Bulmer-Thomas V., Coatsworth J., Cortes-Conde R. The Cambridge Economic History of Latin America: Volume 2, The Long Twentieth Century
Подождите немного. Документ загружается.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
252 Richard Salvucci
uses ostensibly homogeneous and comparable data to provide a frame of
reference. The two sources I employ, Andr
´
eA.Hofman, The Economic
Development of Latin America in the Twentieth Century, and Angus Mad-
dison, The World Economy: A Millennial Perspective, have common roots
at the University of Groningen and share both methodology and sources.
3
Inevitably, data on population and production in Latin America before
1870 are apt to induce a good deal of skepticism. Maddison’s reliance on
the work of Angel Rosenblat for population data is particularly problem-
atic for the specialist. But this has nothing to do with the period after 1870,
where we can proceed with more confidence.
As to Mulhall’s claim that rapid economic progress dated from the
1860s, it is impossible to demonstrate, at least to judge from Hofman
and Maddison. The earliest data for which we have standardized estimates
are for 1870.Tables 7.1 and 7.2 provide estimates of the average annual
compound growth of merchandise exports and real gross domestic product
(GDP) for some seven countries. The median growth rate of merchandise
exports in 1870–1913 was 3.4 percent in this, the Augustan age of the “export
economies.” But, perhaps surprisingly, the median rate in 1913–29 is even
higher, 3.8 percent, which suggests (imperfectly, to be sure) that it is pos-
sible to exaggerate the long-run consequences of World War I. Argentina,
Mexico, and Venezuela all experienced lower rates in 1913–29 than pre-
viously, and in Argentina, the decline was grievous, some 2.5 percentage
points per year. But Mexico surely reflects the impact of the revolution
that lasted from 1910 through 1920 as well as the precipitous decline in
oil production and export in the 1920s. Surprisingly, the substitution of
Venezuelan for Mexican production that occurred at this point (Venezuela
succeeded Mexico as the world’s largest exporter because of discoveries in
the Maracaibo basin) did not arrest the fall in the rate of Venezuelan export
growth, which was 0.3 percentage points per year higher in 1870–1913 than
in 1913–29. Whereas Brazilian export growth remained steady at 1.9 per-
cent in 1870–1929, and Brazil came to be the largest coffee producer in
the world, Chile, Colombia, and Peru all experienced increases in exports
in 1913–29, ranging from the merely noticeable to absolutely staggering.
In Chile, the increase reflected the substitution of copper for nitrates after
the war and substantial exchange depreciation as well, although the Great
Depression brought the growth of the 1920stoadecisive end. Peru’s exports
3
Andr
´
eA.Hofman, The Economic Development of Latin America in the Twentieth Century (Cheltenham,
UK, 2000); Angus Maddison, The World Economy: A Millennial Perspective (Paris, 2001).
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
Export-Led Industrialization 253
Table 7.1. Merchandise export growth average
annual compound rate (1990 prices)
1870–1913 1913–1929
Argentina 5.22.7
Brazil 1.91.9
Chile 3.43.9
Colombia 2.06.8
Mexico 5.42.7
Peru 1.76.2
Venezuela 4.13.8
Median 3.43.8
Sources: Andr
´
eHofman, The Economic Development of
Latin America in the 20th Century (Cheltenham, UK,
2000), 23; and Angus Maddison, The World Economy: A
Millienial Perspective (Paris, 2001), 361.Hofman draws
on an earlier publication of Maddison’s, Monitoring the
WorldEconomy, 1820–1992 (Paris, 1995) for 1870–1913,
p. 236. The base year, 1990,isconsistent throughout.
were driven by a sharp change in their composition away from sugar and
cotton toward more rapidly growing petroleum, particularly in the 1920s.
In Colombia, the expansion was led by coffee, the exports of which dou-
bled in volume in the 1920s, according to Jos
´
e Antonio Ocampo and Mar
´
ıa
Mercedes Botero.
