Professional Documents
Culture Documents
St Antony’s Series
Series Standing Order ISBN 978-0-333-71109-5
(outside North America only)
You can receive future titles in this series as they are published by placing a standing order.
Please contact your bookseller or, in case of difficulty, write to us at the address below with
your name and address, the title of the series and the ISBN quoted above.
Edited by
Enrique Cárdenas
Professor of Economics and Rector
Universidad de las Américas-Puebla
and
Rosemary Thorp
Reader in the Economics of Latin America
St Antony’s College
Oxford
in association with
St Antony’s College, Oxford
Editorial matter, selection and Chapter 1 © Enrique Cárdenas, José Antonio
Ocampo and Rosemary Thorp 2000
Chapter 3 © José Antonio Ocampo and Maria Mercedes Botero 2000
Chapters 2, 4–10 © Palgrave Publishers Ltd 2000
Softcover reprint of the hardcover 1st edition 2000 978-0-333-91304-8
1 Introduction
The Export Age: The Latin American Economies in the Late
Nineteenth and Early Twentieth Centuries 1
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp
Index 323
vii
List of Tables and Figures
Tables
viii
List of Tables and Figures ix
Figures
The three volumes of this mini-series reflect a collaborative effort that began in
1982, when a group of economists and economic historians met to consider
the effect of the 1929 Depression on Latin America, and the explanation of the
continent's rapid recovery. Carlos Diaz Alejandro was the leading light of
the group until his premature death in 1985. This initial group effort led to
the study Latin America in the 1930s, published in 1984. Subsequent collabora-
tion was made possible by the inspiration of Enrique V. Iglesias, President of the
Inter-American Development Bank, whose idea it was to fund a major project to
write an economic history of Latin America for the twentieth century, access-
ible to a wide audience but based on the best Latin American scholarship. His
invitation to me to lead the project and write the overview volume gave me the
most stimulating 16 months of my academic career. New friends joined the
group, and we came together in the course of 1997 in a series of workshops
organized to support the project. These workshops resulted in the books pub-
lished here as Volumes 1 and 3. Volume 2 is a reprint of the 1930s study. The
overview volume was published by the IDB in 1998 as Progress, Poverty and
Exclusion: An Economic History of Latin America in the Twentieth Century.
(available in English, Spanish, French and Portuguese from IDB bookstore,
www.iadb.org/pub, and in English through Johns Hopkins University Press,
www.press.jhu.edu).
My thanks for the opportunity to lead this stimulating project go first and
foremost to Enrique Iglesias himself, to the Inter-American Development Bank
and to the European Union for funding. Tim Farmiloe was our enthusiastic and
patient editor. I owe a particular vote of thanks to my co-editors, who made
themselves available against exceptional pressures and were remarkable part-
ners. Margaret Hancox, Catalina Sanint and Elvira Ryan bore the huge burden
of the final editing and processing.
ROSEMARY THORP
Latin American Centre
Oxford University and St Antony's College
xi
Notes on the Contributors
MarõÂa Mercedes Botero has a Master's degree from the Universidad Nacional
de Colombia and is a doctoral student at the University of London.
xii
Notes on the Contributors xiii
Antonio SantamarõÂa GarcõÂa has a PhD in history, and was a Senior Associ-
ate Member of St Antony's College, Oxford 1997±98. He was formerly Lecturer
at Universidad Carlos III in Madrid and Researcher at the Instituto Universitario
Ortega y Gasset (Universidad Complutense de Madrid).
This book explores the impact on Latin America of the extraordinary trans-
formation in the international economy ± in kind and degree ± that took place
in the half-century or so which preceded the World Depression of the 1930s.
The period covered in the studies varies, however, from country to country,
reflecting the different experiences in the process of integration to the world
economy. The book also studies the internal transformation that the various
economies experienced as a consequence of successive external shocks. Some of
these had rather similar effects throughout: the First World War and the post-
war boom and crash, and the Great Depression of the 1930s. Others did not: the
Baring crisis of the early 1890s, the collapse of coffee prices at the turn of the
century, other events in specific commodity markets, and even the recovery of
the 1920s, the characteristics of which varied considerably from country to
country. Domestic events which affected economic development throughout
this period are also taken into account. The most remarkable of them was,
obviously, the Mexican Revolution.
Although the export sectors were not necessarily the most relevant in terms
of employment or even share in global economic activity, they usually repre-
sented the most dynamic element of the economies and constituted the chan-
nel by which the countries in the region were related to the external world.
With very few exceptions, the export sectors in the Latin American context of
the nineteenth and early twentieth centuries were intimately related to the base
of natural resources, or to specific products that had world demand and were
thus cash commodities. Foreign interests were involved in the trading, trans-
portation and sometimes in the production of these commodities, as in invest-
ment in some domestic economic activities. The rise of a modern international
monetary system contributed to the expansion of trade and the financing of
1
2 Introduction
The rapid expansion of world trade was one of the outstanding features of the
acceleration of economic growth in the industrial `centre' in the nineteenth
century. This process continued in the early twentieth century, but experienced
a sharp slowdown from 1914, when the eruption of the First World War
generated a significant break in world political and economic history. At
1990 prices, world exports increased from US$7.3 billions in 1820 to
US$56.2bn. in 1870, US$236.3bn. in 1913 and US$334.4 billions in 1929
(Maddison, 1995, table I-4). Trade expanded at a faster rate than GDP in the
industrial centre and its offshoots in North America and Oceania up to 1913;
both experienced a slowdown in 1913±29, which was, nonetheless, sharper
for trade (Table 1.1). The large share of raw materials was a distinguishing
feature of world trade throughout the period of analysis. The dominance of
Western Europe, particularly of Great Britain, was also a major characteristic
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp 3
GDP
Western Europe 1.7 2.1 1.5
Western offshoots 4.3 3.9 3.0
Western Europe and offshoots 2.1 2.7 2.2
Latin America 1.5 3.3 3.3
World 1.0 2.1 1.9
World exports in real terms 4.2 3.4 2.2
Real exports of major Latin American countries 3.4 4.0
in real termsa
a
Argentina, Brazil, Chile, Colombia, Mexico, Peru and Venezuela.
Source: Maddison (1995), tables G-2, I-2 and I-4.
up to the First World War, but the balance gradually shifted towards the USA in
the interwar period (Ashworth, 1975; Kenwood and Lougheed, 1971; Maddi-
son, 1991 and 1995).
Behind the significant growth in world trade lay a revolution in technology
and transport, supported by major institutional developments. The break-
throughs in technology, in particular the steam engine and the use of coke
in the production of iron, that were at the heart of the Industrial Revolution
in Britain, had already taken place by the early nineteenth century. The
applications of steam and iron to transport, through railways and steam-
ships, revolutionized transportation, generating major impacts in the integra-
tion of internal markets and, in due course, of the international economy.
However, the spread of these innovations to industry and transportation was
gradual. Sailing ships, for example, were still dominant in world transportation
in the mid-nineteenth century. Many technological landmarks were added
throughout the nineteenth century, including in the last decades of the century
major breakthroughs in steel processing and chemistry and the discovery of
electricity. The internal combustion engine was also to revolutionize trans-
portation and industry in the early decades of the twentieth century. It was
spreading at a rapid rate worldwide in the 1920s, at the end of our period of
analysis.
However, it took the coming together of these technological developments
with institutional changes to generate the growth in production and trade
which was to have such an impact for Latin America as well as almost every
area of the world. The obvious institutional innovations that allowed produc-
tivity to multiply were in the organization of firms, and credit and capital
markets. Joint stock companies, stock exchanges and banks and the accompa-
nying regulatory legislation spread throughout much of Europe and the USA
and eventually worldwide. Landmarks such as the UK Bank Charter Act of 1844
4 Introduction
and the invention of the Registrar General of Companies were the unglamorous
but necessary background to rising prosperity.
Technological developments meant an extraordinary demand for minerals,
including some which up to then had little economic value. The demand for
nitrates, tin, copper and oil was to have a major impact on Latin American
exports. Increasing division of labour and rising incomes meant an increasing
demand for food, including `exotic' tropical products, as well as for agricultural
and forest raw materials. The growth in the demand for temperate products,
such as wool, wheat and meat, motivated the settlement of new lands, includ-
ing Argentina. The growing demand for cotton, henequen and rubber, among
agricultural raw materials, and sugar, coffee, bananas and cocoa, also pene-
trated the continent. Some of these mineral and agricultural raw materials,
which are viewed today as characterized by low-income elasticities, in fact
had the opposite feature during our period of analysis. For example, the
world production of coffee, a typical `exotic' tropical beverage, grew at an
annual rate of 2.7 per cent in 1830±1900, a rate much faster than that of GDP
and GDP per capita in the industrial centre (Ocampo, 1984, ch. 7).
The response of Latin America to the demand for minerals and both temper-
ate and tropical products required movements of capital and labour. Capital
moved primarily from Britain, Germany, France and the USA. It financed the
trading and transportation of commodities, and sometimes their production,
particularly in minerals and bananas. It also financed investments in railroads,
banking and modern public utilities. By 1913, one-fifth of UK capital overseas
was in Latin America, and half of that in Argentina (ECLA, 1965). But capital
also began to move increasingly with no necessary connection to prospective
export expansion. Its general availability responded to developments in the
centre countries. Fluctuations in capital flows were to become one of the most
threatening aspects of the new international economic order.
Furthermore, between 1820 and 1930, 62 million people emigrated from
Europe and Asia, mostly between 1860 and 1914. Of these, 61 per cent went
to the USA and 12 per cent to Canada, but Argentina and Brazil came next with
10 per cent and 7 per cent respectively (Hatton and Williamson, 1994). Chile
and Uruguay also received many newcomers, at least relative to their smaller
populations. Foreign migrant entrepreneurs and technicians, particularly from
Europe, were also important, even in countries which were not recipients of
massive immigration. Tropical and subtropical agriculture needed labour too,
but when it came from abroad it did so mostly under various types of indenture
to take forward plantation-style agriculture throughout the Caribbean, the
Caribbean coasts of Latin America and Peru. Part of these labour flows
were actually within the continent, as black labour from the Antilles moved
to the new banana plantations of Central America (see Chapter 4 in this
volume), the Cuban sugar industry and the construction of the Panama Canal.
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp 5
A methodological interlude
The analysis of Latin America in the export age brings out interesting meth-
odological issues. Some of them are distinctly economic, and relate basically to
the linkages between the export sector and domestic economic activity. Tradi-
tional `staple theory' and the literature on the `Dutch disease' are important
contributions to this analysis. Other issues are more interdisciplinary in char-
acter. Some trends in economic analysis are relevant in this regard, and the
recent emphasis on institutional development as an essential determinant of
historical performance is particularly useful. `Institutions', it should be remem-
bered, are used in this literature in a broad sense to include not only organiza-
tions, such as firms, producers' associations and state institutes, but also rules of
the game and intangibles like conventions and traditions ± that is, fiscal or
monetary conservatism. Also, the fresh notion of `path dependence' is stimu-
lating, but flesh must be put on this concept. Probably the biggest methodolo-
gical difficulty relates to the interaction between the state and economic
development. In this regard, although economists have abandoned the tradi-
tional approach to the state as exogenous and neutral, the insights on how to
endogenize it are partial and suggestive rather than complete, and in some
cases mechanical and simplistic.
Path dependence turns on two crucial ideas. The first is that the catalysing
role of the export economy operates on an economic and social structure that is
itself complex and changing. The second is that the trajectory which the
economy follows shapes, in turn, its own future. We found many instances in
our case studies of how the impact of external forces was conditioned by prior
history. The role of preexisting rural, social and economic structures is particu-
larly important. As we will see, they largely determined how labour was mobil-
ized to produce for world markets, and to what extent the rising demand for
agricultural goods for domestic consumption was supplied domestically.
A particular case in which path dependence and institutional development
came together was in the development of one of the most important `institu-
tions' of modern capitalist economies, that of a free labour market. As we will
see in our comparative analysis, economic development during the export age
in Latin America was crucially dependent on access to a mobile labour force.
Slavery and indentured labour from abroad had been used by the dynamic
export economies in the past, but these alternatives were no longer available
in the late nineteenth century.
Access to the most developed free labour market in the world at the time, that
of Europe, and thereby to the massive currents of European migration, was thus
crucial to guarantee a dynamic response to the opportunities which the inter-
national economy offered. However, it also determined the ± path-dependent ±
evolution of social institutions and the conflicts which characterized the
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp 7
economies and societies involved. The use of another pool of free labour, that
generated by the abolition of slavery in the Caribbean, was also important
though more limited, and its effects on the domestic economic and social
developments of the recipient countries totally different.
Countries which did not have access to international flows of labour had to
rely entirely on the domestic evolution of institutions to generate mobile
labour flows. Here, preexisting domestic labour conditions were crucial. As we
will see, how these conditions evolved over time and how labour was mobilized
to the dynamic export sector and its complementary activities largely deter-
mined both economic growth in the export age as well as the social and
political conflicts which characterized the different nations or colonies
involved.
Obviously, in traditional rural societies the development of a labour market
was intimately associated with that of land. How land ownership was managed
was, thus, also crucial in determining how economies evolved. In this regard,
the issue was not whether land was plentiful ± it was in the Latin American
economies ± but rather how institutions developed to eliminate traditional
`inflexibilities' in the land market, many of them associated with the role of
the Church in traditional agrarian societies, and others to that of Indian com-
munities in those countries and regions in which traditional Indian structures
survived up to the nineteenth century. Equally crucial was the way new land
was granted by the state to private landowners.
Institution building during the period of analysis involved also the develop-
ment of other domestic markets ± for financial services, for foreign exchange,
and so on ± and a modern state. The development of markets and modern
relations between the state and economic activity was crucial in determining
each economy's `capacity for change'. This concept has purely economic
dimensions, such as supply-side elasticity, capacity to absorb technological
transfers from the advanced countries and capacity to innovate. But change
also requires transactions among social actors, and in this regard the relations
that emerged between, for example, business elites and government from a
particular export experience may or may not have proven functional as struc-
tures adjusted to shifting external conditions or internal pressures. The rela-
tively developed capacity for change of Colombia, Mexico and Brazil in the
1930s is illuminated in the next volume, where we analyse the varying speeds
of recovery from the 1929 crash.
The pivot here is the emerging relationship between key export sectors, the
non-export economy and the state, which includes the ability to shift resources
to new exports or domestic economic activities, and the capacity of the state to
develop and implement new policy instruments that empower it to play an
active role in the adaptation of the economy to changing circumstances.
Change, it must be said in this regard, has to be implemented and this is a
8 Introduction
matter of negotiation, but also of the legitimacy and efficiency of public and
private-sector institutions, as well as of the capacity of the private sector to react
in the face of new conditions. The policy machinery and interest-group chan-
nels created during a specific stage of development, say to manage a drop in
commodity prices, may serve as instruments for change when the next shock
comes. But size and the structure of output is important too: shocks require
supply-side flexibility to produce either exportable or domestic alternatives,
and probably prior investment in these alternatives.
The reverse side of the coin might be usefully termed vulnerability. It may be
associated to legitimate lack of alternatives, as the story of the smaller econom-
ies may reveal. But it may be associated to perverse forms of path dependence:
previous success may build up inertia or dynamics which block the adoption of
alternatives when a shock comes. The success of sugar in Cuba in the first
quarter of the twentieth century had quite adverse effects on that nation's
development from the mid-1920s, given extreme dependence and proximity
to the USA: despite a successful restructuring of her sugar industry, her growth
was limited by US protectionism and by the lack of domestic diversification.
Also, the combination of the loss of the dynamism of external markets since
the First World War and the social conflicts which built up simultaneously
explains the vulnerability of the most successful of all Latin American export
economies, Argentina, as CorteÂs Conde argues in his contribution to this
volume (Chapter 9).
The following sections outline the comparative analysis which emerges from
the different country studies included in this volume. They consider the char-
acteristics of export expansion, the linkages of the export sector to the rest of
the economy, the role of the state and some distributive issues.
Preconditions
As the concept of path dependence makes explicit, export expansion in the late
nineteenth and early twentieth centuries built on various preconditions. The
first one is the independent political status of most of Latin America prior to the
export age. The remarkable exceptions are, obviously, Cuba and Puerto Rico,
which continued to be Spanish colonies until the end of the nineteenth cen-
tury, and Panama, which was part of Colombia up to 1903. Major border
disputes were involved throughout the nineteenth century, many of them
continuing up to the present. At least one of them had a major impact on
export development: the War of the Pacific between Chile, Peru and Bolivia
(1879±81). It meant the gain of nitrate deposits for Chile at the cost of Peru.
Whereas this meant the beginning of a spectacular export boom for Chile
(Chapter 6), for Peru it meant a collapse from which it would take more than
a decade to recover.
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp 9
Export dynamics
Major features of export expansion
Export expansions varied considerably throughout Latin America in their mag-
nitude, timing, stability and product composition. The `commodity lottery', to
borrow a concept used by Carlos DõÂaz-Alejandro, was the joint effect of all these
factors and the major determinant of domestic economic developments.
Indeed, most economic expansions in Latin America up to the 1920s were
`export led', in the sense that export growth was faster than and determined
the cyclical patterns of global economic growth. As we will see in the next
section, diversification of economic activity varied considerably from country
to country, depending on the nature of the linkages of the export sector,
domestic policies and size. However, in only one case, Brazil in the first three
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp 11
Table 1.2 Growth of Latin American exports, 1860±1929 (US$ millions at current prices
and exchange rates)
Value of exports
Latin America a 292 664 1493 2954
Argentina 13 163 437 964
excluding Argentina 279 501 1055 1989
Annual growth
Latin America 2.1% 7.0% 4.4%
Argentina 6.5% 8.6% 5.1%
excluding Argentina 1.5% 6.4% 4.0%
Value of world trade 5132b 9299 17345 32148
Annual growth 2.0% 5.3% 3.9%
Share of third world exports 41.8% 37.4% 38.4% 36.4%
Share of world trade 7.1% 8.6% 9.2%
a
Data refer to developing countries of America excluding the British Antilles.
b
1870.
Sources: Latin American and third-world exports, Bairoch and Etemad (1985), table 5.1; value of
world trade, Maddison (1995), table I-3.
decades of the twentieth century, did GDP growth exceed export growth by a
considerable margin.
Over the long term, export expansion in Latin America was more or less
similar to the growth of world trade and slower than in Asia and Africa in the
last decades of the nineteenth century, but faster than world trade and similar
to other `peripheral' regions in the first three decades of the twentieth century
(Table 1.2, and Bairoch and Etemad, 1985). Indeed, the region as a whole did
not experience the slowdown which characterized world trade between the
First World War and the onset of the Great Depression. In fact, real export
growth may have accelerated (Table 1.1). By the late 1920s, Latin America's
share in world trade had reached 9 per cent, an increase over the 7 per cent
typical in the last decades of the nineteenth century (Tables 1.1 and 1.2).
Growth was accompanied by major variations in the composition of regional
exports (Table 1.3). Prior to the era of export-led growth, exports were domin-
ated by sugar, coffee, precious metals, natural fertilizers (guano), tobacco and
hides. Except for coffee, which continued to represent close to a fifth of Latin
American exports, they experienced a long-term relative decline in the export
age. Export growth in the late nineteenth and early twentieth century was
characterized by the relative growth of exports of temperate zone products ±
cereals, wool and meat ± and minerals ± nitrates, tin, copper and oil. Most of
these products experienced a steady increase; the two exceptions were wool and
nitrates which reached their peak share in Latin American exports just before
the end of the nineteenth century and the First World War, respectively. Also, at
12 Introduction
a
Includes the British Antilles.
Source: Bairoch and Etemad (1985), table 5.3.
the beginning of the twentieth century there was a brief but intense rubber
boom, which was followed by rapid decline in the 1920s.
Export booms did not always lead to steady growth. A first factor which lay
behind this pattern was the failure to develop a stable export base. The most
spectacular case was that of the forest rubber boom which countries in the
Amazon basin experienced in the late nineteenth and early twentieth centuries,
which was followed by collapse when plantations in the east were developed on
a large scale in the 1920s. On a smaller scale, reliance on an unstable export
base was quite common in Colombia and Central America prior to the large-
scale development of coffee (Chapters 3 and 4). The exhaustion of the natural
resource on which the boom depended was also an important factor. On a large
scale, this was the case with oil in Mexico in the 1920s, but it was quite a
common phenomenon in mining countries, which depended on the discovery
and development of new resources to keep up the pace of expansion. Due to an
inadequate management of disease controls, banana production was also char-
acterized in the first decades of the twentieth century by continuous movement
of the production frontier (Chapter 4).
Substitution by new products was also significant. The most important was
the substitution of guano by mineral nitrates in the 1870s, and of the latter
by chemical fertilizers from the 1910s. Substitution by beet sugar, largely
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp 13
140
110
100
90
80
70
60
50
1870
1873
1876
1879
1882
1885
1888
1891
1894
1897
1900
1903
1906
1909
1912
1915
1918
1921
1924
1927
1930
1933
1936
1939
Figure 1.1 Net barter terms of trade between primary products and manufacture, 1871±
1913
14 Introduction
1920s (Chapter 5). Brazilian coffee export growth in the last two decades of the
twentieth century was actually followed by very slow aggregate export growth
in the first three decades of the twentieth century. Thus, whereas the export
quantum increased at an annual rate of 5.5 per cent in the last two decades of
the nineteenth century, export growth was rather slow ± 1.7 per cent a year ± in
the first three decades of the twentieth century (Chapter 2). Coffee expansion
was also rapid in the last two decades of the nineteenth century in Central
America, particularly El Salvador and Guatemala, and in Colombia (Chapters 3
and 4). This period also saw the rise of cocoa exports from Ecuador, which
continued to grow up to the First World War (Chiriboga, 1988).
The first three decades of the twentieth century added several late but rapid
export booms in medium-sized countries. They included Peru, based on a
diversified export base of sugar, cotton, copper and oil (Chapter 8); Cuba,
based on sugar (Chapter 10); Bolivia, based on tin (Chapter 7); Colombia,
based on coffee (Chapter 3); and the latest but most spectacular, the oil boom
which started in Venezuela in the 1920s and continued at very rapid rates in the
following decades. Among smaller countries, banana exports were responsible
for export growth during the first three decades of the twentieth century in
Costa Rica, Panama and Honduras (Chapter 4).
The extent of export expansion relative to the domestic economies can be
best measured by the value of exports per capita at the end of the export age. At
the outbreak of the First World War, a definite pattern of regional disparities
had been built (see Table 1.4). The three Southern Cone countries and Cuba had
then the highest exports per capita in the region, which exceeded by a con-
siderable margin those of other countries. Also, within Central America,
exports per capita in Costa Rica exceeded by a considerable margin those of
other countries (Chapter 4). This picture continued with only slight differences
in the late 1920s, basically associated with the booms of the medium-sized
economies. As we will see below, as a reflection of the dominant role of exports,
this pattern was reproduced in many other economic indicators. Some major
regional differences are, thus, a legacy of relative success during the export age.
Nonetheless, relative success from the 1930s onwards was to have a pattern
quite different from that of the export age (see the next two volumes in this
series).
were thus generally associated with labour issues ± which, as we have seen, were
closely intertwined with land distribution issues. On the other hand, mining
was capital and technology intensive. In this case, the major distributive issues
were associated with the dominant role of a few large national and interna-
tional firms and the control over rents from non-renewable resources.
Different products also have different processing and transportation needs,
and such needs determined the extent of domestic forward linkages. Mineral
ores must generally be processed close to the point of production to minimize
transportation costs. Thus, they generally required the development of smelt-
ing and sometimes refining facilities, which lay behind the early industrializa-
tion efforts of mining economies. This is not true of oil, which does not require
processing before it is exported but needs a special, capital-intensive, trans-
portation network ± pipelines and specific port facilities.
As a rule, agricultural goods also require processing and some of them special
transportation facilities. Sugar shares with mining ores the need to be processed
close to the production of its raw material, a fact which accounts for the
impressive development of modern sugar manufacturing in Cuba since the
nineteenth century. Meat exports are also a remarkable case in this regard,
particularly when refrigerated transportation abroad became widespread in
the early twentieth century. They required the development of packing houses
in the export economies, a major factor behind early industrialization in
Argentina. Modern cotton gins are also crucial for the development of stable
exports of this product, as the experience of Peru since the 1890s indicates. In
other cases, agricultural processing was simpler. Such is the case with coffee,
although processing facilities were, nonetheless, an essential element of early
industrialization efforts in coffee producing countries. By contrast, bananas
require no processing before they are exported but need access to a quick,
capital-intensive transportation network, a factor which lay behind the early
penetration of foreign capital in their production.
Processing requirements may turn the export sector into a major instrument
of technological transmission and diffusion. Where industrial processing was
required, the export sector became a channel for the diffusion of foreign
capital-intensive technologies and the training of skilled labour. Processing
was also a major mechanism for technology diffusion in coffee economies,
particularly when part of the processing was done directly by smallholders, as
was the case of Colombia.
Agricultural exports also directly involved mechanisms of technological inno-
vation. More intensive land use was a major characteristic of agricultural export
developments in Argentina in the late nineteenth and early twentieth century
(Chapter 9). Also, the introduction of new varieties of cotton was crucial for that
sector's development in Peru in the 1910s, just as new varieties of sugar were
crucial for the rapid growth of that sector in Cuba. The introduction of coffee in
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp 17
interesting intermediate case was that of migration of black workers from the
Caribbean to the banana plantations of Central America and the sugar planta-
tions in eastern Cuba, a response to the limited local supply of mobile labour
using up the emerging pockets of free labour markets in the region. Here, in
addition to traditional labour conflicts, frictions also built up with domestic
residents, with racial connotations.
As we saw, although slavery was abolished very late in those countries which
relied heavily on it for their export and domestic activities, no country or
colony could count on it for export expansion even before the middle of
the nineteenth century. Chinese-indentured labour was then imported to
work in Cuban sugar and in Peruvian sugar and cotton plantations, but this
mechanism of labour mobilization ceased to be available by the end of the
century.
Thus, in many countries, labour for export and complementary activities had
to be recruited domestically. This form of labour mobilization was many times
suboptimal, leading to a de facto model of economic growth limited by the
supply of `mobile' labour, a factor which was emphasized by many entrepren-
eurs at the time. In this regard, the stories are strongly regional in character, as
there were often substantial differences between regions within a specific
country. In some areas, a modern labour market of some sort had started to
build up and its development may have been accelerated by the liberal reforms
of the nineteenth century. Population pressures, generally, and pockets of
surplus population in peasant economies, in particular, also facilitated this
development. These `free' workers could be mobilized as temporary or perman-
ent wage labourers or, more frequently in export agriculture at the time, as
tenants. In this case, non-economic restrictions on labour mobility ± including
debt peonage ± were often involved. In the opposite case, labour mobilization
mixed cash incentives with outright coercion. This was generally the rule when
mobilization of the Indian population was involved. Frequently, this system
evolved into a regular flow of labour of the first type. The contrast between
these opposite forms of labour mobilization was present in Mexico, as the
development patterns of the North and the central highlands, on the one
hand, and the South, on the other, indicates (Chapter 5). It was also present,
in Central America, in the contrasting development of coffee plantations of El
Salvador and Guatemala, respectively (Chapter 4).
Either because exploitation tended to increase, as argued by Knight in his
chapter on Mexico (Chapter 5), or because the increasing demand for labour
actually increased the opportunities for workers and restricted the ability of
landlords to exploit them ± as argued, for instance, by Palacios (1983) for the
coffee haciendas in eastern Colombia ± major social conflicts erupted. They
sometimes interacted with those arising from the illegal usurpation of state or
communal lands. The most important of them was obviously the Mexican
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp 19
Revolution. The peasant uprisings of Peru in the 1910s and in Colombia in the
1920s and 1930s followed a similar pattern. Obviously, conflicts of this sort had
major political implications. Alliances could be built up with labour unions,
with Indian communities in their traditional revindication for land, or with
peasants involved in land conflicts in frontier regions, for example, and these
alliances could be used by emerging new political forces. The most remarkable
case was, again, Mexico, where the Revolution led to major change in social and
political relations.
There was, of course, another possible source of domestic labour supply:
peasant labour. Colombia in the early twentieth century demonstrated how a
significant pool of peasant labour could be used to produce for the interna-
tional market. Similar though more restricted phenomena were experienced in
the expansion of coffee in Costa Rica from the early Republican period
onwards, and in the Cuban tobacco sector. It also played a role in the supply
of food for the cities or for the export sector, as for instance in supplementary
activities to copper mining in the Central Peruvian Sierra. For the region as a
whole, however, this form of labour mobilization was rather limited in scope.
The control of rents was also a major source of disputes and outright con-
flicts. In a sense, all export expansion in Latin America at the time was based on
the exploitation of previously underutilized resources, an idea that fits neatly
with the old `vent for surplus' trade models inspired by Adam Smith (Myint,
1958). There were indeed significant rents to be captured. In agriculture, the
capturing of rents by producing firms was seen as `natural'. There was, however,
an indirect way to capture them for domestic agricultural and industrial devel-
opment: protection, which increased the terms of trade at which export
incomes could be converted into other goods.
On the contrary, in mineral economies the issue of control over rents was
always in the forefront of domestic political discussions, particularly as they
necessarily involved the negotiation of rights to the exploitation of specific
mineral deposits, generally with fairly large firms. The willingness to attract
new companies by giving up royalties, taxation of mineral exports or direct
taxation of profits was one strategy; the alternative route generally involved the
taxation of exports and/or profits. The strategy followed determined the
returned value from those exports and the linkages to the rest of the economy
(see below).
Conflicts in export economies were magnified in the case of forest exploita-
tion, when they easily became very violent, as they generally involved at the
same time the extraction of rents, the forceful mobilization of labour ± includ-
ing the return to outright slavery ± and the lack of rule of law typical of frontier
regions. The history of rubber exploitation in all countries of the Amazon basin
is the clearest manifestation of how violent export economies could become
when all these conditions were mixed.
20 Introduction
Domestic linkages
The links between the export sector and other activities in the Latin American
economies involved three major issues. The first was the way the surplus of the
export sector was taxed and how the resulting tax proceeds were used domes-
tically. The second was the macroeconomic instability associated to the pro-
pensity of export economies to face strong boom and bust cycles, which could
be compounded by the instability of foreign lending. The third were the
forward and backward linkages generated by export activities and how they
built up ± or not ± the basis for a more diversified economy. As we will see, this
involved other macroeconomic issues, particularly those emphasized by tradi-
tional `staple' theory on the one hand, and the `Dutch disease' literature on the
other.
A major issue in relation to the export economy model was the concept of the
`returned value' of exports, namely the share of gross output that remained
within national boundaries and therefore had a linkage effect with the domes-
tic economy. This percentage depended on the nature of the product and on
the capacity of the state to effectively tax the producer and extract part of the
rent. The latter point becomes crucial for economic reasons, since such taxation
is usually a major means by which a nation captures the rents from its mineral
resources, and because it exposes the interplay between the state and large
investors, often foreigners. The political economy of this interrelationship
usually shows the means and ways by which external economic and political
interference occurred. Taxing of tin mining in Bolivia, nitrate and copper in
Chile, and silver and oil in Mexico are notable examples, as are the absence of
similar efforts in Peru in the post-guano era and the lack of any taxation of
banana exports in Central America and Colombia. As we pointed out, in the
cases of Chilean nitrates and Peruvian guano, export taxation was facilitated by
a high world market share.
As we pointed out above, domestic protection was another major way of
indirectly taxing export activities, as it thereby reduced the real income of
exporters. It was actively used by several countries and, as we will see, partly
determined the linkages of the export sector to the rest of the economy. In the
case of Brazil, as Chapter 2 in this volume shows, protection was partly trans-
ferred to world coffee prices and thus paid by coffee consumers worldwide. It
also redistributed income from those consumers to coffee producers outside
Brazil, notably Colombia and Central America.
Depending on the commodity lottery, including the degree of diversity,
countries had varying tendencies to boom and bust with resulting variances
in the degree of instability. The variability of Chile's terms of trade was so high
that it was equivalent on average to a shock on GDP of 1.5 per cent a year
during 1865±95, and 3.5 per cent a year during 1886±1929. Three major
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp 21
Central America. The domestic sugar and flour industries were protected by
high tariffs in Colombia in the early twentieth century. Cuba is probably the
most striking case of an economy which was at the same time a large exporter
and importer of food. Competition for land was in this case an important
matter, a fact which was not true in other agricultural economies.
Mineral economies provide even more interesting contrasts. The positive
response in Chile to the consumer demands from nitrate workers is a remark-
able example; it also illustrates the importance of state intervention to guaran-
tee adequate domestic transportation and investments in human capital
(Chapter 6). The Peruvian copper story under the Cerro de Pasco Corporation
is quite different, as it began generating significant linkages to the Central
Sierra peasant communities. The contamination of agricultural land by fumes
from the copper smelter led to the purchase of land, the creation of a cattle-
raising division and a delinking from the regional economy, which led it into
stagnation in the 1920s (Chapter 8). Indeed, in several countries, many large
mining companies became importers of foodstuffs for their workers, a beha-
viour which was also typical of banana plantations.
The story of manufacturing during the export age is an equally fascinating
subject. Rising exports were matched by rising imports, mainly manufactures.
However, as several research efforts in recent decades and the contributions to
this volume indicate, manufacturing development clearly predated the 1930s
and the postwar era of state-led industrialization. Manufacturing growth was
induced in several countries through four different channels. The first was the
forward linkage of export activities: as we have seen, some exports required
significant processing and thus generated numerous manufacturing establish-
ments in many countries. The second was the combination of backward demand
linkages and high transport costs, which generated a `natural import substitu-
tion' of some manufactured foodstuffs, beer, printing and, later, cement, among
other manufactures. Indeed, the term import substitution may be inappropriate
in this case, as manufacturing production directly accompanied in several cases
demand growth without significant imports ever developing.
The two additional channels were associated with tariff policy. As we have
seen, all countries in the continent relied heavily on import tariff revenues
during the export age, and this necessarily provided protection. The tariff
schedules of the time generated some peculiar distortions, which are behind
the scenes in the stories of early manufacturing development. They were either
specific tariffs based on the weight of merchandise, or ad-valorem tariffs paid
on the basis of an official price list. Specific weight tariffs gave high protection
to the production of industrial goods with low value per weight: for instance,
the system gave higher protection to simple rather than sophisticated textiles.
Also, under both systems, the real protectionist effect depended on interna-
tional and domestic processes: inflation eroded protection, but deflation ±
24 Introduction
which was quite common during world and domestic crises up to the 1930s ±
increased protection. This generated a peculiar cyclical pattern of protection,
which was compounded with that of exchange rates in those countries which
did not follow gold-standard rules. During external booms, manufacturing was
discouraged by falling ad-valorem tariffs and real appreciation ± but was pro-
moted by rising demand ± but tariff and exchange-rate protection were auto-
matically increased during crises. The coincidence of import substitution
efforts in several Latin American countries during the First World War was a
precedent to a similar story during the 1930s, which is explored in the next
volume. During the war, the physical scarcity of some manufacturing goods
imported from Europe generated an additional encouragement to produce
them domestically.
Beyond this, protectionism was quite common in Latin America at the time.
This obviously matched worldwide trends, as free trade was in fact an exotic
policy practice even in the industrial centre, as we have seen. Thus, contrary to
modern intellectual trends, export promotion and protectionism were not
viewed in several countries as opposite, but rather as complementary strategies,
elements of one single modernization drive. Brazil, Mexico, Chile and
Colombia are the outstanding examples of Latin American countries which
practised protectionism long before `inward-looking' development, and where
early industrialization was closely associated to protection. Argentina was a
different case, in which early industrialization was associated much more
with the other mechanism we have outlined. Interestingly, the rising labour
movement in this country was distrustful of both protectionism and exchange-
rate activism, both of which were seen as ways to depress real wages.
The rise of the textile industry in this country prior to the Great Depression
provides a possible fifth link: its complementarity with dynamic cotton cultiva-
tion. However, this was not common at the time in the region. Indeed, the
processing of imported low-tariff raw materials ± that is, high effective protec-
tion ± was very important for early industrialization in other countries.
The success of manufacturing development depended, of course, on other
factors, two of which were crucial. The first was size, and the second was
domestic market integration. Indeed, early industrialization critically depended
on this market integration, which was the accumulated effect of export growth,
urbanization and the development of modern infrastructure. Obviously, a large
country such as Brazil only depended on this process taking place around its
most dynamic growth pole, the SaÄo Paulo region, delaying true national integ-
ration. By contrast, in medium-sized economies, such as Colombia, successful
industrialization was delayed by lack of domestic market integration, despite
rapidly rising domestic incomes and protectionism. Late domestic market
integration would, nonetheless, become a facilitator of very rapid industrial
growth in Colombia since the 1930s.
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp 25
Table 1.4 Latin America: economic and social indicators, circa 1913
Notes: Exports per capita: current USA dollars per capita; railways per capita: kms of track per
thousand persons; GDP per capita: GDP per capita based on current prices, in USA dollars; literacy:
proportion of literates in population 15 years and over; life-expectancy: years at birth.
Sources: Exports from Thorp (1998); railways from Mitchell (1993); literacy from Hunt (1998); GDP
figures from Bulmer-Thomas (1994, table A.2.1, p. 439); agricultural labour force from Bulmer-Thomas
(1994, table 5.1, p. 122).
boom, as we have seen, which then improved their relative standing. Moreover,
the basis for the dynamism of Argentina was being eroded. Its rate of growth fell
to 2.3 per cent in 1913±29, slightly below population growth. At the same time,
Brazilian economic growth had accelerated. After a rather dismal record in the
nineteenth century, this country's GDP grew at 4.3 per cent a year between
1900±04 and 1925±29 (Haddad, 1980). Moreover, this was the first case in the
region in which GDP growth was not accompanied by rapid export growth ± 1.7
per cent a year in the first three decades of the century, as we have seen. Inward-
looking development had made its first appearance, facilitated nonetheless by
the rapid growth of coffee in the SaÄo Paulo region in the late nineteenth
century and by the transfer of resources from foreign coffee consumers to
protected industrial producers in Brazil. These were the first signs of the new
era to come.
The state played an essential role in export development and the ways the
export sector was linked to the rest of the economy. Although it was certainly
not the type of interventionist state which built up from the 1930s in Latin
America ± as in the rest of the world ± it also did not fit the laissez-faire
image constructed by some nostalgic analysts viewing the export era in Latin
America.
In passing, it should be said that the traditional `law and order' functions
were performed rather poorly in many cases, particularly with respect to `order'.
As we pointed out above, the building of a national, republican order was
painful in most nations. One of the major consequences of this fact was that
the victory of economic liberalism was not necessarily accompanied in the
region by the development of liberal `political' institutions. Indeed, economic
liberalism happily, even functionally, coexisted with many authoritarian
regimes.
State intervention played an essential role in five different areas. The first
was, as we have seen, land and labour policies. As labour or peasant movements
developed, this also meant either repression, the promotion of social legislation
to accommodate the claims of labour or the peasantry, or a mixture of both.
The second was taxation. Although taxation of exports was adopted in some
mineral economies, the most common way to tax at the time was to impose
import tariffs. Whether intentionally or not, this meant protection for domes-
tic agriculture and manufacturing. Indeed, as we have seen, protectionism was
widely practised in the region long before import substitution became an
official policy. Towards the end of our period of analysis new forms of taxation
were introduced on an important scale in the most successful export
28 Introduction
standard viable, and also required domestic relative price adjustments which
had adverse effects on import-competing and export activities.
Finally, aside from taxation issues, governments were also involved in the
direct regulation of exports. The early and most notable example were, of
course, Brazilian coffee regulations. They included prohibitions of new plant-
ings in the first decades of the century, three valorization schemes ± 1906, 1917
and 1921 ± which aimed to accumulate stocks within Brazil to defend prices,
and the `permanent defence' of coffee introduced by the SaÄo Paulo government
in 1925 which led to other regulations (Rowe, 1932). From the mid-1920s, the
Cuban government was also active in the regulation of domestic sugar produc-
tion to manage its own excess supply, and in the promotion of partial interna-
tional agreements (Chapter 10). Indeed, domestic and international regulation
of commodity markets became an attractive alternative since the First World
War as a way of managing overproduction; it became widespread in the 1930s
(Rowe, 1965, part IV).
The specific response of the state depended obviously on the character of the
export elites and to what extent they were articulated or disarticulated in social,
economic and political terms, among themselves, and with other groups. Some
hypotheses can be advanced in this regard. Strong domestic ownership of a
dominant export product but diverse regional bases led to political trading
mechanisms between regional elites, as in Brazil and Colombia. A system
where concessions for the exploitation of natural resources were required
usually meant a close relationship between the state and the elites, their inter-
play and often their overlapping. For instance, members of the government or
the legislature in Mexico and Chile participated in mining companies as share-
holders, and had interests in other activities, including the commercial banks.
Second, the needs of the sector were influential in the demands the elites
imposed on the state. They may or may not have required some degree of state
intervention that went beyond the guardian state. Moreover, this may or may
not have provoked some degree of institutional development in the state, the
issue of laws, norms and other regulations, the prime positive examples being
Colombia and Mexico. Again, export diversity and a strong presence of foreign
capital substituting for the state made for a weak link in this regard (Peru).
The third area was the elite's ability to respond to the social and economic
challenges of the period, with varying degrees of consensus and violence. This
included the ability to respond to the social conflicts which arose during the
export age, including the development of organized labour or the conflicts over
labour and land relations. In Brazil and Chile, elites showed a response capacity,
and to a lesser degree in Colombia, with considerable degrees of violence in the
latter case. Elite capacity to respond was not a characteristic of Peru or Mexico.
This introduction has attempted to draw out the rich themes stimulated by
the case studies and by the workshop where they were discussed. Each theme is
30 Introduction
significant not only for the understanding of the period which it yields, but for
the way export dynamics, linkages and the consequences for the state, the
evolving political economy of policy-making and public±private relations,
were to condition and shape the decades that followed. This theme is devel-
oped in the overview volume that is also a part of this project (Thorp, 1998).
The rest of the present volume contains the country-specific studies in which
the overview is grounded.
References
Ashworth, W. (1975) A Short History of the International Economy since 1850, London:
Longman.
Bairoch, P. and Etemad, B. (1985) Structure par produits des exportations du Tiers-Monde,
Geneva: Universite de GeneÁve: Centre d'Histoire Economique Internationale.
Bulmer-Thomas, V. (1994) The Economic History of Latin America since Independence, Cam-
bridge: Cambridge University Press.
Chiriboga, M. (1988) `Auge y crisis de una economõÂa agroexportadora: El perõÂodo cacao-
tero', in E. Ayala Mora (ed.), Nueva Historia del Ecuador, Vol. 9, Grijalbo.
Corden, W.M. and Neary, J.P. (1982) `Booming Sector and De-Industrialization in a Small
Open Economy', Economic Journal, December.
ECLA (1965) External Financing in Latin America, New York: United Nations.
Eichengreen, B. (1995) Golden Fetters: The Gold Standard and the Great Depression 1919±
1939, New York: Oxford University Press.
Grilli, E.R. and Maw Cheng Yang (1988) `Primary Commodity Prices, Manufactured
Goods Prices and the Terms of Trade of Developing Countries: What the Long Run
Shows', The World Bank Economic Review, January.
Haddad, C.L.S. (1980) `Crecimiento Economico do Brasil, 1900±76', in P. Neuhaus (ed.),
Economia Brasileira: Uma Visao HistoÂrica, Rio de Janeiro: Editora Campus.
Hatton, T. and Williamson, J.G. (1994) `International Migration 1850±1939: An Economic
Survey', in T. Hatton and J.G. Williamson (eds), Migration and the International Labour
Market 1850±1939, London/New York: Routledge.
Hirschman, A.O. (1981) `A Generalised Linkage Approach to Development, with Special
Reference to Staples', in Essays in Trespassing: Economics to Politics and Beyond, New York:
Cambridge University Press.
Hunt, S.J. (1997) `The Human Conditions in Latin America, 1900±95', mimeo, Consult-
ancy paper for this project.
ÐÐ (1985) `Growth and Guano in Nineteenth Century Peru', in R. CorteÂs Conde and S.J.
Hunt (eds), The Latin American Economies: Growth and the Export Sector 1880±1930, New
York: Holmes & Meier.
Kenwood, A.G. and Lougheed, A.L. (1971) The Growth of the International Economy 1820±
1960, London: George Allen & Unwin.
Lewis, W.A. (1978) Growth and Fluctuations 1870±1913, London: George Allen & Unwin.
Maddison, A. (1995) Monitoring the World Economy 1820±1992, Paris: OECD Development
Centre.
ÐÐ (1991) Dynamic Forces in Capitalist Development: A Long-Run Comparative View, New
York: Oxford University Press.
Enrique CaÂrdenas, Jose Antonio Ocampo and Rosemary Thorp 31
Marichal, C. (1989) A Century of Debt Crisis in Latin America: From Independence to the Great
Depression, 1820±1930, Princeton: Princeton University Press.
Mitchell, B.R. (1993) International Historical Statistics: The Americas, 1750±1988, Basing-
stoke: Macmillan.
Myint, H. (1958) `The Classical Theory of International Trade and the Underdeveloped
Countries', Economic Journal, June.
Neary, J. P. and van Wijnbergen, S. (1984) `Natural Resources and the Macroeconomy: A
Theoretical Framework', in J. P. Neary and S. van Wijnbergen (eds), Natural Resources
and Macroeconomy, Oxford: Basil Blackwell.
Ocampo, J. A. (1984) Colombia y la EconomõÂa Mundial 1830±1910, BogotaÂ: Siglo XXI.
Palacios, M. (1983) El cafe en Colombia 1850±1970: Una historia econoÂmica, social y polõÂtica,
Mexico/BogotaÂ: El Colegio de MeÂxico/El Ancora Editores.
Rowe, J.W.F. (1965) Primary Commodities in International Trade, Cambridge: Cambridge
University Press.
ÐÐ (1932) Studies in the Artificial Control of Raw Material Supplies, No. 3, Brazilian Coffee,
London and Cambridge Economic Survey, January.
Thorp, R. (1998) Progress, Poverty and Exclusion: An Economic History of Latin America in the
20th Century, Washington: Inter-American Development Bank and Johns Hopkins Press.
2
Brazil as an Export Economy,
1880±1930*
Marcelo de P. Abreu and Afonso S. Bevilaqua
Introduction
* Revised version of a paper presented in the workshop on the Latin American Export Economies for
the project on the Economic History of Latin America in the 20th Century, held at Panajachel,
Guatemala, 9±11 December 1996. The authors wish to thank Elisa Reis, Andre Villela, RogeÂrio
Werneck and Fernando Blanco as well as participants in the Panajachel meeting for their help and/
or comments. The competent research assistance of Fernando H. Alvares, Alvaro B. A. Motta and Luiz
Gustavo Cherman is acknowledged. The FINEP (Financiadora de Estudos e Projetos) and the Conselho
Nacional de Desenvolvimento CientõÂfico e TecnoloÂgico are thanked for financial support.
32
Marcelo de P. Abreu and Afonso S. Bevilaqua 33
a
GDP at factor prices, including financial intermediation;
services include government.
Sources: 1910±47: Haddad (1978), sectoral data aggregated
using 1947 weights; 1947±80: Brazil (1990).
a
Data not strictly comparable, as up to 1920 rural population in cities of more than 50 000
inhabitants was also included.
Source: Lopes, 1968, pp. 13±16.
domestic policy and world coffee prices. We than consider the linkages of coffee
agriculture with other sectors, especially infrastructure and industry, followed
by the export economy between 1930 and 1960, included as a coda. But this is
an important coda, due to the role of coffee policy in Brazil's recovery from the
Great Depression and the continued importance of commodity exports in
foreign exchange earnings, even if the economy became increasingly closed.
In spite of the erosion of its dominant past position, in 1960 Brazilian coffee
exports still corresponded to almost 37 per cent of world exports. The final
section concludes the chapter.
Coffee became the leading Brazilian export commodity very early in the nine-
teenth-century; by the early 1830s it already accounted for nearly 30 per cent of
total exports. From the 1850s it reached almost 50 per cent of total exports, on
average, and remained in the 50±65 per cent range until the early 1960s. Its
share fell slightly in the 1860s due to the US Civil War, and more significantly
in the 1940s as the Second World War closed most Brazilian coffee markets
outside the USA (see Table 2.3).
The Brazilian share of world coffee production, which was already in excess of
50 per cent in the 1850s, fell slightly in the 1870s, but then, as competitors in
Asia were hit by plant disease and price support schemes were introduced, it
increased to reach more than 60 per cent in every year from 1896 to 1942 - more
than 70 per cent in almost half of these years and more than 80 per cent in
1906. It fell from the early 1940s, but it was still around 47 per cent in the
1950s. The long-term evolution of coffee prices is shown in Figure 2.1. Figure
2.2 shows the terms of trade between coffee prices and import prices. As the
Marcelo de P. Abreu and Afonso S. Bevilaqua 35
a
10-year average of yearly shares. 1850±51 to 1887±88: fiscal years.
Source: Computed from AnuaÂrio EstatõÂstico do Brasil, various years, and Brasil em NuÂmeros, various
years.
80.00
60.00
40.00
20.00
0.00
1850
1855
1860
1865
1870
1875
1880
1885
1890
1895
1900
1905
1910
1915
1920
1925
1930
1935
1940
1945
1950
1955
1960
Year
Brazilian series of import prices are not very reliable, export prices for the USA
and the UK were used as proxies. Export quantities increased modestly until
1880, at 0.9 per cent yearly, then very rapidly until 1900 at 5.5 per cent yearly,
slowing down to 1.7 per cent in 1900±30. By the end of the 1920s capacity to
import (due to coffee exports) had increased more than fourfold in relation to
1851, at roughly 1.8 per cent yearly (see Bacha and Greenhill, 1992, appendix).
The share of Brazil in the world market was lower than the share in produc-
tion as Brazilian coffees tended to command lower prices, and 78 million bags
corresponding to three years of world consumption were destroyed under
36 Brazil as an Export Economy, 1880±1930
300
UK
250
US
200
150
100
50
0
1850
1858
1866
1874
1882
1890
1898
1906
1914
1922
1930
1938
1946
1954
Year
coffee support programmes, mainly in the 1930s. The share of Brazilian coffee
exports in total world exports was nearly 40 per cent in the 1850s, rose to more
than 50 per cent in the 1860s, receded to 40 per cent in the 1870s, and then
increased to reach almost 75 per cent at the turn of the century and more than
80 per cent in 1911. It then fell, but remained above 50 per cent until 1929.
With the Great Depression it fell further to reach a value lower than 30 per cent
in 1938±39, and then rose again to almost 50 per cent until the early 1950s. In
1960 it was still above one-third of the market.
Brazilian coffee exports increased with the dissemination of world coffee
consumption. Growth was very rapid in the early nineteenth century: fourfold
by volume in the 1820s, twofold in the 1830s, and 1.7 times in the 1840s. Then
there was stagnation in the 1850s, further increase in the 1860s ± by 70 per cent
± and stagnation in the 1870s. Growth was resumed in the 1880s, by 30 per
cent, and spectacularly in the 1890s as the export quantum increased by 170
per cent. From 1821±25 to 1901±05, the volume of coffee exports increased
at 5.2 per cent a year. From 1851±55, when production was already well-
established, to 1901±05 it increased at a yearly rate of 3.2 per cent. In the late
1860s the USA market was already absorbing some 50 per cent of the Brazilian
coffee output. This share was the same in the early 1900s and in the late 1920s.
The British share of Brazilian total exports fell sharply due to the insignificant
British coffee imports. Before the First World War Britain had been overtaken by
Germany as a market for Brazilian exports and this share had been reduced to
13 per cent (see Delfim Netto, 1979, p. 9 and Brazil, 1941a, pp. 299±304).1
Marcelo de P. Abreu and Afonso S. Bevilaqua 37
30.00
25.00
20.00
15.00
10.00
5.00
0.00
1887
1889
1891
1893
1895
1897
1899
1901
1903
1905
1907
1909
1911
1913
1915
1917
Year
The rubber economy was based on the extraction of native rubber with ex-
tremely limited pre-export processing. The boom attracted surplus manpower
from other parts of the country, especially from the declining north-east (see
Furtado, 1965, ch. 23). Its abrupt end resulted in a sharp contraction of eco-
nomic activity in the region as it returned to subsistence activities.
Coffee in Brazil was produced in plantations. Before the 1870s the region
around Vassouras, relatively near Rio de Janeiro in the ParaõÂba Valley, had been
the most important producing area, but the primitive planting methods and
the characteristics of the dominant topography led to the continuous destruc-
tion of productive soils. As free land was available at low cost, the coffee frontier
moved southwards in the direction of SaÄo Paulo leaving behind a trail of
destruction.7 In the Paulista plateau there were very good soils for coffee
together with a good topography. For many decades, from the 1870s until the
Great Depression (1928±33), SaÄo Paulo was the most dynamic coffee region in
the country. The Brazilian native population in expanding coffee culture areas
had already been unfavourably affected by the economic expansion before the
coffee boom, so the link between the expansion of the coffee culture and the
declining standard of living of the Indian population does not seem direct. But
the expansion of the coffee economy directly resulted in the destruction of
most of the native Atlantic forest, as burn and slash methods of clearance
remained the rule for the whole period.8
While land policies nominally protected the interests of all those seeking to
occupy free government land, actual policy prevented the establishment of
small-scale subsistence agriculture by immigrants for a long period and thus
the creation of an alternative to low-wage agriculture. The link between land
policy and the elasticity of supply of labour to the plantation-based coffee
Marcelo de P. Abreu and Afonso S. Bevilaqua 39
a
USA and French.
b
These estimates are particularly fragile given the weight of French investment in 1913 and its sharp
fall in the 1920s.
Sources: See sources in Abreu (1985, 1994 and 1995).
nineteenth-century Brazil, especially in the 1890s and after the 1940s when
rates of growth were two or three times those for the total population (see Table
2.2). It is as if most of the important infrastructure investments had been
completed before the turn of the century and urbanization had become the
focus of attraction of new investment (see Lewis, 1978, ch. 7).
The transition to a Republican regime in 1889 was intertwined with the
institutional strengthening of the military, the abolition of slavery, and the
erosion of the political coalition which supported monarchy and was heavily
tilted in favour of the north-east. The coup of 1889 resulted from an alliance
between the military and the Paulista coffee growers. The transition to the
stable political preeminence of the south-east was achieved towards the end
of the 1890s by a pact which assured the automatic election of the presidential
candidate of the party in power and the continuity of local political control by
Marcelo de P. Abreu and Afonso S. Bevilaqua 41
recorded in the Oxford English Dictionary, in 1907 ± was from 1907 to 1930.
The first Brazilian valorization of 1907±13 was decided in the wake of the
considerable exchange rate appreciation which followed the Murtinho Plan of
1898, when the domestic currency rose from 7 pence to 12 pence, coupled with
a large expansion of coffee output in the 1890s and again after 1905. Provincial
authorities entered into a pact to restrain supply by purchasing stocks and to
freeze production. They needed federal backing to raise finance and added a
clause to their pact which demanded exchange rate stabilization. Foreign debt-
service was to be paid by an export tax which was thought would be mainly
borne by consumers, given the price inelasticity of coffee demand. After many
political difficulties (see Fritsch, 1988, ch. 2), coffee valorization came into
being, mainly controlled by foreign banks and coffee importing houses. It was
a success until it was disorganized by compulsory sales of coffee stocks in the
USA in 1912 due to legal action under the Sherman Act. In parallel with
valorization, Brazil adopted the gold standard, an experiment which survived
until the outbreak of the war in July 1914.
The valorization experiment relying on foreign finance ± half-hearted domes-
tically financed attempts were undertaken during the First World War ± was
repeated in the 1920s. But after the difficulties with USA legislation, stocks were
held near producing regions and responsibility for coffee `defence' was trans-
ferred to the state of SaÄo Paulo so that the federal government could retain a
free hand in relation to initiatives to raise finance in New York without fearing
the USA government veto, as happened early in the 1920s. In the wake of the
international trend to return to gold, Brazil adopted the gold exchange stand-
ard in 1926. So there was a close parallel between economic policy develop-
ments before the First World War and before the Depression, with an active role
played by the government. The lack of a laissez-faire tradition, which was
already well-established when the Republican regime was introduced in 1889,
was only reinforced by developments under the Old Republic.
The vulnerabilities of coffee valorization have been examined in the literature.
To the extent that it relied on foreign borrowing it was vulnerable to the cyclical
behaviour of international capital markets. The success of such schemes
depended on the ability to establish barriers to entry, that is, on successfully
limiting the planting of new coffee trees as this would make oversupply even
more serious and turn valorization into a snowball. The costs of price valoriza-
tion were paid only by Brazil as the dominant supplier; competitors were free-
riders as they simply reaped the benefits of higher prices at no cost. In the long
run Brazil's umbrella effect would naturally entail the erosion of its market share.
It is not easy to assess the impact of the coffee export economy on the
standard of living of the Brazilian population as there is very limited reliable
and comprehensive information on social conditions in Brazil before the 1940s.
But what exists underlines the widespread poverty. Average life expectancy at
Marcelo de P. Abreu and Afonso S. Bevilaqua 43
birth in the 1930s was still 42.7 years (the extremes were 33.5 years in the state
of Rio Grande do Norte and 52 in Rio Grande do Sul), and infant mortality
stood at 158.3 per thousand live births (extremes of 201.1 and 119.4 in the
same states). The coffee states and, specifically, SaÄo Paulo were very near the
average (43.6 years and 154.7 per 1000 live births in SaÄo Paulo). However,
improvement by the end of the 1950s was considerable and concentrated in
the south and south-east. The illiteracy of the population over five years of age
was 61.2 per cent in 1940 (65.3 per cent for women), the first year for which
there is reliable data (see Brazil, 1990, p. 52,).16
There is much controversy on the direct effect of the expansion of coffee
production on the standard of living of agricultural labourers and small land-
owners, but there is no hard evidence favouring any of the contrasting inter-
pretations (see Stolcke, 1986, pp. 78ff ). The main mechanism working in favour
of an improvement of the standard of living was the reduction in the share of
employment in agriculture and a corresponding increase in industrial employ-
ment as the productivity in industry was about 3.5±4 times the productivity in
agriculture both in 1920 and 1940. The increase in industrial employment of
almost 1.3 million between these two years corresponded to more than 8.5 per
cent of the working population in 1940.
The links between world coffee prices and the Brazilian exchange rate, or for
that matter world commodity prices and the exchange rate in any commodity
exporter which holds a substantial share of the relevant market, were ac-
knowledged early in the literature on Brazilian economic history.17 Brazilian
supply had an important impact on world coffee prices; due to the low price
elasticity of coffee demand, supply shifts had significant consequences on the
price level, and increased output meant a heavy fall in prices. Retention of
stocks could hold or raise world prices without much impact on quantities
demanded.
The fact that Brazil was the leading coffee producer and exporter implies that
production costs in Brazil had a significant influence in the determination of
world coffee prices.18 Foreign exchange rate fluctuations affected the supply
of coffee in Brazil and consequently world coffee supply due to the weight of
production in Brazil. In the short run, devaluation tended to weaken world
prices as there was an inducement to dump stocks in the market. In the long
run, the net effect of a devaluation will depend on the relative importance of
the impact on production costs and on revenues of coffee growers both instant-
aneous and expected in the long term.
In order to test the empirical relevance of this hypothesis, a standard
reduced form equation for the determination of world coffee prices was
44 Brazil as an Export Economy, 1880±1930
estimated in logarithmic form, using annual data for 1852 to 1930. The Brazi-
lian share of world exports was substantial throughout the 1852±1930 period.19
In the basic specification, real coffee prices (PRICE) are a function of their own
lagged values and of the lag of a variable that tries to capture the supply±
demand balance in the coffee market (MARKET ). Here, this variable is con-
structed as the ratio of world coffee consumption to the sum of world coffee
supply and world coffee stocks. Therefore, an increase in this variable should
have a positive impact on real coffee prices. The inclusion of this explanatory
variable is justified by the special characteristics of the coffee market. Since the
product can be easily stored and production responds to prices with a lag of
several years, a standard model where supply and demand are functions of
current prices and determine the market price through a clearance condition
is not appropriate for the case of coffee (see De Vries, 1975). An additional
explanatory variable, the real exchange rate (RER), is included to capture the
increase in production costs associated with exchange rate fluctuations. The
variable is defined so that an increase in the index corresponds to a deprecia-
tion of the domestic currency and enters the equation with lags of one and five
years in order to capture both short-term and long-term effects. The reason for
the five-year lag is that at the beginning of the century production started, in
general, five years after coffee trees were planted, and some three-fourths of
total costs in the coffee sector were associated with fixed costs (see Rowe 1936,
p. 37).
Table 2.6 presents the main estimation results. Equation 1 displays the base-
line coffee price equation for the entire 1852±1930 period. The short-term
effects of the real exchange rate are not significant, neither are the effects of
the variable MARKET. Long-term real exchange rate fluctuations are shown to
have a significant effect on real coffee prices, with an estimated coefficient
significantly different from zero at standard confidence levels. A real devalua-
tion results in an increase in world coffee prices. In Equation 2 the short-term
RER variable is omitted and then the significance of the RER variable in the long
term is unfavourably affected. Results for subperiods such as 1852±1913 and
1930±60 proved to be disappointing. For the earlier period it is believed that
measurements may have been affected by the shortcoming of the statistical
series. Alternatively, the recession in the 1870s and early 1880s may also have
affected the results. For the period after 1930 the main difficulty is that the
exchange rate ceases to be as relevant as before, since foreign exchange controls
became the rule. However, for the golden age of coffee, the subperiod 1880±
1930 which is the core of this study, the results are much better. In Equation 3
of Table 2.6 it is shown again that the RER variable with a lag of one period is
not statistically significant. When RER
1 is omitted (Equation 4), all estim-
ated coefficients have their expected signs and are statistically different from
zero at standard confidence levels.
Marcelo de P. Abreu and Afonso S. Bevilaqua 45
a
Corrected for first-order serial correlation.
Notes: Figures in brackets are the t-statistic values of the estimations of the coefficients. Data are
adjusted to calendar years.
Sources: Coffee prices: imports into USA, United States (1975); world coffee production,
consumption and stocks: Bacha and Greenhill (1992), statistical appendix; domestic prices USA:
United States (1975); exchange rates: Brazil (1941b); average tariffs: computed from Brazil (1941b),
Brazil (1990) and Fritsch (1988); domestic prices Brazil: Goldsmith (1986), CataÄo (1992), Haddad
(1978) and Brazil (1990).
There was little scope for forward and conventional backward linkages in a
coffee-based export economy. Soluble coffee became important only in the
1950s and there is no mass market for processed coffee products. Traditional
backward linkages were also weak due to the rudimentary techniques used in
agriculture.20 Consumption linkages, however, were powerful and of special
interest. From rather early days there was involvement of coffee growers in the
building-up of export infrastructure, especially railways from the early 1870s,
and somewhat later, from the 1890s particularly, in the development of a
broadly-based import substitution industrialization process. This is a substan-
tial qualification of traditional views which underline the more or less perman-
ent opposition between the interests of coffee growers and industrialists as
providing the economic rationale for the accumulation of tensions, which
were eventually solved by a shift of policies in 1930 in favour of emerging
domestic industry and against coffee interests.
It is true that industrial interests in Brazil had considerably more political
weight in the Old Republic than in other developing economies. Industrial
interests were able to obtain important concessions from the government, in
particular a very protective tariff.21 It has been shown that due to the weight of
Brazilian coffee exports in world exports and to the price elasticity of world
coffee demand a high tariff in Brazil was transferred with a lag to world coffee
prices.22 The basic mechanism at work is through the increased production
costs in Brazil which, given Brazil's very high share of the world coffee market,
affect world prices.
Tariffs were also, of course, very important as a share of government revenue,
as alternative forms of taxation such as taxes on land and on income were
politically unpopular with the agricultural oligarchy, or difficult to implement
as in the case of excise or consumption taxes. During the later years of the
Empire the share of import duties in total central government revenue fluctu-
ated around 60 per cent. It fell slightly in the beginning of the century and
more substantially during the First World War. In the 1920s it was still around
40 per cent of total revenue.
What is argued here is not that high import tariffs were sought because it was
perceived that increased production costs would be transmitted to world coffee
prices. It is rather that once the political economy of protection, as for instance
lobbying by industrial interests, resulted in increased tariffs, there was no
significant deterioration in the performance of exports to justify countering
such tariff increases. Coffee growers were generally keen on maintaining low
production costs, as exemplified by their stance on labour costs since very early
in the nineteenth century. One cannot thus claim that a high tariff was adopted
because it was perceived that the foreign consumer would bear the increased
Marcelo de P. Abreu and Afonso S. Bevilaqua 47
a
Figures in brackets are the t-statistic values of the estimations of the coefficients.
Sources: See Table 2.6.
costs. It is simply that the policy of high protection was not countered by an
export lobby interested in maintaining low production costs as was the case in
most big primary commodity exporters.
The empirical relevance of this hypothesis for the period of this study is
tested with an extension of the framework for the determination of world
coffee prices used in the previous section. The analysis is based on annual
data for the subperiods 1880±1930 and 1880±1913.23 Real coffee prices are
explained by their own lagged values, the lag of the supply - demand balance
variable, the real exchange rate lagged five years, and the implicit tariff (TRF )
also lagged for five years. The lag in the TRF variable is explained by the same
reasons as the lag in the RER variable. As the results in Table 2.7 indicate, all
signs are as expected and all variables are significant at standard significance
levels for both periods. This means that Brazil was able to pass increased import
costs to higher world coffee prices.
moved away from the export economy and in the direction of industry. The
rationale for maintaining a single overvalued exchange rate was complex. The
traditional fiscal argument was, of course, that a sizeable share of government
expenditure was indexed to the exchange rate and thus increased automatically
with devaluation, while government revenues relied importantly on imports
which had been curtailed following the Great Depression. The government
through its monopoly power in the foreign exchange market had access to the
relatively cheaper (official) foreign exchange rate. The other reason was the fear
that a further devaluation would weaken coffee prices even more.
