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
0521812909c13 CB929-Bulmer 052181290 9 October 6, 2005 10:44
582 Stephen Haber
necessarily a bad thing to do. The problem was that, in practice, they
tended to invest a disproportionate share of their resources in large firms
that lost money. In Mexico, for example, Nacional Financiera typically
took a minority shareholder position in companies to which it was also
making loans. If the firm performed well, Nacional Financiera’s stake would,
over time, become purely nominal. If the firm performed badly, however,
Nacional Financiera typically bought out the other shareholders. From the
point of view of the government, many of these enterprises were simply too
big to fail, or had unionized workforces that were politically crucial. This
policy of bailouts not only prevented the bank from deploying its capital
more effectively elsewhere, it also encouraged moral hazard: knowing that
they would be bailed out, manufacturers undertook enterprises of doubtful
profitability.
101
The relentless logic of subsidies ultimately produced government own-
ership of a wide range of manufacturing industries, financed of course
through government development banks. Government ownership tended
to cluster in the production of so-called strategic goods, such as steel, basic
petrochemicals, and fertilizers. Government ownership in these industries
was rationalized on the basis that these were high-risk, capital-intensive
industries characterized by large minimum efficient scales of production
and capital indivisibilities. Although this is true, it was also the case that
these firms tended to produce inputs for other industries. Given the fact
that there were no private shareholders, the prices of the firms’ outputs
were determined on political grounds: government-owned firms wound
up charging below-market prices. In other words, taxpayers subsidized the
privately owned enterprises that purchased the steel, electricity, chemicals,
and other inputs produced by so-called parastate firms.
By the late 1960s and early 1970s, it was becoming increasingly clear that
protectionism and “state-led development” had created endless distortions.
Consumers paid high prices for badly made goods. Entrepreneurs found
it difficult to evaluate projects and make investment decisions because so
much of their calculus turned on political, rather than market, signals. Gov-
ernments were providing increasingly larger subsidies, via the exchange rate
system, government banks, and parastate firms – and the benefits of many of
those subsidies were being captured by foreign-owned manufacturing firms.
By the 1970s, Latin America’s governments were faced with a set of hard
decisions: end subsidies and market distortions by opening up to foreign
101
Enrique C
´
ardenas, “The Process of Accelerated Industrialization in Mexico, 1929–1982,” 195.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c13 CB929-Bulmer 052181290 9 October 6, 2005 10:44
The Political Economy of Industrialization 583
trade, or take advantage of the boom in international capital markets to
subsidize a system that had run into decreasing marginal returns. With
the notable exception of Chile, governments throughout Latin America,
“democratic” and authoritarian alike, opted to buy time by borrowing
abroad. The outcome of that experience is well known: the debt crisis of
the 1980snot only produced an economic and social disaster, it also brought
about a dissolution of the political coalitions that had supported protec-
tionism and subsidization since the 1890s. The result was the end of trade
protection, the privatization of government-owned firms, and the shifting
of Latin American manufacturing increasingly toward export markets. The
full details of that transformation are the subject, however, of the chapter
in this volume by Victor Bulmer-Thomas.
CONCLUSIONS
Beginning in the 1890s, the largest economies of Latin America conducted a
100-year experiment in which they built substantial manufacturing sectors
behind barriers to trade. This experiment was not carried out because public
officials were ideologically committed to industrialization or because of
economic theories that challenged free trade. Indeed, the growth of large-
scale manufacturing industry predated structuralist theories by sixty years.
Rather, Latin America’s industrialization was kick-started by an endogenous
process of economic development, the roots of which were found in the
growth of the so-called export economy. Once industrialists got a foothold,
they lost no time in seeking tariff protection. Governments granted this
protection because it came at virtually no political cost: the one class that
stood to lose from trade protection, consumers, had no political voice. The
result was that by 1914, all of the larger countries of Latin America had
become increasingly protectionist and were undergoing a rapid process of
industrial growth.
