Regional Cleavages and Political Patrimonialism in Brazil, Simon Schwartzman, Chapter 4: The Old Republic, Regional Imbalances and Decentralization

REGIONAL CLEAVAGES AND POLITICAL PARTICIPATION IN BRAZIL

CHAPTER 4
THE OLD REPUBLIC: REGIONAL IMBALANCES AND DECENTRALIZATION

1. From Provinces to States

With the end of the imperial regime, in 1889, the regional provinces of the country, now states, were able to play a more active role in the shaping of the national power system. Once consolidated, the new Republican regime was said to be based on what was called the "Política dos Governadores," the politics of the governors, according to which the state oligarchies of the two main states, São Paulo and Minas, decided between them- selves who the Presidents should be. Several scholarly works on the Brazilian "Old Republic" have been published recently, and there is no point in reconstructing its history here.⁽¹⁾ What matters, however, is a discussion of this period within the theoretical framework developed thus far. This discussion will help to overcome several conceptual difficulties in understanding that period, and, consequently, later periods.

In what is probably the only comprehensive study of a provincial administration during the imperial period, Francisco Iglésias offers in a study of Minas Gerais' administration an overview of what political government at the state level was like at the time⁽²⁾. The first element to be noticed is the hierarchical and centralized system of authority. The provincial President, as he was called, was nominated by the Emperor; his loyalty and allegiance was therefore directed entirely to the central government. He did not have to be born in or somehow related to the province under his government, and it was common to have the same man occupy the presidency of several provinces throughout his political career. "During the Monarchy," says Iglésias, "one does not feel the spirit of the regions influencing the government, as it does in the Republic. The statesmen of that time were national men: although identified with their lands, carrying with them the traits of Pernambuco, Minas or Rio Grande do Sul, they did not play the game of regional interests in public life."⁽³⁾

This extreme centralization, however, was combined with high turnover, undefined tasks, and the absence of explicit policies of any kind. Figures for Minas Gerais show that during the 65 years of this regime, the province had 122 administrative periods, an average of a little less than seven months for each administration. These short periods were steps in the political career of the politicians of the time; these men belonged to one of the parties which alternated in the imperial cabinets, and their main task as governors was to assure the victory of their party in the Congressional elections of their provinces.⁽⁴⁾

This system was inefficient in terms of administrative proficiency, and Iglésias gives abundant evidence of the criticism it received throughout the imperial regime. However, it was efficient enough for what mattered most to the centralized and patrimonialistic imperial state, that is, keeping the Central State free from regional demands, assuring the smooth alternation

of the two-party system which worked so well within a tiny section of the population, and so on. No effective command over the social and economic life of the provinces could be exerted, but this was not necessary; no function of interest aggregation or even effective brokerage between local and central interests could be performed, but this was exactly what was needed to assure the autonomy of the central government.

This means that systems of local, autonomous leadership, based on the property of land and family ties, could flourish, but could seldom organize and articulate themselves into effective regional political bodies. The lack of communication between political authority and local leadership led, sometimes, to violent clashes, typified by the celebrated Canudos rebellion of the late nineteenth century.⁽⁵⁾ In terms of political theory, this led to a serious misunderstanding concerning the nature of the Brazilian political system: namely, the notion that the local bosses were the basis and source of political power at the regional and national levels, through a pyramidal series of cumulative aggregation of interests and political articulation. According to this view the local bosses were represented at the regional and national political levels by their educated sons and relatives) lawyers and doctors trained in Rio, São Paulo or abroad, who could absorb all the rhetoric of European liberalism without relinquishing their rural and traditional origins. The consequence was supposedly a kind of political schizophrenia, which separated what was said and written in laws of the political system from the locus where political power really rested. The well known rhetoric and abstract traits of the country's constitutional and legal systems, as well as the country's dominant political discourse, tended to be attributed to this kind of discrepancy between a façade of political integration and institutionalization, and an actual dispersion and atomization of power. The political system thus appeared to be based on the national integration and centralization of power, but this was only a "tiny superstructure" veiling a system of familistic and private power.⁽⁶⁾

The theoretical mistake underlying this notion is that it tends to dismiss the national political structure as practically insignificant. Nevertheless, this political structure maintained the territorial integrity of the country and subdued all attempts of local autonomy and rebellions, which occurred in the country since its independence. The alternative notion that power was actually centralized and concentrated in the executive accounted for these facts, but excluded from the picture the evident manifestations of local private and familistic power. In short, the debate between theory of centralization and that of decentralization seems to be misplaced. It would probably make more sense to consider that the gap was not as much one between apparent centralization and actual local and private power, but rather one between a supposedly European - type representative political system and an actual patrimonial and hierarchical system of authority. The fact that the scope of the political system was limited to a small elite of enfranchised voters and politicians does not mean that some kind of political representation could not have worked within this limit, providing the political system with a real, even though limited, system of interest representation. The end of the Imperial regime in 1889 can be seen as the consequence of Imperial government's inability to accept and integrate a progressively active local and regional leadership.

2. Regionalism and centralization in the Republican movement

The first Republican regime, which lasted until 1930, did not lead to an enlargement of the scope of the political system, in terms of an increase in political participation. It is remarkable that the political participation system was able to remain stagnant while practically all other indicators of social and economic development increased exponentially, as can be seen in the graph below.⁽⁷⁾ Before 1930, the percentage of voters over the country's total population never went above 3.5 per cent and the figures on Congressional elections in imperial times were not much lower; only in 1945, as a matter of fact, did around 15 per cent of the country's population vote in a national election⁽⁸⁾.

This fact, combined with countless tales of fraud and electoral corruption, led to the notion that the Republican period represented a full implementation of the oligarchic, state-based power system, as a substitute for the centralization of the monarchy. Edgar Carone, in his scholarly study of the First Republic, is well aware of the role of the military and the old monarchists in the affairs of the Republic, but seems reluctant to include these important actors in his generalizations on the nature of the system. He refers to "the people" as the most "outstanding absence" in the Old Republican period, but seems to marginalize the military as well:

The establishment of the Republic is the gesture of a class, the demand of an ascending group... The First Republic is a period in which the coffee lords rise to power, reach their climax, and later decline... The government is the representation of only one [class]; the others live in a marginalization process.⁽⁹⁾

Carone is well aware of the difficulties of linking a straightforward class interpretation to well known facts, such as the presence of the military and the monarchists in the political life of the time. His answer tends to be historiographic and casuistic. He says, for instance, that the military "despises civilians" (Part XIII), that they were divided between those who wanted to respect the constitutional rules of civilian power and those who aimed to "co-participate" in power; he finishes by considering the military a segment of the "middle classes." (Part XVI).

