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
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
272 Richard Salvucci
Summerhill’s social savings calculations are a piece of this argument,
ranging from 6 to 11 percent of GDP in 1913 on freight, and a surprisingly
large 4 percent for passenger service (versus, say, something on the order of
1.5 percent for Mexico). Summerhill’s conclusions, although cautious, point
to an obvious fact. In Latin America, railroads mattered for growth and, in
some places, they mattered a great deal. That this finding emerges should
perhaps not surprise, for the tradition of structuralist economic thinking
so long current in the region has always emphasized sluggish adjustments
in supply as a primary obstacle to economic development, and public, as
opposed to private enterprise as its agent. These putatively large returns to
railroad construction in the presence of insecure property rights, political
upheavals, and undeveloped capital markets go some way to explaining why
railroads as, so to speak, engines of economic growth, went nowhere for
many years: the divergence between private and social rates of return was
simply too great. Domestic investors, capital-constrained, and, perhaps,
alert to the true risks of investing, were unable or unwilling to seize the
opportunities offered. Foreign investors, inevitably cautious, regarded some
of these markets with great trepidation. For most of the nineteenth century,
or at least after default in 1827, the risk premium on Mexican bonds was
rarely less than 10 percentage points above the yield on British consols, and
when the yield on Mexican bonds did converge to the consol rate in 1887,it
was for the first time in nearly a quarter century. For Argentina, significantly,
the perceived risk was considerably lower, and, according to calculations by
Cort
´
es Conde, country risk was usually less than 10 percentage points after
1864,except in moments of severe strain. Hence, Argentina’s earlier start
in railroad construction, and a larger network in 1877 than Brazil, Chile,
Peru, and, of course, Mexico, which had almost nothing. Because the bulk
of railroad construction that occurred after 1870 depended on capital from
Great Britain, the United States, and France, the attitude of foreign investors
toward the principal railroad-building countries was, essentially, a decisive
consideration.
Colombia was somewhat different. Here, recent estimates by Mar
´
ıa
Te resa Ram
´
ırez for 1927 place the social savings for Colombia between
2 and 4 percent of GDP, by far the smallest estimate thus far for
Latin America. Perhaps the simplest explanation is that the measure of
social savings depends on the existence and cost of alternative means of
transportation. Because railroads were constructed with some delay in
Colombia, road and highway transport was already available, reducing the
share of coffee transported by railways and raising the responsiveness of
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
Export-Led Industrialization 273
shippers to fluctuations in the price of rail transportation. And finally, the
railroad in Colombia did not form a true network, as it did in Argentina, or
to a lesser extent, in Brazil, Chile, and Mexico. Most were built to link up
to the Magdalena River, and the effect they had in knitting together mar-
kets across the country – one with very difficult geography – was blunted.
For this reason, then, railroads in Colombia highlight the conditions under
which rail development was important for economic growth in Latin Amer-
ica. In Central America, too, the railroad has been considered less important
from the standpoint of social savings than elsewhere, even if the region was
said to have experienced a “transportation revolution” in the second half
of the nineteenth century. Other places, such as Ecuador, had little or no
railroad development, and in Venezuela, most railroad building took place
between 1880 and 1900, with little occurring thereafter.
An issue that is closely related to the impact of railroads on economic
development, but perhaps requiring separate comment, is the transfer of
public or communal lands to private control that often occurred in response
to or in conjunction with the extension of railways. In Bolivia, Mexico,
and Peru, at various times from the 1870s and 1880s (in Mexico) to the late
nineteenth and first decade of the twentieth century (Bolivia and Peru), the
economic growth experienced by regions first linked by rail to new markets
provided both the incentive and the means to expropriate the lands of free
villages. In Mexico, the spread of the rail network and the privatization
of public lands by survey companies were clearly related, if only as two
expressions of a general policy, the goal of which was to promote Mexico’s
commercial development. More concretely, there is visually compelling
correspondence between those parts of the country (in the north) where the
survey companies were very active and the construction of rail trunk lines.
The figures are striking. During the Porfiriato, nearly 11 percent of Mexican
territory was given over to the land-survey companies in compensation.