The next six decades brought long periods of prosperity and short periods
of stagnation or decline. Over the long term, however, industry outgrew the
rest of the economy. Circa 1975, when this style of industrial development
had reached its zenith, the manufacturing sector’s share of Latin American
GDP was comparable to that in the United States.
Over time, governments played larger roles in the process of industrial
growth. They devised increasingly complicated systems designed to simul-
taneously tax imported manufactures and subsidize the cost of imported
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c13 CB929-Bulmer 052181290 9 October 6, 2005 10:44
584 Stephen Haber
inputs to the production of those goods. They also became financiers,
partners, and even direct owners of manufacturing firms.
This does not mean, however, that Latin American post-1930 industrial
development was “state led” or that it was somehow fundamentally dif-
ferent from the industrial growth of the pre-1930 period. The process of
Latin American industrial development was always ad hoc. Governments
did not grant trade protection because they were committed to a particular
vision of industrial development. They did it because they were rewarding
politically powerful constituents, some of whom were industrialists and
some of whom were industrial unions. They also did it because balance
of payments crises often presented them with difficult political choices:
suffer the negative fallout of devaluation (the effects of which are obvi-
ous and immediate) or impose restraints on trade (the effects of which are
more difficult to detect and understand for the average voter). Similarly,
governments did not create industrial finance banks and buy equity in
manufacturing firms because they had some ideological commitment to
government ownership. They did it because government financial policies
repressed the development of banking systems and securities markets that
could mobilize adequate capital for industrial development. Charting a
different course in terms of financial market regulation would have come
at sizable short-run political and economic costs: governments would have
had to adhere to stable fiscal policies, create independent central banks, and
undo the segmented monopolies and oligopolies that characterized Latin
American banking. Governments, therefore, chose the second-best solu-
tion of government financing of private firms. Once this started, however,
government loans became like trade protection: it was a permanent prize to
be captured. Finally, governments did not become direct owners of man-
ufacturing firms because public officials were convinced of the superiority
of state ownership. They did it because many private firms, especially ones
that the government had lent money to, were simply too big to fail. In short,
governments increasingly intervened in support of industry because it was
in the self-interest of industrialists, industrial unions, and public officials
to do so. This was an outcome of the ad hoc nature of Latin American
industrial development, not the product of carefully thought-out plans or
strategies. This was as true of Latin America’s industrial development from
1890 to 1930 as it was from 1930 to 1980.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
14
POVERTY AND INEQUALITY
miguel sz
´
ekely and andr
´
es montes
INTRODUCTION
Poverty and inequality have been deeply rooted in Latin American societies
since the early colonial era. With the arrival of Spanish and Portuguese
conquerors, the diverse resources of the region in terms of land and other
factor endowments were carved up in ways that favored the few at the
expense of the many. Later, the intensification of colonization brought
the development of economic structures that cemented these inequalities
in place. Both the diversity of factor endowments in the region and the
way in which they have been distributed have therefore played a heavy
hand in predisposing Latin America toward certain paths of development –
paths characterized by wide-ranging degrees of inequality in wealth, human
capital, and political power.
1
The severity of inequality has varied across the continent, depending
on many factors. Generally speaking, however, patterns of production,
land distribution, and schooling that Latin America has followed have
consistently fostered high degrees of inequality. Many of these patterns
The authors would like to thank Adriana Argaiz for her excellent research assistance. The opinions
expressed in this chapter are the authors’ and should not be attributed to the institution to which they
are affiliated.
1
Kenneth L. Sokoloff and Stanley L. Engerman, “History Lessons Institutions, Factor Endowments,
and Paths of Development in the New World,” Journal of Economic Perspectives 14, 3 (Summer 2000):
217–32.
585
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
586 Miguel Sz
´
ekely and Andr
´
es Montes
were formed in the early colonial era, when indigenous populations were
denied rights while a select number of elites of European descent were given
the lion’s share of wealth, land, and power to administer their conquered
territories.
