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
292 Richard Salvucci
regime, fixed or flexible, depends on a country’s individual circumstances.
Certainly, during 1870 through 1930,itappears that Latin America had
no ideal choice. Flexible exchange rates did have their advantages, partic-
ularly for those countries whose exports and industrialization were aided
by the depreciation of silver. But, as Casasus complained in Mexico, the
price in volatility was very high and it may well be, following contemporary
thinking, that the volatility of the international environment created suf-
ficient uncertainty to act as a significant constraint on economic growth.
In any event, price stability and macroeconomic stability are not the same
thing, and macroeconomic stability became increasingly difficult to come
by between 1870 and 1930.
Certainly, we find only limited evidence that factors affecting aggregate
supply and demand ultimately acted as obstacles to growth. The railroad
had truly massive effects on supply, and not only of exportables. In country
after country, a revaluation of resources to reflect changed productivities
followed in the wake of railroad building. The railroad was not merely a
highway to external dependency. And the construction of such infrastruc-
ture also made a decisive beginning possible for industrialization, especially
in Mexico and Brazil, although the true bulk of industrial modernization
would await the end of the Great Depression and World War II. But a start
toward structural transformation had been made, export age or no.
As far as demand goes, we find no real evidence that there was a gen-
eralized deterioration in the terms of trade, although deterioration in the
net barter terms of trade did occur in some instances. More impressive was
the response of the export sector in bringing larger volumes of exporta-
bles to market and, hence, raising the income terms of trade, which points
in the direction of productivity increase. As was Bulmer-Thomas, we are
left with an impression that it was the speed or generality through which
productivity gains were transferred to the rest of the economy that was the
critical issue – the mechanism that equalized productivity across sectors that
proved sluggish. Again, this may have stemmed from uncertainties in the
international environment, especially those that impeded the formation of
or interfered with consistent market signals. Slow population growth may
well have acted as an obstacle to economic growth in some regions, but
the study of the connection between demographic change and economic
growth in Latin America is still in its very early stages.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c08 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:32
8
THE DEVELOPMENT OF
INFRASTRUCTURE
william r. summerhill
Between 1870 and 1930, Latin American nations experienced an unprece-
dented expansion of physical infrastructure. The economies of Latin
America, most of them laggards by the standards of North Atlantic eco-
nomic performance, began to grow with a new-found rapidity, as invest-
ment in modern transportation increased during the late nineteenth cen-
tury. For many nations of the region, this era became one of genuine
economic transformation. Increasing exports of primary commodities, ris-
ing imports of capital goods, the expansion of activities drawing directly
and indirectly on overseas investment, the rising share of modern man-
ufacturing in output, and a generalized increase in the pace and scope
of economic activity were all tied closely to the timing and character of
the region’s infrastructural development. The causation in this process ran
both ways. Infrastructure development fostered economic growth, as mar-
keting and transaction costs fell. Economic growth led to new sequences of
infrastructure investment, as the demand for improvements increased, the
costs of provision fell, and new bottlenecks emerged, creating commercial
and political pressures for follow-on investment. In 1850,nomajor coun-
try in Latin America possessed any of the trappings of modern transport
technology. Half a century later, most of them did, and had experienced
considerable economic change as a result.
This chapter draws on the findings of studies over the last several
decades that have sought to isolate, with increasing empirical and ana-
lytical rigor, the implications of infrastructure development for Latin
American economic growth. The core component of infrastructure in this
era necessarily involved the transportation revolution. Transport invest-
ment raised incomes and output by integrating markets and providing
293
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c08 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:32
294 William R. Summerhill
new opportunities for specialization in agriculture and manufacturing. Rail-
way construction, port improvements, and the eventual spread of motor
roads collectively propelled economic growth across much of the region by
the early twentieth century. Sequences of increasing infrastructural devel-
opment and higher levels of economic activity fed back into each other,
accelerating, in several cases quite dramatically, the rate of material progress.
Infrastructure that facilitates the production and exchange of pri-
vate goods occupies a strategic position in the capital stock of modern
economies. By infrastructure, economists usually mean transportation facil-
ities, though the term at times is employed to encompass publicly owned
capital in the energy sector. The transformative potential of investments in
infrastructure is widely recognized by economic analysts and stems from
infrastructure’s ability to raise the productivity of existing resources, expand
the supply of new resources, and foster both extensive and intensive eco-
nomic growth. Empirical estimates of the economic returns to the for-
mation of infrastructure capital, even in the late twentieth century, are
astoundingly high.