Table 7.2. Real GDP average annual compound growth
(1980 international dollars)
1870–1913 1913–1929 1900–1931
Argentina 6.02 3.56.4
Brazil 2.38 4.74.5
Chile 2.93.7
Colombia 4.74.2
Mexico 3.38 0.82.6
Venezuela 2.88.23.3
Source: Hofman, The Economic Development of Latin America in the
20th Century,pp.30 and 167.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
254 Richard Salvucci
As to the geographical composition of trade, Victor Bulmer-Thomas has
put the matter well, writing that “Britain and Latin America appeared to be
made for each other in the nineteenth century.” Britain, of course, supplied
manufactured goods such as cottons in which it had strong comparative
advantage, and Britain could finance its exports as well. Latin America, for
its part, met a growing demand for food and raw materials in Great Britain.
ButbyWorld War I, Britain had been overtaken as a trade partner by the
United States in most countries, although Argentina remained a crucial
British market down to the 1930s.
Finding comparable growth rates of GDP in 1870–1930 is not easy.
Drawing on Maddison’s data, which are expressed in terms of 1990 inter-
national dollars, for 1870–1913, Argentina, the fastest grower, registered
an average annual compound rate of 6 percent. Mexico was slower, but
still substantial, at 3.4 percent. Venezuela managed a quite respectable 2.8
percent. Brazil was the laggard at about 2.4 percent.
Comparing this same group from World War I through the Great
Depression is instructive. The substantial reduction in Argentine exports
paralleled a sharp decline in output growth of 2.5 percentage points per year,
or about 40 percent, confirming that World War I was a caesura in the coun-
try’s economic history. The war’s effect on other “export economies” was by
no means uniform. Brazilian output increased even as exports stagnated.
Venezuelan output rose by 5.4 percentage points, which is to say, nearly
doubled, even as the rate of export growth declined. Growth in Mexico
collapsed, of course, but distinguishing between the international costs of
World War I and the loss in output exacted by civil war and revolution
in Mexico is not straightforward. By the outbreak of World War I, a strict
correspondence between movements in exports and total output no longer
held, another reason for regarding facile characterizations of the “age of the
export economies” with a grain of salt.
THE GOLD STANDARD AND THE LATIN
AMERICAN PERIPHERY
This is not the place for an exhaustive treatment of the history, theory, or
practice of the gold standard and its variants. The reader may consult the
bibliographical essay for a representative selection of items that do so. Here,
perhaps, all that is needed is a basic sketch.
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
Export-Led Industrialization 255
Basically, the gold standard was a monetary regime in which the value
of the home currency was defined in terms of gold. What this meant, in
effect, was that the value of the currency was defined not only in terms of
domestic circulation but also in terms of foreign currencies fixed to gold.
Because the price of gold was broadly stable, a prerequisite for defining any
currency in terms of a commodity, this meant, in essence, that countries
“on the gold standard” were linked by a system of fixed exchange rates.
Crucially, because the home currency was defined in gold, one could imag-
ine it circulating beyond a country’s frontiers and within the boundaries
of other countries linked to gold, which was, indeed, sometimes the case.
In the simplest model, fluctuations in a country’s international position,
which, by assumption, are limited to merchandise trade alone, would pro-
duce variations in the domestic money supply: namely, an export surplus
gained money; an import surplus lost it. Because the domestic price level
depended in part on the supply of money, and because prices are inversely
related to scarcity, a rise in the domestic money supply (reduced scarcity)
raised prices; a fall in the money supply (increased scarcity) reduced them,
assuming an unchanged demand for money in either case. Again, in the sim-
plest case, famously associated with the Scottish philosopher David Hume
(1711–76), export and import surpluses would be self-correcting because
the rise or fall of domestic prices would reduce or improve the competi-
tiveness of home goods relative to substitutes produced abroad. Thus, an
export surplus would induce an inflow of cheaper substitutes produced
abroad, whereas an import surplus would lower prices and spur exports.
For this reason, the Humean model of the gold standard was said to be
self-regulating or self-adjusting, although it is important to realize that
the export- and import-competing sectors would experience real variations
in their production. Thus, even under the simplest account of the gold
standard, domestic economic performance and a country’s international
position are inextricably linked.
The system becomes somewhat more complicated when borrowing and
long-term lending between countries takes place because the balance of
payments must now reflect capital as well as merchandise movements. The
adjustment to disturbances in the balance of trade may change as well.