The early differences in the extent of inequality persisted and were
reproduced in the way economic and social institutions evolved over
time; these institutions, in turn, exerted their own influence on the path
of economic development. In cases of extreme inequality, institutions
favored elites and limited the access of much of the population to eco-
nomic opportunities. Having attained high socioeconomic status, elites
were thus in an advantageous position to maintain that status over time,
with the rest of society paying the price of underdevelopment. Over-
all, in regions in which wealth, human capital, and political influence
were highly concentrated in the hands of the elites, competition was
restricted.
During the past century, most Latin American countries, in one way
or another, pursued policies aimed at addressing these social inequalities.
These policies were not created in a vacuum but were strongly shaped by the
macroeconomic environment. As the macroeconomic environment evolved
over the course of the century in fundamental (as well as unpredictable
and unforgiving) ways, these changes were mirrored in the evolution of
development policies.
This chapter takes a close look at the historical progress of poverty and
inequality in order to grasp the reasons behind their stubborn persistence.
Only with this knowledge can sound recommendations for future poli-
cies be made. The chapter is organized as follows. The next section sum-
marizes the available evidence on poverty and inequality in the region,
underscoring the limitations of those data. The third section provides
an overview of the main explanations of long-run evolution of poverty
and inequality given in the literature. The fourth section outlines the
conceptual framework used to analyze the evolution of income-earning
assets that form the basis of this chapter. Using this framework, sec-
tions four, five, and six take an in-depth look at the behavior of assets
in Latin America over the course of the past century, delving into issues
surrounding their accumulation, utilization, and returns. The following
section draws lessons on the historical roles that assets have played in
Latin American development in the past, and the last section provides the
conclusion.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
Poverty and Inequality 587
HISTORICAL TRENDS: THE EVOLUTION
OF POVERTY AND INEQUALITY DURING
THE TWENTIETH CENTURY
data limitations
An important challenge in analyzing the evolution of poverty and inequal-
ity in Latin America is data quality and availability. Clearly, any study
attempting to assess the progress of poverty and inequality over long peri-
ods should be aware of the shortcomings of the existing information – and,
consequently, of the assumptions made by its analysis – and strive to tackle
this issue.
There are at least five different sources of information on income for
Latin America: tax records, social security records, establishment statistics,
population censuses, and household surveys. Still, there is no reliable source
on income because most of these sources present various problems, includ-
ing coverage, accuracy, detail, and the specificity of sectors. Because of these
problems, it is normally agreed that the ideal source for estimating poverty
and inequality is household surveys. However, most countries only started
implementing these surveys at the beginning of the 1970s and many are
not even comparable within countries.
Recently, there has been a renewed interest in the use of these kinds of
income and expenditure surveys because they present an invaluable source
for examining changes in income distribution and observing household
behavior relating to expenditure and different forms of income (e.g., finan-
cial assets, wages, bequests). Nevertheless, the quality of surveys containing
more or less detailed information on expenditure and income varies across
a wide spectrum. Furthermore, various methodological differences make
the comparison between surveys even more difficult.
It is essential that inequality analysis take three factors into account,
which are, as described in Deaton: (1)the characteristics of the sample,
(2) differences in survey quality and coverage of population groups, and
(3) differential coverage of income sources and geographic areas.
2
Regarding
the first factor, among the most important elements that have to be taken
into account when using income and expenditure surveys are the sampling
2
Deaton Angus, The Analysis of Household Surveys: A Microeconometric Approach to Development Policy
(Baltimore, 1997).
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
588 Miguel Sz
´
ekely and Andr
´
es Montes
frame and size. This is relevant because weights can overstate or diminish
the importance of population subgroups, leading to inaccurate results. The
same is valid for the sample size of a survey because the larger the sample
size, the smaller the standard error of any estimate.
The quality and coverage of the surveys are also of vital importance for
making cross-country comparisons. In this respect, income misreporting
and underreporting are the most common problems. Additionally, house-
hold surveys use samples in which some groups are likely to be underrepre-
sented. In particular, wealthier individuals are not normally sampled; hence,
the true level of inequality is usually higher than what surveys inform.