1
Investments in the Latin American infrastructure in an
earlier period, the critical decades of transformation before 1930, turn out
to be no less important to the process of economic growth.
Beyond highlighting the principal characteristics of infrastructure in
Latin America and sketching its development, this chapter argues that
the accelerated pace of infrastructure investment after 1870, particularly in
steam locomotion, provided the chief impetus to the transition to modern
rates of economic growth in much of the region. Two qualifications are
in order from the outset. First, assertions about the primacy of one factor
necessarily collide with a number of problems, including one of endo-
geneity. Latin America’s export earnings paid for imported railroad capital,
yet the railroad was required to break through the barriers of geography
that restricted export growth, and the availability of export earnings did
not necessarily ensure that railroads would be built or were built where
they were most needed. The resolution of this causal puzzle depends on
the circumstances of each particular case. Second, the particular impact
of any single infrastructure project is an empirical question that requires
1
Alicia Munnell, “Infrastructure Investment and Economic Growth,” Journal of Economic Perspec-
tives 6:4 (Autumn 1992): 189–98;Peter J. Wylie, “Infrastructure and Canadian Economic Growth
1946–1991,” Canadian Journal of Economics, 29:Special Issue: Part 1 (April 1996): 350–5;and
Edward M. Gramlich, “Infrastructure Investment: A Review Essay,” Journal of Economic Literature 32
(September 1994): 1176–96 provide summaries of the main empirical results of assessments of infras-
tructural development in the high-income North American economies in the latter decades of the
twentieth century.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c08 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:32
The Development of Infrastructure 295
careful, theoretically informed investigation. A railway extension, the con-
struction of a feeder roadway, or the financing of a port improvement are all
actions that involve an investment of resources with implications for both
allocative efficiency and the distribution of wealth. These effects can vary
considerably across specific projects, regions, decades, and countries, and
accurately documenting them for all of Latin America requires far more
research among primary sources than has been done to date. The asser-
tions in this chapter about the impact of infrastructure are thus necessarily
incomplete, yet they provide a set of useful characterizations against which
the findings of future work will have to be compared. Although durable
economic expansion cannot be attributed to a single cause alone, it now
seems unlikely that any other technological or institutional innovation was
more important in the transition to economic growth in Latin America
before 1930.
Moreover, the history of internal improvements in Latin American
nations has a very strong policy component that is not yet adequately
studied. Political factors reigned supreme in the processes of infrastructure
concession, subsidy, and regulation. Yet these remain poorly understood.
By the late nineteenth century, transport improvements became a cen-
tral feature of the political-economic landscape throughout Latin America.
That infrastructure policies interacted with other features of the polity is
not surprising and did not necessarily work the same way from country to
country. Because policy is a product of political preferences and the insti-
tutions that aggregate those preferences into political outcomes, the formal
institutions governing decisions on transport subsidy and regulation were
important. Even in economies where a pure price system exists, the large
minimum-efficient scale of infrastructure investment, uncertainty over the
current or future level of demand for infrastructure services, and worries
on the part of investors over the prospect of excessive regulation or expro-
priation mean that infrastructure investment may languish at levels much
lower than are warranted by prevailing conditions, and less than necessary
for future growth. In short, because of the public-goods character of much
infrastructure, well-functioning markets may get things wrong, resulting in
a poor allocation of infrastructure projects. Perceptions of the public inter-
est, even in Latin American polities where the politically relevant public
was limited by a highly restricted franchise or semi-authoritarian govern-
ment, created heavy pressure for intervention to remedy market failures
and better serve the particular pecuniary interests of politically influential
groups. Market interventions by the state often succeeded in ameliorating
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c08 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:32
296 William R. Summerhill
such failures but typically did so in a way skewed to suit particular interests
stemming from the thirst for profits. Unfortunately, chronicles and narra-
tive descriptions of infrastructural successes and failures in Latin America,
although informative, cannot substitute for careful analysis of the insti-
tutions governing the choice of economic policies. Dissecting the politics
of infrastructure requires a much greater understanding of actual policy
making – including both the political incentives of policy makers and
the institutions in which they operate – than presently exists for most of
nineteenth-century Latin America.
