Exports were the lifeblood of Latin America until at least the 1930s. The world economy provided consumers, capital, labor, and machinery, which eventually spurred export-led growth. Consequently, a great deal of research has focused on the external sector. This understandable interest has led, however, to the relative neglect of the internal sector, where the majority of the population labored.

It has generally been assumed that export producers, such as planters, ranchers, and mine owners, together with their commercial and financial allies, dominated the economies of Latin America up to the Great Depression. They translated economic might into political power, employing the state apparatus to favor the export sector. Many scholars agree with Theotonio dos Santos that in Brazil the coffee planter class was “capable of overcoming all opposition to its dominance.”1

The resultant overreliance on exports retarded production by the farmers, ranchers, industrialists, and artisans who supplied the domestic market.2 This had serious consequences for Brazil, as well as for the rest of Latin America. The weak internal sector has been blamed for contributing substantially to an uneven distribution of income, regional imbalances, low productivity, underdeveloped capital markets, chronic inflation, and unhealthy dependence on foreign markets and capitalists.3

The size and nature of the internally oriented economy as well as the causes of its underdevelopment are important issues that demand careful study. The Brazilian state’s contribution to the growth of the domestic sector has received little empirical examination. This article will address that omission by examining Brazilian monetary and fiscal policies between 1850 and 1930.

This study also seeks to shed light on a related controversy: the extent to which the export oligarchy dominated the state. Since the 1970s a growing number of scholars have challenged the notion that export planters and intermediaries were “in complete control of the nation.”4 Roderick and Jean Barman, José Murilo de Carvalho, and Eul-Soo Pang and Ron Seckinger have argued that during the empire (1822-89) a semiautonomous “mandarin” caste, with relatively weak links to the economically dominant latifundists, ruled as well as governed.5 Developing a similar argument, David Fleischer, Joseph Love, and John Wirth have maintained that during the republic the export oligarchy had to share power.6 These scholars based their arguments to a considerable degree on collective biographies that demonstrate a lack of congruence between the export oligarchy and office holders. Since one need not originate within a given class in order to defend that class’s interests, such evidence is not completely convincing. A systematic study of who benefited from public policy should help to clarify the controversy.

There are, unfortunately, several methodological difficulties that obscure the extent to which the exporting group controlled the state. First, one must distinguish between intention and realization; the actual beneficiaries of measures were not always the intended beneficiaries. Second, as various ECLA economists and students of Brazil have pointed out, the expansion of exports often also encouraged production for the home market.7 Hence policies aimed at one sector often rewarded the other as well. Third, it is sometimes difficult to discriminate between producers for foreign and for home markets. Some products, like sugar and cotton, had a significant number of consumers both at home and abroad. Moreover, many capitalists who participated in the international economy also invested in consumption agriculture and industry, though the bulk of their investment was usually in the international sector.8

These difficulties prevent an exact delineation between the export and internal sectors. They do not, however, prevent a general appreciation of the state’s contribution to the growth of production for the domestic market. That information will enhance our understanding of who controlled the state.

Monetary Policy

Imperial monetary policy reflected the complex pressures on the state. Questions concerning the quantity and nature of money tended to divide not the producers for the home market from producers for foreign markets, but consumers of foreign products or capital from consumers of domestic products, capital, and labor. Consequently it is difficult to ascertain whether exporters or internally oriented growers and manufacturers most profited from monetary policy. Nevertheless, it is possible to determine the extent to which public measures coincided with the wishes of export planters.

Export and internal agriculturalists as well as manufacturers and domestic bankers generally favored easy money to facilitate credit. Although most of them accepted contemporary economic orthodoxy on the advisability of the gold standard, they were usually willing to risk devaluing the country’s currency, the mil-réis, if such a result were necessary to increase the money supply.9 After all, most planters recognized that a devaluation of the mil-réis in times of declining coffee prices dampened the adverse effects on fazendeiros; their mil-réis earnings did not fall as sharply as their pound-sterling income, and the plantation’s wage bill climbed at a rate slower than the devaluation of the currency since they had access to a large reserve of un- and underemployed laborers who were not fully integrated into the money economy.10 A slumping mil-réis protected farmers and industrialists who sold domestically from foreign competition as imports became more expensive for the Brazilian consumer. The inflation that usually accompanied a growth in the money supply also tended to aid industrialists since the price of their goods rose faster than their labor cost. Domestic bankers often received government loans or the privilege to issue currency when money became more abundant; this normally boosted profits despite a real depreciation of outstanding loans because of inflation. With relatively few foreign obligations, they were not much affected by a drop in the value of exchange.11

On the other side of the monetary policy debate were importers, urban consumers, foreign investors, and defenders of the federal and provincial treasuries. These groups were willing to restrict the money supply to maintain a strong exchange rate and stable prices.12

Importers feared that a devaluation would price their goods out of the market. Urban consumers, who depended to a large extent on imported foodstuffs and to a lesser degree on other consumer durables, also opposed a weak mil-réis.13 Foreign direct investors were concerned about the exchange rate because they had to convert profits made in mil-réis into pounds before repatriation; a falling mil-réis reduced their profit rate. European holders of the Brazilian debt also favored a valuable Brazilian currency because it facilitated debt repayment.

The financial interests of the central government and the provinces also recommended a conservative monetary policy that would defend the mil-réis and fight inflation through a restricted money supply. The central government increased the pace of its foreign borrowing in an attempt to limit the drain on the domestic capital market and to avoid raising taxes. The foreign debt climbed from about $30 million in 1850 to $150 million in 1889. By 1930 the foreign debt reached approximately $1.3 billion. A strong currency meant that the government had to spend less of the mil-réis that it took in through taxes to service the foreign debt. The interest rate on domestic state borrowing, which was the principal source of credit during the empire and still important under the republic, was set by the inflation rate; with a restrictive monetary policy, and hence fairly constant prices, the treasury could raise capital paying only 5 or 6 percent interest.14

The monetary policy of the empire, dominated by the Conservative party’s vision of tight money and high exchange rates (except during Liberal interludes of 1853-57, 1866-69, and 1888-89), favored the people tied to European goods and capital. Indeed, between 1870 and 1888 the real per capita money supply declined by about a quarter, despite the fact that Brazil was making the transition from slave to wage labor. When additional money was issued, it was usually used to service past debts and to pay government employees rather than to aid exporters. The expansion of the money supply during the Paraguayan War, for example, helped to cover the expense of supplies, which benefited the internal sector more than it did export producers.15

In fact, agriculturalists’ clamor for mortgage credit went largely unheeded during the empire. While the government authorized mortgage bonds beginning in the 1870s, such bonds never served as an effective means to secure credit, in good part because of the competition of treasury bonds. Parliament extended a few loans to banks to encourage mortgages. In response, the Banco do Brasil advanced a significant number of loans in the Paraíba Valley, but it ceased lending in 1884 because of the approach of the abolition of slavery.16 Only in 1888, when the treasury promised to lend banks 87,000 contos to ease the effects of abolition, did the measures adopted by the state come close to meeting the needs of the situation. Most of the credit of the 1888 measure, however, failed to reach export planters; a greater share went into financing the growth of the domestic market.17

Imperial statesmen also continuously attempted to return to the gold standard. Although they never succeeded, they did stabilize the mil-réis between twenty and twenty-eight pence in the empire’s last forty years. Even those who supported the efforts to expand the money supply in 1853 and 1888 insisted that the new banks of issue be willing to convert their notes into gold in order to maintain the value of the mil-réis close to par. Indeed, the Liberals, while expanding currency to compensate for the abolition of slavery in 1889, signed a contract with a private bank to replace the treasury’s inconvertible notes with gold-backed convertible bank notes.18

Monetary policy during the First Republic, between 1889 and 1930, was by and large more beneficial to export and domestic producers than the empire’s had been. The money supply grew almost twenty-four fold, the mil-réis fell from almost twenty-seven pence in 1889 to an average of under seven in the 1920s, and a large publicly owned bank offered increased mortgage credit.19 This would seem to demonstrate the triumph of coffee planters; but while their economic power is unquestionable, their complete command of monetary policy is more doubtful.