In Central America, the history of the expropriation of village lands
after 1870 is well known, and in Guatemala, for instance, the wholesale
transfer of “unproductive” lands given over to maize being switched to
the increasingly profitable cash crop of coffee has been studied in great
detail. Yet, whatever one makes of these developments, it is very difficult to
credit them with the character of an ostensibly commercial phenomenon –
that resources and output that were previously mobilized and allocated
outside a market framework were now more “efficiently” employed. A
persuasive analysis by H
´
ector Lindo Fuentes of price dispersion among
subsistence commodities such as maize shows substantial uniformity in
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
274 Richard Salvucci
prices across villages in El Salvador, which Lindo Fuentes takes to be evi-
dence of the market embeddedness of transactions. The same held true
in Colombia, where peasant producers brought tobacco, sugarcane, cot-
ton, and cacao to markets throughout the countryside. Whatever else elites
accomplished during this burst of entrepreneurial energies, they did not
bring the discipline or efficiency of the market to bear on these agricul-
tural economies for the first time. Thus, it seems that a simple equation
of the spread of private property in land with material progress confuses
changes in distribution with changes in output, or at the very least, over-
states the growth in product, and obscures the deterioration in the distri-
bution of wealth. Generalization is difficult because the impact of these
policies on vast expanses of largely empty land on the northern Mexican
steppes would be very different from what occurred, in, say, the coffee
plantations of the densely populated regions in highland Guatemala, or in
Santander and Antioquia in Colombia. Yet even in Mexico, pressure on
village lands became increasingly powerful as the commercial potential of
crops such as sugarcane came to bear on village lands in Morelos during the
Porfiriato.
Nor was the phenomenon restricted solely to the indigenous heartlands
of Mesoamerica. McGreevey, although fixing no dimensions on the pro-
cess, concludes that the collective appropriation of village, public, and
church lands in the second half of the nineteenth century in Colombia was
extremely significant. Here, the transfer of such lands was linked to the
rise of an export commodity, tobacco, in the lowlands and to the expan-
sion of cattle ranching in the highlands for the domestic market. More-
over, because the scale requirements of tobacco production were large, the
spread of tobacco cultivation raised income and land values dramatically,
but concentrated them as well. Catharine LeGrand emphasizes that much
of Colombia’s national territory was, strictly speaking, not titled until well
into the nineteenth century and that the dimensions of the transfer of
public lands into private hands was enormous, something like 2.4 million
hectares of land between 1870 and 1930.InArgentina, too, as Carl Solberg
has written, much new land came on to the market in the 1860s and
1870s, when the provinces rolled back the Indian frontier and acquired
vast new expanses of public land, much of which ended up in the hands
of the cattlemen who comprised the backbone of the rural oligarchy, the
whole undertaking perhaps symbolized by General Julio Roca’s “conquest of
the wilderness” in 1879, and substantial sales of public land would continue
well after that date.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
Export-Led Industrialization 275
Although these pressures (and usurpations) reflected changing valuations
that producers placed on land, they also highlight the importance of popula-
tion for economic growth. Historically, efforts to dispossess villages of land
in Colombia, Mexico, and Central America were also efforts to drive their
populations into a labor market by rupturing the ancient nexus between
Indians and peasants, and their lands. There was nothing new about these
efforts in the period after 1870 save, perhaps, the aggressiveness with which
they were undertaken under the stimulus of increased international com-
modity demand. Labor rather than land had, by and large, become the
scarce factor of production in the massive depopulation that followed what
Alfred Crosby termed the “Columbian Exchange.” Where attachment to
village life made peasants reluctant to enter the rhythms of commercial
production, or where population centers lay distant from the demands of
highland mines or lowland plantations – as they did, for example, in Peru (in
cotton growing) and Central America (in bananas) – extra-market devices
featuring a variety of names had become common means of acquiring labor.
In an economic sense, these were structural problems, a mismatch between
labor supply and demand. In the period after 1870,ascommercial opportu-
nities and pressures intensified, new means of acquiring or enlarging labor
forces became widespread. Because African slavery faced extinction in the
Spanish Caribbean in 1870 by the Moret Law, or in Brazil by abolition in
1888, the response was essentially limited to the use of free labor, although
freedom in this context must be interpreted circumspectly. By the same
token, because the natural rate of population growth in Latin America was
relatively slow, immigration frequently became a crucial source of labor,
especially in Argentina, where inexpensive labor was a critical element in
the success of pampean agriculture. It is, therefore, necessary to spend some
time considering the importance of population for economic development
in this period.