It is reasonable to believe that after 1930 Brazil continued to be able to pass
through increased production costs entailed by high tariffs to coffee prices, but
empirical work is prevented by the lack of data on the tariff equivalents of non-
trade barriers, particularly import rationing. As a last resort, the foreign
exchange reserves±imports ratio was used as a proxy for such non-tariff barriers,
but the empirical results were not satisfactory.
Multiple foreign exchange regimes were introduced in recognition of the
undesirable impact of a single official exchange rate regime on non-traditional
exports and on the cost of essential imports. Since the 1930s the Brazilian
government tried to stimulate non-traditional exports by assuring sales of
foreign exchange at more favourable rates.
The Brazilian exchange rate remained nominally stable from 1939 to 1953 in
spite of a difference in inflation relative to the USA which reached almost 140
per cent. This was possible, in spite of balance of payments problems from 1947
on, because of the sharp rise of coffee prices after 1945, but more especially after
1949. Coffee oversupply in the 1930s and early 1940s had become coffee
scarcity by the late 1940s, and the high coffee prices were to persist until
1954. After 1953 the multiple exchange rate regime also included multiple
export and multiple import rates and the government also used the exchange
regime as a rudimentary import tariff, since the Brazilian specific tariff had been
eroded by inflation. The wedge between the average import and export rates
was an important instrument fiscal for the government in the second half of
the 1950s, especially so as it escaped congressional control.
In the second half of the 1950s, as exports declined continually, foreign direct
investment was attracted by heavy subsidies and room to reap monopoly or
oligopoly profits behind a wall of absolute protection. Coffee prices fell after
1954 and the coffee industry was to again face a period of oversupply. While
there was no return to the policy of burning coffee stocks, there was a policy of
fostering the eradication of coffee trees in an effort to control supply. At the
turn of the decade, coffee control was to become an international concern after
many years of single-handed supply restraints by Brazil.
It is not easy to assess the impact of overvalued exchange rates on coffee
interests. In the short term it was likely that at the same time there was a
50 Brazil as an Export Economy, 1880±1930
Conclusions
Brazil was a price-maker in the international coffee market for the whole period
during which the economy can be reasonably described as an export economy.
This is a crucial element for understanding how policies were designed and
implemented. Even before coffee oversupply and the adoption of coffee valor-
ization policies there was no laissez-faire tradition in Brazilian economic his-
tory. Land and labour policies were designed and implemented to maintain the
profitability of the plantation-produced coffee export economy; valorization
depended directly or indirectly on financial guarantees by the federal govern-
ment. There was also complementarity between coffee price support and
Brazil's adoption of the gold and gold-exchange standards immediately before
the First World War and the Great Depression. There was, however, an inbuilt
mechanism in valorization which made inevitable the steady decline of Brazil's
share of the international coffee market as the support of artificially high prices
attracted high-cost competitive suppliers which otherwise would not have been
able to compete.
Government intervention was already important before 1930 also in relation
to protection of the domestic industry against the competition of imports. A
commercial policy based on a very high tariff made import substitution already
relevant from the turn of the century. The impact of such tariffs on coffee
production costs had as a consequence higher world coffee prices due to Brazil's
market power as the main producer and the price inelasticity of world coffee
demand.
The diversification of economic activity was the main source of improve-
ment of the standard of living. This was related to urbanization, the increased
role of government and the rise of industry. It was thus not surprising
that improvement in living standards affected regions very heterogeneously
with the north-east trailing badly behind the south-east, especially in the
1940s and 1950s. Until 1930, however, disparities were surprisingly small,
at a very low standard-of-living level. Coffee culture remained based on low-
wage manpower and concentrated in plantations with very little change in the
techniques used since the nineteenth century. This also applied to the use made
Marcelo de P. Abreu and Afonso S. Bevilaqua 51
of land resources as there were no limits to the advancing land frontier incor-
porated in the coffee economy within the relevant time span. Traditional tech-
niques which involved slash and burn remained the rule.
The scope of state intervention became even wider after 1930 when the
federal government took over coffee policies. Exchange regimes based on
exchange rate overvaluation and multiple rates became the rule, and the
wedge between average export and import exchange rates became important
for fiscal reasons. In some periods the government imposed additional levies on
exporters of traditional commodities, especially coffee, sugar and cotton.
Until the 1930s the protection to industry relied until the 1930s on high
tariffs, but then increasingly on access to low-cost imported inputs and capital
goods coupled with absolute protection assured by exchange controls. Brazil
`made the consumer pay' for its commercial policy based on high tariffs before
1930, and it is likely that it continued to do so after 1930. Exports remained
concentrated in a few agricultural commodities, especially coffee, until very
late, as industrial exports except in very special conditions such as during the
Second World War were unable to compete in international markets. There was
no introduction of new commodity exports until the soya boom in the late
1960s and early 1970s.
Notes
1 The British share of Brazilian total imports fell less markedly, but in the 1900s it was
down to 30 per cent compared to more than 50 per cent before 1880.
2 Brazil imported some foodstuffs, especially before 1930, but only wheat was an
important item of the import bill.
3 Our own estimates. For pre-1913 GDP and deflator data see Goldsmith (1986), pp. 22±
3 and 82±3.
4 See Eisenberg (1974), ch. 1 for a general view of sugar production and exports in
Brazil.
5 See Canabrava (1984) on the short-lived cotton boom in SaÄo Paulo in the 1860s and
1870s.
6 Weinstein (1983) and Schurtz et al. (1925) are the standard sources on the rubber
boom and collapse in the Amazon region.
7 See Stein (1957) for the boom and downfall of Vassouras as a coffee region in the
nineteenth century.
8 See Dean (1995) chs 8±9, for the environmental consequences of coffee culture
expansion.
9 See Dean (1971) and Reis and Reis (1988) for land policies and labour policies in Brazil
and the fragmentation of oligarchic interests.
10 Rowe (1936), p. 29 suggests a much lower threshold of 4000 trees as compatible with
operation of a farm by a family without hiring labour. The importance of the small-
scale coffee production sector would be even smaller if this was the case.
11 See Hatton and Williamson (1994), pp. 7, 44. For immigration to SaÄo Paulo see
Holloway (1980), especially ch. 3.
52 Brazil as an Export Economy, 1880±1930
12 See LalieÁre (1909), part III, for a comprehensive explanation of machinery used in
best-practice farms for the preparation of coffee.
13 Even if the suspiciously high stock of French capital is taken at face value.
14 See Cardoso (1975) for the classic treatment of this issue.
15 See Reis and Reis (1988), for a discussion of the sharply contrasting stance of different
segments of the agricultural oligarchy on land policies, slavery and immigration in
late nineteenth century Brazil.
16 To put these numbers in historical comparative perspective, infant mortality in
England and Wales was 154/1000 live births in 1900, and life expectancy at birth
was 48 years.
17 Granger causality tests show that for the 1850±1930 period as well as for all the
relevant subperiods one cannot reject the hypothesis that there is no causal link
between changes in real coffee prices and changes in the Brazilian real foreign
exchange rate. For early perceptions of the importance of the exchange rate in the
determination of coffee prices see Gudin (1933) and Williams (1934). As noted by
Furtado (1965), Wileman (1896) detected the link between coffee prices and the
exchange rate. Such a link was investigated empirically by Cardoso (1983).
18 For conditions of coffee planting and costs in the 1880s see LaeÈrne (1885), ch. X; for
current costs in the 1900s see LalieÁre (1909), Part 4; on the importance of fixed over
current costs see Rowe (1936), p. 37, and for detailed costs for plantations at different
stages of maturity of coffee trees in the 1920s see Rowe (1932), pp. 88±9.
19 During 1852±1930 the average Brazilian share of world coffee exports was about 51
per cent. It was about 55 per cent in 1880±1930.
20 See Hirschman (1958), ch. 6, and Hirschman (1981) for linkage taxonomies.
21 For the history of protection under the First Republic (1889±1930) see Villela (1993).
22 See Abreu, Bevilaqua and Pinho (1997) for a treatment of import substitution in Brazil
from the 1890s to the 1960s.
23 After the balance of payments crisis following 1930, tariffs are a poor measure of
protection since exchange controls became the rule.
24 There was no scarcity of big spenders, but EpitaÂcio Pessoa and the building of dams in
the north-east in the early 1920s can be singled out.
25 As it was in 1931±34, 1937±39 and 1947±53.
26 This was so in 1934±37, 1939±46 and 1953±64.
References
LalieÁre, A. (1909) Le cafe dans l'eÂtat de Saint Paul (BreÂsil), Paris: A. Challamel.
LeCointe, P. (1922) L' Amazonie breÂsilienne, 2 vols, Paris: A. Challamel.
Lewis, W.A. (1978) The Evolution of the International Economic Order, Princeton: Princeton
University Press.
Lopes, Juarez (1968) Desenvolvimento e MudancËa Social. FormacÄo Ëa da Sociedade Urbano-
Industrial no Brasil, SaÄo Paulo: Companhia Editora Nacional.
Mitchell, B.R. and Deane, P. (1971) Abstract of British Historical Statistics, Cambridge:
Cambridge University Press.
Ocampo, J.A. (1984) `La economõÂa colombiana en la deÂcada del treinta', in J.A. Ocampo
and S. Montenegro, Crisis mundial, proteccioÂn e industrializacioÂn. Ensayos de histoÂria
econoÂmica colombiana, BogotaÂ: CEREC.
Palacios, M. (1983) El Cafe en Colombia. 1850±1970. Una HistoÂria EconoÂmica, Social y
PolõÂtica, Mexico: El Colegio de MeÂxico/El A Â ncora.
Reis, E. and Reis, E. (1988) `As Elites AgraÂrias e a AbolicËaÄo da EscravidaÄo no Brasil', Dados,
vol. 31 (3).
Rowe, J.W.F. (1963) The World's Coffee. A Study of the Economics and Politics of the Coffee
Industries in Certain Countries and the International Problem, London: HMSO.
ÐÐ (1936) Markets and Men, A Study of Artificial Control Schemes in some Primary Indus-
tries, Cambridge: Cambridge University Press.
ÐÐ (1932) Studies in the Artificial Control of Raw Material Supplies. No. 3 Brazilian Coffee,
Memorandum No. 34, Royal Economic Society, London, February.
Schurtz, W.L., Hargis, O.D., Marbut, C.F. and Manifold, C.B. (1925) Rubber Production in
the Amazon Valley, US Department of Commerce, Bureau of Foreign and Domestic
Commerce, Washington, DC: Government Printing Office.
Stein, S. (1957) Vassouras, A Brazilian Coffee County, 1850±1900, Cambridge, Mass.:
Harvard University Press.
Stolcke, V. (1986) Cafeicultura. Homens, Mulheres e Capital (1850±1980), SaÄo Paulo: Brasi-
liense.
Stone, I. (1987) The Composition and Distribution of British Investment in Latin America, 1865
to 1913, New York: Garland.
Summerhill, W.R. (1995) `Railroads and the Brazilian Economy before 1914', doctoral
dissertation, Stanford University.
United Kingdom (1987) Central Statistical Office. Social Trends no. 17, London: HMSO.
United States Department of Commerce (1975) Historical Statistics of the United States.
Colonial Times to 1970, Washington, DC: Government Printing Office.
Villela, A.A. (1993) `PolõÂtica comercial e importacÄes Ëo na Primeira Repu blica: 1889±1930',
M.A. dissertation, Department of Economics, Catholic University, Rio de Janeiro.
Weinstein, B. (1983) The Amazon Rubber Boom, 1850±1920, Stanford: Stanford University
Press.
Wileman, J.P. (1896) Brazilian Exchange. The Study of an Inconvertible Currency, Buenos
Aires: Galli Brothers.
Williams, J.H. (1934) `American Foreign Exchange Problems in Brazil, Argentina, Chile
and Uruguay, 1934', reprinted in Foreign Relations of the United States 1934, Washington,
DC: Government Printing Office.
3
Coffee and the Origins of Modern
Economic Development in Colombia
Jose Antonio Ocampo and MarõÂa Mercedes Botero
Introduction
In comparison with Brazil, Mexico and the Southern Cone economies, Colom-
bia may be characterized as a latecomer to export and industrial development.
Throughout the nineteenth century, Colombia struggled to establish a solid
export base. Its efforts bore little fruit, however, partly because of the nearly
continuous political instability the country experienced throughout the cen-
tury, associated with the difficult process of building a national state after
achieving political independence in 1819. Moreover, the notorious geograph-
ical barriers which characterized the country tended to fragment the economy
into a set of semi-autonomous regions.
In a departure from the nineteenth-century legacy, the twentieth century
produced sustained economic development from the start. The basis for devel-
opment was the establishment of a coffee economy made up of small and
medium-sized producers in the west of the country. Based on coffee production,
and to a lesser extent on a few enclave sectors, Colombia experienced a major
export boom during the two decades prior to the Great Depression of the 1930s.
The expansion of coffee represented not only an appreciable increase in
foreign trade volumes, but also a qualitative change in Colombian develop-
ment patterns. The country's unstable export trade, which in the nineteenth
century depended on exceptional, transitory conditions in international mar-
kets, gave way to a stable and competitive export sector. More importantly,
coffee finally created a cluster of activities that were to break away definitively
from the pre-capitalist economic pattern the country had inherited from the
colonial period, a pattern which the export expansion of the nineteenth cen-
tury had barely touched.
Thanks to the expansion of domestic demand generated by the export boom,
the aggressively protectionist policy of General Rafael Reyes' Administration in
the second quinquenium of the century and the partial shutdown of import
55
56 Coffee and Economic Development in Colombia
markets during the First World War, the industrialization process took off.
However, manufacturing development was still in its infancy during the
1920s. Geographic fragmentation was undoubtedly the reason why manufac-
turing lagged behind in Colombia compared to other large and medium-sized
countries in the region. Nevertheless, the coffee boom, the establishment of the
basis of a modern transport network and the manufacturing experience accu-
mulated prior to the Great Depression created the foundations for the rapid
process of structural change that the Colombian economy was to experience
after 1930.
In this chapter, we analyse the features of Colombia's export expansion and
its relationship to economic development in the closing decades of the nine-
teenth century and the opening decades of the twentieth. The chapter is
divided into five sections, the first of which is this introduction. The second
briefly reviews nineteenth-century development. The third section explores the
most noteworthy aspects of the process of transformation that the Colombian
economy experienced in the first three decades of the twentieth century. The
fourth and fifth examine in greater detail the development of the coffee and
manufacturing sectors in the latter period.
a
Based on purchasing power of exports in terms of textiles, with respect to 1855±58.
b
Excludes 1844/45.
c
1855/56±1858/59.
d
Excludes 1864/65.
e
Based on volume of imports for 1845±49 and trends in textile prices.
f
1892±95.
g
1896±98.
Source: Ocampo (1984). The value of imports for the first period has been reestimated using original
sources.
East. From the late 1880s to 1898 there was another phase of rapid export
growth (5.5 per cent annually between 1888±91 and 1898) based on the only
two products that showed sustained growth in the second half of the century:
precious metals and coffee. Both booms ended in bust around the turn of the
century. Nonetheless, the first played a key role in the economic development
of Antioquia, which from the end of the century onwards was to become one of
the pillars of modern industrial development in Colombia. By bringing coffee-
growing to the west of the country, the second was to lay the foundations for
export development in the twentieth century.
The expansion of gold and silver production in Antioquia and other regions
of the country in the 1870s reversed the downward trend in precious metals
extraction that had begun at the end of the colonial period. The freeing of the
slaves in mid-century had dealt the final blow to the already weakened slave-
based economies of Choco and Cauca in the Colombian Pacific. By contrast,
starting in the eighteenth century, gold-mining had become the mainstay of
economic expansion in Antioquia. This process, far from slowing down after
independence, continued throughout the nineteenth century and underwent
significant modernization. According to Brew (1977), vein-mining was unques-
tionably a seedbed of technical innovation and a channel for modern industrial
development in Antioquia.
58 Coffee and Economic Development in Colombia
abroad. This division of labour worked to the advantage of the big MedellõÂn
merchants, who even ended up managing the ore produced by foreign firms.
The only exception to the rule were some of the large locally-owned mining
companies, who sold gold bars (and/or bills of exchange) in MedellõÂn to finance
their operating costs or exported them to their own agents abroad. There is, on
the other hand, no indication that foreign capital was involved in the market-
ing of the gold.
Along with the expansion of mining came development in other areas of
activity. As the region's capacity to pay for imports increased, other commercial
firms appeared in the state ± later departmental ± capital and some other
localities. MedellõÂn became an important mercantile and financial centre, and
between 1870 and 1883, three banks serving the public and seven private bank-
ing firms were established. In addition, commercial cattle-raising underwent
substantial development, since meat formed part of the miners' staple diet.
The laws and official policy of the state of Antioquia offered great support to
the mining industry. The regional governments set about improving and devel-
oping the transport network, constructed a telegraph system linking MedellõÂn
with the mining districts, promulgated a mining code and established a mint.
Lastly, and perhaps most importantly, the development of mining gave impetus
to technical education and the training of a workforce with the skills to handle
modern machinery. These elements were crucial for the industrial development
of MedellõÂn later on.
As for coffee, in 1860 it was still a minor item in Colombia's foreign trade,
representing only 4 per cent of total exports.1 From then on, however, coffee
production underwent three successive booms in the nineteenth century
which generated a significant shift of the production frontier in the interior
of the country. The first of these expansions took place during the 1860s.
Exports increased to some 100 000 bags, but production remained concentrated
in North Santander, near the Venezuelan border, where coffee-growing had
been introduced in the early decades of the Republic.
The second expansion occurred during the 1870s. Coffee exports increased
from 100 000 to 200 000 bags, and became one of the country's major export
staples. By the end of the decade it accounted for more than 20 per cent of
Colombia's external sales. This period of expansion involved a major shift in
the production frontier. Although North Santander continued to increase acre-
age planted and raised its output to 120 000 bags per year, most of the new
production came from Santander and Cundinamarca, which by mid-decade
were producing 60 000 and 40 000 bags, respectively (Table 3.2). In both these
regions the expansion was interrupted by falling international prices in the
early 1880s and later by civil war.
Soon after public order was restored in the mid-1880s, the largest coffee
boom of the century began. The country's coffee production expanded rapidly;
60 Coffee and Economic Development in Colombia
a
Included in Antioquia.
b
Included in Cundinamarca.
Sources: 1874±1900: Ocampo (1984); for 1913: Bell (1922); for 1932: Coffee census, Revista Cafetera de
Colombia, nos 34±35, January±February 1932.
USc/lb (1980 dollars)
240
220
200
180
160
140
120
100
80
60
40
1830
1833
1836
1839
1842
1845
1848
1851
1854
1857
1860
1863
1866
1869
1872
1875
1878
1881
1884
1887
1890
1893
1896
1899
1902
1905
1908
1911
1914
1917
1920
1923
1926
1929
1932
1935
1938
1941
1944
by the time the last and most serious of the nineteenth century civil wars broke
out, the Thousand Days' War (1899±1902), coffee output had reached 600 000
bags and represented nearly half of Colombia's exports.
The expansion coincided with a period when real international coffee
prices were at their highest levels ever (Figure 3.1), which was no doubt the
Jose Antonio Ocampo and MarõÂa Mercedes Botero 61
fundamental reason for the boom. Other factors, however, influenced the
process. As a reflection of the general crisis in Colombia's foreign trade in
the early 1880s, the declining price of silver in international markets and the
subsequent introduction of paper money, the peso underwent a sharp devalua-
tion. During the initial phase of this process, coffee production costs tended to
decline in terms of gold, which encouraged planting.2 No less important was
the fact that landowners had access to credit from foreign trading companies
on terms that were attractive at that time: an average interest rate of 6.6 per
cent per annum ± with 1.5 or 2 additional points by way of commission ±
maturity of two years and repayment in coffee.
The boom of the 1890s was the most diversified of the century in regional
terms. The key expansion area was western Cundinamarca ± and localities in
neighbouring Tolima ± within the sphere of influence of BogotaÂ. In addition,
production began to expand appreciably in Antioquia, which was producing
some 90 000 bags by the end of the century (Table 3.2).
In Santander and North Santander, coffee production took a variety of
forms and lacked some of the semi-servile aspects that characterized labour
relations in the large plantations in Cundinamarca. This divergent develop-
ment reflected the history of each of these regions. In Cundinamarca, the
evolution of social relations was imprinted by three centuries of domination
of the indigenous population. In the Santanders, in contrast, smallholders
and craftsman had played an important role in the economy since colonial
times.
In Cundinamarca, although small plantations were not unknown, the pre-
dominant model was the large estate, or hacienda, with a permanent workforce
consisting primarily of tenant farmers who were in some respects in a servile
position vis-aÁ-vis the landowner. The tenants constituted the core of the
hacienda workforce. In exchange for a dwelling and an allotment of land for
subsistence farming, the tenant was required to work a certain amount of time
on the plantation, usually at half of what would ordinarily be a day's pay. For
the seasonal work of harvesting, the plantation depended on temporary day
labourers, most of whom, like the tenants, came from the high mountain
plateaux, or altiplano (Palacios, 1983).
Although this form of social organization had elements of serfdom and relied
upon coercion of the labour force with the support of local political authorities,
it cannot be described as a step backward with respect to the labour arrange-
ments on the altiplano haciendas, as some authors have suggested (Kalmanovitz,
1985). On the contrary, Palacios (1983) has shown that, due to the mix of
permanent labour scarcity with an open agrarian frontier, the basic tendency
in western Cundinamarca was towards the strengthening of a peasant economy
on the plantations, a situation that was to lead to the agrarian conflicts of the
1920s (see below).
62 Coffee and Economic Development in Colombia
Exports, economic policy and their impact on the rest of the economy3
At the start of the twentieth century, both the large precious metal mining
operations and the large coffee plantations suffered a collapse, bringing to a
close the longest-lasting export experiments of the nineteenth century. The
inability to develop a solid export base was undoubtedly the main cause of
Colombia's slow economic development in the last century, which left the
country on the periphery of international flows of trade, capital and labour.
Despite these limitations, the expansion of exports in the second half of the
Jose Antonio Ocampo and MarõÂa Mercedes Botero 63
60
% 55
50
45
40
35
30
25
20
15
10
1867/8
1871/2
1875/6
1881/2
1850/1–54/5
1865/6
1873/4
1886
1888
1890
1892
1894
1896
1898
1900
1902
1904
1906
1908
1910
1912
1914
1916
1918
1920
1922
1924
1926
1928
1930
1932
1934
1936
1938
1940
1942
1944
1832/3–35/36
1883/4
1879/80
1877/8
1869/70
1840/–43/4
200
% 180
160
Raw cloth
140
120
Ordinary
cloth
100
80
60
40
20 Coloured
0
1911
1931
1921
1941
1828–31
1848–51
1905
1929
1913
1915
1933
1935
1937
1919
1925
1945
1917
1927
1939
1885–86
1943
1871–73
1890–92
1923
1856–59
1947
1842–44
1865–67
1896–98
1877–80
1835–37
In fact, all the major liberal tariff reforms ± 1847, 1861, 1870 and 1873 ± were
followed a few years later by large customs surcharges that in effect consider-
ably raised taxes on imports. The basic reason for the contradiction between
free-trade principles and the actual trend in tariffs was the increasing fiscal
dependence on customs revenues (Ocampo and Montenegro, 1984). Except
for a few brief periods during which the revenues from tobacco, salt, confis-
cated church property or issuance of money ± during the Thousand Days' War ±
exceeded customs revenues in importance, over the rest of the period in ques-
tion customs were the national government's chief source of revenue.
In the 1880s, under Rafael Nu n Äez' Regeneration movement, protectionist
ideas reemerged with renewed force. Contrary to what is generally believed,
however, the reforms carried out during the Regeneration in the long run
actually reduced existing duties. The basic explanation is that the specific ad-
valorem duties levied on imports were eroded by the inflation that the country
experienced following the devaluation first of silver and later of paper money.
Although the ad-valorem duties reached their height in the period 1885±86,
the average tariff and most of the ad-valorem duties stabilized in the rest of the
decade at levels comparable to those at the close of the 1870s and fell in the
1890s. The supposed protectionist effect of tariff reforms under the Regeneration
was thus largely illusory, although it did result in greater tariff dispersion, which
generated greater effective protection for the same level of nominal protection.
Trends in Colombian tariff policy in the nineteenth century thus had very
little to do with modern controversies about free trade and protectionism. In
the absence of an alternative fiscal base, the need for a sufficient level of public
revenue was the main factor determining tariff movements.
Economic policy played a very modest role in the decline of the handicraft
industry, especially textiles, during the nineteenth century. The backwardness of
the transport system and the effective increase in tariff protection tended to
counteract the effect of sharp reductions in international prices for some
imported goods, especially textiles. Moreover, as some nineteenth century ana-
lysts pointed out at the time, economic growth offered an opportunity of expan-
sion for handicraft activities that were not in direct competition with imports to
expand ± dressmaking, furniture production and printing, to name a few.
The marked fragmentation of the domestic market, the protection generated
by high transport costs and the attempts at protectionism in the early years of
the Republic encouraged a number of experiments in industrialization. Most of
these pioneering efforts were unsuccessful or prospered only on a modest scale.
The first initiatives of importance were undertaken in and around Bogota in the
1830s, when a number of small factories ± producing textiles, pottery, glass,
paper and iron ± were set up. Nearly all these enterprises quickly failed owing to
a lack of markets, adequate raw materials and skilled labour (Ospina VaÂsquez,
1955; Safford, 1965).
66 Coffee and Economic Development in Colombia
The return to peace and the rapid economic stabilization achieved after the
Thousand Days' War, together with the vigorous export promotion and protec-
tion policy introduced by General Rafael Reyes (1904±09), began a period of
rapid economic growth which lasted until the late 1920s and laid the founda-
tions for the structural changes that were to take place in the country's eco-
nomy from the Great Depression onwards.
The economic growth that took place during this period was based on a
strong export expansion. Between 1905±09 and 1925±29, real exports grew at
an annual rate of 7 per cent. The basis of that expansion was the export of
coffee, which grew by 7.3 per cent per year during this period (Table 3.3). Two
other export products contributed to the expansion: bananas, from the first
decade of the century, and oil, starting in the 1920s. Other export products,
including precious metals, lagged behind. The export structure which had
remained diversified until the turn of the century thus became much more
concentrated which, of course, made the whole economy much more vulner-
able to coffee price fluctuations but reflected the great competitiveness of the
coffee industry that developed during the period.
There are no precise estimates of overall economic growth during the period.
On the basis of very incomplete information, Maddison (1995) and Hofman
(1998) estimate that the Colombian economy grew at 3.9 to 4.5 per cent a year
between 1900 and 1929,6 which would indicate that exports made up 13.5 to
15 per cent of GDP in 1905±09 since existing figures show that share having
reached 24 per cent in 1925±29 ± estimated at 1975 prices. Growth was prob-
ably in the higher part of the range;7 it was certainly higher than the regional
and worldwide and regional averages for the first three decades of the century ±
2.2 per cent and 4.1 per cent respectively, according to the same authors ±
although it was lower than that of the fastest-growing Latin American eco-
nomies.
However, economic expansion was uneven over time; its ups and downs were
basically due to fluctuations in the external sector. The growth which charac-
terized the period from 1905 to 1913 was reflected in annual growth of 22 per
cent in deposits in the three major banks of Bogota (Romero, 1994). The period
ended with a sharp increase in the terms of trade and external purchasing
power (see Figure 3.1 and Table 3.3). During that period, as had been the case
during the boom periods of the nineteenth century, there was an influx of gold
coins which facilitated the growth of currency in circulation.
From 1914 to 1922, on the other hand, the country went through a period of
slowdown and sharp fluctuations associated with political and economic
events at the international level. The terms of trade were highly unstable during
68
Period Coffee exports Export volumes Purchasing Composition of exports (in Import volumes Machinery
(thousands of (1925±29100) power of exports dollars at current prices)b (1925±29100) imports
bags) (1925±29100) (1925±29100)
Coffee Mineralsa Other
1900±04 542
1905±09 604 25.7 39.0 19.7 41.3
1910±14 837 37.3 38.4 47.4 16.3 36.3 29.0
1915±19 1 244 50.2 39.9 55.1 7.9 37.1 25.4 6.8
1920±24 1 906 74.3 52.8 75.5 5.6 8.9 41.1 26.4
1925±29 2 451 100.0 100.0 67.7 17.9 14.3 100.0 100.0
a
Oil, coal, nickel and gold.
b
Excludes other.
Sources: Coffee exports: 1900±10, Ocampo (1984); 1910±29, National Federation of Coffee Growers; export volumes and exports' purchasing power: Ocampo and
Montenegro (1984); exports as % of GDP: CEPAL (1957); composition of exports: Foreign Trade Yearbooks and records of purchases of monetary gold by the Banco de
la RepuÂblica; import volumes and imports of machinery: Ocampo and Montenegro (1984).
Jose Antonio Ocampo and MarõÂa Mercedes Botero 69
that period, and external purchasing power was on average relatively stagnant.
The annual rate of growth in deposits at the main Bogota banks was only 2.6
per cent, at a time when their productive assets were stagnant (Romero, 1994).
In critical years, particularly 1914 and 1920, they suffered severe contraction.
During the first years of that period, surpluses accumulated by the Amortization
Board were used to finance public spending, postponing the possibility of
giving metal backing to the currency in circulation. Export restrictions imposed
by the belligerent states were reflected in a balance of payments surplus, but the
inconvertibility of their currencies, which continued during the postwar per-
iod, prevented the surplus from increasing the currency in circulation; a tem-
porary appreciation of the peso therefore ensued. Faced with fiscal problems
again in 1919 and 1921, the government issued treasury bonds. In addition, the
sharp postwar fluctuations in international prices led to the bankruptcy of
major coffee traders and some banks.
Beginning in 1923, the economy entered a phase of faster expansion led by
strong export growth and improved terms of trade, together with the payment
by the USA between 1923 and 1925 of US$25 million in compensation for the
independence of Panama, and the receipt of US$200 million between 1926 and
1928 in net foreign borrowing. The latter was equivalent to 35 per cent of
exports for 1925±29. The country finally had access to the international capital
market, a source of income that had been unavailable for more than a century
owing to the country's inability to pay off its foreign debt left over from the War
of Independence (Junguito, 1995). All this was reflected in an unprecedented
boom in imports, and in levels of public and private capital formation which
had never before been experienced in the country (Table 3.3). Despite increased
imports, the balance of payments surplus led to substantial growth in interna-
tional reserves and currency in circulation; the latter grew by about 20 per cent
per year from 1923 to 1928. Annual GDP growth was 7.4 per cent in 1925±29,
the first years for which such estimates are available. Thus, the country ended
this phase of `outward-oriented development' with a spectacular surge in for-
eign borrowing, imports and investment (CEPAL, 1957; Ocampo and Montene-
gro, 1984).
Urbanisation progressed rapidly during this period: while the country's popu-
lation grew at an annual rate of 2.1 per cent between 1912 and 1938, the 12
main cities grew at a rate of 3.9 per cent. There was strong growth in all urban
activities, including manufacturing (see later), commerce, banking, construc-
tion and government services.
Transport also made notable progress. The railway network expanded con-
siderably, steam-powered shipping flourished and the first highways were built.
In the early 1930s, Colombia's railway network (3262 km) was 6.6 times more
extensive than in the late nineteenth century (McGreevey, 1971). The Pacific
and Caldas railways had created the first integrated network, linking the
70 Coffee and Economic Development in Colombia
coffee-growing areas to the Pacific coast. The opening of the Panama Canal in
1914 justified the development of a modern seaport on the Pacific coast, which
was the natural outlet for the new coffee-growing areas in the west of the
country. Steam-powered shipping on the Magdalena and Cauca rivers and the
gradual introduction of automobile transport also contributed to this revolu-
tion in domestic transport systems. However, most infrastructure investments
came late, following the 1925±29 investment boom. Meanwhile, although the
modern transport network was still far from complete and its separate compo-
nents had been developed mainly to link the various regions of the country to
the outside world, Colombia had, for the first time, the basis for a minimum
level of domestic market integration.
In any event, the country's productive structures were still very backward at
the end of the coffee boom. In 1925±29, manufacturing accounted for less than
8 per cent of GDP, estimated in 1975 prices, whereas the share of the agricul-
tural sector was still 48 per cent. In comparison with international patterns
there was relative `over-expansion' of agriculture and a considerable lag in
manufacturing (London Äo, 1989). This was doubtless attributable to the geo-
graphical complexity of Colombia and to the very late development of modern
means of transportation. Manufacturing could not thrive in an economy
which, aside from the low standards of living of the population, was still
characterized by a considerable fragmentation of the domestic market.
The expansion of government activities was a decisive factor in the trans-
formation process. However, as was common at the time, the state was still
relatively small and had limited means of action. The most significant state
reform was that which took place in 1923 as a result of the Kemmerer Mission.
In that year, the Banco de la RepuÂblica was founded; it was the first central bank
in the modern sense, in that it had a monopoly on the issue of money, acted as
the central repository for bank reserves and was the lender of last resort (Banco
de la RepuÂblica, 1990). Its creation was also the culmination of a long transition
towards a stable gold standard, which had begun after the Thousand Days' War;
the process had involved various mechanisms for guaranteeing the convert-
ibility of the currency and had involved a number of initiatives to replace the
nineteenth century Banco Nacional and the Banco Central created in the early
twentieth century by General Reyes (Avella, 1987). The founding of the new
central bank was accompanied by the creation of a banking supervisory body,
the Superintendencia Bancaria; until then, private financial institutions had not
been regulated. Finally, the Office of the Controller-General of the Republic was
created as a modern instrument of fiscal supervision.
In social terms, the country also underwent significant changes during the
period. The relative size of the wage-earning workforce increased, providing the
basis for trade-union movements which began to form in the 1910s and experi-
enced their first period of expansion in the 1920s. The main areas in which
Jose Antonio Ocampo and MarõÂa Mercedes Botero 71
trade union struggles developed were steam-powered shipping and the export
sectors (Urrutia, 1969; Archila, 1989). The most significant event in terms of its
political implications was, unquestionably, the 1928 banana strike (White,
1978; LeGrand, 1989), immortalized by Gabriel GarcõÂa MaÂrquez in his classic
novel, One Hundred Years of Solitude. At the same time, the availability of alter-
native employment exacerbated conflicts which had long existed in certain
rural areas; as we will see, these were particularly significant in the large
plantations of Cundinamarca, which had been the heart of the expansion of
coffee-growing in the late nineteenth century.
There are few quantitative indicators of the trend in the living conditions of
the population. The first available data for most social indicators date from the
mid-twentieth century and are all unsatisfactory. It is clear that limited progress
was made in the first half of the century. Nevertheless, the acceleration in
population growth from 1.5 per cent per year in the nineteenth century to
2.1 per cent in the first decades of the twentieth century was due in part to
falling death rates, no doubt a reflection of an improvement in living condi-
tions. The increase in the average height of the population since the turn of the
century is also evidence of a gradual improvement in nutrition among the
population as a whole (Ordon Äez and PolanõÂa, 1992).
The coffee boom of the early 1900s entailed significant changes in regional
terms and in the organization of production. As we have seen, most of the
traditional large plantations in eastern Colombia were profoundly affected by
the Thousand Days' War and by the collapse of international prices at the turn
of the century and were unable to expand output once peace was restored and
coffee prices recovered. As a result, production in the east of the country, where
plantations were concentrated, rose very slowly from some 500 000 bags in
1890 to 825 000 bags in 1932, a 1.6 per cent annual increase (Table 3.2). The
main problem they faced was their inability to change their labour relations,
and they remained trapped in servile land tenancy arrangements. In the 1920s,
these plantations became sites of agrarian conflict and, in the following decade,
one of the key targets of the first wave of agrarian reforms, introduced by the
reformist liberal governments of the time (Bejarano, 1979; Palacios, 1983).
At the same time, the coffee cultivation began to flourish in the west of the
country, in the steep but rich volcanic soils of the central mountain range
where Antioquen Äo migration had settled down during the course of the nine-
teenth century. Indeed, the expansion of production centred in the Antio-
quenÄo-settled areas in Caldas, the south of the department of Antioquia, and
the north of Tolima and the Cauca valley. The settlers in those areas had
originally employed slash-and-burn methods of agriculture similar to those
72 Coffee and Economic Development in Colombia
used in other parts of the country, while also engaging in some supplementary
income-producing activities such as pig-raising and searching for archaeolog-
ical treasures. Coffee-growing had not been a major part of their livelihood,
even in those regions that were settled late (QuindõÂo). Nevertheless, from the
turn of the century onwards, coffee-growing expanded rapidly in the small and
medium-sized holdings that had developed in the region as a result of a pro-
tracted struggle over land between settlers and the owners of the idle land.
Coffee-growing expanded at a remarkable pace in the region, with coffee pro-
duction increasing at rates close to 10 per cent a year during the first three
decades of this century so that by 1932 the three main coffee-growing depart-
ments in western Colombia ± Antioquia, Caldas and Valle del Cauca ±
accounted for 57 per cent of domestic output (Table 3.2).
In sharp contrast to the patterns that characterized the central and eastern
parts of the country, production in the west concentrated in small and med-
ium-sized holdings. Consequently, in 1932, 59.5 per cent of coffee bean pro-
duction came from plantations of 12 hectares or less. The proportion was
higher in Caldas and the Cauca valley, while in Cundinamarca and Santander
larger-sized holdings continued to be the norm at that time (Tables 3.4 and 3.5).
In any event, the egalitarian nature of production should not be overstated. The
majority of producers (76 per cent in 1932) had coffee plantations of less than 3
hectares, but their share of production was relatively small (26 per cent accord-
ing to Table 3.4). Social differentiation in the rural economy was as important a
reality as growth in the number of small-scale producers, as evidenced by the
increasing importance of medium-sized farms (Machado, 1988).
The workforce on the smallest holdings was made up exclusively of family
members, although the use of family labour was not highly intensive, except at
harvest time. Coffee-growing methods were simple, and coffee was supplemen-
ted by other crops and animal husbandry. On the largest farms outside labour
was used, especially during harvesting. Hiring practices in Caldas included
Jose Antonio Ocampo and MarõÂa Mercedes Botero 73
Table 3.5 Concentration of production in main departments, 1923 and 1932 (by size of
mature coffee trees)
payment on a per diem basis and `company systems'. The latter were not very
different from the share-cropping arrangements in eastern Colombia. The
workers received 50 per cent of the coffee harvested, with an additional deduc-
tion if it were delivered to the landowner unprocessed. As in the east of the
country, a company system was also used for coffee planting. Under this
arrangement, the worker could make full use of the land during the term of
the contract (generally four years) and would be required to sell the mature
trees to the owners.
One of the most important innovations in the coffee industry in the first
decades of the twentieth century was the advent of the manual pulper. Pulping
machines became the main product of the Amaga ironworks in Antioquia and
became indispensable on all coffee farms. In 1915, 70 per cent of Antioquian
farms had such a machine, a figure which had risen to 97 per cent by 1922
(Palacios, 1983). This innovation made it possible to shift the main processing
tasks other than hulking ± depulping, washing and drying of the beans ± from
the large plantation estates and small processing centres to the smaller farms,
thus bringing the farmers into direct contact with urban traders.
In the areas settled by AntioquenÄos, the relationship between the peasant
and the store-owner ± fondero ± replaced that between tenant and plantation
owner more typical of the eastern part of Colombia. The store-owner would
advance the peasant the credit he required for purchases throughout the year
and in return would take the coffee crop at a significant discount. The interest
rates applied in these transactions reached usurious levels and were a major
bone of contention until the Agrarian Bank and the National Federation of
Coffee Growers intervened in the 1930s to put an end to this unfair practice.
The store-owner was only the first link in the commercialization of coffee.
Coffee beans were delivered to the agents of export firms located in the more
important towns; these agents would then send the beans on to commercial
74 Coffee and Economic Development in Colombia
coffee trade, the manufacture of bags and pulping machines and the growth
of the hulking industry. The critical difference, however, compared to the
nineteenth century, was the development of a modern transport network.
The competition between MedellõÂn, Manizales and Cali to control the coffee
business depended on upgrading transport systems. In this process, MedellõÂn
increasingly lost ground to Cali, thanks to the integrated transport net-
work developed between the coffee-growing area and Buenaventura on the
Pacific which was controlled by Cali interests. This certainly put pressure on
MedellõÂn to seek out new economic activities, especially in the manufacturing
sector.
The creation in 1927 of the National Federation of Coffee Growers consolid-
ated coffee development by facilitating public action in favour of the sector.
However, the Federation's intervention in the domestic coffee market only
really began during the early years of the Great Depression, when it built the
first warehouses and stepped in to purchase crops in regions where traders'
profit margins were excessive. The Federation was a private organization, but
this did not prevent it from gradually assuming public functions. From the time
the Federation was founded, it adopted a special mechanism which enabled the
government to participate in policy-making; the National Coffee Committee,
on which the government was represented, took policy decisions which the
Federation subsequently carried out.
The impetus generated by the export sector was sufficient to consolidate early
attempts at industrialization prior to the crisis of the 1930s. By that time, as the
eminent historian of manufacturing and protectionism in Colombia pointed
out, the era of industrial pioneering was already a thing of the past (Ospina
VaÂsquez, 1955). Thus, as in other Latin American countries similar or larger
in size, the first phase of industrialization was closely linked to export
expansion.10
As one might expect, the relationship between the growth of the coffee
industry and the origins of industrialization has been one of the most popular
topics in Colombian economic historiography. The most widespread theory has
emphasized the role of small and medium-sized coffee producers in the west of
the country in modern industrial development. According to this view, the
widespread distribution of the benefits of expanded coffee cultivation to
the peasant economy in the west of the country laid the foundations for the
development of a domestic market which was to sustain the successful indus-
trialization efforts pursued in Colombia in the early decades of the twentieth
century (Nieto Arteta, 1975; Urrutia, 1979, ch. 3). Arango (1977) has countered
this theory by showing that traders' profit margins in areas of small-scale coffee
76 Coffee and Economic Development in Colombia
production areas were very high, a state of affairs not effectively corrected until
the beginning of the 1930s when the National Coffee Growers Federation
intervened. Hence, according to Arango, the contribution of the coffee expan-
sion was not so much a matter of generating a peasant market for industrial
goods as it was of accelerating capital formation. Coffee also contributed
directly to the development of manufacturing firms and a modern proletariat
in the cities through the spread of industrial hulking, which tended to be
concentrated in urban areas in the west of the country in contrast to the typical
pattern in the regions in which large plantations predominated.
These alternative interpretations of the relationship between coffee produc-
tion and manufacturing focus on certain key aspects of Colombian economic
development, but ignore other equally important elements. With a few excep-
tions ± textiles may have been one ± most consumption of industrial goods was
essentially urban from the start.11 Moreover, there is no clear evidence that
commercial capital associated with coffee participated significantly in the
founding of industrial enterprises in Antioquia (F. Botero, 1984). Lastly,
although employment in the coffee-hulking plants accounted for a large per-
centage of manufacturing jobs in the initial stages of industrialization, espe-
cially in the west of the country, the number of workers involved was too low to
be considered the source of the modern industrial proletariat.
Thus, coffee expansion boosted early attempts at industrialization not so
much through the creation of a rural market for industrial goods or capital
formation in activities directly related to marketing of coffee beans, but
through essentially macroeconomic linkages: capital formation in general, urba-
nization, the development of a modern road system and other factors asso-
ciated with this period of prosperity. Moreover, other factors having to do with
economic policy and various external phenomena contributed to the industrial
take-off.
Among the policy factors, the most important was undoubtedly the impetus
the Reyes Administration gave to early manufacturing. Protectionism, which
had been firmly rooted in the nation's economic thinking since the Regenera-
tion, found one of its most thorough expressions during the Reyes Administra-
tion (Ospina VaÂsquez, 1955; Ocampo and Montenegro, 1984). The tariff
reforms of 1905 and 1913 reinforced and in effect made explicit the protec-
tionist tendency that had been manifest since the 1880s; the reforms raised the
average tariff collected to the highest level yet seen in the country's history
(Figure 3.2). At the same time, the spreads between duties on goods for final
consumption and on inputs widened, and the wave of protectionism extended
to some processed agricultural products ± sugar and flour, among others. In
addition, the Reyes Administration fostered the infant industries ± both import
substitution and export ± through government purchases and loans to enter-
prises, direct subsidies and tariff exemptions on imported machinery.
Jose Antonio Ocampo and MarõÂa Mercedes Botero 77
Table 3.6 Share of net worth of industrial firms in 1945 by date of establishment (%)
1880 or From 1881 From 1901 From 1911 From 1921 From 1931 From 1941 Net worth
earlier to 1900 to 1910 to 1920 to 1930 to 1940 to 1945 in 1945
Food products 70.8 1.5 11.0 16.9 29.8 18.8 27.0 19.0
Beverages 0.0 87.3 9.9 10.8 12.7 2.5 3.0 18.1
Tobacco 0.0 0.0 1.2 14.4 1.1 0.4 1.3 3.3
Textiles 0.0 0.0 67.7 27.6 6.1 30.9 13.1 23.3
Clothing and leather goods 0.1 0.2 1.0 7.6 6.4 11.1 11.2 6.9
Wood and furniture 0.6 0.3 0.1 0.9 3.5 2.9 7.2 2.3
Paper 0.0 0.0 0.0 0.0 0.0 1.4 0.7 0.5
Printing 2.7 3.1 3.7 3.0 2.5 2.1 1.3 2.6
Chemicals 6.8 2.3 0.8 3.7 6.6 5.6 10.4 4.9
Rubber 0.0 0.0 0.0 0.0 0.2 2.2 1.4 0.8
Petroleum products 0.0 0.0 0.0 12.7 0.0 0.0 0.0 2.4
Non-metallic minerals 1.0 4.5 3.8 1.5 17.3 10.3 15.0 8.8
Metallurgy and metal working 17.9 0.8 1.0 0.8 13.6 10.7 7.0 6.9
Other 0.0 0.0 0.0 0.0 0.1 1.1 1.6 0.5
Departments
Antioquia 10.4 11.2 70.1 41.8 12.0 30.2 15.0 28.9
Atlantic 0.0 0.4 0.1 14.9 14.9 19.1 15.2 12.4
Cundinamarca 15.0 79.5 9.6 16.8 30.8 17.1 27.1 28.0
Valle del Cauca 68.3 1.4 2.7 4.8 20.4 18.1 17.3 13.0
Rest of country 6.3 7.5 17.5 21.7 21.9 15.4 25.5 17.7
Investment in each period as %
of net worth in 1945 1.7 13.3 10.4 18.7 18.8 28.6 8.4 100.0
the early decades of the century, especially in textiles, they still lagged far
behind for a country of Colombia's size and level of development. National
production, in fact, supplied only one-third of apparent demand for industrial
goods and less than one-fifth of national consumption of textile products (Chu,
1972; Ocampo and Montenegro, 1984). Moreover, in the period of prosperity
the country enjoyed in the 1920s, industrial production tended to lag behind as
a result of a crisis in the textile industry and other important branches of
manufacturing.
In contrast to some successful nineteenth-century firms founded by foreign
immigrants, the capital of the chief industrial enterprises in the early part of the
century was mainly provided by Colombian nationals. These investments satis-
fied the need of the country's richest families to diversify their fortunes, a
pattern with firm roots in the Colonial and nineteenth-century past. Even in
the case of the principal industrial investors, the percentage of family wealth
invested in industrial ventures appears not to have exceeded 40 per cent. In
many cases, an interest in starting up domestic production had its roots in the
particular investor's previous experience with marketing imported products, a
circumstance that may have helped to reduce the risk entailed by the new
investment. In keeping with a pattern firmly established in some regions of
the country, particularly Antioquia, in the nineteenth century, most of the large
enterprises were initiated as corporations even though they were closely family-
controlled. Moreover, a tendency towards concentration was evident early on
in some sectors ± tobacco, beer, matches, carbonated beverages. The acquisition
of other firms or the establishment of subsidiaries in other regions of the
country had generated a relative concentration in these industries even before
the crisis of the 1930s (EchavarrõÂa, 1992; Ospina VaÂsquez, 1955; Botero, 1984).
From the labour standpoint, the most important trend in manufacturing
during the early decades of the century was the rapid decline in handicraft
employment. Up to at least 1918, handicraft employment remained very
important, particularly among female workers (Table 3.7). Although we cannot
pinpoint the start of the decline, there was a sweeping change between 1918
and the middle of the century during which period the percentage of handicraft
workers in the total workforce dropped from 24 per cent to 7.5 per cent. The
decline particularly affected women's jobs, many of them located in rural
areas or smaller urban centres. The process was most noticeable in the regions
of the country that had had strong handicraft traditions but had been at the
periphery of modern industrial development (BoyacaÂ, Santander, Huila, Cauca
and Narin Äo).
As the handicraft tradition was dying out, manufacturing development was
generating demand for labour on a very limited scale, employing some 20 000
or fewer workers by 1918, that is, a fraction of those employed in the handicraft
sector.13 Up to the 1920s, labour law placed no great restraints on labour
80
Table 3.7 Employment in handicraft and manufacturing activities, 1870 and 1918a
M F M F M F M F M F M F
Antioquia 8 435 9 767 22 899 23 015 7.2 44.9 6.6 69.1 4.7 5.3 3.7 3.6
Boyaca 32 135 70 681 41 135 72 890 21.8 74.7 21.7 49.4 13.8 29.6 13.2 21.3
Cauca 15 138 42 110 28 744 115 688 11.3 39.5 9.2 63.0 7.2 18.8 6.3 24.6
Cundinamarca 12 345 17 080 20 562 53 782 10.9 49.4 9.2 43.8 6.3 8.0 5.3 12.6
Santander 12 781 66 922 20 481 53 564 10.3 88.3 10.1 79.5 6.3 30.2 6.2 15.4
Tolima 3 192 20 765 8 079 50 084 5.2 71.9 5.6 49.5 2.9 17.3 3.4 18.1
a
The departments shown correspond to those existing in the period 1863±86.
b
Artisans, manufacturers and artists in 1870; arts, crafts, manufacturing and industrial production in 1918.
Source: Population censuses.
Jose Antonio Ocampo and MarõÂa Mercedes Botero 81
Notes
equipment for what would become, decades later, Colombia's leading textile firm
(Coltejer) to MedellõÂn (in Antioquia).
6 This would be equivalent to growth in per capita GDP of 1.9 to 2.5 per cent, higher
than that estimated by Bulmer-Thomas (1994, appendix 3) for 1913±28 (1.4 per cent),
which seems too low. This estimate of per capita growth takes as a reference a
population growth of 2 per cent, which is consistent with Colombian population
censuses but lower than that used by Maddison and Hofman.
7 Some recent indirect estimates from the Banco de la RepuÂblica place GDP growth in
1905±29 at 4.9 to 6.7 per cent a year, but these calculations seem high, particularly
the latter.
8 For a more extensive analysis of the coffee economy in the early twentieth century,
see Arango (1977), Machado (1988), Ocampo (1989) and Palacios (1983).
9 The low monetary cost of labour was the result of the fact that peasants also grew
their own food on their land. The importance of monetary cost minimization as a
source of competitiveness in precapitalist economies has been emphasized by Kula
(1974), in a brilliant essay on Polish farms in the sixteenth and seventeenth centuries.
This source of competitiveness for Colombia's coffee industry has been pointed out
by many modern authors, but was mentioned by Mariano Ospina PeÂrez as head of the
National Federation of Coffee Growers as far back as 1934. See Ospina PeÂrez (1934).
10 Alongside different texts quoted in this section, the early history of Colombian
industrialization has also been analysed by Berry (1983), Brew (1977) and Poveda
(1970).
11 According to ECLAC estimates (ECLAC 1957), in 1953 when less than 40 per cent of
the population lived in cities, urban centres accounted for 64 per cent of industrial
consumption. Only in the case of textiles did rural consumption (61 per cent) exceed
urban consumption. Rural consumption of tobacco products was only slightly lower
than urban consumption.
12 This fact is corroborated in the 1916 Ministry of Finance Annual Report, which
presents a brief, incomplete inventory of the industrial firms at that time. According
to that source, less than 30 per cent of the capital invested in industrial activities at
that time was in Antioquen Ä o firms.
13 In reality, this figure could well be too high. The incomplete list drawn up in the
Ministry of Finance Annual Report of 1916 included 121 establishments employing a
total of 12 223 workers. Of those, however, 5000 were employed by the Sincerin sugar
refinery, most of whom must have been agricultural labourers. On the other hand,
the inventory did not provide information on employment in some large enterprises
(Bavaria, Manuelita and two of the biggest textile firms of Antioquia) or in some
categories of industrial activity, especially the more traditional industries (coffee
hulking, distilling, sugar mills and printing).
References
LoÂpez Toro, A. (1970) MigracioÂn y cambio social en Antioquia durante el siglo diez y nueve,
BogotaÂ: CEDE±Universidad de los Andes.
Machado, A. (1988) El cafeÂ: de la aparcerõÂa al capitalismo, 2nd edn, BogotaÂ: Tercer Mundo.
Maddison, A. (1995) Monitoring the World Economy, 1820±1992, Paris: OECD Development
Centre.
McGreevey, W. P. (1971) An Economic History of Colombia, 1845±1930, Cambridge: Cam-
bridge University Press.
Melo, J. O. (1987) `Las vicisitudes del modelo liberal (1850±1899)', in J.A. Ocampo (ed.),
Historia econoÂmica de Colombia, BogotaÂ: Siglo XXI±Fedesarrollo.
Nieto Arteta, L. E. (1975) El cafe en la sociedad colombiana, BogotaÂ: Ediciones Tiempo
Presente.
ÐÐ (1973), EconomõÂa y cultura en la historia de Colombia, revd edn 1996, BogotaÂ: Editora
Viento del Pueblo.
Ocampo, J. A. (1994) `RegõÂmenes monetarios variables en una economõÂa preindustrial:
Colombia, 1850±1933', in F. SaÂnchez (ed.), Ensayos de historia monetaria y bancaria de
Colombia, BogotaÂ: Tercer Mundo±Fedesarrollo±Asobancaria.
ÐÐ (1989) `Los orõÂgenes de la industria cafetera, 1830±1929', in Nueva Historia de Colom-
bia (vol. V, ch. 9), BogotaÂ: Planeta.
ÐÐ (1984) Colombia y la economõÂa mundial, 1830±1910, BogotaÂ: Siglo XXI±Fedesarrollo.
Ocampo, J. A. and Montenegro, S. (1984) Crisis mundial, proteccioÂn e industrializacioÂn:
ensayos de historia econoÂmica colombiana, BogotaÂ: CEREC.
Ordon Ä ez Plaja, A. and PolanõÂa, D. (1992) `Cambios de estatura en Colombia durante el
presente siglo', Coyuntura Social, no. 6, June.
Ospina PeÂrez, M. (1934) Informe del Gerente de la FederacioÂn al Sexto Congreso Nacional de
Cafeteros, BogotaÂ: FederacioÂn Nacional de Cafeteros.
Ospina VaÂsquez, L. (1955) Industria y proteccioÂn en Colombia, 1810±1930, BogotaÂ: Editorial
Santa Fe.
Palacios, M. (1983) El cafe en Colombia, 1850±1970: una historia econoÂmica, social y polõÂtica,
2nd edn, Mexico/BogotaÂ: El Colegio de MeÂxico/El Ancora Editores.
Parsons, J. J. (1968) Antioquen Ä o Colonization in Western Colombia, Berkeley and Los Angeles:
University of California Press.
Posada CarboÂ, E. (1996) The Colombian Caribbean: A Regional History 1870±1950, Oxford:
Clarendon Press.
ÐÐ (1987) Una invitacioÂn a la historia de Barranquilla, BogotaÂ: CEREC.
Poveda, G. (1970) `Historia de la industria en Colombia', Revista Trimestral, no. 11, ANDI.
Romero, C. A. (1994) `La banca privada en BogotaÂ, 1870±1922', in F. SaÂnchez (ed.), Ensayos
de historia monetaria y bancaria de Colombia, BogotaÂ: Tercer Mundo±Fedesarrollo±Asocia-
cioÂn Bancaria de Colombia.
Safford, F. (1965) Commerce and Enterprise in Central Colombia, 1821±1870, Doctoral thesis,
Columbia University.
Tovar, B. (1984) La intervencioÂn econoÂmica del Estado en Colombia, 1914±1936, BogotaÂ:
Banco Popular.
Urrutia, M. (1979) Cincuenta an Ä os de desarrollo econoÂmico colombiano, BogotaÂ: La Carreta.
ÐÐ (1969) Historia del sindicalismo en Colombia, BogotaÂ: Universidad de los Andes.
White, J. (1978) Historia de una ignominia: la United Fruit Co. en Colombia, BogotaÂ: Editorial
Presencia.
4
The Economies of Central America,
1860±19401
HeÂctor PeÂrez Brignoli*
Introduction
The size and basic characteristics of the countries of Central America make
them an ideal case study of the development of export economies. They have
relatively simple economic structures, which, in the majority of cases, have
been dominated for a considerable time by one or two export products. In other
words, the effects of internal diversification that were quick to appear in larger
economies have taken several decades to appear here. In particular let us note
that not so long ago, from 1960±75, the Central American food, drinks and
tobacco industries underwent a rate of growth ± measured in respect of the
added value of the whole industrial sector ± similar to that experienced by the
same industries in the Argentine economy from 1900±29 (PeÂrez Brignoli and
Baires MartõÂnez, 1983, p. 380).
This means that they are export economies in which the development of
import substitution was late and incomplete. For this reason they have been
consistently open to the world market since the middle of the nineteenth
century and broadly favourable to foreign investment with only a small inter-
ruption from 1960 to 1985.
Given that the region is dependent upon just a few export products, there is
also a wide discrepancy with regard to income distribution and social structure.
It would seem that the coffee economy, the very one that sustained long-lived
dictatorships such as that of Manuel Estrada Cabrera in Guatemala (1898±
1920), also helped in other instances, as was the case in Costa Rica, to establish
an equally long chain of democratic and stable regimes. In other words, in so
* The author acknowledges with thanks the comments of Victor Hugo Acun Äa, Enrique CaÂrdenas,
Jorge LeoÂn SaÂenz and Mario Samper. The ideas contributed by Rosemary Thorp, contained in the
convocation documents and summary of the seminar held in Panajachel (Guatemala) in December
1996, have been a constant source of inspiration and encouragement. Thanks to her and members of
the export economies team, work on this project has been as enjoyable as it has been demanding.
85
86 Central American Economies, 1860±1940
far as the relationships between economy, state and business are concerned, the
history of the region offers examples as varied and complex as those of any
Latin American country.
Preconditions
In the middle of the nineteenth century the integration of the Central Amer-
ican economies into the world market continued to pose problems (Lindo-
Fuentes, 1993, pp. 141±201). Products like cocoa, tobacco, indigo, cochineal,
sugar and cotton were produced for the local market; others, like balsam,
sarsaparilla and high-quality timber, seemed very promising with regard to
possible international demand. The main difficulty in establishing an effective
relationship with external markets was still exactly the same as in colonial
times: production costs were not competitive. This was due to very expensive
maritime transport, a primitive internal transport network, which was depend-
ent almost entirely upon mule transport, and the cost and quality of labour.
Moreover, the very size of the countries and producing regions limited the scale
of production and discouraged the interest of exporters, at least in the medium
term. The history of Central America has been characterized ± as officials at the
Guatemalan Consulate (Informe del Consulado, 1983, pp. 123±41) never tired of
repeating ± by an almost infinite succession of short cycles in which a promis-
ing product would appear, production and trade would flourish, and then it
would all disappear as quickly as it had begun. Against this background it is easy
to explain the success, though short-lived, of products with a high unit price as
regards volume, like indigo or cochineal, and of those cheap and easy to obtain,
such as timber from the coastal areas. This colonial model, prevalent since the
sixteenth century, was still common in the middle of the nineteenth century.
What were the factors exerting such a disastrous influence on Central Amer-
ica for the situation to continue for three hundred years? Above all, it was the
cost of transport by sea. In the sailing-ship age, not only was access to the
Pacific coast of Central America difficult but the region was also an extremely
long way from European markets. Besides, the isthmus was underpopulated,
with a number of isolated settlements in the central highlands, making com-
munication very difficult. The rugged, forested, mountainous terrain, which
imposed such a barrier to north±south circulation, also made it difficult to
reach the Caribbean coast. These geographical features, and the fact that
there were no precious metals on the isthmus that could be very profitably
exploited, condemned the region to a marginal and peripheral role within the
Spanish colonial empire. In fact, the Central American isthmus' only compet-
itive `value' was its strategic importance for crossing from one ocean to the
other (PeÂrez Brignoli, (1999)). However ± and in spite of the many attempts and
indubitable potential of other routes ± this strategic importance was only really
HeÂctor PeÂrez Brignoli 87
of value in Panama, and for one simple economic reason: the crossing there was
shorter than anywhere else.
It is therefore easier to understand why the countries of Central America,
which together formed the Audiencia de Guatemala in colonial times (1570±
1821) and went on to independence as a Federal Republic (1824±38), remained
marginal and peripheral for so long. At the same time, these enduring circum-
stances make possible an assessment of the revolution sparked off by the
sustained success of the export of coffee from the second half of the nineteenth
century.
10.0
millions of kg
0.1
0.0
1940
1840
1844
1848
1852
1856
1860
1864
1868
1872
1876
1880
1884
1888
1892
1896
1900
1904
1908
1912
1916
1920
1924
1928
1932
1936
(b) Nicaragua and Honduras
100
10
millions of kg
Nicaragua
Honduras
0.1
0.01
1840
1844
1848
1852
1856
1860
1864
1868
1872
1876
1880
1884
1888
1892
1896
1900
1904
1908
1912
1916
1920
1924
1928
1932
1936
1940
yields, pruning and weeding were as important as abundant and regular rainfall.
As the loss of fertility of the soil was to a certain extent compensated for by the
presence of leguminous trees used for shade and a moderate use of organic
fertiliser, it took several decades before the yield was affected. The decrease in
yield became apparent in the 1920s and continued until the 1950s. As cultiva-
tion techniques underwent little change it followed that there were few techno-
logical changes in the way coffee was cultivated. As this was the case, the only
effective way of dealing with decreasing yields would have been a system of
gradual crop rotation and a phased replacement of the oldest coffee plants, but
during this period international coffee prices never reached the high levels
recorded before 1897. Therefore the principal incentive to push for these innova-
tions and thereby achieve a new expansion of production was lacking. In the
1930s the Central American coffee industry had two obstacles to overcome:
there was no further supply of good soil and most coffee plants were too old.
Harvesting the coffee was extremely labour-intensive, with no possibility of
mechanization. Furthermore, only the system of `picking out' ± several pickings
selected according to the ripeness of the coffee bean ± guaranteed a product of
the highest quality. There were two methods ± commonly called beneficio ± of
processing the coffee bean: the dry method, which was suitable for small-scale
production, and the labour-saving wet method, which, however, required
costly installations and an abundant supply of water. In Costa Rica the wet
method predominated from the 1840s whilst in the other Central American
countries both methods were used. Once again harvesting and processing are
labour-intensive, as already examined when discussing cultivation techniques.
It would be safe to assert that the secret of the quality of Central American
coffee, and therefore its competitiveness in the world market, had much to do
with the quality and organization of the workforce.
The expansion of the coffee industry was closely associated with the triumph
of liberalism ± the so-called `liberal reforms' ± in the final quarter of the nine-
teenth century. The changes in the organization of the state were as much a
requirement for, as a consequence of, the expansion of the coffee plantations
themselves. In Costa Rica liberal reform took place gradually, from the 1830s to
the 1880s. In Guatemala and El Salvador this reform brought about a much
more rapid and violent change between 1871 and 1884, whilst in Honduras and
Nicaragua the reforms started later and were incomplete. In Honduras, despite
the government's attempt to achieve a systematic transformation, the develop-
ment of agro-exports was limited and did not produce a true entrepreneurial
class. In Nicaragua, the liberal regime of Jose Santos Zelaya (1893±1909) ended
with a long period of civil wars and US intervention (1910±34).
In Costa Rica (Cardoso, 1973; Hall, 1976) the expansion of the coffee industry
did not compete with other products but coexisted with the agricultural colon-
ization of the Central Valley for almost a century. The social structure created
90 Central American Economies, 1860±1940
much in the land market but rather in the labour force. In El Salvador, more
bound to the world market by indigo and more densely populated than any of
the other countries, the development of private property in the period prior to
the expansion of the coffee industry had been greater and had already created a
very unequal distribution of land, with the coining of the well-known binomial
latifundio-minifundio (large estate/smallholding) (Lindo-Fuentes, 1990, pp. 149±
51). Therefore, the expropriation of communal lands not only affected land
suitable for coffee but was responsible for the creation of a landless peasantry
which had no alternative but to work on the new coffee plantations. The coffee
industry in El Salvador underwent a more rapid development between 1880
and 1910 than anywhere else in the isthmus.
Coffee-growing in Nicaragua developed in two quite distinct areas (see
Radell, 1964, Dore and Gould in PeÂrez Brignoli and Samper, 1994). The first
was located in the uplands to the south of Managua, very near the Pacific coast.
Although this was a relatively small area, the low cost of transport had, with
government help, enabled the coffee industry to expand there since the 1860s.
In the second area ± which was far from the coast, in the mountainous region of
Matagalpa ± coffee production expanded at the end of the nineteenth century.