In general, government officials did not increase the money supply in an attempt to undermine the value of the mil-réis. Moreover, manipulation of currency was not intended to assist export producers. The administrations headed by Marshal Deodoro da Fonseca between November 1889 and November 1891 increased the money supply in an attempt to boost credit for urban industrialization and real estate speculation. They were dismayed when the mil-réis fell by more than 50 percent.20 The government of Marshal Floriano Piexoto (1891-94), often considered the administration most favorable to industrialization of the First Republic, sought to reduce the money supply and to raise the value of exchange, despite the fact that these actions were likely to hurt domestic producers. Floriano’s supporters argued that the flood of money, rather than increasing productive capacity, benefited just a handful of monarchist speculators and raised the cost of living, of labor, and of imported industrial machinery and raw materials. They did not succeed in stemming the growth of the money supply, however, because of the huge costs of fighting a civil war.21

The lack of direct planter control of government policy was even more evident between 1894 and 1906 when three Paulistas with close ties to coffee cultivation presided over the nation. Despite vociferous protests by São Paulo’s principal planter organization and the Paulista Republican party, the presidents cut back currency issues and loans to banks while raising the value of the mil-réis. The impact of an appreciating currency and dwindling credit at a time when coffee prices had plummeted drove many fazendeiros and manufacturers into bankruptcy. In 1906 the third of the Paulista presidents, Francisco de Paula Rodrigues Alves, almost destroyed the coffee valorization scheme designed by the planters of the three leading coffee-producing provinces when he refused to create the Caixa de Conversão. The institution was necessary, growers believed, to depress the value of the mil-réis and thus facilitate planters’ repayment of valorization loans. Brazil’s precarious finances, which had led to a declaration of a moratorium on foreign debt repayment in 1898, forced the Paulista presidents to break with their fellow fazendeiros.22

Planters did finally receive the Caixa de Conversão and easier money after Rodrigues Alves left office; but the three presidents responsible were more concerned with bolstering production for the domestic market than with aiding exports. Two of them came from provinces in which coffee production was stagnating; in response, these provinces were reorienting their production to domestic markets. The third president represented Rio Grande do Sul, which sold almost totally to Brazilian markets. At the same time that the government encouraged more ample credit by increasing the amount of money in circulation, it also returned to the gold standard and, hence, a less flexible exchange rate by issuing convertible currency through the Caixa. Thus it undercut the redistributive effect of a floating exchange rate that would have benefited export planters.23

A liquidity crisis, provoked by the exodus of foreign capital just before World War I, led to the abandonment of the Caixa de Conversão and convertible currency while causing the money supply to almost double. Some of the currency was issued to finance the second coffee valorization program and to aid banks, which employed sizable government loans to expand short-term credit to agriculture, industry, and, most of all, commerce. Still, the majority of the new money issued was created to cover the federal budget deficit, not to help Brazilian producers. Despite the growth of manufacturing during the war, it received virtually no direct government loans.24

During the 1920s, when coffee planters enjoyed unprecedented prosperity, they often encountered monetary policies that ran counter to their best interests. As governor of Minas Gerais, Arthur Bernardes had worked hard to stimulate the expansion of coffee exports. As president (1922-26), however, he was more concerned with the country’s shaky finances than with planter demands. He terminated the principal institution responsible for financing agriculture, reduced the money supply, and aborted his predecessor’s attempt to establish a federal mortgage bank and a federal coffee defense program.25

The last president of the republic, Washington Luís Pereira da Sousa, former governor of São Paulo, began his administration by helping the export sector. He issued a considerable amount of currency and lowered the exchange rate to assist São Paulo’s Permanent Defense of Coffee price support program. On the other hand, he also sought to reestablish the gold standard, though with a par level for the mil-réis of less than one-quarter of the earlier one. The contradiction between defending coffee and the currency at the same time became evident by the second quarter of 1929 when coffee prices dropped suddenly. Bather than spend the country’s gold to support coffee, he contracted the means of payment and reduced agricultural loans to protect convertible currency. The discontent that this aroused among export producers helped bring the revolution of 1930.26

It is apparent from this overview of monetary policy between 1850 and 1930 that export producers often did not secure the treatment that they desired. Even when public monetary measures did benefit from them, such an effect was usually not the principal consideration of state officials. It is noteworthy that growers for international markets seemed to have fared worst when politicians from the state of São Paulo headed the country.

The frequent inability of export planters to win battles over monetary policy does not demonstrate the political strength of domestically oriented producers. It is generally agreed that the government’s attempts to follow an orthodox monetary model favoring the gold standard retarded industrialization.27 Even during World War I, when traditionally laissez-faire regimes in Europe and North America were forced to intervene to step up domestic production, the Brazilian government did little. Growers for the home market were also not particularly favored.

The orthodox measures employed by public officials during the empire and the republic demonstrate, instead, the political strength of domestic urban consumers who held the threat of the riot and strike, the influence of foreign investors who held the purse strings, and a grave concern in Congress with maintaining the country’s foreign credit. These measures also demonstrate a limited conception of the state’s economic role. Monetary policy was guided by a concern to encourage the internal and international circulation of goods and capital, and to enforce the market forces, but did not conceive of modifying them by expanding productive capacity. Nonetheless, the protection afforded domestic producers by a depreciating mil-réis, and industrial capital accumulation spurred by inflation-driven forced savings, did help domestic production grow faster than exports.

The Taxation System

Brazil’s fiscal system placed a disproportionate tax burden on the international sector, though this lessened over time. The cost of government during the 1850-1930 period was borne primarily by international commerce. Import and export taxes supplied the central government of the empire with almost 80 percent of its revenues in 1850 and still around two-thirds of its income after 1875. The greatest share of income was derived from import duties, which were responsible for two-thirds of all taxes in 1850 and for more than half in 1888. When the republican Constitution of 1891 awarded the provinces complete control of the export tax, the share of the central government’s revenues contributed by foreign trade declined; after World War I it fell to less than 40 percent.28

The central government’s declining reliance on taxes on international commerce was mirrored by the provincial governments; the provinces had always depended unevenly on foreign trade as a source of revenue because of their unequal relationship to the world economy. During the 1908–12 period, for example, four provinces were responsible for 81 percent of Brazil’s exports. Their percentage was approximately the same at the end of the republic. Much of the rest of the country sent very little abroad. Even provinces that did participate actively in the world market, such as São Paulo and Minas Gerais, eventually reduced their imposts on goods shipped abroad. Initially they raised about three-quarters of their income from taxes on foreign exports; by 1930 they declined to one-third. With both the central government and the provinces turning increasingly to domestically oriented production for revenue, foreign commerce’s share of total public revenues fell from 80 percent in 1850 to 68 percent in 1888 to 33 percent in 1930.29

Curiously, the shifting tax burden reflected the opposite of what one might expect. It was during the empire and early republic, when coffee planters contributed as much as a third of GDP and faced little challenge from producers for the underdeveloped internal market, that they contributed the most to the treasury. Industry’s share of taxes grew with its growing economic and political importance. The reason for this is that the distribution of taxes reflected not only relative political power but also the practicalities of tax collection. Exports were taxed disproportionately because it was easy to evaluate and tax their value at the ports. It would have been much more expensive and cumbersome to send tax collectors into the interior to appraise and charge products destined for diffused markets and even subsistence. Although planters grumbled about export taxes, they preferred them to land taxes, which could have made their vast plantations untenable. Few states had a rural land tax, and where one was in effect, it accounted for less than 10 percent of the state’s income. Furthermore, planters were not subject to an income tax. Also, many planters believed that given Brazil’s monopoly position in the world coffee market, the export tax simply raised the world price of the commodity; hence North American consumers ultimately paid it, not Brazilian producers. Planters could also bear import taxes because they fell disproportionately on the urban consumers of imported goods.30

The gradual shift away from taxing international commerce represented a partial victory for planters. As urban markets grew, and with them industry and domestic agriculture, the taxation of home producers was facilitated. In São Paulo, the percentage of revenue generated by the sale of coffee fell from about 75 percent in 1909 to half that in the early 1920s. Since coffee exports, however, contributed less than one-quarter of São Paulo’s GDP in the late 1920s, the export sector was still proportionately overtaxed.31 By the same measure, international trade was also overtaxed nationally. By 1926-30, import duties descended to 40 percent of federal revenues but exports provided only 15 percent of GDP.32 In this case, however, most of those paying import duties were not the same ones profiting from exports.

The rise of federal taxes on domestic production also represented a victory for internal producers. The tax that became the most important source of internal revenue, that on consumption, first became significant in 1897 when it was levied to compensate for the loss of import duties because of tariff protection. The growth of the consumption tax in subsequent years, sometimes accounting for more than one-quarter of total federal income, reflected the increasingly protective level of some customs duties. By 1914, Brazil’s duties on consumer durables had created, in the estimation of a United States congressional investigator, the highest tariff in the Western Hemisphere. Although Congress raised some duties in order to provide additional revenue and others to reduce imports and safeguard the balance of payments, many increases reflected political victories for domestic manufacturers and agriculturalists. There was a growing consensus in Brazil on the need for some tariff protection. Domestic producers provided jobs, consumed Brazilian raw materials, and helped the balance of trade. Moreover, many planters were themselves industrialists. Besides, fazendeiros calling for a public coffee price support program could not deny the legitimacy of protection for other sectors.33 To compensate the financially strapped federal treasury for revenue lost through protective duties, internal producers were willing to pay greater taxes—as long as they did not negate the buffer of the import duties.

