Bulmer-Thomas V., Coatsworth J., Cortes-Conde R. The Cambridge Economic History of Latin America: Volume 2, The Long Twentieth Century
Подождите немного. Документ загружается.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
592 Miguel Sz
´
ekely and Andr
´
es Montes
Table 14.1. Poverty measures for Latin America, 1970–1995
Moderate poverty Extreme poverty
Year
Head
count
ratio
Poverty
gap
FGT(2)
index
Million
of poor
Head
count
ratio
Poverty
gap
FGT(2)
index
Million
of poor
1970 43.618.711.2 117.119.26.54.051.4
1975 36.215.59.3 110.114.84.53.444.9
1980 27.511.16.993.810.52.62.636.0
1985 28.311.35.8107.811.42.91.343.3
1990 35.216.49.7147.917.46.33.673.1
1995 33.115.49.2 152.516.26.13.774.5
Note: Foster, Greer, and Thorbecke index with parameter value = 2.
Source: Juan Luis Londo
˜
no and Miguel Sz
´
ekely, “Persistent Poverty and Excess Inequality:
Latin America 1970–1995.”
As argued by Londo
˜
no and Sz
´
ekely, the 1970swere characterized by an
expansion of the incomes of the poor and the middle classes at the expense
of the richest 20 percent of the population.
7
The 1980s show the opposite:
the income share of the poorest 90 percent decreased considerably while
the income share of the richest 10 percent expanded by about 10 percent.
The 1990s show yet another picture, with the poorest and the richest deciles
losing part of their shares, and the middle class expanding theirs.
The economic expansion in Latin America during the 1970s helped
reduce not only inequality but also poverty. The data observed in Table 14.1
indicate that between 1970 and 1980, both moderate and extreme poverty
declined from 117.1 million to 93.8 million people and from 51.4 million
to 36 million, respectively. For the most part, this trend and improvement
in income distribution are attributed in various studies to the high growth
rates experienced during those years. The poverty gap and the Foster, Greer,
and Thorbecke (FGT) index, which is more sensitive to the changes taking
place at the bottom of the distribution, also declined during the 1970s,
indicating that there were not only fewer poor but that these were less poor
than before.
8
In order to assess the impact on poverty of economic expansion and
changes in inequality, Londo
˜
no and Sz
´
ekely decompose changes in poverty
7
See note 5.
8
James Foster, Joel Greer, and Erik Thorbecke, “A Class of Decomposable Poverty Indices,”
Econom
´
etrica 52 (1984): 761–6, designed a family of poverty measures that are given by the
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
Poverty and Inequality 593
into growth and distribution effects (Figure 14.3 presents the results).
Poverty declined during the 1970s because of the high growth rates and
more equitable income distribution; however, this was not the case for the
1980s. Most of the rise in poverty during that decade can be attributed to
changes in inequality and not, contrary to conventional wisdom, only to
the economic stagnation observed during those years.
Despite the fact that GDP per capita increased by almost 6 percent in
real terms between 1990 and 1995 in Latin America, poverty levels appeared
to remain stubbornly persistent. These results are, in part, attributable to
the low impact of growth on poverty reduction because of the lack of
distributive progress.
Table 14.2 shows changes in absolute poverty for the period between 1970
and 1995,bycountry. During the 1970s, Latin America achieved a decrease
in the numbers of moderately poor of 23.3 million as well as in extreme
poverty of 15.5 million. In particular, Brazil, Colombia, and Mexico made
important strides, although most of this reduction came more from the
upper ranks of the poor in Colombia and Mexico than from the poorest of
the poor as was the case in Brazil. Nevertheless, the economic crisis of the
lost decade of the 1980sbrought a considerable increase in the number of
poor, both at the moderate and the extreme levels.
Between 1980 and 1990, moderate poverty increased by 54.1 million
people, and 37.1 million individuals slipped into the category of extreme
poor. The most alarming increase in the numbers of extreme poor happened
in Brazil, where 23.8 million people became poorer during this period.