The fact was that military participation in politics, at that time, had more to do with regional cleavages and changes in the government structure than with the incorporation of middle sectors into the political process. When considered in terms of the Republic opposition to the old monarchic establishment, these regional cleavages appeared as two clearly differentiated and conflicting Republican ideologies. It would be worthwhile to delineate these ideologies and their regional implications in some detail.

The formal beginning of the Republic movement in Brazil can be situated in 1870, when a Republican Manifesto was issued in Rio de Janeiro.⁽¹⁰⁾ The Manifesto blamed the monarchic regime for all the country's ills and, of course, stated that the Republican system was the solution for all of them. Resides the substitution of the Emperor by a president, however, it proposes little in terms of specific changes of the country's political and social structure. The only item which did appear, and which became central to most of the Republic movement, was the issue of federalism, that is, of granting more autonomy and independence to the provinces, according to the North American model.

The "Manifesto" was, from the beginning, an effort to bring maximum possible support to the movement; therefore, it avoided the most controversial issues. But these issues, which were latent in the differences among the Republican movements in the different region of the country, were to emerge in a series of crises suffered by the Republican movement from its very beginning.

The Republican movement in Rio, as expressed by the newspaper A República, tended to absorb all the rhetoric of the political elite at the time, according to which things were conceptually good or bad, proper or improper, but never convenient or inconvenient for a given interest or group. Thus, the monarchic government was, supposedly,

So bad that, no matter how good the man, royalty forces him to lose whatever qualities he has; monarchy is bad for the country, spoils men, ties their hands, corrupts the King himself.⁽¹¹⁾

This seemed to be reason enough to set one against the monarchic regime, but the Republican opposition was in fact much more specific and concrete than that.

A content analysis of newspapers in Rio and São Paulo during the last five years of the Empire gives initial evidence of two very different types of opposition to the Imperial government.⁽¹²⁾ One of the newspapers was A Província de São Paulo, the leading paper in that state; the other was 0 Paiz, from Rio de Janeiro. The patterns of opinions against the monarchic regime were very similar in both opposition papers, but the political models suggested by them were quite different. The Rio newspaper, which, according to the author, "was connected to some elements of the urban elites (businessmen, military and intellectuals)," tended to favor a military solution, and actually called for army intervention against the Empire; the São Paulo newspaper, connected with "Republican intellectuals and rural proprietors of São Paulo," was explicitly against a military solution.

The reason for this difference is simple: São Paulo had at that time strong regional interests related to the expansion of coffee, and a capacity for interest aggregation which was much higher than in other provinces. A change from a monarchic to a military regime would not enhance the political autonomy they sought, and could actually reduce their chances of getting it. When a military solution came, the conflict between the Paulista Republican Party and the military government broke out immediately, and a pattern of conflict was to reappear intermittently in the years to come.

The Republican ideology, which appears in the content analysis of A Província de São Paulo, can be spelled out in a series of well known traits of the Republican movement in that province. First, the issue of federalism was central, and sometimes even more important than the Republican banner itself. One of the Paulista's Republican leaders, Prudente de Morais, was elected to the Provincial Congress by the monarchist Liberal Party in 1877 and, to justify his acceptance of the nomination, he stated that, once elected,

I will be a truthful Paulista in the first place, only accepting or suggesting those acts which lead to the satisfaction of actual needs and contribute to the grandeur and prosperity of our province...⁽¹³⁾

Second, the Paulista Republicans sidestepped the abolition issue, an explosive topic among the more radical Republicans in Rio and other more urbanized areas. In a formal declaration issued in 1872, the Republicans of São Paulo stated very clearly that they would not push forward a banner such as the end of slavery, which was not "inspired by the nation itself," the "nation" being, of course, the economic and political establishment at the state level.⁽¹⁴⁾ Coffee plantations in São Paulo were moving very rapidly from slave to free labor, and the issue of abolitionism was not so sensitive in the State as it was in other areas of the country. But a careful, non conflictive attitude prevailed. In a statement issued in 1873, at the First Congress of the Paulista party, the principle of regional autonomy was established in order to deal with the slave problem "according to the varying facility of substituting slave with free labor,"⁽¹⁵⁾ with due respect to property rights. Third, the Republican movement was well behaved and non-violent, and it accepted the rules of the political game at the time. They not only disputed positions in the legislative houses of the provinces and the country, but even entered into alliance with the Liberal and Conservative parties. We have seen how Prudente de Morais joined the Liberal ticket in 1877; in 1881 several conservative candidates were elected with Republican support,⁽¹⁶⁾ and in 1884, Campos Sales and Prudente de Morais, both leaders of the Republican Party, were elected to Congress with conservative support.⁽¹⁷⁾ This type of electoral participation continued, and it is estimated that the Republicans had about one fourth of the electoral votes in the province when the Republican regime finally began.

The other type of Republicanism, which was probably best typified by Silva Jardim, a politician from Pernambuco, was drastically different. This was a Republicanism inspired more by August Comte than by Jefferson, and what it took from positivism was the notion of a centralized, rational, modernizing and dictatorial regime. The legitimization for this kind of regime was to be essentially plebiscitarian. In a manifesto issued in 1889, which opposed the Republican Party, Silva Jardim advocated "a strong presidency created by acclamation of popular will, subjected afterwards to universal suffrage..." In another manifesto, it was said that

The Republican regime works in practice through the concentration of political forces, that is, by dictatorship, which is as strong as it is responsible... In the Republican dictatorship, the ruler is a representative of public opinion, which elects or sanctions him.⁽¹⁸⁾

There was no place for regional federation and decentralization of power in this type of political model; since it was free from the direct influence of land-owners and coffee growers, it could easily agree with the abolitionist movement, which was burning in the country's capital in the 1980s. Silva Jardim was in favor of the immediate termination of the slave system.

It was only natural that this kind of radical republicanism did not find support among the majority in São Paulo's or Minas Gerais' Republican movement; as a matter of fact, Silva Jardim entered into open conflict with the party. But he had another powerful ally, which became more important than the regional Republican parties in the events that followed: the national army.

Positivist ideas had been taught in the Military School of Rio de Janeiro since about 1850⁽¹⁹⁾, and notions such as the value of scientific knowledge, rationalization, anti- clericalism, political centralization and effective government were widespread among the military intellectuals of the 1880's. Silva Jardim did not miss this connection when, in 1888, in a meeting at Santos, he openly asked for military support for the Republican cause.⁽²⁰⁾ More important than Silva Jardim was the role of Positivist Republicanism in Rio Grande do Sul: the province was under the direction of Julio de Castilhos, a confirmed positivist himself, and the connection between civilian and military leadership in the province was much closer than in the remaining parts of the country. The Republic began in Rio Grande, establishing a pattern of regional cleavages closely related to the issues of centralization, regional autonomy and civilian vs. military politics, which were to pervade the country's political life in the decades to come.