Because of the distance, the cost of transport was very high and although there
were plans to extend the railway system this never happened. With the expan-
sion of the coffee industry came a certain amount of privatization of communal
and common lands. It is important to note that in Nicaragua coffee never
achieved the dominant status that it did in Guatemala, El Salvador and Costa
Rica. Instead it coexisted with other important industries such as animal
husbandry and left room for the development of a sizeable racially-mixed
peasantry involved in the growing of subsistence crops. This was particularity
linked to resource availability. Because of the altitude and composition of the
soil the amount of land naturally suited to coffee-growing was relatively small
and confined to the two above-mentioned areas, which meant that the lack of
labour caused quite a problem. In order to ensure a supply of labour,
the landowners, supported by the government, employed a debt-bondage
system. Thus there was a contrast between the social profiles of both areas:
large estates ± which had already existed before the advent of coffee ± predom-
inated in the region around Managua, with a coerced labour force, whilst in the
mountains of Matagalpa family-based labour in medium and small enterprises
abounded.
In Honduras coffee was introduced systematically but not very successfully
during the 1870s and 1880s. This can be easily explained. In the rugged moun-
tains of the Honduran hinterland there was no large, continuous area with
volcanic soil such as there was in Guatemala, El Salvador, Costa Rica and in part
of Nicaragua ± only to the south of Managua. As a result, coffee could only
be produced profitably in certain valleys of the Honduran hinterland. The
92 Central American Economies, 1860±1940
isolation of these areas from one another and the high cost of transport
explains why coffee only expanded in Honduras after 1945, precisely during a
period of very high international prices when the routes of communication
were first being modernized.
With the coffee boom the transport system was totally revolutionized. In
particular it became essential to improve and extend the internal road network.
Until well into the twentieth century the ox-drawn cart was the favourite
means by which to transport the coffee beans from the farms to the processing
centres and from there to the railway station. Suddenly, the construction of the
railway to the ports became essential to reduce transport costs from the central
highlands to the coast. Because of the geographical location of the coffee-
growing plantations the Pacific ports were the most accessible, which meant
that the first railways in Guatemala, El Salvador and Nicaragua converged there.
However, in the absence of the Panama Canal ± completed in 1914 and open to
commercial traffic in 1919 ± the Pacific route did not seem the most economical
option. As soon as they were able, the governments of Costa Rica and
Guatemala built railways to the Caribbean coast, or, as they then preferred to
say, `to the Atlantic'. This work was long and difficult, and the cost was invari-
ably underestimated. However, it was finally completed successfully. Around
1910 both Costa Rica and Guatemala could rely on a rail network that passed
through the central highlands and provided an efficient route towards both
oceans. Years later the railways of El Salvador were still connected to those of
Guatemala.
Coffee entrepreneurs were a heterogeneous group. As well as traditional
families firmly rooted in the colonial era there were recently `risen' creoles
and mestizos, who invariably came from political and military backgrounds.
Moreover, from earliest times many foreign immigrants entered the coffee
sector as exporters. In Guatemala there was a majority of German entrepren-
eurs, whereas in El Salvador the immigrant community spanned many nation-
alities. In Costa Rica they were predominantly English and German, while in
Nicaragua there were fewer foreign immigrants than in the rest of Central
America. The profile of the coffee-growing elite was, in any case, unmistakable.
Their power was derived first and foremost from the export trade and from
their control over processing, as well as handling important credit arrange-
ments for the producers; in Guatemala and El Salvador this dominant position
was reinforced and extended by the existence of large estates. In El Salvador the
coffee-growing elite was not composed of only 14 families as commonly
thought, but comprised a little over 100. In the 1920s it is clear that in Guate-
mala, Costa Rica and Nicaragua there was an elite of 100±200 families, in each
country, at the top of the coffee-growing industry. Their interests were not of
course limited to coffee, but included other agricultural, commercial, financial
and industrial activities.
HeÂctor PeÂrez Brignoli 93
The banana export business (Kepner and Soothill, 1957; Posas, 1993; pp. 111±
65; and Ellis, 1983) was established towards the end of the nineteenth century
and soon became as important as coffee (see Figure 4.2). Geographically, ba-
nanas were grown in an `empty' area, namely some alluvial valleys along the
Caribbean coast.
The introduction of the cultivation and export of bananas in Costa Rica was
closely related to the problems of the construction of the railway, during the
many years that it took to link the central highlands of Costa Rica to the
Caribbean coast. In the 1880s Minor Keith, the North American entrepreneur
responsible for the construction of the railway, also became involved in the
cultivation and export of bananas and managed to obtain a full share of both
businesses. Soon afterwards he extended his plantations to Bocas del Toro in
Panama and Santa Marta in Colombia. The business worked closely with the
railways: plantations sprang up along the riverbanks and the trains could thus
quickly transport the fruit to the port. From there this perishable and relatively
fragile product was transported by steamships, which were soon equipped with
refrigeration, to the consumer markets of New Orleans and the east coast of the
USA.
However, the `banana-railway' diptych did not only refer to the technical
organization of production. Governments signed contracts with the railway
100 000
10 000
Thousands of bunches
1 000
of bananas
100
Costa Rica
10 Honduras
Guatemala
Nicaragua
1
1883
1885
1887
1889
1891
1893
1895
1897
1899
1901
1903
1905
1907
1909
1911
1913
1915
1917
1919
1921
1923
1925
1927
1929
1931
1933
1935
1937
1939
and banana companies granting them ownership of large tracts of virgin territ-
ory along the main railway lines and branchlines, and tax exemptions for the
import of rolling stock required by the trains. Thanks to these facilities, within a
few years there were several extraordinarily prosperous companies involved in
the production and export of bananas.
The expansion of the plantations on the Caribbean coast of Central America
began in Costa Rica and Panama in the 1880s, spreading into Honduras and
Nicaragua and reaching Guatemala a little later. The lushness of the new soil
facilitated a rapid expansion with high yields; however, by the beginning of the
twentieth century diseases which attacked the plantations had already
appeared. As agronomic research took so long to find solutions for this menace,
banana glut and famine occurred in close succession, affecting practically all
the areas suitable for banana-growing on the Caribbean coast. This never-
ending search for new territories, and the abandonment of the old, had already
given rise to the introduction of banana plantations on the Pacific coast of
Panama (Puerto Armuelles). The same phenomenon occurred, years later, in
Costa Rica and Guatemala.
It is possible to identify three distinct phases in the organization of the
banana industry. The first, from 1870±98, was characterized by the importance
of small producers and the appearance of several companies involved in trans-
porting and commercializing the product. Towards the end of this period
several commercializing companies merged and the market started to become
more focused. The second phase, 1899±1929, is defined by the domination of a
few companies headed by the United Fruit Company which managed to draw
together the vertical and horizontal strands of the banana industry. Just one
company controlled all the stages, from plantation to embarkation, using boats
with refrigeration chambers. After 1930 the number of banana companies
contracted even further and the onset of economic depression and banana
disease brought about a significant decline in the industry. It was not until
the 1950s that there were very important changes in this industry, such as the
introduction of disease-resistant species and the packing of the fruit in boxes
instead of exporting them in bunches.
Honduras was the only country where integration into the world market was
almost exclusively dependent on bananas. As already mentioned, the liberal
governments at the end of the nineteenth century attempted, unsuccessfully,
to extend the cultivation of coffee. Silver mining still enjoyed only limited
success, so nothing changed until the banana boom at the end of the nine-
teenth century. In the 1920s Honduras became the largest producer of bananas
in the world.
The presence of banana companies struck a novel note in the relationship
between the state and entrepreneurs; very serious situations of corruption and
confrontation were caused by granting new concessions or renewing those
HeÂctor PeÂrez Brignoli 95
already in existence. The worst example was when the conflicts between the
United Fruit Company and the Cuyamel were reflected in the Honduran civil
wars of the 1910s and early 1920s. Some writers have described these episodes,
marked by the occasional deployment of North American marines, as `banana
wars'. Viewed in perspective this description does not seem exaggerated. In any
case, it is clear that the Honduran civil wars ended once the Cuyamel and the
United Fruit Company merged into a single company in 1930.
In Costa Rica the situation was different (Casey, 1979; Carcanholo, 1978).
There was only one banana company (the United Fruit Company) operating in
the country, but with several national producers engaged in the industry.
Although they could only export through the United Fruit Company, their
political and social influence was considerable. Overall, the Costa Rican state,
which had been shaped and dominated by coffee-growing interests, was less
sensitive to the pressures of the banana company, and this is reflected in the
relatively early renegotiation of the contracts with the company. Moreover, in
the case of Costa Rica the United Fruit Company only managed to control the
railway between San Jose and LimoÂn: the San JoseÂ-Puntarenas line, completed
in 1910, always remained state property and was under its control. In Nicar-
agua, the North American banana companies ± in particular the Standard Fruit
Company ± combined such activities with timber exploitation and mining. But
continual civil wars (1909±34) had a very damaging effect, culminating in 1931
when the Sandinistas set fire to plantations and installations in the Puerto
Cabezas area. Moreover, diseases like the so-called `Panama disease' suddenly
took hold. In this context it is no wonder that the banana companies virtually
went out of business. In Guatemala (Dosal, 1993) the United Fruit Company
negotiated an agreement in 1904 and immediately, in addition to the cultiva-
tion and export of bananas, entered into the railway business, with almost
immediate results: by 1912 the International Railways of Central America
(IRCA), a subsidiary of the United Fruit Company, controlled the railways,
unloading bays and port installations in Guatemala.
As the banana plantations were in areas that had only recently been opened
up to agricultural exploitation, there were very few disputes over land. Labour
on the plantations generally came from Jamaica. Black workers had come to
build the railways and continued to arrive in large numbers during the con-
struction of the Panama Canal (for the French phase as well as the US). Once
this work had been completed, large contingents of black workers moved from
Panama to Honduras and Costa Rica.
The nature of social relations on the banana plantations is noteworthy. On
the one hand, there was a large wage-earning proletariat which soon developed
syndicalist grievances. On the other hand, the officials of the banana compan-
ies practised and spread a type of racism not very different from that which
existed in the USA. Immigration from Jamaica ceased around 1930, and after
96 Central American Economies, 1860±1940
that date many peasants of mixed race from the central highlands joined the
workforce on the banana plantations.
From a socioeconomic point of view, the main difference between the way in
which the coffee and banana industries were organized was in the introduction
of two or three foreign banana companies with a very high degree of vertical
and horizontal integration. Furthermore, the fact that the plantations were
situated on low-lying land previously virtually uninhabited made the banana
settlements seem disconnected from the rest of their respective countries. Thus,
many writers have described the banana-growing areas as `enclave' economies.
Although there is an undeniable element of truth in this image, it is also true
that there is a large degree of exaggeration. The banana settlements were not
static and the tremendous initial isolation with respect to the rest of the
country was undergoing rapid change.
From the point of view of the social structure the contrasts were also worthy
of note. In the banana plantations, two or three very large foreign companies
competed with a few national producers and with a large number of wage-
earning labourers. Everything was organized around the large foreign company
versus salaried workers relationship. Further, the fact that during the early
decades this workforce was largely black and of Jamaican origin introduced a
significant element of cultural diversity. The relationship between these black
workers and their mixed-race companions ± whose number was steadily
increasing ± was complex and forever swinging between solidarity and conflict.
In addition, companies threatened by strikes and movements for workers'
rights did not hesitate to exploit these ethnic and cultural differences to disem-
power the workers.
In the coffee communities the social structure was much more complex.
There were at least two distinct opposing relationships: (i) the elite made up
of exporters, beneficiadores, and big landowners versus a large number of me-
dium-sized producers; (ii) all the producers, opposed in their turn to the
permanent agricultural workers who had settled on the estates, and also to
the workers who helped with the harvest and processing. Coffee production
required a sizeable number of bosses and workers but left plenty of opportunity
besides for family work in all stages of production and processing. This also
contrasted with the banana plantations where there was room only for single
men and where middle-ranking workers were much fewer in number.
Something common to both activities/industries was the absence of forward
linkages within the national economies, and the requirement of a relatively
large stream of investments. The coffee plantations required constant reinvest-
ment in work both to keep the bushes producing ± in the 1930s many coffee
HeÂctor PeÂrez Brignoli 97
plantations in Central America were already more than 50 years old ± and to
replace them. In the banana plantations, on the contrary, they normally moved
to other areas once the yield dropped or there was evidence of disease. As was to
be expected, the strategy for dealing with both situations was quite different.
The coffee growers tried, by all means possible, to reduce the monetary cost of
permanent workers and so explored other systems of settlement and payment
in kind with rations of food produced on the estate itself. The banana compan-
ies, for their part, were anxious to obtain additional concessions from public
lands, always leaving a considerable amount of uncultivated land in reserve.
the 1950s. It is not possible to list the problems of each case here, although it is
worth pointing out some dates and processes.
The government of Costa Rica adopted the gold standard in 1896 (Soley
Gu È ell, 1926), just one year before the beginning of the long coffee crisis caused
by overproduction in Brazil and which continued until 1907. This created
difficulties which meant that the reform did not become effective until 1900.
It enjoyed remarkable success although the government, drained of funds, had
to hand over the Costa Rican railway to the Northern Railways Company (a
subsidiary of the United Fruit Company) in 1905.
The second country to adopt the gold standard was Nicaragua (Cumber-
land, 1928; Hill, 1933) in 1912, as a result of the fall of Zelaya and US
intervention. This reform followed the Dawson Accords (1910) and the
Knox±Castrillo treaty (1911). Nicaragua received an initial loan of US$ 1.5
million guaranteed by customs revenue, and proceeded to implement the
changes. The new currency had parity with the dollar, and the paper notes
issued during the governments of Zelaya, Madriz and Estrada started to be
withdrawn from circulation. During the years that followed, the system
was underpinned by new loans and similar guarantees. As in the case of
Costa Rica, the successful adoption of the gold standard implied an arrange-
ment for paying foreign debt. However, in the case of Nicaragua it implied
North American intervention in the collection of revenue and the presence of a
detachment of marines. In other words, the settlement of the debt and the
adoption of the gold standard took place within the framework of a virtual US
protectorate.
El Salvador (Fonseca, 1924) tried to adopt the gold standard in 1892 but the
plan failed. This reform was eventually successful in 1919 ± a rather ideal
moment because the War led to a rise in the price of silver. For a country that
was still on the silver standard it was the best time to change. There were
difficulties when the price of coffee fell in 1920 and 1921 and the subsequent
suspension of the debt interest payments. The situation was redeemed in 1922±
23 with the approval of a US$21.5 million foreign loan guaranteed by 70 per
cent of customs revenue.
Guatemala adopted the gold standard in 1923, immediately after a prolonged
period ± over 25 years ± of devaluation of the silver peso and with inconvertible
notes in wide circulation. Within three years the new currency had completely
replaced the old notes; the creation of the Central Bank of Guatemala in 1926,
with a 10±year guaranteed monopoly for the issuing of currency, completed the
framework of reforms.
Honduras was the last country to adopt the gold standard and to have, as
previously mentioned, a unified system of monetary circulation. There were
several attempts at monetary reform in the 1920s but the government was not
able to achieve this until 1931. The new monetary unit, called Lempira, began
HeÂctor PeÂrez Brignoli 99
Table 4.1 Years in which there was no payment of interest on external public debt
to circulate, although not without problems. Note that the IMF delegation that
visited Honduras in 1949 included the following in their report:
The adoption of the gold standard and the organization of a unified national
monetary system was, amongst other things, tied up with the settlement of
foreign debt. The countries of Central America began to get into serious debt in
the second half of the nineteenth century when their governments sought
loans from Britain in order to boost railway construction. Both the embezzle-
ment of funds and overly optimistic estimates of the final cost of the works to
be financed were responsible for the spiralling of debt: within a few years, all
the governments found it difficult to pay the interest and fell into arrears. The
record of arrears in repaying foreign public debt in the period 1827±1928 (Table
4.1) more than illustrates this point. Costa Rica shows the best management of
arrears, not paying its debt for 23 years, whilst Honduras at the other extreme
spent 93 years without paying. As all countries always negotiated the terms and
arrangements for payment, arranging in their turn new loans, the above-
mentioned table is a complementary indicator of the relative success of each
country in its relationship with the world market.
Table 4.2 State intervention in banking development: central bank and rural credit
`De facto' central banka Central banksb Rural and long-term creditc
Costa Rica 1914 International 1950 Central Bank of 1914 Rural boards
Bank of Costa Rica Costa Rica (International Bank)
1937 Banking reforms 1937 Section of Agricultural
and Industrial Credit
Guatemala 1926 Central Bank of 1946 Bank of Guatemala 1933 National long-term
Guatemala Credit
1933 Mutual Savings
El Salvador 1934 Central Reserve 1961 Central Reserve 1935 Long-term Credit
Bank Bank of El Salvador Bank of El Salvador
1940 Rural Savings Banks
Nicaragua 1941 National Bank of 1960 Central Bank of 1930 Long-term Credit
Nicaragua Nicaragua Bank of Nicaragua
Honduras None 1950 Central Bank of 1950 National
Honduras Development Bank
Notes: In 1924 the National Insurance Bank was created in Costa Rica as a state monopoly.
a
Mixed capital institutions controlled by the state.
b
Public institutions (autonomous and decentralized).
c
Mixed capital institutions controlled by the state.
Guatemala 6.8
Honduras 9.0
El Salvador 6.1
Nicaragua 7.1
Costa Rica 19.7
are apparent if available figures on banking credit per capita in 1950 are exam-
ined: 10 dollars in Guatemala, 7 in Honduras and 51 in Costa Rica.5
Table 4.4 El Salvador and Costa Rica: composition of imports, 1946 (%)
a
Included in intermediate goods.
Sources: AlbarracõÂn y Brignoli (1977); Bovedani (date n/a.), Costa Rican External Trade
Statistics (1907±1946).
from earliest times in countries like Argentina, Brazil and Colombia, does not
occur in Central America until after the period under discussion. In other
words, during the early part of the twentieth century there was little economic
diversification in Central America. During the 1930s depression the typical
response was the expansion of agriculture for the domestic market (Bulmer-
Thomas, 1987, pp. 79±82).
The role of imports, which were mainly composed of consumer goods, was
confined to completing the integration of a national market of goods and
services. Integration may seem a word little suited to describing, for example,
a market which as far as the workforce was concerned displayed a great deal of
diversity and had many imperfections ± systems of forced labour, remuneration
in kind ± and that as far as credit was concerned included relatively primitive
systems for paying advances. However, we should not overlook an even more
fundamental fact: the continuing export boom and the scale of production
meant that all other economic activities were subjected in some way to the
dynamics of the external sector. The distribution of imported consumer goods
reached and affected, in one way or another, all sectors of the population. This
was the first time this had occurred in the history of the isthmus.
price of coffee and bananas was generally stagnant; thus there was no incentive
to bring about technological change and increase efficiency ± it was enough to
maintain the capacity of the production installations and extend them, if at all
possible, thereby minimizing monetary costs. If reinvestment in the same busi-
ness was not very attractive, neither were there great incentives to invest in other
industries. The tiny size of markets, determined as much by the small population
as by their lack of purchasing power, was a definite obstacle in the rise of import
substitution industry. Under these conditions it is no wonder that a large part of
the income obtained by the exporters should be spent on extravagant consumer
goods or invested abroad. In the case of the banana companies, investment
abroad had always been an integral part of their investment strategies.
In this context, it is not surprising that state expenditure constituted the
main source of investment for future development. Indeed, in the final decades
of the nineteenth century, governments began to encourage the construction
of railways, ports and roads, with recourse to foreign loans guaranteed by the
future income promised by foreign trade. In spite of slipping up on more than
one occasion, it has to be said that this strategy, at the time, produced good
results. The way in which public expenditure is distributed can be examined in
Figures 4.3 and 4.4, in which such information is presented for Guatemala and
Costa Rica.
100%
Debt
War
Public sector
Development
Education
80% Administration
60%
40%
20%
0%
1911
1883
1886
1890
1893
1895
1898
1900
1902
1904
1906
1908
1913
1915
1920
1922
1924
1926
1928
1930
1932
1934
100%
80%
60%
40%
Debt
War
20% Health
Public sector
Development
Education
Administration
0%
1870
1873
1876
1879
1884
1887
1890
1893
1896
1899
1902
1905
1908
1911
1914
1917
1920
1923
1926
1929
1932
1935
1938
Figure 4.4 Costa Rican government expenditure, 1870±1940 (classified by destination)
Note: Information is missing for the years 1880±1. These years are ommitted from the horizontal axis.
Sources: RomaÂn Trigo (1995).
These two cases were selected because they are representative of the contrast-
ing trends which can be observed in the region. In Guatemala ± and the
Guatemalan model is repeated in El Salvador, Honduras and Nicaragua ± the
burden of military spending, debt-service and administration costs conspired
continually against increasing the percentage spent on public works and educa-
tion. In Costa Rica, where to start with there is a very similar spending structure
to the one mentioned above, it is clear how, on the contrary, by about 1920 the
trend is reversed and expenditure on public works and education gain in
importance over those on defence and administration. As we shall see below,
this is what finally and in no uncertain terms differentiated Costa Rica from the
rest of Central America. Why this should be so can be explained: in economies
such as those of Central America the only element in the early part of this
century that could be effectively left aside for the future, apart from those
investments in infrastructure already in existence, was investment in public
works and human capital.
In Central America the social foundations of the liberal state came from coffee-
growing. In its political history this corresponds to the so-called `liberal
HeÂctor PeÂrez Brignoli 105
reforms'. In Guatemala and El Salvador the process was very rapid and adopted
the style of a `revolution from above'. The new liberal rulers, under the
notorious aegis of the Guatemalan Justo Rufino Barrios, seized power by
force and in a few years developed a full programme of coffee-export promo-
tion, changes in agrarian ownership and construction of railways and ports.
The result was striking. Coffee exports increased rapidly and a powerful
business elite came into being straight away, joined by some foreign nationals.
National wealth was thus identified with coffee and the social structure
provided the vehicle by which that image could also be shared by the lower
classes through a complex combination of paternalism and violence. The
`from above' character of this revolution needs stressing. In practice things
were not very different, in political terms, from the previous conservative
period. Patronage and paternalism went hand in hand in the exercise of
power, which had very little to do with the impeccable framework of liberal
laws and regulations.
With less purity, but with similar results, liberal reform also took place in
Honduras and Nicaragua. In the case of the former, the programme of eco-
nomic reform failed and only more than twenty years later, with the expansion
of the banana plantations, did integration into the world market bear the
expected fruits. In the Nicaraguan case, the fragmentation of elite interests
put a stop to liberal reform, whilst the civil war and US intervention created a
long parenthesis of twenty years (1909±34). At the end of this period it could be
said that Anastasio Somoza completed the work begun by Zelaya. From the
point of view of the exercise of power, the situation in Honduras and Nicaragua
was very similar to that in Guatemala and El Salvador.
If in the rest of Central America it is possible to talk of a liberal revolution `from
above', in Costa Rica, on the other hand, the process of political transformation
adopts the form of a revolution `from below'. This means that liberal reform
took place gradually over almost a century (1821±1905), in a process of con-
tinual interaction between the coffee-growing entrepreneurs, civil society and
military leaders. The extension of the channels of political participation was
possible and even natural in a social context in which the middle-sized and
small agricultural producers enjoyed a strongly established position.
The later development of the banana export business had special implica-
tions of its own. In Costa Rica and Guatemala banana interests developed in a
political arena already marked out by coffee-growing interests. In Honduras on
the other hand ± owing to the failure of liberal reform ± the latter appeared and
expanded in opposition to an outdated landowning oligarchy whose interests
had difficulty transcending the local and regional scale, with a familiar out-
come: a very weak state confronting the powerful ambitions of foreign banana
companies, characterized by a political economy frequently summed up by the
derogatory, caricaturist term `Banana Republic'.
106 Central American Economies, 1860±1940
Long-term development
Now that we have described the context and its most striking structural char-
acteristics, we can review the long-term evolution of the economies of Central
America and try to obtain some synthetic and comparable measures of its
successes and failures in the long term. In order to analyse the temporal trend
we shall closely examine the data gathered on the volume and value of the
exports.
The physical quantities of coffee exported (Figure 4.1) increased in value at a
constant average rate of between 0.7 per cent per annum in the case of Costa
Rica and 1.7 per cent in the case of El Salvador.
In the period under discussion, Central America exported all the coffee it
could produce and never stockpiled. The export curve therefore reflects the
internal capacity for growth ± that is, the rates of incorporation of land and
labour in the coffee-growing sector. The differences express, at the same time,
important national contrasts. The slow growth of Costa Rica and Nicaragua can
be easily explained in terms of the market of factors of production. In the first
case, labour shortage is to blame: because of the predominance of small and
medium producers, the rate of incorporation of lands suitable for coffee was very
slow. In this sense it is worth recording that in the main coffee-growing area of
Costa Rica, the Central Valley, with a total area of approximately 2700 sq km, the
ecological limit for this crop was reached in about 1930 ± that is, a century after
the coffee-growing industry started to expand. In Nicaragua, both the availabil-
ity of land (the most extensive areas lacked cheap transport) and the availability
of labour (there was strong competition from other agricultural and cattle-
raising activities) were to blame. The export curves of El Salvador and Guatemala
show much stronger rates of increase than those of Nicaragua and Costa Rica. In
both cases this can also be explained by the behaviour of the market of factors of
production: coercive methods ± Guatemala ± or the relative abundance of land-
less workers ± El Salvador ± assured the labour supply necessary for a much more
accelerated incorporation of land suitable for coffee-growing.
In purely economic terms it seems clear that the Salvadorean entrepreneurs
were much more successful in achieving a higher rate of growth than any of
their Central American counterparts. However, the brilliance of these achieve-
ments fades into insignificance in so far as extraregional comparisons are con-
cerned. Colombia ± a coffee-growing country wholly similar as regards the type
of product, the techniques of cultivation and the characteristics of the natural
environment ± saw its coffee exports grow (in volume) at a rate of 9.7 per cent
per annum between 1887 and 1899, and of 2.7 per cent between 1906 and 1940
(Samper, 1948, pp. 85±9). A systematic comparison of the factors that explain
this more than apparent Colombian `superiority' could throw new light on the
profound nature of the Central American coffee-growing economies.
HeÂctor Perez
 Brignoli 107
35.0 Guatemala
EL Salvador
30.0 Costa Rica
25.0
20.0
15.0
10.0
5.0
0.0
1864
1867
1870
1873
1876
1879
1882
1885
1888
1891
1894
1897
1900
1903
1906
1909
1912
1915
1918
1921
1924
1927
1930
1933
1936
1939
(b) Nicaragua and Honduras
30.0
Nicaragua
25.0
Honduras
20.0
15.0
10.0
5.0
0.0
1864
1867
1870
1873
1876
1879
1882
1885
1888
1891
1894
1897
1900
1903
1906
1909
1912
1915
1918
1921
1924
1927
1930
1933
1936
1939
Table 4.5 Rates of growth of physical volume of coffee exports and variability of annual
fluctuations
ra vb
a
Average annual rates of growth, in percentage terms. Calculated from the data in Figure 4.1 using an
exponential function adjustment All rates are statistically significant at a 95 per cent level.
b
Standard deviation of the first differences ± the absolute change from one year to the next. As the
average of the series is approximately zero, comparison of deviations does not present a problem and
can be regarded as an approximation of the annual fluctuations of each series.
Source: Data from Figure 4.5.
Let us now consider the annual fluctuations, using the same coffee export
curves. The variability of the quantities exported from one year to the next was
noteworthy but changed little in the long run. In other words, if the trend of
the series is eliminated its variance remains stable. Under these conditions the
standard deviation of the first differences (Table 4.5, 3rd column) clearly
describes the annual variability of each one of the series. Within this model,
exports from El Salvador and Guatemala show greater fluctuations than Costa
Rica and Nicaragua. In all cases the annual fluctuations can be explained by
pointing to two factors: climatic change (especially rainfall) and random varia-
tions; the limited data prevents, however ± at least for now ± the elaboration of
a strict econometric model to prove this hypothesis.
Examination of the physical volume of banana exports (Figure 4.2) is equally
illustrative as far as the dynamics of this sector are concerned. Costa Rica first
and then Honduras stand out as large producers compared with the much more
modest contributions of Guatemala and Nicaragua. Apart from the export of
coffee, banana exports experienced a rapid cycle of growth followed by another,
no less rapid, fall; annual fluctuations, for their part, are less noteworthy than
in the case of coffee. There is a simple explanation for these cycles as they can
be attributed entirely to diseases affecting the banana plantations ± the so-
called `Panama disease' and the sigatoka. In fact, the appearance of these
diseases caused a rapid drop in production which could only be halted, given
the nature of such pests, with the incorporation of new land. Given the parti-
cular concentration of banana-growing in the hands of a few companies, this
occurred as often within each country as in neighbouring countries. The
national framework is revealed as insufficient when these, as well as with
other conditions, are studied in the banana sector.
We now move on to consider the value of exports (Figure 4.5). It is import-
ant to explain first of all the origin and nature of the information used.
HeÂctor PeÂrez Brignoli 109
Government data regarding exports and imports have existed since the middle
of the nineteenth century; there are often gaps in these sets of data and they
cannot be considered very reliable until the end of the last century when the
customs register was modernized and statistics improved. However, as we do
not have other sets of data at our disposal we have to make do with the picture ±
at times diffused and blurred ± conveyed to us by these data. The data regarding
value, expressed originally in national currency, were first converted into US
dollars, and then into US dollars of 1926. Market exchange rates were used for
the conversion and ± for want of a better deflationary measure ± the wholesale
price index of the US Bureau of Labor Statistics with 1926 100. This option is
reasonable given that a large part of Central American imports came from the
USA.
Exports from Guatemala, El Salvador, Costa Rica and Nicaragua (Figure 4.5),
based above all on coffee, show a similar sustained upward trend from 1864 to
the end of the nineteenth century; stagnation and contraction until 1921; and
a short rapid rise until 1930, followed by an equally sharp fall from then until
1940. If even longer trends are sought, two periods can be identified: one of
rapid rise until 1897 in which the average value of exports grew by at least a
million US dollars per annum to approximately US$ 12 million, and a second ±
until 1940 ± of relative stagnation. During this last period the average value of
the series remains static but broad movements of rise and fall can be observed
around it.
Annual fluctuations, measured in percentage of change from one year to the
next, are illustrated in Figure 4.6. These variations were relatively strong and
had immediate repercussions on the amount of imports. Although these con-
tractions did not put the agro-export economy in danger owing to the dampen-
ing effect of the non-monetized sectors, they did endanger fiscal income and
public investment. As indicated above, state resources depended strictly on
customs revenue (60 per cent or more).
The performance of international coffee prices (Figure 4.7) may explain
almost point by point the trends just analysed. This was to be expected in
economies where the relative weight of coffee exports was overwhelming and
helps us to better understand a famous phrase by Ricardo JimeÂnez, who
affirmed not without humour that there was no better Chancellor of the
Exchequer than a good coffee harvest. As Ricardo JimeÂnez was President of
Costa Rica three times one has to concede that he knew very well what he
was talking about: the country was a price-taker in the international market so
that the value of exports ± and hence of fiscal revenue ± basically depended on
the size of the coffee harvest.
It is important to underline the following aspect of the evolution of the
international price of coffee in real terms (Figure 4.7): in 1897 a large upward
cycle came to an end, interrupted only briefly by the low prices of 1881±85. In
110 Central American Economies, 1860±1940
(a) Guatemala
160
Change %
140
Moving average (5)
120
100
80
60
%
40
20
–20
–40
–60
1891
1921
1864
1867
1870
1873
1876
1879
1882
1885
1888
1894
1897
1900
1903
1906
1909
1912
1915
1918
1924
1927
1930
1933
1936
1939
(b) EI Salvador
160
140
Change %
120 Moving average (5)
100
80
60
%
40
20
–20
–40
–60
1870
1873
1876
1879
1882
1885
1888
1891
1894
1897
1900
1903
1906
1909
1912
1915
1918
1921
1924
1927
1930
1933
1936
1939
1867
1864
Figure 4.6 Annual fluctuations of exports (US$, 1926) (a) Guatemala (b) El Salvador (c)
Costa Rica (d) Nicaragua
Sources: Series as in Figure 4.5.
%
%
0
20
40
60
80
–60
–40
–20
100
120
140
160
40
20
60
80
100
0
120
140
–40
–20
160
–60
1864 1863
1867 1866
1870 1869
Change %
1882
1885 1884
1888 1887
(d) Nicaragua
(c) Costa Rica
1909 1908
1912 1911
1915 1914
1918 1917
1921 1920
1924 1923
Â
1926
Change %
HeÂctor Perez
1927
1930 1929
1933 1932
Moving average (5)
1936 1935
1939 1938
Brignoli 111
112 Central American Economies, 1860±1940
35.0
30.0
25.0
(US cents per pound)
20.0
15.0
10.0
5.0
0.0
1863
1869
1872
1875
1878
1881
1884
1887
1890
1893
1896
1899
1902
1905
1908
1911
1914
1917
1920
1923
1926
1929
1932
1935
1938
1866
the first forty years of the twentieth century coffee prices never again reached
those of that golden epoch.
Let us now consider the case of Honduras as shown in Figure 4.5(b). In the
first place, we do not have many data before 1901; secondly, banana prices
correspond not to market prices but to a fixed value, estimated by the govern-
ment based on that stated by the banana companies; and, thirdly, there were
exports that were not recorded by the Honduran customs. Although this last
factor tends to disappear in the 1920s, the second lasts until 1950. Only after
the creation of the Honduran Central Bank do we have at our disposal a
valuation of the banana exports adjusted in accordance with the FOB quotation
of that fruit. Thus, for example, during the period 1931±50 the declared valua-
tion remained fixed at US$0.50 whilst in 1947±48 the FOB price per bunch was
calculated at US$2.05 (Central Bank of Honduras, 1957, p. 34). In the rest of the
banana-exporting countries there was a similar situation. In Costa Rica the
valuation declared in the period 1895±1920 fluctuated between US$0.4 and
US$0.5 per bunch, rising to US$0.75 during the 1920s. When in 1950 the
Central Bank decided to calculate the real value (FOB) of the banana exports
the same thing happened as in Honduras: the value declared by the banana
companies was multiplied by three (Central Bank of Costa Rica, 1951, p. 88).
This being so, it is obvious that before 1950 we lack an adequate market
valuation of such exports.
HeÂctor PeÂrez Brignoli 113
Are these sets of data of no use? In fact that is not the case as at least they
reveal the physical volume of the exports and the resources seized by the
government through the tax per bunch exported (one US cent per bunch in
the case of Honduras). On the other hand, whilst the banana companies may
not have been obliged to bring into the country the foreign exchange resources
generated by the exports ± and this only happened after 1950 ± we can assume
that the series of undervalued exports reflect the resources generated by this
sector that were available within the country. In other words, if we had been
able to obtain a set of real values before 1950, what this set would reveal would
probably be company profits that, in general, never reentered the country.
With these observations in mind we can return to Figure 4.5(b), which even
with these problems is simple to analyse: the curve of the value of Honduran
exports shows a sharp rise in the 1920s followed by an equally dramatic fall in
the following decade.
The well-known phrase of Ricardo JimeÂnez quoted earlier is now in its true
context: for state resources and the general well-being of the population ba-
nanas count for much less than coffee. Moreover, it must be remembered that
in Costa Rica the total value of exports was shared almost equally by coffee and
bananas. In Honduras, where the state basically depended on the banana
sector, the situation was much more extreme. An International Monetary
Fund mission that studied the Honduran economy in 1950 was in a position
to assert, amongst other things, that
considerably. The product per capita is shown to match the other indicators per
capita. Costa Rica shows better figures than those of the rest of Central America
as regards the value of exports and imports as well as in kilometres of railways in
operation or government expenditure. The socio-demographic indicators also
reveal similar discrepancies: in Costa Rica it is typical to find less illiteracy,
a
In years, own estimates using birth and death registers, censuses and `inverse projection' method.
b
Per thousand births, Idem.
c
In thousands, data from Collver (1965), and own estimates.
d
Average annual growth in percentages, data from Collver, op. cit., and Statistical Abstract of Latin America, vols
20 (1980), 24 (1985) and 26 (1986), and own estimates.
e
Population living in towns of more that 5000 inhabitants, census data.
f
As a percentage of the total population, census data
g
As a percentage of the total population, census data and own estimates.
h
Average 1924±26 in 1970 dollars, data from Bulmer-Thomas (1987), statistical appendix. To calculate the per
capita indices we have used population estimates which appear in (c) and not those of Bulmer-Thomas.
i
As a percentage of gross domestic product, average 1924±26, data from Bulmer-Thomas, op. cit.
j
In 1970 dollars, Bulmer-Thomas, op. cit.
k
Railway kilometres in 1921±22: data from Long (1925).
l
In millions of dollars, 1928, data from Winkler, quoted in Jones, op. cit. pp. 293±5.
m
In millions of dollars, data from Jones, op. cit.
n
Current dollars.
o
Current dollars.
HeÂctor PeÂrez Brignoli 115
greater life expectancy and a higher increase in population. In 1925 the per-
centage of illiteracy in Costa Rica (24 per cent) was similar to that of Chile and
Italy, and substantially less than that of Spain (43 per cent) and Mexico (59 per
cent) (Bunge, 1940, pp. 418±19).
Note that if we aggregate the indicators corresponding to Panama its values
would be more similar to those of Costa Rica than to those of the rest of Central
America. At least from 1920 it is clear that the isthmus is characterized by a
notable north±south divide. To the north of the San Juan river the demographic
and economic indicators reveal less wealth per capita, a large rural population
and high mortality rates. To the south of the San Juan river, in Costa Rica and
Panama, the situation is different. These countries can be seen to be in a much
more favourable situation given their position in the majority of the given
indicators, which is above the Latin American averages although below coun-
tries like Argentina, Chile, Uruguay and Cuba (Bulmer-Thomas, 1994, p. 444).
The data on Honduras merit a special mention. In the 1920s some indicators
show higher values than in subsequent years. I refer to the product and exports
per capita, and to the kilometres of railways per thousand inhabitants. During
the 1930s the Honduran indicators again resemble those of Guatemala owing
to the deep recession caused by the fall in banana exports as shown in Figure
4.5(b). In this sense the Honduran economic boom of the 1920s has all the air
of an exceptional opportunity, with no match before or since.
To sum up, the data in Table 4.6 reveal appreciable discrepancies in the nature
and consequences of economic growth, and the situation did not change much
in the following two decades. The levels of the internal product per capita
varied little. In Honduras, Guatemala, El Salvador and Nicaragua more than
60 per cent of the population were still illiterate. From 1910, when the laying of
the main railway lines was virtually complete, until the time they began to
construct the inter-American highway (1945), the road and rail networks and
transport experienced very little change.
The persistence of profound social inequalities was rather more serious and
much more difficult to change. The boom in agricultural exports expanded and
in a certain way `froze' the ethnic divisions of Guatemala at the same time as it
created new ethnic barriers in the banana areas of Honduras and Costa Rica. In
El Salvador the ethnic conflicts and the polarization of classes led to the great
peasant rebellion of 1932 with its tragic consequence of death and repression
(PeÂrez Brignoli, 1995, pp. 232±61). Political instability and the weakness of the
state in the face of foreign interests was another inheritance of the agro-export
model, present particularly in Honduras and Nicaragua. Behind the caricature
of `Banana Republics' hid perhaps a much deeper reality to which the region
was condemned by geopolitics and the reduced size of its countries.
In brief, it may be affirmed that the development of agricultural exports inte-
grated the Central American economies into the world market, incorporating a
116 Central American Economies, 1860±1940
Notes
1 On the Central American economies in general, see Cardoso, Ciro and PeÂrez Brignoli
(1977), Torres Rivas (1971) and Munro (1918), which is still indispensable.
2 Coffee was introduced to Costa Rica from the Caribbean where the earliest techniques
of cultivation and processing also had their origin. However, once commercial cultiva-
tion began there was a marked process of technological adaptation and invention
which encompassed the phases of cultivation, harvesting and processing.
3 In Central America the term refers to the owners of coffee-processing plants.
4 One example among many: an English commercial agent who visited El Salvador
around 1910 found annual interest rates fluctuating between 18 and 40 per cent. See
MartõÂn (1985).
5 Guerra Borges (1994, p. 47); for Honduras the author's calculations are based on the
IMF report, p. 54 in Central Bank of Honduras (1957).
References
AlbarracõÂn, P. and PeÂrez Brignoli, H. (1977) EstadõÂsticas del comercio exterior de Costa Rica
(1907±1946), San JoseÂ: Avances de InvestigacioÂn, Universidad de Costa Rica.
Adler, J., Schlesinger, E.R. and Olson, E.C. (1952) Las finanzas puÂblicas y el desarrollo
econoÂmico de Guatemala, Mexico: Fondo de Cultura EconoÂmica.
ANACAFE (1975) BoletõÂn EstadõÂstico, Guatemala.
Belli, P. (1975) `ProlegoÂmenos para una historia econoÂmica de Nicaragua, 1905±1966',
Revista del Pensamiento Centroamericano, vol. 146 pp. 2±30.
Bingham, J.L. (1974) Guatemalan Agriculture during the Administration of President Manuel
Estrada Cabrera, 1898±1920, masters thesis, Tulane University.
Bovedani, B. (no date) El Salvador's Foreign Trade and Development Problems, San Salvador:
Ministerio de EconomõÂa (no date available).
Browning, D. (1971) El Salvador. Landscape and Society, Oxford: Clarendon Press.
Bulmer-Thomas, V. (1987) The Political Economy of Central America since 1920, Cambridge:
Cambridge University Press.
ÐÐ (1994) The Economic History of Latin America since Independence, Cambridge: Cam-
bridge University Press.
Bunge, A. (1940) Una Nueva Argentina, Buenos Aires: Editorial Kraft.
Cambranes, J.C. (1985) Cafe y campesinos en Guatemala, 1853±1897, Guatemala: Editorial
Universitaria de Guatemala.
HeÂctor PeÂrez Brignoli 117
All the chapters in this book concern individual countries;1 this chapter con-
cerns Mexico. Using countries as units of analysis, however, is neither inevit-
able nor unproblematic, especially for the period c. 1900±30. Latin American
nation-states of that period were usually weak states,2 ruling highly imperfect
nations. Though states were tending to get stronger ± hence the emergence of
Wolf and Hansen's order and progress dictatorships (Wolf and Hansen, 1967,
pp. 168±79) ± their political and (a fortiori) economic powers were quite limited;
they could repress, but not represent; their burgeoning military capacity was
not matched by an equivalent capacity to mould their societies or control their
economies.3 This partly reflected the familiar condition of dependency, which
early twentieth-century governments could do little to mitigate.
But, more important, it reflected the fragmentation of political economies.
Porfirian Mexico was ostensibly more stable and integrated than many Latin
American states, yet its politico-economic cohesion ± hence, its validity as an
analytical unit ± were both limited (Reynolds, 1970, pp. 6±7). Porfirian Mexico
lacked a central bank;4 in the far south, Guatemalan pesos circulated; in some
villages markets were based on barter rather than money transactions.5 The
YucataÂn peninsula was cut off from the rest of Mexico, logistically, economic-
ally and psychologically: there was no railway connection, trade was skewed
towards the USA and (to a lesser extent) the Caribbean, and when a thirsty
customer ordered a `foreign' beer in MeÂrida he got a glass of Dos Equis ± from
Orizaba (Gamio, 1982, p. 12). Northern border ports ± still small, but fast
growing ± were already closely integrated with the US economy (hence, during
the Revolution, wild talk of northern secession flourished); northern company
towns ± such as Cananea ± enjoyed a large measure of both economic and
political autonomy (Sariego, 1988). Government spending in 1910 stood at
4.4 per cent of GDP (table 5.1 below); while the Revolution pushed up this
119
120 Export-led Growth in Mexico, c.1900±30
percentage (1925: 5.9 per cent), the real `rise of the active state' came post-
1930.6 Revenue was also a reflex of trade, especially foreign trade, above all
imports, hence fiscal dependency mirrored trade dependency.7 Although rail-
ways had dramatically affected the economy, helping to forge a more integrated
market (see below), this process chiefly favoured foreign trade (Coatsworth,
1984, ch. 5); many parts of Mexico ± towns bypassed by the railway as well as
more remote rural districts ± remained economically isolated, awaiting the
road-and-truck revolution of the 1930s and 1940s.8 Of 246 villages in 10 states
surveyed in 1922±24, only 5 per cent could be reached by rail, 2 per cent by car,
and 1 per cent by water transport; ox-carts served 27 per cent, while the
remainder (64 per cent) depended on horses, burros or foot transport (Tannen-
baum, 1968, pp. 85±7).
Throughout Mexico, but particularly in the centre, the south, and the sierras,
the ancient tlameme could still be seen footslogging along primitive roads, back
bowed, the tumpline biting into a calloused forehead.9 About 20 per cent of the
Mexican population were reckoned to be literate;10 while this emphatically did
not mean that 80 per cent of Mexicans were ignorant `parochials', scarcely
aware of the existence of the patria,11 it did mean that networks of economic
exchange were patchy; that Mexico was still ± to borrow a Durkheimian meta-
phor ± in a phase of `mechanical' economic solidarity, which perhaps embraced
two-thirds of the population.12 Contrary to crude Marxist analyses, it would
seem, some sort of loose national consciousness preceded the formation of an
integrated national market and a dynamic national bourgeoisie. Certainly the
domestic market was thin and fragmented; in some remote communities, we
have noted, barter prevailed; while even in northern Mexican communities ±
major cities apart ± furniture was scant, spoons were rare, and glass windows a
luxury (McKellar, 1994, pp. 51, 66). Merchants lamented the `damned want-
lessness' of the Mexican people, their chronic failure to consume.13 These were
old and familiar complaints,14 which were paralleled by equally old and famil-
iar complaints about Mexico's feckless working class, its chronic ± even con-
genital ± failure to respect the time and work discipline required by modern
capitalism.15 Consumer and labour markets alike, we may conclude, were
highly imperfect; much of the spadework of state-building, infrastructural
investment and/or `primary accumulation' had yet to be done (Marx, 1957,
vol. 2, ch. 24).
We have to be careful, then, when we generalize about `Mexico'. National
statistics are not only inherently unreliable; they also offer a very partial picture
of the country, since they can better capture ± for example ± foreign trade rather
than subsistence agricultural production; and, of course, they can tell us little
about subjective notions such as nationalism, business confidence, class antag-
onism or political legitimacy ± all of which the `new institutional economics'
would wish us to take into account (see Appendix). The notion of simple
Alan Knight 121
Thus, apart from its well-known social reforms ± which were often slow in
coming ± the Revolution brought significant socioeconomic change, albeit
change that is hard to calibrate, given not only the lack of reliable statistics
but also the `impressionistic' nature of many of these changes. We cannot
measure, we can only suggest, an enhanced entrepreneurialism and work
ethic; a decline of deference; class polarization; greater social and spatial mobil-
ity; and perhaps greater regime legitimacy.26 What can be said with some
confidence is that the Revolution was a genuine revolution which had import-
ant economic consequences; but that these consequences affected the domestic
political economy much more than they affected Mexico's external economic
relations ± even though those relations had helped provoke the Revolution in
the first place. `Dependency' (to use a loose catch-all) endured, perhaps it was
even enhanced; but the threads of dependency now wove their way through
both a transformed state (not, as yet, a `Leviathan') and a transformed civil
society. On that level, 1910 marked a decisive break which had no parallel in
the rest of Latin America. Hence the peculiarity of Mexican periodization which
must embody two distinct ruptures: the rupture in the domestic political econ-
omy in 1910, and the rupture in external relations in 1930. The former is
unique, the latter bears comparison with the rest of Latin America. But the
fact of the first rupture helped determine the Mexican reaction to the second.
Thus, as I have argued elsewhere, Cardenismo was the peculiar result of the
intersection of a distinctive domestic (political economy) trajectory and a
shared external (world depression) conjuncture (Knight, 1991a, ch. 5, especially
pp. 245±6).
If, as I have just argued, the domestic political economy of the Porfiriato
entered into crisis in 1910, and its associated external economic relations in
1930, thus giving us two clear terminal points, when did the process of export-
led growth begin? It is obvious that 1900 is, for these purposes, a meaningless
date, another case of irrational hectohistory (Fischer, 1970, p. 145). To under-
stand Mexico's phase of desarrollo hacia afuera, we have to go back into the
nineteenth century, as we do in most other Latin American cases; indeed, one
of the interesting puzzles concerns the timing of this phase, which clearly
varied from country to country.27 In Brazil it was an incremental story, lacking
sudden surges. In Peru it was a more episodic, stop±go saga. In Chile it came
relatively early, in Venezuela relatively late. The question of timing is clearly
linked to impact: incremental growth was less destabilizing than sudden booms
and busts; the latter depended, to a large degree, on the nature of a country's ±
or a region's ± engagement in the new global division of labour. DõÂaz Alejandro's
`commodity lottery' was a key determinant of economic trends long before the
Alan Knight 125
concession was given in 1837, the line was not completed until 1873. Indeed,
in 1860 ± when the USA had 30 000 miles of track ± Mexico had 15, three of
which served to carry pilgrims from the capital to the shrine of La Guadalupe.
No national network existed until the 1880s: Mexico lagged behind not only
the USA, but also Cuba, Brazil and Argentina.
If domestic capital was scarce and timid, foreign investment was no less leery.
The foreign loans of the 1820s ran into the sand; in the absence of export
markets, foreign direct investment was scant; and its pioneers ± such as the
British Real del Monte Mining Company ± found the problems of transport and
labour recruitment daunting (Randall, 1972, especially ch. 7). The domestic
market was shallow, hence attempts at industrialization failed: the govern-
ment's bold attempt to foster textile manufacturing in the 1840s failed (Potash,
1959). Mexican entrepreneurs, following the trail blazed by Mexican landlords,
strove to make a profit on a low turnover, capturing local markets, and relying
heavily on political favours: tariffs, credit, government contracts (Haber, 1989,
pp. 5, 62). The economy was therefore highly politicized: governments, though
shifting and unstable, could reward their friends and punish their enemies; and
recurrent civil wars gave ample opportunity for punishment and bred a climate
of business desconfianza.30 This in turn encouraged a suspicious and introverted
business ethic, an entrepreneurial equivalent of Banfield's `amoral familism' ±
or, to take an example closer to home, Foster's `limited good' (Banfield, 1958;
Foster, 1965, pp. 293±315). Business was a zero-sum game, played with fixed
stakes. Technological innovation and expansion were risky, even counterpro-
ductive.
In such a climate, families stuck together; high prices and low turnover were
the norm; and the kind of trust-cum-confidence which favours risk-taking and
joint-stock ventures was lacking. These features survived well into the twenti-
eth century.31 Meantime, although it is impossible to estimate welfare for this
period, impressionistic evidence suggests that popular living standards, though
low, were fairly stable.32 Hence, amid recurrent civil wars, popular mobilization
seems to have been directed against political and clerical abuses, not economic
exploitation (Coatsworth, 1988). The clearest exception was YucataÂn, where
the massive rebellion known as the Caste War derived, in part, from the region's
precipitate entry into world markets and the agrarian commercialization which
this provoked (Reed, 1964; Rugeley, 1996).
In a sense, YucataÂn pointed the way forward. During the last quarter of the
nineteenth century Mexico achieved political stability and experienced rapid
export-led growth. The old vicious circle of instability and desconfianza gave
way to a virtuous circle: trade and investment ! government solvency !
government stability ! confidence ! trade and investment. Foreign and
domestic factors intertwined in ways that are difficult to disentangle analytic-
ally. The onset of the Porfiriato (1876) coincided roughly with the Second
Alan Knight 127
Industrial Revolution and the new Kondratieff cycle associated with it (Car-
doso, 1980, pp. 259±67; on Kondratieff (and other) cycles, see Maddison, 1991,
pp. 89±123.). Demand for new industrial raw materials (rubber, fibres, copper,
zinc, lead) combined with increased consumer demand in Europe and the USA
± for coffee, sugar, tropical fruits and gum (chicle). Surplus capital sought foreign
outlets, with Latin America a favoured ± `informally imperial' ± outlet.33
Meanwhile, the newly-established government of Porfirio DõÂaz (1876±1911)
capitalized on Mexico's war-weariness to create a stable, authoritarian regime
which made light of the liberal provisions of the 1857 Constitution and, like
contemporary regimes throughout Latin America, adhered to a positivistic
belief in the primacy of economics. Without economic development, positiv-
ists reasoned, political liberalism was a sham and a recipe for instability; devel-
opment in turn required a strong, centralized government which would favour
foreign trade and investment. Although Porfirian policy-makers ± notably the
celebrated CientõÂficos ± revered the market, they argued, realistically enough,
that in a society where the market was weak the state had to create the appro-
priate conditions for market-led growth.34 Hence, in pursuit of a distant eco-
nomic liberalism ± and a still more distant political liberalism ± the Porfirian
state resorted to a variety of direct interventions designed to promote growth. It
introduced a new Commercial Code in 1884, and a Mining Code in 1892
(Sariego et al., 1988, pp. 28±32); promoted a banking system; suppressed inter-
nal customs (the old colonial alcabalas, 1896) and hiked external tariffs;
balanced the government budget and consolidated the national debt; subsid-
ized railway building and, in the 1900s, consolidated the bulk of the railways in
the Ferrocarriles Nacionales de MeÂxico (CalderoÂn, 1965, pp. 613±22); modernized
ports and promoted telegraph communications; legislated to encourage the
colonization of virgin land and the break-up of corporate village landholdings;
impeded labour organization and repressed social protest; and wooed foreign
business and carefully cultivated Mexico's image in the external world.35 Like
other `late-industrializers', therefore, Mexico pursued a strategy of development
which took the market to be central, but the role of the state to be vitally
supportive (cf. Gerschenkron, 1962).
By its own lights, the Porfirian regime was strikingly successful. It preserved
political stability for 35 years during which time GDP grew some 2.6 per cent;
or, given a population growth rate of 1.4 per cent, at 1.2 per cent per capita per
annum (Reynolds, 1970, pp. 16, 23). Foreign investment rose from minimal
levels to 3.4 billion pesos in 1911 (Rosenzweig, 1965c, vol. 2, pp. 987±90, 1153±
7). Exports, growing at 6.1 per cent per annum, were the motor of growth; and,
within Mexico's diverse export basket, precious metals declined in relative
importance as non-precious metals (copper, lead, zinc) and agricultural raw
materials (coffee, henequen, rubber, cotton) gained, with exporters benefiting
from the long-term depreciation of the Mexican silver peso.36 The new railway
128 Export-led Growth in Mexico, c.1900±30
system ± which grew from 400 to 12 500 miles during the period ± made this
export performance possible (freight rates actively favoured exports); it also
tended to accentuate US dominance of Mexico's markets (Coatsworth, 1984,
chs 4, 5). Railways made bulk production and transport feasible: the cost of
shipping a ton of textiles the 130 miles from QuereÂtaro to Mexico City fell from
US$61 to US$3.37 Such savings made possible the dramatic growth of mining
and metal-refining in northern Mexico; and of aggressive cash-crop production
throughout the country: henequen in YucataÂn, sugar in Morelos, coffee in
Veracruz, cotton in the Laguna, cattle and vanilla in the Huasteca, wheat,
chickpeas and vegetables in the north-western river valleys (CossõÂo Silva,
1965; Knight, 1986a, vol. 1, p. 80ff.). In some cases (minerals, henequen)
production was almost entirely exported; in others (coffee, cotton, sugar), the
domestic market was significant ± hence it would be wrong to dichotomize the
two. In addition, Mexican manufacturing ± of textiles, steel, cement, glass,
dynamite, newsprint, soap ± grew rapidly during the 1890s, which decade is
now rightly seen as the first phase of import substitution industrialization
(ISI).38 ISI and desarrollo hacia afuera were not, therefore, incompatible.39
There can be little doubt, however, that exports were the motor and that the
domestic market lacked the dynamism of the export sector: while agricultural
exports boomed, domestic production lagged ± though by how much it is hard
to say.40 The most dynamic sector of domestic cash-crop production was that of
pulque, which could now be rail-freighted from the huge hacienda pulqueras of
the plains of Apam, into the maw of a thirsty capital whose burgeoning popula-
tion had abundant reason to drown its sorrows.41
For, with the renewed `compression' of the Porfiriato, living standards tended
to fall and social tensions to rise.42 The social impact of Porfirian growth can be
plotted both spatially and chronologically. First, the impact varied by regions.43
In general, northern Mexico ± economically dynamic, underpopulated, mes-
tizo, and more closely tied to the US economy ± benefited from Porfirian
growth. The mines of Sonora and Chihuahua flourished, attracting US capital;
and at the same time ± stimulated by the McKinley tariff of 1890 ± a Mexican
smelting industry grew up dominated by the ASARCO (Guggenheim) interests
(Nava Oteo, 1965, p. 241; Sariego et al., 1988, pp. 36±7). Commercial agricul-
ture, benefiting from the new railways though restricted by water shortage,
sprang up in the Laguna, the Sonora river valleys and the lower RõÂo Grande
valley. Land prices rose (CossõÂo Silva, 1965, pp. 73±4). Monterrey became the
preeminent industrial city of the north, producing steel, glass and beer, to the
advantage of its dynamic entrepreneurial elite (Saragoza, 1988, chs 1±3). Long
familiar with free wage labour, the north experienced no revolution in its
relations of production: the northern bourgeoisie could rely on cash incentives
to attract migrant workers from the populous centre/south (French, 1996, pp.
30±1, 43, 120); hence coercion was at a discount, save in those pockets of the
Alan Knight 129
north ± such as the Yaqui River Valley ± where recalcitrant peasant farmers had
to be dispossessed.44 The north thus conformed to the classic Marxist syndrome
whereby market growth generated free wage labour and an entrepreneurial
bourgeoisie.
At the other end of the country ± the trans-Isthmian south-east ± a very
different outcome ensued. Again, population was scant (and dispossessed pro-
letarians were even scantier); landlords who benefited from buoyant markets ±
for coffee, timber, rubber and henequen ± combined coercion and cash incent-
ives in their search for labour.45 The coffee finqueros of Soconusco on the
Chiapas Pacific coast used the enganche system to draw highland Indians
down from the sierras; with time, a regular flow of labour leading to a perman-
ent labour force developed; and the relations of production in coastal Chiapas
thus roughly resembled those of coastal Peru (GarcõÂa de LeoÂn, 1985, vol.1 pp.
191±8; cf. Albert, 1985). Elsewhere in the south, outright coercion was more
apparent: in the harsh monterõÂas of highland Chiapas, in the notorious coffee
fincas of Oaxaca's Valle Nacional, and on YucataÂn's henequen estates where the
`Divine Caste' ± the Yucateco oligarchy ± blended traditional paternalism with a
more coercive commercial exploitation (Knight, 1986a, vol.1, pp. 97±9; Wells,
1985). Though large subsistence Indian populations remained ± in the Yucatan
lowlands and the highlands of Chiapas and Oaxaca ± it did not prove easy to
`articulate' their subsistence mode of production with commercial farming: not
because Indians were indifferent to market opportunities, but rather because
the available market opportunities ± poor wages for bad jobs ± were unattract-
ive. Thus, in south-eastern Mexico, as in Guatemala, the plantocracy ruled;
labour was coerced and coercion was justified in racist terms: el indio no oye sino
por las nalgas.46
Central Mexico, the most densely populated part of the country, also pre-
sented the most complex economic picture. The rural sector, long dominated
by the hacienda (and rancho),47 eagerly responded to new market opportunities:
the pulque of Hidalgo and the sugar of Morelos were classic cases. The progress-
ive expropriation of peasant village lands ± and the proletarianization of arti-
sans ± helped ensure an adequate supply of land, water and labour; hence ± at
the price of considerable social tension ± landlords increasingly relied on the
swelling free labour market to meet their needs.48 Indeed, some landlords
sought to rationalize their `internal' labour relations by replacing costly `pater-
nalist' perks with a straight cash wage ± a rational option in times of inflating
food prices and growing labour surplus (Bellingeri, 1978, pp. 121±36). In
doing so, they found ± like their Peruvian counterparts ± that rationalizing
`traditional' labour patterns could provoke peon protest (cf. MartõÂnez Alier,
1977, pp. 82±90). `Feudal' relations acted as a brake on commercialization and
a focus for subaltern resistance. Manufacturers in central Mexico ± the country's
industrial heartland ± also benefited, individually and as employers, from
130 Export-led Growth in Mexico, c.1900±30
falling real wages;49 but, collectively and as sellers, they found their products ±
typically, cheap cotton textiles ± hard to sell in Mexico's shallow, constricted,
domestic market. In general, indeed, firms which catered to the domestic
market ± the trams, light and power companies, and some branch railways ±
experienced a less profitable Porfiriato than their exporting counterparts
(Haber, 1989, ch. 7).
Chronologically, too, there were significant differences. During the first
twenty years of the Porfiriato demand for labour rose, encouraged by the rail-
way boom. After the mid-1890s, however, demand slackened and, as popula-
tion growth conspired with both urban and rural proletarianization, the labour
market came to favour buyers (Reynolds, 1970, p. 25; Simpson, 1937, p. 712).
Given both the weakness of the early sindicatos and the hostile stance of the
state ± evident in the celebrated cases of Cananea and RõÂo Blanco ± the counter-
vailing force of unionization could do little to arrest the erosion of real wages.50
Food prices and food imports also rose during the 1900s, squeezing consump-
tion across the board.51 Poor Mexicans probably ate less, appear to have drunk
and smoked less, and certainly bought less.52 They also procreated less or, more
likely, died earlier: population growth fell from 1.6 per cent per annum in 1877±
1900 to 1.1 per cent in 1900±10 (Reynolds, 1970, pp. 22±3). Demand for cheap
textiles ± perhaps the best barometer of popular living standards53 ± declined,
leading to a crisis in Mexico's chief manufacturing sector (Rosenzweig, 1965a,
pp. 330, 338; Knight, 1986a, vol. 1, pp. 134±5). Finally, in 1905 the government
embarked on an ambitious currency reform designed to place Mexico on the
gold standard and thus avoid the vicissitudes of a silver-based currency (BaÂtiz
VaÂzquez y Canudas Sandoval, 1980, pp. 421±31). The desired monetary stability
proved elusive; and deteriorating terms of trade after 1905, coupled with the
coincidental US depression (1907) and poor harvests in 1909±10,54 aggravated
social tensions at a time when ± largely independently ± a vigorous political
opposition was mounting an unprecedented challenge to the DõÂaz dictator-
ship.55
The Revolution which ensued, starting in 1910, obeyed economic motives ±
among others. It was also intimately associated with the preceding phase of
desarrollo hacia afuera. That is not to say that it was inevitable, `overdeter-
mined', or proof of Mexican exceptionality. Most of the socioeconomic ingre-
dients were to be found elsewhere in Latin America: an authoritarian, racist and
positivistic regime bent on export-led development; a dynamic commercial
agriculture linked to peasant dispossession and a degree of labour coercion; a
growing manufacturing sector ± especially in textiles ± catering to a shallow
domestic market;56 and a falling death rate and hence a growing population
which, coupled with urban and rural proletarianization, eased the labour mar-
ket to the advantage of employers who usually had the ear of the state. Mexico
differed only in degree: its growth came late and was unusually rapid, spurred
Alan Knight 131
continued, reaching its nadir in 1916±17 (Knight, 1986a, vol. 2, p. 406ff.). Now
not only confidence and exports collapsed; the very human fabric of the
country suffered as hyperinflation ruined firms and families, as entire regions
retreated into a barter economy (or dealt in US dollars) (tables 5.1, 5.2), as
hunger, sickness and destitution stalked the land. Aggregate national statistics,
it is true, paint a rosier picture; but that picture is heavily coloured by the
petroleum boom which, impelled by the First World War and facilitated by
the enclave character of the industry, ensured a crucial flow of foreign exchange
and tax revenue well into the 1920s.63
While oil financially bolstered the infant revolutionary government, a grow-
ing war-weariness undermined armed resistance, weakened the radical
elements of the labour movement, and, perhaps, engendered a collective
ethic of `reconstruction' ± a phenomenon witnessed in other traumatized
societies (Knight, 1986a, vol. 2, pp. 524±6). As the regime of the Sonorans
(1920±34) embarked on economic reconstruction and renewed state-building,
they could therefore count on broadly favourable domestic conditions, coupled
with grudging US support.64 The `conjunctural' costs of the Revolution were
fairly quickly covered: oil production peaked in 1921, the same year in which
textile production reached its prerevolutionary level (Meyer, 1977, p. 9; Haber,
1989, p. 125). Electricity consumption ± which had been fast rising in the late
Porfiriato ± recovered as early as 1918; beer consumption by 1920 (Reynolds,
1970, p. 28; Haber, 1989, pp. 126, 138). By 1925, GDP was 44 per cent greater
than in 1910 ± 41 per cent in per capita terms (Reynolds, 1970, p. 16). Even
population made a relatively rapid recovery (table 5.1) helped by incremental
improvements in public health.65 In short, as Reynolds observes, `the opportu-
nity cost of the Mexican Revolution was much less than might have been
supposed' (Reynolds, 1970, p. 326).
In this sense, the Revolution produced few dramatic economic consequences.
At best, prerevolutionary trends were somewhat accelerated, or retarded. Popu-
lation growth briefly halted; but rural±urban migration ± and, even more,
Mexico-US migration ± accelerated.66 Manufacturing grew in importance and
± for reasons which will be mentioned shortly ± domestic consumption appears
to have increased, thus alleviating the under-consumptionist crisis of the later
Porfiriato.67 Petroleum production declined following the providential boom of
1910±21; but mining, benefiting from US investment, technological change
and concentration of ownership, enjoyed a prosperous Indian summer.68 In the
agricultural sector, subsistence farming appears to have stagnated,69 but com-
mercial production was buoyant, particularly in respect of certain crops ±
cotton, henequen, sugar, coffee, bananas, chickpeas ± and certain regions
(notably the north-west: the homeland of the Sonoran dynasty) (Reynolds,
1970, pp. 103±5). In the short term, therefore, the Revolution did not reverse
the subsistence/commercial dualism evident in Porfirian agriculture.