In theory, all changes in production could be attributed to changes in
the size of the labor force (an increase in total output), or in productivity
per member of the labor force (an increase in per capita output), or both.
Practically speaking, population increase is a way of estimating growth in
the labor force because, as things stand, the data from this period are not
likely to yield much more. It is also, perhaps, useful to consider the quality
of labor and, particularly, the extent to which education did or did not
augment productivity. Recent findings suggest that Latin America, in this
regard, had also fallen well behind more advanced nations of the Atlantic
economy by the late nineteenth century. Finally, where it is possible to
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
276 Richard Salvucci
Table 7.5. Population growth rates
(percent per year)
Argentina, 1889–1930 3.1
Brazil, 1890–1929 2.1
Chile, 1870–1930 1.6
Colombia, 1843–1912 1.5
Mexico, 1895–1930 0.75
Peru, 1896–1928 2.2
Venezuela, 1870–1930 1.1
Source: Based on author’s calculations.
identify them, changes in demographic structure may also shed light on
larger patterns of economic growth, as they crucially do in Argentina.
As the data in Table 7.5 suggest, population growth in Latin America
in this period was highly uneven. For purposes of discussion, it is possi-
ble to distinguish between those countries whose populations grew quite
slowly, those that increased at a more typical rate, and those whose demo-
graphic growth was truly rapid. The median rate of growth, given the
uncertainties in the data, seems to be about 2 percent or so, with Mexico
and Colombia at the lower end of the scale, Argentina at the upper, and
Brazil, Chile, Peru, and Uruguay at various intermediate levels. There
is naturally a reasonably strong association between the rate of popula-
tion growth and GDP growth in this period (see Figure 7.1). The scatter
Growth of Per Capita Income
Growth of Population
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0 0.5 1.0
1.5 2.0 2.5
3.0
3.5
Figure 7.1.Growth of population and per capita income.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
Export-Led Industrialization 277
diagram displays the association for those countries for which we have
data, and it is strongly positive, with a simple correlation between popu-
lation and output growth of 0.92.Perhaps more important, there is also a
positive association between population growth (in logs) and per capita
income growth (the difference between the logs of output and popu-
lation growth), with a simple correlation of 0.78. Countries with more
rapidly growing populations also had higher income growth per head.
The result is not as unusual as it might seem. In the long run, with the
extension of rail networks and, hence, the frontiers of commercial agricul-
ture, more people had more, rather than less, land and capital with which
to work, so diminishing returns did not occur. Juan Bautista Alberdi’s
(1810–84) celebrated aphorism, “to govern is to populate,” makes perfect
sense in this context. Nor is the result dependent strictly on the data we have
collected. A similar exercise linking per capita GDP in 1913 to the growth
rate of population between 1870 and 1913 for eight countries (Argentina,
Brazil, Chile, Colombia, Mexico, Peru, Uruguay, and Venezuela) in 1990
international dollars drawn from Angus Maddison’s The World Economy
produces similar results (see Figure 7.2). Population growth raised rather
than lowered per capita income in the period between 1870 and 1930.
Where natural increase substantially comprised the bulk of population
increase, as it did in Mexico, demographic change was painfully small.
Robert McCaa has estimated that the annual growth rate of the population
Population Growth Rate
GDP Per Capita
4000
3500
3000
2500
2000
1500
1000
1.0 1.5 2.0
2.5 3.0 3.5
500
Figure 7.2. GDP per capita (1990 international dollars) in 1930 versus population growth
rate 1870–1913.
Source: Maddison, The World Economy, 193, 195.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
278 Richard Salvucci
there between 1876 and 1910 was no more than 1.71 percent, and after 1900,
the rate slowed to 1.09 percent. The Revolution, according to McCaa,
produced a demographic disaster, perhaps the worst since the sixteenth
century, with population falling by one million (total population in 1910 was
about fifteen million). Total growth from 1910 through 1930 was no more
than 1.4 million, or less than the increase for 1900–10.Demographic growth
per se could have therefore done little or nothing to raise total output, and
Mexico was not, and had never been, a magnet for foreign immigration.