On the provincial level, the decline of the export tax often did not indicate a shift in the tax burden from international production to domestically oriented goods. In the provinces that sold few of their products abroad, export duties were largely levied on goods shipped to other parts of Brazil. For these provinces the decline of export duties and the decline of the common, though unconstitutional, provincial import duties, encouraged a widening of the national market.34


The imperial and republican states spent a substantial part of their budgets on economic activities; they rewarded the internal sector to a rather surprising degree. Central government expenditures under the empire grew 500 percent between the 1840s and the 1880s, consuming perhaps as much as one-third of GDP. The funds went increasingly into economic areas as administration and military expenses fell.35

Real central government spending continued to grow under the republic despite the advent of federalism; total state spending grew even faster. Total real per capita state spending expanded more than 150 percent between 1885-88 and 1926-30.36 After a decline in economic investment in the 1890s because of falling revenues, political chaos, and the need to buy the allegiance of public employees, public capital investments resumed their importance in the early twentieth century. After the economy recovered and foreign capital became easily available, fixed capital investments averaged over one-fifth of federal spending until 1922. Although Brazil’s per capita state expenditures were fifth in Latin America (behind Uruguay’s, Argentina’s, Chile’s, and Cuba’s) in absolute terms they were by far the greatest in Latin America (surpassing, by the way, those of twelve European countries). Total Brazilian state capital investments were also the greatest in Latin America during those years.37 The republic’s last years witnessed a dwindling of federal economic outlays because of a financial crisis; this was partially offset by the increased capital investments of the wealthiest provinces.38

Federal spending, which substantially outweighed the spending of all of the provinces combined, was relatively more generous in provinces that produced little for export abroad. By far the major recipient of federal largesse was the Federal District, which did not produce exports. It frequently absorbed two-thirds of the union’s expenditures, often twice as much as it contributed to revenues. Thus the central government concentrated wealth from the rest of the country in the city of Bio de Janeiro. Although it is true that many wealthy fazendeiros lived in Bio, most of the funds helped stimulate the internal economy. Rio was the country’s largest manufacturer and banking center until World War I; it continued to be by far the largest market for domestic production and the financial center throughout the republic.39 Although much spending in Rio served to fuel imports, demand was increasingly met by domestic producers since tariff policy, the weak mil-réis, and the expansion of the railroad allowed them to outcompete importers. Some provinces also received more back from the federal treasury than they contributed. Rio Grande do Sul and Mato Grosso, which were minor exporters, were so rewarded in large part because of military and transportation expenses in these border areas.40

On the other hand, the country’s principal exporter, São Paulo, found itself subsidizing other parts of the country. Paulistas opposed the empire in part because they paid into the monarchy’s treasury three and one-half times as much as they received. Under the republic the outflow from São Paulo worsened to nine to one in some years.41 The only other area of the country to suffer a net drain to the federal treasury of similar magnitude was the other main exporting area: the North. Once its rubber boom subsided before World War I, however, its share of federal funds became more commensurate with its contribution.42

These figures do not, however, demonstrate simply that the federal government spent preferentially in the nonexporting areas of the country. São Paulo, in addition to providing as much as 80 percent of the country’s exports, was also an industrial center by the 1890s and an important producer of internally oriented agriculture. The federal treasury siphoned off resources from São Paulo’s internal sector as well as its external one, though the large market in the Federal District was much more valuable to São Paulo’s manufacturers and farmers than to its export planters. The distribution of federal funds was dictated more by reasons of state than by a plan to develop the internal economy. The capital received the bulk of funds simply because it was where the government was centered, while some provinces received money because of national defense concerns. The richest provinces, which were the exporters, were the most capable of contributing and so they did. In return they demanded and received a considerable amount of local autonomy. Still, one assumes that if coffee planters held complete hegemony over the national government, they would have secured an arrangement that was less costly to themselves.

Indeed, little of state spending was directly invested in the export sector. Agricultural experimental stations, seed, and schools received only 1 percent of the direct economic investment during the empire. No money was spent on fertilizers or irrigation. The major benefit that agriculture gained was government subsidies of five hundred contos to modern sugar mills, one-fifth the sum spent on the telegraph. The imperial government did offer agriculture long-term loans through the Banco do Brasil and through a subsidized loan program in 1888; but since the treasury was by far the principal borrower in the Brazilian market, its net impact was probably to siphon resources away from agriculture.43

Little spending during the empire was directed toward helping internally oriented producers. Although there were a few cases of government loans to industry, they were small and sporadic.44

The republican federal government also spent little to aid directly the international sector. Despite calls from export producers for more assistance, federal officials, aside from providing the transportation infrastructure, the necessary legal structure, and attracting foreign investment, gave little directly to exports. The federal government never provided a mortgage bank, though there were several failed attempts to do so. The defense of coffee programs, the most obvious examples of aid to exports, received little federal assistance; all federal funds provided to coffee producers were repaid with interest. Nor did the federal government come to the rescue of rubber exporters when prices collapsed in 1912. Sugar similarly gained little; the national government’s subsidization of modern central sugar mills was transferred to the provinces in the 1890s.45 Even federal spending on European immigration was phased out in the early twentieth century. Besides, immigration had been important for the internal economy as well as the international one since Europeans provided a disproportionate share of the work force of industry and retail commerce.46 The low level of direct support for export producers was symbolized by the Ministry of Agriculture, Industry, and Commerce. It did not even exist between 1892 and 1909. After it was resurrected in the latter year, Congress awarded it an average of only about 3 percent of the budget for the rest of the republic.47

At the provincial level, exporters were better able to assure ample return. In São Paulo they were able to secure public investment in railroads, immigration subsidies, and a mortgage bank that loaned primarily to coffee planters. Even here, a disproportionate share of investment flowed naturally to the cities since the impact of social services and public utilities could be maximized in urban areas, and a substantial organized group of people existed to demand them. As fazendeiros increasingly moved to the cities themselves, they supported the europeanization of São Paulo and smaller towns. These expenditures ultimately had a much greater multiplier effect on the internal market than on exports.48

Most of the benefits that accrued to domestically oriented producers from public spending were similarly indirect and sometimes unintended. Little direct federal spending consciously intended to build up the internal sector. There were no internal price supports and scant attention was paid to irrigation, fertilizers, and machinery. The major federal effort was a substantial, though woefully inadequate, drought relief program in the Northeast, beginning in 1909 and lasting until 1924, which led to the construction of many dams and reservoirs.49

Industry received more direct help, but total aid remained meager. There were some loans to particular industries during the empire but they were infrequent. The major effort during the First Republic was an 1892 bond issue. In actuality 80,000 bonds were issued and a significant portion of these funds were loaned to manufacturers. The loans were intended, however, less to finance increased capacity than to rescue overextended companies. There were some smaller short-term loans granted in 1915, 1918, and 1919 but again the amounts were generally little. During World War I only about one thousand contos were loaned directly to factories; and rather than stimulating the economy through countercyclical spending to compensate for slack imports, the Bras administration actually reduced real federal spending by 24 percent during the war. After the war some strategically important industries received federal loans because of increased concern with the national defense; four iron companies received 13,300 contos between 1918 and 192.3; coal mines benefited from the equivalent of $1.5 million between 1918 and 1930; and the federal Geological Service purchased an oil rig for Brazilian companies to use in oil exploration.50 While these investments reflected growing public concern with industry, they were inadequate to bring about any substantial change. Producers for the home market continued to function without a development bank or any sort of serious state planning and coordination.