Other countries where poverty increased significantly in contrast to the
previous decade were Guatemala and Peru, where the number of moderately
following formula:
P
α
=
1
n
y
i
< p
p − y
i
p
α
where P is the poverty line, α is the sensitivity measure, n is the total population, yi is the income of
the i-th individual, and p − yi is the poverty gap. For this index, widely known as the FGT index,
α can take different values, giving place to various poverty measures. In the case α = 0, the index
is the head-count ratio, which indicates the share of the population whose income or consumption
is below the poverty line. In the case α = 1,weobtain the poverty gap, which provides information
regarding how far off households are from the poverty line. In other words, the poverty gap captures
the mean aggregate income or consumption shortfall relative to the poverty line across the whole
population and is obtained by adding up all the shortfalls of all the poor and dividing it by the total
population. In the case α = 2,weobtain the poverty severity, also denominated FGT(2), which takes
into account not only the distance separating the poor from the poverty line but also the inequality
among the poor, that is, a higher weight is placed on those households who are farther away from
the poverty line.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
-50
-40
-30
-20
-10
0
10
20
30
40
Growth
-22
-2
Redistribution
-15
21
Residual
-2
1
Total Change
-39
1970-1980
1980-1985
1990-1995
(%) Change in the Head-Count Ratio
0
0
-2
6
28
Figure 14.3.D
ecomposition of the change in poverty into growth and redistribution
effects for Latin America:
1870–1995.
Source:
Juan Luis Londo
˜
no and Miguel Sz
´
ekely, “Persistent Poverty and Excess Inequality: Latin America
1970–1995.”
594
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
Table 14.2.
Changes in absolute poverty in Latin America, by country
Moderate poverty
Extreme poverty
Country
1970–1980 1980–1990 1990–1995 1970–1980
1980–1990 1990–1995
Latin America
−23.30
54.11
4.56
−15.46
37.14
1.14
Bahamas
−0.01
0.01
0.01
−0.01
0.01
0.01
Brazil
−13.12
34.75
1.24
−10.10
23.75
0.51
Chile
0.38
1.47
−1.04
0.23
0.21
0.62
Colombia
−4.21
0.67
−0.21
−1.51
0.57
−0.52
Costa Rica
0.19
0.11
0.01
0.23
0.00
−0.03
Dominican Republic
−0.51
1.34
−0.14
−0.61
0.86
−0.20
Guatemala
0.43
2.38
0.33
0.17
1.57
0.20
Honduras
0.52
0.97
0.40
0.21
0.49
0.24
Jamaica
0.10
−0.12
−0.01
0.01
−0.15
−0.03
Mexico
−2.23
−1.25
4.00
−0.12
1.49
0.42
Panama
0.00
0.49
−0.07
−0.16
0.41
−0.06
Peru
1.41
3.18
−0.70
−0.33
0.86
0.02
Venezuela
−0.88
1.10
0.10
−0.14
0.72
−0.01
Other Countries
−5.37
9.00
0.63
−3.32
6.36
1.2
∗
Millions.
Source: Juan Luis Londo
˜
no and Miguel Sz
´
ekely, “Persistent Poverty and Excess Inequality: Latin America
1970–1995,”
Journal of Applied
Economics
31 (May 2000):
93–134.
595
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
Table 14.3.
Poverty and inequality in Latin America,
1989–2000
Country
Index
1989–1990 1991–1992 1993–1994 1995 1996 1997
1998 1999 2000
Argentina
Gini Index
0.4771
0.4935
Head-Count Ratio
18.40
17.90
FGT(2)Index
2.27
2.26
Bolivia
Gini Index
0.5449
0.5323 0.5274 0.5877 0.5890
0.6014
Head-Count Ratio
65.63
63.40 63.60 62.14 62.34
61.36
FGT(2)In
dex
20.19
18.27 17.37 24.84 24.18
27.40
Brazil
Gini Index
0.5728 0.5952 0.5911 0.5907 0.5919 0.5901 0.5847
Head-Count Ratio
48.26 49.68 44.66 41.55 41.25 41.92 41.