3. The regional basis of militarism: Rio Grande do Sul.

The military tradition of Rio Grande do Sul did not fail to have a decisive influence on the establishment of the First Republic and its development. This military tradition, which had its roots in the establishment of the Sacramento colony near the Rio de la Plata in 1680, was clearly visible in the nineteenth century. The country's biggest secessionist movement, the Farroupilha War (1835-45), and three foreign confrontations - the Cisplatine campaign, 1817-28, the Platine Campaigns of 1849-52 against Rosas and the Paraguayan War of 1864-70 - all took place either in Rio Grande or near it. Joseph Love gives some estimates of Rio Grande's share in the military efforts at the time: about three fourths of the military men in the conflict against Rosas were from Rio Grande, and 34,000 men from that state were mobilized for the Paraguayan War - more than one fourth of the total Brazilian effort. About 15,000 men, more than one fourth of the Brazilian army in the pre and post Paraguayan campaign, were regularly stationed there. Still according to sources quoted by Love, "more officers holding the rank of Brigadier general and above were from Rio Grande than from any other province."⁽²¹⁾

The close relationships between the civilian and the military elites in Rio Grande is personified in the leading political figure of the state during the nineteenth century, Manuel Luis Osório, who was also one of the most prestigious military leaders in the Farroupilha, Cisplatine and Paraguayan campaigns. His successor in the leadership of Rio Grande politics, Silveira Martins, was not a military man, but ran his Rio Grande Radical Liberal Party in a tight military fashion: '"The Riograndense Liberal Party moves like a regiment of Frederick [ the Great]."⁽²²⁾ When the Republican movement began in Rio Grande, under the leadership of young lawyers trained in the São Paulo Law School, (Assis Brasil, Julio de Castilhos, Borges de Medeiros and Pinheiro Machado),(23⁾ it almost immediately assumed the characteristics of Silva Jardim's radical Republican wing - strong opposition to slavery, Comtian positivism, revolutionary rhetoric, and military involvement. Only on the issue of political decentralization did they join the Paulistas and the remaining Republican movement: they followed the Farroupilha tradition of regional independence, and actually proclaimed the Farroupilha revolution as the root and inspiration of Rio Grande's Republican tradition. This secessionist inclination should not be taken, however, as a federalist ideology; once in national power, the Gaúcho Republicans became strongly favorable to governmental centralization and intervention in other states.

The relationships between Rio Grande's Republicanism and the military establishment can be seen very clearly in the sequence of events which led to the end of the Empire in 1889. A disciplinary matter involving a lieutenant colonel soon became a national crisis involving questions of military honor, subordination of military to civilian leadership, and so on. The "military question" of 1883 arose during a period in which the conservatives were governing, and a series of political cleavages were superimposed - Liberal vs. Conservative, military vs. civilian leadership, military professionals vs. military political leaders, and finally, Republicans vs. Monarchists. It is difficult to outline the full picture, since the old notion that "the enemy of my enemy is my friend" does not seem to hold when institutional, regional and generational cleavages converge into a single crisis. The Republican movement led by Julio de Castilhos and his newspaper, A Federação, was in open conflict with the state's Liberal leadership, represented by Silveira Martins; but, at the same time, it was on Silveira Martins' side, strongly supporting military officialdom against civilian monarchic authority. Rio Grande's bearing in the whole "military question" was remarkable. Visconde de Pelotas, a liberal Senator from Rio Grande and an army marshal, took a strong stand on the issue, in a speech on the Senate floor in 1886; Sena Madureira, the military man who began the whole affair in 1883, spoke out again in 1886, while serving in Rio Grande, condemning the Minister who had punished him in 1884. In late 1886, the governor of Rio Grande, also a military marshal from that state, took a public stand in defense of Sena Madureira; this man was Deodoro da Fonseca, who was to lead the military coup against the Empire in 1889. Julio de Castilhos and his newspaper were present everywhere, giving support and setting fire to the conflict between the military establishment and the civilian government; an engraved plaque given to him by the officers of the Porto Alegre garrison honored him for his "unsurpassable patriotism in the defense of the sacred rights of the military class."⁽²⁴⁾

It is important to stress that the close relationships between military and civilian leadership in Rio Grande did not mean that the two sectors were merged and fully integrated. The Brazilian army, since at least the Paraguayan War, was taking on all the characteristics of a professionalized and institutionalized body within the government, and the whole "military question" should not be explained away as simply a matter of conflict between Rio Grande Republicans and the monarchy. According to Boehrer, the more extreme Republican propagandists, including Silva Jardim and Julio de Castilhos, were excluded from the movement which placed Deodoro at the head of the Provisional Government in 1889⁽²⁵⁾:; after all, this was a military affair, in which civilians should not interfere. Once the new regime started, however, civilians would have to be listened to; but this turned out to be a difficult and complicated task.

Republican history in Rio Grande is marked by a protracted and bloody conflict between the successors of the old Liberal Party, who organized a new Federalist Party under the leadership of Silveira Martins, and the Republican Party led by Julio de Castilhos Julio de Castilhos came to power in Rio Grande under Deodoro' s government, thanks to his faithful support of the central government. He fell when Deodoro was ousted in 1891 by Floriano da Fonseca (another general, also born in Rio Grande); but Floriano was soon to give his support to Castilhos, who came back to power in the state, thanks to the passive support of Porto Alegre's military garrison. Strong control of the state government, organized along the more authoritarian dimensions of the positivistic model, and continuous support of the Federal government in exchange for political and military backing on the internal front are the main items of Castilho's political strategy and his source of strength. On the other hand, the Federalists, or Maragatos,(26⁾ had their political base in the rural cattle-raising frontier area of Bagé. Having been left out of power, they developed. the more radical, caudillo-like and independent minded traits of the Gaúcho tradition. In general, however, the two sides of Rio Grande's political elite shared a tendency towards constant appeal to armed insurrection, close civilian and military ties, and a drive towards centralization and strong control, when in power, and towards decentralization and federalism when out of power; whoever had a given political stand at any given point depended very much on who was in power. In spite of the heavy European migration into Rio Grande, which made it one of the most modern and Europeanized states in the country, the line of political cleavage ignored ethnic boundaries and did not incorporate in any significant way the German or other immigrant groups.⁽²⁷⁾ In this aspect, Rio Grande's political experience is similar to what occurred for a long time in Argentina, and differs sharply from the North American experience, which brought immigrant groups into political life as soon as they were settled.