Alan Knight 133
Millions of current pesos Millions of pesos of 1950 1950 pesos per capita
Population (millions)
1900±01 13.8
1910±11 15.2
1921 14.3
1925 15.2
1930 16.6
1994, chs 6, 8, 9). Further proof came with the flow of new US manufacturing
investment to Mexico: Ford, Dupont, Colgate-Palmolive.70 Pace some versions,
this did not represent a betrayal of the Revolution. The Sonorans took a `busi-
nesslike' attitude to government, were themselves successful entrepreneurs,
and had never indulged in militant anti-Americanism. Nor had the Revolution
in general been premised on xenophobic anti-Americanism. Porfirian desarrollo
hacia afuera, involving hefty US investment and closer US±Mexican trade rela-
tions, had made a mighty contribution to social tensions and, eventually, to
revolutionary upheaval; but the causal links were complex and indirect, hence
foreign trade and investment were rarely perceived as inherent evils. Revolu-
tionaries targeted proximate victims ± landlords, old regime officials, the
military, the Church ± and, with some exceptions, allowed the major Anglo-
American companies to survive, even to prosper. These companies were not
much liked, but neither were they hated; and, of course, they paid taxes and
brought in foreign exchange. At worst ± in the case of the oil industry ± the
1920s witnessed sporadic skirmishing between companies and government as
the latter sought to increase tax revenue and curtail the autonomy and arrog-
ance of over-mighty corporate subjects (Meyer, 1977, chs 4±7). But there was no
dramatic deÂnouement; that awaited the depression, the rise of Cardenismo, a
new labour militancy, and a new official commitment to desarrollo hacia adentro
(Knight, 1991a).
Revolutionary change was more apparent in the political realm ± a realm
which, in contrast to the economic, was amenable to brusque and purposive
transformations.71 Here, certain formal changes are well-known: the new 1917
Constitution provided for agrarian and labour reform, affirmed national own-
ership of subsoil deposits ± thus `nationalizing' minerals and hydrocarbons ±
and envisaged a more interventionist, social-engineering state which would
arbitrate class conflict, Mexicanize material resources and modernize Mexican
culture (CoÂrdova, 1988, ch. 5). These, however, were distant often rhetorical
goals which required enabling legislation if they were to be implemented. They
also depended on the shifting balance of political forces, both Mexican and
foreign. As regards land reform, progress was slow. Prior to 1933, 4090 ejidal
communities comprising 754 600 ejidatarios had been established; 22 per cent
of all rural communities had benefited from the programme; and 4 per cent of
the area of the Republic, 10 per cent of actual and potential crop land, were to
be found in ejidos (Simpson, 1937, chs 10±11). Sweeping reform therefore
awaited the CaÂrdenas presidency of 1934±40 (Knight, 1991a, pp. 256±64).
However, three important caveats are in order. First, the Revolution severely
weakened the labour-repressive ± even quasi-servile ± systems which had flour-
ished in the Porfirian south. Coercive debt-peonage was ended, or mutated into
free wage labour (Vos, 1988, pp. 228±9; Joseph, 1988, pp. 103±5; Gruening,
1928, pp. 139±40). Second, the agrarian reform ± however gradual and grudging
Alan Knight 135
± represented a marked change from the Porfiriato; the peasantry, far from
being proletarianized, were now being reconstituted under the aegis of the
state. This was, of course, unique in Latin America in this period. Third, the
bare numbers tell only part of the story. Land distribution varied by region:
some states, such as Morelos, were disproportionately affected.72 Even land-
lords who escaped expropriation faced higher taxation and became nervously
aware that the good old days of the Porfiriato were irrevocably gone; some
prudently moved into trade, industry and banking ± or married their daughters
off to powerful revolutionary generals. The character of Mexico's economic
elite began to change as agrarian oligarchs gave way to entrepreneurial busi-
nessmen.
Meanwhile, behind the impersonal statistics of the reparto lay a host of local
stories: of conflict, mobilization, murder, land invasion, litigation and repres-
sion, complete with a cast of agraristas, rojos, bolcheviques, CROMistas, caciques,
pistoleros, defensas sociales, and guardias blancas.73 This, even more clearly,
represented a departure from the Roman peace of the Porfiriato. It did not, of
course, involve a calm, rational and disinterested division of property rights,
but rather a messy and violent process of class conflict with the state ± in its
several manifestations, national, regional and local ± playing an ambiguous
role: sometimes encouraging reform, sometimes stymying it (Knight, 1991b,
pp. 73±104). The outcome of this complex dialectic was therefore to a large
degree unplanned and unforeseen: between 1930 and 1933 it seemed as if the
agrarian reform had been halted, at least at the national level (Simpson, 1937,
chs 8, 24); after 1934, it forged ahead as never before. Popular pressure and
political rivalries determined the outcome, and the `Leviathan state' was more
end-product than first cause of this decisive and unique process.
Agrarismo therefore had a socioeconomic impact out of proportion to the
mere number of hectares distributed. Labour reform, too, though of less struc-
tural significance, also had implications beyond simple ± or not so simple ±
questions of wages, hours and conditions. Concerning the latter, the evidence is
mixed and imperfect. Article 123 ± the Magna Carta of Mexican labour ± was not
regulated until 1931 (Saragoza, 1988, pp. 155±67; Carr, 1981, p. 175). Through
the 1920s, therefore, labour reform displayed an `inescapable variation', by both
region and sector (Gruening, 1928, p. 364; note also Carr, 1981, p. 136). Some
points are clear, however. Unionization, feeble during the Porfiriato, forged
ahead; and by 1926 the major labour confederation, the CROM, claimed 2
million members (Gruening, 1928, p. 339; Carr, 1981, p. 158). In some indus-
tries ± printing, railways, oil, textiles, tramways ± unions were now well-
established, and both employers and political authorities had to reckon with
syndical power. The latter depended on political clout, tactical alliances
with major caudillos, a union presence on the new labour tribunals ± Juntas de
ConciliacioÂn y Arbitraje ± and a gradual trend towards collective contracts
136 Export-led Growth in Mexico, c.1900±30
markedly': Yates (p. 275) notes discrepancies to the tune of 2.7 million cattle,
1.5 million horses, 1.4 million mules, 1.1 million hogs and 3 million sheep.
Rival estimates of the increase in total agricultural production for the period
1950±57 range between 130 to 180 per cent. In general, it seems, estimates are
`downwardly sticky'; officials prefer to inflate rather than deflate. Thus, `on one
occasion . . . the Secretary of Agriculture took the undersecretary's figures for the
corn harvest and added 30 per cent before making his announcement' (Yates, p.
274). In light of these post-1930 problems, we can hardly take pre-1930 figures
at face value. It is quite likely, for example, that estimates exaggerate hacienda
production (much of which was marketed) at the expense of peasant produc-
tion (much of which was locally consumed), thus overestimating the negative
impact of the post-1917 agrarian reform.
The state's `cognitive' failure is sometimes deliberate, the result of design
rather than official incapacity or civic evasion. Corruption is the most obvious
and pervasive example: customs officials turn a blind eye; army officers pad
their payrolls (see note 78); and tax-collectors underrate properties (Knight,
1986a, vol.1, p. 93). If, in the 1990s, the black economy constitutes ± it has been
said ± between 20 per cent and 30 per cent of Mexican GDP ± what was its
extent in 1900 or 1930?
Even if data-collection is honest and efficient, our retrospective reading of
the data may be error-prone. Major misapprehensions have arisen from histor-
ians' (mis)understanding of contemporary categories of analysis: for example,
haciendas, ranchos and pueblos (see notes 47, 57). Tannenbaum (and others)
readily equated peon with `debt peon' and `debt peon', with serf, concluding
(Tannebaum, 1968, p. 30) that millions of Mexicans groaned in rural servitude
in 1910 (which they did not). Errors of this kind easily occur when data series
are collected and collated by analysts who are not closely familiar with the
original sources and who may be taken in by deceptive semantic similari-
ties. The higher the level of aggregation, the greater the danger of such false
assimilation.
There is a final problem, of a more grand theoretical kind. By virtue of seeking
quantifiable data, the economist places a premium on what can be measured.
Thus, as I have said, hacienda production may be exaggerated at the expense of
subsistence production; and industrial output may show up more readily than
artisanal output. Monetized market transactions are therefore readily access-
ible, other transactions are obscured. If `the economy' is defined in terms of
market relations, then the problem is tautologically removed. If it is more
broadly defined ± that is, in Polanyi's terms, as the material means-provisioning
system ± the problem remains. For example: despite vicissitudes the Mexican
economy grew substantially during the twentieth century; per capita income
rose. To the extent that, say, Mexicans lived longer, this was a real transforma-
tion. But in many other respects inferring welfare from income is clearly risky.
Alan Knight 141
Notes
1 With the exception of HeÂctor PeÂrez Brignoli's chapter on `Central America'. However,
Central America (i) has witnessed repeated efforts at political and economic integra-
tion and (ii) is, in its totality, smaller than many individual republics, Mexico
included.
2 For a useful discussion, involving Mexico, see Migdal (1988).
3 In Michael Mann's terminology, such states, for all their `despotic' power, lacked
`infrastructural' power ± `the capacity of the state actually to penetrate civil society
and to implement logistically [sic] political decisions throughout the realm' (Mann,
1988, p. 5).
4 The Banco Nacional, established in 1884, was closely tied to the state and was respon-
sible for 46 per cent of the currency in circulation in 1910. It did not, however, enjoy a
monopoly of emission: the Banco de Londres y MeÂxico accounted for 16 per cent ±
down from 29 per cent in 1900 ± and a swathe of provincial banks scattered through-
out the states of the Federation were responsible for a further 38 per cent: Rosenzweig,
(1965b); ZebaduÂa (1994, pp. 44±5).
5 `They make their sales in the primitive form . . . they do not use money, only by means
of exchanging maize, beans, dried chile and tortillas': report from Coyotepec, Puebla,
1926, cited in Tannenbaum (1968, pp. 85±6). The use of Guatemalan pesos ± which
142 Export-led Growth in Mexico, c.1900±30
p. 65) broadly concurs in respect of a fall in per capita GDP of 37 per cent between
1800 and 1860.
29 Tutino (1986) offers a good overview of these successive phases of `compression' and
`decompression'.
30 Walker (1986); Haber (1997, pp. 17±18) compares the `repressive regulatory' ± and
chronically irresponsible ± financial policies of nineteenth-century Mexican govern-
ments with the more sound and business-friendly policies of their Brazilian counter-
parts, which fostered a more developed capital market.
31 Haber (1989, ch. 6; and pp. 68±83), which discusses industrial finance, noting the
close links between merchant-financiers and industrialists; Haber possibly exagge-
rates the role of foreign finance and entrepreneurship, on which, note Camp (1985,
p. 56 and passim).
32 Travellers accounts of Mexico in the period c. 1820±50 frequently comment on its
rustic primitiveness; but few or none describe poverty, inequality and destitution in
the bleak terms of Humboldt (1988).
33 Knight (1999).
34 The CientõÂficos thus anticipated the `orthodox paradox' of the neoliberal 1990s ± that
is, a `state-led transformation in favor of the market . . . the use of the state to reduce
and diminish the economic influence of politics': Centeno, 1994, p. 34 (the phrase is
Miles Kahler's). Of course, we might better say the neoliberals emulated the CientõÂficos
(mutatis mutandis).
35 Cardoso (1980), second part, offers a good overview based on CosõÂo Villegas' monu-
mental Historia Moderna de MeÂxico: El Porfiriato, La Vida EconoÂmica.
36 CossõÂo Silva (1965) `La agricultura' and Nava Oteo (1965) `La minerõÂa', in CosõÂo
Villegas (1965, vol. 1, chs 1, 3); CaÂrdenas (1997, pp. 80, 83), on the effect of exchange
rate depreciation.
37 Haber (1989, p. 16). `Social savings resulting from the advent of the railroad may have
been as high as 38.5 per cent of GDP for Mexico': Haber, `Introduction', p. 16, citing
William Summerhill, `Transport Improvements and Economic Growth in Brazil and
Mexico', in Haber (1997).
38 Haber (1989, pp. 7±8, 30ff); Rosenzweig (1965a, pp. 329, 342). Manufacturing
industries, like exporters, were helped by exchange rate depreciation (CaÂrdenas,
1997, pp. 83±4); meanwhile, both federal and state government concessions
helped offset the higher cost of imported capital goods: Rosenzweig (1965a, pp.
467±72).
39 As Haber generalizes for Latin America as a whole: Haber (1997, p. 13).
40 Coatsworth (1976, pp. 167±87) rebuts the old myth of a massive decline in per capita
food production during the Porfiriato; but even his own evidence shows a deteriora-
tion in the final years of the regime, which required sizeable food imports to cover
Mexico's shortfall at a time when food prices were rising significantly. See also CossõÂo
Silva (1965, pp. 21±2).
41 Leal and Huacuja Rountree (1982, chs 2 and 3) offers good analysis of the pulque
economy.
42 Knight (1986, vol. 1, ch. 3); Simpson (1937, p. 712), offers a (rare) index of real
agricultural wages for the late Porfiriato, showing a decline from 1903100 to
191068.
43 Katz (1980, pp. 25±48) offers a good, conventional, breakdown (north, centre, south),
which I will follow; Reynolds (1970, pp. 7±8, 94±5) prefers a five-region model.
44 The dispossession of the Yaquis provoked intense conflict, leading to mass deporta-
tion of Yaqui families which some landlords ± dependent on Yaqui labour ± sought to
Alan Knight 145
avert; but racism and Realpolitik prevailed over economic interest: Knight (1986, vol.
1, pp. 111±12).
45 Vos (1988); Joseph (1988, chs 1, 2). The ratio of coerced to voluntary labour is a
matter of debate: see Knight (1986b, pp. 41±74).
46 `Indians only hear through their backsides': Joseph (1988, p. 70).
47 Recent research has rehabilitated the rancho ± roughly, the small farm ± as an
important element in Mexican rural history, thus qualifying the old dichotomy of
hacienda versus village; it has also confirmed ± rather than discovered de novo ± that
categories like hacienda and rancho were often used arbitrarily or inconsistently: see
Tannenbaum (1968, pp. 30±1) and Guerra (1985, vol. 1, pp. 325±9 and vol. 2, annexe
5). A general overview such as this does not permit a full discussion; I shall focus on
the hacienda ± roughly, the large profit-seeking private farm ± although much of what
is said also applies, to some degree, to the rancho ± roughly, the small profit-seeking
private farm. Apart from blurring this (`emic') distinction, I shall avoid making any
strict (`etic') distinction between hacienda and `plantation'; what for some are `plan-
tations', are for me simply haciendas producing `tropical' crops, usually for export.
48 Womack Jr. (1968, ch. 2) is the classic study. On artisanal decline: Knight (1986, vol. 1,
pp. 131±2); Haber (1989, p. 58); Rosenzweig (1965a, pp. 320±2, 340±1), which charts
an 83 per cent decline in artisan consumption of raw cotton during the Porfiriato.
49 Labour may have been cheap, but it was also reckoned to be feckless and inefficient:
Haber (1989, p. 35); Gruening (1928, pp. 350, 385).
50 In fact, the most common form of prerevolutionary working-class organization was
not the trade union but the mutualist society; self-help seemed more realistic than
syndical militancy: Knight (1986, vol. 1, pp. 133±50); French (1996, pp. 133±7).
51 Reynolds (1970, p. 105); Rosenzweig (1965b, pp. 882±3). Cf. Yates (1981, p. 3), which
presents Angulo's optimistic and contentious figures for the 1900s.
52 Cigarette packets sold: 1906±07: 524 m; 1908±09: 505 m; CervecerõÂa CuauhteÂmoc
production: 1900: 4.87 m litres; 1907: 14 m; 1909: 11.58 m. Haber (1989, pp. 49, 53).
53 Rawski (1996, p. 51).
54 Between 1905 and 1910 Mexico's net barter terms of trade fell from 124 to 93
(1937100): Reynolds (1970, p. 43 n. 24). The US recession, of course, hit the north-
ern mining economy hardest: French (1996, pp. 30±2).
55 There is considerable debate concerning the relationship between the economic
downturn of the late Porfiriato and the growth of political opposition. I would
question any neat causal link, partly on theoretical grounds, partly because the
opposition (i) had deep roots, antedating 1907, and (ii) followed a politico-electoral
± rather than a business cycle ± timetable: Knight (1986, vol. 1, pp. 64±5).
56 Shallow in comparison with, say, Argentina, where, though the population was half
that of Mexico, manufacturing output in 1913 was 67 per cent greater: Victor Bulmer-
Thomas (1994, p. 137).
57 I am referring chiefly to Mexico's `external peasantry' ± that is, those who remained
in villages and had not been absorbed into hacienda (`internal peasant') communities.
Contrary to some received wisdom, they were still a sizeable majority in 1910: Jean
Meyer (1986, pp. 477±509).
58 By no means all peasant rebels were agraristas, seeking land reform; some were
motivated by the desire for local political autonomy, which Porfirian state-building
± partly funded and facilitated by export-led growth ± had severely undermined: I
have referred to such rebels (loosely) as serranos: Knight (1986, vol. 1, pp. 115±27).
59 Again, a necessary caveat, given that revisionist scholarship would question the
`revolutionary' impact of the Revolution (see note 22 above).
146 Export-led Growth in Mexico, c.1900±30
60 European economies during the Second World War are a case in point; within our
own field, the case of Colombia ± both now and during the Violencia ± springs to
mind.
61 Military spending as a percentage of total government spending: 1910, 20 per cent;
1913, 26 per cent; 1917, 70 per cent; 1925, 31 per cent; 1935, 21 per cent: Wilkie
(1970, p. 102).
62 Marichal (1989, pp. 178, 213). Mexico's default helped accelerate the flow of US funds
to South America: Stallings (1987, pp. 125±6).
63 Womack (1978, pp. 80±123). Oil production rose from 3.6 m barrels in 1910 to 26.2 m
in 1914, 63.8 m 1918, and 193.4 m in 1921 (the peak); petroleum taxes contributed
11 per cent of federal government receipts in 1918, 21 per cent in 1920 and (a freak)
34 per cent in 1922, after which decline set in; during this period oil production
represented between 6 and 7 per cent of GNP: L. Meyer (1977, pp. 8±9, 15, 16).
64 US policy towards the institutional revolution ± which combined a desire for stability
with antipathy to radical reform ± is well-described by Smith (1972 ) and ZebuduÂa
(1994, chs 5, 8, 9). Erfani (1995, pp. 5±6, 10ff.), argues that the revolutionary state
fortuitously benefited from the emergence of `new norms about state sovereignty'
(that is, respect for national self-determination) which came to govern international
± especially US-Mexican ± relations in the 1920s (or before: if the inspiration is
Wilsonian, we should backdate the shift to c. 1913). There is some truth in this
argument, though not as much as the author believes.
65 The 1921 census gives a total population of 14.33 m, compared to 15.16 m in 1910;
natural increase ± in the absence of revolution ± would have resulted in a population
of around 16.91 m in 1921. However, the 1921 census undercounted; Reynolds
estimates (1970, p. 18, 26) to the tune of nearly half a million, which does not strike
me as excessive. Thus, the absolute population loss was between 300 000 and 400 000;
the demographic deficit ± taking into account natural increase that did not occur ±
nearer 2 millions. Since birth rates do not appear to have changed much over time
(see Reynolds, 1970, p. 18 and Cook and Borah, 1974, vol. 2, p. 319), the chief
determinant was the death rate which, having gradually declined during the Porfir-
iato (reaching 3.29 per cent in 1910), shot up during the Revolution, then, in the
1920s, resumed its gentle descent, initially due to the elimination of smallpox and
typhus: 1921, 2.84 per cent; 1930, 2.67 per cent. Revolutionary mortality was very
largely a matter not of fighting but of dearth and disease ± in particular, the Spanish
influenza which ravaged a weary world in 1918±19 (Knight, 1986, vol. 2, pp. 420±2);
thus, younger males did not suffer disproportionately. Arguably, disease killed off the
older, weaker members of society, leaving a younger, more productive postrevolu-
tionary population.
66 The rural population was 71.2 per cent of the population in 1910, 68.8 per cent in
1921, and 66.5 per cent in 1930 (Reynolds, 1970, p. 93). During the decade of
revolution, Mexico City grew by a quarter; Ciudad Juarez by 140 per cent (Knight,
1986, p. 523).
67 For example, cotton textile production: 1910, 34 700 tons (cotton consumed); 1918,
20 300; 1921, 31 700; 1925, 43 700: Haber (1989, p. 125).
68 Meyer (1977, pp. 8±9); Bernstein (1964); Sariego et al. (1988, pp. 63±74). Brown (1985,
pp. 362±85) convincingly argues that subsoil depletion was more important than
radical reform in obliging US interests to shift from Mexico to Venezuela.
69 Again, this is tricky question. According to Reynolds (1970, p. 105) production of
corn and beans fell by 40 per cent and 31 per cent respectively between 1907 and
1929, while population grew 9 per cent. For graphic confirmation, see Simpson
Alan Knight 147
(1937, p. 499). While there was no doubt some switch of consumption away from
these staples, the fall seems excessively large and may reflect the old problem of
adequately measuring subsistence production ± especially during a period when a
peasant revolution had recently concluded and endemic rural violence, including the
Cristero rebellion, was continuing.
70 Haber (1989, p. 143); Krauze (1977, pp. 289±93). The development of the Mexican
consumer market, which was obviously linked to processes of import substitution,
urbanization and US direct investment in manufacturing, has been little studied; as a
glance at newspaper adverts suggests, consumerism ± incipient during the later
Porfiriato ± forged ahead during the 1920s and (even) the 1930s.
71 Socialist revolutions ± for example, Russia 1917, Cuba 1959 ± undertook rapid and
radical economic transformation; but Mexico's Revolution was not socialist and ± in
the manner of other `bourgeois' revolutions, perhaps ± it placed more emphasis on
political and cultural change, especially in the 1920s.
72 Simpson (1937, p. 622), shows that ejidos controlled as much as 87 per cent of
Morelos crop land in 1933: compare a national average of 10 per cent. Not that
reform guaranteed well-being: Gruening (1928, pp. 162±3).
73 Good studies are: Friedrich, (1977) and Craig (1983).
74 Carr (1981, p. 134), quoting a US embassy report of April 1924.
75 Simpson (1937, p. 712): real agricultural wages (1903 100): 1910: 68; 1920: 120;
1925: 135; 1930: 118.
76 For example, Carrillo Puerto's brief radical experiment in YucataÂn, or Adalberto
Tejeda's Veracruzano agrarianism: Joseph (1988, part 3); Fowler Salamini (1977).
77 Government spending in relation to GDP: 1910: 4.4 per cent; 1925: 5.9 per cent;
1930: 6.7 per cent; 1935: 6.8 per cent; 1940: 8.2 per cent. From Wilkie (1970, p. 22)
and Reynolds (1970, p. 16). Comparable, though slightly different, figures are given
in Table 5.1.
78 Wilkie (1970, pp. 58, 160±1). The process of demilitarization not only incurred
praetorian protest and revolt, but also involved considerable resort to graft and
peculation by way of buying off dissident officers: for a good example, see ZebaduÂa
(1994, p. 343). Gruening (1928, 319±31) presents a sustained indictment of `scanda-
lous conduct and public debauchery. . . in the army'; the economic costs of graft are,
of course, hard to quantify.
79 Cook and Borah (1974, pp. 432±5); Gruening (1928, pp. 544±50), which shows the
federal health budget rising from around half a million pesos p.a. in the late Porfiriato
to 8.4 m in 1927.
80 ZebaduÂa (1994, pp. 242, 255ff.), which notes (p. 258) the parallel with the Kemmerer
missions to South America: Kemmerer was not called upon to reform Mexico's finan-
cial system ± although he did play an advisory part in the currency stabilization of
1916±17, from which flowed his valuable study of Mexico's revolutionary inflation:
Inflation and Revolution: Mexico's Experience of 1912±17 (Princeton, 1940); however, the
Mexican banking reforms of the 1920s were distinctly Kemmerer-ish in character. See
also Krauze (1977, pp. 35±52, 134±45) (irrigation); Wionczek (1967, pp. 64±5, 191).
References
Albert, B. (1985) `External Forces and the Transformation of Peruvian Coastal Agriculture',
in C. Abel and C. Lewis (eds), Latin America, Economic Imperialism and the State, London:
Athlone Press.
148 Export-led Growth in Mexico, c.1900±30
Balderrama, F.E. and RodrõÂguez, R. (1995) Decade of Betrayal. Mexican Repatriation in the
1930s, Albuquerque: University of New Mexico Press.
Banfield, E. (1958) The Moral Basis of a Backward Society, Illinois: Glencoe.
BaÂtiz VaÂzquez, J.A. and Canudas Sandoval, E. (1980) `Aspectos financieros y monetarios',
in C. Cardoso (ed.), MeÂxico en el siglo XIX, Mexico: Nueva Imagen.
Bellingeri, M. (1978) `Del peonaje al salario: el caso de San Antonio Tochatlaco de 1880 a
1920', Revista Mexicana de Ciencias PolõÂticas y Sociales, vol. 24, no. 91.
Bernstein, M. (1964) The Mexican Mining Industry, 1890±1950, Albany: State University of
New York.
Brenner, R. (1975) `Agrarian Class Structure and Economic Development in Pre-Industrial
Europe', Past and Present, vol. 70.
Brown, J. (1985) `Why Foreign Oil Companies Shifted their Production from Mexico to
Venezuela during the 1920s', American Historical Review, vol. 90.
Bulmer-Thomas, V. (1994) The Economic History of Latin America since Independence, Cam-
bridge: Cambridge University Press.
CalderoÂn, F.R. (1965) `Los ferrocarriles', in C. Villegas (ed.), Historia Moderna de MeÂxico: El
Porfiriato, La Vida EconoÂmica (vol. 1), Mexico: Editorial Hermes.
Calles, P.E. (1927) Mexico Before the World, New York: Academy Press.
Camp, R.A. (1985) Entrepreneurs and Politics in Twentieth-century Mexico, New York: Oxford
University Press.
CaÂrdenas, E. (1997) `A Macroeconomic Interpretation of Nineteenth-Century Mexico', in
S.H. Haber (ed.), How Latin America Fell Behind, Stanford: Stanford University Press.
ÐÐ (1987) La industrializacioÂn de MeÂxico durante la Gran DepresioÂn, Mexico: El Colegio de
MeÂxico.
Cardoso, C. (1980) MeÂxico en el siglo XIX (1821±1910): Historia EconoÂmica y de la Estructura
Social, Mexico: Nueva Imagen.
Carr, B. (1981) El movimiento obrero y la polõÂtica en MeÂxico, Mexico: Ediciones Era.
Cedillo VaÂzquez, V. (1964) Principales factores del desarrollo agrõÂcola en MeÂxico (1925±60),
Mexico: Universidad AutoÂnoma de MeÂxico.
Centeno, M.A. (1994) Democracy Within Reason. Technocratic Revolution in Mexico, Univer-
sity Park, Penn.: The Pennsylvania State University, Press.
Chase, S. (1931) Mexico: A Study of Two Americas, New York: The Literary Guild.
Coatsworth, J.H. (1988) `Patterns of Rural Rebellion in Latin America: Mexico in Com-
parative Perspective', in F. Katz (ed.), Riot, Rebellion and Revolution: Rural Social Conflict in
Mexico, Princeton: Princeton University Press.
ÐÐ (1984) El impacto econoÂmico de los ferrocarriles en el porfiriato, Mexico: Ediciones Era.
ÐÐ (1976) `Anotaciones sobre la produccioÂn de alimentos durante el Porfiriato', Historia
Mexicana, vol. 26(2).
Cook, S.F. and Borah, W. (1974) Essays in Population History. Mexico and the Caribbean, 2
vols, Berkeley: University of California Press.
CoÂrdova, A. (1988) La ideologõÂa de la RevolucioÂn Mexicana, first pub. 1973, Mexico: Edi-
ciones Era.
CosõÂo Villegas, D. (ed.) (1965) Historia Moderna de MeÂxico: El Porfiriato, La Vida EconoÂmica,
2 vols, Mexico: Editorial Hermes.
CossõÂo Silva, L. (1965) `La agricultura', in CosõÂo Villegas (ed.), Historia Moderna de MeÂxico.
El Porfiriato, La vida econoÂmica, Vol. 1, Mexico: Editorial Hermes.
Craig, A.L. (1983) The First Agraristas, Berkeley: University of California Press.
DõÂaz Fuentes, D. (1993) Las polõÂticas fiscales latinoamericanas frente a la gran depresioÂn,
Madrid: Ministerio de EconomõÂa y Hacienda.
Alan Knight 149
Erfani, J.A. (1995) The Paradox of the Mexican State. Rereading Sovereignty from Independence
to NAFTA, Boulder: Lynne Rienner.
Farris, N.M. (1984) Maya Society under Colonial Rule, Princeton: Princeton University Press.
Fischer, D.H. (1970) Historians' Fallacies, New York: Harper.
Florescano, E. (1986) Precios de maõÂz y crisis agrõÂcolas en MeÂxico, 1708±1810, Mexico:
Ediciones Era.
Foster, G. (1965) `Peasant Society and the Image of the Limited Good', American Anthro-
pologist, vol. 67.
Fowler Salamini, H. (1977) Agrarian Radicalism in Veracruz, 1920±38, Lincoln: University of
Nebraska Press.
French, W.E. (1996) A Peaceful and Working People. Manners, Morals and Class Formation in
Northern Mexico, Albuquerque: University of New Mexico Press.
Friedrich, P. (1977) Agrarian Revolt in a Mexican Village, Chicago: University of Chicago
Press.
Gamio, M. (1982) Forjando Patria, first pub. 1916, Mexico: Editorial PorruÂa.
GarcõÂa de LeoÂn, A. (1985) Resistencia y utopõÂa, MeÂxico: Ediciones Era.
Gerschenkron, A. (1962) Economic Backwardness in Historical Perspective, New York: Praeger.
GonzaÂlez, L. (1979) Historia de la RevolucioÂn Mexicana. PerõÂodo 1934±40. Los artõÂfices del
Cardenismo, Mexico: El Colegio de MeÂxico.
Gruening, E. (1928) Mexico and its Heritage, London: Stanley Paul.
Guerra, F.X. (1985) Le Mexique. De l'Ancien ReÂgime aÁ la ReÂvolution, 2 vols, Paris: L'Harmat-
tan.
Haber, S.H. (ed.) (1997) How Latin America Fell Behind: Essays on the Economic Histories of
Brazil and Mexico, 1800±1914, Stanford: Stanford University Press.
ÐÐ (1989) Industry and Development: The Industrialisation of Mexico, 1890±1940, Stanford:
Stanford University Press.
Hassig, R. (1985) Trade, Tribute and Transportation. The Sixteenth-Century Political Economy
of the Valley of Mexico, Norman: University of Oklahoma Press.
Huber, E. and Safford, F. (eds) (1995) Agrarian Structure and Political Power. Landlord and
Peasant in the Making of Latin America, Pittsburgh: University Pittsburgh Press.
Humboldt, A. von, (1988) Political Essay on the Kingdom of New Spain, first pub. 1811,
Norman: University of Oklahoma Press.
Joseph, G.M. (1988) Revolution from Without. YucataÂn, Mexico and the United States, 1880±
1924, Durham: Duke University Press.
Katz, F. (1980) La servidumbre agraria en MeÂxico en la eÂpoca porfiriana, Mexico: Editorial Era.
Kemmerer, E.W. (1940) Inflation and Revolution: Mexico's Experience of 1912±17, Princeton:
Princeton University Press.
Knight, A. (1999), `Britain and Latin America in the Nineteenth Century', in W. R. Louis,
(ed.), The Oxford History of the British Empire, Vol. 3, Oxford: Oxford University Press.
ÐÐ (1997), `The Modern Mexican State: Theory and Practice', paper presented at the
LASA congress, Guadalajara, April.
ÐÐ (1992a) `The Peculiarities of Mexican History: Mexico Compared to Latin America,
1821±1992', Journal of Latin American Studies, quincentenary edition.
ÐÐ (1992b) `Revisionism and Revolution: Mexico Compared to England and France',
Past and Present, vol. 134, pp. 180±8.
ÐÐ (1991a) `The Rise and Fall of Cardenismo', in L. Bethell (ed.), Mexico since Indepen-
dence, Cambridge: Cambridge University Press.
ÐÐ (1991b) `Land and Society in Revolutionary Mexico: The Destruction of the Great
Haciendas', Mexican Studies/Estudios Mexicanos, vol. 7(1), pp. 73±104.
150 Export-led Growth in Mexico, c.1900±30
ÐÐ (1987) US±Mexican Relations, 1910±40: An Interpretation, San Diego: Center for US-
Mexican Studies.
ÐÐ (1986a) The Mexican Revolution, 2 vols, Cambridge: Cambridge University Press.
ÐÐ (1986b) `Mexican Peonage: What Was It and Why Was It?', Journal of Latin American
Studies, vol. 18(1).
Krauze, E. (1977) La reconstruccioÂn econoÂmica, Mexico: Colegio de MeÂxico.
Leal, J. F. and Huacuja Rountree, M. (1982) EconomõÂa y sistema de haciendas en MeÂxico,
Mexico: Editorial Era.
Leroy Ladurie, E. (1969) Les paysans de Languedoc, Paris: Flamarion.
Maddison, A. (1991) Dynamic Forces in Capitalist Development, Oxford: Oxford University
Press.
Mann, M. (1988) `The Autonomous Power of the State: Its Origins, Mechanisms and
Results', in Mann, M., States, War, and Capitalism, Oxford: Blackwell.
Marichal, C. (1989) A Century of Debt Crises in Latin America, Princeton: Princeton Uni-
versity Press.
MartõÂnez Alier, J. (1977) Haciendas, Plantations and Collective Farms, London: Frank Cass.
Marx, K. (1957) Capital, 2 vols, London: Dent.
McKellar, M.M. (1994) Life on a Mexican Ranche [sic], Cranbury, NJ: Associated University
Presses.
Meyer, J. (1986) `Haciendas y ranchos, peones y campesinos en el Porfiriato: algunas
falacias estadõÂsticas', Historia Mexicana, vol. 35(3).
Meyer, L. (1977) Mexico and the United States in the Oil Controversy, 1917±42, Austin:
University of Texas Press.
Migdal, J.S. (1988) Strong Societies and Weak States. State±Society Relations and State Cap-
abilities in the Third World, Princeton: Princeton University Press.
Nava Oteo, G. (1965) `La minerõÂa', in CosõÂo Villegas (ed.), Historia Moderna de MeÂxico. El
Porfiriato, La vida econoÂmica, Vol. 1, Mexico: Editorial Hermes
North, D. (1990) Institutions, Institutional Change, and Economic Performance, Cambridge:
Cambridge University Press.
Potash, R.A. (1959) El banco de avõÂo en MeÂxico. El fomento de la industria, 1821±46, Mexico:
Fondo de Cultura EconoÂmica.
Randall, R.W. (1972) Real del Monte: A British Mining Venture in Mexico, Austin: University
of Texas Press.
Rawski, T.G. (1996) `Issues in the Study of Economic Tends', in Rawski et al., Economics and
the Historian, Berkeley: University of California Press.
Reed, N. (1964) The Caste War of YucataÂn, Stanford: Stanford University Press.
Reynolds, C.W. (1970) The Mexican Economy: Twentieth-Century Structure and Growth, New
Haven: Yale University Press.
Rosenzweig, F. (1965a) `La industria', in CosõÂo Villegas (ed.), Historia Moderna de MeÂxico. El
Porfiriato. La Vida EconoÂmica, 2 vols, Mexico: Editorial Hermes.
ÐÐ (1965b) `Moneda y bancos', in D. CosõÂo Villegas (ed.), Historia Moderna de MeÂxico. El
Porfiriato. La Vida EconoÂmica, 2 vols, Mexico: Editorial Hermes.
ÐÐ (1965c) `Las inversiones extranjeras', in D. CosõÂo Villegas (ed.), Historia Moderna de
MeÂxico. El Porfiriato. La Vida EconoÂmica, 2 vols, Mexico: Editorial Hermes.
Rugeley, T. (1996) YucataÂn's Maya Peasants and the Origins of the Caste War, Austin: Uni-
versity of Texas Press.
Saragoza, A.M. (1988) The Monterrey Elite and the Mexican State, Austin: University of Texas
Press.
Sariego, J.L. (1988) Enclaves y minerales en el norte de MeÂxico, Mexico: Editorial de la Casa
Chata.
Alan Knight 151
Sariego, J.L. Reygadas, L., GoÂmez, M.A. and Farrera, J. (1988) El Estado y la minerõÂa
mexicana. PolõÂtica, trabajo y sociedad durante el siglo XX, Mexico: Fondo de Cultura
EconoÂmica.
Simpson, E.N. (1937) The Ejido: Mexico's Way Out, Chapel Hill: University of North
Carolina Press.
Smith, R.F. (1972) The United States and Revolutionary Nationalism in Mexico, 1916±32,
Chicago: The University of Chicago Press.
SolõÂs, L. (1971) La realidad econoÂmica mexicana: retrovisioÂn y perspectivas, Mexico: Siglo XXI.
Stallings, B. (1987) Banker to the Third World: US Portfolio Investment in Latin America, 1900±
86, Berkeley: University of California Press.
Summerhill, W. (1997) `Transport Improvements and Economic Growth in Brazil and
Mexico', in S.H. Haber (ed), How Latin America Fell Behind, Stanford, Stanford University
Press.
Tannenbaum, F. (1968) The Mexican Agrarian Revolution, first pub. 1929, Hamden, Conn.:
Archon Books.
Taylor, W.B. (1985) `Between Global Process and Local Knowledge: An Inquiry into
Early Latin American Social History, 1500±1900', in O. Zunz (ed.), Reliving the Past.
The Worlds of Social History, Chapel Hill: University of North Carolina Press.
Turner, F.C. (1968) The Dynamic of Mexican Nationalism, Chapel Hill: University of North
Carolina Press.
Tutino, J. (1986) From Insurrection to Revolution in Mexico: Social Bases of Agrarian Violence,
1750±1940, Princeton: Princeton University Press.
Victoria, C.S.J. and VelaÂzquez RamõÂrez, S. (1988) `El estado y las polõÂticas econoÂmicas en el
porfiriato', in C. Cardoso, (ed.), MeÂxico en el siglo XIX. Historia econoÂmica y de la estructura
social, Mexico: Nueva Imagen.
Vos, J. de, (1988) Oro verde. La conquista de la Selva Lacandona por los madereros tabasquen Ä os,
1822±1949, Mexico: Fondo de Cultura EconoÂmica.
Walker, D.W. (1986) Kinship, Business and Politics: The MartõÂnez del RõÂo Family in Mexico,
1823±67, Austin: University of Texas Press.
Wells, A. (1985) YucataÂn's Gilded Age: Haciendas, Henequen, and International Harvester,
1860±1915, Albuquerque: University of New Mexico Press.
Wilkie, J.W. (1970) The Mexican Revolution: Federal Expenditure and Social Change since
1910, Berkeley: University of California Press.
Wionczek, M. (1967) El nacionalismo mexicano y la inversioÂn extranjera, Mexico: Siglo
Veintiuno Editores.
Wolf, E.R. and Hansen, E.C. (1967) `Caudillo Politics: A Structural Analysis', Comparative
Studies in Society and History, vol. 9(1), January.
Womack, J. (1978) `The Mexican Economy During the Revolution, 1910±40: Historio-
graphy and Analysis', Marxist Perspectives, vol. 1(4).
Womack, J. Jr, (1968) Zapata and the Mexican Revolution, New York: Knopf.
Yates, P. L. (1981) Mexico's Agricultural Dilemma, Tucson: University of Arizona Press.
ZebaduÂa, E. (1994) Banqueros y revolucionarios: la soberanõÂa financiera de MeÂxico, 1914±29,
Mexico: El Colegio de MeÂxico.
6
Peru, 1884±1930: A Beggar Sitting on a
Bench of Gold?
Paulo Drinot
* I am grateful to all the participants at the conference held at Lake AtitlaÂn, Guatemala, for their
comments and suggestions on an earlier version of this chapter. Special thanks go to Shane Hunt,
Nelson Manrique, Rosemary Thorp and the editors who have commented extensively. The usual
caveats apply.
152
Paulo Drinot 153
After a brief glance at the background to the period under review, I begin by
studying the period from a macro or national perspective. This first section
examines both the exceptional circumstances arising in the 1890s, which
resulted in import substitution and industrialization and the potential for
autonomous development, and the failure to sustain this `model' of develop-
ment, linked to both endogenous and exogenous factors including interna-
tional price fluctuations and the penetration of foreign, especially US, capital.
The second section of the paper disaggregates the national economy into its
regional components. Here emphasis is placed on the effect of the export
economy at the regional level and its interaction with and impact on pre-
existing conditions. The end of the War of the Pacific, 1884, has been chosen
as the starting date, although as will become evident the multiple processes
under study both predate and postdate this particular year. The end year, 1930,
is self-explanatory, although again, many of the processes or themes that arise
in this period, impervious to academic periodization, naturally flow through to
the following period. Unlike Brazil or Argentina, in Peru the Depression did not
result in substantial structural change, or, perhaps more to the point, in desar-
rollo hacia adentro.
Peru's geography makes three regional spaces immediately obvious: the coast,
the sierra and the selva. However, the country's history adds further complexity
to this neat geographical triptych. For example, it is worth noting the coin-
cidence between the regions under study and distinct pre-Inca cultural areas
(Lumbreras, 1980). Similarly, the historical trajectory of colonial rule varied
substantially from one region to the next, leaving markedly different legacies
(Lockhart, 1974, ch. 2). The contrast between the central sierra, where the
colonial legacy was one of strong peasant communities and weak landlords,
and the south, where the opposite held, is instructive. The early republican
period saw further evidence of disarticulation. For the first two to three decades
of republican rule, a succession of civil wars were fought to determine the
country's political centre of gravity. In the end, coastal forces and interests
overcame southern interests. The coast's victory and the subsequent pacifica-
tion and consolidation of a Lima-based state rule under Castilla was in no small
way aided by the discovery of undreamed-of wealth piled up on islands off the
mainland (Gootenberg, 1989).
Between 1840 and 1879 Peru exported 11 to 12 million tons of guano, worth
some 750 million pesos (Bonilla, 1980). Hunt has argued that guano created a
rentier economy that, in combination with poor state policy, resulted in a lost
opportunity (Hunt, 1985). However, as Hunt himself shows, the guano prosper-
ity through its varied linkages to the national economy, some of greater benefit
than others, was central to Peru's mid-century revival. Indeed, guano revenues
contributed to the pacification of the country, to investment in ± largely
unneeded ± transport infrastructure, and to the creation of a modern, albeit
154 Peru, 1884±1930
Following the war, the CaÂceres government undertook the difficult task of
rebuilding the country's economic foundations. In 1889, the CaÂceres regime
signed the Grace contract which handed over Peru's railways and other
assets to British bondholders organized around the Peruvian Corporation,
created to run the railways, in exchange for the settlement of Peru's foreign
Paulo Drinot 155
debt and the restoration of its international creditworthiness (Miller, 1976 and
1983). Incipient institutions were created to integrate a country whose disunity
was singled out in contemporary debates as a major factor in Peru's wartime
defeat. In addition to identifying and mapping Peru's disputed international
boundaries, the members of the Sociedad GeograÂfica de Lima, founded in 1884,
set out to map the country's natural resource endowment, braving unexplored
Amazon rivers and the highest peaks of the Andes (LoÂpez-OcoÂn Cabrera, 1995).
With Pierola's defeat of CaÂceres in the Civil War of 1894±95, the Civilista party,
in government when the War of the Pacific broke out, returned to power in
alliance with an old foe. This event marked the beginning of a new era in
Peruvian history, known as the Aristocratic Republic. Although not as homo-
geneous or hegemonic as much of the literature indicates (see Miller, 1982), the
Civilistas promoted an export-oriented economy with minimal state interven-
tion and which, in practice, favoured coastal sugar and cotton planters. Produ-
cer associations like the Sociedad Nacional de Agricultura and the Sociedad
Nacional de Industrias, were created to fulfil a parastatal coordinating role and
to promote development.
It was in this context that an exceptional export boom occurred. As Tables 6.1
and 6.2 show, between 1895 and 1900 almost all export products experienced
strong increases in volume and value. In all, exports by value rose by 137 per
cent in this period. As Thorp and Bertram (1978) have shown, export growth
went hand in hand with economic diversification and industrialization. Several
factors combined to create such an outcome, including a highly favourable
depreciating exchange rate ± between 1890 and 1897 the depreciation was
nearly 40 per cent ± and the high reinvestment propensity of certain export
sectors, particularly coastal agriculture and mining. To the natural protection
afforded by the exchange rate was added a high import tariff policy, adopted
Table 6.1 Peru, growth rates of exports by volume, 1884±1930 (Five-year averages)
Source: Up to and including 1910 the estimations are based on Thorp and Bertram (1978), p. 330.
From 1911 to 1930, the estimations are based on Portocarrero, BeltraÂn and Romero (1992), pp. 131,
135, 142. All oil figures are taken from the latter source and are expressed in thousands of barrels.
156 Peru, 1884±1930
Table 6.2 Peru, growth rates of exports by value, 1884±1930 (five-year averages)
1884±1890 12 160 58 61 22 69 2
1890±1895 16 24 11 8 26 133 0.4 7
1895±1900 115 8 8 36 99 3 028 115 137
1900±1905 17 22 61 48 72 17 72 6 479 7
1905±1910 8 132 18 283 282 84 41 864 53
1910±1915 139 50 46 17 134 47 988 68
1915±1920 26 530 53 12 5 208 11 73
1920±1925 21 18 6 5 39 87 88 19
1925±1930 8 30 50 55 14 64 35 4
Source: Up to and including 1910 the estimations are based on Thorp and Bertram (1978), pp. 334±5.
From 1911 to 1930, the estimations are based on Portocarrero, BeltraÂn and Romero (1992), p. 133, 137,
142. All oil figures are taken from the latter source.
not for protection but as the only source of revenue available for a government
too weak to impose export duties on its export-oriented oligarchy.3 Indeed, in
1900 customs duties represented around 60 per cent of all tax income (Porto-
carrero, BeltraÂn and Romero, 1992, p. 120).
Elite investment groups fuelled diversification through a dynamic realloca-
tion of export revenues. Coastal planters, for example, strongly diversified their
business interests. In 1927, RamoÂn Aspillaga, owner of the CayaltõÂ plantation,
held 60 per cent of his investments in banks and insurance companies; both the
Pardo and Prado families established partnerships with Italian immigrant
industrialists to invest in the emerging textile industry;4 and in addition the
Prado family invested in electric and transport infrastructure (the Empresas
EleÂctricas Asociadas, which ran the Lima tramway system) and finance (the
Banco Popular, as well an insurance company) (Gilbert, 1982, pp. 127, 158±60;
Portocarrero, 1995; Quiroz, 1988; Portocarrero and TorrejoÂn, 1992).
As a result, the share of manufactured consumer goods in total imports fell
from 58 per cent in 1891±92 to 39 per cent in 1907. Further evidence of import
substitution is reflected in the rise of local production of cotton textiles as a
percentage of total supply from 5 per cent in 1891 to 47 per cent in 1908. By
1905, one survey claimed there were 173 urban manufacturing firms in Lima.
Similarly, Lima saw the development of large utilities companies and a financial
and banking sector. Moreover, the export boom had important indirect
demand effects on the local capital goods industry: until the early 1900s most
sugar mills as well as the smelters used in mining were produced by local
foundries (Thorp and Bertram, 1978, pp. 28±36).
Eventually, the diversification process was reversed with the adoption of the
gold standard in 1897, and the later introduction of export duties. The industrial
impulse of the 1890s was not sustained. Much of the industrial investment had
Paulo Drinot 157
1890 28 9 15 33 1 13 ±
1895 35 7 15 26 1 14 ±
1900 32 7 7 22 18 13 ±
1905 32 7 8 6 10 16 ±
1910 20 14 7 10 18 18 2
1915 26 11 5 5 17 5 10
1920 42 30 2 5 7 1 5
1925 11 32 4 10 8 1 24
1930 11 18 3 4 10 ± 30
At the same time, Peru underwent a shift in external dependence as the USA
became its principal trading partner, a role played by Britain since the early
republican period (Portocarrero, BeltraÂn and Romero, 1992, pp. 152, 161).
Needless to say, both the copper and oil sectors were overwhelmingly con-
trolled by US capital.
Until the 1920s, the state played a marginal part in the processes outlined
above. The Peruvian state was small and weak, and its income depended
almost entirely on import taxes and taxes on tobacco, alcohol, salt and sugar.
For much of the period, export or income taxes were out of the question. By
the time export taxes were introduced, in the mid-late 1910s, international
commodity prices had begun their downward spiral. State expenditure was
limited: in 1914 only 210 000 and 100 000 Peruvian pounds were spent on
education and development respectively, whilst by comparison expenditure on
the army and the police amounted to 477 000 and 314 000 Peruvian pounds
respectively (Portocarrero, 1983, p. 25). The state was particularly weak in
physically outlying areas, such as the highlands of the Puno department,
Cajamarca and Chachapoyas in the north, or the selva, where power was priva-
tized in the hands of gamonales, local political bosses or rubber barons. Fiscal and
monetary policy and the rules of national accounting fit uneasily in these cases.
Taxation on rubber production was difficult if not impossible and the taxation of
exports highly inefficient. In the altiplano, most economic transactions were
carried out in Bolivian currency (Mitre, 1986, p. 51).
In the 1920s, government expenditure, and the general presence of the state,
increased significantly. Revenues increased nearly fourfold between 1919 and
1927, and the number of public employees increased over sixfold between
1920 and 1930 (Collier and Collier, 1991, p. 139). Most of the expenditure went
to finance irrigation works, road building, railways and the modernization and
expansion of Lima. For once, education received some attention: investment
doubled and enrolment rose significantly (Portocarrero and Oliart, 1989, pp.
58±9). These changes resulted in a significant increase in GDP (Seminario and
BeltraÂn, 1998). However, the expansion of the LeguõÂa state was largely financed
by foreign borrowing and subject to high degrees of corruption and fraud;5 local
investment was hardly encouraged, and credit systems remained inadequate. The
creation of the Peruvian Reserve Bank in 1922 did little to alter the bleak financial
situation, but, nevertheless, infrastructural investment, particularly in roads, led
to greater national economic integration and in particular to the demographic
expansion of Lima ± by 1931, nearly 40 per cent of the capital's population
originated in the provinces (Stein, 1980, pp. 56±7). Meanwhile, the growth in
the commercial sector and the state bureaucracy contributed to the emergence of
an urban middle class (Parker, 1998), which, along with the incipient working
class, constituted the basis for the emergence of new political alternatives, and
especially of the Alianza Popular Revolucionaria Americana (APRA) (Stein, 1980).
Paulo Drinot 159
Juliaca, Ayaviri and Sicuani developed around the railway line,7 and Arequipa
merchant houses were thus able to penetrate into the highlands through their
agents, undermining the role previously played by itinerant merchants and
muleteers. The replacement of annual fairs by Sunday fairs further intensified
the pace of the wool economy.
Merchants played a central role, as credit providers for both producers and
traders, in the development of the wool trade and the regional economy.
Indeed, the commercial houses' ventures into wool contributed to the monet-
ization of important sectors of the altiplano and generated a strong demand for
a variety of goods: by 1914±19, the Ricketts house's rural sales as a percentage of
total sales amounted to 50 per cent. Merchants helped articulate a regional
economy that extended beyond Peru's borders. Ricketts bought foodstuffs,
coca, cocoa, coffee and palillo in Cuzco and sold them in Iquique, Antofagasta
and the wool regions of Puno and even Bolivia. These regional products could
account for more than 20 per cent of income. The development of trading
centres along the railway line enhanced this process (Burga and ReaÂtegui, 1981,
pp. 32, 37; Jacobsen, 1988, p. 150).
The wool economy had important linkage effects at the regional level. Fol-
lowing the arrival of the Southern Railway in 1908, the city of Cuzco experi-
enced marked growth, including a degree of industrialization.8 Both the city's
growth and higher demand from wool producers increasingly drawn into the
market stimulated regional agriculture.9 Arequipa merchants were largely
responsible for much of the new investment; as well as providing imports of
all descriptions to the population of Cuzco, from basic tools to foie gras and
champagne, they played a central role in the commercialization of Cuzco's
products. Cuzco's growth, therefore, was a part of Arequipa's growing domin-
ance of the southern economy.10
However, at one level the boom had negative consequences. As wool prices
soared, demand for Cuzco products followed suit. The city of Cuzco was soon
facing drastic food shortages and became strongly demonetized. The sale of
metal currency, indispensable for wool transactions in the highlands, became
highly profitable. `Turco' merchants, minoristas of Arab origin, speculated in the
sale of coins to wool traders. Wool transactions in the highlands could only be
carried out in low denomination coins, usually Bolivian coins; and as a result
the high denomination Peruvian currency was strongly devalued locally.
Attempts first to limit and then stop the exodus of both currency and foodstuffs
failed. On 13 January 1918 Cuzco was shaken by popular protest. With the end
of the war in Europe prices began to drop (GuilleÂn MarroquõÂn, 1989, p. 53;
Burga and ReaÂtegui, 1981, p. 39).
Arequipa's main function in the wool trade was that of an entrepoÃt city. As the
obligatory stopping point between the highlands and the port of Mollendo, the
city inevitably grew. However, devoid of Cuzco's geographically determined
Paulo Drinot 161
natural protection, Arequipa did not industrialize to the same extent as the old
Inca capital. Nevertheless, some factories were erected,11 and Arequipa's growth
stimulated local agricultural production in the city's hinterland.12
The effect of the wool trade on the regional economy was unevenly distrib-
uted; whole departments appear to have been left behind. Ayacucho, stuck
between a south stimulated by wool exports and a central sierra under the
influence of copper, did not participate much in the regional trade flows and
therefore did not benefit from the indirect effects of the south's insertion into
the world economy. The city of Ayacucho exhibited the lowest population
growth rate of all sierra cities between 1876 and 1940. The result appears to
have been the strengthening of landlord power and an increased exploitation
of the peasantry (Gamarra, 1992, pp. 104±6; Degregori, 1990, pp. 29±31; Galdo
Gutierrez, 1992, pp. 107±8).
Perhaps the most important direct social consequence of the wool trade was
the usurpation of peasant land. This process has received the attention of many
scholars and need only be briefly sketched out here. Underlying this process
was gamonalismo.13 The gamonales, often intermediaries in the wool trade, used
local political bosses, or indeed became themselves political bosses, encroach-
ing on community lands, and thus helping to form the avalanche of hacienda
expansion that characterized this period. The number of haciendas, many of
which were created by newcomers, rose from 705 in 1876 to 3219 in 1915.
Correspondingly, the percentage of land held by peasant communities fell from
half during the early republican period to a third by the 1920s ( Jacobsen, 1988,
p. 152).
Pressures internal to the peasantry contributed to the encroachment process.
Jacobsen estimates that the population of communities may have increased by
as much as 50 to 60 per cent between 1876 and 1940; social differentiation
within the peasantry gradually undermined preexisting solidarity mechanisms;
and, finally, the peasantry's increasing insertion into regional commercial cir-
cuits resulted in greater vulnerability ± the possibility of withdrawing into self-
sufficiency in times of crisis was partially eroded. All three factors combined to
facilitate the transfers of land. Independent peasants were incorporated, as
colonos, into expanding haciendas.14 In the early 1920s, however, the avalanche
of hacienda expansion met a solid barrier: peasant rebellion.15
The end of the war in Europe precipitated a major drop in world commodity
prices. The price of Peruvian wool fell by 50 per cent in England and by 80 per
cent in Arequipa between its peak in 1918 and its nadir in 1921. Merchant
houses in Arequipa withdrew from the market, and at the same time, these
merchants acquired land as bankrupt landowners were forced to sell. The
region was plunged into crisis.16 Burga and ReaÂtegui (1981, pp. 34, 48±9) have
estimated that the peasantry's purchasing power fell by as much as 80 per
cent.17 As landowners' incomes from wool fell, the impetus to overexploit the
162 Peru, 1884±1930
haciendas's colonos grew and attempts to extract greater free labour services from
the peasantry broke the delicate balance which the patron±client relationship
was built on and created the conditions for peasant uprisings. By 1923, early
state tolerance and veiled support of peasant claims was replaced by active
repression, either through direct intervention or by allowing landowners to
respond with predictable violence (Kapsoli and ReaÂtegui, 1972; Flores Galindo,
1993b, pp. 365±9; Burga and Flores Galindo, 1994, pp. 177±200).
Some landowners tried to counteract the effect of falling prices. Encouraged
by the success of the sociedades ganaderas in the central sierra, some haciendas in
the south moved towards the introduction of higher yielding sheep varieties,
improvements in grazing and enclosures. However, attempts like the experi-
mental farm at Chuquibambilla were thwarted by the frequent and sometimes
systematic opposition from both the peasantry, for whom rationalization of
production meant the elimination of their huaccho flocks, and Arequipa mer-
chants. Similarly, attempts by the latter to enter production, as in the case of
the Gibson merchant house and the creation of the Sociedad Ganadera del Sur, in
association with Puno landowners, failed as a result of conflicting interests. The
merchants wanted to increase control over alpaca flocks for commercialization,
while the landowners were eager to venture into more capital-intensive sheep
wool production (Bertram, 1977, pp. 8±15).
In summary, the insertion of the south into the world economy produced a
commercial dynamic in the region. Merchant networks spread out of Arequipa
into wool-producing highlands and beyond into the fertile valleys and the ceja
de selva regions of Cuzco and the coastal valleys of Arequipa. Moreover, the
wool economy stimulated urban and industrial growth. These processes
resulted in increasing conflicts over land, aggravated by demographic growth,
and which, despite resistance, typically reduced the peasantry's share. More-
over, the growing dependence on a single export product increased the region's
vulnerability to shocks. Thus, when the price of wool collapsed in the 1920s,
the regional economy was plunged into crisis. Isolated attempts to rationalize
production and increase productivity were resisted by, on the one hand, a
peasantry for whom these changes represented further reductions in income
and, on the other, a regional elite uneasy about the prospects of increased
competition and change in general.
The decentralist movement of the early 1930s became the political expres-
sion of the changes undergone by the south. An alliance between the urban
middle class and progressive landowners, the decentralist movement was com-
posed of those groups most affected by falling wool prices in the 1920s, and for
whom the rising centralism of the LeguõÂa period was an obstacle to regional
autonomous development (Jacobsen, 1988, p. 167; Deustua and ReÂnique, 1984,
pp. 97±113). The decentralists' political programme was full of good intentions
vis-aÁ-vis the emerging proletariats of Cuzco and Arequipa and towards the
Paulo Drinot 163
Indian masses, however, unlike APRA the decentralists did not present them-
selves as an `alliance of intellectual and manual workers'. Their political base
was small and restricted to specific southern interests with little hope of expan-
sion, and although the movement actually held office during the brief Samanez
Ocampo government in 1931, it never really took power.
Coastal agriculture
The Peruvian coast is a long and narrow desert, and agriculture can only take
place in narrow valleys irrigated by rivers flowing from the Andes. Thus,
although both sugar and cotton in Peru benefit from particularly favourable
natural conditions, the successful exploitation of both products has tradition-
ally been conditional on overcoming the constraints imposed by water and
labour shortages.18
In Peru sugar is subject to substantial economies of scale and requires vast
quantities of water. Until the 1920s sugar planters showed little interest in
modern irrigation works, and irrigation laws were generally in their favour
(OreÂ, 1990, pp. 18±20). The modus operandi was to gain access to water rights
through land acquisition, thus unleashing a process of hacienda expansion
beginning in the 1850s. In the nineteenth century sugarcane had been grown
in most valleys in the coast. In our period, however, sugar production became
increasingly concentrated on the north coast, which by 1932 accounted for 83
per cent of production. It was on these plantations, concentrated in the valleys
of La Libertad and Lambayeque departments, that newcomers to the sugar
industry initiated a process of modernization in the 1860s and 1870s. These
men, and some women, often widows, had usually made their fortunes in the
guano and nitrates booms, or were merchants and financiers. Some of these
new planters were recent migrants, like the Italian Larco family, but overall
foreign ownership in the sector was limited.19 Steam ploughs and railways were
imported, as was sugar-making machinery which increased processing capacity
and raised productivity. For the most part, however, milling remained anti-
quated (Albert, 1976, p. 67a). Nevertheless, installed capacity increased just in
time to meet growing international demand.
In particular, two foreign events favoured the development of the industry. In
1893, Chile gave preferential treatment to imports of unrefined raw sugar to
protect its refining industry, and at the same time Peruvian sugar exports did
particularly well during the Cuban Revolution of 1895 and the subsequent
Spanish±American War, following which the USA became a major market for
the first time.20 As Table 6.2 indicates, sugar exports by value rose by 115 per
cent between 1895 and 1900. In the 1900s, however, the adoption of the gold
standard and changing relative prices spelled a downturn for the sugar indus-
try.21 By contrast, the First World War years marked a new boom period for
sugar. As a result of significant investments from 1908, Peru's sugar sector was
164 Peru, 1884±1930
able to respond to the rise in international prices,22 and export growth reached
139 per cent in the period 1910±15.23 The 1920s, again, was a decade of
stagnation. Peru's competitors had also responded to the rising prices in the
mid±late 1910s. By the 1920s, the beet sugar producers had recovered. In the
central and southern valleys, many planters switched to cotton, which until
1925 continued to benefit from relatively good prices as a result of a boll-weevil
plague in the USA in 1921. In the north, where switching to cotton was not
economically feasible, planters attempted to reduce costs to a minimum and
boost production to compensate for falling prices. Smaller mills closed. One
major plantation, Roma, went bankrupt and was bought by another, Casa
Grande. Wage cuts led to industrial unrest: in 1921 a strike at the Roma
plantation resulted in violent repression from a state still shaken by the coun-
try's first general strike two years earlier (Blanchard, 1982, pp. 126±35; Burga
and Flores Galindo, 1994, pp. 235±50).
Indeed, labour played a critical in role in shaping the Peruvian sugar industry.
The abolition of slavery did not itself create labour shortages: sugar planters
complained of `lack of hands' well before emancipation. It is true that many
slaves moved to the cities, particularly Lima, and by 1876 blacks accounted for
9.3 per cent of Lima's population (Cuche, 1975, p. 34). But an emancipated
slave population could hardly have provided the labour to fuel, a sugar export
boom in the second half of the nineteenth century. Independent peasant
communities on the coast could have provided the labour and some were
indeed incorporated via the process of hacienda expansion. However, on the
whole, the coastal peasantry was reluctant and more often than not could not
be forced to leave its land to work in the sugar or cotton plantations (Bell, 1985,
pp. 41±2; GoÂmez and Bazan, 1981, pp. 53, 55). Eventually, the devastation of
the War of the Pacific, deteriorating economic conditions, and particularly the
growth of the internal market and a growing internal differentiation would
draw the peasantry into the export industries. In the 1850s and 1860s, however,
labour was drawn from an unexpected source: China. Between 1849 and 1874,
when British pressure at Macao put an end to the trade, some 90 000 coolies
were brought to Peru. Some were put to work on the guano islands, but the
bulk were destined for the sugar and cotton plantations on the Peruvian coast.
Yet, as with slavery (to which coolie labour bore a close resemblance in terms of
work conditions), the ending of the trade effectively sounded the death knell of
coolie labour. Indentured Chinese labourers had to work off debts over a period
of eight years. Though planters tried a number of strategies to keep the Chinese
on the plantations after the eight years, typically the Chinese left. During the
War of the Pacific, many Chinese fled the plantations, and in some cases
collaborated with the Chilean army (RodrõÂguez Pastor, pp. 95±111; Peloso, pp.
26, 39, Stewart, 1951). In the 1890s, Japanese workers were contracted. Again,
however, this type of labour proved unsuitable to the planters' needs. Like the
Paulo Drinot 165
Chinese, the Japanese left the plantations as soon as they fulfilled their con-
tracts. Moreover, planters complained that the Japanese could only perform
light work. Unaccustomed, or perhaps unwilling to accept the harsh treatment
meted out by plantation overseers, the regular protests by Japanese workers
sometimes escalated into full-blown strikes (Gardiner, 1975; Morimoto, 1979,
Fukumoto, 1997, ch. 3 and 4).
From the 1880s onwards workers from the northern sierra, particularly from
the departments of Cajamarca and the sierra of La Libertad provided the bulk of
labour on sugar estates. A form of labour contracting, known as enganche, was
adopted as the mechanism to draw labour to the coast. Contratistas (often local
merchants or muleteers), hired by the estates, provided peasants with cash
advances in exchange for labour on the coast. Though enganche is often por-
trayed as a demoniac system of oppression, it represented a response to the
needs of both planters and the sierra peasantry. For the peasants, enganche
labour on the coastal sugar plantations represented a seasonal income adjust-
ment. Wages were high by comparison; and the seasonal work could be carried
out between harvests. Moreover, many used enganche to escape from political
turmoil and banditry, conscription into the army, debts or the law. From the
planters' point of view, enganche had a number of advantages, too, not least of
which was the low militancy among workers whose primary interests remained
elsewhere (Gonzales, 1985, pp. 84±95, 118±23; GoÂmez and Bazan, 1981, pp. 53,
59±60; Albert, 1984, pp. 102±11; Blanchard, 1979; KlareÂn, 1973, ch. 2).