2
The upshot of the experience with infrastructure development between
1870 and 1930 is best summarized along the following lines. Many of the
areas in Latin America with strong economic potential did not possess
extensive arrays of roads, networks of canals, or even interlocking systems
of natural inland waterways on the eve of the railway age. Freely navigable
rivers were abundant in some regions, such as portions of the Southern
Cone, but virtually absent from others, as in the case of interior Mexico
and northern South America. There was, of course, no mechanism that
guaranteed that the waterways that were navigable would be located where
human populations could make best use of them commercially. Regions
that suffered from backward pre-rail transport systems, limited inland nav-
igation, and thus often extraordinarily high long-distance transport costs
stood to gain considerably from infrastructure investment.
In these areas, railroads were truly necessary for both the export sector
and domestic markets. By the late nineteenth century, railway construc-
tion provided a specific form of infrastructure capital that offered greater
efficiency and affordability of overland transportation for goods and pas-
sengers alike in Latin America. By providing for low-cost inland movement,
steam locomotion effected a tremendous increase in the supply of trans-
port services to Latin American economies before 1930. This increase was
complemented by innovations in ocean shipping and the associated cre-
ation of port facilities. Infrastructure investment embodied in transport
improvements freed up capital and labor previously employed in produc-
ing transport services, permitting these factors of production to be used in
other activities, raising the productivity with which they were employed,
bolstering the growth of both agriculture and industry, and thereby raising
2
One effort in this direction is William Summerhill, “Institutional Determinants of Railroad Subsidy
and Regulation in Imperial Brazil,” in Stephen Haber, ed., Political Institutions and Economic Growth
in Latin America: Essays in Policy, History, and Political Economy (Stanford, CA, 2000).
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c08 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:32
The Development of Infrastructure 297
national income. Where the demand for transport services was sufficient to
compensate the cost of infrastructure investment, the net benefits were sub-
stantial. In this way, railways served as an important stimulus to economic
growth in Latin America. Indeed, the sheer magnitude of the aggregate
economic gains several Latin American countries derived from railway cap-
ital was unprecedented by international standards. By 1930, when the pace
of railway expansion had already slowed, transport services along nascent
motor roadways built on the higher levels of economic performance pro-
vided by the previous generations of infrastructural investment and, in
many areas, served to sustain the process of intensive economic growth
that was underway.
This chapter proceeds by providing an overview of four categories of
infrastructure investment in Latin America before 1930.Itfirst discusses
the relatively limited development of terrestrial transportation before the
advent of the steam-powered locomotion. It then traces the course of railway
expansion and provides a summary of the main findings on the railway’s
impact. In reviewing the consequences of railway development, it focuses
on the cases of Argentina, Mexico, Brazil, and Colombia. Three factors
motivate this restricted examination. First, it is impossible to do justice to
even the basic details of all Latin American countries in a single chapter.
Second, these countries include the most important cases in terms of the
extent of their infrastructure investments. Indeed, by 1900, the first three
of these countries accounted for fully 75 percent of all railway track in Latin
America. Finally, to date, rigorous econometric assessment of the railway’s
impact remains limited to these four cases.
Although railway expansion was by far the most important component of
infrastructural development, it did not proceed in isolation from advances
in other forms of infrastructure. The third section of the chapter, therefore,
takes up the topic of port improvements in the age of export growth and
mass migration. The rise of motor roads in the early twentieth century
briefly occupies the conclusion.
ROADWAYS BEFORE STEAM
Large-scale investments in infrastructure came to Latin America relatively
late. In partaking of the transportation revolution in a late-comer fashion,
Latin America in effect largely skipped the turnpike era of internal improve-
ments. Even in the best of cases, only very limited improvements to inland
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c08 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:32
298 William R. Summerhill
roads preceded railway development. Descriptions of transport conditions
in Latin America by contemporaries in the first part of the nineteenth
century provide historians with a sense of the strength of the brake on
economic growth arising from backward pre-rail transport systems. Early
roads throughout Latin America reflected in varying degrees the combina-
tion of two features: pre-contact indigenous routes of movement and the
location of commercial centers that appeared during the colonial era. In a
good many cases, colonial roads had their origins in indigenous trails when
the latter proved useful to the evolution of commerce and trade.
3
In north-
eastern Brazil, trails from coastal ports evolved into crude, unpaved roads
for animal traffic and carts that carried sugar for export. Elsewhere, roads,
at times paved with stone, were constructed to facilitate the movement and
monitoring of gold shipments from Minas Gerais to the coasts of Rio de
Janeiro and S
˜
ao Paulo.
