.8 3.3

1.11.6

13.2

4.7

3.44.1

1.8 1.6

0

2

4

6

8

10

12

14

MENA CEE-CIS Sub-Saharan Africa E.D. Asia LAC Advanced

Per

cent

of G

DP

Energy price subsides as a percentage of GDP

Pre-tax

Tax-subsidies

Post-tax

20

Bottom quintile, 19.0

Second quintile, 19.7

Third quintile, 20.6

Fourth quintile, 20.1

Top quintile, 20.6

Kerosene

Bottom quintile, 7.2

Second quintile, 11.4

Third quintile, 16.2

Fourth quintile, 22.5

Top quintile, 42.8

Bottom quintile, 3.0

Second quintile, 5.7

Third quintile,

9.7

Fourth quintile, 19.4Top quintile,

61.3

GasolineBottom

quintile, 7.3

Second quintile, 11.7

Third quintile, 16.3

Fourth quintile, 22.6

Top quintile, 42.0

Diesel

Bottom quintile, 3.8

Second quintile, 7.6

Third quintile,

12.6

Fourth quintile, 20.8

Top quintile, 53.8

LPG

Fuel subsidies benefit upper income groups the most… …across all products

21

Fiscal policy accounts for nearly 3/4 of Europe vs. Latin America Gini difference

Re-distributional impact: Europe vs. Latin America0.23

0.17

0

0.05

0.1

0.15

0.2

0.25

Disposable gini difference

Fiscal impact

Right-side axis

0.52

0.46

0.5

0.27

Latin America Europe0

0.1

0.2

0.3

0.4

0.5

0.6

Gross income Disposable income

Source: Goñi, López and Servén, 200822

Benefit Incidence of Education and Health Public Sp ending(share of bottom 40 percent)

In-kind public spending has been found to be regressive in many developing economies

0

10

20

30

40

50

60

70

Chi

le (

1996

)R

oman

ia (

1997

)A

rgen

tina

(199

3)D

omin

ican

Rep

ublic

…B

ulga

ria (

1997

)M

alaw

i (19

97)

Mac

edon

ia (

1996

)M

exic

o (1

996)

Per

u (1

994)

Mon

golia

(19

95)

Kyr

gyz

Rep

ublic

(19

93)

Cot

e d'

Ivoi

re (

1995

)G

uyan

a (1

993)

Pan

ama

(199

7)T

anza

nia

(199

3)S

outh

Afr

ica

(199

3)M

ozam

biqu

e (1

996)

Ban

glad

esh

(200

0)E

cuad

or (

1998

)V

ietn

am (

1993

)A

rmen

ia (

1996

)K

azak

hsta

n (

1996

)M

adag

asca

r (1

993)

Nep

al (

1996

)N

icar

agua

(19

93)

Gui

nea

(199

4)

Education Benefit Incidence for Bottom 40%

0

10

20

30

40

50

60

70

80

Arg

entin

a (1

993)

Chi

le (

1992

)C

olom

bia

(199

2)C

osta

Ric

a (1

992)

Jam

aica

(19

93)

Hon

dura

s (1

995)

Bra

zil (

1990

)P

eru

(199

7)M

ongo

lia (

1995

)T

anza

nia

(199

3)B

angl

ades

h (2

000)

Sou

th A

fric

a (1

993)

Egy

pt (

1994

)K

enya

(19

92)

Mad

agas

car

(199

3)M

ozam

biqu

e (1

996)

Bul

garia

(19

95)

Vie

tnam

(19

93)

Indo

nesi

a (1

990)

Rom

ania

(19

97)

Gha

na (

1998

)In

dia

(199

6)E

cuad

or (

1998

)G

uine

a (1

994)

Health Benefit Incidence for Bottom 40%

23

Conditional cash transfers

� The recent expansion of “conditional cash transfer” programs provides a promising approach for enhancing the distributive power of public spending in developing economies

� The largest programs, in Brazil and Mexico, have reduced the Gini by 2.7 percentage points (Soares et al., 2007)

� However, these programs need to be targeted to the poorest households

24

V. Lessons for the design of fiscal policy in developing countries

25

Lessons for the design of fiscal policy

� In developing economies, the capability of fiscal policy to address income inequality needs to be enhanced

� This requires improvements on two fronts:

� The level of tax and spending needs to be increased

� The redistributive impact of tax and spending needs to be improved

26

Enhancing role of tax policy

� Strengthening resource mobilization capacity

� Improvement in administrative capacity

�Expansion of corporate and personal income tax bases (addressing exemptions, loopholes, and tax compliance)

�Expansion of tax policy instruments (VAT plus excises)

27

Enhancing role of public spending

� Higher and better targeted spending

�Expansion and improved targeting of social assistance (eliminate universal price subsidies)

�Expansion of health and education

�Expansion of conditional cash transfers

28

THANK YOUTHANK YOUTHANK YOUTHANK YOU

29

Change in Gini Coefficient, 1990 to 2005

ChangeLarge Increase

(Change ≥ 5)

Medium Increase

(3 ≤ Change < 5)

Small Increase

(0 < Change < 3)

Small Decrease

(-3 < Change < 0)

Medium Decrease

(-5 < Change ≤ -3)

Large Decrease

(Change ≤ -5)

Latin America

and Caribbean1990-2005

Colombia, Honduras,

Paraguay, Venezuela

Bolivia, Costa Rica,

Uruguay

Argentina, Dominican

Republic, Guatemala,

Jamaica

El Salvador,

Panama

Brazil, Chile, Ecuador,

Nicaragua, PeruBelize, Mexico

Sub-Saharan

Africa1990-2005

Cote d'Ivoire, Ghana, Niger,

Rwanda, South Africa

Mozambique,

Tanzania

Burundi, Madagascar,

Zambia

Cameroon,

Nigeria, UgandaGambia

Burkina Faso, Central

African Republic,

Ethiopia, Guinea,

Guinea-Bissau, Kenya,

Lesotho, Malawi, Mali,

Namibia, Senegal,

Swaziland

Asia and Pacific 1990-2005

China, Indonesia, Rep. of

Korea, Lao PDR, Nepal, Sri

Lanka

Bangladesh,

Cambodia, TaiwanIndia, Mongolia,

Philippines, Vietnam

Thailand Malaysia

Middle East

and North Africa1990-2005

Kyrgyz Republic, Tajikistan,

Turkmenistan, UzbekistanDjibouti

Egypt, Mauritania,

Morocco, TunisiaPakistan Iran, Jordan

30