Inequality: Where are we? Jaime Saavedra Poverty Reduction and Equity

THE WORLD BANK

Oslo May , 2011



Measuring and analyzing inequality of outcomes and opportunities.



What is the equity impact of policies, or…how to make policies more equity enhancing? The PSIA agenda.



What is the role of fiscal policy? Are taxes and transfers working in reducing inequality? Is the social contract effective?

Where are we on inequality?  Persistent

and high across regions. Small reduction in some regions, some countries

 Reductions

in inequality of opportunities in some regions

 Strong

case for public policy to address inequalities

60

Inequality high and persistent in Latin America

Colombia Bolivia Honduras Guatemala Paraguay

Ecuador

Nicaragua

Brazil Panama

50

Chile Mexico Peru Costa Rica Dominican Republic Argentina Uruguay

Gini Index 40

El Salvador Jamaica

20

30

Venezuela, R.B. de

6

7 8 Log GDP per capita (PPP) Other Countries

9

LAC Countries

Scatter plot log per capita GDP (PPP) and Gini coefficient. Most recent since 2002.

10

Gini Index 40

50

60

Europe and Central Asia countries are more equal on average.

Russian Federation Macedonia, FYR Turkey

Georgia

Uzbekistan

Montenegro Latvia Bosnia and Herzegovina Lithuania Estonia Poland

Kyrgyz Republic

Albania

30

Tajikistan

Moldova

Romania Kazakhstan

Armenia

Slovenia Hungary Croatia

20

Bulgaria Belarus Serbia Ukraine

Azerbaijan

6

7 8 Log GDP per capita (PPP) Other Countries

ECA Countries

9

10

Also high in Africa 60

Comoros

Liberia

50

Lesotho Zambia Kenya Madagascar Gambia, The

Gini Index 40

Mozambique Congo, Dem. Rep. of

NigerCentral African Republic Guinea Uganda Sierra Leone

Ghana

Cote d'Ivoire Congo, Rep.

Nigeria Gabon

Chad Burkina Faso Mali Benin

Malawi

Senegal

GuineaBissau Togo

30

Burundi

20

Ethiopia

6

7 8 Log GDP per capita (PPP) Other Countries

SSA Countries

9

10

Uzbekistan

Albania

Macedonia, FYR

Bosnia and Herzegovina

Lithuania

Tajikistan

Russian Federation

Georgia

Kyrgyz Republic

Romania

Moldova

Kazakhstan

Ukraine

Belarus

Turkey

Armenia

Bulgaria

Azerbaijan

-6

And also in East Asia and the pacific Jamaica

Nicaragua

Honduras

Uruguay

Costa Rica

Colombia

Mexico

Guatemala

Panama

Peru

Dominican Republic

Chile

Bolivia

Brazil

El Salvador

Argentina

Gini Annual Change (p.p) 0 2 4 -4 -2

ECA

Paraguay

6

Mixed pattern in Eastern Europe and Central Asia

Venezuela, R.B. de

Ecuador

-6

Gini Annual Change (p.p) -4 -2 0 2 4

6

Changes in Inequality (Early 2000’s-Late 2000’s) have varied across countries Small reduction in most Latin America countries LAC

Diverging patterns in Brazil and China



Manmohan Singh, : “Large scale poverty is the foremost challenge facing our country.. We have to be acutely conscious of regional disparities and imbalances within the country, and address the inequalities that exist.” (The Hindu, March 27, 2011)



Dilma Rouseff, "The most determined struggle will be to eradicate extreme poverty … we can be a more developed and fairer country... I will not rest while there are Brazilians without food on their table, homeless in the streets, and poor children abandoned to their luck." (Inauguration address, January 2011)



Jacob Zuma, "While many South Africans celebrate the delivery of houses, electricity or water, there are yet many others who are still waiting... " (2011 State of the Nation Address)



Min Zhu, IMF “The increase in inequality is the most serious challenge for the world ... I don’t think the world is paying enough attention.” (Davos, May 2011)



Premier Wen “…many had not seen the benefits of China’s dynamic growth” ...[we need to ]... “combat rising inequality and corruption” (Washington Post, 3/12/11)



Thai Democrat Party campaign statement "Need to address income inequality which (is) causing social problems in Thai society” (BKK Post, 2/28/20)

Another poverty indicator: malnutrition Underweight (0-3 years) 60 50

%

40 30 20 10 0 1988

1990

1992

1994

India

1996

1998

2000 Nigeria

2002

2004

2006

2008

2010

Peru

Reduction in % of children underweight in India, Nigeria and Peru

Source: DHS, STATcompiler

And again large and persistent gaps across income levels Underweight (0-3 years) by Wealth Quintile 100 90 80 70

%

60 50 40 30 20 10

India Lowest

Nigeria Lowest

Peru Lowest

India Highest

Nigeria Highest

Peru Highest

2000

1996

1991

2008

2003

1990

2006

1998

1992

0

But little or no convergence between the richest and poorest – in fact widening of rich-poor gap for India and Nigeria

Source: DHS, STATcompiler

Two Latin-american children Probability of finishing 6to grade on time Jamaica Argentina México Chile El Salvador Venezuela Panamá Bolivia Paraguay Uruguay Honduras Ecuador Costa Rica R. Dominicana Colombia Perú Guatemala Nicaragua Brasil 0.0

Boy with 4 siblings in a rural home , uneducated head of household and income at extreme poverty levels

Child with one sibling un an urban home, household head with secondary education in family income per capita of US$25 a day.