4
In Mexico, cart roads connected the capital both
to the port of Veracruz and to the silver-mining regions, and in Central
America, much of the colonial trade in indigo moved along simple trails.
Throughout the Andes, footpaths frequently proved resistant to any trans-
formation into roads, given the difficult terrain. In all too many areas, the
costs of shipping goods were wildly exorbitant, and only those products
with relatively high value-to-weight ratios were worth moving over any
appreciable distance. An additional part of the challenge posed by slow,
inefficient transportation in nineteenth-century Latin America was a fre-
quent lack of physical security and the dissipation of public funds that could
have been used to maintain and improve the highways. In much of main-
land Spanish America, the factionalized struggles to capture control of the
state made it difficult to consistently pursue and support policies designed
to enhance communications. Moreover, warfare, brigandage, and hostile
indigenous populations further increased the risk and thereby exacerbated
the costs of overland movement.
5
A geography highly unfavorable to inland movement was the chief cul-
prit in the high costs of transport that plagued nineteenth-century Latin
America. The steep mountain ranges of Mexico, Brazil, Colombia, and
the other Andean regions physically partitioned large expanses, providing
3
John H. Coatsworth, Growth Against Development: The Economic Impact of Railroads in Porfirian
Mexico (DeKalb, IL, 1981), 17–18.
4
Richard Momsen, Rates Over the Serra do Mar (Rio de Janeiro, 1964).
5
Stephen Haber, Industry and Underdevelopment: The Industrialization of Mexico, 1890–1930 (Stanford,
CA, 1989), 13–15, 24–5;Bert J. Barickman, “‘Tame Indians,’ ‘Wild Heathens,’ and Settlers in Southern
Bahia in the Late Eighteenth and Early Nineteenth Centuries,” The Americas 51:3 (1995): 325–68.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c08 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:32
The Development of Infrastructure 299
natural barriers that were extremely difficult to breach and leaving mar-
kets geographically fragmented. Lacking navigable rivers, mid-nineteenth-
century Mexico had long relied on a combination of wagons, animals, and
human porters to move freight through the country. The Spanish colonial
government, seeking to improve basic transport conditions, had grafted
aroad network on top of the pre-contact indigenous transport routes in
order to link the major population centers, mines, and ports. Yet, by the
end of the eighteenth century, Mexico’s colonial transport system had fallen
into disrepair, and there were not many stretches of road that would permit
the passage of wheeled vehicle traffic. Wagons continued to carry freight
wherever possible, but the bulk of the goods shipped in the nineteenth
century was hauled over Mexico’s steep terrain by mule.
6
In Colombia, the
backbone of the limited transport system in the nineteenth century was
the Magdalena River. Seasonal variations of water level and the presence of
rapids impeded through passage. The cost of movement was elevated by the
requirement of portage for a full twenty kilometers along one stretch of the
river.
7
The introduction of steamboats later in the century helped reduce
costs, but the river route did not otherwise enjoy any improvements in
navigability. Nor was there much relief provided by turnpikes. Colombia’s
roads at independence were the paths inherited from colonial era, and no
new roads were built before 1850. The steep mountains of the interior at
many points did not even permit mules to pass, so that human porters were
an especially important means of moving freight. Shipment by porter cost
six times that by mule.
8
The little road building that did occur after 1850
was limited to linking the lowlands with the coast and did not involve the
interior.
9
In Brazil, few rivers were navigable in any commercially significant sense,
and many were of limited relevance to all but the most local versions of
trade. The absence of cheap, long-haul waterborne transport in inland areas
with rich agronomic endowments and the virtual lack of an active colonial
road-building scheme conditioned early settlement patterns. Unsurpris-
ingly, at the start of the nineteenth century, much of Brazil’s population
lived within a few days travel of the coast. Interior settlement was lim-
ited to areas that were relatively accessible and to regions possessing readily
marketable, high value-to-weight goods, such as the gold and diamond
6
Coatsworth, Growth Against Development, 17–24.
7
William Paul McGreevey, An Economic History of Colombia, 1845–1930 (Cambridge, 1971), 42.
8
Ibid., 42–8.