Occasionally Congress gave Brazilian producers preference in federal purchases. To qualify for this aid, however, the Brazilian product had to be at least equal to the import in price and quality, and this generally nullified the advantage.51

Publicly owned industries bolstered the manufacturing sector, but they failed to have many linkages with other industries. The central government established an iron smelter in Ipanema, São Paulo, in 1818, which continued to produce iron in small quantities until it closed in 1895. The empire also established a munitions and cartridge factory that together employed about six hundred people. During World War I a federally owned shipping line, the Lloyd Brasileiro, purchased a large share in a coal mine. In addition, the Central Bailroad, Sorocabana Bailroad, and Lloyd Brasileiro maritime company employed close to seven thousand people in their workshops, which were responsible for manufacturing parts as well as for making repairs on the equipment. All of these companies worked on a relatively rudimentary level and were dependent in great degree on imported raw materials, parts, and machinery.52

Most state economic spending was in railroads, which constituted indirect subsidies to producers. Even in infrastructure investment, domestic producers benefited arguably more than exporters. The state was extremely active in railroad building during the empire and much of the republic. Initially the railroads aided the export sector, but over time they helped to expand the internal market. The empire was influential in railroad construction beginning with the first line built in 1854. The scarcity and timidity of willing private investors forced the state to enter the economy on an unprecedented scale. By 1889, 38 percent of Brazil’s rail system, then the largest in Latin America, belonged to the central government. Another third enjoyed federal profit guarantees, and 15 percent more, provincial guarantees. In the 1888 budget, railroad expenses consumed one-third more than the country’s army.53

The principal public railroad in the country, the Dom Pedro Segundo (renamed the Central do Brasil during the republic) was also its largest coffee carrier during most of the empire. Its original intention was to connect Rio’s coffee hinterland with its port. Thus the central government stimulated the expansion of the export economy. Even during the empire, however, there was a counter-tendency to invest in nonexporting areas. True, the exporting Center gained more than half of all public railroad capital by 1889; but another major exporting area, the North, received none. Meanwhile, the South, where exports were minimal, held close to one-quarter of all public railroad capital. Similarly, the Northeast, which by this time contributed only about one-ninth of Brazil’s exports, also held one-quarter of public railroad capital. Railroad guarantees from the central government intensified the movement of funds away from exports. The Center was awarded only about one-third of all central government guarantees in capital, while the Northeast, which contributed only 16 percent of federal revenues, held 42 percent of the concessions.54

Provincial railroad policy differed markedly from that of the central government. Since only exporting provinces had sufficient revenues to offer profit guarantees, only the Center provinces offered concessions. The railroads built with aid from these funds served exports to a greater extent than the central government’s lines.55

Federal, and, to a lesser extent, provincial, railroad policy during the republic was even more partial to internal producers than had been imperial policy. Increasingly, public railroads catered to the domestic market and fulfilled strategic objectives rather than serving primarily to carry exports.

Much of the Union’s railroad spending was devoted to public enterprises. By 1930 it had expanded its share of the system to 59 percent, owning railroad capital of almost $500 million. The key federal line continued to be the Central, which had expanded impressively. The decline of coffee in the Paraíba Valley, however, and the growth of the Federal District as a manufacturing and consumption center meant that the Central was no longer principally an export line. Although the federal government took over an important coffee railroad in 1914, the Noroeste, little of the country’s coffee traveled on federal trains. In 1920, when federal cars carried 37 percent of all rail freight, they only moved 5 percent of the coffee.56

The increased orientation of federal railroads to the internal economy was not only a result of the swelling internal market and geographic shift of export crops; strategic considerations also played an important role. Although Brazil did not develop an integrated rail system before 1930, there were efforts to push rail into remote areas to promote migration and make them militarily defensible. As a result there was a disproportionate number of federal lines built to the West and South. Congress also planned to extend the Central from Rio up to the Amazon, but the line never reached farther north than Pirapora on the São Francisco River.57

The regional distribution of federal railroad spending displays the preference given to domestic producers. As a gauge, I have taken the percentage of federal rail capital in each region and divided it by the percentage of each region’s contribution to the federal budget. The Center’s share of national railroad capital was only 60 percent of its revenue contribution, while the South’s railroad capital was 1.6 times its contribution, the Northeast 2.8 times, and the North 4.8 times. The federal government was giving preferential treatment to the areas with the lowest exports (by 1930 the North was no longer an important exporter). This trend was particularly noticeable in São Paulo, the largest revenue contributor in the Union. During the republic, the state ranked only fifth in the amount of federal railroad funds it received. In per capita terms, it ranked sixteenth out of twenty provinces.58

The railroad spending of the provinces was more oriented to exports than federal spending because São Paulo was the principal province engaged in railroad construction and financing. While in 1920 they carried only 14 percent of all freight, provincial railroads transported fully 44 percent of all coffee. Even in São Paulo, however, railroads increasingly carried domestically destined goods.39

The overall trend was for public trains to serve internal producers. In 1910, 20 percent of all cargo was intended for export; by 1920 that had fallen by almost a half. In comparison, it has been calculated that over 50 percent of all traffic on Mexico’s lines in 1910 was exports. By the end of the republic, at least two-thirds of all Brazilian rail cargo was neither exports nor imports.60

As public railroads turned ever more to serving the internal market, they accentuated the benefits that domestic producers received from the state. Public railroads, particularly the Central, charged lower freight rates than private ones. While exporters enjoyed this advantage as much as internal producers, as public export traffic declined so did export’s rewards. Consumption goods in the cities deemed “of primary necessity,” such as basic foodstuffs, accrued special privileges since federal authorities sought to contain the cost of living in the cities to prevent riots.61

The state also supported the expansion of shipping, which gradually fostered internal commerce. Until 1862, because of the importance of the merchant marine to the creation of internal markets and the national defense, all ships engaged in the coastal trade had to be owned and captained by Brazilians, and at least two-thirds of the crew members also had to be Brazilians. This policy was unsuccessful. The national fleet was outmoded and inefficient; in the middle of the nineteenth century only 7 percent of shipping commerce, measured by value, took place between Brazilian ports. In 1867 all Brazilian ports, including the Amazon, were thrown open to foreign vessels while the central government subsidized the national fleet to encourage coastal trade. By the 1880s the treasury was spending almost three thousand contos a year on subsidies. By 1886–87 interprovincial commerce had ballooned twentyfold from midcentury and had grown to 17 percent of all shipping.62

The republican regime continued to spur the growth of interprovincial commerce. The more than $40 million that the federal government awarded to shipping companies between 1908 and 1930 went, with one exception, to Brazilian firms. Since Congress reinstated the prohibition of foreign participation in the coastal trade in 1897, the great majority of the shipping conducted by subsidized companies was in interprovincial commerce. An important part of this trade in the nineteenth century consisted of re-exports of European goods; but by the twentieth century 90 percent of the coastal cargo was nationally produced. The subsidies and monopoly allowed interprovincial trade to grow threefold between 1908 and 1930 and to come to average about one-third of all shipping; they also helped the Brazilian commercial fleet to become the largest in Latin America, though it was still only one-fiftieth the size of the British merchant marine.63

The state, of course, also aided international shipping. The first regular mail line between Brazil and England had received an imperial subsidy. Under the republic, the federal government nationalized the largest maritime company in South America, the Lloyd Brasileiro, which became the single largest carrier of coffee for many years. But the great majority of its business was conducted in the internal trade. Brazilian shipping, even at its height during World War I, was responsible for less than one-quarter of Brazil’s international commerce.64

The major public assistance to international shipping came in the form of port construction projects. But many of the largest international ports, such as the ones at Santos, Manaus, and Pará, were privately owned and financed. The port of Rio, which was modernized at considerable expense to the federal government and then leased to a French company, was not only the largest importer from abroad, but also the principal entrepôt for domestic production. The ports at Rio Grande do Sul and Forteleza, two of the republic’s costliest building projects, were primarily engaged in interprovincial trade.65


The domestic sector of the Brazilian economy grew considerably in the 1850–1930 period. According to a recent estimate, the Brazilian economy as a whole grew faster after 1900 than did the economies of Western Europe and the United States,66 while the internal sector expanded faster than exports. Even though exports may have contributed more than one-third of GDP in 1860, by 1930 they had fallen to around 15 percent.67 And while exports dominated most provinces in the nineteenth century, by 1912 most provinces depended principally on the home market; over 80 percent of the country’s exports originated in four provinces. As a result, probably less than one-sixth of Brazil’s population worked in areas directly related to exports. Indeed, the number of industrial workers had come to rival the number of export workers.68 The country became virtually self-sufficient in such important products as textiles, clothing, and shoes; many food items, such as rice, dairy products, lard, and beans; and raw materials, like cotton.

A good part of this growth, of course, resulted from factors largely independent of state action. Exports usually give birth to a host of activities in the internal economy because of multiplier effects and fiscal consumption, and forward and backward linkages.

Still, in numerous ways the state favored the expansion of the domestic sector. As has been shown, inconvertible currency and a depreciated exchange protected domestic producers. Taxes placed a disproportionate burden on international trade, while customs duties defended national goods. The patterns of expenditures redistributed funds from exporting regions to nonexporting ones, and from the countryside to the cities, while railroads and ships stimulated the formation of a national market.

The fact that the state’s economic policies so favored the domestic sector does not necessarily reflect substantial power on the part of industrialists and farmers. After all, there was no conscious plan to industrialize or to promote consumption agriculture, no development bank, and few government loans. The forces of the market, not state intervention, were to bring prosperity. Much of the state intervention that did occur was initially intended to encourage exports. The benefits that accrued to domestic producers were often side-effects of measures aimed at resolving public financial crises brought about by recessions in the world economy; expansions of the money supply and increases of some customs duties often derived from such an impetus. Some of the internal economy’s gains seem to have reflected strength in weakness; best exemplified by the federal government’s subsidies for railroads in areas where the lack of profitability discouraged private investors. Similarly, public assistance to international shipping was not necessary; the market would see to growth in this sector. But an investment was necessary to foster interprovincial commerce.