26
FGT(2)In
dex
15.03 15.51 12.91 11.98 11.91 11.52 11
.11
Chile
Gini Index
0.5470 0.5220 0.5558
0.5638
0.5587
Head-Count Ratio
32.37 19.78 22.70
18.32
16.11
FGT(2)In
dex
6.12 2.77 3.69
2.91
2.60
Colombia
Gini Index
0.5670 0.6038 0.5697
0.5756 0.5679 0.5620
Head-Count Ratio
42.39 44.67 38.79
38.37 37.79 39.37
FGT(2)In
dex
10.73 11.94 8.83
10.78 9.94 10.14
Costa Rica Gini Index
0.4596 0.4598 0.4549 0.4570
0.4589 0.4612
Head-Count Ratio
35.89 34.23 29.20 28.70
30.86 30.47
FGT(2)In
dex
8.87 8.07 6.11 6.11
6.33 5.88
Dominican Republic Gini Index
0.4810
0.4778
Head-Count Ratio
38.13
34.56
FGT(2)In
dex
8.05
5.32
Ecuador
Gini Index
0.5600
0.5616
Head-Count Ratio
49.53
47.98
FGT(2)In
dex
17.15
15.41
596
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
El Salvador Gini Index
0.5052
0.5195 0.5589 0.5455
Head-Count Ratio
58.60
61.25 63.98 63.98
FGT(2)Index
14.54
15.84 21.34 21.29
Honduras
Gini Index
0.5703 0.5489
0.5284 0.5908 0.5852 0.5843
Head-Count Ratio
77.20 75.94
76.30 74.73 74.85 75.25
FGT(2)I
ndex
32.25 31.48
30.15 35.43 34.88 35.42
Mexico
Gini Index
0.5309 0.5341 0.5361
0.5276
0.5377
Head-Count Ratio
19.74 16.17 15.34
21.22
21.17
FGT(2)I
ndex
3.28 2.29 2.13
3.62
4.15
Nicaragua
Gini Index
0.5669
0.6024
Head-Count Ratio
70.67
72.68
FGT(2)I
ndex
28.88
28.14
Panama
Gini Index
0.5625
0.5602
0.5755 0.5652 0.5631
Head-Count Ratio
47.75
47.81
43.53 38.05 36.61
FGT(2)I
ndex
17.04
16.43
14.37 11.92 10.80
Paraguay
Gini Index
0.5700
0.5692 0.5942
Head-Count Ratio
52.09
51.00 61.12
FGT(2)I
ndex
17.96
19.48 23.28
Peru
Gini Index
0.4643 0.4832
0.5055
0.4933
Head-Count Ratio
41.86 43.98
43.23
42.43
FGT(2)I
ndex
10.33 10.82
11.40
12.22
Uruguay
Gini Index
0.4064 0.4319
0.4209
0.4300 0.4388
Head-Count Ratio
23.15 19.55
16.61
11.69 13.59
FGT(2)I
ndex
2.70 2.91
2.09
1.46 1.78
Venezuela
Gini Index
0.4396
0.4288 0.4669
0.4863 0.4705 0.4675
Head-Count Ratio
12.55
8.68 15.23
17.95 18.87 20.63
FGT(2)I
ndex
2.13
1.08 2.53
3.48 3.46 4.08
Source: Miguel Sz
´
ekely,
The 1990sinLatin America: Another Decade of Persistent Inequality, but with Somewhat
Lower Poverty
(Mexico City,
2001).
