Nora Lustig
Samuel Z. Stone Professor of Latin American Economics
Dept. of Economics
Tulane University
Nonresident Fellow, Center for Global Development
and Inter-American Dialogue
Meeting of Western Hemisphere Think Tanks in
Support of the Summit of the Americas
Bogotá, March 14, 2011
1

Using Gini coefficient,
 19 percent more unequal than Sub-Saharan
Africa
 37 percent more unequal than East Asia
 65 percent more unequal than developed
countries
2
Gini Coefficient by Region (in %), 2004
60.0
55.0
53.2
Gini coefficient
50.0
44.7
45.0
40.0
35.0
32.2
38.9
38.9
39.1
South Asia
North Africa
and the
Middle East
East Asia and
the Pacific
33.6
30.0
25.0
20.0
High Incom e
Europe and
Central Asia
Sub-Saharan Latin Am erica
Africa
and the
Caribbean
3
Gini Coefficient for Latin America: early 1990s-mid 2000s
(Gasparini et al., 2009)
60
Gini coefficient
55
50
45
40
0
35
=
Early 90s
Mid. 90s
Early 2000s
Mid. 2000s
5

Inequality in most Latin American countries
(12 out of 17) has declined (roughly 1.1% a
year) between (circa) 2000 and (circa) 2007.

Except in two cases, decline is statistically
significant. (rectangles in grey in next slide)
6
-1.27
-2.00
-0.83
-0.24
Total 17 countries
Total 12 countries
0.98
Nicaragua
0.89
Costa Rica
0.79
Uruguay
1.00
Honduras
Guatemala
-0.29
Venezuela
-0.36
Ecuador
-0.50
Bolivia
Panama
-0.88
Paraguay
-0.99
Peru
-1.00
Dominican Rep.
-1.12
-1.02
Argentina
-1.00
Chile
Brazil
-1.40
Mexico
-1.50
El Salvador
1.50
1.02
0.50
0.00
0.05
-0.56
-0.37
-0.83
7
8
The decline took place in:
 Persistently high inequality countries (Brazil)
and normally low inequality countries
(Argentina)
 Fast growing countries (Chile and Peru), slow
growing countries (Brazil and Mexico) and
countries recovering from crisis (Argentina
and Venezuela)
 Countries with left populist governments
(Argentina), left social-democratic
governments (e.g., Brazil, Chile) and centerright governments (e.g., Mexico and Peru)
10
 In-depth analysis in four countries:
 Argentina (Gasparini and Cruces)
(urban; 2/3 of pop)
 Brazil (Barros, Carvalho, Mendoca &
Franco)
 Mexico (Esquivel, Lustig and Scott)
 Peru (Jaramillo & Saavedra)
11
Inequality has continued to decline in Brazil
The Gini Coefficient reached its lowest level in the last three decades.
Evolution of the degree of inequality in per capita income:
Brazil, 1995-2009
1981-2009
0.65
0.634
0.64
0.63
0.615
0.62
0.612
Gini coefficient
0.61
0.599
0.60
0.588
0.589
0.582
0.594
0.600
0.602
0.600
0.598
0.596
0.59
0.58
0.599
0.592 0.594
0.580
0.587
0.587
0.581
0.569
0.566
0.57
0.56
0.560
0.552
0.55
0.544
0.54
0.539
0.53
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
0.54
45
0.52
40
0.50
35
0.48
30
0.46
0.44
25
0.42
20