Enganche is best understood as a byproduct of the long-established commer-
cial circuits linking the north coast and the northern sierra. From the 1860s
onwards, the coastal economy had had a strong demand effect on two types of
sierra goods: pastoral (cattle) and human (labour). Sierra landowners responded
to the first demand vigorously. Despite poor transport infrastructure, a strong
commercial nexus was established, articulated around a network of merchants
and muleteers. Cattle exports were soon on the rise, and attempts were made to
expand properties and capitalize production, especially through the introduc-
tion of better breeds (Taylor, 1983, pp. 13±17). The local merchants were often
representatives of coastal commercial houses, and typically would export local
foodstuffs, cattle products and some artisan goods to the coast, and import salt,
rice, sugar, agricultural tools and imported textiles. Moreover, some merchants
were also involved in the commercialization of Cajamarca's ceja de selva pro-
ducts, including coca, cocoa, coffee, cotton, rice and aguardiente (RodrõÂguez
Doig, 1986, pp. 24±5; Puga de Losada, 1893, p. 126).
Between the late nineteenth century and the 1940s, haciendas held around 60
to 70 per cent of the land in Cajamarca. However, over 75 per cent of the rural
population lived from subsistence agriculture on its own land. The peasantry
was highly mobile and strongly inserted in the market through artisan produc-
tion. Landowner attempts to access the labour of this free peasantry, either
166 Peru, 1884±1930
to haciendas. In the twentieth century, the local peasantry, with credit from
local mill-owners, was able to engage in production for both regional and
national markets. By 1918 rice had become the principal peasant crop, benefit-
ing from a high import tariff (GoÂmez and BazaÂn, 1981, pp. 36±45). A similar
process occurred in the Jequetepeque Valley (Burga, 1976, pp. 189±95).
On several counts, cotton production varied substantially from sugar. First,
cotton was less circumscribed to a single area, with production taking place in
Piura, the Sur Chico ± Chincha, Ica and Pisco ± and from the 1900s onwards the
ex-sugar producing central valleys of Can Äete and Chancay (Thorp and Bertram,
1978, pp. 44, 57; Albert, 1976, p. 138a; Peloso, 1999, ch. 3; Arroyo, 1981, p. 112;
Pachas Castilla, 1976). Moreover, land tenure in cotton agriculture, which
requires less water and is not subject to economies of scale, has been more
varied, especially during the period under review. Finally, although enganche
labour was a feature of cotton agriculture, it did not predominate nor did it
lead, like in the sugar sector, to the formation of a rural proletariat. Rather,
cotton planters turned to sharecropping, or yanaconaje, in the 1870s. As enter-
ing a sharecropping agreement did not substantially alter the living and work-
ing conditions of the coastal smallholding peasants, many were easily drawn to
cotton cultivation. For the planter, yanaconaje, combined typically with sea-
sonal wage labour, provided a series of benefits including the sharing of risks
and costs of production, and low militancy (Matos Mar, 1976, pp. 40±51; Bell,
1985, p. 45; Peloso, 1999, ch. 6).
In the 1890s and particularly the 1900s, the cotton export boom had less to
do with a disruption to a major world producer, as it had in the 1860s, and more
with the particular qualities of Peruvian cotton, better suited to the fine spin-
ning techniques developed in the 1890s (Bell, 1985, p. 36). In the 1900s,
production and exports rose dramatically as area under cotton increased and
Peru's markets diversified. Much of the increase in production resulted from a
switch to cotton, either from sugar in the central valleys, or vine in Pisco and
maize in Chancay (Bell, 1985, pp. 37±9). As Table 6.1 indicates, exports by
volume rose by 32 per cent and 87 per cent in 1900±05 and 1905±10 respect-
ively. Favourable prices translated these increases into a 132 per cent rise in
exports by value in the period 1905±10. Irrigation projects, on the whole
limited to Piura, did not greatly increase production. On the other hand, the
installation of modern cotton gins in the 1890s appears to have made a more
substantial contribution. Although Britain remained Peru's top market, by
1912±14 Peru had become the USA's fourth major source of supply. In addition,
in the 1890s cotton benefited from the development of a Lima textile industry
and the emergence of markets for its byproducts like cotton seed oil, used for
soap and candle manufacturing, and cotton seed cake used for cattle-feed.
Unlike the sugar sector, cotton offered few avenues for mechanization,
although some innovation did occur. During the First World War, for example,
168 Peru, 1884±1930
some planters switched to Mitafifi cotton which was used for heavy tyres
and aeroplane fabric. More important was the development of a new variety,
introduced in 1912 by a planter from Pisco called FermõÂn TanguõÈs, a Spaniard
born in Puerto Rico, who revolutionized the Peruvian cotton industry. TanguõÈs
cotton was both wilt-resistant and yielded 25 per cent more lint than other
varieties. In 1918 it accounted for eight per cent of Peru's total cotton produc-
tion. By 1931, the figure stood at 91 per cent. In Piura, where TanguõÈs did not
adapt, the introduction of the Pima variety from Arizona had similar con-
sequences. Pima accounted for 5 per cent of Piura's cotton exports in 1925
and 98 per cent in 1930, thus completely displacing the indigenous Aspero
cotton (Cueto and Lossio, 1999; Low, 1979, pp. 146, 148; Thorp and Bertram,
1978, p. 58).
Immigrants like TanguõÈs played an important role in the success of the cotton
sector. Calixto Romero, a Spaniard, reinvested much of the wealth accumulated
in the production and commercialization of Panama hats in the Catacaos
area of Piura into cotton cultivation (Rean Äo and VaÂsquez, 1988, ch. 2 and 3).
In the area around Lima, Italian merchants and petty traders diversified into
cotton, often entering the sector through reclamation of debts (Bonfiglio, 1995,
pp. 65±6). In Chancay, the Japanese came to dominate cultivation both as
landowners and yanacones (Matos Mar, 1976, pp. 82±102; Fukumoto, 1997,
pp. 203±7).
In contrast to the diversification in production, foreign merchant houses like
Duncan Fox, Graham Rowe and Locketts were firmly in control of commercial-
ization. Moreover, the merchants exercised significant influence over the pro-
duction process through their control of the credit system.25 Ginning, too, was
increasingly dominated by these merchants, although until the 1930s most
large haciendas continued to gin their own cotton (Albert, 1976, p. 159a;
Thorp and Bertram, 1978, p. 52).
Like sugar, cotton experienced the boom and bust effect of the First World
War. As Tables 6.2 and 6.3 indicate, cotton exports by value rose by 530 per cent
between 1915 and 1920, while cotton's share within the export quantum rose
from 11 to 30 per cent in the same period. Acreage under cotton grew, as did
productive units and employment in the sector (Albert, 1985, p. 239; Low,
1979, p. 120). Following the introduction of a sliding-scale cotton export tax
in 1915, the cotton sector became a major source of government revenue: by
1919 the tax yielded 21.9 per cent of total government revenue. In the 1920s,
the bust was delayed by the effect on world prices of a boll-weevil plague that
wiped out much of the US crop. Between 1920 and 1925 exports by value only
fell from some US$18 million to US$15 million; but by 1930 they had fallen to
US$ 10 million. Indeed, the recovery of the US cotton crop in 1925 coincided
with floods which wiped out the Peruvian crop. With rock-bottom world prices,
the cotton sector fell into depression, although it would recover in the 1930s.
Paulo Drinot 169
Oil
In the 1900 to 1930 period, oil production was limited to a small area on the far
northern coast.26 Although developed initially in the 1860s, it was not until the
first decade of the twentieth century that the industry really took off. In 1904,
internal demand suddenly rose as both the Central Railway and the Peruvian
Steamship Company switched from coal to oil. At the same time international
demand was rising; the industry expanded rapidly and by the late 1920s oil had
become Peru's leading export.
Oil was produced on three oilfields, Negritos, Lobitos and Zorritos. However,
the bulk of production originated in Negritos, which was owned from 1913
onwards by the International Petroleum Company (IPC), a subsidiary of
New Jersey Standard Oil. Indeed, Negritos' output as a percentage of total
output rose from 59 per cent in 1890 to 63 per cent in the 1900s, 80 per cent
in the 1920s and 83 per cent in the 1930s. Both Lobitos and Zorritos were
forced to enter into working alliances with the IPC. Lobitos, owned by a
British company, exported all its oil through the IPC's refinery at Talara. By
the 1920s exports from Zorritos, owned and exploited by the immigrant
Faustino Piaggio, were insignificant. Rather, it concentrated on satisfying a
growing local demand made highly profitable by a price fixing agreement
with the IPC.
Oil created almost no linkages to the national economy. Indeed,
for the IPC the profits on its sales of products within Peru were sufficient to
meet the entire local-currency cost of production, with the result that the
IPC, despite its role as the single leading export enterprise of the 1920s,
played virtually no part in the actual supply of foreign exchange to the
market.
(Thorp and Bertram, 1978, p. 103)
Moreover, the lack of linkages was a direct result of state policies. In 1922 the
new Petroleum Law imposed a 10 per cent royalty on crude oil output. How-
ever, following negotiations the IPC was exempted from payment of royalties
and a series of land and export taxes in exchange for US$1 million in cash and
assistance in the placing of US$2.5 million in petroleum bonds in the USA. In
other words, the LeguõÂa government exchanged the potential benefits of devel-
oping an oil industry for short-term financial support.27 Such short-termism
was not irrational but corresponded to the time-horizon of Peruvian decision-
makers (Thorp and Bertram, 1978, p. 110).
Given the lack of empirical studies, it is difficult to establish the oil industry's
impact at the local and regional level. In many respects the oilfields were
typical enclaves; workers drawn from nearby villages are likely to have spent
most of their wages within the oil camps. Their typical demands during strikes
170 Peru, 1884±1930
Not surprisingly, they conclude: `On any reasonable assumptions, Peru would
have been better off without the IPC' (Thorp and Bertram, 1978, p. 107).
commercial centre, channelled trade with the Loreto region and Brazil as well
the ceja de selva regions of Huanuco, Huaraz, Pataz and Chachapoyas. However,
trade was limited by prohibitive transport costs (Roux, 1994, p. 157).
Rubber was first used in the manufacture of raincoats and rubber shoes in the
1830s. The economic prospects of the material were transformed in 1839 by the
discovery of the vulcanization process, which made rubber resistant to tem-
perature changes. From this date onwards, world demand for rubber increased
steadily with consumption in Europe and the USA rising from 400 tons in
1850 to between 50 000 and 60 000 tons in 1900 (Thorp and Bertram, 1978,
p. 67). The Amazon area was soon producing around 50 per cent of world
rubber; and the rapid rise in prices generated a rubber-rush, not unlike its
more famous (but less infamous) gold-rush cousins (Roux, 1994, p. 104; Pen-
nano, 1988, p. 110).
As prices rose the logic of rubber changed. Before the 1890s, local and foreign
entrepreneurs had exploited rubber in an artisan fashion. Increasingly, rubber
production came to be dominated by a handful of rubber barons like the
notorious Fitzcarrald and Arana (Pennano, 1988, pp. 157±68). By the 1890s
foreign merchant houses, often with the backing of international banks, had
taken control of commercialization. Finally, the 1900s saw the arrival of large
foreign companies, which attempted to take over at the level of production
(Roux, 1994, p. 114). As Table 6.2 indicates, the changes in the production
process resulted in substantial increases in exports: rubber exports by value rose
115 and 72 per cent in the periods 1900±05 and 1905±10, respectively. In 1898,
in an attempt to regulate and, more important, to tap the benefits from the
boom, the government passed a law granting land concessions to rubber tap-
pers in exchange for a small lease and a royalty on the rubber produced (Thorp
and Bertram, 1978, p. 69). In 1902 the tax amounted to two soles per hectare
(Roux, 1994, p. 115). Taxation at the level of production was made necessary by
the difficulties in taxing the export of rubber. Pierola had introduced export
duties on rubber in 1896, but the region's isolation as well as local officials' high
propensity for corruption made the enforcement of these taxes extremely
difficult and contraband highly enticing: in 1899 only one-third of total rubber
exports were taxed (Roux, 1994, p. 107). A few companies ± like the Inca Rubber
Company and the Peruvian Amazon Company ± took advantage of the new
law. However, these enterprises proved incapable, as Weinstein has suggested,
of introducing a capitalist mode of production in the selva, and breaking the
patron-client relations on which rubber exploitation was based (Weinstein,
1983, pp. 131±2).
The rubber boom gave birth to a highly cosmopolitan urban centre in the
middle of the jungle, Iquitos (San RomaÂn, 1994, p. 150), whose population in
the 1900s stood at around 20 000 (Thorp and Bertram, 1978, p. 68). Iquitos's
high import propensity and opulence is reflected in the level of import tax
172 Peru, 1884±1930
revenues which in 1910 stood at 217 600 Peruvian pounds, up from 167 373 the
previous year (Pennano, 1988, p. 204).
With the rubber boom, a complex commercial circuit was quickly put in
place. The commercialization of rubber occurred at three separate levels. The
first or highest level was that which linked the merchant houses in Iquitos,
Manaus and Belem and the European and US markets. In 1906, 59 per cent of
the rubber exported through Iquitos went to Britain, 37 per cent to France and 3
per cent to the USA. By 1914, however, exports to the USA had overtaken
exports to the European continent (Bonilla, 1977, p. 130; Pennano, 1988,
p. 188). The second level linked the merchant houses and the rubber entrepre-
neurs through the system of aviamiento, whereby the merchant provided the
entrepreneur tools and foodstuffs in exchange for rubber. Finally, the third
linked the rubber patron and the peon, either mestizo enganchados or semi-
enslaved Indians, through the stores owned and run by the rubber entrepren-
eurs, where rubber and goods were exchanged (Roux, 1994, p. 110; Weinstein,
1983, p. 134). At all these junctures, exorbitant profits were made. One final
agent was the itinerant trader, often of Armenian or Jewish extraction, who
undercut the rubber patron's monopoly sales to the peons. For most of the
period trade was funnelled through Iquitos and then the two Brazilian ports,
Manaus and Belem. Peru's failure to develop a merchant fleet in the Amazon
handed the control of river transport to Brazil. The latter's monopoly was only
broken when rubber production expanded to the south-eastern regions of
Ucayali and Madre de Dios, where articulation occurred with the neighbouring
regions of Cuzco and Puno through the Southern Railway.29
Although the bulk of rubber production was undertaken in the Loreto region,
a proportion was also harvested in the ceja de selva regions of the northern,
central and southern sierra and in what became, in 1913, the Madre de Dios
department.30 In 1908 the government ceded selva lands to two foreign com-
panies interested in exploiting the rubber in the jungle regions of the Cuzco
department. The demand for foodstuffs from these and other `plantations'
stimulated foodstuff production in the north of La ConvencioÂn province
(GuilleÂn MarroquõÂn, 1989, pp. 32±3). This coincided with Cuzco's `orientalist'
period when, frustrated by its isolation from the rest of the country, the old Inca
capital turned its attention to the east and the possibility of an Amazonian link
to the Atlantic. The construction of a railway to a navigable point on the
Urubamba River was subject to pressure from the aguardiente producers of
La ConvencioÂn province. The narrow-gauge railway reached Macchu Picchu
in 1930 and La ConvencioÂn's capital, Quillabamba, only recently (Tamayo
Herrera, 1981, pp. 100±1).
The rubber boom ended almost as abruptly as it had started. In the 1880s
Britain established rubber plantations in its Asian protectorates of Ceylon,
Malaysia and Indonesia. Stimulated by rising prices, Asian plantation rubber
Paulo Drinot 173
production rose sharply. By 1910 these plantations were exporting over 8000
tons annually (Thorp and Bertram, 1978, p. 68). Prices plummeted (Pennano,
1998, p. 110), and the higher costs involved in wild-rubber production had
made much of Peruvian rubber uncompetitive by 1912. Plantation rubber came
to dominate the world market: thus, while at the beginning of the 1910s the
Amazon produced over half of the world's supply, by the end of the decade it
produced less than one-tenth (Weinstein, 1983, p. 132). Peru's share of world
rubber production fell from 2.4 per cent in 1910 to 0.05 per cent in 1925
(Pennano, 1988, pp. 122±30). The First World War generated a short-lived
boom, but the industry's days were numbered. As Table 6.2 indicates, rubber
exports by value fell by 87 per cent in 1920±25 and 64 per cent in 1925±30. A
few alternate industries were developed; wood exports tripled between 1923
and 1935, and coffee, cotton and tobacco began to be grown by colonos, or
settlers (San RomaÂn, 1994, pp. 180±2). Politically, the decline of the region
found expression in a failed separatist rebellion in 1922 in Iquitos (Pennano,
1988, p. 168; Basadre, 1983, vol IX, pp. 274±6).
By the late 1910s, Cerro had begun to reassess its system of production, forced
to do so, by several factors. Higher copper exports from Chile and Mexico were
making the copper industry more highly competitive; the mineral content of
the ores from Cerro de Pasco had fallen off sharply; and finally, the continued
acquisition of mining properties had made it necessary to integrate operations
and expand smelting capacity (O'Brien, 1996, pp. 121±2). The company's
response was to build a gigantic smelter at La Oroya, completed in 1922. The
smelter could produce 2500 tons of copper a day, as much as the two smelters
that were closed. The Oroya smelter profoundly changed the rhythm and
rationale of production, and its increased smelting capacity multiplied the
demand for ores. Moreover, the new technology made possible the commercial
production of new metals like lead and zinc.
The transformations, which corresponded to a transition from a manufactur-
ing to an industrial rationale (Mallon, 1983, ch. 6), had significant con-
sequences for the mining workforce. Prior to the construction of the smelter,
Cerro's production regime was based on contractors who employed enganchado
labourers. In the general changes of 1920s, however, contract mining was
curtailed and relegated to peripheral activities. Enganche labour, an expensive
and irregular source of labour, was removed from most operations. The new
production rationale required a stable and skilled labour force, which enganche
could not provide. Typically, enganchados resisted regularization of work-time
and preferred to work for very long shifts in order to pay back their advance and
return to their fields as quickly as possible (Mallon, 1983, p. 210; Flores
Galindo, 1993a, ch. 3). Gradually the company moved to the direct hiring of
labour, which in the 1920s came to represent just below 50 per cent of the
company's labour force.
The ecological disaster created by the fumes of the Oroya smelter pro-
foundly altered Cerro's links with the regional economy. The company omitted
to install filters, and as a result of the ensuing pollution vast amounts of land
were left in an unproductive state and many livestock were killed. Faced with
protests from both communities and landowners, Cerro promptly undertook
the purchase of contaminated land. In the process it accumulated over 200 000
hectares, and in the long run with the installation of filters Cerro benefited
greatly from this. It invested heavily in what became the DivisioÂn Ganadera
(cattle-raising section). It imported cattle and sheep and was able to raise
yields and achieve high levels of meat and wool production for local, national
and international consumption. Moreover, the DivisioÂn Ganadera produced
meat for the mining work force thus permitting the company to keep
wages low.
In conjunction with the gradual phasing out of contract mining and
enganche, the DivisioÂn Ganadera undermined the role of the local petty and
middle commercial class in the articulation of the mining centres with the
Paulo Drinot 177
Conclusion
From many perspectives in the 1884±1930 period Peru was indeed, as the
Italian explorer Antonio Raimondi first noted more than a century ago, a
beggar sitting on a bench of gold. Indeed, Peru's natural resource, or export
commodity, endowment was unashamedly large. The loss of guano and nitrate
revenues following the War of the Pacific was mitigated by the rapid recovery of
other export sectors. In a sense, Peru's loss of nitrates was not as economically
important as Chile's gain of them. Thereafter, a strongly diversified export
quantum, with relatively high growth rates, came to characterize Peru's eco-
nomic development until the 1930s. However, as shown above, the relative
importance of each component within the export quantum changed over the
period. As a result of international price variations, locally-owned sectors with
high returned values lost their pole position to foreign-owned sectors, which in
the case of copper over time strongly reduced its returned value, or in the case
of oil contributed next to nothing to the national economy. These changes
were inscribed within a context of underdeveloped financial systems and
inadequate state management, which contributed to the general reduction in
the export sector's contribution to the national economy. It is therefore quite
natural that, as Portocarrero and Oliart (1989, pp. 108±12) have argued, the
non-establishment view of Peruvian history attributes the inconsistency
between the existence of natural riches and the existence of poverty and under-
development to a combination of bad government and imperialism. However,
this general view of Peru's export experience in the period 1884±1930 is insuffi-
cient, while not entirely false. Peru's export economy was strongly regionally
circumscribed, creating regional and local export economy histories character-
ized by their diversity and complexity.
It is clear from the preceding discussion that Peru's insertion into the world
economy profoundly transformed its regional economies. Each region exhib-
ited specific social, economic and political ± even cultural ± conditions prior to
the export economy period. At the same time, each region was inserted into the
world economy through different export products. Two questions arise from
this. First, to what extent was the nature of change determined by preexisting
regional conditions? And second, to what extent were the changes determined
by the export product itself?
With the exception of oil, all the export products created linkages in their
respective regional economies. In some cases, these linkages were articulated by
a preexisting commercial network, as in the northern sierra. In others, new
networks were created. These either competed with the traditional networks, as
in the south, or coexisted with them, as in the central sierra. The main railways
were built ahead of demand. The Central Railway only became viable through
the increased bulk needs of the copper trade; the Southern Railway played a
Paulo Drinot 179
minor role in the wool trade; but the smaller often private railways in the sugar
sector played a major role in cutting costs and releasing labour for other
activities. However, none of these railways contributed substantially to the
formation of an internal market.
The commercial networks served to stimulate urban growth and incipient
industrialization, oiled by merchant credit systems. Both the wool trade and
cotton production worked on cash advances in exchange for future crops. Non-
economic ties were often created alongside more strictly financial ones. On the
coast, foreign merchants established links with coastal planters through mar-
riage with the latter's daughters.
The implications for the indigenous population varied from region to region.
Changes at the level of the peasant economy were strongly linked to the ra-
tionale imposed by the export economy in each region. In the central sierra and
the coast, copper and sugar contributed to the proletarianization of the peasan-
try, and demographic growth played a central role in this process. The wool
sector, on the other hand, did not see any changes in the relations of production.
Cotton, paradoxically, adopted a precapitalist mode of production, although
sharecroppers became a rural proletariat in every other sense. The transition to
proletarian status resulted in new grievances and new forms of struggle. How-
ever, overall, this transition was largely incomplete at the end of our period.
In physical and capital terms the size of the export sectors and units varied
from region to region. All regions shared a common vulnerability to interna-
tional price fluctuations. The selva's rubber boom ended suddenly when prices
fell. The south, perhaps, remained the least vulnerable sector; its export quan-
tum was somewhat more diversified than that of other regions. On the other
hand, the power structure in the south meant that the effects of cyclical crises
were transmitted most directly to the most vulnerable elements in society. On
the coast, vulnerability was conditional on the degree to which sugar and
cotton dominated in each valley; in those valleys where export crops could be
substituted with domestic-use agriculture, it was possible to weather the worst
effects of the crises. By contrast, in the Chicama Valley the downturn in the
sugar process in the 1920s had far-reaching consequences, contributing to the
emergence of APRA.
To conclude, both preexisting conditions and the export product shaped and
fashioned the insertion of Peru's regions into the world economy. In all regions,
the interplay between the two factors had substantially different results.
Although all regions shared the intrinsic vulnerability represented by depend-
ence on an export product, the effects of this vulnerability were themselves
dependent on the prevailing economic, social and political conditions. The
prevailing conditions set the parameters of each region's insertion into the
world economy and, more important, of its consequences for regional and
national development.
180 Peru, 1884±1930
Notes
1 For comparative export commodity concentration ratios in Latin America see table 3.2
in Bulmer-Thomas (1994, p. 59).
2 Merchants stocked wool during the war period. At the end of the war, total wool
exports rose from 1460 metric tons in 1883 to 5256 metric tons in 1884.
3 By 1891, import items were subject to a 39 per cent duty on average. In 1893, an 8
per cent surcharge on all import duties was added (Thorp and Bertram, 1978, p. 30).
4 The Italian community played a key role in the development of industry. When the
Sociedad Nacional de Industrias was founded in 1896, 25 of the 76 founding members
were Italians (Bonfiglio, 1994, p. 179). On the contribution of Italian immigration to
the Peruvian economy, see also Worrall (1990), Chiaramonte (1983) and Alfaro Valle-
jos and Chueca Posadas (1975).
5 Almost all of the 77 million soles spent on public works were financed by loans
from the USA. The proportion of state revenues deriving from loans rose from around
5 per cent in 1919 to around 45 per cent in 1926±28. By 1931, the total public debt
stood at 500 million soles of which 360 million corresponded to the external debt
(Burga and Flores Galindo, 1994, p. 212; Collier and Collier, 1991, p. 139; Portocarrero,
1983, pp. 23±6).
6 The railway's effect on the structure of trade was greater than its impact on the volume
of trade. Transports costs accounted for a relatively small percentage of the total FOB
value of wool ± perhaps no more than 15 per cent ± and therefore the switch
from llama and mule transport to rail transport represented only a marginal saving
(Jacobsen, 1993, p. 185).
7 A similar commercial circuit was established along the shores of Lake Titicaca follow-
ing the introduction of steam navigation (Tamayo Herrera, 1982, p. 83).
8 By 1905 the city boasted four beer factories and four cocaine factories. In 1913, Italian
immigrants built a hydroelectric plant. The same industrial group took advantage of
the new energy supply to build two textile factories in the city. A flourmill was built in
1921. `Modernization' was a priority for a city that considered itself ± and in many
ways physically remained ± isolated from the rest of the country. In the late nineteenth
century the telegraph was installed, followed by the tramway in the 1900s. The
telephone arrived in 1921. The first cars for public service arrived in 1919. More
significantly for a city separated from the coast by some of the highest mountains in
the world, the first aeroplane arrived in 1921 (Flores Galindo, 1993b, p. 353; Tamayo
Herrera, 1981, pp. 108±9, 130, 144).
9 Three types of agriculture existed in the Cuzco area. The provinces of Cuzco, Quispi-
canchis, Calca and Urubamba were characterized by small haciendas producing corn,
wheat, barley, oats, rye, quinua as well as vegetables, potatoes, fruits and meat, milk and
dairy products. Significantly, these haciendas were for the most part directly exploited
by their owners. The principal markets were the highland region and to a lesser extent
the city of Cuzco. In this way, these haciendas were intimately linked to price fluctua-
tions in the wool markets. The highland areas of Cuzco were characterized by wool-
producing haciendas, exhibiting many of the same features of the Puno haciendas.
Finally, the ceja de selva region, in particular the province of La ConvencioÂn, has been
characterized by Hobsbawm as neo-feudal. These haciendas specialized in the produc-
tion of sugar ( for aguardiente ( and coca, palillo, annatto, woods, fruits, coffee and other
tropical products for the highland markets. Throughout the period, demand for
these products, especially alcohol and coca, remained high (Tamayo Herrera, 1981,
pp. 102±6; GuilleÂn MarroquõÂn, 1989, pp. 127, 130±7; Hobsbawm, 1970).
Paulo Drinot 181
10 However, at least in one respect, Arequipa's hegemony was challenged. Local land-
owners undertook the commercialization of their haciendas' production. By 1910,
three textile factories had been built in the Cuzco hinterland; and cocoa-growing
landowners from the ceja de selva region set up chocolate factories in Cuzco. One
local hacendada, Maria Romainville widow of Latorre, had perhaps the most diversi-
fied `concern' of all. As well as a textile mill in her hacienda in Urcos, she owned
another two haciendas in La ConvencioÂn Valley and one in Quispicanchis. Her
almaceÂn in Cuzco sold textiles, aguardiente, sugar, cocoa, coffee, corn and wheat
(Mo È rner, 1979, p. 28).
11 By 1920, a biscuit factory, a sweet factory, a beer factory, a soap and candle factory, a
shoe factory, as well as a textile factory, dotted the Arequipa landscape (Flores
Galindo, 1993b, pp. 351±2).
12 Unlike Puno or Cuzco, land tenure in most of Arequipa has traditionally been of the
minifundio kind. These minifundios produced foodstuffs, but there appears to have
been a growing amount of land planted with corn for animal fodder. At the same
time, sugar plantations were built in the Tambo Valley on Arequipa's coast and in the
Camana and Majes Valleys viniculture gave over to cotton cultivation (Flores
Galindo, 1993b, pp. 348±51).
13 Manrique (1997, p. 355) defines gamonalismo as `a type of exercise of power and social
structure of domination whose crystallization played a critical role in the accumula-
tion of wool'.
14 According to Jacobsen (1993, pp. 259±90), the transition to colono status did not
significantly alter the peasant family's living and working conditions: the labour
services performed for the landowner replaced those previously performed for local
political authorities. Thus, although occasionally peasants sold land as a result of
pressures of a coercive and sometimes violent kind, more often peasants sold their
land willingly, if reluctantly. Although this is likely to have been the case for the
poorest peasants, for the average peasant the transition to colono status represented a
decline in both economic and social status. Moreover, land was not merely a source of
economic well-being for the Peruvian highland peasant; the links with land were also
spiritual. These elements combine to explain why peasants resisted the expansion of
haciendas.
15 According to Burga and Flores Galindo (1994, ch. 4), to contrast in nineteenth-
century rebellions, the Gran SublevacioÂn del Sur acquired a political millenarian, rather
than merely fiscal context. The `rebels' called for the reimposition of the Inca Empire.
However, recent research indicates that these uprisings, which were far less isolated
and spontaneous than previously assumed, had little to do with millenarianism.
Rather, they were tagged millenarian by local landowners intent on showing that
the Indians were seeking to subvert the national order through calls for the restora-
tion of the Tawantinsuyo and by allying with Indians from Bolivia to establish an
alternative anti-national state (Manrique, 1996, p. 227).
16 One exemption appears to have been domestic-use agriculture, which benefited from
the demand generated by a growing urban population. This situation allowed some
Arequipa merchants to weather the crisis ( Jacobsen, 1988, pp. 161±3).
17 Of course, this is not to imply an oversimplified Pavlovian correlation between
falling incomes and peasant rebellion. As Burga and ReaÂtegui have noted, the Rumi
Maqui rebellion of 1915 occurred during a period of rising wool prices, with corre-
sponding increases in the peasantry's living standards. However, there is little doubt
that by the early 1920s, the altiplano peasantry was undergoing a process of acute
pauperization.
182 Peru, 1884±1930
18 Sugar, which in Peru is particularly resistant to pests, can be cut and processed
throughout the year because its production is independent of rainfall. Moreover, the
indigenous varieties of cotton grown in Peru are high-yielding: cotton growers can
expect to produce two crops a year. In addition, Peruvian cotton is typically of a long
staple variety, a characteristic that commands a premium in international markets.
19 In 1912, foreign ownership accounted for no more than 25 per cent of output. By
1930, this percentage had fallen (Albert, 1985, pp. 240±1; Thorp and Bertram, 1978,
pp. 42±3).
20 Exports to the USA rose from 1299 tons in 1897 to a peak of 58 731 tons in 1901. This
represented 1.2 per cent and 51.2 per cent respectively of Peruvian exports (Albert,
1976, p. 22a).
21 The percentage increase in sugar production was 83 per cent in the period 1889±1900,
and 38 per cent in 1900±10 (Thorp and Bertram, 1978, p. 43).
22 Albert (1976, p. 138a) estimates that some one million pounds sterling were invested
in 1912±14.
23 The downturn in the 1915±20 period is accounted largely by the fall from US$16
million in 1919 to US$5 million in 1920 in sugar exports by value.
24 During the expansionary period of the war years, Casa Grande, like other major
coastal plantations, purchased two highland haciendas in Cajamarca, and later a
rice hacienda in the Jequetepeque Valley. In Cajamarca, cattle rearing was combined
with foodstuff production on a sharecropping basis. Casa Grande was thus able to
produce cheap foodstuffs for its growing plantation labour force, and thus maintain
low wages. Moreover, by requiring that highland hacienda tenants pay off their labour
services on the sugar plantation, it was able to secure cheap peasant labour (Taylor,
1983, p. 34±42; Deere, 1990, pp. 48).
25 The situation changed somewhat in 1918 with the entry of the Banco Italiano into
the market for short-term loans to cotton growers. However, at first the Banco Italiano
restricted its credit facilities to the Italian community (Quiroz, 1993, p. 66).
26 A small oilfield in Puno operated in the first decade of this century (Thorp and
Bertram, 1978, p. 96).
27 On the other hand, Lobitos was not able to secure such concessions from the
government. Its contribution to the Peruvian exchequer was therefore significantly
higher than the IPC's (Miller, 1982b, pp. 421±3).
28 According to Michael Taussig (1987, p. 100), commenting on the well-documented
Putumayo events: `Torture and terror were not simply utilitarian means of produc-
tion; they were a form of life, a mode of production, and in many ways, for many
people, not least of whom were the Indians themselves, its main and consuming
product'. See also Varese (1973, pp. 169±209, 235, 247).
29 In 1912, 112 tons of rubber from the Ucayali River were commercialized through
Iquitos, while 127 tons were sent to the Pacific port of Mollendo (Roux, 1994, pp.
158±68).
30 See Flores Marin (1987, ch. 3) for an account of the workings of the Tambopata
Rubber Syndicate Ltd in the Madre de Dios region.
31 The Mantaro Valley provided flour or unprocessed cereals, as well as meat. The
Huaylas region specialized in sugar and sweets. Huanuco was a provider of fruits,
coca and aguardiente. Wood was brought from Paucartambo, mercury from Huanca-
velica and Chonta, coal from Yanahuanca and ChayacaÂn, and salt from the Salinas de
Huaura and later from San Blas, on Lake JunõÂn (Contreras, 1988, pp. 40±3).
32 In the 1870s, the Sunday fair at Huancayo brought together between 25 000 and
30 000 traders (Long and Roberts, 1984, pp. 32±5; Manrique, 1987, p. 255).
Paulo Drinot 183
33 The company built a 125km railway from Cerro de Pasco to La Oroya and a new
smelter at Tinyahuarco, the largest in Peru. By 1907 the Tinyahuarco smelter was
capable of processing a total of 1000 tons of ore a day, more than five times the
amount processed by the next three largest smelters (Mallon, 1983, p. 171).
34 For a more detailed treatment of these issues see Samaniego (1974).
References
Albert, B. (1985) `External Forces and the Transformation of Peruvian Coastal Agriculture,
1880±1930', in C. Abel and C.M. Lewis (eds), Latin America, Economic Imperialism and the
State: The Political Economy of the External Connection from Independence to the Present,
London/Dover, New Hampshire: The Athlone Press, pp. 231±49.
ÐÐ (1984) `The Creation of a Proletariat on Peru's Coastal Sugar Plantations, 1880±1920'
in B. Munslow and H. Finch (eds), Proletarianisation in the Third World, London: Croom
Helm, pp. 99±120.
ÐÐ (1976) An Essay on the Peruvian Sugar Industry, 1880±1920 and The Letters of Ronald
Gordon, Administrator of The British Sugar Company in Can Ä ete, 1914±1920, Norfold: Uni-
versity of East Anglia.
Alfaro Vallejos, J. and Chueca Posadas, S. (1975) El proceso de hacer la AmeÂrica: Una familia
italiana en el PeruÂ, Lima: Pontificia Universidad CatoÂlica del PeruÂ.
Arroyo, E. (1981) La hacienda costen Äa en el PeruÂ: Mala-Can Äete, 1532±1968, Lima.
Barclay, F. (1989) La Colonia del PereneÂ: Capital ingleÂs y economõÂa capitalista en la configuracioÂn
de la regioÂn de Chanchamayo, Iquitos: Centro de Estudios Teologicos de la Amazonia.
Basadre, J. (1983) Historia de la RepuÂblica del PeruÂ, 1822±1933, 10 vols, Lima: Editorial
Universitaria.
Beals, C. (1934) Fire in the Andes, Philadelphia/London: J.B. Lippincott.
Bell, W.S. (1985) An Essay on the Peruvian Cotton Industry, 1825±1920, working paper no. 6,
Centre for Latin American Studies, Liverpool: University of Liverpool.
Bertram, G.I. (1977) `ModernizacioÂn y cambio en la industria lanera del PeruÂ, 1919±1930:
un caso frustrado de desarollo', Apuntes 6, pp. 3±22.
Blanchard, P. (1982) The Origins of the Peruvian Labor Movement, 1883±1919. Pittsburgh:
University of Pittsburgh Press.
Blanchard, P. (1979) `The Recruitment of Workers in the Peruvian Sierra at the Turn of the
Century: The Enganche System', Inter-American Economic Affairs, vol. 33, pp. 63±83.
Bonfiglio, G. (1995) `Los italianos en Lima', in A. Panfichi and F. Portocarrero (eds),
Mundos Interiores: Lima, 1850±1950, Lima: Universidad del PacõÂfico, pp. 43±73.
ÐÐ (1994) Los italianos en la sociedad peruana: Una visioÂn histoÂrica, Lima: SAYWA edi-
ciones.
Bonilla, H. (1980) `Guano y crisis en el Peru del XIX' in Nueva Historia General del PeruÂ,
Lima: Mosca azul editores.
ÐÐ (1977) `El caucho y la economõÂa del oriente peruano', in Los mecanismos de un control
econoÂmico, (Vol. V), Instituto de Estudios Peruanos/Fondo del Libro del Banco Industrial
del PeruÂ, pp. 123±34.
ÐÐ (1974) Guano y burguesõÂa en el PeruÂ, Lima: Instituto de Estudios Peruanos.
Bulmer-Thomas, V. (1994) The Economic History of Latin America since Independence, Cam-
bridge: Cambridge University Press.
Burga, M. (1976) De la encomienda a la hacienda capitalista: El valle del Jequetepeque del siglo
XVI al XX , Lima: Instituto de Estudios Peruanos.
184 Peru, 1884±1930
Burga, M. and ReaÂtegui, W. (1981) Lanas y capital mercantil en el sur: La Casa Ricketts,
1895±1935, Lima: Instituto de Estudios Peruanos.
Burga, M. and Flores Galindo, A. (1994) Apogeo y crisis de la RepuÂblica AristocraÂtica, in
A. Flores Galindo, Obras Completas Vol. II, Lima: FundacioÂn Andina/SUR Casa de
Estudios del Socialismo.
Chiaramonte, G. (1983) `La migracioÂn italiana en AmeÂrica Latina. El caso peruano',
Apuntes, vol. 13, pp. 15±36.
Collier, D. and Collier, R.B. (1991) Shaping the Political Arena, Princeton: Princeton Uni-
versity Press.
Consejo Provincial de Paita (1935) Memoria leõÂda por el Alcalde Titular Sen Ä or Rodolfo Alfa-
geme en la sesioÂn de Concejo celebrada el 2 de enero de 1935, dando cuenta de la marcha de la
InstitucioÂn durante el an
Ä o anterior, Paita: Imprenta Mercantil.
Contreras, C. (1988) Mineros y campesinos en los Andes: Mercado laboral y economõÂa cam-
pesina en la sierra central, siglo XIX, Lima: Instituto de Estudios Peruanos.
Cuche, D. (1975) Poder blanco y resistencia negra en el PeruÂ: un estudio de la condicioÂn social
del negro en el Peru despueÂs de la abolicioÂn de la esclavitud, Lima: Instituto Nacional de
Cultura.
Cueto, M. (1992) `Sanitation from Above: Yellow Fever and Foreign Intervention in Peru,
1919±1922', Hispamic American Historical Review HAHR, vol. 72(1), pp. 1±22.
Cueto, M. and Lossio, J. (1999) InovacioÂn en la agricultura: FermõÂn Tanguis y el algodoÂn en el
PeruÂ, Lima: Universidad del PacõÂfico/Cosapi OrganizacioÂn Empresarial.
Davila Apolo, D. (1976) Talara, los petroleros y la huelga de 1931, Memoria para optar al
grado de Bachiller en Ciencias Sociales con mencioÂn en SociologõÂa, Lima: Pontificia
Universidad CatoÂlica del PeruÂ.
Deere, C.D. (1990) Household and Class Relations: Peasants and Landlords in Northern Peru,
Berkeley: University of California Press.
Degregori, C.I. (1990) El surgimiento de Sendero Luminoso: Del movimiento por la gratuidad de
la ensenÄ anza al inicio de la lucha armada, Ayacucho 1969±1979, Lima: Instituto de Estu-
dios Peruanos.
Deustua, J. and ReÂnique, J.L. (1984) Intelectuales, indigenismo y descentralismo en PeruÂ,
1897±1931, Cuzco: Bartolome de Las Casas.
Espinosa Bravo, C.A. (1936) Facetas de Jauja, Jauja: C.E. Sanguinetti.
Flores Galindo, A., (1993a), `Los mineros de Cerro de Pasco, 1900±1930 (Un intento de
caracterizacioÂn social)', in A. Flores Galindo, Obras Completas, (Vol. I), Lima: FundacioÂn
Andina/SUR Casa de Estudios del Socialismo.
ÐÐ (1993b) `Arequipa y el sur andino: Ensayo de historia regional, Siglos XVIII±XX', in
A. Flores Galindo, Obras Completas, (Vol. I), Lima: FundacioÂn Andina/SUR Casa de
Estudios del Socialismo.
Flores Marin, J.A. (1987) La explotacioÂn del caucho en el PeruÂ, Lima: Consejo de Ciencia y
Technoloqia del Peru CONCYTEC.
Fukumoto, M. (1997) Hacia un nuevo sol: Japoneses y sus descendientes en el PeruÂ, Lima:
AsociacioÂn Peruano Japonesa del PeruÂ.
Galdo Gutierrez, V. (1992) Ayacucho: Conflictos y Pobreza ± Historia Regional (siglos
XVI±XIX), Ayacucho: Universidad Nacional San CristoÂbal de Huamanga.
Gamarra, J. (1992) `Estado, modernidad y sociedad regional: Ayacucho, 1920±1940',
Apuntes, vol. 31, pp. 103±14.
Gardiner, C.H. (1975) The Japanese and Peru, 1873±1973, Albuquerque, University of New
Mexico Press.
Gilbert, D.L. (1982) La oligarquõÂa peruana: Historia de tres familias, Lima: Editorial Hori-
zonte.
Paulo Drinot 185
GoÂmez, J. and BazaÂn, I. (1989) Capitalismo y formacioÂn regional: Chiclayo entre los siglos XIX
y XX, Chiclayo: Poblacio n y Desarollo/Instituto de Investigacio n y CapacitacioÂn.
ÐÐ (1981) Capitalismo y regioÂn en Lambayeque, 1879±1930, Chiclayo: Gaviota Roja
Editores.
Gonzales, M. (1985) Plantation Agriculture and Social Control in Northern Peru, 1875±1933,
Austin: University of Texas Press.
Gootenberg, P. (1993) Imagining Development: Economic Ideas in Peru's `Fictitious Prosperity'
of Guano, 1840±1880, Berkeley: University of California Press.
Gootenberg, P. 1989, Between Silver and Guano: Commercial Policy and the State in Postinde-
pendence Peru, Princeton: Princeton University Press.
GuilleÂn MarroquõÂn, J. (1989) La economõÂa agraria del Cusco, 1900±1980, Cusco: Centro de
Estudios Rurales Andinos Bartolome de las Casas.
Hobsbawm, E. (1970) `A Case of Neo-Feudalism: La ConvencioÂn, Peru', Journal of Latin
American Studies, vol. 1(1), pp. 31±50.
Hunt, S. (1985) `Growth and Guano in Nineteenth Century Peru', in R. CorteÂs Conde, and
S. Hunt (eds), The Latin American Economies: Growth and the Export Sector, 1880±1930,
New York/London: Holmes and Meier, pp. 255±318.
Jacobsen, N. (1993) Mirages of Transition: The Peruvian Altiplano, 1780±1930, Berkeley and
Los Angeles: University of California Press.
ÐÐ (1988) `Free Trade, Regional Elites, and the Internal Market in Southern Peru,
1895±1932', in J. Love and N. Jacobsen (eds), Guiding the Invisible Hand: Economic
Liberalism and the State in Latin American History, Westport: Praeger, pp. 145±76.
Kapsoli, W. and ReaÂtegui, W. (1972) El campesinado peruano: 1919±1930, Lima: Universi-
dad Nacional Mayor de San Marcos.
KlareÂn, P. (1973) Modernization, Dislocation and Aprismo: Origins of the Peruvian Aprista
Party, 1870±1932, Austin/London: University of Texas Press.
Kruijt, D. and Vellinga, M. (1979) Labor Relations and Multinational Corporations: The Cerro
de Pasco Corporation in Peru, 1902±1974, Assen: Van Gorcum.
Lockhart, J. (1974) Spanish Peru, 1532±1560: A Colonial Society, Madison/London: Univer-
sity of Wisconsin Press.
Long, N. and Roberts, B. (1984) Miners, Peasants and Entrepreneurs: Regional Development in
the Central Highlands of Peru, Cambridge: Cambridge University Press.
LoÂpez-Ocon Cabrera (1995) `El nacionalismo y los orõÂgenes de la Sociedad GeograÂfica de
Lima', in M. Cueto (ed.), Saberes Andinos: Ciencia y tecnologõÂa en Bolivia, Ecuador y PeruÂ,
Lima: Instituto de Estudio Peruanos.
Low, A. (1979) Agro-Exporters as Entrepreneurs: Peruvian Sugar and Cotton Producers,
1890±1945, doctoral thesis, University of Oxford.
Lumbreras, L. (1980) `El Peru prehispaÂnico', in Nueva Historia General del PeruÂ, 2nd edn,
Lima: Mosca Azul.
Mallon, F. (1983) The Defense of Community In Peru's Central Highlands: Peasant Struggle and
Capitalist Transition, 1860±1940, Princeton: Princeton University Press.
Manrique, N. (1997) Circuitos mercantiles y acumulacioÂn en escala regional: Las provincias
altas de Arequipa, siglos XVIII-XX, manuscript.
ÐÐ (1996) Historia de la RepuÂblica, Lima: Cofide.
ÐÐ (1988) Yawar Mayu: Sociedades terratenientes serranas, 1879±1910, Lima: Instituto
FranceÂs de Estudios Andinos/DESCO.
ÐÐ (1987, Mercado interno y reÂgion: La sierra central, 1820±1930, Lima: Centro de Estudios
Teologicos de la Amazonia.
Matos Mar, J. (1976) Yanaconaje y reforma agraria en el PeruÂ: El caso del valle de Chanca,
Lima: Instituto de Estudios Peruanos.
186 Peru, 1884±1930
Mc Evoy, C. (1994) Un proyecto nacional en el siglo XIX: Manuel Pardo y su visioÂn del PeruÂ,
Lima: Pontificia Universidad Cato lica del Peru .
Miller, R. (1983) `The Grace Contract, The Peruvian Corporation and Peruvian History'.
Ibero-Amerikanisches Archiv vol. 9(3/4), pp. 319±48.
ÐÐ (1982a) `The Coastal Elite and Peruvian Politics, 1895±1919', Journal of Latin American
Studies, vol. 14(1), pp. 97±120.
ÐÐ (1982b) `Small Business in the Peruvian Oil Industry: Lobitos Oilfields Limited Before
1934', Business History Review, vol. LVI(3), pp. 400±23.
ÐÐ (1982c) `The Wool Trade of Southern Peru, 1850±1915', Ibero-Amerikanisches Archiv,
vol. 8(3), pp. 297±311.
ÐÐ (1976) `The Making of the Grace Contract: British Bondholders and the Peruvian
Government, 1885±1890', Journal of Latin American Studies, vol. 8(1), pp. 73±100.
ÐÐ (c. 1976) `Railways and Economic Development in Central Peru, 1890±1930', in
R. Miller, C.T. Smith and J. Fisher (eds), Social and Economic Change in Modern Peru,
Liverpool: University of Liverpool, pp. 27±52.
Mitre, A. (1986) El monedero de los Andes. RegioÂn econoÂmica y moneda boliviana en el siglo
XIX, La Paz: Hisbol.
Morimoto, A. (1979) Los inmigrantes japoneses en el PeruÂ, Lima: Universidad Nacional
Agraria.
MoÈ rner, M. (1979) Notas sobre el comercio y los comerciantes del Cuzco desde fines de la colonia
hasta 1930, Lima: Instituto de Estudios Peruanos.
Nugent, D. (1997) Modernity at the Edge of Empire: State, Individual, and Nation in the
Northern Peruvian Andes, 1885±1935, Stanford: Stanford University Press.
O'Brien, T.F. (1996) The Revolutionary Mission: American Enterprise in Latin America,
1900±1945, Cambridge: Cambridge University Press.
OreÂ, T. (1990) Riego y organizacioÂn: EvolucioÂn histoÂrica y experiencias actuales en el PeruÂ, Lima:
Intermediate technology Development Group.
Pachas Castilla, R. (1976) EconomõÂa y sociedad en el valle de Chincha: 1860±1918, Lima:
Universidad Nacional Mayor de San Marcos.
Parker, D. (1998) The Idea of the Middle Class: White-Collar Workers and Peruvian Society,
1900±1950, University Park, Penn.: The Pennsylvania State University Press.
Peloso, V. (1999) Peasants on Plantations: Subaltern Strategies of Labor and Resistance in the
Pisco Valley, Peru, Durham/London: Duke University Press.
Pennano, G. (1988) La economõÂa del caucho, Iquitos: Centro de Estudios Teologicos de la
Amazonia.
Portocarrero, F. (1995) El Imperio Prado: 1890±1970, Lima: Universidad del PacõÂfico.
Portocarrero, F. and TorrejoÂn, L. (1992) Las inversiones en valores nacionales de la eÂlite
econoÂmica, PeruÂ: 1916±1932, Lima: Universidad del PacõÂfico/Consorcio de InvestigacioÂn
EconoÂmica.
Portocarrero, F, BeltraÂn, A. and Romero, M.E. (1992) Compendio estadõÂstico del PeruÂ,
1900±1990, Lima: Universidad del PacõÂfico/Consorcio de InvestigacioÂn EconoÂmica.
Portocarrero, G. (1983) De Bustamante a OdrõÂa: El fracaso del Frente DemocraÂtico Nacional,
1945±1950, Lima: Mosca Azul Editores.
Portocarrero, G. and Oliart, P. (1989) El Peru desde la escuela, Lima: Insituto de Apoyo
Agrario.
Puga de Losada, A., (1893), Ensayos literarios, Lima: Imp. de la IlustracioÂn Sud-Americana.
Quiroz, A. W. (1993) Domestic and Foreign Finance in Modern Peru, 1850±1950: Financing
Visions of Development, London: St Antony's/Macmillan.
Quiroz, A.W. (1988) Financial Leadership and the Formation of Peruvian Elite Groups,
1884±1930, Journal of Latin American Studies, vol. 20, pp. 49±81.
Paulo Drinot 187
ReanÄ o, G. and VaÂsquez, E. (1988) El Grupo Romero: Del algodoÂn a la banca, Lima: Universi-
dad del PacõÂfico/Centro de Investigacio n y Promocio n del Campesinado.
Roberts, B. (c. 1976) `The Social History of a Provincial Town: Huancayo, 1890±1972', in
R. Miller, C.T. Smith, and J. Fisher (eds), Social and Economic Change in Modern Peru,
Liverpool: University of Liverpool, pp. 130±97.
RodrõÂguez Doig, E. (1986) El Camino de los Enganchados, Lima: AsociacioÂn Peruana para el
Fomento de las Ciencias Sociales.
RodrõÂguez Pastor, H. (1989) Hijos del Celeste Imperio en el Peru (1850±1900): MigracioÂn,
agricultura, mentalidad y explotacioÂn, Lima: Instituto de Apoyo Agrario.
Roux, J.C. (1994) L'Amazonie peÂruvienne: Un Eldorado deÂvore par la foreÃt, 1821±1910, Paris:
L'Harmattan.
Samaniego, C. (1974) Location, Social Differentiation and Peasant Movements in the Central
Sierra of Peru, doctoral dissertation, University of Manchester.
San RomaÂn, J. (1994) Perfiles histoÂricos de la AmazonõÂa peruana: Iquitos, Centro de Estudios
TeoloÂqicos de la Amazonia, Centro AmazoÂnico de Antropologia e AplicacioÂn Pinctica,
Instituto de Investigacioes de la Amazonia Peruana.
Seminario, B. and Beltran, A. (1998) Crecimiento econoÂmico en el PeruÂ: nuevas evidencias
estadõÂsticas, Lima: Universidad del PacõÂfico.
Stein, S. (1980) Populism in Peru, The Emergence of the Masses and the Politics of Social
Control, Madison: The University of Wisconsin Press.
Stewart, W. (1951) Chinese Bondage in Peru, Durham: Duke University Press.
Tamayo Herrera, J. 1981, Historia social del Cuzco republicano, Lima: Editorial Universo.
Taussig, M. (1987) Shamanism, Colonialism, and the Wild Man: A Study in Terror and Healing,
Chicago/London: The University of Chicago Press.
Taylor, L. (1986a) Bandits and Politics in Peru: Landlord and Peasant Violence in Hualgayoc,
1900±1930, Cambridge: Centre of Latin American Studies.
ÐÐ (1986b) Estates, Freeholders and Peasant Communities in Cajamarca, 1876±1972, work-
ing paper no. 42, Centre of Latin American Studies, Cambridge: University of Cam-
bridge.
ÐÐ (1983) Cambios capitalistas en las haciendas cajamarquinas, 1900±1935, working paper
no. 39, Centre of Latin American Studies, Cambridge: University of Cambridge.
Thorp, R. and Bertram, G. (1978) Peru, 1890±1977: Growth and Policy in an Open Economy,
London: Macmillan.
Varese, S. (1973) La sal de los cerros: Una aproximacioÂn al mundo Campa, Lima: Retablo de
Papel Ediciones.
Weinstein, B. (1983) `Capital Penetration and Problems of Labour Control in the Amazon
Rubber Trade', Radical History Review, vol. 27, pp. 121±43.
Werlich, D. (1978) Peru, A Short History, Illinois: Southern Illinois Press/ Carbondale and
Edwardsville.
Wilson, F. (1987) `The Conflict between Indigenous and Immigrant Commercial Systems
in the Peruvian Central Sierra, 1900±1940', in R. Miller (ed.), Region and Class in Modern
Peruvian History, Liverpool: University of Liverpool, pp. 125±62.
Worrall, J.E. (1990) La inmigracioÂn italiana en el PeruÂ, 1860±1914, Lima: Instituto Nacional
de Cultura.
Yepes del Castillo, E. (1972) Peru, 1820±1920: Un siglo de desarrollo capitalista, Lima:
Instituto de Estudios Peruanos/CampodoÂnico ediciones.
7
Bolivia, 1900±39: Mining, Railways and
Education
Manuel E. Contreras
Introduction
During the first 40 years of this century, the Bolivian economy was a classic
example of rapid export-led growth, based on tin exports. This new export
expansion consolidated Bolivia's entrance into the international market in
the last third of the nineteenth century with the export of silver. However,
Bolivia's export base at the onset of the twentieth century was more diversified
than it was to be by the end of the 1930s. In 1900, silver represented 36 per cent
of Bolivia's total exports, rubber represented 26 per cent, while the tin industry
was responsible for the remaining 24 per cent. The relative importance of tin
surpassed the importance of silver and rubber by 1903. Silver exports decreased
significantly from 1905 onwards. Rubber exports remained stable until 1910,
when they accounted for 37 per cent of total exports, but thereafter decreased,
reaching 5 per cent in 1920.1
Yet it is tin that has caught the attention of scholars both within and outside
Bolivia. According to Almaraz (1976): `The twentieth century [in Bolivia]
arrived on the shoulders of the tin industry', and the destiny of this mineral
has marked Bolivia's development until recently. Since 1910, tin has repres-
ented more than 50 per cent of the value of total exports of the country, and
since 1930 it has represented 70 per cent. Of the remaining 30 per cent, two-
thirds has been composed of other minerals including silver, tungsten, anti-
mony, copper, lead, zinc, bismuth and sulphur, while agricultural and farming
products such as rubber, coca and cattle have represented only 8 per cent
(GutieÂrrez Guerra, 1940, pp. 6, 23, 26±7). Tin formed the economic base for
Bolivia's growth and modernization, and it was the exploitation of this mineral
that both attracted the greatest amount of investment and paid the highest
taxes. In order to maximize the exploitation of tin, an extensive infrastructure
including railways and roads was constructed. At the same time, public services
such as drinking water, sewage systems, electricity, health and public education
188
Manuel E. Contreras 189
were created. Finally, new financial institutions were established, including the
Banco Nacional de Bolivia in 1911, the Banco Central, and the General Comp-
troller of the Republic in 1929 and 1930, respectively. The military was profes-
sionalized and an important process of urbanization and modernization took
place.
In 1900 Bolivia had a population of 1.8 million. By 1905 this had been
reduced to 1.7 million as a result of the territories recently conceded to Brazil
and Chile. At the time the country had 1.19 persons per square kilometre,
placing Bolivia last in population density in South America, while occupying
third place in territorial expansion and seventh place in total population (Ofi-
cina Nacional de InmigracioÂn, EstadõÂsticas y Propaganda GeograÂfica, 1905, pp.
235±7). Due to its mountainous topography, Bolivia was, in addition, a land
with little internal communication and where access to many of the country's
areas was quite difficult. The construction of communication systems, there-
fore, was vital for the growth of the export sector and its general development
and modernization.
We will study the evolution of the Bolivian economy between 1900 and 1939
by examining the development of the tin mining industry. We will attempt to
weave together the political economy of the export sector and the association
between the mining industry and the other productive sectors. In addition, this
chapter will study the effects of the Great Depression (1929±33) and the Chaco
War (1932±35) on the tin market as well as on the relation between the Bolivian
state and the export sector through fiscal and exchange policy. The evolution of
the tax structure and its significance in the Bolivian economy represents a vital
aspect in understanding the development of the state (Gallo, 1990).
The chapter also briefly analyses the construction of railways and road sys-
tems and the establishment of the national education system. These develop-
ments reflect the process of state consolidation that was carried out during this
period of `export-led growth'. Railways and education were part of the liberal
legacy ± the Liberals governed Bolivia between 1899 and 1929 ± and were
effectively financed by the export sector. Railways were not only essential
for mining, they also formed part of the liberals' modernizing project.
Construction of roads replaced railways and constituted, at first glance, the
government's response to the 1930s Depression. Although the connection of
education with the export sector was less direct, it reflected the positivist ideals
of the Liberals and the need to establish a national education system, a goal
unfulfilled since the nineteenth century. As such, the export sector represented
perhaps the most important element in the modernization of the country as it
established the foundations of the Bolivian education system.
Indeed, the main argument of this chapter is that the export sector's
resources allowed investment in both physical infrastructure and human capi-
tal, factors now recognized as central to development. The `export-led growth'
190 Bolivia, 1900±39
model, in its expansive phase and in its moments of crisis, had a significant
internal impact: it effectively established the foundations on which future
development would be built. The chapter maintains that the various govern-
ments' capacity to extract the surplus coming from the mining sector was
greater than is traditionally accepted. This chapter also analyses the variations
in the mechanisms of financial extraction by the state and argues that at those
times when considerable resource accumulation took place, either at the gov-
ernment level or at the level of companies, other factors helped determine its
impact on the country's development. In the case of the government, military
expenditure, investment in non-productive projects, and corruption played a
dominant role in minimizing the potential impact of such capital accumula-
tion. However, in the case of the mining companies, the ineffective use of
resources was primarily due to the absence of opportunities and conditions
for productive investment.
The chapter is divided into six parts. Following this introduction we analyse
the chronological development of the tin mining industry. We then focus on
the influence of the mining industry on the national economy through fiscal
policies, followed by a summary and synthesis of the development of railways
and roads. After this we describe the evolution of the education system and the
efforts towards modernization during this period, and in the final part review
the development of the export economy, emphasising the role of the state in
relation to the interests of miners and their efforts towards modernization
during this period.
Table 7.1 Bolivia: tin exports in four decades; annual averages, maximum, minimum,
and annual rates of growth, 1900±39
Source: Calculated from export data in Walter GoÂmez-D'Angelo (1978), pp. 218±20.
expansion benefited from the industry's capacity to resolve the problems posed
by internal transportation, the high cost of fuel, and an insufficiently large
technical and managerial workforce (Contreras, 1994).
The methods of production during this first period (1900±15) were primitive
and not unlike those utilized during the eighteenth and nineteenth centuries.
With the exception of certain mining companies, exploitation was not sys-
tematic: there were no geological maps, and production was undertaken with-
out prior planning, exploiting only the richest and most accessible veins while
using as little capital as possible. This unsystematic process often resulted in the
deterioration of the ore deposits' economic value, as a great percentage of
the mines' potential was ignored. The rise in prices, the advent of railroads,
and the increasing availability of capital led to large-scale imports of machinery
and, from 1908 onwards, concentration mills (Contreras, 1980, pp. 27±9).
At first, merchants were the main source of capital, mostly advances and
credits. However, these sources were able to satisfy the demand for capital only
in the short term. Given that the financial participation of local banks was
marginal, investment capital was sought internationally. After 1906 ± following
the Peace Treaty with Chile, which solved the territorial problem ± there was a
significant flow of Chilean capital into the industry, thereby increasing the
number of companies that were floated in the Santiago Stock Exchange. The
main Chilean purchase, the Llallagua mine (the richest in Bolivia), was made in
1906. In 1912, investment from Chile reached £2.6 million. By 1913, foreign
capital exceeded £10 million, composed primarily of Chilean, British and, to a
lesser extent, US capital (Contreras 1980, pp. 33±7).3
The capitalization of mining made possible the concentration of production
by a group of `modern' companies characterized by a high level of mechaniza-
tion in the exploitation and concentration of tin, modern administration,
greater access to working capital, and the direct selling of minerals. This
group accounted for 55 per cent of total tin production in 1912, and 65 per
cent by 1917 (Contreras, 1984, p. 14).
192 Bolivia, 1900±39
Table 7.2 Bolivia: exports and imports by country of destination (% of total value)
(a) Exports
(b) Imports
Sources: 1900: Oficina Nacional de InmigracioÂn, EstadõÂsticas y Propaganda GeograÂfica (1905), pp.
457, 460±1; 1913: Victor Bulmer-Thomas (1994), pp. 74, 76; 1920: LeoÂn E. Bieber (1984), p. 96; 1930:
DireccioÂn General de EstadõÂstica (1936); and 1939: Antonio Mitre (1996), p. 119.
194 Bolivia, 1900±39
Following the War, the saturation of tin stocks in the US led the government
to prohibit imports until 1919, significantly affecting Bolivian mining. Once
this problem was overcome, tin mining continued to expand, as the short post-
war boom materialised.
The 1920s
After reaching record levels in 1920, tin prices began to fall in step with the
world recession and growing tin stocks (Fox, 1974, p. 115). The depression hit
both the mining industry and the economy at large: in 1921 tin exports
plummeted to 19.086 FMT, down from the previous high of 29 280 FMT in
1918 (see Table 7.1).
Once this crisis was overcome, from 1924 to 1928 tin experienced substantial
growth, with both consumption and production booming. Production rose
throughout the period: in 1929 a record 47 191 FMT were exported, and during
the 1920s Bolivian production came to represent 18 per cent (1921) and 24 per
cent (1929) of world production (LoÂpez Rivas, 1955, p. 115). Between 1920 and
1929, exports increased at an annual rate of 5.9 per cent, surpassing the rate
achieved in the previous period (1910±19, see Table 7.1).
This growth was directly linked to huge investment flows. Between 1920 and
1926, investment in new operations and in the capitalization of existing mines
amounted to US$27 million. Of this total sum, US$12 million stemmed from
the investments of the Guggenheim Brothers in the Caracoles Tin Company in
1920 (Ness, 1938, table XIV, p. 130, and appendix IV, p. 274). In 1925 this
company was one of the most important producers in Bolivia (5960 gross tons),
second only to the Patin Äo Mines (18 827 gross tons) (Mitre, 1993, table 19, p.
116). The Guggenheim Brothers were equally interested in acquiring other
companies, including those of Aramayo in which their investment reached
almost one million dollars invested in 70 000 shares of Aramayo Company
stocks.5 Finally, in 1928, the Guggenheim Brothers reached an agreement with
Aramayo in exchange for a share in the profits. This contract lasted only until
1933, when Aramayo took advantage of the tin crisis and the heavy losses in
Caracoles to buy the mine from the Guggenheim Brothers (Crespo, 1981, pp.
247±8).
The continued growth of the tin mining industry increased its importance in
the national economy, leading to disputes about the contrast between the
relative boom of this sector and the state of crisis of the country. Between
1918 and 1925 there were many criticisms in Congress over the outflow of
capital from the mining sector, criticisms which went as far as suggesting the
need to `nationalize Bolivian mining' and recommending the direct participa-
tion of the state in the production companies (Prado, 1988, pp. 70±1).
In 1925, the mining industry formed the Bolivian Association of Industrial
Miners, which grouped together the principal mining companies along with tin
Manuel E. Contreras 195
1930 and 1939, tin exports decreased at 3.6 per cent a year, reaching in 1933 a
level of exports similar to 1901±03 (see Table 7.1). Bolivian production fell to 12
per cent of world production in 1937 ± its lowest level since 1904 ± to finish the
decade with no more than 17 per cent (Lopez Rivas, 1955, p. 107).
Towards the end of 1929, mining interests demanded a reduction in railroad
cargo freights and the freedom to readjust salaries. In 1930, the AsociacioÂn de
Industriales Mineros formally requested the end of state intervention in wage
levels, a reduction in railroad freights for mineral exports and imports of inputs,
the creation of the Caja de Seguros e Indemnizaciones in order to alleviate the
burden of social benefits on companies, and the devaluation of the national
currency. The government of Hernando Siles (1926±30) had pledged to support
these measures, but was prevented from carrying them out by its overthrow by
the Military Union in 1930 (Lofstrom, 1972, pp. 63±4).
The Depression led to important changes in monetary and fiscal policy. In
1931, after Great Britain had abandoned the gold standard, Bolivia followed
suit and in addition suspended external debt payments. It was now necessary
for the government to fix the exchange rate. At the same time, given the lack of
foreign loans, it had to find a way of solving its constant budget deficits.