Analyzing poverty and inequality using cross-country comparisons leads
to several problems, such as the definition of economies of scale at the
household level, the treatment of missing or zero incomes, accounting
for regional price variations, and so on. Sz
´
ekely and Hilgert engage in a
detailed discussion of these issues and argue that country rankings may be
strongly influenced by parameter choices rather than genuine distributive
differences.
3
An additional complication is that there are only a few countries in Latin
America for which data are available to document the evolution of poverty
and inequality over long periods of the twentieth century. The earliest
comparative study of poverty trends in Latin America and the Caribbean
was done by Altimir in the late 1970s.
4
The author identifies considerable
variations regarding poverty levels in urban households. These ranged from
5 percent in Argentina to 38 percent in Colombia. On the other hand, rural
poverty registered even more dramatic variations. It varied from 19 percent
in Argentina to 73 percent in Brazil. Altimir’s estimates show a generalized
improvement in the 1960s and 1970s that halted, and in some cases reversed,
in the 1980s.
As is discussed later, the 1980s and 1990swere years of economic volatility
and increasing poverty and inequality in the region. There is no country
in Latin America – based on the limited data available – for which it can
confidently be said that income inequality improved during the 1990s. The
main factor driving the lack of distributive progress was wage inequality.
Increasing income inequality was driven by the widening wage gaps. Fur-
thermore, macroeconomic shocks such as the debt crisis and economic
3
Sz
´
ekely, Miguel, and Marianne Hilgert, “What’s Behind the Inequality We Measure? An Investigation
Using Latin American Data” (OCE Working Paper Series no. 409,Inter-American Development
Bank, 1999).
4
Oscar Altimir, La dimensi
´
on de la pobreza en Am
´
erica Latina (Santiago, 1979).
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
Poverty and Inequality 589
20
30
40
50
60
South Asia
Eastern Europe
Developed East Asia Africa Latin America
Gini Coefficient
World Average =40
India
Pakistan
Sri Lanka
Poland
Yugoslavia
Romania
Hungary
Australia
New Zealand
USA
Thailand
China
Korea
•
Zaf
Gambia
Senegal
Zambia
Madagascar
Uganda
Tanzania
Nigeria
Ghana
Egypt
Norway
Sweden
Canada
Belgium
Finland
Mozambique
Tunisia
Peru
Venezuela
Argentina
El Salvador
Nicaragua
Colombia
Regional
Average
Brazil Paraguay
Figure 14.1.Gini index for countries with good-quality data for the 1990s.
Sources: Miguel Sz
´
ekely and Marianne Hilgert, “What’s Behind the Inequality We Mea-
sure? An Investigation Using Latin American Data” (OCE Working Paper Series no. 409,
Inter–American Development Bank, 1999); Klaus Deininger and Lyn Squire, “A New Data
SetMeasuring Income Inequality,” WorldBank Economic Review 10, 3 (September 1996):
565–91.
reforms of the 1980s also played a role in driving a larger wedge between
the wages of some groups vis-a-vis others.
latin america: the most unequal region in the world
Figure 14.1 shows the Gini index for a set of Latin American countries,
in addition to countries worldwide, for which information around 1995 is
available. We note that the average world Gini index is approximately 0.4
(the Gini ranges from 0 to 1, with 1 being the highest and most unequal
value). On average, South Asia had the lowest inequality level, with a
Gini index of 0.3, followed by Eastern Europe. Next are the developed
countries, with an average of 0.32.Although these three regions have similar
average inequality indices, it is interesting to note that the variance is greater
among developed countries. Although all Eastern European and South
Asian countries for which data are available are found to have lower-than-
world-average inequality levels, rich countries such as the United States,
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
590 Miguel Sz
´
ekely and Andr
´
es Montes
Australia, and New Zealand are found to have inequality levels that are very
similar to the world average. The situation of the East Asian countries is
comparable, with inequality levels below the world average, though with a
higher variance. Furthermore, the latter region includes Thailand, which
has a relatively high level of inequality.