Decline in inequality is statistically
significant and significant in terms of order
of magnitude
There is Lorenz dominance (unambiguous
decline independently of choice of
inequality measure)
Robust to income concept (e.g., monetary
vs. total)
16
 Representative sample of Latin
American diversity:
 high/medium/low ineq
 high/low growth
 Populist/social democratic/centerright governments
17
Sample Representative of High and Low
Inequality Countries
(Latin America: Gini Coefficient by Country; circa 2007; in percent)
65
59.3
60
54.8 54.9
54.3 54.4
55
53.2
51.8 52.3
50
48.2 48.3
48.9
51.6
49.7 49.7 49.9
46.8
45
43.5
TOTAL COUNTRIES
Honduras
Bolivia
Panama
Brazil
Guatemala
Ecuador
Paraguay
Nicaragua
Chile
Mexico
El Salvador
Peru
Costa Rica
Dominican Rep.
Argentina
Uruguay
40
Venezuela
Gini coefficient
57.2
18
All four countries experienced substantial market-oriented reforms in the 1990s (and since the 1980s in the
case of Mexico). In particular, trade and foreign investment regimes were liberalized, many state-owned
enterprises were privatized and, more generally, markets were deregulated. These four countries also faced
significant macroeconomic crises between 1995 and 2006 and, except for Argentina, have pursued broadly
prudent fiscal and monetary policies in particular since 2000. Starting in 2003, Argentina and Peru benefited
from very favorable terms of trade as a result of the boom in commodity prices. As a result, Argentina and
Peru have enjoyed high per capita growth rates between 2003 and 2006: 7.8 and 5.2 percent a year,
respectively. In contrast, in Brazil and Mexico, GDP per capita growth was modest: 2.7 and 2.8 percent a
year, respectively
Curvas de incidencia del crecimiento del Ingreso per cápita (percentiles): Argentina
(zonas urbanas). Años 2000-2006
Ingreso per cápita
Crecimiento Promedio
Promedio tasas de crecimiento
Tasa de crecimiento (en %)
40.0
35.0
30.0
25.0
20.0
4.75
PIB pc 12.8
15.0
10.3
10.0
4.7
5.0
0.0
-5.0
-10.0
1
10
19
28
37
46
55
64
73
82
91
100
Fuente: elaboración propia en base a SEDLAC (CEDLAS y Banco Mundial).
20
Curvas de incidencia del crecimiento del Ingreso per cápita (deciles):
Brasil. Años 2001-2006
Ingreso per cápita
Crecimiento Promedio
Promedio tasas de crecimiento
Linear (Crecimiento Promedio)
Curvas de incidencia del crecimiento del Ingreso per cápita (percentiles): Brasil. Años
2001-2006
Ingreso per cápita
Crecimiento Promedio
Promedio tasas de crecimiento
30.0
28.0
35.0
28.9
25.6
25.0
22.3
20.3
20.0
19.0
15.6
15.0
10.0
18.4
12.1
8.6
11.3
PIB pc 9.4
5.0
0.0
4.0
Tasa de crecimiento (en %)
Tasa de creciemiento (en %)
35.0
30.0
25.0
20.0
18.5
15.0
11.3
10.0
PIB pc 9.4
5.0
0.0
1
Fuente: elaboración propia en base a SEDLAC (CEDLAS y Banco Mundial).
10
19
28
37
46
55
64
73
82
91
100
Fuente: elaboración propia en base a SEDLAC (CEDLAS y Banco Mundial).
21
Over the last years, the income of the Brazilian poor has been growing as
faster as per capita GDP in China. By the other hand, the income of the
richest has been growing as faster as per capita GDP in Germany.
Distribution of countries according to the average per capita GDP
growth rate between 1990 and 2005
15
Average annual growth rate (%)
13
11
China
9
Brazilian bottom 10%
7
5
3
Germany
1
Brazilian top 10%
-1
-3
Haiti
-5
0
5
10
15
20
25
30
35
40
45
50
55
60
Distribution of countries (%)
65
70
75
80
85
90
95
100
Income per capita for each decile
Average income per capita
Average of income per capita growth rates
Lineal (Average income per capita)
60.0
Income per capita for each percentile
Average income per capita
Average of income per capita growth rates
50.0
80.0
40.0
36.6
28.7
30.0
27.0
23.3
20.0
20.0
17.6
17.4
16.3
24.5
14.6
9.6
10.0
Rate of growth (in %)
70.0
60.0
50.0
40.0
30.0
25.1
20.0
16.3
10.0
De
cil
e9
De
cil
e1
0
De
cil
e8
De
cil
e7
De
cil
e6
De
cil
e5
De
cil
e4
De
cil
e3
De
cil
e2
0.0
De
cil
e1
Rate of growth (in %)
50.0
0.0
0
10
20
30
40
50
60
70
80
90
100
23
100.0
70.0
60.0
50.0
50.0
37.8
40.0
31.2 26.4
30.0
22.9
19.9 16.9
20.0
9.9
De
cil
e9
De
cil
e1
0
De
cil
e8
De
cil
e7
De
cil
e6
De
cil
e5
De
cil
e4
De
cil
e3
0.0
14.8
17.5.
8
De
cil
e2
10.0
30.9
Rate of growth (in %)
80.0
78.7