Moreover, as we will see below, the need to finance the Chaco War was an
additional stimulus for an increased participation by the state in the national
economy. In both cases the tin mining industry played an important role and
its behaviour was subject to public scrutiny. The rate at which the exchange rate
would be fixed was a source of conflict between the miners and both importers
and industrialists. Importers were keen on sticking to a parity exchange rate
and lobbied for an adequate provision of foreign exchange. Industrialists,
naturally, would have benefited from a devaluation. However, they also wanted
adequate provision of foreign exchange in order to cover their import needs.
The mining industry, on the other hand, sought a currency devaluation, thus
reducing local costs, and opposed foreign exchange controls. This confronta-
tion effectively alienated the mining companies ± especially the larger corpora-
tions such as Patin Äo, Hochschild and Aramayo ± from the importing industries
and from the manufacturing sector (Whitehead, 1972, pp. 78±9).
Perhaps one of the most important effects of the Depression was that it
created a greater role for government in the mining industry. The government
fixed and enforced export quotas, and the small mining companies and later
the medium-sized mining companies joined forces in the CaÂmaras Departamen-
tales de MinerõÂa and the AsociacioÂn Nacional de Mineros Medianos founded in 1939
in order to carry out the distribution of quotas. Quotas became a source of
conflict both between the larger companies (Patin Äo, Hochschild and Aramayo)
and the medium and small mining companies. Through exposure in the
national press, these disputes helped to destroy the illusion of a cohesive
mining sector in the eyes of both the government and the public. At the
Manuel E. Contreras 197
same time, they contributed to a greater awareness within the urban middle
classes of the problems facing the mining sector.
The internal distribution of export quotas was based on the actual tonnage
exported in 1929 by each group or company: the quota was set by the ITC to
coincide with this percentage. During the 1938±39 period, the government
used the 1937 production as the base. These internal quotas were distributed
by the government to each mining group in the case of the large mining
groups; to each company in the case of the medium mining groups ± as defined
by those which produced more than 60 FMT per year, but less than Patin Äo,
Hochschild and Aramayo ± and to the Camaras in the case of small mining
groups (Hallowell, 1948, chs I, II).
The large mining groups closed their most expensive mines and concentrated
their quotas in the highest yielding mines. Despite these measures the restric-
tions placed on production by the export quotas had severe effects on them.
PatinÄo Mines' production fell from 20 800 FMT in 1929 to 8000 FMT in 1932.
According to its general manager, more than half of their mines were closed.
They returned to manual drilling in order to employ the greatest number of
workers, and the lowest levels of mines were allowed to flood in order to avoid
the costs of pumping. The mills worked only one eight-hour shift, five days a
week. Salaries were reduced in step with the reductions in production (Quere-
jazu Calvo, 1978, p. 174).
After 1933, the government took control of 50 per cent of any increase in the
Bolivian quota. This allowed the government to distribute its quota between
the different mining sectors, thereby compensating for any distortions result-
ing from the use of the base year (1929) in the distribution, and/or directly
exporting any increase in order to generate foreign exchange (Hallowell, 1948,
p. 55). This last measure was taken only when foreign exchange was scarce and
as the need to cover the costs of the Chaco War with Paraguay (1932±35)
emerged. The distribution of quotas was also subject to financing contracts
between the mining industry and the state. Between 1931 and 1935, the
mining industry's financing of the state increased to 29.5 per cent of the
national budget.8 In many cases the contracts stipulated that the export
quota of the mining group would increase or would remain the same.
The internal distribution of export quotas was only necessary between 1931
and 1933 when Bolivia exceeded its production quota as dictated by the ITC,
giving way to overexporting resulting from the increase in small and medium
mining production starting in 1932, when prices started to improve. This
occurred to such an extent that in April 1933, for example, the Patin Äo and
Aramayo groups voluntarily conceded part of their quotas in favour of the small
companies. By contrast, between 1934 and 1937 Bolivia was unable to fulfil its
quota and several causes can account for that failure. First, the quota assigned
by the ITC was too large in relation to the long-term production capacity of
198 Bolivia, 1900±39
Generally, taxes have been used as the main indicator of mining's contribution
to the national economy. However, we should not lose sight of the fact that
mining activity also had a regional impact: it created a demand for national
goods and for imports that, upon being introduced into the country, paid
import taxes and created wealth in the importing sectors. An example of the
boom promoted by tin mining was the urban, commercial and financial devel-
opment of Oruro, described in 1912 by a French observer as the `home of
industrial and commercial progress', with nine banks, two agencies and various
factories and commercial centres (Walle, 1914, pp. 204±5). Moreover, the role of
the mining industry in the introduction of a cash economy to rural areas has
yet to be fully studied. There is no doubt, however, that tin mining was not the
great driving force for development in Bolivia that it could have been, but
detailed explanation is beyond the scope of this work. However, two of the
main causes were the government's failure to extract greater taxes from
the mining industry, in addition to the inadequate use of these funds by the
government and the main mining companies' failure to reinvest domestically
(see GoÂmez-D'Angelo, 1978, pp. 186±7; Contreras, 1990a, pp. 265±87).11 Some
of the reasons for this second phenomenon are examined later in the chapter.
The value of export taxes as a per cent of total mining exports has been used
as an indicator of mining's contribution to the economy. This percentage
increased from a ten-year average of 2.7 per cent in the 1900s to 4 per cent in
the 1910s, reaching 9.3 per cent in the 1920s.12 However, if we calculate the
percentage as a ratio of taxes to profits, the percentage for 1925 rises from 9.1 to
17.7 per cent (Ali Ayub and Hashimoto, 1985, table 3, p. 12). Similarly, the
value of export taxes as a percentage of total fiscal revenues increased from an
average of 6 per cent between 1901 and 1904 to an average 22 per cent for the
1916±20 period (Contreras 1994, p. 8).13 However, there was debate around
mining's contribution to the economy and on how to address this issue.
Until the early 1920s, export taxes, on a sliding scale in relation to price
fluctuations, accounted for the bulk of taxation levied from the mining indus-
try. In 1920, faced with financial difficulties, the government introduced the
200 Bolivia, 1900±39
first tax on profits. But it was not until 1923 that the government of Bautista
Saavedra (1921±25) introduced a new profit tax to cover the financial obliga-
tions of the `Nicolaus' loan, taken out in 1922 to help refinance the public debt.
This tax clearly established the amount that could be deducted in order to
determine profits. The introduction of this law and the existence of a capable
entity for tax collection (the Permanent Fiscal Commission), significantly
increased taxation on the mining industry (the percentage of total taxes ± of
exports and profits ± in relation to the value of exports doubled: from 7.4 per
cent in 1923 to 15.6 in 1924).14 According to GoÂmez-D'Angelo, the 1920s stand
out as the decade during which the tax system was able to extract an ever-larger
proportion of increases in production (GoÂmez-D'Angelo, 1978, pp. 142±6).
However, as the government turned to taxes on profits in the 1920s, and
foreign exchange controls in the 1930s and thereafter, as the main sources of
resource extraction from the tin industry, the relative importance of the export
tax on tin decreased gradually.
The increased taxes on the mining industry during the 1920s and 1930s
reflected the shift in the government's fiscal policy away from import tariffs, a
feature of the 1900±10 period, towards a greater reliance on export taxes.
Although this policy was due, in part, to the state's greater financial needs, it
is still significant that there were no attempts by the government to increase
taxes on agriculture which was ± and still is ± the principal economic activity of
the country (Gallo, 1990, pp. 102±4).
Table 7.3 illustrates the relative importance of the taxes paid by the mining
industry and the greater absolute amount `extracted' from the sector by the
state during the 1930s. Based on data from the Patin Äo Mines, Table 7.3 demon-
strates that taxes on exports represented 50 per cent of total taxes paid during
the 1920s and 30 per cent of the total taxes over profits. The relative weight of
both taxes decreased significantly in the second half of the 1930s, when taxes
Äo mines, 1924±39
Table 7.3 Bolivia: detail of average taxes for selected periods in Patin
(£ 000s)
£ % £ % £ % £ % £ %
a
Includes import taxes on machinery and production goods, on edible goods, and also on urban
properties.
Äo Mines Archives, COMIBOL (1942±43), `Cuadro
Sources: Calculated from information in the Patin
demostrativo de impuestos fiscales', file 1323.
Manuel E. Contreras 201
on the exchange rate represented more than half of total taxes collected. This
reinforces our assertion that export tax revenue ± whose importance decreased
significantly after the 1920s ± is an inadequate proxy for estimating the true
contribution of the mining industry to fiscal resources.
Not surprisingly, the mining industry was opposed to all of these measures.
Precisely when it was thought taxes would be increased, Patin Äo, taking advant-
age of the shortage of funds and the regional pressures to complete the railway
between Sucre and PotosõÂ, obtained a five ±year moratorium on tax increases in
exchange for a £600 000 loan (Contreras, 1990a, pp. 270±2).
Äo offered the loan to the government for the first time on 6 August 1924,
Patin
after purchasing the Llallagua Tin Company from Chilean interests and regis-
tering PatinÄo Mines and Enterprises Consolidated, Inc. in Delaware, USA. This
acquisition completed the purchase of Llallagua Company stocks, a process
begun in 1914, through the British firm Duncan Fox and later through the
Anglo South American Bank. Indeed, the purchase of the Llallagua Tin Com-
pany ± the principal tin producer of the country ± and the later establishment of
PatinÄo Mines in the USA are two landmarks in the development of the tin
mining industry and of Bolivian capitalism. Patin Äo was the only Bolivian
capitalist (and one of the few from the periphery) who was able to build a
true transnational company ± with interests in four continents and more than
five countries ± starting from his own country.15
Bolivia was a country without a significant external debt before 1908. From
1909 to 1930, government indebtedness rose to more than US$80 million
(Ness, 1938, table 8, p. 104). More than 80 per cent was acquired in the
1920s, as the public debt tripled during this period and external loans replaced
internal loans. From 1930 to 1935 ± and especially after Bolivia defaulted in
1931 ± internal debt in Bolivia replaced the external debt, and the total debt
tripled. The final destination of the externally acquired funds cannot be deter-
mined with exactitude: Ness's estimation is that between 1909±30, approx-
imately 40 per cent of the net financial resources was utilized for the
construction of railways and 16 per cent was directed to what he terms `general
purposes of the state' ± which included everything from the Vickers contract for
military armaments, to the nation's fiscal deficit, to the payment of public
loans: 8 per cent went to public services, 5 per cent to `financing purposes' ±
in order to establish the Banco Nacional Boliviano in 1911, which was converted
into the Banco Central in 1929 by recommendation of the Kemmerer mission,
and the rest to refinance the internal debt (Ness, 1938, pp. 104±9). In contrast,
during the 1920s, railways were responsible for 80 per cent of acquired loans.
The mining industry was opposed to the external debt acquired by the
government, since it associated loans with increased taxes (US National
Archives, File 824.51/307, 1 April 1925). In 1924, Congress was contemplating
imposing an integral tax on exports, thereby eliminating the recently created
202 Bolivia, 1900±39
profit tax. Congress argued that by increasing their declared capital, mining
companies were paying between one-third and one-half of the taxes they
owed.16 In order to defend themselves from future increases in taxes ± particu-
larly now that mechanisms were in place to forcefully impose and collect them
± mining companies created lobby groups in Congress that opposed tax legisla-
tion through the AsociacioÂn de Industriales Mineros. The Association hired pro-
minent members of the Senate as lawyers (US National Archives, 28 November
1924, p. 23). While the 1920s experienced the development of an efficient
system to tax the mining industry, in the 1930s the state was able to substan-
tially decrease the export of profits through the mandatory payment of foreign
exchange, and, in paying the mining industry for its foreign exchange at the
official (overvalued) rate, managed to impose an additional tax on them. This
implicit tax rose from 12 per cent of the value of mining exports in 1936 to 17
per cent in 1938, decreasing to 14 per cent in 1939, which meant that the
percentage of taxes coming from mining during this decade increased to 14
from 9 per cent in the previous decade.17 In Patin Äo Mines, the tax resulting
from exchange controls during the second half of the 1930s exceeded the
export and profit taxes, and represented 59 per cent of total taxes paid (see
Table 7.3). In addition, the mining industry had to take a loss over its accumul-
ated balances due to the devaluation of the Boliviano, from Bs.19.6 per pound
sterling in 1933 to Bs.141 in 1939.
However, the actual impact of these measures for the country's development
was not as expected. Large quantities of money in Bolivianos were accumulated
in Bolivia without apparently finding productive investment. For example, in
1933 the Banco Mercantil had accumulated £650 000 in current accounts and
had to suspend interest payments on deposits `regardless of their time period or
condition' (Banco MercantõÂl, 1934, p. 2). The increase in foreign exchange
payments allowed Patin Äo Mines, for example, to accumulate Bs.19 million in
Bolivia in 1935 ± equivalent to 34 per cent of income and 10 per cent of
national expenditure ± which it could not put to productive use (Patin Äo
Mines and Enterprises Consolidated, 1935; Banco Central de Bolivia, 1937, p.
104). PatinÄo Mines tried to invest these accumulated balances in several areas:
in mining, in diversification and in loans. The investments made in the mining
sector were concentrated in the exploration of new mining properties and new
minerals, especially in gold, and in the purchase of stocks in existing compan-
ies. The principal efforts at diversification consisted of investment in the colon-
ization of the Chapare which, it was hoped, would yield valuable woods for
mining struts and a cattle industry to provide meat for the mining industry ±
meat had until then been imported mainly from Argentina. Unfortunately,
these projects met bureaucratic obstacles and on the whole proved to be fail-
ures. PatinÄo Mines also attempted to invest in the cement factory of Viacha, but
was not able to achieve a satisfactory agreement with its American owner, the
Manuel E. Contreras 203
demobilization after the war and, principally, for public works. Likewise, the
state's entry into the productive sector `as the result of the political need and
the ideas for renovation of the country that had vigorously surfaced as a
reaction to the structural faults that the war had emphasised' (United Nations,
1958, ch. IV, p. 62) was the cause of greater public expenditure. In the following
section we analyse the investments made by the state.
During the second period, after 1915, more railroad systems were initiated
than were finished. The railway between La Paz and Yungas ± with Beni as final
destination ± begun in 1915, was abandoned in the 1950s after less than 80
kilometres of track had been laid. The Sucre±PotosõÂ railway, initiated in 1916,
was only finished in 1943. Finally, the Cochabamba±Santa Cruz railway, begun
in 1928, was also left incomplete. The last two railways were the source of
intense regional and national debate since they represented important regional
aspirations (RodrõÂguez, 1992). It is evident that these railways are markedly
different from those built during the first period, since the former had the
underlying motivation of linking domestic areas.
Railroads completed in this period included the Atocha±VillazoÂn railway,
begun in 1915 and finished in 1925, and the Oruro±Cochabamba railway ±
although initiated during the first period (1906), it marked a turning point in
railway construction between the phase of externally-oriented railways and the
phase of internal integration. Indeed, this railway was the first effort to estab-
lish a line of communication between two important cities.
Finally, there were several frustrated attempts to construct railways of `integ-
ration' (in contrast to `export' railways), primarily financed by private capital;
the notable exception being Patin Äo's Cochabamba±Chimore railway project
aimed at colonizing the Chapare. After almost two years of discussion, the
Bolivian Congress decided to deny Patin Äo the railway concession (Geddes,
1972, pp. 152±67). We underline this fact because it illustrates the private
sector's interest (in this case the mining sector's) in developing railways away
from Bolivia's borders and focused, instead, on promoting regional develop-
ment, an aspect that is ignored by historiography.
Another interesting example is the Sucre±PotosõÂ railway. Begun with domes-
tic funds, in the 1920s it became partly responsible for growing foreign indebt-
edness. It is singled out because it a good example of a `political' railway. It was
not built as a response to economic criteria (it was not very profitable); rather,
construction was carried out in order to satisfy political interests: important
regional interests demanded that a railway be built, and it was initiated in 1916
following many debates. By 1922 it was evident that the volume of traffic was
not sufficient to pay for the interest on the loan to cover construction costs
(McQueen, 1925, p. 108).
It is perhaps not surprising, once we take into account the political and
economic necessity of railway construction, that railways were responsible for
approximately 40 per cent of Bolivia's external debt between 1900±30 (Ness,
1938, table 8, pp.107). However, the railways were not only financed with
external debt: compensations received by Bolivia following the wars in the
Pacific and Acre were an additional source of finance. As a product of the
PetroÂpolis Treaty (1903) the country received £2 million, which was utilized
by the Bolivian Railway (formed in 1906) to construct the main railways of the
206 Bolivia, 1900±39
country. Similarly, through the Peace Treaty with Chile in 1904 the Chilean
government pledged to finance the railway between Arica and La Paz, which
was finished in 1913. Finally, as we mentioned previously, private capital also
invested in railroads.
Railroad construction stopped following the suspension of external credit to
Bolivia and the advent of the 1930s economic crisis. Roads took over. Bolivia
had no road system at the beginning of the century, and the few roads that did
exist had been constructed towards the end of the nineteenth century around
the country's principal cities and the adjacent rural populations. These were the
municipal roads. The national roads connected the principal cities and the
mining centres. The majority were dirt roads. By 1941, however, there were
11 000 kilometres of roads, although only a fraction of these could be used
throughout the year (Oficina Nacional de InmigracioÂn, EstadõÂsticas y Propa-
ganda GeograÂfica, 1905, pp. 464±7; Contreras, 1990b, p. 193).
In the 1930s, the Central Bank served as the principal driving force behind
road construction. In 1933 it presented a plan of road and public works to the
legislative body to `mobilise capital that would strengthen businesses in gen-
eral, giving work to thousands and increasing the possibility for wealth in the
diverse regions of the motherland'. It was hoped that the initiation of the
public works would be `a powerful stimulus for the productive districts, that
would intensify interdepartmental commerce of products, thereby invigorating
the activities of businesses in domestic markets' (Banco Central de Bolivia,
1934, p. 14). This initiative lasted until 1939, by which time the Central Bank
had handed out more than Bs.100 million in municipal and departmental loans
(Banco Central de Bolivia, 1940, p. 74).
Education
Political instability and slow economic growth were responsible for the limited
development of Bolivian education in the nineteenth century. By the twentieth
century there were only a few hundred schools and less than 50 000 students in
a country of 1.7 million people. The rate of illiteracy was 83 per cent; and the
education budget amounted to 5.9 per cent of the total budget. It was in this
context that the Liberals came to power, armed with an education project.
Ismael Montes, who governed the country from 1904 to 1909 and then from
1913 to 1917, is perhaps the clearest example of the new thinking on education
in this period. With the support of various Ministers of Education who pledged
to change the Bolivian education system, Montes achieved the first education
reform of the century. The principal aspects of the reform were,
A characteristic of this reform process was the search for educational models in
Europe, going as far as hiring Belgian pedagogues to direct the first teacher-
training schools.
The administration of the education system was strengthened with the crea-
tion of the Direcciones de EducacioÂn, and with the state's gradual takeover of
educational responsibilities from both the municipalitiesand the Catholic
Church. Until then, primary schools were under the sole guidance of the
municipalities, and there was no national coordination of these schools.21 For
the directors of the reform it was not simply a question of increasing infra-
structure, but also of organizing the little that existed. In some cases the state
was able to recover the municipal schools with municipal funds, but in
the majority of cases the state had to dedicate new funds to education. During
this period the education budget increased from Bs.382 724 in 1900 to
Bs.4.083 million in 1924. The number of public schools rose from 84 in
1900 to 614 in 1924 (Contreras, 1990b, p. 316). Meanwhile, the reformers did
succeed in developing standard curricula at the national level.
Bolivia had no teacher-training system until 1909. There were few teachers,
and most had been educated outside the country and/or were foreigners ±
usually Chilean or European. In order to increase the number of teachers, the
liberal government first sent Bolivians on scholarships to train abroad, until in
1906 it formed the Normal School for Teachers in Primary Education in Sucre,
in 1915 the first Rural Normal School in Umala, La Paz, and in 1916 the first
institutes to train teachers in secondary education. In effect, the `Normal
School of Sucre was the base of the entire Bolivian education reform' (Minis-
terio de EducacioÂn, 1917, pp. 58±59). Not only did it help develop teacher-
training in Bolivia, but also, in hiring European pedagogues to do so, a new
teaching system was developed that sought to instil in pupils a `solid and
extensive' knowledge, to form their `scientific spirits' and to prepare `men
and women of initiative' (Ministerio de EducacioÂn, 1917, pp. 58±59).
From another perspective, the education of indigenous individuals contained
a clear `civilizing' concept and sought to `Castilianize' them in order to incor-
porate them into national life. The first action by the liberal government was to
establish a system of travelling schools with specially chosen teachers to visit
various ayllus, or indigenous communities. In order to give greater continuity
to this process, permanent schools were established; indeed, many were created
and developed by the indigenous populations' own initiative.
208 Bolivia, 1900±39
The reform established mixed education in the standard schools, and under
liberal rule the first lyceums for girls were set up in La Paz in 1917. The reform
also established technical education in the cities, and commerce schools
were set up to train accountants and skilled office workers. In PotosõÂ and
Oruro, the first surveying and mining schools were created which would later
serve as the foundation for mining engineering schools.22
The liberal reforms addressed the principal aspects of Bolivian education.
It was a top-down reform, in which European influence was significant,
and arose out of decisions taken by the governing elite and visionary education
ministers. The liberals established the foundations of the education system
that was later subjected to the fluctuations of the Bolivian national economy.
Although the budget for the education system increased in the 1920s, in
the 1930s it was tightly restricted by the economic crisis. However, in 1930
an administrative reorganization of the education sector was attempted
with the creation of the National Council of Education. This council was
composed of the Director General of Education and three general inspectors,
and presided over by the Education Minister. Its mandate was to direct public
education, supervise private education, appoint and promote teachers and
manage the education budget. This attempt gave considerable autonomy to
the education sector by placing policy in the hands of individuals with
educational or administrative qualifications, instead of in the hands of a
politically nominated minister. However, the main weakness of the proposal
was that the Council never achieved financial autonomy from the Ministry.
Nevertheless, the attempt is a clear indication of pro-education initiatives at
this time. In 1939, the education budget reached levels similar to those at the
end of the 1920s, representing 8.4 per cent of the total budget (Nelson, 1949,
pp. 8, 11).
In the 1930s, indigenous education also grew. In contrast to the 1920s, in
the first half of the 1930s this expansion was due more to the indigenous
population's own efforts and to the efforts of ecclesiastical organizations
than to efforts by the state. Schools were part of `the struggle to brake the
expropriations of communal lands and to restore previously usurped com-
munities' (Soria Choque, in Choque et al. 1992 pp.41±78). The Warisata ayllu
school developed in this context. Founded by Indian communities in 1931,
it reflected the Indian communities' needs by providing practical training in
agriculture and crafts. It became the model of Indian education had a profound
impact on Bolivian education and is widely regarded as a genuine effort at
reform `from below'. The Chaco War interrupted this effort, as it recruited
teachers and Indians, and children left school to take care of the fields
(Contreras, 1999).
After the war, the Toro (1936±38) and Busch (1939) administrations supported
indigenous education vigorously. Despite this support, Warisata was closed
Manuel E. Contreras 209
Conclusions
Although this process was of greatest advantage to urban centres, rural areas
benefited from the expansion of education and the development of infrastruc-
ture in railways and roads. The different governments during this period were
able to increase the resources extracted from the mining sector, despite the
increasing organization of this industry. Once the possibility of depending on
the external sector had gone and confronted with the need to finance the
Chaco War, the Bolivian government was able to shift its taxation policies.
Through the mandatory payment of foreign exchange and through differential
exchange rates, the government was able to extract greater resources from the
mining sector.
Contrary to what we might expect, the reduction in the outflow of dividends
and the accumulation of resources in Bolivianos in the country did not result in
a greater level of development. The true barriers that opposed growth and
development were the lack of viable investment opportunities for the private
sector in the 1930s, and the squandering of public funds caused by corruption
and investment in nonproductive, although prestigious, projects in the 1920s.
Both cases point to the structural problems fundamental to the Bolivian eco-
nomy, which to a greater or lesser extent persist even today.
In changing its fiscal policies, the government demonstrated its capacity to
redirect its actions and adapt to new situations, both in the formulation of new
policies and in their implementation. Nevertheless, it is fitting to ask if there
were alternatives to the policies carried out during this period. According to
Gallo (1990), confronted by the external debt the government had little alter-
native but to increase taxes on the mining industry. Clearly this is correct, but
in order to understand the manner in which it was done ± through a single
differential exchange rate and the mandatory payment of foreign exchange ± it
is necessary to understand the influence that the Chaco War had in the case of
Bolivia. This conflict was a powerful stimulus for the state to take on a more
active role, and once exchange controls were imposed during the war they were
not rescinded after the end of the conflict, as they were perceived to be effect-
ive. In addition, there existed an institutional mechanism to carry out such a
policy ± the Banco Central ± while the apparatus for collecting taxes remained
deficient.
Moreover, there is little doubt that the state found it easier to use this
mechanism than to create new taxes for the mining sector. From this perspect-
ive, the Bolivian case seems to confirm Hirschman's dictum that where the
export sector is powerful and influential, it is politically difficult to increase
taxes and, in this context, the great advantage of an `inflation-overvaluation'
arrangement is not only that it has an indirect influence on politically and
socially influential groups, but that the mechanism also has greater legitimacy
given that it has the respectable aim of `defending the national currency against
depreciation' (Hirschman, 1971, p. 118).
Manuel E. Contreras 211
Notes
1 Data taken from Bieber (1984, tables II.5 and II.10, pp. 20 and 32); on the develop-
ment of rubber production see Gamarra (1993, pp. 23±76).
2 Fine metric tons (FMT) is the actual metal content of the ore, as opposed to gross
tons.
3 Mitre (1993, table 11, p. 212) estimates that for 1928, national capital represented 41
per cent, US capital 39 per cent, and Chilean capital only 20 per cent, of total mining
investment.
4 In effect, `bringing Bolivian minerals to the US not only meant cheaper tin and
employment for thousands of American workers, it also meant the creation of a
market for products from the US that . . . could reach $ 50 million annually' Daily
Commerce Reports, vol. 1, no. 4 (6 January 1916), p. 53, cited in Contreras (1984, p. 34).
5 See the correspondence between Carlos VõÂctor Aramayo and Carlos Navarro in the
Navarro Papers in the La Paz Archives. These papers contain many letters covering the
period from 1919 to 1928.
6 Calculation from Bolivia, 1936, El estanÄ o en Bolivia 1935, La Paz, table 7.
7 Concerning the effects of the Great Depression on the Bolivian economy, see White-
head (1972), Pacheco (1990) and Contreras (1990a).
8 Calculated from data from the General Budget from the Banco Central de Bolivia
(1937, p. 104).
9 Archivo Patin Äo Mines-CorporacioÂn Minera de Bolivia, `Turnover of labor'. Concern-
ing the role of women in Bolivian mining during the nineteenth and twentieth
centuries, see Gustavo RodrõÂguez Ostria (1991, ch. IV).
10 See Altshuler (1939, pp. 41±7) for a technical description of the mine. Letter from the
vice-president of Mauricio Hochschild S.A.M.I to the president of the Commission of
Nationalization of Mines dated 8 August 1952, in Informe de la ComisioÂn de Naciona-
lizacioÂn de Minas.
11 During the 1920s and 1930s, Contreras notes that `the real barriers to the growth
were the absence of opportunities for investment for private capital in the 1930s, and
the waste of public resourses due to corruption and the investment in prestigious yet
non-productive projects during the 1920s' (1990a, p. 287).
12 Calculation from United Nations (1958, p. 10). The value of exports is an estimate
from customs and includes part of the added value of transport and smelting. This is a
proxy indicator.
13 The later decrease is explained by the fact that after the 1920s, export taxes were no
longer the only taxes levied on the tin mining industry. First a tax on profits on the
industry was imposed and later an `implicit tax' was introduced by establishing the
mandatory payment of foreign exchange using an overvalued rate of exchange, as
detailed in the following paragraphs.
14 Concerning the `Nicolaus' loan and the ComisioÂn Fiscal Permanente, see Contreras
(1990a); data from GoÂmez-D'Angelo (1978), table 13, p. 204.
15 See Contreras (1994, pp. 26±8), for the emergence of the Patin Äo `empire'.
16 The tax rate, according to the Law of 1923, was fixed by establishing the relation
between profits and the declared capital. The greater the relation, the greater the tax
rate. Therefore, by increasing their declared capital the mining companies could
effectively reduce their tax rates.
17 GoÂmez-D'Angelo (1978) implicit tax in table 18, p. 217, averages calculated from
table 13, pp. 203.
Manuel E. Contreras 213
References
Soria Choque, V. (1992) `Los caciques apoderados y la lucha por la escuela (1900±52)', in
R. Choque et al., op. cit.
United Nations (1958) AnaÂlisis y proyecciones del desarrollo econoÂmico, Vol. IV: El desarrollo
econoÂmico de Bolivia, Mexico City.
US National Archives (1925) File 824.51/307 `Roswell to State Department', Washington
DC, 1 April.
ÐÐ (1924) Roswell Barker to State Department, 28 November.
Walle, P. (1914) Bolivia: Its People and its Resources, its Railways, Mines and Rubber-Forests,
London: T. Fisher Unwin.
Whitehead, L. (1994) `State Organization in Latin America since 1930', in L. Bethell (ed.),
The Cambridge History of Latin America, Vol. IV, Part 2, 1930 to the Present, Cambridge:
Cambridge University Press.
ÐÐ (1972) `El impacto de la Gran DepresioÂn en Bolivia', Desarrollo EconoÂmico vol. 12,
no. 45.
8
Trying to `Tax and Spend' Oneself out
of the `Dutch Disease': The Chilean
Economy from the War of the Pacific to
the Great Depression
Gabriel Palma*
It might be argued that in general one should guard against the tendency to
overestimate the importance of economic factors behind any war; however, in
the particular case of the War of the Pacific perhaps the danger lies in under-
estimating these factors. Firstly, the Peruvian government nationalized the
nitrate deposits in the mid-1870s, affecting a number of important interests,
both Chilean and British.1 Among the former were those of the past President
of the Republic, Manuel Montt, and the future President Jose Manuel Balma-
ceda, as well as those of a number of leading national entrepreneurs.
Second, Chile had been experiencing economic crisis since 1873; and
towards the end of the decade this crisis was fast becoming unmanageable.
Between 1872 and the outbreak of the War of the Pacific the export price index
fell to the lowest level recorded since foreign trade statistics were first collected
(1844). The collapse of export prices was a consequence of the international
economic crisis which followed the end of the `Great Victorian Boom' ± a cycle
of rapid economic growth which lasted from about 1850 to 1873. Largely
because of the high levels of external public debt ± which in real terms had
trebled in the decade leading up to 1873 ± the decline in Chile's exports was
accompanied by an even greater retraction in imports. These fell by half
between 1875 and 1879, producing a chain reaction throughout the national
economy.2 Public revenues were also seriously affected, mainly as a result of the
* I am grateful for comments received at the conference of AtitlaÂn and at various seminars, especially
in Santiago, CancuÂn, Bilbao, Cambridge, London and Sussex. Edna ArmedaÂriz, Enzo Faletto, Daniel
Hahn, Francisco LeoÂn, HernaÂn SaÂez, IgneÃs SodreÂ, the editors of this book and particularly Geoff
Harcourt and Luis Ortega have also made valuable contributions. The collaboration of Margaret
Hancox has been invaluable. The responsibility for the final result is, of course, mine.
217
218 The Chilean Economy and `Dutch Disease'
high proportion drawn from foreign trade activities, falling by nearly 40 per
cent in 1874. By 1878 revenues were still 28 per cent below the 1873 level.3
Contemporary commentators were unanimous in affirming that the eco-
nomic crisis was the worst that the country had experienced in the whole of
its independent political life.4 In 1878, in ValparaõÂso alone, seven of the leading
commercial houses went bankrupt leaving debts in excess of £ 500 000;5
importing and trading concerns, some as important as that of A. Gibbs, closed
down a significant proportion of their operations; some of the country's leading
exporters, such as Jose TomaÂs Urmeneta and Maximiano ErraÂzuriz, were unable
to meet their obligations; while the government itself was hard pressed to pay
public employees. In 1877 the public deficit reached a level equivalent to one-
third of public expenditure.
Given the critical nature of the financial situation, on 22 July 1878 the
Treasury Minister presented the Chamber of Deputies with a bill temporarily
suspending the convertibility of the peso; so evident was the crisis that Parlia-
ment approved the law in less than 24 hours.6 The outlook for the following
year was gloomier still; for example, on 31 January ± that is, only a fortnight
before the outbreak of hostilities ± the newspaper El Ferrocarril predicted in an
editorial that 1879 `would be a year of mass business liquidation'.
Perhaps the most telling account of the situation facing the country is found
in President Pinto's personal diary. Just three months before the beginning of
the war he wrote in a way that left no room for doubt:
The economic situation of the country is very bad and the forecast is of a
deterioration rather than an improvement. The harvest has been abomin-
able and the price of copper in Europe is falling as never before. Another bad
year with a situation as delicate as the one we have, could have nothing but
bad consequences . . . If a new mining discovery or other news in the same
style does not come to improve our position, the crisis that has been affect-
ing us for years will get worse.
(Personal Diary of President Pinto, 18 November 1878, p. 309; quoted in
Ortega, 1979, p. 455)
by its Parliament and was therefore not in force.7 Chilean companies which
were affected refused to pay the new tax on the grounds that it violated the
treaty, and the Bolivian government ordered an auction of their assets. As a
result the Chilean government occupied Antofagasta and thus initiated the
bloody War of the Pacific. Chilean troops disembarked and occupied Antofa-
gasta on 14 February 1879, the very day fixed for the auction of the assets of the
CompanÂaÄõ de Salitres y Ferrocarriles de Antofagasta, a company in which three of
the four civil ministers in President Pinto's first war cabinet were important
shareholders.
It is obviously impossible to guess how the Chilean government would have
reacted to these events had there not been an internal economic crisis of such
magnitude, and if the economic interests of such important politicians had not
been affected so directly. In any case, from an economic point of view Bolivia's
suicidal violation of the treaty could not have occurred at a more critical time
for Chile and, from a political point of view, it could not have attacked more
influential Chilean interests.
There were also other political problems, dating from earlier years. In the
forefront of these was the long struggle between Chile and Peru for hegemony
along the Pacific coast. This was stated clearly by Jose Manuel Balmaceda when
he was a deputy:
On the Pacific coastline in South America there are only two centres of
attraction and progress: Lima and Callao, and Santiago and ValparaõÂso.
One of those centres must succumb so that the other can arise. For our
part we need Tarapaca as a source of wealth and Arica as a frontier outpost
on our coastline.
(Chile, Congreso Nacional, Chamber of Deputies, Extraordinary Session 18,
9 August 1881; author's own translation)
The territories of Antofagasta and the nitrate territory of Tarapaca were the
real and direct cause of the war. . . [and therefore] to return to the enemy
dominion over the very cause of the conflict, after our triumphs and the
occupation of those territories, would show unjustifiable negligence, and an
absolute lack of the knowledge which affairs of state presume.
(Chile, Memoria del Ministerio de Relaciones Exteriores, 1881, docto. 16;
author's own translation)
Finally in the AncoÂn Peace Treaty between Chile and Peru, signed in October
1883, which brought the War of the Pacific to an end, nine of the 14 clauses
220 The Chilean Economy and `Dutch Disease'
made reference to nitrates and to guano; this, together with the facts described
above, has caused many historians to refer to this war as `The Fertiliser War'
(Loveman, 1988).
In the economic history literature there is consensus that in Chile in the period
between 1830 and the First World War the `engine' of economic growth was
located in the export sector. However, this consensus does not extend to the
type of economic, political and institutional dynamics which emerged from this
export-led growth; there are strong divergences in the literature with respect to
the way in which the `energy' or `stimulus' generated by this export engine was
transmitted to the rest of the economy, in particular to manufactures.
The engine of growth was located in the export sector because of both the
quantity and diversity of exportable natural resources, and the fact that these
were the only activities not constrained by the reduced size of the internal
market. Constraints on exports were given rather by the capacity to mobilize
productive resources ± national as well as foreign ± in order to prospect, extract,
transport and sell these natural resources.
The controversy in the literature is centred on two areas. The first concerns
the kind of productive structure and political institutions that resulted from
this export-led growth. The second relates to the effectiveness with which the
potential for economic growth generated by the export engine was used; in
particular, how effectively the injection of aggregate demand from the export
sector was oriented towards the internal market. The first area concerned the
type of problems analysed by the theory of linkages proposed by Hirschman,
the second those of the Keynesian multiplier.
In the case of nitrates, the basic economic question raised by the first con-
troversy is whether, through input±output relationships, the nitrate industry
put into motion significant investment activities in the rest of the economic
structure, or whether it became an `enclave' with little productive impact on
the country's economy. Concerning the second controversy, the main question
is whether the rapid expansion of exports tended to promote or restrict the
increase in the economy's non-export tradable activities, particularly in manu-
facturing industry; that is, whether the resulting structure of relative prices and
expenditure was favourable or unfavourable for the development of these
activities. In the post-oil-crisis terminology of the 1970s this controversy ques-
tions whether the nitrates expansion caused an outbreak of `Dutch Disease'.
In the analysis which follows we will examine various aspects of both con-
troversies by combining the model of Hirschmanian linkages with that of the
Keynesian multiplier in a generalized linkages approach; that is, one of direct
Gabriel Palma 221
and indirect linkages generated by the export sector with the rest of the eco-
nomy.8 From this point of view, the dynamics of the export sector stimulate
direct linkages with the rest of the economy, be they backwards or forwards,
through purchases of inputs and machinery, the development of infrastructural
work needed for their operation, and the demand for agricultural products for
those working in the export sector, or through the processing required by the
primary product prior to its being exported.
The indirect linkages between the export sector and the rest of the economy,
in turn, have two sides; one is related to the generation of foreign exchange, and
the other to the injection of aggregate demand from the export sector. As for the
foreign exchange generated by the export sector, equivalent to the returned
value of exports (total value of exports minus profit repatriation by the
foreign firms which controlled nitrates and later copper), such revenues lift
the foreign exchange constraints on the rest of the economy. As far as the
injection of aggregate demand from the export sector is concerned, to the direct
component ± related to the production process of the exported goods ± must be
added the indirect component, linked mainly to the public expenditure
financed by export taxes and to the local consumption and investment expend-
iture of the exporting entrepreneurs who returned their surplus to the country.
The dynamics of the export sector depend, on the demand side, on the
volume, composition and stability of international demand for the country's
exports. On the supply side they depend on the export sector's capacity to
respond to the demand stimuli it receives. This capacity to react is a function
of the availability of natural resources and of the possibility of mobilizing
national and foreign factors of production. The direct linkages depend on
such specific characteristics of the export sector as its geographical location,
ease of transport and convenience of processing prior to export, on the tech-
nology used (which affects, for example, the labour intensity of the activity),
and on the input-purchasing policy of the companies involved in the export
sector.
The indirect linkages depend on the way in which the export sector is linked
to the rest of the economy. On the demand side, three factors stand out. The
first is the volume of the aggregate demand generated by this sector. The second
is the composition of this aggregate demand, particularly the proportion of the
surplus retained by the government in the form of taxes and the structure of
public expenditure ± in particular the proportion destined for investment ± and
the propensity to save, invest and consume of those controlling the export
sector. The third is the effectiveness with which the injection of aggregate
demand from the export sector is oriented to the internal market, which is a
function of relative prices ± in particular of the structure of import taxes, the
rate of exchange, the difference between internal inflation and that of the
international market and the behaviour of transport costs.
222 The Chilean Economy and `Dutch Disease'
On the supply side, the most important factors enabling the indirect linkages
are given by the proportion of the export surplus invested in the country's non-
export activities, their technological complexity, entrepreneurial initiative and
the availability of skilled labour.
The dynamic of the (new) export sector and its direct linkages
The volume, composition and stability of international demand for Chile's
exports
After 1873 the rapid fall in international demand for copper, silver, wheat and
flour ± Chile's main exports before the War of the Pacific ± was quickly com-
pensated by a large international demand for nitrates, an essential input for the
explosives industry and for agriculture in more developed countries (at a time
when, furthermore, the guano industry was in rapid decline). Chile's nitrate
exports from occupied territories grew so quickly that by the time Chilean
troops took Lima in 1881, these exports had already reached a level higher
than that for any of the other prewar export products mentioned. After the War
of the Pacific, Chile, in possession of the provinces of Tarapaca and Antofagasta
± annexed from Peru and Bolivia respectively ± became the only country in the
world with economically exploitable reserves of natural nitrate salts.
Nevertheless, the concentration of international demand for Chile's exports
on a single commodity brought an important element of instability. It also led
to a decline in the degree of manufacturing value-added in Chile's exports, as
nitrates, unlike copper and silver, did not require significant local processing in
view of the simplicity of the process of extraction and the small gains possible
in time and cost of transport.
As regards the stability of international demand for this new leading export
product, we find an accentuation of the cycle which characterized export
products in the preceding period. An initial boom in international demand ±
with local production responding rapidly ± was followed by a decline towards
the end of the 1890s, a subsequent recovery followed by another rapid decline
after the First World War, and then by a brief increase in the 1920s, and finally a
fall after the 1929 crisis. The first downturn in demand towards the end of the
1890s was due to the problems in the world economy up to 1896±97.9 The
second took place after the end of the First World War and was caused as much
by the development of both natural and synthetic substitutes for nitrates ±
which exposed the low level of productivity of Chilean exports (Figure 8.1) ± as
by the state in which the war left the world economy.10 The third drop, from
1929, was one from which the natural nitrate industry would never recover.
The industrial production of synthetic nitrates began in Germany at the
beginning of the twentieth century, but as the cost was initially significantly
higher than that of Chile's natural nitrates, the first major threat to the
Gabriel Palma 223
monopoly which Chile enjoyed came with the discovery of the fertilizing
potential of another mineral, ammonium sulphate. This was found in Europe,
and although the cost of production was higher than that of Chile's nitrates, it
enjoyed a number of technical advantages. Thus, while Chile's production
of nitrates grew by 73 per cent between 1903 and 1913, global production of
ammonium sulphate grew by 157 per cent. Nevertheless, the final blow to
Chile's nitrate producers came with the subsequent introduction of a number
of technical improvements in the production of synthetic nitrates. Whether or
not the nitrate industry in Chile was ever capable of transforming its methods
of production, or whether this was simply left too late, the fact is that ammo-
nium sulphate and synthetic nitrates swiftly displaced natural nitrates from the
market.11
However, in Chile's history not only were errors repeated, but also strokes of
fortune. If some leading export products collapsed, others appeared with equal
celerity to replace them; in this particular case, paradoxically, there were vari-
ous elements common to the technological changes which made the produc-
tion of synthetic nitrates economically viable, and those which enabled the
large-scale exploitation of copper minerals of low concentration, of which
Chile had one of the largest reserves in the world. Thus, with the opening of
the mines of El Teniente and Chuquicamata in 1912 and 1915, and of Potrerillos
in 1927, large-scale copper-mining, the Gran MinerõÂa del Cobre, was to become,
with the collapse in demand for Chilean nitrates, what nitrates had been to
copper, wheat, flour and silver at the end of the 1870s.
Nevertheless, the crisis of the 1930s was to change this export cycle entirely;
it would not be the problem of productivity which in the final analysis would
reveal the vulnerability of the new sector; Chilean exports were not now to be
replaced in the international market by new and more efficient producers; new
export products would no longer be relied upon to rescue the national eco-
nomy from successive crises. The drop in international demand for Chile's
exports would be of such magnitude and so long-lasting,12 and its effects on
the national economy so profound, that in the four following decades the
remedy was to be sought in a change in the `style of development', rather
than in that of the export product of the moment ± a process that had begun
around the time of the First World War. This new and profound `export pessim-
ism' ± which began with the nitrate problems and worsened with the 1929 crisis
± was to be the common denominator of the economic policies of the eight
governments following the crisis, in spite of the fact that ideologically they
ranged from conservatives to socialists.
The response capacity of the export sector and its direct linkages
In the year before the outbreak of hostilities ± 1878 ± Chilean exports were at
their lowest level since 1866; in 1882, while hostilities still continued, they had
224 The Chilean Economy and `Dutch Disease'
reached the highest level ever recorded in the country's history. Nitrates' share
of total exports was to grow rapidly, to peak at 81 per cent a year before the First
World War broke out. Nitrates' relative decline thereafter was swift. By 1929 the
share was on a par with smelted and refined copper, and fell sharply through
the 1930s and into the 1940s so that by 1940 nitrates accounted for only 20 per
cent of the total.
In terms of the quantum of production, nitrate exports from Bolivian territ-
ory before the war ± a significant part of which was carried out by companies
with Chilean capital ± had risen continuously during the 1870s, while those
from Peruvian territory had declined by 18 per cent between 1875 and 1878,
mainly as a result of the Peruvian nationalization of the deposits. The war
barely affected production in the province of Antofagasta, which achieved a
record level only one year into the conflict. The recovery of production in the
province of Tarapaca began in 1880, and in 1882 also reached its highest level
of production ± in spite of the fact that fighting was still going on in Peruvian
territory. In 1890 the combined volume of production in Tarapaca and Antofa-
gasta was already 3.3 times higher than in 1878, but almost all the growth had
taken place in the province of TarapacaÂ. If we compare the level of production
in 1890 with 1875, the highest prewar level, we find that the former was 3.2
times higher. Later, as a result of the rapid drop of the price of nitrates, the main
exporters ± in particular the British speculator John Thomas North who by then
had become the so-called `Nitrate King' ± pursued a policy constraining produc-
tion which led to relative stagnation of exports between 1890 and 1897. How-
ever, when the nitrate price began to rise again, towards the end of the decade,
production responded rapidly causing exports of this mineral to increase 2.6
times between 1897 and 1913. Production fell rapidly in 1914 and 1915, but
recovered quickly, reaching over three million metric tons in 1917, its highest
level, never since exceeded. Production fell again in 1919, but by 1929 had
almost reached its record level of 1917. The 1930s crisis brought about the
industry's final collapse.
The downturns of the nitrates cycles were not a product of the exhaustion of
natural resources or of an increase of production costs brought about by a
significant reduction in the quality of resources. They were a product of turbu-
lence in the international economy, the producers' reaction to this, and of the
development of alternative sources of supply which exposed the local produ-
cers' inability to maintain competitiveness ± when ammonium sulphate and
synthetic nitrates brought about this situation, around the time of the First
World War, the Chilean nitrate industry was still employing the same technol-
ogy used in the 1880s. In fact, as Figure 8.1 shows, productivity in the nitrate
industry had been declining rapidly since the mid-1880s.
It is indisputable that the combined effect of the lack of substitutes for
nitrates before the First World War and of the fact that Chile owned the only
Gabriel Palma 225
100 100
80 80
60 60
40 40
20 20
0 0
1880 1890 1900 1910 1920 1930
Figure 8.1 Chile: productivity in the nitrate industry, 1880±1930 (output per worker, 3-
year moving average)
Source: Statistical appendices in Palma (1979). Unless otherwise stated, this is source for all graphs in
this work.
1878=100
1000 1000
800 800
600 600
400 400
200 200
0 0
1865 1875 1885 1895 1905 1915 1925 1935
(Palma, 1984). Its subsequent recovery was slow and difficult. The export
volume at the end of the 1920s was not achieved again for three decades14 ±
and in per capita terms, not for half a century!
As regards the direct backward linkages of the nitrate export sector with the
rest of the economy, this reveals a panorama similar (if not worse) to that
shown by wheat production and extractive mining in the previous period;
despite its enormous importance to the national economy, its direct capacity
to stimulate local manufacturing via purchasing inputs was limited.15 However,
if demand for food from the nitrate area is included among the direct backward
linkages, this was not the case for agriculture.16 It has been argued that this
demand provided a strong stimulus for agricultural production (Kirsch, 1973
and Cariola and Sunkel, 1982), in particular because it was linked to successive
governments' policies for supplying the nitrate area with domestic production
and not with imports. It should be taken into account, however, that the war
obliged a significant proportion of the Bolivian and Peruvian population of the
occupied territories to emigrate to their respective countries. Therefore, an
important part of the demand for agricultural products from the nitrate areas
was in order to feed workers transferred from other parts of Chilean territory,
which indicates a reallocation of demand rather than a net increase. Further-
more, in the case of products such as wheat and flour, part of the additional
demand from the nitrate areas simply absorbed production which had
Gabriel Palma 227
previously been exported. In other words, the rapid increase in demand for
food in the nitrate area was probably more a stimulus for activities associated
with the transport, storage and commercialization of agricultural products,
than for agricultural production itself.
As for the direct forward linkages, nitrates did not require significant process-
ing before export. However, because of its crystalline structure nitrate has a low
density, so its export required a large fleet since each metric ton exported
needed a cargo space of approximately 1.5 cubic metres.
As far as manufacturing industry is concerned, as nitrate exporting did not
call for local processing activities it did not generate connected local manufac-
turing activities; nitrates also did not require large amounts of manufacturing
inputs or machinery. Nevertheless, we should not pass over the purchase of
some locally manufactured inputs, or, with greater reason, the investment in a
railway network to facilitate nitrate exports, and the port works that the export
traffic required. Although the largest oficinas imported a substantial proportion
of their capital equipment, there is good evidence that the nitrate installations
came to be among the major customers of many of the foundries, workshops
and establishments producing transport materials in the period, and the
immediate causes of the creation of a number of similar establishments.
Among the latter, we may pick out as an example the creation in Antofagasta
of the FundicioÂn Orchard Hermanos, which produced a wide range of metal goods
for the nitrate works, and also installed a steel converter of the Bessemer type.
Again, in the first years of this century Siemens-Schuckert Ltd ± the world's
largest conglomerate producing machinery and electrical equipment ± opened
a subsidiary in Antofagasta exclusively for the purpose of supplying the oficinas
with the products they required (Kirsch, 1973; Palma, 1979; and Pinto and
Ortega, 1991).
In relation to the development of railways in the nitrate zone, an idea of the
transport requirements of the nitrate industry can be derived from the fact that
the deposits were found discontinuously along a strip of land more than 4000
sq kms in extension. Railways were essential if production was to be increased,
as otherwise it would only have been possible to extract deposits near the ports
of exit.17 Between 1868 and 1875 a Peruvian company built two important
railway lines, those of Iquique and Pisagua; these later became part of the
Empresa Nacional de Ferrocarriles del PeruÂ.18
After the War of the Pacific, with the rapid expansion of the nitrate business a
number of new railway companies were formed, most of them British-owned.
The total length of private railway line in the country, which in 1875 had
amounted to 650 kms, exceeded 1000 kms within a decade, to reach 1254 kms
in 1887 and 1614 kms in 1890; by this time, however, the private rail network
was not confined to the mining areas of the north but included the coal fields of
the south. By the time the 1930s crisis struck, Chile had 9000 kms of track.19
228 The Chilean Economy and `Dutch Disease'
The articulation of the export sector with the rest of the economy
via indirect linkages
The volume of the demand injection from the export sector
From the point of view of the real value of exports, the export profile is even
more erratic than that of the volume exported, as the export price index
followed a cyclical pattern which tended to magnify changes in the volume
exported.
Figure 8.3 shows that the real value of Chilean exports tripled between 1878
and 1888. Exports then barely rose until 1900, almost doubled in just 8 years
and continued to grow until the First World War. As a result, the real value of
exports was 6.6 times higher in 1913 than in 1878 ± representing an annual
average growth rate of 5.6 per cent. The collapses after 1913 and 1919 are also
worth noting, as is the rapid recovery until 1929 ± now led by the Gran MinerõÂa
copper exports. Exports later fell even more spectacularly, and in 1932 only
achieved a sixth of their 1929 level.20
This cyclical behaviour was even more striking in the case of nitrates. The real
value of nitrate exports grew initially at an annual average rate of 15 per cent in
the 1880s; this dropped to 4 per cent in the 1890s, to be revitalized between
1899 and 1913 at 10 per cent. The real value of exports of these nitrates was 13
times higher in 1913 than in 1880. In 1919, on the other hand, the value
exported was only a seventh of that of 1913, but recovered in the 1920s.
Gabriel Palma 229
million £ (1900)
35 35
30 30
25 25
20 20
15 15
10 10
5 5
0 0
1865 1875 1885 1895 1905 1915 1925 1935
Nevertheless, the real value of nitrate exports for 1929 was only three-quarters
of the 1913 value. With the recession and the long-term effect of the develop-
ment of substitutes, in 1935 the value of nitrate exports went back to a level
one-seventh that of 1913, and this time did not make a significant recovery.
The injection of aggregate demand into the rest of the economy did not
depend on the total value of exports but on the returned value of exports and
the structure of its expenditure. A series of attempts has been made to try to
quantify the loss of surplus from the export sector due to the repatriation of
profits by the foreign companies which controlled nitrates and, later, copper. It
is agreed that on the subject of nitrates gross value of production was divided
into three broadly similar parts: production costs, export taxes and profits
(Mamalakis, 1976; FernaÂndez, 1978). As nitrate entrepreneurs were predomin-
antly foreigners, a significant part of profits ended up outside Chile. Further-
more, one must bear in mind that an important part of the cost of nitrate
production corresponded to rail transport and the provision of water to the
oficinas, either for drinking or for leaching the caliche, and that these activities
were also mainly British-controlled and hence an additional source of remit-
tance abroad. This enormous loss of surplus from the export sector ± estimated
by Mamalakis at 7 per cent of GDP (1976, p. 56) ± undoubtedly substantially
diminished the aggregate demand injection into the local economy.
230 The Chilean Economy and `Dutch Disease'
There has been enormous speculation as to what would have happened to the
Chilean economy if this syphoning off to the exterior of the surplus from the
export sector had not existed; in particular, regarding what would have hap-
pened if this surplus had been returned to the country and invested. Mamala-
kis, for example, hypothesises that `had these resources been used for
investment, the investment coefficient, which ranged from 15 to 20 per cent
of GDP, could have increased by the possible critical amount necessary to
achieve a self-propelling transformation' (1976, p. 60).
However, this type of counterfactual analysis ignores the institutional frame-
work of the Chilean economy at the time. It is one thing to argue (rightly in my
opinion) that if nitrates had been in the hands of national entrepreneurs, these
might well have been able to respond to international demand in the same
degree as foreign entrepreneurs ± at least, there do not appear to be technical or
financial reasons to suggest the contrary.21 It is another, very different, thing
not to realise that given the historical propensities to consume, save and invest
of the national exporting entrepreneurs, it would have been highly unlikely
that the additional portion of the surplus remaining in the country would in
fact have been directed towards investment (rather than consumption).
Furthermore, some doubt must arise as to whether in such a hypothetical
situation the government would have imposed the taxes on nitrate exports
that were in fact introduced, particularly if we take into account what hap-
pened in the previous period, when export activities were mainly in the hands
of national entrepreneurs.22 Thus, public investment would never have reached
the high levels actually achieved during this period (see later, p. 244 ff ).
Finally, we should not forget the difficulties that attended the marketing of
copper exports before the War of the Pacific, when they were not in the hands
of British companies ± a similar problem could have been faced by nitrate
exports if these had been in Chilean hands.23
All this leads to three conclusions regarding the loss of a substantial part of
the nitrate surplus: first, that there is no convincing evidence that the presence
of British entrepreneurs in the nitrate industry was a `necessary' factor for the
development of this industry; that is, other than in relation to commercializa-
tion, the nitrate industry does not seem to have required North's oligopolistic
control, either for its technological development, the financing of its invest-
ment, or the development of its infrastructural works. Second, given all this,
the large repatriation of profits by British entrepreneurs could hardly be justi-
fied as payment for a net contribution made by these entrepreneurs. Third,
although there can be little doubt that any use to which the lost surplus
might have been put would have been more positive for the local economy ±
including its consumption by the local oligarchy ± than its direct loss,24 it is not
particularly productive to speculate nostalgically (as much of the literature
does) upon the hypothetical situation which would have prevailed if all had
Gabriel Palma 231
remained unchanged bar the assumption of national control, and if all the
surplus had been invested locally.
Returning to the indirect linkages of the nitrate industry, what matters from
the point of view of the analysis of the injection of aggregate demand is not
only the size of the surplus that did return to the country, but also the way it
was spent; in particular, whether this returned surplus was spent in its entirety ±
either in consumption or in investment ± and whether the private or the public
sector had a larger propensity to save than to invest. One of the causes of the
`Dutch Disease' is that it is often difficult to absorb all the additional income
generated by a rapidly expanding export sector, especially its surplus compon-
ent; this leads to an excess level of savings ± either in the private or the public
sector. The counterpart of this excess savings is a surplus in the balance of
payments which puts pressure on the rate of exchange to revalue.
The available evidence shows that the part of the nitrate surplus that was
appropriated by the government was spent in its entirety. In fact ± as we will
discuss in detail later (p. 234 ff) ± in 31 of the 37 years between the War of the
Pacific and the First World War the public sector was actually in deficit. We do
not have sufficient information, however, regarding the expenditure of the
surplus returned and retained by the private sector, although we do know
that the oligarchy's propensity to consume was particularly high, and that
though the level of investment made by nitrate entrepreneurs in the nitrate
industry was low, their investment in other economic activities ± especially
agriculture and manufacturing industry ± was significant.
Thus, even though the returned value in the nitrate industry was signific-
antly lower than the total value of its exports, at least the level of the injection
of aggregate demand made by this industry and the expenditures associated
with it in the rest of the economy ± including the government ± seems to have
been similar to that returned value.
The degree of stability of aggregate demand coming from the export sector
As discussed above, exports during the nitrate era followed a strong cyclical
pattern. Part of the reason for this phenomenon lies in the growing instability
of export prices, and the resulting terms of trade and its growing cyclical
pattern are shown in Figure 8.4.
This shows that from the War of the Pacific to the end of the difficult period
through which the world economy passed before 1895±96, the export price
index fell at a similar rate to the estimated import price index; afterwards,
however, the import price index began to fluctuate more. The growing degree
of sensitivity of the prices of Chile's export products to fluctuations in the
international economy (which were already significant in themselves) meant
that even if the volume of the demand injection from the export sector reached
particularly high levels, the growing instability produced an equally significant
232 The Chilean Economy and `Dutch Disease'
1878=100
150 150
125 125
100 100
75 75
50 50
1865 1875 1885 1895 1905 1915 1925 1935
degree of uncertainty in the local economy. For example, from the point of
view of the instability of the terms of trade, the average annual percentage
variation ± in absolute terms ± in this relative price index more than doubled
between 1865±95 and 1896±1929.
Figure 8.5 shows the increase in the annual percentage variation of the terms
of trade from an average of 4.3 per cent per annum between 1865 and 1895 to
one of 10.2 per cent between 1886 and 1929. As Chile exported more than one-
third of its output, the increased instability resulting from terms of trade varia-
tions ± assuming everything else constant ± was equivalent to an average
random shock in Chile's national income of about 1.5 per cent per annum in
the first period and 3.5 per cent in the second.
Moreover, as the stability of the injection of aggregate demand coming from
the export sector depends on the stability of the returned value of exports ± and
not on its total value ± the situation was made even more uncertain by the
changing share of foreign ownership of the nitrate industry. For example,
British ownership of nitrate production in Tarapaca and Antofagasta increased
from 14 per cent the year before the War of the Pacific to 60 per cent in 1895, to
decrease again to 39 per cent before the First World War (Cariola and Sunkel,
1982, table 16).
Gabriel Palma 233
40% 40%
20% 20%
0% 0%
–20% –20%
–40% –40%
1865 1875 1885 1895 1905 1915 1925 1935
taxation approximately half the surplus generated by this activity (or one-third
of the gross value of production: Palma, 1979).
In constant prices of 1900, while export taxes before the War of the Pacific
never reached £100 000, in 1890 they provided more than £2.2 million, to
reach £5.5 million in 1910. After the First World War the revenue from these
taxes began to decline, reaching a level half that of the beginning of the
century. This was not only the result of the fall in nitrate exports, but also
because taxes on the Gran MinerõÂa del Cobre were proportionately four or five
times lower than those levied on nitrate exports.27
Thus the relative contribution of the export sector to public revenues
increased almost ten-fold between 1878 and 1885 ± from 2.7 per cent to 25.4
per cent of `ordinary' public revenues ± while that of all the external sector
(taxes on exports and imports), grew from 44 per cent to 68 per cent. It
continued to grow, coming in 1895 to represent more than 80 per cent of
Chile's ordinary public revenues, then levelled off to around 75 per cent until
the First World War. Thereafter, with the reduction in the volume of foreign
trade, as well as in the level of export taxes, its contribution fell steadily to reach
48 per cent in 1925 and 41 per cent in 1930. As a tesult, internal taxes had to be
increased rapidly from their insignificant levels during the period of booming
foreign trade, when they fell to less than 1 per cent of public revenues.
its deficit financing) was fast enough to help counteract the possible `Dutch
Disease' effect of the rapidly growing exports: that is, whether public expendi-
ture increased at the pace needed to absorb the foreign exchange supplied by the
returned value of exports, so as to avoid revaluation. The other is related to the
quality of investment: whether enough of this public expenditure went into
investment in physical and human capital ± that is, whether it also helped to
increase the country's productive capacity, particularly of non-tradables, so that
domestic output could increase at the pace of expenditure.
The first issue is discussed later (p. 244 ff). As far as the second is concerned, as
a result of the economic crisis toward the end of the 1870s, public expenditure
devoted to investment in physical capital and education was falling in both
absolute terms and in relation to total public expenditure. After the war ±
particularly as a result of the vast plan for public works and education put
into effect by President Balmaceda ± public expenditure on investment in
physical capital quadrupled in real terms in only a decade, while that on educa-
tion increased eight-fold. As a result, the total devoted to investment grew five-
fold in these years ± or at a rate of 17 per cent per year. In consequence, almost
half of public expenditure went on investment in these two areas, and they
reached, in all probability, their highest relative levels in the history of the
country (Palma, 1979). As Blakemore correctly observes,
For this purpose, in 1887 Balmaceda created the Ministry of Public Works and
Industry, and endowed it with a rapidly expanding budget. Education also
received a substantial boost, for while ± according to Blakemore's estimates ±
there were 79 000 pupils in the country at primary and secondary level when
Balmaceda came to power in 1886, there were over 150 000 by the time he was
brought down (1974, p. 87). In the five years following his defeat public
investment fell by 25 per cent, partially recovering thereafter ± in a cyclical
fashion ± to reach about one-third of public expenditure towards the end of the
1920s, with the public works programme of President Iban Äez.
Transmission mechanisms
The effectiveness with which demand from the export sector was oriented to-
wards the internal market depended in essence upon fiscal policy and on the
Gabriel Palma 237
structure of relative prices; the latter, in turn, was a function of the structure of im-
port tariffs, the exchange rate, the relationship between the internal and interna-
tional rates of inflation, and the movement of transport costs. As fiscal policy was
discussed above, this section will study the determinants of relative prices.
The growing political unity of the two powerful economic groups took con-
crete form in 1897 with the creation of the Centro Industrial y AgrõÂcola. This
association immediately declared that `the flag of the Centro Industrial y AgrõÂcola
is the protectionist flag' (quoted in Wright, 1975, p. 52), and began a vigorous
campaign to raise the tariffs on a wide range of products.30
The campaign met with immediate success, for on 23 December of the same
year Law 980 was approved raising the import tariff on a number of agricultural,
livestock and manufactured products competing with local production. There
were four economic developments which helped to shape this event. The first
relates to the return to the gold standard in 1895. This had placed the producers
of tradable goods in a difficult situation because it had been implemented with
an overvalued nominal exchange rate, and with an internal inflation higher
than the international. Between 1894 and 1897 the peso had revalued with
respect to the pound sterling by 22 per cent in real terms. The second relates to
the fast fall in import prices, which that year had reached their lowest level
since statistics on foreign trade were first elaborated (1844). As imports had a
high price-elasticity of demand, both the revaluation of the real rate of
exchange and falling import prices had led imports to grow in real terms by
44 per cent in just two years ± producing in the process one of the few trade
deficits during the nitrate cycle, and a large deficit in the current account.
The third development relates to the fact that the return to the gold standard
with an overvalued nominal exchange rate and relatively high internal inflation
had also strained the internal macroeconomic equilibrium. The government
was obliged to reduce its deficit ± by 1896 30 per cent of public expenditure ± by
reducing its expenditure in 1897 by 20 per cent. In these circumstances, the
government obviously saw as positive the possibility of raising its revenues
through an increase in import tariffs. Finally, the protectionist lobby found an
unexpected ally in the oreros (those in favour of the gold standard); the political
accord which had brought the gold standard back after 20 years of inconvert-
ibility was very fragile, and those in favour of this regime were forced to support
the change in tariffs as a way to consolidate the political support for the gold
standard ± even though, given their monetarist view of the balance of pay-
ments, they would normally have been against a tariff increase of this nature.
The law placed ad valorem taxes on a wide range of agricultural and manu-
facturing products and a fixed tax on 28 products, including cattle.31 The
increase in tariffs was so drastic and demand for imports was so price-elastic
that the real value of imports fell by 30 per cent in just one year. Furthermore, as
exports continued to grow, the trade deficit of 1897 was reversed the following
year into the largest trade surplus the country had seen.32 Afterwards imports
did recover, but in real terms it took them a whole decade to return to their
1896 value; meanwhile, exports grew by 80 per cent. This asymmetric beha-
viour generated a sustained trade surplus.
Gabriel Palma 239
It is important to note in this context that the structure of the new import
tariffs ± in particular the low level of the tariff, or the exemption from it, for a
wide range of intermediate inputs and machinery ± and the fall in imports, led
to a relatively small increase in public revenues. What was substantial was the
increase in the ratio between revenues from taxes on imports and the value of
imports. All this is another indication that the main aim of the 1897 law was
not so much an increase in public revenues ± welcome though this might have
been ± but an increase in effective protection for agriculture and existing
branches of manufacturing. Subsequent changes in tariffs confirm the exist-
ence of an industrial policy and a trade policy of selective protection.