It does, however, seem odd that a hegemonic export oligarchy would voluntarily redistribute resources to the poorer, weaker sectors of the country. One could argue that export producers did not reap greater benefits from the state because they did not particularly wish it. Shane Hunt, in an analysis of Peruvian spending and taxation policy at approximately the same time, also found that the export oligarchy did not win the preponderant share of returns from the state. He suggested that this may have occurred because “the oligarchy’s favored economic position is created and maintained by its control over natural resources, particularly land, and by its monopoly of contact with the world economy. The oligarchy is the prime beneficiary of capitalism’s invasion into traditional society and for this reason has remained satisfied with a state that is merely permissive rather than directly augmenting of oligarchic income. ”70 While the dynamic in Brazil was different because it did not have a sizable indigenous society, the thrust was quite possibly the same. Capital accumulation still occurred in the private sphere. Monopoly capital had not yet arisen to demand government regulation and coordination. Little public assistance was necessary beyond the maintenance of relatively unfettered markets for land, labor, capital, and the transportation infrastructure. The state was primarily a tool for maintaining a political pact, not redistributing wealth to the oligarchy.

Still, even while keeping down state economic interventions, one assumes that planters would seek to ensure that they received the majority of the benefits. Nathaniel Leff has argued that planter hegemony should have brought about a large state economic role. Only the lack of sufficient resources prevented that.71 Yet, as has been shown, the problem for the oligarchy was not only the total amount of resources available to the state, but also their distribution. It is clear that although export producers were indeed opposed to an interventionist state, they frequently sought measures that were refused.

Part of the planters’ failures may have stemmed from the fact that most government officials were liberal professionals, not exporters. José Murilo de Carvalho has shown the large presence of government bureaucrats and magistrates in positions of power during the empire. Tobias Monteiro found in 1908 that only 9 of the federal government’s 63 senators and 8 of the 212 deputies were agriculturalists. J. W. F. Rowe has argued that even in São Paulo the large-scale planters were not the politicians: “On the contrary, politics throughout Brazil is a separate and well-defined sphere of activity. The political leaders are far from being puppets of what may be termed the ‘political planters.’’’72 Yet even when planters such as Campos Sales and Rodrigues Alves were in the presidency, planters were defeated on key issues. People in power believed that they had to represent the long-run interests of the nation, not just the short-run interests of the export oligarchy.

Politicians were able to impose their view of the state’s proper role in good part because of divisions within the international community. Export producers and European holders of the Brazilian debt were often on the opposite side of the debates. The foreigners wanted monetary and fiscal conservatism, planters often did not. European leaders frequently carried the day. Brazil had by far the largest foreign debt in Latin America, and the economy was vulnerable to any crisis in the world economy. It had a large money economy rather than a sizable subsistence peasant sector and was dependent on exports. Exports in Brazil constituted a larger share of GDP than in most other more developed Latin American countries and they were concentrated in only two products, themselves highly susceptible to cyclical crises. Consequently, the Brazilian state faced frequent financial problems that required foreign funding to resolve.

Export producers were also divided among themselves. Planters in less prosperous areas often found the schemes of the wealthier exporters uninteresting. The Treaty of Taubaté failed to be enacted for just such a reason. When rubber prices tumbled, coffee growers remained deaf to rubber exporters’ pleas for assistance. In addition, many export producers also had interests in industry, public utilities, or domestically oriented agriculture. Hence they had divided loyalties.

This is not to say that producers for the Brazilian market were politically helpless, merely the fortuitous beneficiaries of divisions in the dominant fraction of the ruling class. Several important social forces operated in their behalf. Urban consumers could threaten to riot or strike; farmers could promise cheaper food to secure urban peace and lower wages. In addition, domestic producers could step up import substitution to help the balance of payments. Industrialists helped make the country self-sufficient, a growing concern to the military and to other sectors of the government. They could also make Brazil “modern,” more like Europe and the United States, which was the ambition of many members of the ruling class.

This overview of the distribution of public economic policy benefits demonstrates that the Brazilian state between 1850 and 1930 was not simply the tool of the export oligarchy. Politicians and bureaucrats faced complex and conflicting pressures. The international sector was fragmented and the domestic sector growing in importance. While producers for internal markets did not yet dominate the state and did not have gold, they had the numbers and, increasingly, they had the future.


Theotonio dos Santos, “Brazil” in Ronald H. Chilcote and Joel C. Edelstein, eds., Latin America: The Struggle with Dependency and Beyond (New York, 1974) p. 425. For similar views, see, for example: E. Bradford Burns, A History of Brazil, 2d ed. (New York, 1980), p. 375; André Gunder Frank, Lumpenbourgeoisie, Lumpendevelopment: Dependence, Class, and Politics in Latin America (New York, 1972), pp. 59-60; Wolfgang Hein and Konrad Stenzel, “The Capitalist State and Underdevelopment in Latin America—The Case of Venezuela,” Kapitalistate (San José, Calif.), 1:2 (1973), 33; Caio Prado Júnior, História Econômica do Brasil, 15th ed. (São Paulo, 1972), p. 167.


For example, see: Mircea Buescu, Evolução Econômica do Brasil, 4th ed. (Rio de Janeiro, 1979), p. 137; Nelson Werneck Sodré, História da Burguesía Brasileira, 3d ed. (Rio de Janeiro, 1976), pp. 107, 137, 178.


Nathaniel H. Leff, in Underdevelopment and Development in Brazil, vol. I: Economic Structure and Change, 1822-1947 (London, 1982), pp. 115, 116, 121, 135, points out many of the ramifications of a weak internal sector. The Economic Commission for Latin America has stressed the retarding effect of export dependence on industrialization; see its Development Problems in Latin America: An Analysis (Austin, 1970).


Burns, History of Brazil, p. 375. There had been earlier studies that denied planter hegemony. Joaquim Nabuco, Abolitionism: The Brazilian Anti-Slavery Struggle, trans, and ed. by Robert Conrad (Urbana, Ill., 1977), p. 128, argued this during the empire. Tobias do Rego Monteiro, Funccionarios e Doutores (Rio de Janeiro, 1919), p. 37, expressed a similar view about the republic. Fernando Henrique Cardoso and Enzo Faletto, Dependência e Desenvolvimento na America Latina: Ensato de Interpretação Sociológica, 4th ed. (Rio de Janeiro, 1977), pp. 63–76; Raimundo Faoro, Os Donos do Poder: Formação do Patronato Político Brasileiro, 2d ed., 2 vols. (São Paulo, 1975), passim; Florestan Fernandes, A Revolução Burguesa no Brasil: Ensato de Interpretação Sociológica (Rio de Janeiro, 1974), p. 208; Celso Furtado, Formação Econômica do Brasil (Rio de Janeiro, 1959), p. 204, and Sodré, Burguesía, p. 180, argued in the 1960s that planters had to share power, especially during the republic; but these authors did not empirically demonstrate their case.


Roderick Barman and Jean Barman, “The Role of the Law Graduate in the Political Elite of Imperial Brazil,” Journal of Inter-American Studies and World Affairs, 18:4 (Nov. 1976), 423–449; José Murilo de Carvalho, A Construção da Ordem, A Elite Política Imperial (Rio de Janeiro, 1980), pp. 73–91; Eul-Soo Pang and Ron L. Seckinger, “The Mandarins of Imperial Brazil,” Comparative Studies in Society and History (London), 14:2 (Mar. 1972), 215-244.


David Fleischer, “Political Recruitment in the State of Minas Gerais, Brazil, 1890–1970” (Ph.D. Diss., University of Florida, Gainesville, 1972), p. 75; Joseph L. Love, “Um Segmento da Elite Política Brasileira em Perspectiva Comparativa,” in Centro de Pesquisa e Documentação de História Contemporánea do Brasil, ed., A Revolução de 30, Seminário Internacional (Brasilia, 1983), pp. 47-96; John D. Wirth, Minas Gerais in the Brazilian Federation, 1889-2937 (Stanford, 1977), p. 145.


See, for example: Wilson Cano, Raízes da Concentração Industrial em São Paulo (São Paulo, 1977); João Manuel Cardoso de Mello, O Capitalismo Tardío: Contrihuição á Revisão Crítica da Formação e do Desenvolvimento da Econoìnia Brasileira (São Paulo, 1982); Warren Dean, The Industrialization of São Paulo, 1880-1945 (Austin, 1969); Sérgio Silva, Expansão Cafeeira e Origens da Indústria no Brasil (São Paulo, 1976); Oswaldo Sunkel and Pedro Paz, El subdesarrollo latinoamericano y la teoría del desarrollo (Mexico City, 1970).