De
cil
e1
Rate of growth (in %)
90.0
Income per capita for each percentile
Average income per capita
Average of income per capita growth rates
Income per capita for each decile
Average income per capita
Average of income per capita growth rates
140.0
130.0
120.0
110.0
100.0
90.0
80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0 17.5
0.0
0
10
32.3
20
30
40
50
60
70
80
90
100
24
Are the results of declining or slowly rising incomes at the top credible? A well-known fact is
that household surveys do not capture well incomes at the top. That can be due to limitations of
the representativeness of the sample at the top and/or non-sampling errors such as underreporting of incomes at the top, particularly non-wage income. In order to check the extent to
which surveys may be underestimating incomes at the top, we calculated the average income of
the two “richest” households. In 2006, the monthly total household income in current dollars was
equal to 14,779 for Argentina, 70,357 for Brazil, 43,148 for Mexico and 17, 563 for Peru. Do
these numbers “make sense”? If we had access to tabulations based on income tax returns as
they exist for all of the advanced countries, we would be able to estimate the average income of
the top .01 and .001 percent of the distribution. Regretfully, governments in Latin America do
not make this information available (we can guess why).
The Rich in Latin America: Estimates of Monthly Income
(circa 2007-2009)
Merrill Lynch High Net Worth Individuals (HNWI) and Ultra-High Net Worth Individuals (UHNWI): 2007
(in millions of dollars otherwise specified)
LATAM
WORLD
WEALTH**
POP
AVE. WEALTH*** Monthy Inc***
WEALTH
POP
AVE. WEALTH*** Monthy Inc***
HNWI (US$1 m or more in assets)*
$6,200,000
400000
$15,500,000
$64,583 $40,700,000 $10,100,000
$4,029,703
$16,790
UHNWI (US$30 m or more in assets)*
$2,200,000
4400
$500,000,000
$2,083,333 $14,300,000
$103,300
____________________________________________________________________________________________________________________________
FORBES Billionaires in Latin America
(in millions of dollars otherwise specified)
30 billionaires (US$1 b or more in assets)
$115,000
30 $3,833,333,333
$15,972,222
Carlos Slim
$35,000,000,000
$145,833,333
____________________________________________________________________________________________________________________________
Source: top panel Merrill Lynch and Capgemini World Wealth Report (2009); bottom panel Forbes, April 2009
Note: Population figures in units.
* Investible assets exclude primary residence, collectibles, consumables and consumer durables.
** total wealth for UHNWI in Latin America was assumed to be the same proportion of HNWI' wealth as for the world which equalled 35 percent.
*** in dollars per month. Author's estimates based on Merril Lynch and assumption **;
monthly returns were calculated assuming a yearly 5 percent return on investible assets.
$138,431,752
$576,799
Top income shares in Argentina and Gini Coefficient in Greater Buenos Aires, 1997-2004
top 1% income sharetop 0.1% income share
Argentina (% )
Argentina (% )
(1)
(2)
1997
1998
1999
2000
2001
2002
2003
2004
12.39
12.57
13.53
14.34
12.91
15.53
16.85
16.75
4.27
4.37
5.22
5.68
5.22
6.92
7.40
7.02
1
2
Gini Coeff. G* Gini Coeff. G Gini Coeff. G
Greater Buenos Aires
(3)
(4)
(5)
0.469
0.485
0.470
0.486
0.511
0.519
0.509
0.488
0.492
0.508
0.497
0.516
0.536
0.552
0.546
0.524
Source: Alvaredo and Piketty, Ch 4 in Lopez-Calva and Lustig. 2010.
Notes : Top shares in columns (1) and (2) are taken from Alvaredo, 2010.
G* denotes the Gini coefficient of individual income based on the Greater Buenos Aires
household survey, own calculations. All results correspond to
October surveys, except for 2003 (May). Only income earners with positive income were considered and no
further adjustments were applied. G1≈S+(1-S)G*, where S is the estimate of the top 0.1% income share.
G2≈S+(1-S)G*, where S is the estimate of the top 1% income share.
The Greater Buenos Aires households’ survey is taken as representative of Argentina as a whole (see chapter
CRUCES-GASPARINI in this volume).
0.535
0.550
0.541
0.560
0.574
0.594
0.592
0.573