For example, one of the additional changes in tariffs, which took place the
very next year, 1898, increased the fixed duty on refined sugar. The higher
duty permitted local producers to raise the internal price, but the ferocity of
the resulting popular reaction forced the government to reverse its original
measure ± reducing this time the tariff on imports of refined sugar by
32 per cent. However, as the government also reduced the import tariff
on unrefined sugar by 50 per cent, it was able to achieve two goals simul-
taneously: to please consumers by forcing a reduction in the internal price
of sugar, and to appease the refiners by increasing their effective protection ±
as the tariff on unrefined sugar fell by significantly more than that on the
final product. Similar changes took place in regard to textiles and cotton
thread.
Later, in 1916, a further general revision of import tariffs took place, which,
according to a report from the Superintendent of Customs, `raised the import
tariff on ``essential'' goods by 50 per cent, other manufactured foods by 86 per
cent, and various manufactured goods by an even greater proportion' (quoted
in Wright, 1975, p. 56). The US Federal Trade Commission calculated that
taking into account import duties and other customs charges, American-
preserved peaches, for example, were now paying a tariff equivalent to 250 per
cent of the CIF value of this product (US Federal Trade Commission, 1916, p. 52;
quoted in Wright, 1975, pp. 56±7).33
The tariff structure was revised again in 1921, 1925 and 1928, but as a
result of the already mentioned fall in export taxes after the First World War,
these industrial and trade tax changes had fiscal considerations as well as
forming part of a policy of giving priority to local production of inputs and
machinery, while the tariff changes of 1897 and subsequent years had favoured
the producers of final consumer goods.
However, the Sociedad Nacional de Agricultura was still not satisfied despite all
the increases in the tariff on the importing of competing agricultural and
livestock products; it also pressed for more `effective' protection from foreign
competition, by preventing the construction of the Chilean stretch of the
railway from Salta to Antofagasta, in order to keep Argentine products out of
240 The Chilean Economy and `Dutch Disease'
the nitrate area. As a result, the Chilean government did not complete the
portion of track on its territory until 1940 ± by which time work on the line
had been suspended for two decades.
Despite the nature of the 1897 law, that of the subsequent revisions of the
import tariff and the widespread political support for increased protection ± in
particular the agro-industrial alliance that brought about the 1897 law ± most
writers on this period persist in the claim that the nitrate-led expansion only
consolidated the commitment to free trade, frustrating ever-more decisively the
development of manufacturing in the country. For example, Claudio VeÂliz's
famous `three-legged table' hypothesis insists on the validity of the theory of
the permanent conflict between the economic interests of landowners and
manufacturers: the former in favour of free trade, the latter for protectionism
(VeÂliz, 1963, p. 104). Other authors have also argued that the level of import
tariffs before the 1930s crisis was very low and that its structure was not
designed to stimulate specific productive activities. In their opinion, the rapid
expansion of nitrate exports, and the related growth of public revenues, gave a
final blow to those who tried to develop local manufacturing industry through
protection. Thus, for example, AnõÂbal Pinto dismisses the protectionist inten-
tions of the 1897 law as `a timid and temporary fit of protectionism' (Pinto,
1959, p. 40). Teresa Janneret, in turn, wrongly states that the tariff policy in
Chile before 1930 `did not discriminate between different sectors of produc-
tion' (1972, p. 64), and that `protection in Chile [before 1930] was limited to
general and relatively modest tariffs on imports (25 to 30 per cent)' (ibid., p. 70;
see also Hinkelammert, 1970, p. 62). Finally, two historians of very different
ideologies ± Encina, a conservative, and RamõÂrez Necochea, a communist ±
make the same claim (1942±50 and 1958, respectively).
Although it is obvious that the tariff structure after the crisis of the 1930s
increased the degree of discrimination and selectivity between different eco-
nomic activities, it is equally certain that it did not initiate the process. The
heterogeneity of the tariff structure brought about by the 1897 law and sub-
sequent revisions are a clear indication of the more selective approach adopted;
and the levels of the tariffs on many imported products imposed by the 1897
law (60 per cent, or even more in the case of many specific tariffs), can hardly be
called `timid' or `modest'. Obviously, this is not the right way of labelling a law
which resulted in huge increases in effective protection, a drop of almost 30 per
cent in imports ± a fall that took a decade to be reversed ± and which produced a
record surplus in the balance of trade. Also, subsequent changes in the tariff
structure clearly show that the law was not a `temporary fit of protectionism'.34
`Dutch Disease' phenomenon is rapid export growth which tends to revalue the
real rate of exchange and, through this, to move relative prices against tradable
production. As Figures 8.6, 8.7 and 8.8 show, available evidence suggests that
this effect did not occur during the nitrate cycle.
The average real exchange rate during the nitrate cycle (1878±1914) was 13.9
pesos to the pound sterling, with a standard deviation of only 1.17 pesos; this
stability is surprising, since this period includes not only the initial export
boom but also one foreign and one civil war. The stability of the rate of
exchange can also be seen in the fact that in 1913±14, at the end of the nitrate
cycle, the real rate (12.7 pesos to the pound sterling) was almost the same as
before the War of the Pacific, and this in spite of the fact that in real terms
exports had increased seven-fold. Even in the first decade of this cycle, when
exports tripled, the real rate of exchange remained stable.
Furthermore, the only years where there was some variation correspond to
specific periods of short duration, such as the two wars and the unsuccessful
return to the gold standard between 1895 and 1898, years which included the
1897 tariff reforms and the resulting sharp drop in imports. Other ways of
measuring the real rate of exchange, such as the ratio between tradable and
non-tradable prices, show a similar picture.
If the price indices for tradable and non-tradable goods used in Figure 8.7 are
an accurate reflection of what happened to prices of these goods and services,
the real rate of exchange (defined in this way) demonstrates, first, a revaluation
50 50
40 40
30 30
20 20
10 10
0 0
1865 1875 1885 1895 1905 1915 1925 1935
Figure 8.6 Chile: real rate of exchange, 1865±1935 (pesos per £ at constant 1900 prices)
Source: Rate of exchange, Palma (1979).
244 The Chilean Economy and `Dutch Disease'
1880=100
200 200
150 150
100 100
50 50
0 0
1865 1870 1875 1880 1885 1890 1895 1900
Figure 8.7 Chile: real rate of exchange, 1865±1900 (prices of tradables divided by those of
non-tradables)
Note: A fall in the index indicates a revaluation. The prices of tradable goods are measured as the
average of the indices of `British parity' (Wagner) and of wholesale agricultural products (Latorre,
recalculated by Wagner); the prices of non-tradable goods are measured by the index of Clavel and
Ferraro (which is available only until 1900). In the figure, 1880 is shown as the base year (and 1889,
the beginning of the civil war, as a reference point) because relative prices in both 1879 and 1891 were
badly affected by the two wars which the country experienced during this period.
Source: Wagner (1992) and Clavel and Ferraro (1990); see also end-note 35.
1880=100
200 200
150 150
100 100
50 50
0 0
1865 1870 1875 1880 1885 1890 1895 1900
Figure 8.8 Chile: real rate of exchange, 1865±1900 (prices of tradable divided by nominal
wages)
Note: A fall in the index indicates a revaluation.
Source: Prices of tradable goods are measured as in Figure 8.7; wages, according to Rojas' index (1982).
Gabriel Palma 245
during the five year-period of Santa MarõÂa (when the index dropped by 25 per
cent), only to recover almost entirely during Balmaceda's government (until
the beginning of the civil war) ± in 1889 this relative price index reached a value
95 per cent that of 1880. That is to say, Balmaceda's fiscal, commercial and
industrial policies seem to have managed to reverse almost all of the earlier
drop in relative prices between tradables and non-tradables.
Another way of measuring the real rate of exchange involves using nominal
wages as an indicator of prices of non-tradable products. The resulting real rate
of exchange is shown in Figure 8.8. According to this definition, the real rate of
exchange did not revalue during the decade of high initial export growth; it did
not even experience the initial revaluation during the five years of Santa Maria
shown in the previous index. Furthermore, towards the end of the century this
real rate of exchange was devalued significantly.
The evidence available thus far shows that by the end of the first decade
of the nitrate cycle, when exports tripled, the real rate of exchange ± by any of
its definitions ± had a value similar to that for the beginning of the cycle.
This relative stability of the real rate of exchange requires more detailed
analysis, as export growth of such magnitude would normally produce a
significant surplus in the balance of payments which would put pressure
on the real rate of exchange to revalue. This is due to the fact that although
growth in exports brings about an increase in imports (through its effects
on income), this increase tends to be lower than that in exports ± unless there
is a revaluation of the real rate of exchange or a compensating increase in
expenditure.
For example, given the usual assumptions of the theory of the Keynesian
multiplier (including constant coefficients) in an economy without a public
sector and an autonomous private investment, a growth in exports increases
income by an amount equivalent to this growth in export multiplied by the
multiplier [1/(s m)] (where s is the average propensity to save and m the
average propensity to import). The increase in income, in turn, raises imports
by an amount equivalent to this growth in income multiplied by the propensity
to import; this amount is equivalent to the initial growth in export multiplied
by [m=
s m]. As this last multiplier is necessarily less than one, the resulting
increase in imports is always lower than the growth in exports.
If the public sector is introduced into the analysis through autonomous
public expenditure and income equivalent to a proportion t of income, the
difference between export growth and import growth is even more marked ± as
direct taxation on income introduces a new `leakage' to the circuit of income
and expenditure. In this case, the multipliers become [1=
s m
1 t t] and
[m=
s m
1 t t] respectively; the second multiplier is not only less than
one, but lower than in the earlier case ± so the difference between the growth in
exports and imports is even larger.
246 The Chilean Economy and `Dutch Disease'
That is, if private and public investment are autonomous, the mechanism
which adjusts income when faced with an export expansion is not able to
generate equilibrium in the balance of payments ± the `income effect' generated
by the increase in exports is not sufficient to maintain the internal and external
equilibria (that is, between the increase in income and that of absorption and
between imports and exports).
In turn, if investment ± private and public ± is not autonomous but rather
induced, the final equilibrium would depend on the value of the accelerator.
One apparently paradoxical result of this multiplier±accelerator equilibrium is
that even if the accelerator induces an increase in investment large enough as
to absorb all the savings generated by the growth in exports, the resulting
increase in expenditure would still not be sufficient to achieve macroeconomic
equilibria ± and thus avoid a revaluation. In order to achieve this equilibrium
(without revaluing) it would be necessary to have an additional (compensatory)
increase in expenditure (over and above the one induced by the previous
accelerator), and so, at least initially, a deficit expenditure either in the public
or in the private sector.38 This deficit expenditure was precisely what fiscal
policy provided. Without this, a primary-export boom would have tended to
generate an insufficient increase in expenditure and thus a surplus in the
balance of payments, which would have put pressure for a revaluation of the
real rate of exchange. The role of this revaluation would have been to recover
internal and external equilibria through an inducement to increase both
imports and expenditure ± that is, it would have added a `price effect' to the
insufficient (export-led) `income effect' on imports.39
This phenomenon can be seen more clearly through the macroeconomic
relation between the surplus ± or deficit ± in the current account and that
between savings and investment. The income and expenditure accounts yield
the following relationship between both surpluses (or deficits):
X M
S Ipr
S Ipu ], where X is total exports, M is total imports
(including the repatriation of profits and the servicing of the foreign debt),
S Ipr is the difference between savings and investment in the private sector,
and
S Ipu is the same for the public sector ± equivalent also to the difference
between income and expenditure in each sector. If exports grow more rapidly
than imports (including factor payments), it is because investment is growing
less rapidly than savings (or expenditure less rapidly than income) whether in
the private or the public sector. One way of reducing both surpluses is to make
imports increase more rapidly than the level induced by the growth in export
(via the `income effect' of the latter). This can be achieved either by a compens-
atory increase in expenditure, or through the `price effect' of a revaluation, or
through a combination of the two.40
What is unusual about the case of Chile in this period, as shown in the earlier
graphs and in Figure 8.9, is that the solution to the above-mentioned problem
Gabriel Palma 247
1878=100
1000 1000
800 800
600 600
400 400
200 1 200
2
0 0
1865 1875 1885 1895 1905 1915 1925 1935
Figure 8.9 Chile: exports and imports, 1865±1935 (at constant 1900 prices)
[1] Imports [2] Exports
balance of payments. After the tariff reforms of 1897, this tendency was
reversed for another decade ± both due to a slower growth of imports and
a recovery in exports. In the following decade, however, the opposite pheno-
menon appeared.
Adding together these cycles one can see that what really characterized the
nitrate cycle was the fact that instead of producing a surplus in the current
account and pressure for a revaluation of currency, and domestically an excess
of savings, the cycle was associated, though unstably, with a deficit in the
current account, a shortage of internal savings, an increase in absorption larger
than that of income, and relative stability in the real value of the currency.
Figure 8.10 looks at one of the key aspects of this phenomenon, comparing the
growth in public expenditure with that of exports.
Figure 8.10 shows how fiscal policy caused public expenditure to grow,
though unstably, at a comparable rate to that of exports between the War of
the Pacific and the First World War. In fact, this expenditure grew somewhat
faster than exports. Furthermore, at the crucial moment when the nitrate
industry was incorporated into Chile's economy ± that is, when the danger of
the `Dutch Disease' was at its greatest ± between 1878 and 1891 public expen-
diture increased 42 per cent faster than exports. As public expenditure in 1878
was equivalent to three-fifths of exports, this faster growth rate of public
expenditure implies that, in terms of absolute value, this expenditure grew by
1878=100
1200 1200
1000 1000
800 800
600 600
400 400
200 1 200
2
0 0
1865 1875 1885 1895 1905 1915 1925
Figure 8.10 Chile: public expenditure and exports, 1865±1930 (at constant 1900 prices)
[1] Exports [2] Public expenditure
Gabriel Palma 249
50% 50%
25% 25%
0% 0%
–25% –25%
–50% –50%
1870 1880 1890 1900 1910 1920 1930
It is obvious that the fiscal policy of this period generated some of the
negative effects which Meller identifies ± which can be seen, for example, in
the public sector's difficulty in adjusting its expenditure after Balmaceda when,
as a result of lower growth of exports, the high levels of public expenditure
created an excess of absorption and a deficit in the current account. In turn,
public revenues became overly dependent on a very unstable external sector.
However, this fiscal policy also had the positive effects already analysed. A
further effect, ignored by critics, was that by eliminating internal taxes it helped
to increase the value of the multiplier; and, by doing so, helped to maximize
the growth of income arising from an increase in exports ± as from other
demand injections such as public expenditure and private investment. We
have already mentioned that the multiplier of an autonomous growth of
exports in such an economy is [1=
s m
1 t t]; however, in an economy
without a public sector, this rises to [1=
s m], as one of the income leakages
(t) is eliminated. What seems paradoxical, however, is that this higher multi-
plier for an economy without a public sector is also relevant for one with a
public sector, but in which its expenditure is financed (in Meller's words)
`through the taxation of foreigners (who control exports)'; that is, through
254 The Chilean Economy and `Dutch Disease'
taxes which are outside the income-expenditure circuit. So although this fiscal
policy did generate institutional problems, it also had its successes. It increased
national income ± forcing foreign entrepreneurs to return about half of their
earnings to the country, increasing public investment in physical and human
capital, and increasing the value of the multiplier. It also increased absorption
(an essential factor in avoiding `Dutch Disease').46
Thus approximately a third of nitrate exports ± equivalent to the cost of
production ± was absorbed through the productive process, and another third
(taxes on exports) through fiscal policy (public expenditure and transfers to the
private non-export sector). As for the remaining third, profits after taxes, the
share that went to the national nitrate entrepreneurs ± or foreigners who had
become resident in the country ± was absorbed in consumption or investment;
while the share taken by foreign entrepreneurs who repatriated their earnings
was absorbed, from a balance-of-payments point of view, by this process of
repatriation.
All of the above helps to explain how in this period in Chile it was possible to
combine a particularly rapid increase in primary commodity exports with
development in manufacturing. It is well-known that the income elasticity of
expenditure on manufactures is greater than that of agricultural and non-
tradable products. However, from a `Dutch Disease' point of view, it is not a
question of how the aggregate demand for manufactures increases, but of how
the demand for nationally-produced manufactured goods increases. Apart from
the quality of these products, this depends mainly on relative prices and the
level of expenditure and its structure. We already know that on average absorp-
tion grew more rapidly than income, that there was no revaluation of the real
rate of exchange and that import tariffs increased for a wide range of manufac-
tured goods. As a result, in spite of domestic inflation and the drop in transport
costs, the evidence shows that effective demand for nationally-produced man-
ufactures grew at least at a rate more or less similar to that of income. Studies of
manufacturing production for this period tend to confirm this fact (Kirsch,
1973; Palma, 1979 and 1984). In sum, the demand-side evidence, as well as
the supply-side, shows that national manufacturing production was stimulated
by the nitrate boom: exactly the opposite of what one would expect if this had
been a case of `Dutch Disease'.
This also helps us to understand why, contrary to what is usually argued (see
for example Behrman, 1976) the extraordinary expansion of exports did not
cause a shrinking of traditional exports, at least for the first decade of the nitrate
cycle. Although wheat exports fell during the following decade, this was not the
case for wheat production. Production of other crops also grew significantly, as
in the case of barley, beans, potatoes, peas and chickpeas.47 In the next decade,
wheat production practically doubled and for a few years exports again reached
the level of the end of the 1880s.
Gabriel Palma 255
Copper production after the end of the first decade of the nitrate cycle fell
gradually at the beginning of the twentieth century, to recover slowly after-
wards. So the beginning of the Gran MinerõÂa of copper in 1912 found this
industry with an export level similar to the figure for 1888 ± and the second
half of the 1870s. However, studies show that this stagnation was also not
directly due to the nitrate boom ± that is, to problems of relative internal prices
± but to a growing technological backwardness, because in Chile the copper
industry had been unable to follow the particularly rapid technological devel-
opment which characterized this industry internationally from the end of the
nineteenth century (Reynolds, 1965; Culver, 1982; and Mayo, 1982).
To sum up this section: the growth in nitrate exports was linked to a rate of
growth of absorption at least similar to, if not higher than, that of income. It
did not generate surpluses in the current account or in national savings, or a
real revaluation of the peso. It also saw the growth of effective demand for
nationally-produced manufactures at a rate similar to that of income, and did
not cause a mechanical shrinking of traditional exports. The axis for this
phenomenon was fiscal policy, which taxed exports heavily and injected a
high level of public spending ± with a strong investment component ± into
an economy characterized by an elastic supply of labour and non-tradable
goods.
Conclusions
We have seen how the nature of production in the export sector, fiscal policy
and the relative price structure were essential factors determining how effect-
ively demand from the export sector was oriented towards the domestic mar-
ket. Concerning the first of these factors, the production process of nitrates
and, later, that of copper, had a low level of direct linkages ± forward as well as
backward. For this reason, neither nitrates nor copper greatly helped the pro-
cess of absorption, nor did they provide a significant stimulus to other product-
ive activities. The exceptions to this rule are the railway network for the mining
area, the development of infrastructure related to both products and, in the
case of nitrates, agriculture (and related services). As far as the second factor is
concerned, the fiscal policy of `tax, transfer and spend', did help to generate the
kind of external and internal macroeconomic equilibria necessary for avoiding
Dutch Disease, especially during the first decade of the nitrate boom. Regarding
the third factor, the relative price structure ± the main factor which determines
the export sector's indirect linkages ± we know that, of its four principal
components, two operated in one direction and two in the other. The nominal
rate of exchange and the level and structure of import tariffs provided an
incentive for the orientation of aggregate demand towards the domestic
markets. Meanwhile, the other two factors ± the large gap between internal
256 The Chilean Economy and `Dutch Disease'
and external inflation and the reduction of transport costs ± operated in the
opposite direction. Furthermore, it seems that until 1913 the difference
between the internal and external rate of inflation was similar to that of the
depreciation of the peso. The final effect depended on the balance between the
reduction in transport costs and the increase in the level of import tariffs and
the change in their structure. Although the available empirical evidence is
insufficient, we can conclude that during the nitrate cycle producers of tradable
goods did not have to face either the economic policy or the changes in relative
prices that one would expect in a rapidly expanding export economy. Therefore
they did not suffer Dutch Disease. This conclusion is confirmed by the estim-
ates for manufacturing production in this period.
Later, between the First World War and the return to the gold standard in
1925, producers of tradable goods benefited from the rapid real devaluation of
the peso, only to be harmed by the rapid revaluation which took place during
the short period of convertibility ± 1925±31. They were also harmed by having
to face two very different exchange rate polices in a short period.
One striking aspect of the policies of this period was that the rapid growth in
public revenues from high taxes on the export sector did not induce a reduction
in tariffs on those imports in competition with domestically produced pro-
ducts. On the contrary, the abundance of public revenues enabled the imple-
mentation of a tariff structure which increased effective protection for many
tradable goods. Together with an increase in tariffs on these products, it allowed
a reduction in tariffs on the inputs and machinery necessary for their domestic
production, in spite of the drop in public revenues that this implied. This
orientation of tariff policy gradually produced a horizontal rather than a ver-
tical structure of manufacturing. The conditions which favoured producers of
consumer goods worked against producers of manufacturing inputs, and until
the First World War fiscal, commercial and industrial policies favoured the
former at the expense of the latter. Later, after the crisis in the external sector
and the country's political and ideological changes, these policies changed,
increasingly favouring inputs.
As far as agriculture is concerned, the cost of the tariff policy was the reinfor-
cing of attitudes which were anti-export and anti-specialization. However,
indiscriminate protection of agriculture seems to have been the necessary
price for the support of rural producers for a policy of industrial protection.
Finally, since the policies protecting both sectors, manufacturing and agri-
culture, were not conditioned to increases in productivity, or limited in time,
their effect, given the reduced size of the domestic market, was a significant
reduction in the level of competition faced by producers. Protection thus
eliminated perhaps the most important factor necessary for the efficiency of
the market: the discipline which only competition can provide. So Chile
escaped Dutch Disease but not without a cost.
Gabriel Palma 257
Notes
1 The fact that the nationalization of Peruvian nitrates affected British investment and
that political relations between Bolivia and Great Britain went through a period of
tension, has led to a general tendency to overestimate the role played by Britain in
the war, and to pay insufficient attention to the political and economic problems of
Chile itself which lay behind it. This view is expressed very bluntly by the US
Secretary of State James G. Blaine, who stated that the War of the Pacific was `an
English war on Peru, with Chili [sic] as the instrument . . . Chili [sic] would never have
gone into this war one inch but for the backing of English capital' (quoted in Sater,
1990, p. 39). Among the numerous accounts of British involvement in the War of the
Pacific, see Collier (1993), Dennis (1931), FernaÂndez (1978), Ortega (1979) and Sater
(1990).
2 Unless otherwise stated, all data in this chapter have as a source the statistical
appendices in Palma (1979). Figures in sterling are at 1900 prices, and all rates of
growth are expressed in real terms.
3 When the crisis began in 1873 import taxes made up some 51 per cent of what was
called `ordinary' public revenues, and export taxes contributed another 4 per cent; the
rest was made up of the revenues from public enterprises (32 per cent), internal taxes
(11 per cent), and other smaller taxes (2 per cent). `Extraordinary' public revenues were
made up of external and internal loans which financed the fiscal deficit.
4 For an analysis of the literature of the period, see Ortega (1979, ch. 6).
5 See Archives of A. Gibbs, ValparaõÂso to London, 28 December 1878.
6 The day following the passing of this law, President Pinto wrote in his diary: `the
banks would have gone broke that same day if inconvertibility had not been
declared', quoted in Ortega (1979, p. 453).
7 See Dennis (1931, especially ch. 3). The Bolivian government needed this tax because
its finances were going through an extremely difficult period since the effects of the
international economic crisis were exacerbated by the disastrous effects of the recent
civil war. It has been argued that the main reason for the dictator Daza deciding to
increase public revenues in violation of the treaty was as a result of his perception
that Chile was in no position to go to war both because of its own serious internal
economic problems and because of its border tensions with Argentina ± which had
recently obliged them to give Patagonia up to Argentina and to agree to joint control
of the Magellan Strait.
8 Corden and Neary called these two types of linkages between the export sector and
the rest of the economy the `movement of resources' effect and the `expenditure'
effect (1982).
9 The price of nitrates fell by 44 per cent just between 1884 and 1888. The end of the
`Victorian Boom' in 1873 was followed by a generalized drop in prices of primary
commodities and manufacturing products in the world economy until 1896±97.
According to S. B. Saul (1976), this generalized drop in prices over 14 consecutive
years was the most significant economic phenomenon of the nineteenth century.
10 1919 was a particularly difficult year for the Chilean economy. The value of exports,
in real terms, fell to about one-third the 1918 levels, with the attendant effect on
public revenues from this source. This fall in export revenues was owed to the fact
that the price of nitrates fell by 17 per cent that year and the quantum exported by
more than two-thirds. The main reason for this decline was the fact that European
countries and the USA had accumulated large stocks of nitrates during the war, and
258 The Chilean Economy and `Dutch Disease'
these stocks were being used in 1919. Exports to the USA in 1919, for example,
reached only one-fifth of the 1918 level ± 370 000 tons and 1.7 million, respectively
± while those to Britain were equivalent to just 5 per cent of the 1918 volume (25 000
tons and 535 000 tons, respectively).
11 Between 1913 and 1923 the share of Chilean nitrates in world production of nitro-
genous fertilizers fell by almost half ± from 55 per cent to under one-third. This rapid
drop and the problems with the price of nitrates caused the Chilean nitrate industry
to negotiate the formation of a cartel with the world's two largest producers of
synthetic nitrates, I. G. Farbenindustrie and Imperial Chemical Industries. These
three producers controlled 70 per cent of world nitrogen production. US producers,
which controlled around 10 per cent of production, were unable to become a part
of this cartel because of their country's anti-trust legislation (see Keynes, 1985, pp.
565±7).
12 For many economic historians what was exceptional about the crisis in the 1930s was
not so much its initial magnitude but its long duration. The world economy had
already got through various abrupt falls in prices relatively quickly ± one of which
took place shortly before the 1929 crisis (1921); what was unusual about the 1929
crisis was that prices not only failed to recover, but they actually continued to fall
throughout the world for many years. For an analysis of this phenomenon and of its
possible causes see specially Kindleberger (1984).
13 In the case of copper its initial increase was also followed by a period of extreme
instability and eventual relative stagnation after 1929. Thus in 1954, 25 years later,
exports from the Gran MinerõÂa ± after having being through a cycle of abrupt swings ±
were just 12 per cent higher than in 1929. This aspect of the copper export cycle
mirrored the characteristics of early export cycles, except that in this case it was
necessary to wait four decades for the appearance of new export products to revitalize
the export sector ± in this case agricultural, wood and fish products.
14 In the period from 1929 to the second decade of the 1950s, only 1942 shows the
quantum of exports higher than those of 1929. This is not to overlook, of course, the
problems inherent in quantum indices (Laspeyres) when used over periods as long as
those studied here.
15 The nitrate industry was even practically self-sufficient for one of its principal inputs,
explosives; these were necessary for the removal of the caliche and were produced
from the nitrates. The industry was also self-sufficient for one of its fuels, wood from
the tamarugo tree, although indiscriminate felling of these trees led to massive
ecological devastation.
16 The majority of nitrate mines paid with tokens instead of money. These could only be
exchanged for food and other essentials in the mines' own pulperõÂa. For this reason,
labour conflict of this period ± some of which were repressed with particular brutality,
like the massacre in the Santa MarõÂa de Iquique School ± were characterized not only
by the amount of tokens paid as salary, but also by the workforce's desire to improve
the `rate of exchange' between tokens and food.
17 In 1835, after visiting the nitrate deposits in TarapacaÂ, Darwin wrote: `The chief
expense is its transport to the sea coast. The mine consists of a hard stratum, between
two and three feet thick, of the nitrate mingled with a little of sulphate of soda and a
good deal of common salt. It is close beneath the surface and follows for a length of
150 miles the margin of a grand basin or plain' (1845, p. 41).
18 These ironically proved very useful in the Chilean advance into TarapacaÂ. After the
war, Colonel North, the `Nitrate King', incorporated them into his nitrate empire,
when in April 1887 he purchased 58 per cent of their shares.
Gabriel Palma 259
19 To get an idea ± in relative terms ± of the quantity of resources which were allocated
to this work, it may be noted that relative to population the Chilean network was 2.8
times larger than Great Britain's at this time.
20 The fall of exports was so extraordinary that, according to a League of Nations study,
the 1929 Depression affected the Chilean economy not only more than any other in
Latin America (as was the case for volume exported), but more than any other in the
world. See Ellsworth (1945, p. 23).
21 Even the financial bases upon which North constructed his nitrate `empire' were of
local origin; it was funds provided by the Iquique branch of the Bank of ValparaõÂso
which allowed North to purchase the `keys to his kingdom', the famous nitrate
certificates which gave him control of the oficinas Primitiva, Peruana, RamõÂrez,
Buen Retiro, Jazpampa and Virginia, among others. For a detailed account of North's
manoeuvres, see Blakemore, (1974).
22 From 1840 onwards agricultural and livestock exporters paid no export taxes at all,
and of their connected processing activities only leather paid a 6 per cent tax, while
mining exports also came to pay a maximum 6 per cent ± this being the rate paid by
processed copper. This was the reason why during the 1870s, for example, export
taxes were, on average, equivalent to just 1.4 per cent of the value of exports, and that
they only contributed 3.3 per cent of `ordinary' public revenues. See Palma (1979).
23 For the problems faced previously by Chilean exporters of copper, particularly in the
British market, see Palma (1979).
24 To be spent financing North's rapid upward social mobility, his purchase of honorific
titles, his sumptuous gifts ± such as a famous piano for Queen Victoria ± or his lavish
entertaining of the British oligarchy. Blakemore (1974) provides a vivid account of
this side of North's life.
25 For an analysis of economic ideology in Chile since Independence, particularly with
respect to the role of manufacturing in economic development, see Palma (1979).
26 The characteristics of the production costs of nitrates can be found in RamõÂrez
Necochea (1958), Blakemore (1974), FernaÂndez (1978), Palma (1979), Cariola and
Sunkel (1982), and Pinto and Ortega (1991).
27 The taxes applied to exports from the Gran MinerõÂa in this period were only equival-
ent to approximately 7 per cent of the gross value of production. See Reynolds (1965,
pp. 363, 365).
28 Created in 1883, this association of manufacturing entrepreneurs was the successor of
the Junta de Fabricantes.
29 This was a direct response to the opening of the Trans-Andean Railway through the
Uspallata pass. As a result of this new railway, import of cattle had increased by 75 per
cent in just one year (to 74 000 head). For a justification of the need for this tax, see
Sociedad Nacional de Agricultura, 1886, p. xvii. For an analysis of the situation, see
Wright (1975. pp. 50±1).
30 For an interesting study of the influence of Litzen's ideas on tariff policy in the
period, and on the economic thinking of some politicians, MalaquõÂas Concha in
particular, see GarcõÂa-Huidobro (1972).
31 On 59 of these agricultural and manufactured products, this law placed an ad valorem
tax of 60 per cent; it also placed a tax of 35 per cent on manufactured products which
mainly served as inputs for other products of the same sector. It left a standard tariff
of 25 per cent for products not otherwise specified, and divided raw materials and
machinery into three categories, placing 54 at 15 per cent, 66 at 5 per cent, and
leaving 118 free of all import duties. Finally, specific taxes were placed upon 28
products; the taxes in these cases, as Kirsch reminds us, `were usually much higher
260 The Chilean Economy and `Dutch Disease'
than the 60 per cent ad valorem category' (1973, p. 221). For the text of the law see
Anguita, 1912 (vol. III), pp. 400±7. For an analysis of the political situation which led
to the passing of this law, see Eyzaguirre (1957). For an analysis of the law, see
Sociedad de Fomento Fabril (1901) and Ho È rmann (1901).
32 Imports of cattle, for example, which had reached an average of over 100 000 head in
1895±96, dropped to just 43 000 in 1898 and less than 28 000 in 1899.
33 The protection of the preserved food industry seems to have stimulated local produc-
tion to such an extent that Chile became an exporter of these products. For example,
NicolaÂs Rubio's FaÂbrica Nacional de Conservas Alimenticias (located in his hacienda El
Estanque) exported a significant proportion of its output of two million tins of
preserved foods both to Europe and Latin America; see Palma (1979, p. 297).
34 One criticism of the tariff policy before the First World War that could be made is that
in protecting manufacturing activities producing mainly final consumption goods, it
discriminated against the production of inputs and machinery; but this would be a
criticism against the specific type of protectionism followed at the time and not
against the existence of a protectionist policy in itself.
35 There are several estimates of the movement in the level of prices of the period (for an
analysis of them, see Wagner, 1992). From these indices we have constructed a new
one which is the average of three indices: that of wholesale agricultural prices of
Latorre (expanded by Wagner); Wagner's `British parity' index (which measures the
changes in the nominal exchange rate and prices in Great Britain); and Clavel and
Ferraro's 1990 index (which measures consumer prices, mainly of food, household
items and clothing). In our opinion our index is a better approximation of the general
level of domestic inflation than the individual indices, as it includes a wider variety of
products and services, and because it averages different biases in previous indices (the
first two indices mentioned have a wholesale bias and the third one a consumer-
price-index bias).
36 There is an extensive literature on some aspects of this inflationary phenomenon
in Chile, in particular as regards the monetary anarchy which prevailed through-
out (see, for example, Hirschman, 1963). The growing need to impose some order on
the monetary chaos of the country led to the recourse in 1924 to the famous
Kemmerer Mission; this created the Central Bank and engineered the return
to the gold standard in 1925 (see Fetter, 1931). For the legendary controversy between
oreros (the proponents of the system of convertibility of the peso) and papeleros (their
opponents), see GarcõÂa-Huidobro (1972), Bustos (1988) and Drake (1989).
37 As already mentioned, as a result of the War of the Pacific, Chilean exports trebled in
real terms in just a few years. To put this into perspective, this expansion of exports is
similar to the one oil-exporting countries had in 1973±74. In the debate of the effect
of the discovery of oil on the North Sea in the British economy, an important
distinction was made between the possible outbreak of a `Dutch Disease' resulting
from a primary-export expansion based on a product already exported by a country
and that, as in Great Britain, based on a new export product. Although in Chile the
export boom was not caused by the `discovery' of a new product, the Chilean case
resembles the British in that it was based on the incorporation of one into the
national economy. See, for example, Worswick (1980), Byatt et al. (1982) and Pesaran
(1984). For studies of export booms contemporary to nitrates in other Latin American
countries, see Baham and Coomes (1995).
38 For a proof of these points, see Palma (1998).
39 The tendency to revalue the real exchange rate could be made worse by speculative
pressures on the currency based on good prospects for exports and the balance of
Gabriel Palma 261
payments (see Corden, 1977). However, the fact that national and international
financial markets at the time were not properly developed reduced the possibility
of this situation taking place.
40 Added to this, of course, is the fact that throughout this period the Chilean economy,
as with most economies in Latin America, was increasing its degree of integration into
the world economy. This process was mostly the result of technological and institu-
tional changes. This long-term process was increasing its income elasticity to import, a
process which also helped to absorb additional foreign exchange produced by nitrates.
41 As already mentioned, following the international financial crisis of 1873 it was not
possible to get new foreign loans; in fact, the only one which was possible to acquire
between 1873 and the War of the Pacific in 1875, was due to the fact that it was
necessary for the amortization of an already existing loan ± secured in 1867. As a
result, despite the crisis, the country had to generate a surplus in its trade account in
order to accumulate the foreign exchange needed to service the existing foreign debt.
42 Yet another aspect of public expenditure which proved to be very conflictive, parti-
cularly with the rural oligarchy, was its effect on the labour market in the country-
side. Until then, wages in the agricultural sector were not only extremely low but
were paid in kind ± as part of a system that tied the peoÂn to the hacienda. As a result of
the rapid growth of public expenditure, particularly in infrastructure, demand for
labour in the rural sector increased so fast that landowners not only had to increase
wages to retain their labour, but also had to begin to pay at least part of wages in
money. This rise in agricultural wages, their partial monetarization, and the increased
labour mobility in the countryside marked the beginning of capitalist relations of
production in the agricultural sector; but these were `modernizations' that big land-
owners never forgave Balmaceda for.
43 In the special case of nitrates, this measure did not act as a disincentive to foreign
investment for three reasons. Firstly, as already mentioned, foreign investment
proper was not significant in nitrates anyway ± North financed the purchase of his
oficinas with a loan secured in Chile. Secondly, even after this level of taxation, profits
of foreign entrepreneurs were still very large ± equivalent to approximately one-third
of the value of production. Thirdly, these entrepreneurs, in any case, could not
produce nitrates in another part of the world as Chile was the only country with
economically exploitable nitrate deposits.
44 Among the internal taxes abolished by Balmaceda were excise and mining licences.
When he took over the presidency, these two taxes contributed almost half of internal
taxes. Yet another source of conflict between Balmaceda and big landowners was the
fact that Balmaceda (despite being himself a large landowner) did not abolish (or even
reduce) the other major internal tax, that on agriculture ± in 1885 this contributed
about 40 per cent of internal taxes. Given the nature of the political coalition that
deposed Balmaceda, it should not be surprising to find out that one of the first
measures taken by the new government was to abolish this tax on agriculture.
45 For a detailed study of these arguments, see Palma (1979).
46 The main difference between Meller's and our understanding of the role of fiscal
policy in this period is that, in our opinion, whatever might have been its subjective
motivation, its actual effect was positive for the Chilean economy. In other words,
what really matters is not whether the reduction in internal taxes ± and in the tariffs
of some public services ± took place due to `crowding out', `populism' or by the
conscious design of fiscal policy (in our opinion, at least in Santa MarõÂa's and
Balmaceda's governments it was due fundamentally to the last), or if the rapid
increase in public expenditure created a culture of lax fiscal `discipline'. What matters
262 The Chilean Economy and `Dutch Disease'
is that in the short and medium term the effect of this fiscal policy was clearly
positive ± and in the long term, as Keynes used to say, the protagonists of this fiscal
policy would not be around . . .
47 Between the middle of the 1880s and 1900, for example, production of potato doubled
in volume and that of barley and beans grew by 12 and 17 times, respectively.
References
Albert, B. (1976) `An Essay on the Peruvian Sugar Industry, 1880±1920', mimeo, Norwich,
University of East Anglia.
Anguita, R. (1912) Las Leyes Promulgadas en Chile desde 1810 hasta el 1 de junio de 1912, 4
vols, Santiago: Imprenta Barcelona.
Barham, B.L. and Coomes, O.T. (1995) `Reinterpreting the Amazon Rubber Boom: Invest-
ment, the State, and the Dutch Disease', Latin American Research Review, vol. 29, no. 2,
pp. 73±109.
Behrman J.R. (1976) Foreign Trade Regimes and Economic Development, Chile/New York:
Columbia University Press.
Bethell, L. (ed.) (1993) Chile since Independence, Cambridge: Cambridge University Press.
Blakemore, H. (1974) British Nitrates and Chilean Politics, 1886±1896: Balmaceda and North,
London/Santiago: The Athlone Press.
Blomstrom, M. and Meller, P. (eds) (1990) Trayectorias Divergentes: comparacioÂn de un siglo
de desarrollo econoÂmico Latinoamericano y Escandinavo, Santiago: CIEPLAN-Hachette.
Bustos, C. (1988) Monetary Problems of an Export Economy. Seventy Years of Monetary Debate
in Chile, 1844±1914, M.Phil dissertation, University of Hull.
Byatt, I. et al. (1982) `North Sea Oil and Structural Adjustment', working paper no. 54,
London: Government Economic Service.
Cariola, C. and Sunkel, O. (1982) `Expansio n salitrera y transformaciones socioecono- Â
micas en Chile, 1860±1930', mimeo, University of Sussex.
Chile, EstadõÂstica Comercial (various years), ValparaõÂso: Superintendencia de Aduanas.
Chile, Congreso Nacional (various years), Santiago: Imprenta del Ferrocarril.
Chile, Memoria del Ministerio de Relaciones Exteriores (various years), Santiago: Imprenta
Nacional.
Clavel, F. and Ferraro, L. (1990) `Una nueva versioÂn de los precios en el siglo pasado',
Encuentro Annual de Economistas de Chile, 1990, Santiago: University of Santiago.
Collier, S. (1993) `From Independence to the War of the Pacific', in L. Bethell (ed.), Chile
since Independence, Cambridge: Cambridge University Press.
Corden, W.M. (1977) Inflation, Exchange Rates and the World Economy, Oxford: Clarendon
Press.
Corden, W.M. and Neary J.P. (1982) `Booming Sector and de-Industrialization in a Small
Open Economy', Economic Journal, vol. 92, pp. 825±48.
Culver, W. (1982) `The Rise and Decline of Mining Regions and Mining Policy: Chile's
Role in the Nineteenth Century World Copper Trade', mimeo, New York.
Darwin, C. (1845) The Voyage of the Beagle, Journal of Researches into the Natural History and
Geology of the Countries Visited during the Voyage Round the World on HMS `Beagle',
London: Amalgamated Press Ltd.
Dennis, W.J. (1931) Tacna and Arica, New Haven: Yale University Press.
Drake, P. (1989) The Money Doctor in the Andes: The Kemmerer Missions, 1923±1933,
Durham, NC/London: Duke University Press.
Ellsworth, P.T. (1945) Chile, an Economy in Transition, New York: Macmillan.
Gabriel Palma 263
Encina, F.A. (1940±52) Historia de Chile desde la Prehistoria hasta 1891, Vols. 1±20, Santiago:
Editorial Nacimiento.
Eyzaguirre, J. (1957) Chile durante el Gobierno de ErraÂzuriz EchaÂurren, 1896±1901, Santiago:
Zig-Zag.
FernaÂndez, M.A. (1978) The Development of the Chilean Economy and its British Connections,
1895±1914, doctoral Dissertation, University of Glasgow.
Fetter, F.W. (1931) Monetary Inflation in Chile, Princeton: Princeton University Press.
Garcia-Huidobro, G. (1972) El Desarrollo EconoÂmico Chileno durante el Siglo XIX: Una
Historia CrõÂtica, BA dissertation, University of Chile.
Gibbs, A. Archives of Messers Antony Gibbs and Son Ltd., Guildhall Library, London.
Girard, L. (1965) `Transport', The Cambridge Economic History of Europe, Vol. VI, London:
Cambridge University Press.
Hinkelammert, F. (1970) El Subdesarrollo Latinoamericano: Un caso de Desarrollo Capitalista,
Santiago: Ediciones Nueva Universidad.
Hirshman, A.O. (1963) Journeys Towards Progress: Studies of Economic Policy Making in Latin
America, New York: Twentieth Century Fund.
HoÈ rmann, J. (1901) `Efecto de la ley de impuestos del 23 de diciembre de 1897; el progreso
de la industria nacional', El Mercurio, ValparaõÂso, 30 August.
Janneret, T. (1972) `El sistema de proteccioÂn a la industria chilena', in O. Mun Äoz, Proceso a
la IndustrializacioÂn Chilena, Santiago: Ediciones Nueva Universidad.
Keynes, J.M. (1985) Collected Works, Vol. 15, London: Macmillan.
Kindleberger C. (1984) `The 1929 World Depression in Latin America ± From the Outside',
in R. Thorp (ed.), Latin America in the 1930s: The Role of the Periphery in World Crisis,
London: Macmillan.
Kirsch, H. (1973) The Industrialisation of Chile, 1880±1930, doctoral dissertation, Univer-
sity of Florida.
Loveman, B. (1988) Chile, the Legacy of Hispanic Capitalism, 2nd edn, New York: Oxford
University Press.
Mamalakis, M. (1976) The Growth and Structure of the Chilean Economy: From Independence
to Allende, New Haven: Yale University Press.
Meller, P. (1990) `Una perspectiva de largo plazo del desarrollo econoÂmico chileno, 1880±
1990', in M. Blomstrom, and P. Meller, Trayectorias Divergentes: comparacioÂn de un siglo de
desarrollo econoÂmico Latinoamericano y Escandinavo, Santiago: CIEPLAN-Hachette.
Mayo, J. (1982) `Commerce, Credit and Control in Chilean Copper Mining before 1880',
mimeo, Barbados.
Mun Äoz, O. (1972) Proceso a la IndustrializacioÂn Chilena, Santiago: Ediciones Nueva Uni-
versidad.
Ortega, L. (1979) Change and Crisis in Chile's Economy and Society, 1865±1879, doctoral
dissertation, University of London.
Palma, J.G. (1998) `A Keynesian-Macroeconomics View of the Dutch Disease', mimeo,
Cambridge. Department of Applied Economics.
ÐÐ (1984) `From an Export-Led to an Import-Substituting Economy: Chile 1914±1939',
in R. Thorp (ed.), Latin America in the 1930s: The Role of the Periphery in World Crisis,
London: Macmillan.
ÐÐ (1979) Growth and Structure of Chilean Manufacturing Industry from 1830 to 1935:
Origins and Development of a Process of Industrialisation in an Export Economy, doctoral
dissertation, Oxford University.
ÐÐ (1978) `Dependency: A Formal Theory of Underdevelopment, or a Methodology for
the Analysis of Concrete Situations of Underdevelopment?' World Development, vol. 6,
nos 7/8, July/August.
264 The Chilean Economy and `Dutch Disease'
At the beginning of the First World War, Argentina had gone through several
decades of continuous growth ± only briefly interrupted during the two years of
the 1890s crisis. At the time, Argentina's 7 per cent overall growth rate ± 3.8 per
cent per capita ± was higher than that of the USA and Australia and the majority
of European countries.
It is common currency that the availability of fertile land and the incorpora-
tion of capital and labour were the bases for growth. However, in explaining it,
institutional factors, preconditions for the arrival of capital and labour, have
not received equal attention. Political stability and judicial security, the respect
of property rights ± which guaranteed that earnings and profits would not be
touched ± all came into effect some time between the inauguration of the
Constitution and the country's definite organization in 1880. It would seem
that the explanation given by DõÂaz Alejandro, in a classic book, continues to be
valid:
Pre-1930 Argentine experience fits the broad outlines of the staple theory of
trade and growth . . . it appears that growth then was related closely to succes-
sive booms in the export of land-intensive commodities, with land having a
very low opportunity cost. The economic usefulness of pampean land was not
discovered overnight, as in the case of an oil deposit, but arose as a result of a
combination of growing European needs for primary goods, technological
progress in transport and an increasing interest by Argentine policy makers in
promoting exports, foreign investments and immigration.
(DõÂaz Alejandro, 1970, p. 9)
[and moreover]: Pre-1930 growth can be said to have been `export led'
not because exports and their associated capital inflows provided growing
265
266 Argentina 1875±1930
also been alleged that investment declined. Two reasons are usually given. First,
the end of the frontier meant that the possibilities for exceptional profits
ended. Second, the downturn in the international flow of savings was not
compensated by a greater capacity for domestic savings, due to the high
dependency ratio of a young population (Taylor, 1992). On the side of demand,
it is argued that the new trends of world trade and their negative effect on
countries that relied on exports of primary products are largely to blame ±
although this factor was more influential in the period after 1930.
Other arguments point to rigidity-inducing institutional factors, which ham-
pered the adaptation to new market conditions and to changed resource avail-
ability. Argentine social structure and patterns of land ownership, particularly
the absence of family farming, have been generally blamed (Solberg, 1987).
Moreover, it has been claimed that the sectors traditionally linked to agricul-
ture had opposed protectionist policies which favoured industrialization. In
this chapter we will take up some of these arguments.
Growth trends
world crisis in 1929. Growth in the 1920s was only slightly slower than between
1892 and 1912. We now turn to analyse in detail both the period of accelera-
tion, 1875±1913, and the slowdown, 1913±27.
Deceleration, 1913±27
In 1913, the upward phase was interrupted. In 1914, GDP fell 13 per cent. The
following year witnessed a recovery, GDP growing 7.6 per cent. However, in
1916 and 1917 GDP experienced falls of 5.5 per cent and 15 per cent respect-
ively, but from that year onwards GDP grew at a rate of 4.5 per cent per annum ±
measured as a three-year average ± until 1929 when, due to the world crisis, this
trend was again interrupted.
What factors brought about the fall of total GDP by 2.3 per cent per annum
between 1913 and 1917? What factors contributed to the subsequent recovery?
Was it a recovery, or a quite independent growth spurt? If so, why was it
interrupted? Could it be because the long phase of Argentine decline had
already begun in 1913? Was it due to an ephemeral recovery in the 1920s,
unsustainable in the context of the existing economic structure? Were there
Roberto CorteÂs Conde 269
Terms of trade
A further element we need in order to study the evolution of GDP and of
income ± for its effect upon welfare ± is the effect of the terms of trade. In
Figure 9.1 we adjust GDP, deflating export values (which are not at constant
prices) by their purchasing power (the ratio of export prices to import prices).
From 1915 to 1927 the terms of trade were unfavourable to exports. Therefore
income ± defined here as GDP adjusted by the terms of trade ± lagged behind
GDP in the postwar period. At the point of greatest divergence, 1920, income
was 14 per cent below GDP: a consequence of the contraction brought about by
the increase in USA interest rates. It is perhaps important to point out that
Argentina responded to the price fall by producing in greater volume: thus,
neither output nor income fell. From 1922 onwards, the terms of trade
improved, returning to prewar levels in 1928.
9500
GDP in millions of pesos
GDP adjusted
8500
7500
Million pesos
6500
5500
4500
3500
1913 1915 1917 1919 1921 1923 1925 1927 1929 1931 1933 1935 1937 1939
Product % %
decline began. This argument simplifies a more complex process. Let us explore
whether there existed supply-side limitations to growth and the validity of the
`end of the agricultural frontier' argument.
Between 1875 and 1900, wheat and corn, followed by alfalfa and linseed,
were responsible for agricultural growth. In the years 1914, 1916 and 1917,
three crops failed. In the following year, production recovered owing to better
weather conditions, but the area under cultivation did not increase as fast as in
the first decade. Production of the principal crops in the pre- and postwar years
developed as shown in Table 9.1.
According to Bunge, the peak in extensive growth was reached in 1908. His
argument was based on a comparison of the divergence in the rates of cultiv-
ated land expansion and population (Bunge in Llach 1985). In fact, however,
the slower incorporation of land for cereal farming coincided with the expan-
sion of other more intensive crops ± and of greater value per ton ± like cotton.
Data on cultivated land do indeed point to the end of the frontier. A similar
conclusion is reached if we compare output growth between 1913 and 1929.
Nonetheless, if the peak of agricultural expansion is taken into account ± 1924,
not 1929 ± the situation is different. Compare the evolution of wheat ± a cereal
± and cotton, a new industrial crop, in Table 9.2.
Between 1913 and 1924, the value of the combined production of wheat and
cotton grew by 10.6 per cent per annum. Moreover, the extension of cotton
1913 2 850 88.3 251 655 1.8 145 261 251 916
1924 5 202 143.3 745 447 51.1 335 17 119 762 565
1929 4 425 87.9 388 958 115.4 215 24 811 413 769
Source: Author's elaboration using data from the Anuario GeograÂfico (1941).
272 Argentina 1875±1930
Source: Banco de la NacioÂn Argentina, Report and Balance Sheet (1923) for `Cultivated area'; author's
own elaboration of textile production data.
future of the industry was also to a degree under threat from Britain's imperial
preference (O'Connell, 1984). When external markets closed, domestic avail-
ability increased, prices fell and domestic consumption increased because the
supply of cattle from natural pastures is relatively inelastic ± it is not possible to
fatten the cattle forever, because the grass runs out. When stocks diminished
and external demand rose ± as at the end of the war ± there was a retention of
cattle, supply diminished, prices went on increasing and domestic consump-
tion decreased. Nonetheless, the contribution of cattle to the increase in GDP
did not result from an increase in production, but from the value of its product.
This phenomenon occurred in the 1900s and was over by the First World War. A
further change of this nature did not happen after the war.
Turning to capital endowments in the two subperiods, these depended upon
investment and both domestic and foreign savings. In Argentina, from the last
two decades of the nineteenth century to 1913, the negative current-account
balances coincided with positive balances in the capital account produced by
massive foreign investment inflows. In 1910, the Argentine Minister of Eco-
nomy ± Jose MarõÂa Rosa ± ordered the preparation of an inventory of foreign
capital (MartõÂnez and Lewandowski, 1912). The inventory estimated that some
5059 million pesos had entered the country: 1572 million corresponded to
investments in Argentine bonds and 1827 million to railways. Banks, ports,
tramways and gas and electricity companies followed far behind. A further 364
million pesos were invested in land companies and mortgages. In 1917, Bunge
estimated that foreign capital amounted to 8813 million pesos of which 3055
million were in railways and 1493 million in bonds (Table 9.4). The bulk of
railway investment took place before the war. In 1914, the national census,
estimated private ± foreign ± capital in railways as 3051 millions pesos and the
property of the state as 320 million (see Table 9.5 below).
After the war, technology was changing and new forms of transport and
communication were appearing; the railway and the telegraph began to be
replaced by automobiles, trucks and telephones. The number of inhabitants
per automobile in 1910±14 was 680, by 1939 it had dropped to 25. Moreover,
there had been changes in communications. Telegraph wires stagnated: from
31 000 kilometres in 1910±14, the figure had increased to 41 000 in 1915±19 and
45 000 in 1928, in contrast to the 293 per cent increase from 1910±14 to 1931.
Prior to 1935, the estimates of the balance of payments are not reliable
enough to deduce the movement of capital. However, it does appear that up
to 1914 there was a wide gap in the current account which was covered by
capital inflows.3 However, after the war, when foreign, mainly British, savings
ceased to perform the same role, was there not domestic saving? Was there a
decline in postwar domestic saving? Bank saving deposits, public bonds, mort-
gage bonds ± the income bond most used at that time ± and the volume of
shares in the stock exchange (Figure 9.2) can all be used as indicators of the
development of internal saving. We do not take into account here the saving
which stems from reinvested profits.
Domestic saving is subject to great fluctuations, influenced by special circum-
stances such as the financial deepening in 1903 after convertibility or the
withdrawals which coincided with the 1907 crisis. However, it cannot be said
that domestic saving diminished in the postwar period compared to prewar ± in
1,200
1,000
800
Million pesos
600
400
200
0
1900 1902 1904 1906 1908 1910 1912 1914 1916 1918 1920 1922 1924 1926 1928 1930
Figure 9.2 Argentina: saving trends in public bonds, mortgage bonds, shares and deposits
in savings accounts (three year average ± at 1914 prices)
Source: Author's own calculations using data from the Anuario GeograÂfico (1941).
Roberto CorteÂs Conde 275
this case we refer to institutional savings. On the contrary, there seems to have
been an important increase between 1915 and 1923.
Of course, the magnitude of domestic saving is not comparable to that
required for the installation of the railway network financed by foreign
savings in the 1900s. These investments had a long-term amortization, which
meant that the initial rhythm could not be maintained. This is a factor to take
into account when the investment of the first decades of the century is com-
pared to that of the postwar period. On the other hand, in the case of the
railways there was a problem of indivisibility. This explains why, at the begin-
ning, the effect of railway productivity on total capital productivity was low.
Here, the capital/output relation was 9.7 in 1885±89, 11.4 in 1900±04, 9.7 in
1910±14 and 6.2 in 1925±29.4 The paradoxical result was that as the quality of
capital increased, its marginal productivity did not diminish, but rather
increased. At the beginning installed capacity was underutilized, but later
increased as land entered production. The levels of monetization of the eco-
nomy were very high; around 50 to 60 per cent, with deposits at about 30 to 40
per cent.
The greatest investment was in infrastructure and principally in railways: in
1895 investment reached 1.2 billion pesos ± 1.3 billion pesos in 1914 prices ±
and in 1914 increased to 3 billion pesos. No less surprising were the changes in
the quality of stock and cattle installations whose value increased from 1.1
billion to 4.7 billion pesos between 1895 and 1914; in this case the investment
was local. The results were lower than the exceptional results obtained when
the quality was improved. In 1914 the process had been completed. There were
transport networks available which covered the production areas and an
amount of cattle stock capable of supplying the world market demand. The
bulk of investment had been carried out, and what remained required less
effort. Nonetheless, the investment carried out after the war in other sectors ±
industrial machinery, textiles, automobiles, telephones ± was just as important,
albeit of a different magnitude. Postwar Argentina still had to fully utilize its
enormous installed capacity.
Without doubt, the variations in the capital stock point to a slowdown in the
rate of investment. It must be taken into account, however, that it has not been
possible to compare the prewar period with the period which ended in 1930,
due to problems of information. In Table 9.5 the levels of stock in 1914 are
compared with those of 1935, when the country was still suffering the effects of
the 1930s crisis. Our estimations cover only some of the items shown in the
1914 census (Table 9.5). However, the slowdown does not necessarily imply
that domestic saving shortages are to blame. Instead, at least in this case,
savings were greater than investment. We conclude that the positive commer-
cial balances in the decade were a consequence of previous of trade deficits
financed by capital inflows (see Table 9.6).
276 Argentina 1875±1930
Table 9.5 Argentina: variation in the capital stock, 1914±35 (millions of pesos)
Annual
1914 1935 variation (%)
Source: Author's own compilation based upon publications in the Revista de EconomõÂa Argentina and
the censuses of 1914 and 1935.
1920 1921 1922 1923 1924 1925 1926 1927 1928 1929
Exports 716.31 677.45 767.00 947.74 903.56 792.50 960.95 1 048.90 1 001.05 661.30
Imports 712.02 567.74 761.00 822.59 850.84 744.70 816.60 843.50 881.70 778.90
Balance 4.29 109.71 6.00 125.15 52.72 47.80 144.35 205.40 119.35 117.60
Servicing of debt 145.50 115.58 157.50 169.33 165.27 153.35 177.55 187.75 197.10 180.00
Total of services 201.50 161.08 201.50 218.33 202.27 190.35 236.55 241.75 248.10 223.00
Current account 197.21 51.37 195.50 93.18 149.55 142.55 92.20 36.35 128.75 340.60
Increase in debt 179.21 49.92 193.60 53.18 158.55 144.20 122.00 186.85 4.24 276.60
Var. in Reserves 18.00 1.45 1.90 40.00 9.00 1.65 29.80 150.50 124.51 64.00
GDP current prices 9 388 8 646 8 073 8 871 9 806 9 197 9 301 10 097 10 084 10 352
Var. reserves as a % of GDP 0.44 0.04 0.05 1.02 0.21 0.04 0.73 3.39 2.81 1.41
Net foreign outflow 22.29 111.16 7.90 165.15 43.72 46.15 114.55 54.90 243.86 53.60
Net external balance (% of GDP) 0.54 2.92 0.22 4.23 1.01 1.14 2.80 1.24 5.50 1.18
a
Five gold pesos equalled one pound sterling.
Source: Based on Phelps (1938).
277
278 Argentina 1875±1930
How did market conditions affect the price of labour? Real wages were mainly
affected by exogenous factors. First, at the end of the 1880s and beginning of
the 1890s inflation had a negative effect ± even though mitigated by the
international decrease in food prices (CorteÂs Conde, 1979). By contrast, during
the second half of the 1890s, deflation resulted in an important increase in real
wages. In the 1900s, in spite of exchange rate stability, wages were subject to the
negative impact of the international increase in prices. From 1880 to the First
World War they increased by 30 per cent, 1 per cent a year in real terms (CorteÂs
Conde, 1979). These figures undermine the widespread thesis which holds that
wages fell during the period of massive immigration. During the same period
the economy was growing by 3.7 per cent per capita a year, probably owing to
the great flexibility in labour supply from abroad and to the fact that, comparat-
ively, Argentine wages were still higher than those in Italy and in Spain. This, as
well as much else, changed during the postwar period.
The wartime recession affected wages catastrophically. Industrial recession,
construction, the government and international inflation were all to blame.
Between 1914 and 1918, real wages fell by 11 per cent a year. It is likely that
employment was not as harshly hit, perhaps because of wage flexibility.
From 1918, the postwar recovery contributed to an increase in real wages, not
just as a result of greater activity but because prices fell and in general nominal
wages remained stable. From 1918 to 1930 wages rose by an incredible 7.5 per
cent a year. This is notable because GDP growth was around 3 per cent and per
capita growth was slightly negative. Partly because the immigration flow had
decreased, workers acquired more negotiating power and were able to improve
their share of income distribution. However, once again, the decrease in prices
was largely responsible for these changes. The cost-of-living index in 1920 was
186
1914 100; in 1930 it had fallen to 133 ± falling on average 3.7 per cent a
Roberto CorteÂs Conde 279
year. As had happened in the 1890s, in the 1920s the decrease in prices was the
determinant factor. As a result, during the second half of the 1920s the condi-
tions of Argentine workers improved much more than those of British workers
in relation to the immediate postwar period (Figure 9.3) while the gap vis-aÁ-vis
Italian workers was even wider, as real salaries were reduced by Mussolini's
policies (Figure 9.4).
180
140
120
Index (1914 = 100)
100
80
60
40
20
0
1885
1887
1889
1891
1893
1895
1897
1899
1901
1903
1905
1907
1909
1911
1913
1915
1917
1919
1921
1923
1925
1927
1929
1931
1933
1935
1937
Figure 9.3 Real Wages in UK and Argentina, 1885±1938
Source: CorteÂs conde: Ei Progreso Argentino.
180
140
120
Index (1914 = 100)
100
80
60
40
20
0
1911
1931
1901
1921
1885
1887
1889
1891
1893
1895
1897
1899
1903
1905
1907
1909
1913
1915
1917
1919
1923
1925
1927
1929
1933
1935
1937
Between 1917±29 it went up again by 7.5 per cent, and even though that
growth was lower than the prewar rate, it was nonetheless important. But the
difference was that the cargo transported increased more than the amount of
railway lines constructed. This allowed a greater utilization of installed facil-
ities. The network had increased by 5.2 per cent annually between 1900 and
1913 but slowed down to 0.5 per cent between 1913±27. Capital invested
amounted to 6.3 per cent and 1.6 per cent respectively. In 1900±14, the cargo
transported was 1.18 tons/km, and in 1925±29 it increased to 1.36 tons/km.
From the final decades of the nineteenth century onwards, textiles and food-
stuffs had been the leading sectors of industrial growth. However, the evidence
on production alone does not reflect the depth of their transformation.
The food industry began with modern factories and high capital intensity.
The textile sector, on the other hand, included thousands of dressmakers ±
the putting-out system, which was counted as manufacturing in the GDP
estimates. The change towards factory activity took place during the 1920s,
simultaneously with the extension of cotton cultivation which provided low-
cost inputs.
Concerning the evolution of industrial production there are data available
from the censuses of 1895, 1914 and 1935. This information presents a variety
of problems, which led us to carry out annual estimates of production and
industrial product.5 Also, the census only takes into account factory activity
whilst the GDP estimates take into account everything which implies trans-
formation of raw materials. According to the census, between 1895 and 1914
the capital stock in industry grew by 9 per cent annually ± our estimates ± while
GDP growth was below 7 per cent. Between 1914 and 1935 industrial produc-
tion increased by 2 per cent a year. Between 1895 and 1914, capital stock in the
industry increased by an annual 5.8 per cent and from 1914 until 1935 by 2.4
per cent. This would contribute to maintaining the picture of a declining trend.
However, some qualifications are in order. In the food sector the slowing-down
process was more pronounced. The 10 per cent yearly growth before the war fell
to 0.3 per cent a year. In textiles the rate fell from 11 per cent in 1895±1914 to
4.4 per cent in 1913±35. However, if we take the peak trend of 1913 and that of
1928, industrial output grew 1.4 per cent a year according to our estimates.
Nonetheless, food increased 3 per cent a year and textiles 3.5 per cent. We
obtain these results because the production of the whole industrial sector was
adjusted by the census data, which includes less dynamic sectors which do not
appear in our sample.
The textile industry suffered an important fall in 1914. However, it recovered
very rapidly because of the availability of domestic inputs, and during the
recovery phase the growth rates were higher. From 1917 to 1927 industry
increased by 3.6 per cent a year. Between 1914 and 1926 food production
grew by 3.3 per cent, whilst textiles grew at an incredible 7.2 per cent a year.
282 Argentina 1875±1930
The problem was not demand but supply, and the recovery was based on a
partial substitution of local production for raw materials which prior to the war
had been imported. This process continued after the war: cloth was imported
whilst cotton was exported. It should be noted that the import substitution
which took place in the textile industry consisted of a substitution of inputs
rather than final products. The postwar growth, as we have seen, was notable
and took place in the factory sector.
Between 1895 and 1914, capital productivity increased ± the K=Y relationship
fell from 1.51 to 0.76 ± a result of the greater use of installed capacity in food,
where at the outset there had been substantial investment in industrial plants,
meat-packing plants, sugar refineries and mills. Capital productivity went
down slightly again by 1935, probably because of investments in important
factories in the textile industry which was witnessing the transition from put-
ting-out to the factory system. As a result of the transformations, the textile
industry became more organized, and production increased from 13 000 pesos
per establishment in 1895 to 80 000 in 1935.
Demand
Between 1910±14 and 1920±24, exports doubled. Nonetheless, the five-year
growth rates 1900 to 1929 show a notable deceleration, as in Table 9.8.
Until 1929, however, foreign demand remained stable, despite a strong
slowdown in world trade in the postwar period which particularly affected
primary products. World cereal demand no longer had the dynamizing effect
of the first decade, and the meat trade began to face protectionist restrictions
(see Figure 9.5).