9
Ibid., 162.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c08 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:32
300 William R. Summerhill
fields of Minas Gerais. Hilly and mountainous terrain, especially along the
coast, extended from the northeast to the south, presenting an obstacle
to overland communication with the interior that was just as imposing as
inland mountain ranges elsewhere in the hemisphere. Even with its natural
endowment of rivers, Brazil’s pre-rail transport system was likely more rudi-
mentary as that of Mexico. Argentina in particular, and the Rio de la Plata
region in general, did not suffer to the same degree the topographic condi-
tions that impeded overland communications in Mexico, Colombia, and
Brazil. Up-country terrain grew more hilly than what was found on the
littoral, and as one approached Andean ranges, the physical impediments
to movement became quite obvious. Yet the major inland water routes
did suffer barriers that could not be easily remedied by nineteenth-century
channeling and canal-building techniques. The Paran
´
a-Paraguay river sys-
tem that connected the Rio de la Plata to the Mato Grosso region north of
Asunci
´
on was far too shallow in some spots to allow for year-round move-
ment.
10
Both the Uruguay and Paran
´
a rivers had fall barriers that made
unimpeded through passage impossible. Although the region did not have
the mountainous terrain that complicated road building in other places,
overland movement on roadways still suffered the normal vagaries of poor
maintenance, washout, and minimal improvements in the half-century
following independence. No figures exist to indicate the extent of road
mileage or investment in roadways by public authorities in either the late
colonial or early national period. In general, Latin America as a whole suf-
fered from limited road capacity, and the capacity that did exist was crude,
often little more than ruts, and barely suitable. In a few cases, turnpike com-
panies were granted concessions, licenses, and even public subsidy by the
mid-nineteenth century, though these improved roads were exceptional.
11
Although case studies of road companies are almost completely absent
from the literature, it is a fair assessment that such roads, while improv-
ing the conditions of transit, did not offer much of a cost advantage for
shippers. From the colonial Spanish Caribbean to the mainland republics,
complaints by public officials and commercial interests alike commonly
attributed economic stagnation to the lack of internal improvements. Given
the physical location of productive possibilities, overland shipment became
10
Thomas Whigham, The Politics of River Trade: Tradition and Development in the Upper Plata,
1780–1870 (Albuquerque, NM, 1991), 4–7.
11
Domingos Giroletti, “A Companhia e a Rodovia Uni
˜
ao e Ind
´
ustria e o Desenvolvimento de Juiz de
Fora, 1850 a 1900” (typescript, Belo Horizonte, 1980).
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c08 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:32
The Development of Infrastructure 301
the predominant means of inland freight transportation in many regions of
Latin America that had the brightest economic prospects. Yet road build-
ing, debilitated by the factors sketched previously, did not provide much
in terms of new infrastructure at any time in the nineteenth century, did
not in the aggregate contribute much to the transition to modern rates of
economic growth, and never captured much tangible interest from policy-
makers until after the implantation of railways.
THE EXPANSION OF RAILWAY
INFRASTRUCTURE
Railways were the most palpable of all of the infrastructure investments
that characterized the economic modernization of Latin America in the
nineteenth century.
12
They were widely heralded throughout the region
even before their construction as the superlative solution to the problems
posed by the high cost of overland interior transportation in the region. In
many cases, that promise was realized, as proposed projects gradually came
to fruition and brought major reductions in transport costs. The limited
public and private resources of the newly independent Latin American
countries made it difficult to improve the quality or capacity of the transport
sector in the first half of the nineteenth century. Concessions in countries
such as Brazil and Mexico were first made in the 1830s, but only Cuba
executed its plans to build lines before mid-century. By and large, it was
not until the second half of the century that Latin American nations began
to construct rail networks in earnest. In Colombia, railways first traversed
the Isthmus of Panama in 1855, but the country did not place another line
in operation until 1871.
13
Until then, freight was carried by river wherever
possible, but most of it was shipped overland across steep terrain on the
backs of humans and mules.
14
Mexico had no significant railway until the
completion of the line between Veracruz and Mexico City in 1873.
15
Brazil
enjoyed its first railway in 1854, but the early major lines were not in place
until the 1860s.
16
12
Roberto Cort
´
es Conde, The First Stages of Modernization in Spanish America (New York, 1974).
13
McGreevey, Economic History of Colombia, 253.
14
Ibid., 43–5.
15
Arthur P. Schmidt Jr., Social and Economic Effect of the Railroad in Puebla and Veracruz, Mexico,
1867–1911 (New York, 1987), 1–11.
16
Manoel da Cunha Galv
˜
ao, Not
´
ıcia sobre as Estradas de Ferro do Brasil (Rio de Janeiro, 1869).