597
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
598 Miguel Sz
´
ekely and Andr
´
es Montes
-0.004
0.001
0.006
0.011
Trend for the Gini Index
-0.02 -0.015 -0.01 -0.005 0 0.005 0.01 0.015 0.02
Trend for the Head Count Ratio
Figure 14.4.Trends in poverty and inequality in Latin America during the 1990s.
poor increased by 2.4 million and 3.2 million, respectively. The 1990sproved
to be a period of recovery, except for Mexico, where more than 4 million
individuals slipped into the ranks of the moderately poor because of the
1994–5 economic crises.
Sz
´
ekely finds that despite the stabilization and macroeconomic growth
enjoyed during the second half of the 1990s, the region’s well-being did
not improve.
9
Both inequality and the number of poor increased between
the early and late 1990s, not only in the region as a whole but also on a
country-by-country basis. As shown in Figure 14.4,poverty and inequality
grew hand in hand during the 1990s. The figure plots the trend in the Gini
index on the vertical axis versus the trend for the head count ratio, showing
a clear positive relationship between increases in inequality and increases
in poverty.
Furthermore, Table 14.3 shows how inequality increased considerably in
a number of countries. The Gini index grew by four points between 1993
and 1998 in El Salvador, Nicaragua, Peru, and Venezuela, and by about
six points in Bolivia during the decade. In other countries such as Chile,
Colombia, Ecuador, Mexico, and Panama, inequality showed no trend.
9
Miguel Sz
´
ekely, The 1990sinLatin America: Another Decade of Persistent Inequality, but with Somewhat
Lower Poverty (Mexico City, 2001).
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
Poverty and Inequality 599
However, despite economic growth and a stabilization of prices, inequality
was not reduced in most of the region.
The same can be said for poverty. In some countries, such as Paraguay and
Venezuela, poverty increased sharply during the 1990s. Paraguay underwent
a 10 percent increase between 1995 and 1999, whereas in Venezuela the
increase represented by the head count ratio was of the order of almost 8 per-
cent over the course of the decade. Moreover, the FGT index confirms that
the increase in poverty not only increased the number of poor in those
countries, but also made the poorest worse off than before. This is also the
case for other countries, such as El Salvador, Mexico, and Peru, where even
though poverty remained more or less constant, the FGT index increased,
signaling a worsening of the situation of the poor. This could explain
the relationship between inequality and poverty for those countries where
poverty did not increase while inequality did. That is, although poverty did
not rise in absolute terms, the poorest suffered further welfare losses.
Despite a strong increase in inequality in Latin America and the persis-
tently high inequality levels, there were a number of exceptions to these
trends. Brazil, Chile, and Uruguay saw a reduction in the numbers of poor
during the 1990s, as illustrated in Table 14.3.Brazil and Uruguay reduced
poverty by almost 10 percent during the decade, and in Chile poverty
halved, from 32.4 percent at the beginning of the 1990sto16.1 percent in
1998.Still, it is too early for optimism because the only success story of
overall economic progress in Latin America seems to have been the case of
Chile.
ECONOMIC GROWTH, POVERTY,
AND INEQUALITY
Why has Latin America been plagued by persistently high inequality and
poverty levels? This section offers and critiques some of the traditional
explanations for the long-term prevalence of these phenomena, particularly
the Kuznets inverted U hypothesis, and presents a brief assessment of the
growth-inequality-poverty nexus.
the kuznets hypothesis
The widespread adoption of the belief that economic growth is the main
vehicle for poverty reduction and improvement in living standards spawned
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
600 Miguel Sz
´
ekely and Andr
´
es Montes
many different theories regarding poverty and inequality. Among the most
well known and influential is the one elaborated by Nobel laureate, Simon
Kuznets, in 1955.Kuznets’s studies of distributional changes during the
development process hinged on population shifts from traditional to mod-
ern activities, specifically the dynamics of the transition during the initial
stages of development to a modern industrial sector with higher wages.
Kuznets concluded that at low income levels, economic growth tended
to create more inequality. Once a critical threshold level of income was
reached, however, further economic growth tended to reduce inequality.