Strictly speaking, it would be difficult to find any international financial
historian who would agree that the gold standard operated in the sim-
ple Humean way or by following what have been traditionally called “the
rules of the game.” From the standpoint of the major European powers
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
256 Richard Salvucci
that participated in the gold standard, namely Great Britain, France, and
Germany, and from that of the United States as well, the operational details
of the gold standard, not to mention its much vaunted “automatic” adjust-
ment process, have been largely relegated to the status of one of the foun-
dation myths of international financial history.
Butifthe view from the center has gradually moved to a somewhat
less orthodox consensus on how the gold standard (or, increasingly after
1880, the gold exchange standard, in which currencies convertible into
gold could also be held as reserves) worked, the view from the periphery
has always been a skeptical one. Perhaps Alec Ford presented the most
eloquent and influential criticism of the operation of the gold standard
in Latin America in his book, The Gold Standard 1880–1914.Britain and
Argentina.
4
Fundamentally, Ford argued that the automatic and smooth
adjustment of macroeconomies under the gold standard was, if not wholly
illusory, more properly attributed to its operation in and impact on Great
Britain. Moreover, for a peripheral economy engaged in exporting primary
commodities, the operation of the gold standard amplified booms and busts
rather than mitigating them, so that domestic adjustment to international
imbalances was anything but smooth. In other words, the gold standard
operated procyclically in the periphery.
For example, Ford argued that a country whose growth was driven by
exports of primary commodities would experience large inflows of specie
under the gold standard. Thus, growth would be accompanied by a mone-
tary expansion, rather than by the automatic tightening the Humean mech-
anism contemplated. For Ford, such tightening was more likely to occur in
countries in which growth followed an autonomous increase in domestic
investment to which no monetary accommodation necessarily occurred –
in other words, in Great Britain rather than in Argentina. Moreover, Ford
believed that the Bank of England could stem an outflow of gold by raising
interest rates, the so-called Bank Rate. Countries with primitive finan-
cial systems and no central bank could do no such thing, which meant
that Latin America under the gold standard lacked a policy instrument
available to the center countries and was thus unable to moderate the
effects of a gold outflow consequent on a trade deficit. Further, because
countries like Argentina (the argument has been made for Colombia by
William McGreevey as well) did not produce substitutes for the goods they
imported from Great Britain (capital goods), their demand for imports
4
Alec Ford, The Gold Standard 1880–1914.Britain and Argentina (London, 1962).
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
Export-Led Industrialization 257
would tend to be relatively inelastic. As a result, trade deficits would per-
sist unless home demand fell dramatically. So the adjustment process in
Latin America would be subject to busts, or deep recessions – an awkward
accompaniment to the monetarily induced booms that occurred during
expansions. The system, in Ford’s view, was far from stabilizing in Latin
America and, in fact, could be viewed as inherently destabilizing or, in the
jargon, “procyclical.”
As if all this were not enough, several other unpleasant facts about the
gold standard in Latin America followed. Because export-oriented elites in
Argentina (and the argument has been made for Brazil and Chile as well)
were paid in gold, they had a natural bias in favor of currency deprecia-
tion. Where domestic monetary systems included a paper component that
fluctuated in terms of gold, export elites naturally favored depreciation, for
their export receipts could then purchase more domestic goods, or non-
tradables, such as land. In the event of an appreciation, or increase in value
of the currency, elites were likely to favor suspension of gold convertibility:
abandoning the gold standard, in other words.
Another unpleasant fact was that commodity and capital shocks in Latin
America would tend to go hand-in-hand. What Ford termed the “cushion”
available to Great Britain of attracting capital when gold flowed out was
not likely to be available to Latin America because lenders would be highly
unlikely to advance funds to countries whose primary source of repayment,
receipts from commodity exports, had fallen.
A third unpleasant fact for Latin America was the very instability of com-
modity prices relative to the price of manufactures. This is not so much an
example of the “declining terms of trade” thesis as it is the simple recogni-
tion that agricultural prices will inevitably be more variable than the price
of industrial goods because of the inherently greater instability of supply.
This meant, in essence, that export receipts would tend to be more variable
than a country’s import bill, particularly given the structure of international
trade in the nineteenth century, when Latin America exchanged a variety
of commodities for wage goods such as cotton cloth or for capital goods as
domestic industrialization there got underway.