The details of Rio Grande's influence during the Old Republic need not be given here, but a few highlights are useful to complete this picture. Joseph Love has calculated the number of years spent in all ministries by natives of each Brazilian state during two periods: from 1889 to 1910 and from 1910 to 1930. During the first period, Rio Grande's participation was insignificant: only 2.56 years, as opposed to 12.64 for Minas Gerais, 9.73 for Bahia and 9.02 for São Paulo. After 1910, Rio Grande jumps from twelfth to first place with 18.13 years, as opposed to 16.09 for Minas Gerais and 12.37 for São Paulo. When he considers only the three major ministries - treasury, transportation and justice - Rio Grande falls to second place (15.14, as opposed to 15.45 for Minas Gerais), while São Paulo, still in third place, falls far behind with only 6.71 years.⁽²⁸⁾

The first competitive presidential election in the country's history was held in 1910 and during his political campaign, the opposition candidate carried the banner of civilism. This was also the first time that a military man, Hermes da Fonseca, presented himself as a regular candidate for the presidency. Hermes da Fonseca was, of course, from Rio Grande, where he got 48,000 of the 64,000 votes cast. Ruy Barbosa, a prestigious civilian from Bahia, won in his own state with 75 per cent of the vote, and in São Paulo with 74 per cent. These figures should not be taken as truly representative of popular preference, because of their manipulation by all kinds of legal and illegal techniques.⁽²⁹⁾ But they do indicate where the central government controlled the electoral process, and where an opposition was able to show up. In spite of his indisputable personal prestige, Ruy Barbosa lost in all other states except Rio and Maranhão. Once in power, Hermes da Fonseca worked closely with the civilian political leader of Rio Grande, Pinheiro Machado, structuring a strong regime and a new political organization, the Conservative Republican Party.⁽³⁰⁾

In São Paulo, Pinheiro Machado and Hermes da Fonseca worked together to curb the state's political leadership. In other states, they clashed in what became known as the politics of "national salvation" - intervention in the state's internal affairs on behalf of military leaders or politicians loyal to Pinheiro Machado.⁽³¹⁾ After Hermes' term as president, only Minas Gerais - which had supported him from the very beginning - Rio Grande and São Paulo continued to play a meaningful role in the country's national politics.

4. São Paulo and Minas Gerais

The standard interpretation of Brazilian politics in the Old Republic holds that power was under the oligarchic control of the café au lait axis, that is, the coffee state of São Paulo and the state of Minas Gerais. In fact, the political role of São Paulo was always less significant than its ever growing economic weight suggests. In spite of the fact that the Partido Republicano Paulista supported every winning presidential candidate since 1898, except Hermes da Fonseca, only Campos Sales (1898-1902) Rodrigues Alves (1902-1906) and Washington Luiz (1926-1930) were from that state. To this absence from central power in the period from 1910 to 1926 corresponds a relatively reduced participation in the ministries, as shown by Love in the figures quoted above.

There are two ways of accounting for this apparent marginalization of São Paulo. One is saying that indicators such as holding the presidency, or cabinet offices, are not good enough as a measure of political marginalization; because if politics is about class interests, the only good indicators are those reflecting actual policy decisions related to economic interests. Thus, Valéria Pena quotes the fact that, at a given point in time, the Brazilian Bank used about 70 per cent of its resources to support the coffee economy in São Paulo: this is supposed to indicate that São Paulo did not really lack political power.⁽³²⁾ The other possibility is to argue that, considering the decentralization of the political system at the time, the control of the national state was relatively unimportant for the economic purposes of São Paulo's economic and political elite. Mário Wagner Viera da Cunha, for instance, argues that the autonomy of the states is very high in the beginning of the Republic, but decreases sharply after the First World War. In the beginning, the states were free to contract loans abroad, to tax exports, to establish customs barriers among themselves and to keep their own armed forces. When the economic ascendancy of the United States substituted former British preeminence in Brazil, the tradition of foreign economic dealings with private banks was substituted for direct dealings among states. The national government became thus a necessary broker and its control, a necessary support for state-based economic interests.⁽³³⁾

It is certain that São Paulo interests controlled much of the economic mechanisms having to do with coffee interests.⁽³⁴⁾ The taxes on exports were laid by the states themselves, and represented more than 40 per cent of the states' revenues in the 1915-1929 period. Imports, however, were taxed by the central government, and used to represent between 40 and 50 per 35 cent of the federal revenue up to 1929.⁽³⁵⁾ Since the ability to import is a function of the ability to export, the difference between the two types of taxation actually represented a process of income transfer from the export areas to those states where political power could influence the allocations of federal resources. This situation was certainly felt in São Paulo, and the notion that the state alone was feeding a dependent and parasitic country was pervasive.

In the early twenties, there were a series of military outbursts against the central government in the states; the first, of course, occurred in Rio Grande do Sul. In 1924, a military revolt, loosely articulated with military groups throughout the country, was started in São Paulo.⁽³⁶⁾ This military rebellion received the full support of the São Paulo Chamber of Commerce, whose President, José Carlos de Macedo Soares, gave a clear picture of the state's grievances regarding the political system:

Did São Paulo have the right to abandon the Federation to the often exclusive dominance of lesser statesmen? Did it have the right to allow the utilitarian politicking of "empreguismo" discouraging all civic courage through its systematic support of regional bosses and unjust expropriation of political mandates? The fact is that São Paulo's absence was not limited to "nominating positions," which have been the goal and the ambition of almost all the country's political men. We totally lost our influence in the legislature, both in the Federal Chamber and in the Senate. We were completely excluded from one of the Republic's powers, since there is not one Paulista in the Supreme Court at this moment... We do not have a single representative in the Superior Council of Trade. In diplomacy, as in the judiciary, in the Navy as in the Army, in the powers of the State, everywhere, São Paulo is systematically excluded from all positions of influence and authority.⁽³⁷⁾

What is remarkable about this statement is the clarity with which it distinguishes the two types of politics which coexisted in the country. One, "the goal and the ambition of almost all the country's political men," was the power to employ people, to hire friends in the civil service, to distribute favors and bring benefits to supporting groups. In other words, this was the power to use the state machinery as something to be had, to be profited from, and to increase one's own prestige and wealth - a kind of private patrimony. What São Paulo's leading businessmen wanted was not that, but the control of the decision mechanisms of the country, in other words, the power to use government resources in support of their own independent economic pursuits. For the Paulistas, politics was a way to improve their business; for "almost all" others, politics was their business.