In other words, Kuznets predicted that inequality would grow during the
first stages of development and industrialization, but decline consistently
after reaching a turning point. This hump-shaped relationship between per
capita income and income inequality has come to be known as the Kuznets
inverted U hypothesis.
The empirical background for the Kuznets hypothesis came from
research on inequality indicators for the United Kingdom, West Germany,
and the United States. During the 1950s, when the research took place,
inequality was falling in those countries, after having risen earlier. The
economic mechanism thought to underlie these nonlinear trend lines was
the transfer of labor from low-productivity (medium inequality) sectors to
high-productivity (low inequality) sectors – namely, the transfer from agri-
culture to industry. The changes in income distribution observed suggested
that inequality between these two sectors was substantially greater than the
inequality within them and, given the data available for that period of time,
that conclusion seemed unassailable.
Kuznets therefore formally introduced the idea of a link between inequal-
ity and development by acknowledging that total inequality is the result
of the combined income distributions of the rural and urban populations.
Kuznets argued that the income gap between these two groups presents
the following two issues: (a) the average per capita income of the rural
population is usually lower than that of the urban, and (b) within-group
inequality for the rural population is somewhat narrower than that for the
urban population.
Based on these two statements, he reached the following conclusions:
first, that all other conditions being equal, the increasing weight of urban
populations means an increasing share for the more unequal of the two
component distributions; and second, the relative difference in per capita
income between the rural and urban populations does not necessarily
drift downward in the process of economic growth. Indeed, there is some
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
Poverty and Inequality 601
evidence to suggest that it is stable at best, and, in fact, tends to widen
because per capita productivity in urban economic pursuits increases more
rapidly than in agriculture. If this were so, inequality in the total income
distribution would increase during the beginning stages of the industrial-
ization process.
After a certain period of time, the theory predicts, inequality will decline
as rural–urban migration drives down the remunerations received by the
expanding modern sector while driving up those received in the traditional
sector. The effects of these two processes narrow the rural–urban income
gap.
Comprehensive tests of the Kuznets hypothesis have been recently made
possible by the compilation, in 1995–6,ofthe highly regarded Deninger-
Squire
10
International Inequality Database, which contains 682 observa-
tions of Gini coefficients and quintile shares for 108 countries. Deninger-
Squire concluded that these data provided little support for an inverted
Urelationship between levels of income and inequality when tested on a
country-by-country basis, with no support for the existence of a Kuznets
curve in approximately 90 percent of the countries researched. In light of the
much more plentiful information available to today’s empirical researchers,
one might expect the support for the Kuznets hypothesis to have largely
eroded.
In fact, with the availability of improved econometric techniques, the
debate on the relation between inequality and economic growth has contin-
ued. For instance, Anand and Kanbur argue that if the econometric speci-
fication is improved, the inverted U shape relationship between inequality
and growth vanishes.
11
Bruno et al., Li et al., and Ravallion and Chen
use an improved data set and argue that there is no systematic relation
between the Gini inequality index and GDP per capita growth.
12
Never-
theless, according to Barro, inequality and growth do follow the inverted
U shape relationship suggested by Kuznets. De Janvry and Sadoulet and
Morley arrive at the same conclusion by using a data set that includes only
10
Klaus Deininger and Lyn Squire, “A New Data Set Measuring Income Inequality,” WorldBank
Economic Review 10, 3 (September 1996): 565–91.
11
Sudhir, Anand, and S.M.R. Kanbur, “The Kuznets Process and the Inequality-Development Rela-
tionship” Journal of Development Economics 40 (1993): 25–52.
12
Michael Bruno et al., “Equity and Growth in Developing Countries: Old and New Perspectives
in the Policy Issues”; Hong Yi Li et al., “Explaining International and Intertemporal Variations in
Income Inequality,” The Economic Journal 108, 446 (1998): 26–43;Martin Ravallion and Shaohua
Chen, “What Can New Survey Data Tell Us about Recent Changes in Distribution and Poverty?”
The World Bank Economic Review 11, 2 (1997): 357–82.