There are many different factors that affect the distribution of
income over time: “… the evolution of the distribution of
income is the result of many different effects—some of them
quite large—which may offset one another in whole or in
part.” (Bourguignon et al., 2005)
Useful framework: to consider the ‘proximate’ factors that
affect the distribution of income at the individual and
household level:
1.
2.
3.
4.
5.
Distribution of assets and personal characteristics
Return to assets and characteristics
Utilization of assets and characteristics
Transfers (private and public)
Socio-demographic factors
29

Proportion of working adults as a share of the total number of
adults (and total household members) rose; partly linked to the
sharp increase in female labor force participation:1990-2006 by
18.1 p.pts in Mexico, 14.2 in Argentina, 12.0 in Brazil and 5.8 in
Peru.
Dependency ratios improved proportionately more
for low incomes.
Working adults (except for Peru) became more
equally distributed (female adults participated
proportionately more for low incomes)

Average years of schooling rose faster for the bottom quintile
than for the top quintile.
=> Distribution of education (human capital)
became more equal in all four countries
30
31
Household per capita income and its determinants
Per capita
household
income
DEMOGRAPHIC
Proportion of adults in
the household
•FERTILITY
•MARKET
•POLITICS/
INST.
•STATE
•DEMOGRAPHIC
•MARKET
Household
income per
adult
Household non-labor
income per adult
•RENTS & PROFITS
•REMITTANCES
•GOV. TRANSFFERS
Household
labor income
per adult
Proportion of working adults
•PARTICIPATION IN LABOR
FORCE
•EMPLOYMENT OPPORT
•DEMOGRAPHIC
•MARKET
•POLITICS/INST./
SOC. NORMS
•STATE
(EDUCATION)
Labor income per working
adult in the household
•WAGES BY SKILL/OTHER
•HOURS WORKED
32

Per capita household income can be written as:
y = a ( u w + o)

This identity relates changes in per capita household
income, y, to its four proximate determinants:
changes in the proportion of adults in the household, a;
changes in the proportion of working adults, u;
(iii) changes in labor income per working adult in the
household, w; and
(iv) changes in household non-labor income per adult, o.
(i)
(ii)
33
ARGENTINA (urban areas): 2000-2006
Marginal Contribution of
Source :
De mographic Factors
(adults pe r hous e hold)
In Pe rce ntage
Points
In Pe rce nt
-0.20
8
Non-labor Income
-0.68
26
Part. in Labor Marke t
-0.43
17
Earnings pe r Worke r
-1.30
50
SUBTOTAL
Inte ractive Te rm (all)
TOTAL
-2.61
100
91
-0.26
9
-2.87
100
BRAZIL: 2001-2006
Marginal Contribution of
Source :
De mographic Factors
(adults pe r hous e hold)
Non-labor Income
Part. in Labor Marke t
Earnings pe r Worke r
SUBTOTAL
Inte ractive Te rm (all)
TOTAL
In Pe rce ntage
Points
In Pe rce nt
-0.23
6.6
-1.61
45.2
-0.15
4.1
-1.57
-3.56
0.61
-2.94
44.1
100.0
120.8
-20.8
100.0
MEXICO: 2000-2006
Marginal Contribution of
Source :
De mographic Factors
(adults pe r house hold)
Non-labor Income
Part. in Labor Marke t
Earnings pe r Worke r
SUBTOTAL
Inte ractive Te rm (all)
TOTAL
In Pe rce ntage
Points
In Pe rce nt
-0.50
10.3
-0.73
-0.44
15.1
9.1
-3.19
65.5
-4.87
100.0
1.79
-3.07
158.3
-58.3
100.0
PERU: 1997-2006
Marginal Contribution of
Source :
De mographic Factors
(adults pe r house hold)
Non-labor Income
Part. in Labor Marke t
Earnings pe r Worke r
SUBTOTAL
Inte ractive Te rm (all)
TOTAL
In Pe rce ntage
Points
In Pe rce nt
-1.43
59.2
-2.29
94.4
0.08
-3.4
1.21
-50.1
-2.42
-1.65
-4.07
100.0
59.5
40.5
100.0
 Demographics: Changes in the ratio of
adults per household were equalizing,
albeit the orders of magnitude were
generally smaller except for Peru.
 Labor force participation: With the
exception of Peru, changes in labor
force participation (the proportion of
working adults) were equalizing. This
effect was stronger in Argentina.
36
Labor income (Earnings): In Argentina, Brazil, and
Mexico between 44% and 65% of the decline in
overall inequality is due to a reduction in earnings
per working adult inequality. In Peru, however,
changes in earnings inequality were unequalizing
at the household level but not so at the individual
workers’ level.
 Non-labor income: Changes in the distribution of
non-labor income were equalizing; the
contribution of this factor was quite high in Brazil
and Peru (45% and 90%, respectively).