Another expression of São Paulo's discontent appears in a book published in 1930, by a high official of São Paulo's Secretary of Finance.⁽³⁸⁾ Based on abundant statistical information, he claims that, during the period 1922-1924, São Paulo contributed about a third of the country's federal budget, whereas Minas was the state which most cost the Union. Going into great detail, he shows, for instance, that in 1928 the State of São Paulo was responsible for 88 per cent of the state's railroad system, whereas the railroad system in Minas Gerais, slightly bigger than the Paulista, was 70 per cent government owned. In that year, Minas Gerais had 28 per cent of the federal railroads in its territory, as opposed to São Paulo, which had only 4 per cent. His analysis covers federal expenditures on the postal system, health, and education; and for all items, the conclusion is the same: Minas' share of federal expenditure is out of proportion to its contribution to the Union's income. In an appendix to his book, the author even challenges the figures that placed Minas Gerais ahead of São Paulo in terms of population. Indeed, he says, considering the great differences in productivity between the two states, the only explanation for the census differences would be the ability of the Minas politicians to increase the figures in order to get more benefits for their own states. "There is no way out of the dilemma, " says he; "either the population in Minas Gerais is smaller or, if it is bigger, it has an unfortunate inertia."⁽³⁹⁾

Brazilian historiography is particularly poor in its analysis of Minas Gerais as a region within the context of the country's political system. In an often quoted study of some twenty years ago, Cid Rebelo Horta showed how the social and economic elite of the state was interwoven in a tight net of about thirty families.⁽⁴⁰⁾ These thirty families controlled state politics from the local to the national level, and extended their influence into the national sphere. Valéria Pena joins Julio Barbosa and other analysts of Minas politics in considering that the bargaining power of the Minas Gerais' political elite can be explained by the "firm organization and institutionalization of their local power sources " and by the state's relatively high demographic density, whose cause can be traced to the gold period of the eighteenth century.⁽⁴¹⁾

In other words, institutional and demographic factors seemed to be more important than simple economic explanations. The close familistic structure of the Minas Gerais elite, and its active participation in national politics are an argument against the theories that explain their political influence by their control of land and the political system at the local level. Politics in Minas Gerais was always very centralized, and a small committee of notables, known as the "Tarasca, " used to make all important decisions within the Partido Republicano Mineiro in the days of the Old Republic.⁽⁴²⁾

5. The 1930 Revolution: Facts and Ideologies

Brazilian society and politics become increasingly more complex and are studied less as we move beyond 1930. In this year, a revolutionary movement brought Getúlio Vargas, a civilian and former governor of Rio Grande do Sul, to the presidency. Vargas was to remain in power until 1945, and be elected again to tile Presidency in 1950. The year of 1930 is usually taken as a landmark of the beginning of modern Brazil, and, as a matter of fact, during the 1930's, there is a significant increase of several indices of social modernization. Hard data are difficult to obtain, since there was no general census in Brazil in 1930, and the censuses of 1920 and 1940 are not quite comparable. It has been estimated, however, that the country's urban population increased from 10 to about 30 per cent from 1920 to 1940; this percentage kept rising as time passed.⁽⁴³⁾ Government expenditures, which remained stable in value per capita from 1907 to 1943, showed nevertheless a substantial increase in absolute terms after 1930.⁽⁴⁴⁾ After 1930, social expenditures started to appear in the federal budget as separate items: they added up to about 10 per cent of the budget around 1940.⁽⁴⁵⁾ The occupational structure of the population did not change much: employment in agriculture went from 69 to 65.1 per cent from 1920 to 1940, whereas employment in manufacturing increased only 1 per cent, from 13 per cent to 14 per cent.⁽⁴⁶⁾

Interpretations of the 1930 Revolution tend to be a central concern of Brazilian historians and political scientists, since there is a shared belief that it is important to know how modern Brazil started if one wants to understand how it is today and why. The actual facts, which are, of course, very complex, can be summarized in a few items. First, the revolution came about after a crisis in the arrangement according to which the presidency should belong to Minas Gerais; that year, the outgoing president, Washington Luis from São Paulo, wanted to pass the presidency to fellow Paulista, Julio Prestes. The leading states were split: Minas Gerais and Rio Grande rallied against São Paulo and the federal government. The official candidate won, but the opposition candidate, Getúlio Vargas, finally got the presidency after a series of military clashes. The victory was not a direct result of the military campaign, which lasted 21 days, from the 3^rd to the 24th of October, 1930. In that year, the high military command finally decided to ask President Washington Luis to step down. Secondly, there was a climate of military rebellion which had started years before, a movement known as "tenentismo." Luís Carlos Prestes, a Gaúcho army captain who was to become years later the head of the Brazilian Communist Party, was the leader of a famous military column which, in 1925, swept the country in open rebellion against the government, marching several thousand miles from Rio Grande to the Northeast and entering, two years later, into Argentina. In 1930, Prestes had already split ideologically with his former military colleagues, and refused to join the military movement in support of Getúlio Vargas. The young officers who gathered around him, however, were to become the strong men of Vargas' regime after 1930.⁽⁴⁷⁾ Third, the electoral campaign in 1930 was carried on by the "Liberal Alliance" which, for the first time, presented a political platform opposing the state oligarchies and government inefficiency.⁽⁴⁸⁾ Fourth, the revolution of 1930 came about in the midst of a very difficult economic situation, because of the effects suffered by the coffee trade during the 1929 world crisis.

There are two leading interpretations of the 1930 movement in Brazilian literature, according to Boris Fausto. The first follows the standard Marxist-type model of development: it begins with a feudal system and a weak and dependent central government; this traditional system enters into contradiction with an emerging urban bourgeoisie, and ends with a bourgeois revolution, which in turn paves the way for working class access to the political arena. Adapted to the context of an export economy, this theory identifies, in its simple form, the "feudal" with large-scale export farming of a colonial or semi colonial kind; the "bourgeois" revolution is also seen as a nationalist and anti-imperialist movement. Many Brazilian writers have tried to see in the 1930 revolution the bourgeoisie's arrival to power, if not directly, at least in terms of its "objective" consequences.⁽⁴⁹⁾

A typical example is given by Octávio Ianni, who says that

The 1930 Revolution should be interpreted as a super structural moment of "primitive accumulation" which provides the basis for the country's subsequent industrialization, and this is so, in spite of the fact that it was not predominantly led or nourished by the emerging industrial and financial bourgeoisie.⁽⁵⁰⁾