For a discussion of the overlap of sectors, particularly agriculture and industry, see: Dean, Industrialization of São Paulo; Darrell E. Levi, A Família Prado, trans, by José Eduardo Mendonça (São Paulo, 1977); Zélia M. Cardoso de Mello, “Contribuições ao Estudo da Natureza do Empresariado Paulista,” in X Encontro Nacional de Economia of the Associação Nacional dos Centros de Pos-Graduação em Economia, Aguas de São Pedro, Rio Grande do Sul, 1982 (Porto Alegre, 1982), 1055–1086; Flávio R. Versiani, “Industrial Development in an ‘Export’ Economy, the Brazilian Experience before 1914,” Journal of Development Economies (Amsterdam), 7:3 (Sept. 1980), 307-329.


The most affluent planters often favored a stable exchange rate and domestic prices because since they lent to other planters and held government bonds, inflation eroded their assets. See: Congresso Agricola. Colleção de Documentos (Rio de Janeiro, 1878), p. 134; Cardoso de Mello, “Estudo do Empresariado,” 1060, 1064, 1068; Warren Dean, Rio Claro, um Sistema Brasileiro de Grande Lavoura, 1820-2920, trans, by Waldivia Marchiori Portinho (Rio de Janeiro, 1977), pp. 41, 44, 59. Some manufacturers also feared devaluations because, as J. P. Wileman, Brazilian Exchange: The Study of an Inconvertible Currency (1896, rpt. New York, 1969), pp. 194, 260, and the Gazeta de Comércio e Finanças, Rio de Janeiro, Mar. 14, 1896, p. 3, observed, the cost of imported machinery and imported raw materials often constituted half the final cost of Brazilian produced goods. Over time the opposition of these groups to easy money diminished as planters diversified from government bonds into industry and banking and industrialists increasingly used domestically produced raw materials.


Furtado, Formação Econômica, p. 196, was the first to recognize this. For the views of various economic groups on monetary and exchange policy, see, for example: Congresso Agricola, 2878; Ata da Seção do Conselho do Estado, Rio, March 12, 1885, in the Instituto Histórico e Geográfico Brasileiro, (IHGB), Lata 545, Pasta 55; Brazil, Ministério da Fazenda, Relatório 1892, p. 24, and 1906, pp. vi, xiii; Associação Comercial de São Paulo, Relatório 2895, pp. 46, 65; Brasil, Congresso, Documentos Parlamentares, A Caixa de Conversio (Paris, 1914); Jornal do Commércio, Retrospecto Comercial 2920, (Rio de Janeiro, 1911), pp. 23, 29; United States Department of Commerce, Arthur Redfield, Brazil: A Study of Economie Conditions Since 2923 (Washington, D.C., 2920), p. 73; Eudes Barros, A Associação Comercial no Império e na República (Rio de Janeiro, 1975), p. 159.


Brasil, Presidente, Mensagem Apresentada ao Congresso Nacional, 1916 (Rio de Janeiro, 1916), p. 243; Departamento Nacional de Estatística, Movimento Bancário, 2927-28 (Rio de Janeiro, 1930), p. 40.


Contemporary economics preached that there was a close relationship between the money supply and the rates of exchange and inflation when currency was not gold-backed. Even though Gustavo Henrique Barroso Franco, in Reforma Monetária e Instabilidade Durante a Transição Republicana (Rio de Janeiro, 1983), pp. 33-43, demonstrates that the relationship was not so strict because the balance of payments heavily influenced exchange and prices, most contemporary Brazilians viewed the money supply controversy through the prism of nineteenth-century economic orthodoxy.


Urban unrest with the cost of living ignited such conflicts as the Praiera Rebellion in Recife (1848-50), the Cabanagem in Belem (1835-39), the Vintim riot in Rio (1880), the Jacobin movement in numerous cities in the 1890s, and enormous strikes, particularly in Rio and São Paulo during and immediately after World War I.


Análise e Perspectiva Econômica (APEC), A Economia Brasileira, 1974 (Rio de Janeiro, 1974), table 1-20.


Money supply figure from Diretoria Geral de Estatística (DGE), Armário Estatístico, 1939/1940 (Rio de Janeiro, 1941), pp. 1417, 1418, and deflated using the index of Claudio Contador and Claudio Haddad, “Produção Real, Moeda, Preços; A Experiência Brasileira no Período 1861-1970,” Revista Brasileira de Estatística (Rio de Janeiro), 36 (1975), 430-434; for a fine monetary and fiscal history, see: Liberato de Castro Carreira, História Financeira e Orçamentária do Império no Brasil, 2 vols. (1889, rpt. Rio de Janeiro, 1980), passim.


Banco do Brasil, Relatório, 1889 (Rio de Janeiro, 1889), p. 121; L.R.O., Quatro Mezes de Administração Financeira (Lisbon, 1890), p. 18; C. F. Van Delden Laërne, Brazil and Java: Report on Coffee-Culture in America, Asia, and Africa to H. E. the Minister of the Colonies (London, 1885), p. 212; Joseph E. Sweigart, “Financing and Marketing Brazilian Export Agriculture: The Coffee Factors of Rio de Janeiro, 1850-1888” (Ph.D. Diss., University of Texas, Austin, 1980), pp. 109-169.


Banco do Brasil, Relatório, 1888, p. 6, 2890, pp. 14, 20, 1891, p. 9; Banco Hypotecário do Brasil to Floriano Peixoto, Oct. 18, 1893, Arquivo Nacional, Arquivo do Presidente Floriano Peixoto, Caixa 8L, Pacote 3; Castro Carreira, História Financeira, II, 776; Amaro Cavalcanti, Resenha Financeira do Ex-lmpério do Brasil em 1889 (Rio de Janeiro, 1890), p. 50; O Dia, Rio de Janeiro, Aug. 30, 1889, Arquivo Casa Rui Barbosa (ACRB), Recortes; Francisco de Paula Mayrink, Finanças do Brasil, Discursos Proferidos na Camara dos Deputados nas Sessões de 25, 26, 28 de Agosto de 1891 (Rio de Janeiro, 1891), pp. 28, 29, 33; Ministério da Fazenda, Auxílios a Lavoura (Rio de Janeiro, 1889), p. 53; Jornal do Commércio, Dec. 1889, ACRB, Recortes.


Contract of Banco Nacional with Ministério da Fazenda, Oct. 2, 1889, ACRB, Pasta Banco do Brasil; Castro Careeira, História Financeira, II, 743.


Calculated from Paulo Neuhaus, História Monetária do Brasil, igoo-45 (Rio de Janeiro, 1975), p. 165; DGE, Anuário Estatístico, 2939-1940, p. 1354, and deflated using the index of Contador and Haddad, “Produção Real,” 430-434.


Diário Oficial, Jan. 30, 1890, in ACRB, Pasta Banco dos Estados Unidos do Brasil; Rui Barbosa, Finanças e Política da República: Discursos e Escriptos (Rio de Janeiro, 1892), pp. 18, 39, 42, 120; Ministério da Fazenda (Barão de Lucena), Exposição de Motivos sobre a Situação Financeira e Ideías de Reforma (Rio de Janeiro, 1891), pp. 18, 19; Ministério da Fazenda, Relatório, 1898, p. 61, 1899, p. xii.


Brasil, Presidente, Mensagem apresentada ao Congresso, May 12, 1892, p. 19; Camara dos Deputados, Anais, 1892, III, 134; A Metralha, Rio de Janeiro, Nov. 23, 1893, p. 2; Gazeta de Commércio e Finanças, Rio de Janeiro, Mar. 14, 1896, p. 3.


Associação Commercial de Sáo Paulo, Relatório, 1895, p. 64; Alcindo Guanbara, A Presidencia Campos Sales, Política e Finanças, 1898-1902 (Rio de Janeiro, 1902), pp. 311-325; M. F. Campos Sales to J. C. Rodrigues, Rio de Janeiro, May 23, 1899, in Instituto Histórico e Geográfico do Brasil, Arquivo J. C. Rodrigues (AJCR), 1-3, 4, 85; J. C. Rodrigues to M. F. Campos Sales, Rio de Janeiro, June 1899, IHGB, AJCR 1-3, 4, 86. For a scathing indictment of the policy by an industrialist, see: Luis Vieira Souto, O Ultimo Relatório da Fazenda (Rio de Janeiro, 1902).