37
=> Decline in labor inc0me
(except for Peru at the household
level) and non-labor income
inequality important
determinants of the decline in
overall income inequality (in per
capita household income)
38
Decline in labor income inequality:
 employment generated by recovery: open
unemployment fell from 14.8% in 2000 to 9.6% in
2006
 shift in favor of more low-skilled, labor-intensive
sectors as a result of the devaluation
 rise in the influence of labor unions which compresses
wages
 fading of the one-time effect of skill-biased technical
change that occurred in the 1990s
Decline in non-labor income inequality:
 more progressive government transfers: Jefes y Jefas
de Hogar program launched in 2002
39
Argentina: Returns to education
Ratio predicted wages
4.0
3.5
3.0
2.5
2.0
1.5
1.0
1980
1986
1992
Secondary/primary
1998
2006
College/primary
40
41

Decline in labor income inequality:
 About 50% accounted for by decline in attainment
inequality (quantity effect) and less steep returns -wage gap by skill narrows—(price effect). Latter
dominant. (See Gini for years of schooling and returns
by skill in next two slides)
 About 25% accounted for by decline in spatial
segmentation; especially, reduction in wage
differentials between metropolitan areas and
medium/small municipalities. Also, decline in sectoral
segmentation.
42
43
Figure 7.2: Evolution of the wage differential among metropolitan regions and different
sized municipalities: 1995-2006
35
30
32,0
30,2
29,5
28,0
26,0
25
Differential between metropolitan
regions and small country
municipalities
25,7
Differential (%)
23,0
Differential between medium and
small country municipalities
20
20,7
19,3
15,4
14,5
17,1
14,5
15
13,3
13,1
13,2
13,6
11,8
10,7
12,9
10
10,1
10,7
0
1995
10,4
9,4
1996
11,4
11,0
9,3
Differential between metropolitan
regions and medium country
municipalities
5
1997
18,4
1998
1999
6,1
6,2
6,4
4,5
2000
2001
2002
2003
2004
5,6
2005
2006
Years
45


Contribution of changes in the distribution of
income from assets (rents, interest and
dividends) and private transfers was
unequalizing but limited.
Most of the impact of non-labor income on the
reduction of overall income inequality was due
to changes in the distribution of public transfers:
changes in size, coverage and distribution of
public transfers. Bolsa Familia accounts for close
to 10 percent of the decline in household per
capita income inequality.
46