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Two African children: Probability of completing 6th grade on time Nigeria Zimbabwe Namibia Ghana Kenya Cameroon Madagascar Malawi Mali Zambia Sierra Leone Congo Dem Rep. Niger Ethiopia Uganda Mozambique Tanzania Liberia Rwanda 0%

10 %

20 %

30 % Poor child

40 %

50 %

60 %

70 %

80 %

Rich child

Note: Rich child is a boy, living in a urban household where the head is a male with 12 years of education, with one additional child in the household and belong to the fifth quintile of wealth. Poor child is a boy, living in a rural household where the head is a male with 5 years of education, with four additional children in the household and belong to the first quintile of wealth.

The equality of opportunity principle Circumstances exogenous to the individual, like birth place, gender, ethnicity, income and education of parents should not determine the persons wellbeing

There are two problems here “In Peru 4 out of 10 children less than 5 years do not have access to clean water” Problem:

access, coverage

“Those 4 children are indigenous” Problem:

circumstances affect children's chances. distribution

Human opportunity index 

Inequality of opportunity-sensitive coverage rate that incorporates: A)The average coverage of a good or service. that should be universal, B) If it is allocated according to an equality of opportunity principle  access to key goods and services should not be determined by “circumstances” outside one’s control E.g. race, gender, parental education, wealth, geographic location).

Africa and Latin America (late 2000s)

HOI - LAC

Chile

Uruguay

Brazil

Argentina

Dominican Republic

Venezuela, R.B. de

Peru

Jamaica

Bolivia

HOI - Africa

Mexico

Panama

Colombia

Paraguay

Costa Rica

El Salvador

Ecuador

Nicaragua

Honduras

Guatemala

namibia

uganda

malawi

zambia

ghana

rwanda

tanzania

madagascar

mozambique

.

0

20

HOI (%) 40 60

80

100

Attending School (10-14 years)

Coverage

Note: All HOI used for these comparisons use the same definition of opportunities and comparable list of circumstances

In terms of school attendance, African countries are comparable with many countries in LAC region – in coverage and HOI

HOI - LAC HOI - Africa

Mexico

Jamaica

Argentina

Chile

Ecuador

Uruguay

Peru

Bolivia

Venezuela, R.B. de

Panama

Costa Rica

Colombia

Paraguay

Dominican Republic

Honduras

El Salvador

Brazil

Nicaragua

Guatemala

namibia

ghana

zambia

malawi

madagascar

uganda

tanzania

mozambique

rwanda

.

0 20

HOI (%) 40 60 80

100

Africa and Latin America (late 2000s) Finished 6th Grade On Time (13 - 15 years)

Coverage

African countries compare poorly with most LAC countries on completion of primary on time - Late entry is a major problem in Africa

Encouraging trends for Africa on school attendance (late 1990s – late 2000s) Annual Change - Attending School (10-14 years) Nicaragua Honduras El Salvador Brazil Mexico Colombia Ecuador Panama Costa Rica Guatemala Peru Venezuela, R.B. de Dominican Republic Chile Paraguay Jamaica tanzania rwanda mozambique madagascar uganda zambia ghana malawi namibia

0

1 2 Percentage Points Change HOI

3

Change Coverage

Large improvements in school attendance for most African countries • In almost all African countries change in HOI > change in coverage  reduction in inequality in attendance •

But mixed picture on trends for primary school completion in Africa Annual Change - Finished 6th Grade On Time (13 - 15 years) Brazil Peru Costa Rica Paraguay El Salvador Nicaragua Honduras Guatemala Mexico Dominican Republic Ecuador Venezuela, R.B. de Colombia Chile Panama Jamaica namibia malawi zambia tanzania madagascar uganda mozambique rwanda ghana

-1

0

1 Percentage Points Change HOI

• Little

2

3

Change Coverage

or no improvement in HOI for 4 out of 9 African countries • Increase in HOI much smaller than increase in coverage for the 6 African countries showing an improvement  rise in inequality in primary school completion

HOI Coverage

Source: World Bank using DHS data; work under progress (do not cite) Chile

Venezuela, R.B de

Mexico

Costa Rica

Brazil

Ecuador

Paraguay

Colombia

Jamaica

El Salvador

Guatemala

Panama

Peru

Nicaragua

Honduras

Ghana

Namibia

Zambia

Madagascar

Uganda

0

20

40

% 60

80

100

Africa and Latin America - Access to Electricity (late 2000s)