Jornal do Commércio, Retrospecto Commercial, 1910, pp. 23-28; Afonso Pena to Davi Campista, Rio de Janeiro, June 18, 1907, AN, Arquivo Afonso Pena, 1.1.335; Rio de Janeiro State, Presidente, Mensagem, Aug. 1, 1905, p. 12; Nícia Vilela Luz, A Luta pela Indusrialização do Brasil, 2d ed. (São Paulo, 1975), p. 193; Ana Maria dos Santos and Sonia Regina de Mendonça, “Intervenção Estatal e Diversificação Agrícola no Estado de Rio de Janeiro (1888-1914),” presented at the meeting of the Associação Nacional de Professores Universitários em História, Niterói, Rio de Janeiro, June 1979; Joan Bak, “Some Antecedents of Corporatism: State Economic Intervention and Rural Organization in Brazil” (Ph.D. Diss., Yale University, 1977), pp. 47-83.


Brasil, Presidente, Mensagem, May 12, 1918, p. 547; Ministério da Fazenda, Contos do Exercício Financeiro de 1925 e Relatório da Contadoria da República (Rio de Janeiro, 1926)> PP. 246. 247; Leopoldo de Bulhões, Os Financistas do Brasil (Rio de Janeiro, 1914), p. 43; Neuhaus, HistóriaMonetária, p. 165; Winston Fritsch, “Brazil and the Great War, 1914-1918, Texto para Discussão No. 62, Departamento de Economia, Pontifìcia Univer-sidade Católica do Rio de Janeiro, Jan. 1984.


Ministério da Fazenda, Relatório, 1926, p. xviii; Winston Fritsch, “1924,” Pesquisa e Planejamento Econômico (Rio de Janeiro), 10:3 (1980).


São Paulo State, Presidente, Mensagem, July 14, 1930, p. 11; Jornal do Commércio, Retrospecto Commercial, 1929, p. 48; João Neves to Getúlio Vargas, Rio de Janeiro, May 23, 1929, in Centro de Pesquisa e Documentação de História Contemporânea do Brasil (CPDOC), Arquivo Getúlio Vargas, GV 1929.0523; Departamento Nacional de Café, Defesa do Café no Brasil, 2 vols. (Rio de Janeiro, 1935), I, 96, 98; Affonso de Escragnolle Taunay, História do Café no Brasil, 15 vols. (Rio de Janeiro, 1939-43), PP. xiii, 360.


Carlos Manuel Peláez and Wilson Suzigan, História Monetária do Brasil: Análise da Política, Comportamento e Instituiçoes Monetárias (Rio de Janeiro, 1976), pp. 369-377; Anníbal V. Villela and Wilson Suzigan, Política do Governo e Crescimento da Economia Brasileira, 2889-2945 (Rio de Janeiro, 1973), p. 32.


Calculated from DGE, Anuário Estatístico, 1939-1940, pp. 1410, 1412, 1417, 1418; Castro Carreira, História Financeira, passim, and Villela and Suzigan, Política do Governo, pp. 418-421.


DGE, Anuário Estatístico, 1939-1940, p. 1417; Castro Carreira, História Financeira, passim; Villela and Suzigan, Política do Governo; Estado de Minas Gerais, Secretaria de Agricultura, Minas Gerais e o Bicentário do Cafeeiro no Brasil (Belo Horizonte, 1929), p. 53; Secretaria de Estado de Negócios da Fazenda do Estado de São Paulo, Evolução Histórico-Funcional (São Paulo, 1978), p. 91.


The Brazilian Yearbook, 1908-1909, J. P. Wileman, ed. (New York, 1909), pp. 371-400.


Secretaria de Estado dos Negócios da Fazenda do Estado de São Paulo, Evolução Histórico, p. 91. Export share of São Paulo’s GDP calculated from Joseph Love, São Paulo in the Brazilian Federation, 1889-1937 (Stanford, 1980), p.286, and estimating the service sector at 19 percent of the total from Villela and Suzigan, Política do Governo, p. 291.


DGE, Anuário Eatatístico, 1939-1940, p. 1417. GDP figure calculated from the Anuário Estatístico, p. 1358, and Haddad and Contador, “Produção Real,” pp. 430-434. Nathaniel Leff, in Underdevelopment, I, 28, uses more recent Haddad data to arrive at the figure of 15.6 percent for exports’ percentage of GDP in the 1911-13 period; my figure was 18.4 percent. This greater estimate for GDP yields an export participation in 1929 of under 13 percent.


United States Department of Commerce, Bureau of Foreign and Domestic Commerce, “Tariff Systems of South American Countries,” Tariff Series No. 34 (Washington, D.C., 1916), p. 14. Leopoldo de Bulhões, a staunch free trader and Minister of Finance, reported in the 1904 Relatório (p. vi) that he “approved of protection for industries . . . that appear viable” and had established high duties since 1896 for that purpose. Maria Teresa Versiani, in “Proteção Tarifária e Crescimento Industrial nos Anos 1906/1914: o Caso das Cervejas,” in ANPEC’s IX Encontró Nacional de Economia, Olinda, Recife (Porto Alegre, 1981) quotes the British Ambassador (p. 1170) saying in 1912: “In respect of customs duties, Brazil is probably the most heavily taxed country in the world.”


The export concentration is demonstrated by Brazil’s Departamento Nacional de Indústria e Comércio, Brazil of Today, Natural Wealth, Economie Forces, Progress (Rio de Janeiro, 1931), p. 161, which shows two-thirds of the country’s exports leaving two ports, Santos and Rio de Janeiro. For more on interprovincial duties, see: Centro Industrial do Brasil, A Campanha Contra os Impostos Interestadoaes (Rio de Janeiro, 1929).


Calculated from Castro Carreira, História Financeira, I, 340-342; II, 627-655; José Murilo de Carvalho, “Elite and State Building in Imperial Brazil” (Ph.D. Diss., Stanford University, 1974), pp. 578–581.


DGE, Armário Estetístico, 1939-1940, pp. 1293, 1410, and deflated using Contador and Haddad, “Produção Real,” 434, 435.


Villela and Suzigan, Política do Governo, pp. 414, 415; DGE, Anuário Estetístico, 1908-1912, Vol. II (Rio de Janeiro, 1927), p. lxxi.


Villela and Suzigan, Política do Governo; Love, Sao Paulo, pp. 206, 302; Robert M. Levine, Pernambuco in the Brazilian Federation, 1889-1937 (Stanford, 1978), 189, 191. John Wirth, Minas Gerais, p. 213.


The Federal District’s receipts calculated from Contadoria Geral da República, Balanço Gerai da Receita e Despesa, 1890, 1900, 1910, 1923, 1930 (Rio de Janeiro: 1893, 1905, 1917, 1924, 1931), passim. Anon., Synthese do Comércio Marítimo Gerai do Brasil de 1873-1874 (NP, ND); Ministério de Agricultura, Indústria, e Comércio, Superintendência do Abastecimento, Movimento de Cabotagem de Diversos Géneros Alimentícios e de primeira Necessidade no Brazil e no Anno de 1919 (Rio de Janeiro, 1920), passim; DGE, Anuário Estatístico, 1939-1940, pp. 1294, 1295, 1320, 1350, 1356, 1413-1415, 1420, 1427.


Contadoria Gerai de República, Balanço Geral.


Ibid.; Castro Carreira, História Financeira, II, 589.


Contadoria da República, Balanço Geral.


Castro Carreira, História Financeira, II, 635, 832; Frederico José de Santa Anna Nery, Le Brésil en 1889, avec une carte de l’empire en chromolithographie, des tableaux statistiques, des graphiques et des cartes (Paris, 1889), p. 452; Ata da Seção do Conselho de Estado, Rio de Janeiro, Mar. 12, 1885, in IHGB, Coleção Instituto Histórico, Pasta 55, Lata 545; Maria Bárbara Levi, História da Bolsa de Valores (Rio de Janeiro, 1979), pp. 107, 108, 280; Ministério da Fazenda, Relatório, 1925, p. 246, DGE, Anuário Estatístico, 1939-1940, p. 1410.


Castro Carreira, História Financeira, I, 379, 468, 541; Eulália Maria Lahmeyer Lobo, História do Rio de Janeiro: Do Capital Comercial ao Capital Industriai e Financeiro, 2 vols. (Rio de Janeiro, 1978), I, 115, 116; Anyda Marchant, Viscount Mauá and the Empire of Brazil: A Biography of Irineu Evangelista de Sousa, 2813-1889 (Berkeley, 1965), pp. 52, 53.


Antonio Delfim Netto, O Problema do Café no Brasil, 2d ed. (Rio de Janeiro, 1979); Roberto Santos, História Econômica da Amazonia (1800-1920) (São Paulo, 1980), pp. 246–259; Barbara Weinstein, The Amazon Rubber Boom 1850-1920 (Stanford, 1983), pp. 225-231; Peter L. Eisenberg, Modernização sem Mudança, a Indústria Açucareira em Pernambuco 1840-1910 (São Paulo, 1977), pp. 111-133. For the debates on federal assistance to coffee and rubber, see: Brasil, Congresso, Documentos Parlamentares, Política Econômica: Defesa da Borracha (1906-1914). (Rio de Janeiro, 1915), and Documentos Parlamentares, Política Econômica: Valorização do Café, 2 vols. (Rio de Janeiro, 1915).