So in some portions of Mexico, such as the Yucat
´
an peninsula, which were
remote from major population centers or from poles of economic growth
to which workers could migrate (the proximity of northern Mexico to
the United States is the usual example), relative labor scarcity in the face
of growing commercial demands produced a history of repression on the
henequen plantations. Yet what was true for henequen was not invariably
the case for other commodities. Sugarcane planters in Morelos may have
despoiled villages of their lands, but they also substituted capital for labor in
producing and processing sugar. Over time, this industry actually required
not more but relatively less labor, a fact reflected in both census data and
in the nature of agrarian protest during the Mexican Revolution. Still,
Mexico’s overall record of growth between 1895 and 1930 suggests that slow
population increase was an obstacle to economic growth.
Colombia, too, experienced relatively slow population growth, with an
overall average between 1870 and 1930 of something just over 1 percent
per year. Catharine LeGrand suggests that immigration into Colombia
was not large, and that the creation of an adequate labor force under the
pressure of export expansion was inevitably difficult, particularly when it
involved peasants who migrated into frontier regions. Hacendados from
western Cundinamarca and Tolima imported workers from the eastern
highlands through a system of labor contracting called enganche.But there
were also widespread conflicts in the interior, in western Colombia, and
along the Caribbean coast between peasants settled on public lands, known
collectively as colonos, and land entrepreneurs interested in developing lands
for cattle ranching, the production of forest products, or coffee plantations.
Again, there is no reason to assume that an increase in the export surplus
measures the increase in domestic output because it is clear that the peasant
population was not entirely outside the market economy to begin with.
Rather, the effort to appropriate previously untitled (but hardly vacant)
lands was designed to mobilize a labor force adequate to the needs of an
export economy in which a rural proletariat was absent and population
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
Export-Led Industrialization 279
growth was slow. Thus, LeGrand concludes, “in order to generate a labor
force (italics mine) ...entrepreneurs logically sought to turn the public lands
into private properties ...The fact that the population of Colombia was
much smaller . . . accounts in part for the difficulties large proprietors had
in acquiring labor.”
7
Chile, too, experienced slow population growth, at a rate of about 1.6
percent per year. Nevertheless, Arnold Bauer has suggested that economic
expansion under the stimulus of external demand, first in wheat and later
in the export mining sector of nitrates and copper, was not significantly
hindered by labor shortages experienced elsewhere in Latin America because
“Central Chile ...was an area with plentiful if idle hands.” Indeed, Patricio
Meller estimates that nitrate exports grew at something over 6 percent per
year between 1880 and 1930, and Carmen Cariola and Oswaldo Sunkel
point to the “spectacular” population growth in the Norte Grande that
facilitated the growth of nitrate exports.
Peru experienced somewhat more rapid demographic increase. The
annual rate of population growth between 1876 and 1940 was in the neigh-
borhood of 2 percent yearly, and after the termination of Chinese inden-
tured immigration in 1874, immigration did not figure importantly. In
Peru, too, the rise of sugar plantations on the north coast in the valleys of
Lambayeque, Jequetepeque, Chicama, and Santa Catalina presented major
problems of labor recruitment, a difficulty addressed to some extent by the
rise of the infamous enganchadores,orlabor contractors, who facilitated
migration from the Sierra to the coast. Peter Klar
´
en refers to the “critical
shortage of labor” in the emerging sugar industry, concluding that an ade-
quate supply of labor had long been a major problem for the region’s sugar
planters.
Like Peru, Brazil’s population grew at roughly 2 percent between 1872 and
1940,orfrom about 9.9 million to 41.2 million people. But a few distinct
features of demographic change require comment. Unlike Mexico and Peru,
Brazil experienced substantial immigration, with the total flow between
1870 and 1920 amounting to about 3.3 million persons, of whom 1.4 mil-
lion, or some 42 percent, were of Italian origin, followed by lesser but still
significant numbers of Spaniards, Japanese, and Arabs. There was a pro-
nounced tendency for immigration to increase in the wake of the abolition
of African slavery. Immigration exceeded 100,000 for the first time in 1888,
7
Catherine LeGrand, Frontier Expansion and Peasant Protest in Colombia, 1830–1936 (Albuquerque,
NM, 1986), 38, 40.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
280 Richard Salvucci
and something in excess of a million immigrants were to follow in the
1890s, with more than 200,000 in 1891 alone. As Nathaniel Leff observes,
the connection between the abolition of slavery and the rise of immigration
was not circumstantial but purposeful because the coffee planters of S
˜
ao
Paulo pressured the central government and the province of S
˜
ao Paulo to
subsidize European immigration and to ensure an elastic supply of labor in
the face of abolition. Thus, in Leff’s view, the abolition of slavery in Brazil
did not have the same effects on the labor supply that it did in the U.S.