This statement is based on the fact that there was a surge of industrialization in Brazil after 1930. This kind of ex post facto explanation leads him into trouble, however, when he has to explain why the strongest center of opposition to the Vargas regime after 1930 was located precisely in São Paulo, which was also the state which benefited most from the country's industrialization: he is led to say that this opposition, and especially the 1932 Constitutionalist Revolution in São Paulo, "was only against the ideals of the non-bourgeois components of the 1930 Revolution."⁽⁵¹⁾

A second, and seemingly similar model switches the explanation from the bourgeoisie to the middle classes. But this switch is not just an alternative to the former one, since it has deeper connotations. Middle-class theorists think less in terms of an economic process of industrialization than in terms of a social process of modernization, and the "middle class," or "middle sectors," is too inclusive a concept, covering all the emerging groups which were neither part of the landed aristocracy and/or the political elite nor completely subordinated to them. The growing unrest of young military men after 1920 is seen by many authors as an indication of a rising middle sector, which had so far been out of the political system and now demanded more political participation. The 1930 revolution was then, for these authors, essentially a middle-class movement, which opened the doors of the political system to new middle groups.⁽⁵²⁾

What strikes one as the main difference between these two models is not only that they point to two different social groups as the main actors in the 1930 Revolution, but that they also have a very different image of the role of the political system in the process of change In the first case, political power is nothing but an epiphenomenon, which is explained and modified by the clash between two sectors of the country's economic system. In the second case, however, the middle sectors are seen less as a socio-economic class than as a social stratum which has consumption, political participation and political power demands. Political participation and political power are sought not as a means of fulfilling the economic interests of a given sector of the economy, but as a goal in itself, from which other forms of social and economic participation follow.

Since neither of the explanations has either empirical support or much theoretical relevance, their interest lies less in what they explain than in what they suggest.⁽⁵³⁾ They indicate two different intellectual and ideological approaches to the understanding of Brazilian history; more important than this, they reflect two tendencies in the development of Brazilian society which are often pointed out as alternatives, but which are never seen as parts of a simultaneous process of contradictory development. In other words, the differences in the interpretations of Brazilian political history are not just a matter of intellectual disagreement; they lead to quite different evaluations and propositions about the country's political experience and political future.

The details of these ideologies are a chapter in the history of Brazilian social thought, which cannot take much of our attention here. It is enough to note that, whereas the theorists of the "bourgeois revolution" tend to come from a Marxist tradition, which shares with liberal ideologies the image of a passive state, the inspiration of the "middle-class" theorists comes from the fascist experiences and ideologies of Europe. Virgílio de Santa Rosa, one of the more lucid contemporary analysts of the 1930 Revolution, himself a middle-class theorist, considers the Bolshevik and the Fascist experiences examples of the creation of an efficient and rational state structure, directed by the intelligentsia and the middle sectors and concerned with breaking down the traditional power structure in their countries. Azevedo Amaral, one of the most outstanding ideologists of the Vargas regime) explicitly denies the relevance of both the fascist and the bolshevik experiences to Brazil, and does not speak of middle sectors. He goes still further, however: he sees in the countryside's local agrarian leadership the tellurian sources of national strength, and blames the regional oligarchies for plaguing the country with an exotic and fictitious liberal state. The 1930 Revolution is seen, in this context, as an effort to approximate the nation to its authentic sources, the beginning of a new era. This romantic component is absent from other authors in the same line of thought, but all share the idea of a central state which could recuperate its autonomy after decades of control by the regional oligarchies.⁽⁵⁴⁾

The Vargas regime was established as a compromise between some of the traditional regional oligarchies and a group of young military, modernizing intellectuals; it appeared in the wake of a general increase in the levels of political participation in the country. But the Vargas regime soon alienated a very significant sector of the Brazilian intelligentsia, who joined the fascist "integralista" movement and even tried to overthrow the regime in 1937, in an effort to push the ''middle-class" revolution still further. In spite of individual differences and diverging views about the historical origins of the country's problems and possible solutions, it is not too difficult to see how the "middle-class" ideologists share the notion (and the ideal) of a governmental structure, free from class and regional constraints, which could lead the country's life according to the will of its leadership.⁽⁵⁵⁾

6. The new centralization

Placing ideological disputes aside, the fact is that the political regime, inaugurated in 1930, signified a radical departure from the previous one in terms of its much stronger centralization and concentration of power. Their leaders were not representatives of the "bourgeoisie," nor of "rising middle classes." They could clearly be linked to the political and military tradition of Rio Grande, and they responded in a diffuse, uncertain and undecided way to the demands of urbanized sectors of the country for some elements of social welfare, an increase in the efficiency of the central state and in its administrative, military and economic strength.

Presidential elections during the Old Republic often had no competition (Rodrigues Alves in 1902, Afonso Pena in 1906, Epitácio Pessoa in 1918, Washington Luis in 1926). When competition existed, cleavages were essentially regional. Ruy Barbosa, twice defeated in competitive elections, had his political base in the state of Bahia, and Getúlio Vargas in 1930 drew his support from Minas Gerais, Rio Grande do Sul and Paraíba Table 11 presents data for a comparison of these competitive elections.

TABLE 11 COMPETITIVE PRESIDENTIAL ELECTIONS DURING THE FIRST REPUBLIC
	1910	1914	1930
States in the opposition	São Paulo, Bahia, Rio de Janeiro and Distrito Federal	Bahia	Minas, Paraíba, Rio Grande do Sul
Average per cent votes for candidate winning presidential election in states where he won.	87.8 (15 states)	95.6 (19 states)	85.4 (only 9 states)
Average per cent votes for candidate losing presidential election in states where he won.	71	62.1	84.8
Per cent of voters of total population	1.64%	2.14%	5.10%
Source. Data gathered by Celina Moreira Franco, Lucia Lippi Oliveira and Maria Aparecida Hime, "O Contexto Politico da Revolução de 30," Dados 7, 1970; complemented by information collected by Irene Moutinho from the Annaes do Congresso Nacional and other sources.

The year of 1910 witnessed the first competitive election in the Republic, with a turnout of about 1.6 per cent of the total population. In 1914, the turnout was 2.14, but in 1926, not shown in the table, there was a single candidate, and the turnout figure fell to 2.06. Only in 1930, at the brink of the Vargas revolution, did the turnout rise to above 5 per cent. The 1930 election was more competitive than the previous ones at the state level, and most evidently so in the city of Rio de Janeiro, where the winning candidate received only 51 per cent of the vote. In general, however, the pattern is the same both in the winning and in the losing states, with one additional characteristic: the level of turnout generally increased in the opposition states. This supports the general proposition that political competitiveness tends to amplify the scope of political conflict; it signifies a departure from the previous pattern of extremely limited enfranchisement.⁽⁵⁶⁾ The political system based on the dominance of the more traditional states could not sustain itself anymore, but power did not shift to the more modern Sao Paulo, which lost out once again. The Vargas regime transferred the source of power still further from the states to the central executive, and increased the participation of the military and the dependence of the regional oligarchies upon the central government.