Decline in labor income inequality:
 Educational attainment became more equal and returns less steep.
 The latter seems to be associated with the decline in relative supply of
workers with low educational levels. Between 1989 and 2006, the share
of workers with less than lower-secondary education fell from 55% to
around 33%.
 It coincides with the period in which government gave a big push to
basic education.
▪ Between 1992 and 2002 spending per student in tertiary education expanded in
real terms by 7.5 percent while it rose by 63 percent for primary education.
▪ The relative ratio of spending per student in tertiary vs. primary education thus
declined from a historical maximum of 12 in 1983-1988, to less than 6 in 19942000 (by comparison, the average ratio for high-income OECD countries is close
to 2).
 Next two slides show: Gini for yrs. of schooling and returns to schooling
47
The equalizing contribution of government
transfers increased over time (both at the
national level as well as for urban and, especially,
rural households). By 2006 transfers became the
income source with the largest equalizing effect
of all the income sources considered.
 Remittances became more equalizing too but
with a smaller effect than government transfers.
 Both more than offset the increasingly
unequalizing impact of pensions.

50


The sharp rise in the role and equalizing impact
of public transfers was a consequence of a
significant policy shift in 1997, when the
government launched the conditional cash
transfer program Progresa/Oportunidades.
During 1996-2006 the size of public transfers
increased; they became more equally distributed
among recipients, and the recipients of transfers
increasingly belonged to relatively poorer
segments of the population.
51
52

Labor income inequality:
 Changes in educational structure were equalizing at
the household and individual workers levels.
 Changes in returns to education, however, were
equalizing at the individual workers level but not at
the household level. Changes in assortative matching
might have been a factor.
 Earnings gap by skill narrowed at the individual
workers level as in the other countries. Fading out of
skill-biased technical change and a more equal
distribution of education/educational upgrading.
 Next two slides show the Gini for years of schooling
and the returns to schooling.
53
56
Educational upgrading and a more equal
distribution of educational attainment have been
equalizing (quantity effect). No “paradox of
progress” this time.
 Changes in the steepness of the returns to
education curve have been equalizing at the
individual workers level (price effect). Except for
Peru, they have been equalizing at the household
level too.
 Changes in government transfers were
equalizing: more progressive government
transfers (monetary and in-kind transfers);
expansion of coverage, increase in the amount of
transfers per capita, better targeting.

57
59

Increase in relative demand for skilled labor
petered out: Fading of the unequalizing effect
of skill-biased technical change in the 1990s:
Argentina, Mexico & Peru.

Decline in relative supply of low-skilled
workers: Expansion of basic education since
the 1990s: Brazil, Mexico and Peru .
61
62
Other effects:
 Decline in spatial labor market segmentation in
Brazil.
 In the case of Argentina, the decline also
driven by a pro-union government stance and
by the impetus to low-skill intensive sectors
from devaluation. In Brazil, increase in
minimum wages.
64

In the four countries government transfers
to the poor rose and public spending
became more progressive
▪ In Argentina, the safety net program Jefes y Jefas de
Hogar.
▪ In Brazil and Mexico, large-scale conditional cash
transfers => can account for between 10 and 20
percent of reduction in overall inequality. An effective
redistributive machine because they cost around .5%
of GDP.
▪ In Peru, in-kind transfers for food programs and
health. Also access to basic infrastructure for the poor
rose.
65
In the race between skill-biased technological
change and educational upgrading, in the last ten
years the latter has taken the lead (Tinbergen’s
hypothesis)
 Perhaps as a consequence of democratization and
political competition, government (cash and inkind) transfers have become more generous and
targeted to the poor
 Will the momentum towards a less unequal Latin
America continue?

66

Despite the observed progress, inequality continues to
be very high and the bulk of government spending is
not progressive.

The decline in inequality resulting from the educational
upgrade of the population will eventually hit the
‘access to tertiary education barrier’ which is much
more difficult to overcome: inequality in quality and
‘opportunity cost’ are high and costly to address.

Making public spending more progressive in the future
is likely to face more political resistance (entitlements
of some powerful groups).
67
68
69

Argentina: current monetary (no imputations
for owner’s occupied housing); assumes it’s after
taxes for wage earners & before for other
(assumption); after monetary government
transfers (in survey).

Brazil: current monetary plus imputed value of
income in kind (no imputations for owner’s
occupied housing); before taxes (assumption)
and after monetary government transfers (in
survey and assumption).
70

From 1930s to 2000s: inverted U to a U

Alvaredo and Piketty, chapter 4, in LopezCalva and Lustig (2010)