These unpleasant facts about the gold standard meant that countries in
the periphery would not necessarily find the operation of the gold standard
beneficial, and that even if they did, their ability to adhere to the gold stan-
dard in the long run would be highly doubtful. And, in fact, the record of
Latin America in this period is one of spotty, imperfect, and variable adher-
ence to the gold standard, an interesting commentary on the instability of
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
258 Richard Salvucci
Table 7.3. Latin America and the gold standard, 1870–1930
Argentina
1876–1881 inconvertible paper standard
1881–1885 gold standard
1885–1903 inconvertible paper standard
1903–1914 gold standard
1914–1927 inconvertible paper standard
1927–1929 gold standard
Bolivia
1825–1914 silver standard
1914–1928 inconvertible paper standard
1928–1931 gold exchange standard
Brazil
1822–1906 inconvertible paper standard
1906–1914 gold standard
1914–1927 inconvertible paper standard
1927–1930 gold standard
Chile
1875–1895 inconvertible paper standard
1895–1898 gold standard
1898–1926 inconvertible paper standard
1926–1932 gold exchange standard
Colombia
1871–1885 gold standard
1885–1903 inconvertible paper standard
1903–1914 gold standard
1914–1923 mixed standard
1923–1931 gold exchange standard
Costa Rica
1871–1900 bimetallic standard
1900–1914 gold standard
1914–1924 inconvertible paper standard
1924–1932 gold exchange standard
Ecuador
1875–1884 inconvertible paper standard
1884–1900 bimetallic standard
1900–1914 gold standard
1914–1926 inconvertible paper standard
1926–1932 gold exchange standard
El Salvador
1883–1892 bimetallic standard
1892–1894 gold standard
1894–1914 silver standard
1914–1919 inconvertible paper standard
1919–1931 gold exchange standard
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
Export-Led Industrialization 259
Table 7.3. (cont.)
Guatemala
1870–1897 bimetallic standard
1897–1925 inconvertible paper standard
1925–1933 gold exchange standard
Honduras
1879–1931 silver standard
1931–1934 gold exchange standard
Mexico
1867–1905 silver standard
1905–1913 gold exchange standard
1913–1925 mixed standard
1925–1931 gold exchange standard
Paraguay
1885–1923 Inconvertible paper standard
1923–1932 Argentine peso standard
Peru
1875–1898 inconvertible paper standard
1898–1914 gold standard
1914–1930 inconvertible paper standard
Uruguay
1865–1914 gold standard
1914–1931 inconvertible paper standard
Venezuela
1879–1931 gold standard
Notes and Sources: The institutional arrangements and monetary regimes
under which Latin American operated between 1870 and 1930 are fraught
with confusion. Few authors provide precisely the same chronology and
detail. The authoritative attempt to sort these matters out is of David
George Cowen, “A World of Difference: Exchange Rate Regime Choice
and Economic Performance in the Interwar Years” (Ph.D. dissertation, The
University of Texas at Austin, 1995), Appendix 3.
monetary regimes during the period. The definitions and dating of being
“on the gold standard” vary widely, but Table 7.3 provides some details of
the adherence of Latin America to the gold standard.
On the whole, then, it suggests that adherence to the gold standard in
Latin America between 1870 and 1930 was difficult, and that the advantages
of maintaining some form of gold convertibility may have been illusory.
In fact, Albert Fishlow has suggested that the merits of remaining on a
floating exchange rate there were substantial, and that these advantages,
rather than the preference of export elites for currency depreciation – not
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
260 Richard Salvucci
a consideration that should be underestimated – explain why Latin Amer-
ica under the gold standard might as fairly be termed Latin America “on
and off ” the gold standard. For, if nothing, else, as Keynes pointed out,
the system of fixed exchange rates efficiently transmitted economic crises
between countries, a dubious benefit for peripheral economies rendered
more, rather than less, stable by the monetary regime.
The principal implication of this discussion is the difficulty of consistent
long-term growth from peripheral countries under the gold standard. If,
in fact, macroeconomic stability were not a strong point of the system,
the ability of the Latin American economies to provide adequate levels
of income and satisfactory rates of growth would be quite questionable.