Benedito Valadares, a legendary politician from Minas Gerais who became the living symbol of "traditional" politics in Brazilian political folklore, gives a candid account of how his political career was started and implemented under the shadow of Vargas, after 1930.⁽⁵⁷⁾ The essential element of his political strategy was his almost absolute personal loyalty to Vargas. In this, he was in open opposition to the leading Mineiro politicians from the pre-Vargas period, such as Antonio Carlos, who attempted to exercise Minas Gerais' right to the presidency under the old agreement of the "politics of the governors'"; and to young leaders, such as Virgílio de Melo Franco or Gustavo Capanema, who tried to play a political role of their own. Valadares' letter of introduction to Vargas was his performance as chief of police in the fight against the Paulista insurrection to 1932. He does not care to justify his stand in defense of the Vargas regime; the fact that he is recompensed with a nomination as the President's deputy in the state of Minas seemed to be enough to demonstrate that he was right. It is interesting to see how Valadares goes to Vargas after the President of Minas Gerais, Olegário Maciel, dies:

His death was a terrible blow, since, besides liking him very much, I was completely lost, without a chief or a guide, which is so necessary to the young in public life. Arthur Bernardes was on the other side, Antonio Carlos had his preferences. The new ones entangled themselves in political competition...⁽⁵⁸⁾

Valadares goes to Rio de Janeiro and asks for an audience with Vargas, from which he emerges, in practice, as Vargas' s personal deputy in Minas Gerais. Once in power, he makes some attempts to behave as a free agent and is especially active in the political maneuvers for the presidential election, which was supposed to be held in 1938. He attempts to obtain an agreement on a single candidate, who would be himself; but the 1937 coup, which was to establish Vargas as dictator, was already under way, and he decided to join. He was to remain at the direction of the state's affairs until 1945, and from then on was a national leader of the Social Democratic Party, which emerges in 1945 after Vargas' fall, bringing together the majority of the local "traditional" politicians in Minas Gerais and other states.

This is, in short, the secret of the politico mineiro which survives the First Republic. Not quite the representative of rural oligarchies, nor the expression of ill dissimulated economic interest, but the open agent of the head of State, working in open confrontations, as in 1932, or more often in undercover maneuvers, but always in a context where the major trump is access to the dominant center of political and economic power, the federal government. Valadares never comes into the open and is the main instrument of the elimination of the old state leadership's aspirations for political autonomy.

If, in Minas, the transition to the new centralization was relatively easy, in São Paulo the situation was much more difficult. First, of course, the state had not joined the revolutionary movement of 1930. But, mostly, there was very little in common between the new political leadership and the state's interests. Warren Dean quotes the fact that, when Vargas' deputy João Alberto comes to São Paulo shortly after 1930,

He was so unaware of the size of São Paulo's industrial park that he considered resolving labor troubles by inviting the owner and one worker from each firm to a meeting. He didn't realize the audience would have numbered 11,000.⁽⁵⁹⁾

In general, the policy of the new government was liberal in economic terms, and the eventual support it gave to populist demands was not particularly to the liking of São Paulo's industrialists.⁽⁶⁰⁾ As Dean summarizes,

The most striking change in the economic environment of the 1930's was the increasing intervention of the government. But this intervention was not designed to accelerate the process of industrialization: the alternatives of the export economy had not yet been played out.⁽⁶¹⁾

When, after 1937, the liberal outlook was changed into an explicit policy of economic growth and industrialization, the path chosen by the government was not to support the Paulista industrial system, but to keep initiative under control of the state. The government could certainly not ignore the resources that were available in São Paulo, and soon a "rapprochement" began between government and industrialists; but initiative and entrepreneurial leadership all belonged to the former.

In 1932, hope had vanished for those who had joined the Revolution, expecting that the new regime would simply re-establish the power and autonomy of the state leadership: many of Vargas supporters in 1930 were now in the São Paulo barricades against him. One of these was the gaúcho João Neves da Fontoura, leader of the Liberal Alliance, which had supported the Vargas campaign. The other was the "Paulista" Julio de Mesquita Filho.⁽⁶²⁾ Once in power, the logic of the situation was such that the new centralization carried out the alienation of the more traditional political leadership in the south. Borges de Medeiros, the leader of Rio Grande's Republican Party, who had placed Vargas in the state leadership and supported his candidacy to the presidency, also joined the 1932 uprising against him, in behalf of state autonomy. Champions of federalism but promoters of centralization--such seems to be the historical predicament of Rio Grande's politicians.

The tendencies towards an increase of executive power, an increase in the role of the central state in the social and economic life of the country, continuous co-optation of autonomous political leadership at all levels, and subordination of the economic to the political process would be firmly established during the Vargas regime and never actually from then on. At the same time, however, interregnal cleavages gradually became intra-regional and national in a process which began in Rio and tended to remain an essentially urban fact. This combination of strong centralization and "plebiscitarian' politics seems to have been at the root of Brazilian populism since the late thirties.

Notes

1. The most comprehensive work on the 1889-1930 period is the trilogy by Edgar Carone (1969, 1970, 1971).

2. Francisco Iglésias (1958).

3. Iglésias (1958) , p. 39.

4. Iglésias (1958), p. 47.

5. Euclides da Cunha (1944).

6. Cf. A. O. Cintra (1971); M. I. Pereira de Queirós (1956-57); Nestor Duarte (1939).

7. Maria Antonieta Parahyba (1970).

8. Figures from Guerreiro Ramos (1961), p. 32; Joseph L. Love (1971), p. 119; and Tribunal Superior Eleitoral (1964).

9. E. Carone (1969), p. 288.

10. The following analysis is mostly based on materials gathered by George C. A. Boehrer (1954). This is a detailed, state-by-state account of the organization of the Republican movement in Brazil, followed by an analysis of the main issues and points of conflict in the Republican platforms. It is curious how the author assumes, from the beginning, that the Republican movement was to become a national party, which was not to be true, even after the Republican regime was inaugurated.

11. A República, December 13, 1870, as quoted by Boehrer (1954), p. 37: "O governo monárquico é tão mau, que por melhor que seja o homem, a realeza força-o a perder o que ele tem de bom; a monarquia é má para o país, estraga os homens, ata-lhes as mãos, corrompe o próprio rei."