Changes in the HOI - Electricity (late 1990’s vs. late 2000’s) Annual Change - Access to electricity (0-16 years) Guatemala Jamaica El Salvador Paraguay Peru Nicaragua Panama Ecuador Brazil Mexico Colombia Chile Honduras Costa Rica Venezuela, R.B. de indonesia cambodia ghana namibia tanzania madagascar malawi mozambique uganda zambia rwanda

-.5

0

.5 1 Percentage Points Change HOI

1.5

Change Coverage

Source: World Bank using DHS data; work under progress (do not cite)

2

What is behind observed inequality of opportunities among children? (Peru 2009) 100% 90%

14

20

80% 70%

28

60%

21

23

26

49

47

45

12

14

15

Adequate sanitation

Electricity

27

50% 40% 30% 20%

33

32

10% 0% Finish primary (6th) on time

School Adequate water attendance supply (ages 10-14)

Gender

Ethnicity (Language)

Education

Single parent

Number of Children

Region

Income

Altitude

What is behind observed inequality of opportunities among children? – the case of sanitation 100% 90%

38

80%

49

52

70% 60% 50%

5

30

40%

23 30%

34

20%

14

27

10% 0%

Cambodia

Household composition Income Geographical Location

Peru

Indonesia

Household head education Gender

The case for public policy to address the equity challenge

Gini after taxes and transfers 0.6 0.5

0.3 0.2 0.1

Market Incomes

Disposable Incomes (after tax and transfers)

Source: Goni, Lopez and Serven 2008

Sweden

Denmark

United Kingdom

Spain

Italy

France

Mexico

Chile

Brazil

0.0 Argentina

Gini Index

0.4

Taxes and Income – 2008 70

Tax revenue (% of GDP)

60

50

40

30

20

10

0 5

6

7

8

9

10

Log GDP per capita (constant 2000 US$)

Source: WDI and OECD

11

12

Taxes as a % of GDP Latin America and OECD Tax collection is low -

Not because of low rates

-

Because exemptions and loopholes

-

Progressive personal income taxes are barely collected

-

Very high informality

…and informality seems to relates to the (perceived) quality and fairness of state services Government Effectiveness

.2 0

CHI

VEN ARG ECU

COL PER BOL HON GUA BRA MEX PAN SAL

CHIJAM COS

-.2

ARG ECU

BOL COL PER GUAHON JAM BRA MEX COS SAL PAN

e(Share of Self-employed / X)

.2 0

VEN

-.2

e(Share of Self-employed / X)

.4

Parcial Correlations controling for the GDP pc at PPP

.4

Parcial Correlations controling for the GDP pc at PPP

-1.5

-1

-.5

0

e(Government Effectiveness / X) coef = -.07803752, se = .02249611, t = -3.47

.5

1

-4

-2

0

2

4

e(Impartiality of Courts / X) coef = -.02090103, se = .00903889, t = -2.31

Impartiality of Courts

Does the social contract work? United Kingdom 30 %

Transfers HH amount as % of GDP per hh

30 %

HH amount as % of GDP per hh

Colombia

20 %

10 %

0%

-10 %

20 %

10 %

0%

-10 %

-20 %

-20 %

-30 %

-30 %

Taxes

-40 %

-40 %

-50 %

-50 %

-60 %

-60 % I

II

III

IV

V

From: Breceda, Rigolini and Saavedra, 2009)

I

II

III

IV

V

Increase in Tax Rate on middle class (US$13 a day +) required to eliminate extreme poverty LAC

8.4

MNA

16.3

ECA

17.6

EAP

45.2

AFR

80.3

SAR

83.1

0

10

20

30

40

50

Marginal Tax Rate (%)

Source: Ravallion 2008

60

70

80

90

The richer do not participate and opt out 

The rich opting out into higher quality private provision (pension, security, private education, etc), leaving as demanders of public services those with less “voice”

...and the poor also feel disengaged. 

Differential access to public goods that should be universal equal access to property rights, to protection under the law, to judiciary services, to basic services

Overall negative views about state services and the political system.  

A system that gives me very little, and gives the others low quality services In some cases continued perception of patronage and corruption in social assistance

Implicit validation of widespread tax evasion  Much more EXIT than VOICE in Hirschman´s terminology

Low taxation and informality reflect a social contract that does not work for all. Bad equilibrium of low taxes, weak states, regulators with low enforcement capacity, insufficient provision of public goods, unequal opportunties

Redistribution through taxes or through transfers? A Tax Reform in Mexico

Progressive on the tax side

Regressive on the tax side…matters?

(% reduction in household incomes due to tax reform)

Some hope on the expenditure side Brazil : Contributions to changes in inequality 

Non-labor income has a more important role in explaining reduction in extreme poverty and inequality : mostly rural pensions, less CCTs

Source: Barros et al (2010).

Thank you