The biggest difficulties created by foreign demand were felt in the structures
of the cereals and meat trades. There were countless agricultural and cattle
producers, and no control over supply. Between them and international
demand stood grain-trading firms and meat-packing plants, in a position to
reach collusive arrangements and control demand. In periods of expansion
nobody complained of the cost of middlemen, but during periods of decline
1900±04 147
1910±14 431 118
1915±19 707 64
1920±24 834 18
1925±29 935 12
50 350
% 45 Index
300
40
35 250
30 200
25
20 150
15 100
10
50
5
0 0
1910 1913 1916 1919 1922 1925 1928 1931 1934 1937 1940
they passed on the effects of lower prices to the producers, provoking protests
and requests for state intervention.6
We turn here to some indirect indicators of the variations of internal
demand. We have already shown that population growth slowed down, how-
ever slightly, influenced by the decline in migration. From 1914 to 1917, there
was also a strong fall in employment. In the Federal Capital, from a base of 100
in 1914 employment fell to 85 in 1917, recovering in the 1920s and rising to
139 in 1929 (National Geographic Committee, 1941). Finally, wages tended to
fall in years of depreciation and rose when the exchange rate revalued.
Investment also experienced important fluctuations, for two reasons. First,
Argentina was a new country and needed to create a production infrastructure.
This is done only once, and has a very long period of amortization. Second,
strong disturbances were felt from the international capital market during the
war and postwar periods. Foreign investment ± in which the railways played the
greatest role ± increased until the war, fell during the years of the conflict and
recovered afterwards. However, the sectors it was directed to had changed,
which also meant that it had a lower magnitude. Public and private construc-
tion slowed down. In part, the war recession and the ensuing slow recovery
were to blame, and in addition the construction of the infrastructure of public
buildings and houses had now been completed. Government expenditure
diminished between 1915 and 1917, though afterwards, during the 1920s, it
grew at slightly above 10 per cent of GDP.
284 Argentina 1875±1930
Policy
Trade policy
DõÂaz Alejandro has rebutted the view that following the First World War Argen-
tina failed to adapt and left domestic industries unprotected, facing unstable
foreign markets and a changed resource endowment. He showed that negative
net protection did not occur and that the tariffs had not only sought a fiscal
objective between 1906 and 1940, but had in some way been protectionist (DõÂaz
Alejandro, 1970, p. 294).
Not many explicit economic policy decisions are documented in this
period. However, even though they may not have been visible, this does not
mean that they did not exist. It is also true that policies did not correspond to a
single objective. Most of the time they responded to different interests, situa-
tions and even pressure from sectors which sometimes coincided and some-
times did not. The government's preoccupation was to avoid increases in the
prices of manufactured goods in the war years ± raw materials or finished
products ± affecting consumers and producers. The cost of imported goods
depended on several factors: the international price, the exchange rate and
tariffs. The effective tariff was a combination of the nominal tariff and the
values which the government fixed on imported merchandise for the purpose
of paying the duty ± the official or aforo value.7 In general, each year up-to-date
lists of aforo values were published which might or might not coincide with
market values; however from 1906 to 1920 the aforo values were not revized. As
the prices of imported goods had increased significantly during the war the
divergence between both values increased, resulting in a decline in protection
(see Figure 9.5 and Jorge, 1971). The exchange rate did not compensate
because, in spite of the suspension of convertibility, the peso appreciated dur-
ing the war years. Instead, protection came from the increase of imported prices
and the strong supply restrictions in the economies dedicated to wartime
production.
The government's aim in reducing the aforos and with that the revenue from
the tariff was to try to relieve the effect of the rising price of goods on con-
sumers and producers. While prices increased up to 1919, the peso appreciated
and the aforos fell ± both reducing protection. Between 1919 and 1920 prices
began to go down and the peso depreciated. In 1920 prices continued to fall; in
1921 prices went up in step with a continued depreciation of the currency. The
depreciation increased the combined effect of the tariff and aforos with protec-
tionist effects.
From 1924, the local currency underwent a revaluation as a result of the effort
to return to convertibility at the prewar parity. At the same time the prices of
imported products fell, which meant that the difference between the real values
and the aforo values was reduced, and eventually eliminated, with a consequent
Roberto CorteÂs Conde 285
50
Argentina prices
40
UK prices
30
20
Variation %
10
–10
–20
–30
–40
1886
1888
1890
1892
1894
1896
1898
1900
1902
1904
1906
1908
1910
1912
1914
1916
1918
1920
1922
1924
1926
1928
1930
1932
1934
1936
1938
Figure 9.6 Prices in the UK and Argentina
Sources: CorteÂs Conde (1979); Anuario GeograÂfico (1941); and Mitchell (1975).
From 1885 to 1899 Argentina was under a regime of floating exchange ± and
forced circulation. A system of national and provincial issuing banks led to a
disproportionate creation of money and debt and the crisis of 1890. Liquida-
tion was followed by a period of deflation with a strong contraction of spending
and money, during which the exchange rate was stabilized and domestic prices
made to converge with international ones (Figure 9.6).
When in 1899 the monetary reform was implemented with the creation of
the Currency Board, the country formally joined the gold standard. Thereafter,
and until 1913, the convertibility regime was applied successfully and became a
positive factor for the development of the private financial system and of a
domestic debt market, which grew rapidly in the first decade of the century.8 In
this new atmosphere it was perceived that property rights would be respected, a
change from the repeated confiscations of the nineteenth century.
However, it cannot be said that the convertibility system was the only de-
cisive factor in attracting foreign capital; there was also a fall in the exchange
rate. The 1880s had witnessed massive capital inflows, but this capital was
subscribed in European stockmarkets and had special guarantees or was issued
in pounds or gold. This meant it was not affected by the volatility of the
exchange rate since the legal jurisdiction which protected it was not that of
Argentina but that of the country where it was subscribed. The fixed exchange
rate offered stability for direct investments in domestic currency ± pesos. The
interest rates had fallen sharply at the beginning of the century, had then risen
until 1914, but then fell again and hovered around 5 per cent in real terms.
Roberto CorteÂs Conde 287
Also, interest rates converged with international rates, the remaining difference
reflecting the country risk factor.
As in the rest of the world, Argentina suspended convertibility at the begin-
ning of the First World War and reestablished it in 1927. It should be pointed
out that whilst the suspension lasted it respected the rules of the gold standard,
issuing only against gold inflows.9 At the beginning the suspension had a
slightly unfavourable effect on the peso, but later, owing to capital inflows
from Europe, not only was the prewar parity maintained but up to 1920 the
peso, appreciated in relation to the pound. The depreciation took place
between 1921 and 1924, and from then on the trend reversed. In 1927 it
returned to the prewar parity, unlike several European countries such as France
and Italy.
In Argentina the money supply and credit under the gold-standard regime
was conditioned by the seasonal gold inflows which took place during the
summer months. Indeed, the so-called Argentine cycle consisted of expansion
of money and credit in the summers and contraction in the winters. This led
to the official bank ± NacioÂn ± partially sterilizing the entrance of foreign
exchange in the summer and expanding credit in the winter, based upon
surplus reserves. To do so it used the bank's own gold reserves which did not
appear in the assets of the Currency Board.
During the war years, in which Argentina suffered international inflation
until 1916, the increase in prices was quite small. From 1913 to 1916 they
increased 4.5 per cent a year, and an increase of 12.7 per cent a year took
place from 1916 to 1929. Thereafter, a strong effort of stabilization took place,
leading to a fall of 4.9 per cent a year.
Public finances
Under the Currency Board regime, the government could not monetize its
deficit. Once Argentina concluded the 1880s debt negotiations it regained
access to foreign credit, but this financing was interrupted during the war
years. In an atmosphere of stability, the domestic market was an important
substitute. Nonetheless, when the government incurred a deficit it used short-
term instruments and created a floating debt which was periodically consolid-
ated. The financing problems stemmed from the reduction of resources in the
war years and the weight of the previous debt. It could be thought that the use
of domestic savings by the state had a crowding-out effect on private invest-
ment; inferred from the relatively low volume of shares traded. However, the
bulk of the debt market was channelled towards construction through mort-
gage bonds. The success of this bond, the enormous population increase in
Argentina's main cities and a regime of joint-stock companies which even
today provides little security to the minority shareholders, explains the lower
preference for private debt.
288 Argentina 1875±1930
In Argentina, the state had by now taken on the functions which were then
considered part of its responsibility, as probably in other new countries also.
Nevertheless, its importance should not be underestimated. It had a major role
in public works, in promoting agriculture and in social policy from 1904
onwards, when the Labour Department was formed and quite advanced social
legislation was implemented.
Until 1930, the tax regime was based on the import tax. Despite its regressive
nature, because it applied to consumption independently of consumers'
income, in Argentina it was widely accepted. This was due to the low cost of
foodstuffs and imported manufactured goods as a result of low tariffs. Because
of that, as long as surplus demand for labour prevailed ± except during the years
of the war recession ± workers were in favour of free trade and monetary
stability, reflected in high real wages. For several decades this maintained an
ample social consensus ± not extended to the political area ± which ended in
the years after the First World War.
by the increasing weight of labour organizations. Until then the rule had
been that the market resolved the issue of payment for goods and services,
but now incomes were negotiated through organizations and interest
groups. The new situation created a problem of collective action.
4 Even though the state in the nineteenth century did not have an insignifi-
cant role ± much greater in the Continent and in the central powers ± the
new conception of war, which required the subordination of all resources to
its objectives, extended its field of influence. Exchange and price controls
were set up, the government intervened in the purchasing of raw materials
and in war production. This produced an institutional change unthinkable
in any other set of circumstances. The civil administration now experienced
the possibility of exercising greater control over society. Even though at the
end of the conflict many of the controls were given up, the lessons learned in
this period would be used repeatedly in the following decades.
5 Along with these circumstances came others which had an effect upon
exporting countries. These included a decrease of the growth rate of trade,
not just due to the restrictive actions of those who sought to secure a regular
supply of foodstuffs and basic raw materials, but also because the prewar
expansion had reached high production levels and in some cases resulted in
inventory accumulation. Also, the countries which were not involved in the
war went through a brief postwar boom which came to an end with a
dramatic crisis of overproduction in 1921, made more serious by policies
in the industrialized countries.
The outbreak of the World War in August 1914 found Argentina particularly
unprepared to withstand the impact of the economic forces which that
catastrophe loosed. Especially susceptible to international economic influ-
ences, because of its close and financial relations with Europe, Argentina was
bound to be severely affected by the outbreak of such a struggle, even under
a normal domestic situation. But in 1914 Argentina was already in the
depths of a depression . . . The immediate cause of this reaction was traceable
to the Balkan War, which tended to check the flow of European capital into
Argentina that had characterized the preceding decade.
(Phelps, 1938, pp. 22±3)
advances in social legislation. Interest groups which sought a better share for
their sector, rather than production increases or general welfare, began to
appear, and the earlier consensus ended. The institutions for the new situation
were not designed until the 1930s and 1940s, but the problem they were
responding to arose during the first postwar period.
To sum up, the country went through strong income fluctuations because of
the depreciation of the peso immediately after the war and its later appreciation
until the return to prewar parity. This led to distributive conflicts which, while
they did not have the magnitude of those of Germany or France, were greater
than that of any other Latin American country. The resolution of these new
conflicts required agreements concerning new institutional rules which were
very difficult to reach. There was little experience and the game itself was to be
resolved in the negotiations. Thus began a long period of uncertainty which
had an effect on investment and, consequently, growth.
Argentina went through important political changes which affected the
governability of the country and, as a consequence, the investment rate and
growth. As a result of the long pressure to widen political participation, in 1910
the electoral system was reformed establishing secret and obligatory universal
suffrage. Limited participation was extended in a rapidly changing society.
With the implementation of the new regime, the opposition party took over
government ending a long period of stability and hegemony of a coalition of
conservative parties which had governed the country since 1880. The general-
ized political apathy of the native population and the fact that an enormous
number of immigrants (a third of the population in 1914) showed no signs of
wanting to become naturalized citizens to obtain political rights ± because to do
so meant having to do military service ± resulted in very low transition costs for
governments whose electoral procedures were constantly under public scrutiny
but which had, nevertheless, an ample consensus on fiscal and economic
aspects.
As participation was extended, the postwar period introduced economic
problems to the realm of policy, thus emphasizing differences and aggravating
conflicts. Argentina initiated an experiment which was soon criticized by those
who left the government, and which in an atmosphere of mutual intolerance in
1930 led to the first military coup of the twentieth century. By 1930 Argentina
had acquired the conflictive elements which would characterize it for the next
half century.
Conclusions
It has commonly been claimed that the end of the frontier was the cause
of the slowdown in growth, with emphasis on the factors which influenced
the supply side. It is true that the rate at which areas were sown with
cereals declined, but this explanation is insufficient: the development of
other crops and the changes in livestock-raising ± which led to the growth in
the first decade ± did not depend upon the increase in the amount of land
exploited but on its different use. Indeed, in the case of meat the problem was
one of demand rather than supply. There was no reason, as was seen in the
immediate postwar period, for the cattle stock and production to fail to
increase.
As far as investment is concerned, it is true that the data show a slowdown in
the rate of capital accumulation, but there are two caveats, one concerning the
statistical effect caused by starting from very low levels, as happened at the
beginning of the century; and a second concerning the magnitude and indivis-
ibility of the investments in infrastructure in a new country with a long
maturing period. Some exceptional circumstances should also be taken into
account, including the overexpansion and overinvestment in the first decade;
the financial deepening which followed the move to convertibility in 1899
during which there was an exceptional increase in savings; and the strong
investment in the railway network in the 1900s.
Another line of explanation of the slowdown comes from the institutional
factors which influenced factor markets once there was a change in the avail-
ability of resources. The industrial sector witnessed a modification in the
intensity of factor use, ranging from more intensive land use in the production
of foodstuffs to more intensive labour use in textiles. However, despite popula-
tion increases as a consequence of immigration, the labour supply was never as
flexible as during the prewar period. This was seen in the different behaviour of
wages in the two periods. In this, two circumstances played a part. One was the
greater organizational power of workers, and the other the halt in immigration
flows during the war years.
In the prewar period, in a context of monetary stability and zero inflation,
the relative factor prices stemmed from the combined action of supply and
demand and from the organizational ability of the workers or entrepreneurs
when negotiating them. By contrast, the war witnessed a phenomenon that,
although not totally new, became decisive. The workers' real incomes were now
affected by governmental regulations on currency, exchange rates and tariffs
more than by negotiations over their nominal values. As a result, it became
necessary to influence the government in order to improve a sector's position.
Thus began a long period of distributive conflicts, which continued into the
following decades. This coincided with significant changes in the political
regime: popular participation was increased with the passing of the law of
universal suffrage, secret and obligatory.10
294 Argentina 1875±1930
With respect to demand-side changes, whilst the initial fall of GDP was
caused by a supply shock ± the decrease in imports ± the weakness of the
following recovery was due to the changing conditions of world demand, a
consequence of protectionist policies in the industrialized countries which
affected the prices of cereals and the meat trade. The displacement of Germany
during the war years and the years that followed until the stabilization of 1924
closed a market which had become important for Argentina during the prewar
period.
How did the country respond to the new demand conditions? Were adequate
policies implemented? The alternative policy and the one most debated in the
literature would have involved abandoning the gold standard instead of trying
to deflate and return to parity, either allowing the exchange rate flexibility or
adopting, like France, an undervalued exchange rate (Temin, 1990). It would
seem that because of the previous success, upon receiving an enormous flow of
capital it was reasonable to focus on the advantages ± from the point of view of
credibility ± of maintaining the country within the rules of the gold standard
and returning to the prewar parity.
In a period of decline of prices, however, the pressures on policy had to do
with the patterns of internal trade. While there were countless agricultural
producers, the commercialization of grains and meats were in the hands of
few firms and, in the phases of falling prices, they were able to pass on the losses
to the producers. Because of this producers demanded state intervention in
commerce, pressing for policies which in the following decades favoured the
internal market (Smith, 1969).
Turning to trade policies, first of all it must be said that, unlike Brazil,
Argentina's market share was not large enough to be able to pass cost increases
on to foreign consumers. Also, in cattle production there was no possibility of
accumulating stocks over long periods. As far as imports are concerned we have
seen that there was de facto protection during the war years when prices of
imported products and the cost of transport increased. The policies tended to
be more protectionist when in the postwar period prices went down. Finally,
Argentina's range of instruments included the nominal tariff, the aforo and the
exchange rate. These policies had as their principal aim the avoidance of price
increases in raw materials and basic products.
In this period external circumstances were largely responsible for slow
growth. They included a supply shock due to the fall in imports which pro-
duced the first fall between 1914 and 1917 and later on the changing condi-
tions in the cereal and meat markets which no longer stimulated growth as
before. In response, a new option emerged: the expansion of intensive crops
such as cotton which provided a cheap raw material for the rising textile
industry. It can be said that the successful initial adjustments of the 1930s to
the conditions of world markets were in some way prepared in the initial
Roberto CorteÂs Conde 295
postwar years. In this sense, it would seem that there was a certain unity in the
interwar period (1914±45). What also seems clear is that the institutional
changes which caused these adjustments were not positive for the subsequent
evolution of the country (DõÂaz Alejandro, 1970). While Argentina seemed at the
end of the 1940s to have successfully isolated itself from external negative
shocks, several decades later it seems that it was isolated from positive shocks,
if we take into account the way in which world trends moved between 1950
and 1970 (Maddison, 1991). Australia, which also suffered the same external
negative shocks between the 1920s and 1940s, experienced notable growth
from 1948 to 1970, in response to world economic expansion, of a very differ-
ent nature from that of Argentina.
The most serious works on investment and capital formation are those of Ganz
and Balboa and Fracchia, which served as a base for the ECLA study on Argen-
tina (Goldsmith and Saunders, 1959). Ganz's estimations were criticized by
DõÂaz Alejandro, because they used 1950 prices. DõÂaz argued that the domestic
price estimates were exaggerated; he converted the estimates into 1935 prices,
closer to the price structure of the first two decades, recognizing that none-
theless they might still be overestimated. Whilst for Ganz (prices of 1950)
investment represented 39 per cent of GDP in 1900±09, 42 per cent in 1910±
19 and 13 per cent in 1915±19 for the same period, DõÂaz's estimates were 17.2
per cent, 18.7 per cent and 6 per cent (DõÂaz Alejandro, 1970). To Ganz, in 1920±
24 the fixed investment represented 26 per cent of GDP and to DõÂaz only 12 per
cent, while the estimates for 1925 also differed, 33 per cent and 15 per cent.
More recently Robert Gallman carried out another revision, this time based
upon 1914 prices, obtaining a result in which investment represented 16.9 per
cent of GDP in 1900±09 and 14.6 per cent in 1910±14 (Davis, and Gallman,
1996). Using Bunge's date (1916) he estimated a capital stock in 1916 (at prices
of 1914) of 12.5 billion pesos while Ganz estimated it at 23.5 billion (at prices of
1914).
ECLA's work on investment (1959) is based on 1950 prices and does not
indicate that deflation has been carried out. Before 1935 there were no surveys
concerning wholesale prices; those published referred to the cost of living. The
Anuario GeograÂfico of 1941 quotes as a source the National Statistics Board for a
series of wholesale prices it publishes, which differ substantially from consumer
prices which, with all its defects, is the only systematic series.
The method used by ECLA is based on the annual flow of domestic produc-
tion and on capital-goods imports. With many difficulties the import data can
be re-done, but there are no annual series of capital-goods production. Our
296 Argentina 1875±1930
Notes
1 Similarly, the displacement of the USA by Argentina in the British meat market, rather
than an important growth in demand, was responsible for expansion in the case of
cereals (CorteÂs Conde, 1979).
2 Davis and Gallman (1996). For further details, see Appendix.
3 Derived from information concerning foreign capital from diverse sources. It confirms
that external savings, until the First World War, compensated for the lack of domestic
saving in the magnitude required by the exploitation of virgin land.
4 For the different estimates of the capital stock, see Appendix.
5 The 1895 census has data on the workforce, industrial establishments and capital, but
not on production. The 1914 census is the most complete; however, that was a year of
deep recession. In 1935, again, the 1930s depression was in full swing. Both these years
are too atypical to be used as markers for establishing trends.
6 Gravil (1985) and Smith (1969). One of the characteristics of the Argentine export
business was precisely that on the one hand the supply was in the hands of many
Roberto CorteÂs Conde 297
producers. On the other, demand was controlled by the trading houses: `the four big
ones', Bunge, Born, Dreyfus and Weil. On several occasions, in connivance with the
railways and thanks to the impossibility of storing the products and the segmented
nature of the credit market, the trading firms made the conditions for producers less
than competitive. In the meat market, at the beginning the price war between British
and USA meat-packing plants favoured the producers. Later, agreements concerning
participation and prices were reached to the detriment of producers. Strong reactions
from the producers ensued, including growing anti-foreign feelings which in the
1920s resulted in demands for state intervention (Smith, 1969).
7 It is assumed that the official value, aforo, should correspond to the market even
though it was not always so because it was easier to correct the tariff by modifying the
aforo values. Thus, if the tariff was 30 per cent and the valuation coincided with the
market's ± called the `real' value ± for example of 100 pesos, 30 pesos were paid in
taxes, but if the market value was 200 pesos and the aforo continued to be 100 pesos,
the 30 pesos which were paid implied an effective rate of only 15 per cent.
8 Until 1890 the issuing, deposit and discount banks had been mainly state-owned.
The private commercial banks (discount) had fewer volumes of deposits.
9 Because of the war, emergency measures came into effect, and even though they were
not used they were useful during the 1930s to make the currency board's regime more
flexible. It consisted of the authorization to issue, by the currency board, based on the
re-discount.
10 The workers' position would change with the devaluations of the 1930s and improve
with Peronism.
References
Alesina, A. (1988) `The End of Large Public Debts' in F. Giavazzi and L. Spaventa (eds),
High Public Debt: The Italian Experience, New York: Cambridge University Press.
Balboa, M. and Fracchia, A. (1959) `Fixed Reproducible Capital in Argentina', in R. Gold-
smith and C. Saunders (eds), The Measurement of National Wealth, Chicago: Quadrangle
Books.
Banco de la NacioÂn Argentina (1923) Report and Balance Sheet.
Bunge, A. (1928) `Los capitales extranjeros en la RepuÂblica Argentina' in Revista de Eco-
nomõÂa Argentina, no. 131.
ÐÐ (1920) Los problemas econoÂmicos del Presente, Buenos Aires. ed. A.E. Bunge.
ÐÐ (1916) Riqueza y Renta en la Argentina, Buenos Aires: Agencia General de Librerias y
Publicaciones.
Comite Nacional de GeografõÂa (1941) Anuario GeograÂfico Argentino, Buenos Aires: ComiteÂ
Nacional de GeografõÂa.
CorteÂs Conde, R. (1997) La economõÂa Argentina en el largo plazo: 1875±1975, Buenos Aires:
Sudamericana.
ÐÐ (1994) `Estimaciones del PBI en Argentina 1875±1935', in Documentos de Trabajo,
Buenos Aires: Departamento de EconomõÂa, Universidad de San AndreÂs.
ÐÐ (1979) El Progreso Argentino, Buenos Aires: Editorial Sudamericana.
Davis, L. and Gallman, R. (1996) `Argentine Savings and Investments, Late XIX Century
and early 20th Century', mimeo.
DõÂaz A.C. (1970) Essays on the Economic History of the Argentine Republic, New Haven/
London: Yale University Press.
298 Argentina 1875±1930
Introduction
This chapter focuses on the analysis of the period we have called the years of
`alteration, crisis and adjustment' in the Cuban economy. Paradoxically, the
interwar period is one of the least-studied of the island's history and perhaps
the most decisive to understand the country's later evolution. Cuba was one of
the countries most affected by the 1930s crisis because of its dependency on the
production and export of sugar, mainly to the USA. Nevertheless, this depend-
ency was maintained long after the depression had passed. Authors such as
Maddison (1988), DõÂaz-Alejandro (1989) and Zanetti (1983) point out that the
island did not undergo a process of diversification as occurred elsewhere.
The link with the US market since independence in 1898, and reinforced during
the Great War, was, they argued, largely to blame. We take a different line. We
will try to prove that the adjustment of the sugar sector to post-First World War
market changes caused a structural crisis in the economic and sociopolitical
system of the island. At the same time, it provided the means to confront this
crisis, a thesis which involves approaching the problem from the supply side
and from a comparative perspective.
* This work was financed by the DGICYT and DGES programmes nos. PD 94-0373 and PB 96-0868,
and by a postdoctoral grant at St Antony's College, Oxford, from the Ministerio de EducacioÂn y Cultura
(Spain). I am grateful for the comments and suggestions made by my thesis director, C.D. Malamud,
by A. GarcõÂa Alvarez, J.A. Ocampo, R. Thorp and M. Lillo.
299
300 The Cuban Export Economy, 1898±1939
Our thesis offers elements to explain the peculiarity of the Cuban economy
and its comparative evolution. Before the interwar period, it presented a level of
development (per capita GDP, alphabetization or infant mortality rates) similar
to that of the most diversified economies in Latin America and, at the same
time, degrees of dependency on foreign trade, on a single product (sugar) and a
single buyer (USA) which only could be compared with the less-developed
countries in the region. This distinctiveness was maintained after the 1930s
crisis (see Bulmer-Thomas, 1994, pp. 59 and 441), in our opinion because it was
the best adjustment option to the changes in the world economic system.
The Great War resulted in a big increase in Cuban sugar production. This
increase continued in the 1920s when world supply recovered. We think the
increase was a direct consequence of the way in which the sugar industry
responded to the reduction in the old continent production during the war,
during what we are calling the period of alteration of the island's export
economy. The market was rapidly saturated and prices began to fall prior to
the 1930s crisis; consequently, the crisis occurred at a time when the Cuban
economy was very dependent on the external sector. However, this was the
necessary but not sufficient condition for the absence of structural change after
the depression. The sufficient condition was the sugar industry's adjustment,
allowing Cuba to reenter the world market as an exporting economy in a better
situation than that of other competitors, to regain a growth rate similar to that
of the most developed countries in Latin America and to restore the internal
sociopolitical order, which had been destabilized by conflicts since the begin-
ning of the 1920s, culminating in a revolutionary uprising in 1933.
No one would question the existence of a direct link between Cuba's independ-
ence and the increase in the supply of sugar, or that access to the US market was
an underlying issue. Although it is difficult to place precise dates on the econ-
omy's evolution, it is fair to say that by the end of the 1830s, when the first
railway was inaugurated, the island had become specialized in exporting sugar.1
In 1847, 50 per cent of Cuban sugar was sold to North America and after 1877
very rarely less than 80 per cent. US demand was concentrated on raw sugar to
feed its own refineries. From the 1860s onwards, island ingenios concentrated on
this product. In SantamarõÂa (1997b) we demonstrated that the commercial
relationship with the USA was accompanied by a strong price convergence
that was altered by the crisis of 1882±83 which depressed the price of sugar
and reduced its buying power by 69 per cent in the next two decades. We also
showed that Spanish policy was consistent with these circumstances. Indeed,
two agreements were signed to facilitate Cuban±US trade. However, the pro-
blem remained unsolved, and in the middle of the 1890s a very high zafra
Antonio SantamarõÂa GarcõÂa 301
Alteration (1914±25): from the First World War to the end of the
sugar boom
With an expanding market and all of its productive capacity now free, Cuban
sugar supply grew by an annual rate of 18 per cent between 1900 and 1913,
rising out of step with North American demand (see Figure 10.1). The alter-
native in a situation such as this would have been to reduce the production or
to look for new customers; however, the First World War eliminated the prob-
lem and replaced it with an urgent need to increase export due to the contrac-
tion of European supply. The island became the Allies' supplier of sugar,
especially after the USA joined the war in 1917 and began to regulate trade.
The way in which the sector responded to the new incentive of demand was
similar to that observed in previous stages: by building 35 factories and, pri-
marily, by increasing output per mill from 147 to 205 tons between 1913 and
1919. This was achieved by acquiring more land and laying more railways ± in
1913 there were 170 000 hectares and 4500 km. respectively; by 1919, 270 000
304 The Cuban Export Economy, 1898±1939
7000
Cuban output
Exports to US
6000 Price
Tons and cts US$/lib × 1000
5000
(thousands)
4000
3000
2000
1000
0
1850 70 90 30
Figure 10.1 Cuban sugar: output, exports to the USA and price of raw sugar, 1850±1939a
a
The price corresponding to 1920 ± 11.95 cts/lb ± is not shown due to its extreme value.
Sources: SACT (1903±36) and CEF (1937±40).
and 11 200. Finally, the endemic problem of labour shortage at harvest time was
solved by importing the workforce from other West Indian islands.12
The speed of sugar expansion during the First World War altered three basic
elements of the Cuban economic and export system, as well as its productive
organization. First, the growth in supply became totally detached from US
demand; second, the process of renewal of capital (vintage-capital) was inter-
rupted.13 Third, the sector had remained in the hands of local owners and the
relative slowness with which it was modernized allowed it to be self-financing,
but the urgent need to produce more sugar during the war resulted in a rapid
concentration and modernization of the structure of ownership in order to
facilitate entry of the necessary investment. Control moved away from industry
towards financial capital, especially that of US origin.14 The alteration of these
elements continued on an even greater scale during the postwar period when
the market situation varied radically. In 1919 Washington stopped regulating
sugar trade, and the price increased from 5.96 to 11.95 cts/lb average between
1919 and 1920. Inflation, however, was caused by a mistaken perception of a
supply shortage. Inevitably, a sudden deflation followed, lowering the price to
3.1 cts/lb in 1921, with serious consequences for the sugar sector and the Cuban
economy. When prices rose, many producers obtained loans on futures pawned
at an average of 10 cts/lb. They were unable to pay them back when they had to
sell at 3.1 cts, and they lost their properties. Generally, these producers had only
started to invest in the business in the war years or were simple speculators.
Banks and financial groups who acquired their industrial and/or agricultural
Antonio SantamarõÂa GarcõÂa 305
a sustained increase in production between 1920 and 1925.16 The sudden rise in
the island's production must therefore be explained from the supply side, and
by analysing the postwar period in the light of the logic of the industry's
growth. Indeed, as we will see below, such an analysis is essential to under-
standing the preservation of the economic system based on sugar exports after
the 1930 crisis. Nevertheless, and even though it may seem contradictory,
before explaining this phenomenon, we must study its outcome.
In the introduction we stated that Cuba did not undergo a process of diversi-
fication similar to that which occurred in other Latin American countries as a
result of the Great War. In general, that event gave rise to increases in exports,
but reduced imports, favouring diversification. The stimulus to this on the
island, where import supply was guaranteed by USA, did not happen in the
same way, and after 1919, because sugar production did not decrease so as to
leave resources for other activities, again the situation did not change.17 The
GDP estimates are constructed with figures heavily dependent on the evolution
of the external sector, and are therefore not reliable as a measure of diversifica-
tion. In 1913 exports appear to have represented 44 per cent of income; 66 per
cent in 1919, and 55 per cent by 1924. In countries such as Chile or Argentina
the per centage was lower ± 11±15 per cent. Meanwhile, the proportion of
industrial output in Argentina's GDP increased from 13 per cent to 17 and 18
per cent in this same period; and in Chile it surpassed 20 per cent by the mid-
1920s exceeding exports in value.
For Cuba, indications such as the employment trend in each sector and the
conclusions of MarqueÂs (1998) show that the lack of structural change indic-
ated in the income estimates reflect the reality. The war and the postwar period
therefore reinforced the dependency of the island's economy on sugar and also
on the US market,18 in spite of the fact that in 1921, due to this double
dependency, the Cuban economic system had shown signs of a structural crisis
caused by the excess of supply of international sugar and the increase in the
North American tariff.
Rising sugar exports from 1922 onwards and production in 1925 had counter-
productive effects. In the medium and long term it worsened the structural
problems of the Cuban economy, and in the short term the opportunity cost of
this strategy was also negative. The following regression equation shows that
the main factor determining the price between 1920 and 1924 was the tariff:19
in the USA; the percentage which the North American tariff represented in this
value, and T and D20, dummy variables to capture the effect of time ± calcu-
lated via the rate of increase in price ± and the alterations which the market
underwent in 1920 due to other factors. According to the equation, ceteris
paribus, and except in 1920, when other factors were of primary importance,
the 500 000 tons sold in New York by the Sugar Export Co. depressed the price
by 13 per cent. In response, the USA raised the tariff from 1.6 to 1.76 cts/lb; in
other words, an increase of 0.76 cts in respect of the 1920 level, which alone
caused a deflation of 50 per cent.
It was to be expected that an increase in Cuban sugar exports would have a
negative opportunity cost in the short term, since the end of this course of
action was to carry out dumping to eliminate competition and maintain the
market achieved during the war. The financial backing from the USA for the
island's industry in 1921 allowed it to happen, and further advantage was
gained from a temporary cutback in world supply and the following price rise
after 1922, due to the crisis in the Rhine which curbed the recuperation of
European production, and to a mosaico plague which devastated many sugar-
cane plantations in Puerto Rico and the North American South.20
However, as can be deduced from the previous equation, it was foreseeable
that protectionism would rule out the success of dumping. Moreover, it would
seem that this situation had a harmful effect on the banks' calculations of the
market trend: as such, they prevented Cuban mills in financial difficulties from
stopping grinding in 1921. Cleveland and Huertas et al. (1985) prove that this
was what happened in the most important case, the National City Bank ± it
eventually controlled 37 factories, which produced 25 per cent of the zafra in
1927. In effect, the island sugar producers knew in 1924 that with the end of
the exceptional conditions of the last triennium, European and US industry
would again grow and depress the price. The curious thing is that they then
decided to increase their own supply by 1 million tons, further saturating the
market and worsening the deflation to the point that historiography considers
1925 as the date marking the end of the upward cycle in sugar production.21 It
was not necessary to continue dumping, since the maintenance or raising of
the level of exports did not require producing more sugar ± there was a large
stock accumulated from earlier years. But, furthermore, that was not the aim,
demonstrated by the fact that immediately after the sharp rise in the zafra, the
same producers asked the Cuban state for the regulation which they had
rejected in 1921.22
A 1926 law stipulated that the mills should reduce their supply by 10 per cent.
There was also an attempt to sign an agreement with European producers and
308 The Cuban Export Economy, 1898±1939
Washington was asked to revise the Reciprocal Trade Treaty. This policy failed
with respect to its external objectives. Cuba was unable to obtain a reduction in
US tariffs, since exports to this country were not regulated, or to reach a
consensus on a compromise in the old continent, to which Java ± a Dutch
colony ± was opposed. Following the establishment of protectionist measures in
Japan and India, both its traditional customers, Java had begun to send sugar to
Europe, and consequently it was not possible to improve the price. In brief, in
1928 it was obvious that the island was capable of swamping the market and
depressing the value of sugar, but incapable of promoting on its own an
improvement in the latter or of reaching the necessary international agree-
ment.23 In the face of failure, regulation was eliminated and in 1929 the mills
were allowed to produce freely again, worsening the saturation of the market.
This occurred just before the beginning of the 1930 crisis which caused a new
increase in North American tariff. Internally, the restrictive legislation was
successful, since by limiting supply it prevented the larger and more efficient
factories ± especially those which were the property of refining companies in
USA ± from monopolizing the demand of these refineries, the main buyers of
Cuban output and payers of better prices.
That the government should consider preventing the largest factories from
monopolizing the market is explained by the changes in the sociopolitical
situation in Cuba in the mid-1920s. We have pointed out above that the
1920±21 crisis undermined the pillars on which the system of domination
rested and the later events contributed to the destabilization of the sociopoli-
tical situation. Faced with growing conflict, the ruling class grouped together
around a project led by G. Machado, who went on to win the elections in 1925.
The project proposed revisions to the Reciprocal Trade Treaty, the regulation of
sugar production to protect the smallest producers, the abolition of the Platt
Amendment, a tariff reform and a public works programme. We have already
discussed the first two objectives and the third did not materialize, but tariffs
and expenditure did change. On average, budgets grew from US$65.3 million in
1919±24 to US$84.4 million in 1925±29 (from 8.2 to 13.6 per cent of GDP); in
1927 a new customs regulation was published and in 1925 the construction of
the central highway began.24
Machado's project was coherent given the difficult situation in which the
country found itself. Paradoxically, however, that very situation prevented
its success in the short term. The alterations which the World War and the
early postwar years had caused in the growth of sugar supply brought about a
structural crisis in the Cuban economic and sociopolitical system. At the same
time, the changes increased dependency on this product and on the North
American market, reducing the options for adjustment. This was the necessary
condition for the maintenance of the system, even after the 1930 crash, but
the sufficient condition was that sugar also afforded the means for dealing
Antonio SantamarõÂa GarcõÂa 309
with the situation. The key was to renew the trade agreements with the USA,
hence the revised bill of the Reciprocal Trade Treaty and the threat to
abolish the Platt Amendment and reform the tariffs to modify the concessions
granted to its imports into Cuba in 1902. This also led to the restriction on
sugar supply, thereby recognizing the responsibility of the island producers in
the strengthening of North American protectionism. However, so that his
policy would work, Machado had to make concessions which thwarted the
achievement of that objective, since what prevented the legislation of 1927±
28 from regulating exports to USA was the pressure exerted by its refiners.25
Machado's bill laid the foundations of what later became the policy to con-
front the depression. However, it also contributed to worsening its effect. The
creation of the oligarchic block and the deferral of the presidential mandate in
1928, in violation of the Constitution, deinstitutionalized the channels for
expressing opposition, especially that of the new social sectors ± the workers'
movement and the urban middle classes. These groups were most affected by
the economic crisis, and they played a leading role in protests which led to the
constitution of the oligarchic block in 1925. Civil confrontation increased as a
result. Furthermore, the failure of short-term economic policy forced a resort to
external credit to finance the rise in expenditure. Between 1900 and 1926 Cuba
obtained US$103 million in loans; between 1926 and 1929, US$189 million and
in 1932 another US$20 million, so that the crisis of the 1930s found the
country in debt and also heavily dependent on the flow of foreign capital,
which broke down in 1929.26
We can now see the significance of the sharp rise in the zafra in 1925, since
basically it caused the adjustment which provided the means to face the crisis
and reinsert Cuba into the reorganisation of the post-depression world market
as an exporting economy. Dye (1998) has shown that the costs of adjustment
inherent in the coordination of the complex sugar production system meant
that mill capacity was underutilized for a few years from its foundation or
technological renewal, and to optimize costs it was necessary to develop this
capacity later. In SantamarõÂa (1995b) we should that the Great War gave rise to
an accelerated increased in industrial infrastructure which had not become
profitable by the end of the conflict. Fifty three percent of the centrales were
in this situation in 1919. They were also, on average, bigger than the rest and
almost all were acquired by new companies founded by the banks, which
averted their closure in 1920±21, or belonged to firms backed by these banks.
From these mills, 63 per cent stepped up their production in 1925 by more than
26 per cent ± a per centage equal to the increase in the island's aggregate supply
with respect to 1924 ± and the rest had in the main developed their optimum
productive capacity before that date. In short, therefore, it can be said that in
1920 Cuba was producing more sugar than the market could absorb, but less
than existing infrastructure allowed.
310 The Cuban Export Economy, 1898±1939
1400.0
Output/refinery
1200.0
Output/refinery/day
1000.0
800.0
Index
600.0
400.0
200.0
0.0
1901 1905 1915 1925 1935 1939
Figure 10.2 Index of sugar output per active mill and effective zafra day, 1901±39
(1903=100)
Sources: SACT (1903±36) and CEF (1937±40).
(until 1928 the yield did not rise above 12 per cent). Between 1913 and 1925,
the cane cut/colony/day rose spectacularly, but not the cane transported by rail
per day, whose growth came in 1925±30. Finally, Zanetti (1996) studied how the
mills adapted their machinery to those changes and points out that in the
1920s most had adopted procedures which allowed them to absorb a greater
inflow of raw material, solving what had been an endemic problem until then
in the sector: intensifying the grinding damaged the yield.
Another reason why the intensivismo of the 1920s went unnoticed is that it
slowed down in the 1930s, although this was due to causes of an institutional
nature. After 1930, state regulation was widened to cover all the elements
involved in the sugar industry. It fixed the dates for the start and end of the
grinding, stipulated workers' wages and cane prices, and asigned colonos and
centrales quotas of cane and sugar prduction and sugar export. This caused a
standstill and factor price rises, which nevertheless were compensated by the
intensification process. After an initial deterioration in the indicators in Table
10.1, even with lower zafras due to the crisis in exports after 1929, the sector
regained efficiency levels greater than those of other competitors. Despite
deflation, it operated only with losses in 1931±32, and after the renewal of
the trade agreements with the USA (1934) it once again produced profits, made
possible by a new reduction in the number of grinding days, the adoption of
small technical innovations, more appropriate selection of raw material,
improvements in the coordination of all the activities to avoid interruptions
in the operation and an increase in the efficiency of work ± these two last
aspects had made little progress during the 1920s.30
312 The Cuban Export Economy, 1898±1939
a
1000 arrobas of cane cut/colono and arrobas of cane transported; sugar in tons; yield: 100 arrobas of
cane/arrobas of sugar.
Sources: SantamarõÂa (1995b, chs. IV and VII); SACT (1913±35); and CEF (1939).
a
Effects measured as per centages of GDP for each year. The tariff and export effects are calculated
estimating the loss of income resulting from failure to supply 50 per cent of US consumption,
following the 1929 tariff; the paralysis in the flow of capital was estimated on the assumption that
Cuba in the 1930s had received the same amount of loans as in the 1920s. The exchange rate was
assumed constant, at the level of 1929.
Source: SantamarõÂa (1995b, ch. VII, and 1998a).
Studying this adjustment of the Cuban sugar industry is the key to under-
standing what happened in its economy in the 1930s. The data in Table 10.2
show that the main depressive factor in this decade and also the most important
obstacle to recuperation was the US sugar tariff, which together with the decrease
of exports reduced the GDP more than the paralysis of the flow of capital or the
fall of exchange rates.32 This is contrary to what occurred in other Latin Amer-
ican countries (see Maddison, 1988; Thorp, ed 1989; and DõÂaz Fuentes, 1994).
This explains why the policy to confront the crisis was also different.
It was rational to expect that the island economy could reinsert itself in the
new post-Depression world order as a sugar exporter. To this end, and in the light
of this new order becoming increasingly determined by bilateral relations, it was
even more necessary to renew the trade treaty with the USA and sign an inter-
national sugar agreement. In 1930 an attempt was made in Brussels, but it failed
because the impact of the crisis on demand was underestimated and importers
were not included. To the USA Cuba proposed a separate pact ± the Chadbourne
Plan ± which was unsuccessful as it did not establish any binding obligations.
Both agreements limited the signatories to supply and export quotas. While
other countries did not comply with them, the island not only complied but
produced and sold below the assigned levels for some years due to the under-
estimation of demand, and maintained this policy after the Brussels Agreement
ended in 1935.33 This form of behaviour, as we said, corresponded to rational
expectations regarding the possibility of renewing the trade treaty with the USA.
To achieve it, Cuban producers recognized their responsibility in the issue:
The fault is ours, the market can carry 3 million tons yet we have sent them 4
million, causing a fall in the price and a defensive move on the part of the
sugar beet producers . . . . [We put forward] as a solution that a commission be
314 The Cuban Export Economy, 1898±1939
sent offering a pact limiting the production to be sold to the US, and
legislation to prevent the producers from exporting above a previously
established quota.
(ANHCC, 1930, pp. 11 and 21; author's translation)
Producers also knew that they had to be united. In 1927, when US refineries
succeeded in preventing the regulation of sugar exports, there had been no
consensus in Cuba. We have already seen that when the island began to refine
they changed their attitude.
The US attitude was also crucial: apart from considerations over the respons-
ibility it acquired in Cuba in 1898, it had to protect many of its citizens'
interests there, while at the same time its exports lost ground in the island
market, one of the most important in Latin America yet one whose purchasing
power had been severely affected by the crisis. Furthermore, a Pacific war was
considered a likely scenario for the 1930s, in which case Cuban sugar would be
necessary since this war would affect two of North America's suppliers ± Hawaii
and the Philippines.
All of these arguments, however, were of little use if US sugar beet producers
did not change their stance. To this end, Cuba limited its zafra and exports in
the hope that, since their increase had been the reason for raising the North
American tariff in the 1920s, their decrease would now serve to reduce it.
Indeed, in this sense there was one further advantage. Beet producers deter-
mined US trade policy, but they were not the ones who benefited most.
Between 1919 and 1930, US sugar production was on average 2.8 million
tons/year, a figure which increased to 4.3 million tons in 1931±9 (55 per cent)
due to the drop in Cuban sales, but sugar beet production only increased from
990 000 to 1.13 million tons (14 per cent). The Philippines, Hawaii and Puerto
Rico were the greatest beneficiaries; they operated with lower costs and even
threatened to push sugar beet out of the market, especially given that the tariff's
depressive effect on the price was forcing Cubans to accept increasingly smaller
payments for their product in order to avoid the customs tariff.
Besides agreeing to stabilize the market, Cuba still needed to fulfil one more
requirement to succeed in renewing the treaty with the USA: the restoration of
the internal sociopolitical order. With the Depression, conflicts increased and
often ended in violence largely because of the deinstitutionalization of the
forms of expression of opposition to the regime. Unable to resolve them,
Machado lost the backing of those who had supported him in 1925, and in
1933 a strike caused his resignation. However, the oligarchy no longer had
control of the state and the revolutionary government which took over
proved equally unable to restore order. Soto (1985) demonstrates that the revo-
lutionaries lacked an alternative to Machado's economic programme. They did
not go far beyond reinforcing its social justice component to meet the demands
Antonio SantamarõÂa GarcõÂa 315
of the middle classes and the labour movement. As a result, the need for agree-
ments with the USA, which would help finance the new social expenditure,
became even more pressing. The key, therefore, was the North American atti-
tude. A speech by Roosevelt before becoming president defined it with precision:
Meanwhile, the social conflict in Cuba had reached extremes which Thomas
(1973) described as an armed stalemate, and the solution arose from that stale-
mate. For Tabares (1973) there were two forces:
The pact was provided by the only institution which was unscathed in the crisis
of legitimacy in the political system: the army; namely its middle ranks who
took control in the so-called RevolucioÂn of the Sergeants (1933). From amongst
them there also came the new strong man of Cuba: Batista, who managed to
reunite the conflicting interests into a common programme:
Now the Republic is born on an irrefutable basis because it will take the form
which the country freely shows it wants. It will not be fascist, or socialist, or
communist, it will take the form that the majority wants to give it.
(Batista, 1933, p. 8, author's translation)
The specific nature of the Cuban economy in the Latin American context at the
beginning of the twentieth century ± a strong dependency on sugar production
and export to the USA, and at the same time levels of development similar to
that of the most diversified economies ± was reinforced during the First World
War and the 1920s, despite the reorganization of the international market
during the postwar period. These changes caused a structural crisis and a
destabilization of the internal sociopolitical order, heightened by the Depres-
sion of 1930. However, the island reinserted itself into the market during the
1930s, maintaining its specialization in the production and sale of sugar, espe-
cially to North America. The lack of diversification was the necessary, but not
the sufficient condition that explains this fact. The sufficient condition was the
adjustment in the sugar industry to confront the structural crisis, and an
economic policy concordant with the specificity of the Cuban economy and
with this adjustment, which allowed the country to reenter the world market in
a better situation than that of other competitors, and to regain a growth rate
similar to that of the most developed countries in Latin America. This allowed
the restoration of the sociopolitical order, thereby promoting a more equitable
income distribution.35
This chapter has thus taken issue with the standard view of Cuba, which
views the significance of the interwar years too simplistically in terms of
continued and extreme dependence on the US sugar market. We have told a
more complex story, which shows how the supply-side response to market
developments in the period from the First World War to the 1930s resulted in
an economic structure in the sugar sector which made expansion of production
an imperative, while this fact and the oligarchy consolidated behind Machado
created sociopolitical tensions only partially resolved after the Revolution of
1933, in a way which allowed political developments to reinforce the tight
relationship with the USA. The successful continuation of the preferential
relation with this country was a result of Cuba's competitiveness, of a stable
political situation, and energetic negotiations. The outcome in the 1930s at
least was a relatively good economic performance.
Notes
1 See Sanz et al. (1998, pp. 373±90), for the relation between the integration of Latin
American economies in the world market and the construction of the first railways;
Zanetti and GarcõÂa (1987, pp. 1±20) and SantamarõÂa (1995a and 1998b) for the specific
case of Cuba.
2 In the 1868±90 period, production stagnated somewhere between 579 000 and 775 000
tons due to the reorganization of the sector in those years. Between 1891
Antonio SantamarõÂa GarcõÂa 317
and 1897 it was always above 800 000 tons, reaching a record of 1.1 millions. The FOB
price in the USA fell from 3.40 cts/lb to 2.07 between 1890 and 1897 and the tariff
rose from 0 cts in 1890, to 40 per cent ad valorem in 1894 and to 1.685 cts/lb in 1897.
See Moreno (1978) Vol. III. The last general published work on the years prior to
Cuban independence is that by GarcõÂa (1996). Ingenio is the word for a Cuban sugar
factory; central is a completely mechanized ingenio; zafra is the word for the cane
harvest and it is utilized as a synonym of sugar production.
3 See PeÂrez de la Riva (1975) and GarcõÂa (1994b) regarding imports of Chinese coolies
and YucataÂn Indians.
4 See Moreno (1978), Vol. I and Deerr (1950) for an account of the technological
evolution of the industry.
5 In Cuba cane takes between eight and 12 months to grow, whilst in other places in
the world it takes 18. Furthermore, the shoots of each plant are normally used for six
or seven years. See MartõÂn et al. (1987).
6 The natural division between agriculture and manufacturing, explains the need for
building a transport system which would communicate between the canefield and
the central ± a private railway. The industrial sugarcane process can be divided into
three parts ± grinding, evaporation and filtering ± which must be perfectly coordin-
ated. A technical improvement which, for example, increases the flow of cane juice ±
guarapo ± from the mill to the evaporators is not efficient and causes a bottleneck if it
is not used in conjunction with innovations to these evaporators to absorb the
increase. For more on this subject see Dye (1998), pp. 102±49.
7 The construction of centrales meant a reduction in the number of factories and an
increase in production per unit ± 1365 ingenios and 328 tons in 1860; 850 and 445 in
1890, and 205 (all of them centrales) and 1464 in 1900. See Le Riverend, (1985: 490).
8 See Le Riverend (1973), Perez (1986), GarcõÂa, (1998) and Zanetti (1989 and 1998), for
more on this subject.
9 For more on the new legal framework, see Jenks, (1928) and Le Riverend (1973). With
regard to the expansion of the sugar industry in the Eastern half of Cuba, Hoernel
(1976) and LuzoÂn (1989). For more on the extension of the railway, Zanetti and
GarcõÂa, 1998 and SantamarõÂa, 1995a and 1998b.
10 The investment of foreign capital mainly in rail rather than in the sugar industry was
not a phenomenon specific to the twentieth century, and nor was it confined to US
investments. In SantamarõÂa (1998b) p. 303, we explain that this was probably due to
opportunities for self-financing which enabled a slow modernization of the mills.
This did not happen in the rail sector, where large injections of capital were necessary,
either to build a complete network of transport ± Eastern provinces ± or to consolidate
in a single company a dozen lines which had emerged in the Western half of Cuba to
satisfy local needs, and which over time had stopped making a profit due to competi-
tion amongst themselves. US investment did not rise above US$ 50 million in 1896
and of that amount 30 was placed in the sugar industry. Around 1911 the first figure
had grown to approximately 200 million, whilst the second only rose to 50. Rail
construction accounted for 40. See Pino, 1984: 342, Jenks, 1928, chs. I-VI, and LoÂpez
Segrera, 1973: 173.
11 Data from Brundenius (1984)
12 According to Zanetti and GarcõÂa, 1976: 212, the first official licence to import West
Indian casual-workers was obtained by the United Fruit Co. in 1913. From that date
and until the mid 1920s more than 500 000 entered Cuba in order to work in the
sugarcane harvest, coming mainly from the other Caribbean islands and from Spain.
See PeÂrez de la Riva (1975).
318 The Cuban Export Economy, 1898±1939
24 See LoÂpez Segrera, 1980, and Thomas, 1973, vol. II. Pollitt, 1989, is the study which
best explains the relation between the crisis of sugar production and the arrival to
power of Machado. See also SantamarõÂa, 1994, 1995b and 1998a, and IHC, 1998, vol.
III: 240.
25 With regard to the tariff, MarqueÂs, 1989 proves that it was a reaction to US trade
policy and its objective was to alleviate the reduction in state income, heavily
dependent on customs earnings which were depressed by the fall in the price of
sugar, rather than a measure to encourage economic diversification.
26 On the effect of the depression on external credit see Kindleberger, 1985, Galbraith,
1989 and Fishlow, 1985; with reference to loans granted to Cuba, IHC, 1998, vol. III:
405.
27 For more on how in the second half of the 1920s the island industry became again the
most competitive in the world, see Dye, 1993: 586, and USTC, 1926; about the sugar
yield in 1925, SantamarõÂa, 1994: 136.
28 MartõÂn et al., 1987: 567.
29 The effective length of the zafra ± the average time taken by factories to do it ± was
reduced from 132 days in 1919, to 107 in 1925, 90 in 1930, 61 in 1935 and 62 in 1939.
See CEF, 1940.
30 MartõÂn et al., 1987: 576 and Dye, 1991 show that the interruptions in the operation
are the worst possible drawbacks for an effective production process.
31 Apart from sugar refined, the Cuban sugar industry began to produce rich molasses
and syrups, and increase the distillation of rum, demanded by US consumers. See
SantamarõÂa, 1995b, ch. VI.
32 Dye and Sicotte, 1999, calculate that nearly 35 per cent of the losses in real export
earnings between 1930 and 1934 was caused by the US sugar tariff.
33 On bilateralism in trade in the 1930s, see Kenwood and Loughed, 1972: 337±340. On
the Brussels Agreement and the Chadbourne Plan, Zanetti, 1989: 133±135 and Bal-
linger, 1971: 37.
34 For more on all these aspects see SantamarõÂa, 1998a and 1995b, ch. VI. In these
studies we analyse the results for the Cuban economy of the international agreement
and the new treaty with the US in a comparative perspective. This analysis confirms
our conclusions about the relative success of the adjustment in this economy to the
post-Depression market situation. Another aspect also analysed is why some mea-
sures thought to be to confront the crisis were maintained subsequently and what
problems this meant.
35 We have not enough space in this chapter to explain in depth the adjustment in the
Cuban economy and its results in comparative terms. In SantamarõÂa, 1998a and
1995b, ch. VI, the reader may find the data about the more equitable income
distribution and the island's macroeconomic statistics compared to those of other
Latin American countries which prove our affirmations.
References
Ballinger, R.A. (1971) A History of Sugar Marketing, Agricultural Report no. 197, Washing-
ton, DC: US Department of Agriculture.
Batista, F. (1933) Al pueblo de Cuba, Havana: Lex.
Bernhart, J. (1948) The Sugar Industry and the Federal Government, a Thirty Years Record,
Washington, DC: Government Printing Office.
Brundenius, C. (1984) Revolutionary Cuba: The Challege of Economic Growth of Equity,
Boulder: Westview Press.
Bulmer-Thomas, V. (1994) The Economic History of Latin America since Independence, Cam-
bridge: Cambridge University Press.
Cayuela, J.G. (ed.) (1998) Un siglo de historia de Espan Ä a: centenario, 1898±1998, Cuenca:
Universidad de Castilla-La Mancha.
Chantez, S. and FernaÂndez, J.M. (1985) `El fenoÂmeno econoÂmico del intensivismo en las
zafras azucareras de la deÂcada de 1940', Islas, no. 80, pp. 3±8.
CEF (Cuba EconoÂmica y Financiera) (1937±58) Anuario Azucarero de Cuba, Havana: CEF.
CII (Cuba Importadora e Industrial) (1932±36) Havana: CII.
Cleveland, H.B., Huertas, T.F. et al. (1985) Citibank, 1812±1970, London: Harvard Uni-
versity Press.
Collazo, E. (1994) Una pelea cubana contra los monopolios (un estudio del crac bancario de
1920), GijoÂn: Universidad de Oviedo.
Cuba ContemporaÂnea (1913±26) Havana: Cuba ContemporaÂnea.
Deerr, N. (1950) The History of Sugar, 2 vols, London: Chapman & Hall.
DõÂaz-Alejandro, C. (1989) `AmeÂrica Latina en los an Äos treinta', in R. Thorp (ed.), (1989)
pp. 13±35.
DõÂaz Fuentes, D. (1994) Crisis y cambios estructurales en AmeÂrica Latina. Argentina, Brasil y
MeÂxico en el perõÂodo de entre-guerras, Mexico: Fondo de Cultura EconoÂmica.
DGC (DireccioÂn General del Censo) (1907, 1919 and 1943) Censo de la RepuÂblica de Cuba,
Havana: DGC.
Dye, A.D. (1998) Cuban Sugar in the Age of Mass Production. Technology and the Economies of
the Sugar Central, 1899±1929, New York: Stanford University Press.
ÐÐ (1993) `ProduccioÂn en masa de azuÂcar cubano, 1899±1929', Revista de Historia Eco-
noÂmica, vol. 11 (1), pp. 563±93.
Dye, A. and Sicotte, R. (1999) `The U.S. Tariff and the Cuban Revolution', umpublished,
New York: Barnad College, and Boston: University of Harvard.
Fishlow, A. (1985) `Lessons from the Past: Capital Markets during the 19th Century and
the Interwar Period', International Organization, vol. 39 (3), pp. 73±87.
Galbraith, J.K. (1989) El crac de 1929, Barcelona: Ariel.
GarcõÂa, A. (1998) `La consolidacioÂn del dominio imperialista', in Instituto de Historia de
Cuba, Historia de Cuba, IHC, (1998) Vol. III, pp. 90±141.
ÐÐ (1996) `Estructuras de una economõÂa colonial en transicioÂn', in C. Naranjo et al.
(eds.), 1996, pp. 195±210.
ÐÐ (1994a) `Una saga azucarera en torno a dos siglos', in J. UrõÂa (ed.), 1994., pp. 43±56.
ÐÐ (1994b) `Los traficantes del Golfo', Historia Social, vol. 18, pp. 33±46.
ÐÐ and GarcõÂa Mora, L.M. (eds) (1998) Textos claÂsicos de la historia de Cuba (CD-Rom),
ColeccioÂn ClaÂsicos Tavera, Madrid, Fundacio n Histo rica Tavera y Digibis.
Hoernel, R.B. (1976) `Social and Social Change in Oriente, Cuba, 1898±1946', Journal of
Latin American Studies, vol. 8(2), pp. 215±49.
IHC (Instituto de Historia de Cuba) (1994±) Historia de Cuba 5 vols, Havana: PolõÂtica.
Jenks, L.H. (1928) Our Cuban Colony, New York: Vanguard Press.
Kenwood, A.G. and Loughed, A.L. (1972) Historia del desarrollo econoÂmico mundial, Madrid:
Istmo.
Antonio SantamarõÂa GarcõÂa 321
financial system 7 Indians 10, 90±1, 121, 170, 179, 181 n 14,
weakness of Peruvian 157 181 n 15, 181 n 17
First World War 1, 2, 11, 74, 77, 122, 224, industrial revolution 3
265 industrialization 56, 65±6, 75±81, 82
effect on Argentine growth 269, 288±90 n 10, 82 n 12, 82 n 13, 77±9, 126, 128,
effect on Bolivian tin 192±4 132, 145 n 56, 156, 180 n 8, 181
effect on Cuban sugar production 300, n 11, 203, 227, 256, 267, 269, 281±2,
303, 306 293
effect on Peruvian cotton 168 inflation
effect on Peruvian wool trade 161±2 66, 132, 203±4, 241, 260 n 35, 260 n 36
foreign debt institutions 3, 5, 6
in Bolivia 201, 205 rigidity in (Argentina) 267
in Central America 99 changes in (Argentina) 290±2
in Chile 217, 235 instability 20±1
in Peru 158 fiscal in Colombia 66
foreign investment monetary in Central America 97
in Argentina 283 political in Central America 115
in Bolivia 194 of Chilean export sector 232±3
in Cuba 304±5 internal migration
in Mexico 126 in Central America 101
in Peru 163, 168, 174 in Colombia 63
frontier 266±7, 271, 293 in Peru 165
role of women and children in labour living standards 126, 128, 130
force (Colombia) 80±1 markets 120
scarcity in tin mining during Chaco oil boom 132
War 198 political instability 121±2, 125, 131
scarcity of (Cuba) 301 population 132
sharecropping (Peru) 167 Porfirian regime 126±30
use of coercion in Central railways 120, 125±6, 127±8, 142 n 8
America 90±1 regional fragmentation 119, 121,
use of Jamaican workers in Central 128±9
America 95±6 relations of production 128±9, 134±5
Lamont, Thomas 133 role of state 136±7, 147 n 77
land similarity to Andean economies 121
changes in use of (Argentina) 270±1 size of state 119±20
ownership of 7 social conflict 135
usurpation of peasant (Peru) 161 social reform 136
liberalism 89, 105, 127, 189, 291 structure of foreign trade 133
linkages, 20, 22±3 technological innovation 126
absence of forward linkages in coffee and timing of export boom 124±5
banana production in Central unionization 135±6, 145 n 50
America 96±7 US influence 123, 134, 143 n 21
coffee in Brazil 46 wages 130
coffee in Colombia 76 mining 9, 57±9, 173±4, 223±4
from nitrates (Chile) 226±7 Montes, Ismael 206
in Bolivia 209±10 Montt, Manuel 217
lack of in Peruvian oil 169±70 Murtinho Plan 42
rise and fall in copper (Peru) 177 Mussolini, Benito 279
shift from direct to indirect (Chile) 228
theory of 220±2 National Federation of Coffee-Growers
literacy (Col.) 73, 74, 75
in Bolivia 206 Nicaragua
in Central America 115 bananas 94
in Mexico 120 coffee 91
liberal reforms 105
Machado, President 308, 309, 314, 316 role of immigrants 92
manufacturing 23 nitrates
Meller, P. 253 annexation of 218
Mexico demand for Chilean 222
Agrarismo 135 exports of 224±5
Cardenismo 137 production of synthetic 222±3
banks 141 n 4 threat from ammonium sulphate
constitutional reforms 134±5 to 223
domestic agriculture 128 North, John Thomas 224, 230, 252
ethnic makeup 121 Nun Äez, Rafael 65
exports 127±8, 131±2
foreign investment 126, 127, 133±4 oil
hyperinflation 132 in Mexico 123, 132, 33
impact of Revolution 122±4, 130±1 in Peru 169±70
industrialization 126, 128, 132, 145 Orozquismo 131
n 56
labour market 130, 142 n 17 Panama 87, 94
limits of quantitative data 137±41 path dependence 6
literacy 120 Äo, SimoÂn I. 193, 201±3, 205, 209
Patin
328 Index
Peru in Brazil 39
agriculture 180 n 9 in Chile 227
Cerro de Pasco Corporation 174±5 in Cuba 302
commercial systems 159, 168, 172, in Mexico 127±8
175, 179 in Peru 160, 163, 174, 175
cotton production 167±8 RamõÂrez Necochea, H. 240
demonetization 160 Reciprocal Trade Treaty (Cuba) 302±3,
domestic investment 156 308
export boom 1890s 155±6 returned value 20, 113, 177, 229±31
impact of guano 153±4 Reyes, Rafael 55, 67, 70, 76, 77
impact of the War of the Pacific 154 Roca Runciman Agreement 285
industrialization 156, 180 n 8, 181 n 11 Roosevelt, President 315
labour 164±6, 177, 179 Rosa, Jose MarõÂa 273
merchants 160 rubber
oil production 169±70 in Brazil 37
Peruvian Corporation 154 in Peru 170±3
politics 161±2, 179
railways 158, 159, 180 n 6, 186 n 33 Santa MarõÂa, Domingo 219, 245, 250, 252
regional linkages of exports 160, 161, Santos Zelaya, Jose 89, 100
166±7, 169, 175±6, 178±9 Sherman Act 42
regional fragmentation 152, 153 slavery 6, 7, 9, 18, 39, 52 n 15, 57, 301
role of immigrants 163, 168 social indicators 113±16
role of state 158 Somoza, Anastasio 105
rubber production 170±3 Standard Fruit Company 95
shifts in export quantum 157 staple theory 6, 20, 22, 265, 266
social conflict 177, 161, 181 n 17 state, role of 27, 48, 70, 76, 103±4,
sugar production 163±4, 182 n 18 119, 123, 127, 136±7, 158, 180 n 5,
US influence 158, 182 n 20 196±7, 199, 207, 211, 217±18, 231,
urbanization 160, 161, 171 231, 234, 235±6, 248±55, 283, 287±8,
usurpation of peasant land 161, 181 289, 311
n 14 statistics, problems with, 137±41
wool production 159 steamships 9
Phelps, Vernon 290 stock exchange 3
PieÂrola, NicolaÂs 155 sugar
Pinto, President 218 in Brazil 37
Pinto, AnõÂbal 240 in Chile 239
Polanyi, Karl 140 in Cuba, 300±16
Platt Amendment 302, 308, 309 in Peru 163±7
population 33, 63, 71, 81 n 4, 86, 132,
189, 267±8, 276, 283 tariffs 9, 23, 46±7, 65, 76, 155±6, 200,
protectionism, 5, 8, 13, 19, 20, 24 234, 237±40, 284±5, 302, 313, 314
in US 272, 307, 309 taxes 9, 27, 171, 199±202, 234±5, 250,
in Colombia 63±4, 76 251
in Chile 237±8 technological innovation 3, 4, 16±17, 66,
in Argentina 267, 280, 285, 291, 294 103, 162, 163, 167±8, 176, 226, 274,
301, 310, 311
quinine 56 terms of trade 13, 34, 69, 231±2, 270
Thousand Days War 60, 62, 65, 66, 67,
racism 121 70, 71
railroads/railways 9 tin, 188, 190±204
in Argentina 274±4, 280±1 crisis and growth in the 1920s 194±5
in Bolivia 204±6 effect of Chaco War 197±8
Index 329