South after the Civil War. As Leff put the matter, labor markets in Brazil
differed because “shifts in the demand curve for labor determined shifts in
the supply schedule of workers,” something that seems quite different from
what occurred in Mexico and Peru. As a result, Brazilian income growth
in 1890–1929 was a respectable 4.3 percent per year. Still, the emergence
of large estates on the Brazilian coffee frontier was accompanied by the
expropriation of settlers who were expelled onto public lands farther in the
interior in a process that LeGrand explicitly compares with Colombia.
THE IMPACT OF DEMAND
Variations in production, prices, and employment in the short run are typ-
ically attributed to fluctuations in demand because its determinants can
change relatively quickly. Political upheavals and monetary disturbances
can occur suddenly and with little warning, the result of the vagaries of
regime changes, an untoward mining discovery (e.g., gold in California
and Australia in 1849), or legislation affecting the supply and demand of
precious metals under the gold and silver standards (e.g., that undertaken
by the Latin Monetary Union between 1865 and 1878). At the same time,
variations in international commodity prices, which could to some extent
be taken as given (the major exception, of course, would be Brazilian cof-
fee and the valorization schemes, wherein Brazil, with its large share of
the global market, sought to support the price of coffee by reducing its
supply), could vary rapidly and unexpectedly from year to year as well, pre-
senting producers with a rapidly evolving picture of demand. Indeed, to an
extent, the leitmotiv of discussions of the commodity cycle is the variabil-
ity of prices, something that Bulmer-Thomas has termed the “commodity
lottery.” Yet a composite picture of these forces, by definition, would be
difficult to provide because they are as varied as the countries whose history
we have chosen to analyze. But some attempt at generalization is necessary
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
Export-Led Industrialization 281
to make the story intelligible and coherent rather than episodic, disjointed,
and anecdotal.
Oneofthe best-known discussions in the literature of economic develop-
ment is the question of the terms of trade, or the relative prices of exports. In
essence, the net barter terms of trade is the measure most commonly com-
puted because it is the one for which data are most readily available, and,
hence, the framework within which comparisons are routinely, if some-
times facilely, made. The subject is of particular relevance to the history of
Latin America because the terms of trade literature is traditionally associ-
ated with the work of the distinguished Argentine economist, Ra
´
ul Prebisch
(1901–85). At a basic level, an “improvement” in the terms of trade – that
is, an increase in the price of exports relative to imports – represents an
increase in real income for the exporting country because the relative price
of exports improves. “Deterioration,” on the other hand, implies that real
income has fallen. In the history of Latin America, the notion of deterio-
rating terms of trade has sometimes served as a justification for the rejec-
tion of liberal economic prescriptions that rely on free trade as an engine
of growth because under such conditions, trade could ostensibly become
impoverishing. By extension, attempts to alter comparative advantage, such
as import substitution industrialization, seem logical if demand for a com-
modity – and hence its relative price – is falling. This explains the historical
importance so often ascribed to the analysis of the net barter terms of
trade.
Whereas accounts of the terms of trade are frequently carried out at a
high level of aggregation, it makes some sense to proceed at the country
level when the commodity composition of exports differs, as it did in
Latin America, even if variations in income in the developed (importing)
countries imposed some measure of correlation on export price movements
in Latin America. And, given the limits of the data, a striking heterogeneity
in country experience in the period 1870 through 1930 is apparent. There
was no overall tendency for the net barter terms of trade to deteriorate,
even if each country did experience episodes of deterioration, and even
if some countries, most notably Mexico and Peru, did experience falling
terms of trade after 1870. This is especially important when one considers
that studies of the period before 1870 reveal much the same thing. In other
words, for roughly the first century of their independence, the nations
of Latin America, considered on the whole and individually, experienced
growing rather than declining demand for their tradables if the relevant
metric is the net barter terms of trade.