Although foreigners constituted only 5 percent of the national population in 1920, they represented, according to Thomas W. Merrick and Douglas H. Graham, População e Desenvolvimento Econômico no Brasil: De 1800 até a Atualidade, trans, by Waltensir Dutra (Rio de Janeiro, 1979), p. 139, 21 percent of the industrial work force and 30 percent of the commercial work force, and only 6 percent of the agricultural work force. In the Federal District and São Paulo, the centers of prosperity, foreigners represented over half of commerce’s employees and over 40 percent of industry’s.


DGE, Armário Estatístico, 2939-1940, p. 1412; Hélio de Alcântara Avellar, HistóriaAdministrativa e Econômica do Brasil (Rio de Janeiro, 1976), pp. 263, 273.


Love, São Paulo, pp. 156, 255-259.


Janice Theodoro da Silva, Raízes da Ideologia do Planejamento: Nordeste (1889–1930) (São Paulo, 1978), pp. 89-94.


Ministério da Fazenda, Relatório, 1895, p. 112; Documentos Parlamentares, Meio Circulante 1893 a 1895 (Rio de Janeiro, 1914), p. 11; Camara dos Deputados, Anais, 1892, vol. Ill, pp. 217, 218; Centro Industrial do Brasil, Relatório, 1928-1920, p. 174; Brasil Presidente, Mensagem, 1918, p. 477; Memorandum of Belgo-Steel Company, January 1923 in CPDOC, Arquivo Raul Soares, RS 23-01-98/3. For the military’s role, see: Stanley E. Hilton, “The Armed Forces and Industrialists in Modern Brazil: The Drive for National Autonomy,” HAHR, 62 (Nov. 1982), 629-673.


The Brazilian Review, Nov. 12, 1901, p. 768; E. F. Central do Brasil, Relatório, 2907, p. 182; Lloyd Brasileiro, Relatório, 2919, p. 35.


Ministério de Obras Públicas e Viação, Relatório, 1897, p. 135; Camara dos Deputados, Anais, 2922, vol. XVI, p. 35; Lloyd Brasileiro, Relatório, 1923, p. 31.


Castro Carreira, História Financeira, II, 635, 832.


Ibid., p. 831; DGE, Armário Estetístico, 1939-40, p. 1379.


Castro Carreira, História Financeira, II, 832.


Inspectoria Federal das Estradas da União, Estatística 1920, p. xxxviii, 1929/1930, p. 308; Leff, Underdevelopment, I, 148; Julian Smith Duncan, Public and Private Operation of Railways in Brazil (New York, 1932), p. 87.


Brasil, Presidente, Mensagem 1910, p. 710.


Calculated from Contadoria da República, Balanço Geral, 1931, passim, and Inspectoria Federal das Estradas, Estatística das Estradas 1929/1930, p. 308.


Duncan, Public Railroads, pp. 87, 108; Inspectoria Federal das Estradas, Estatistica 1920, p. xxxviii.


Inspectoria Federal das Estradas, Estatísticas, 1910, p. xxxii, 1920, p. xxxviii; John Coatsworth, El impacto económico de los ferrocarriles en el Porfiriato, 2 vols. (Mexico City, 1976), II, 19. Brazilian railroad statistics only enumerate exports equal to 80–90 percent of all Brazilian exports. I assumed that the other 10–15 percent weighed the same as the average of the enumerated exports (given in DGE Anuário Estatístico 1939–1940, pp. 1359, 1362) and traveled on railroads with the same frequency.


Calculated from Inspectoria Federal das Estradas, Estatística das Estradas, 1906, 1910, 1915, 1920, 1925, 1930, passim; Joāo Pandia Calógeras, Problemas de Governo (São Paulo, 1920), p. 21.


Anon., Synthese do Comércio Marítimo, pp. 424, 425; A. C. Tavares Bastos, Discursos Parlamentares (Brasilia, 1977), pp. 340–377; Castro Carreira, História Financeira, II, 477; Afonso Celso de Figueiredo (Visconde de Ouro Preto), A Década Republicana (Rio de Janeiro, 1899), p. 63; Richard Graham, Britain and the Onset of Modernization in Brazil, 1850–1924 (London, 1968), p. 15; Waldyr Niemeyer, O Brasile seu Mercado Interno (Rio de Janeiro, 1948), p. 34.


Subsidy figure calculated from DGE, Anuário Estatístico, 1939–1940, pp. 1339-1341, 1357, and interpolating in current dollars for ten years of incomplete data. DGE, Anuário Estatístico, 1908–1912, II, xxi, 19, 20. It should be added that despite the subsidies and government-mandated fare reductions on Brazilian-produced goods, there were numerous complaints that coastal shipping was exorbitantly expensive, hurting the competitive position of national producers.


Wileman’s Brazilian Review, Aug. 29, 1916, p. 551, Dec. 3, 1924, p. 1633; Graham, Britain and Modernization, p. 90. According to Lloyd Brasileiro, Relatório, 1917, p. 7, 1919, p. 35, 1921, p. 6, the federal government subscribed 25,000 contos of the company’s stock and gave it thirteen German ships that the government had seized during World War I.


Ministério da Agricultura, Movimento de Cabotagem, passim; Ministério da Fazenda, Relatório, 1926, pp. 79–85; Ministério da Viação e Obras Públicas, Relatório, 1925, pp. 166–176; Inspectoria Federal de Portos, Rios e Canais, Relatório, 1930, passim.


Claudio Haddad, Crescimento do Produto Real no Brasil, 1900–1947 (Rio de Janeiro, 1978), p. 18.


The 1860 export percentage of GDP is a conservative estimate, based on Contador and Haddad, “Produção Real,” 430–434, which yields an unlikely 44 percent. Leff, Underdevelopment, I, 21, argues that the nineteenth-century figure for exports should be around 15 percent; that is, despite the rapid growth of industry and domestic agriculture and transportation, which as Haddad shows in Crescimento do Produto, pp. 15, 16, grew substantially faster than exports this century, exports accompanied the growth of the internal economy. Given the lack of a money economy and transportation in the middle nineteenth century, it is unlikely that the internal sector at that time was relatively as large as it became by World War I.


The calculation for the export work force is a ball park figure based on Ellis Júnior’s figure for São Paulo’s agricultural workers in A Evolução da Economia Paulista e Suas Causas (São Paulo, 1937), p. 22, and the percentage of arable land in coffee. I assumed that the size of the work force was proportional to the land under cultivation to arrive at the number of coffee workers in São Paulo. I divided that by São Paulo’s percentage oí all coffee production given in V. S. Wickizer, The World Coffee Economy (Stanford, 1943), p. 248. This technique provides a figure for Minas’s coffee-worker population that agrees with the estimate of Minas Gerais’s Secretaria de Agricultura, Bicentenário do Cafeeiro, p. 79. The figure for the total coffee work force I divided by coffee’s percentage of total exports, assuming other exports averaged a similar productivity, which is reasonable since only sugar and cacao, which has small exports, could surpass coffee, from DGE Anuário Estatístico 1939–40, p. 1380, for all export workers and added another 19 percent of the total for service workers. That I divided by the 1920 work force from Merrick and Graham, População, pp. 200, 201. The result was 12 percent for export workers. To be conservative I have used the figure one-sixth, though one-ninth of the labor force for exports is defensible. The export population by this estimate in 1930 was about 1.44 million, while the 1920 industrial work force had already reached 1.2 million.


See Albert O. Hirschman, Essays in Trespassing: Economics to Politics and Beyond (Cambridge, 1981), pp. 59–97.


Shane Hunt, “Distribution, Growth and Government Economic Behavior in Peru,” in Gustav Ranis, ed., Government and Economic Development (New Haven, 1971), p. 414.


Leff, Underdevelopment, II, 118.


Carvalho, A Construção da Ordem, pp. 81, 83; Monteiro, Funcionarios, p. 37; Joseph Love in “Um Segmento da Elite,” p. 54, found in a study of the political elite of Minas, Pernambuco, and São Paulo between 1889 and 1937 that only one-quarter of the group were fazendeiros; even in São Paulo only 30 percent were. J. W. F. Rowe, Studies in the Artificial Control of Raw Materials Supplies: Brazilian Coffee (London, 1932), p, 28.

Author notes


The author is grateful to Roderick Barman, Winston Fritsch, Eugene Ridings, and two anonymous readers for their helpful comments.