Africa is fifth in terms of regional averages. Its average inequality level
exceeds the world average, although with substantial variability among
countries. This area includes South Africa (coded Zaf ), which has the
highest inequality level of all countries in the world, and Egypt, which has
arelatively low inequality level. Therefore, in this case, the average tells us
little about the situation of individual countries.
Last, as shown in the figure, Latin America has the highest average Gini
index, with a value of approximately 0.54.Itisthe only region in which all
countries in the sample exceed the world average inequality level. It is also
interesting to note that South Africa’s high inequality index is followed by
a substantial number of countries, all Latin American.
the evolution of poverty and inequality
in latin america, 1970–2000
It has been made clear that many limitations hinder comparisons of income
distribution and poverty in Latin America. Nevertheless, some efforts have
been made to collect within-country comparable information on income
and expenditure in order to analyze the evolution of inequality and poverty
at the regional level for the last twenty-five years. Londo
˜
no and Sz
´
ekely
gathered a large number of household surveys from the region to analyze
progress from 1970 to 1995.
5
In their analysis, the authors face the standard
comparability problems between countries, as well as the underreporting
for some income variables. However, their analysis allows for a study of the
evolution of poverty and inequality for the largest countries in the region,
which is described herein.
Figure 14.2 illustrates the economic performance after the 1960s.
6
Extreme and moderate poverty declined, while inequality also registered a
5
Juan Luis Londo
˜
no and Miguel Sz
´
ekely, “Persistent Poverty and Excess Inequality: Latin America
1970–1995,” Journal of Applied Economics 3, 1 (May 2000): 93–134.
6
For the purposes of this chapter, two definitions of the poverty line are used: a 1985 Purchasing Power
Parity (PPP)-adjusted dollar-a-day line to measure extreme poverty, and US $2 1985 PPP-adjusted
dollars per capita per day for moderate poverty. Regarding the discussion on the definition of poverty
lines, there are significant differences in consumption patterns across countries and through time,
which make it difficult to establish common criteria. Other scholars, such as Montek Ahluwalia,
Nicholas Carter, and Hollis Chenery, “Growth and Poverty in Developing Countries,” Journal of
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c14 CB929-Bulmer 052181290 9 October 6, 2005 10:49
Poverty and Inequality 591
53
54
55
56
57
58
59
70
71
72
73 74
75
76
77
78
79 80 81 82 83 84 85 86 87 88 89 90 91 92 93
94
95
0.1
0.2
0.3
0.4
0.5
PPP Adjusted GDP Gini Index Moderate Poverty Extreme Poverty
Gini Index Proportion of Poor
Figure 14.2.Inequality and poverty in Latin America: 1970–1995 (GDP per capita).
Source: Juan Luis Londo
˜
no and Miguel Sz
´
ekely, “Persistent Poverty and Excess Inequality:
Latin America 1970–1995,” Journal of Applied Economics 3, 1 (May 2000): 93–134.Purchasing
PowerParity states that exchange rates between currencies are in equilibrium when their
purchasing power is the same in each of the two countries. This means that the exchange
rate between two countries should equal the ratio of the two countries’ price level of a
fixed basket of goods and services. When a country’s domestic price level is increasing (i.e.,
a country experiences inflation), that country’s exchange rate must depreciate in order to
return to PPP.
constant decline that ended in the early 1980s. The early 1980swere char-
acterized first by recession and later by stagnation, whereas the 1990s show
arecovery. Also, income distribution improved substantially from 1970 to
1982 (the Gini index was reduced by four points), whereas the 1980s coin-
cided with a sharp deterioration in income distribution (the Gini peaked
at 58.3 in 1990). Even though the 1990swere years of recovery, extreme
and moderate poverty did not show any improvement during this period.
Inequality increased substantially.
Development Economics” 6 (1979): 299–341;Michael Bruno et al., “Equity and Growth in Developing
Countries: Old and New Perspectives in the Policy Issues,” in Vito Tanzi and K. Y. Chu, eds., Income
Distribution and High Quality Growth (Cambridge, MA, 1998); 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, focus on producing internationally comparable estimates
for each of the world regions.