As Victor Bulmer-Thomas has pointed out, the main reason why income
levels there in the twentieth century are well behind those of the devel-
oped countries in general, not to mention the Organization for Economic
Cooperation and Development (OECD) members (Mexico excluded, of
course), is because consistent, long-term performance is virtually impossi-
ble to find. Volatility, in other words, is as much a cause as a consequence
of “underdevelopment,” and it is difficult to see how the currency regime
described by Alec Ford and others could have done anything other than
amplify domestic volatility, the intrinsic variability of commodity prices
aside. In this regard, as we shall see, the growth experienced by Colombia,
Mexico, and Peru, and by Central America on silver through the mid-1890s,
may well suggest merit to the continuing depreciation of the exchange rate
that the international fall in silver prices brought after 1872.
Recent studies of Brazil and Chile have amplified, augmented, and
extended these findings. Brazil, for instance, was able to adopt the gold
standard when ample capital imports allowed it to do so, but remaining
there proved difficult. The domestic fluctuations in macroeconomic activ-
ity described by Ford occurred with an appreciating exchange rate (which
raised opposition from coffee exporters); rising imports; a deteriorating
balance of payments; and, inevitably, pressure to suspend convertibility
altogether. The gold standardbrought price rather than macroeconomic sta-
bility to Brazil, historians of its operation have concluded. Chile was able to
remain on the gold standard only for brief periods of time (1870–8, 1895–8),
buffeted as it was by large fluctuations in copper and nitrate exports.
Moreover, wages and prices in Chile proved highly inflexible, which
made the country an unlikely candidate for a fixed exchange rate, given
that a decline in external demand required deflation, which was diffi-
cult to achieve. Chile benefited from a floating exchange rate between
Cambridge Histories Online © Cambridge University Press, 2008
P1: GDZ
0521812909c07 CB929-Bulmer 0 521 81290 9 October 6, 2005 12:28
Export-Led Industrialization 261
1880 and 1894,soavoiding financial panics and balance of payment
crises.
In light of this discussion, it is important to notice that opposition to the
gold standard has not been unanimous and that there has been particularly
eloquent dissent from Argentine economic historians, whose acquaintance
with the gold standard, macroeconomic performance, and price stability
has found such recent echo under the “convertibility” system introduced
by Carlos Menem in 1991 to put an end to hyperinflation, and abandoned
under excruciating deflationary pressures in the beginning of 2002.One
of the principal criticisms of Ford, offered by Roberto Cort
´
es Conde, is
twofold. First, Cort
´
es Conde argues that Ford’s statistical evidence was
based to some degree on John Williams’s classic study, Argentine Interna-
tional Trade Under Inconvertible Paper Money,
5
the calculations of the trade
balance and capital flows in which Cort
´
es Conde judged seriously deficient,
and, therefore, likely to exaggerate the degree of adjustment (driven by
changes in the money supply) to unfavorable imbalances demanded under
afixedexchange rate. Yet, perhaps more important, Cort
´
es Conde argued
that the benefits of proceeding under the gold standard, and, particularly,
under the monetary discipline that strict convertibility imposed, more than
outweighed the costs of lost output. In a country with weak financial insti-
tutions or a political environment prone to inflationary finance, Cort
´
es
Conde suggests, it is the benefit of a flexible exchange rate (such as that
offered by depreciating silver) that is questionable, for which Cort
´
es Conde
offered the experience of Argentina with price inflation before 1870,orat
the time of the Baring Crisis (1890), as an object lesson.
Gerardo della Paolera and Alan Taylor have made a similar criticism, tren-
chant and sophisticated, in their study of the Currency Board in Argentina.
It is important to understand that part of the novelty of the argument made
by della Paolera and Taylor is that it relies on the monetary approach to
the balance of payments to make its case. The monetary approach to the
balance of payments regards all adjustment, both in the short and in the
long run, as a monetary phenomenon, because people are free to trade
financial assets across countries. So a change in the wealth of a country’s
citizens leads them to buy or sell financial assets rather than goods interna-
tionally, and money moves between countries immediately rather than as
aresult of changes in the balance of trade. By implication, then, the large
5
John H. Williams, Argentine International Trade Under Inconvertible Paper Money, 1880–1900 (Cam-
bridge, MA, 1920).