12. Cf. Irene Maria Magalhães (1970), p. 173-178.

13. Open letter published in A Província de São Paulo, August 4 1877, as quoted in Boehrer (1954) p. 86: "Se for eleito, na assembléia provincial, procurarei antes de tudo ser verdadeiro paulista, só aceitando ou indicando medidas que importarem a satisfação de necessidades reais e que forem tendentes ao engrandecimento e prosperidade de nossa província...

14. From the 1888 manifesto transcribed in Boehrer (1954), p. 266: "O partido republicano, cujas tendências não são autoritárias, está bem longe de executar reformas que não sejam inspiradas pela própria nação."

15. Boehrer (1954), p. 267.

16. Boehrer (1954), p. 98 ff.

17. Boehrer (1954) p. 103.

18. From the 1888 manifesto, transcribed in Boehrer (1954), pp. 233-34: "...O regime republicano exerce-se no campo da ação prática pela concentração das forças políticas, isto é, pela ditadura, tão forte quanto responsável... na ditadura republicana quem governa é um representante da opinião pública. por ela instituído ou sancionado."

19. Boehrer (1954), pp. 229, 283.

20. Boehrer (1954), pp. 279-80, gives evidence of widespread support for military action among civilian Republicans, and describes the activities of Silva Jardim on behalf of Sena Madureira, an officer who was to become the pivot of the "military question" (see section III, below).

21. Joseph L. Love (1971), pp. 15-16. Much of the following analysis of Rio Grande's political role is based on this excellent historical study.

22. Love (1971). p. 24.

23. Love (1971), p. 28: "All four of these men had graduated between 1878 and 1885; and all came from ranching families; three of them would govern Rio Grande and the fourth, Pinheiro Machado, was to become the state's foremost representative in the federal senate."

24. Quoted by Love (1971), p. 31.

25. George C. A. Boehrer (1954), p. 286, and (1966) pp. 43-57.

26. The nickname "Maragatos," given to the Federalists, seems to have originated in a place called Maragataria, which was a passing point between Brazil and Uruguay. It implies that the Federalists had strong connections with Uruguay (Silveira Martins himself was born in that country) and moved quite freely between the two countries, where they could get supplies or find a sanctuary and an often illegal market for their cattle. For a description of the conflict between Republicans and Federalists, see Love (1971), chapter iii.

27. Love (1971), p. 131: "Despite the increasingly active role cf the colono population in the state's economic affairs, the colonos played a minor role in politics. The traditional estancieiro's economy has its analogue in politics. The vast majority of the establishment. and opposition leaders of Rio Grande were members of the landowning class. The close ties between these landowners and the military elite is what makes of Rio Grande's politics something quite different from what a political system based on large land properties is usually supposed to be."

28. Love (1971), p. 123, table 3.

29. For election figures from the Old Republic I am relying on information gathered by Irene Moutinho from several sources, but mostly in the Annaes do Congresso Nacional ("Apuração da Eleito de Presidente e Vice-Presidente realizada a 1 de Março de 1910"). Election outcomes had to be confirmed by Congress, and the criteria for this confirmation are strictly partisan. On the whole, the analysis of election outcomes during the period suggests that electoral fraud is probably a better indication of political strength than the votes themselves. Ruy Barbosa, after his defeat., was able to draw a picture of the outcomes according to which he actually had won the election throughout the country. Correct figures, in short, seem to be difficult to get and are not too meaningful politically.

30. Carone (1969), p. 256.

31. Carone (1969), p. 265. Pinheiro Machado is a central figure in the analysis, which Love makes of Rio Grande's role in the Brazilian Old Republic. Cf. Love (1971), chapter vi, "Pinheiro and His Party."

32. Maria Valéria Junho Pena, (1971), p. 43. The reference is from an article by Juarez Távora published in O Estado de São Paulo as quoted by Boris Fausto (1970), p. 76. Juarez Távora was a young lieutenant of the revolutionary movement, who brought Vargas to power in 1930, and his article was meant as an attack on São Paulo's power pretensions at the time. As we shall see later, the 1930 Revolution and its aftermath, the unsuccessful 1932 insurrection in São Paulo, are further confirmations of the political marginality of the country's economic center.

33. M. W. Vieira da Cunha (1963), pp. 19-20.

34. The first governmental institution treated for intervention in the coffee economy was the Instituto Paulista de Defesa Permanente do Café, which controlled the flux of coffee towards the Port of Santos and the supply, and financed the stockage of surpluses. This Institute existed from 1924 to 1931, and from then on, there was always a national organization with increasing power to interfere in the coffee economy: the Conselho Nacional do Café (1931-33), the Departamento Nacional do Café (1933-46), the Departamento Econômico do Café (1946-52) and finally the Instituto Brasileiro do Cáfé, which still exists. Cf. Elisa Maria Pereira Reis (1972), p. 13 passim.

35. Fernando Antônio Rezende da Silva (1971), pp. 235-282.

36. For an account of the 1924 Revolt in São Paulo cf. Carone (1969), p. 373 ff.

37. José Carlos Macedo Soares (1925), p. 12: "Tinha São Paulo o direito de abandonar a Federação ao domínio - por vexes exclusivo - de estadistas menos adiantados, de permitir a politicagem utilitária do 'empreguismo, desanimando todas as coarsens cívicas, pelo apoio sistemático aos mandões regionais, pela expropriação injusta dos mandatos? Pois bem: a abstenção de São Paulo não se limitou aos cargos de nomear, que tem constituído o alvo e a ambição de quase todos os homens públicos do país. Perdemos totalmente a influência legislativa, tanto na Câmara Federal quanto no Senado. Fomos completamente excluídos de um dos poderes da República, pois no Supremo Tribunal Federal, a esta hora, não há um único juiz de São Paulo. Entretanto, deles dizia Ruy Barbosa: "Podemo-nos consolar da fraqueza de seus políticos., ao menos, com a serenidade impoluta dos seus magistrados. Não temos um só representante no Conselho Superior do Comércio. Na Diplomacia, como na Magistratura, na Marinha, como no Exército, nos poderes do Estado, por toda parte, em todos os postos de influência e de autoridade, São Paulo está sistematicamente excluído."

38. Manuel Olympio Romeiro (1930).

39. Olympio Romeiro (1930), p. 102.

40. Cid Rebelo Horta (1956).

41. M. Valéria Pena (1971), p. 45: "O alto poder de barganha política das facções mineiras na estrutura do poder nacional parece poder ser explicado em virtude da firme organização e institucionalização de seu poder local, que não se apresentava fracionado em nível externo, além de sua alta densidade demográfica originária do ciclo do ouro."

42. For a summary of Minas Gerais' politics during the Old Republic, cf. David V. Fleischer (1972), chapter iii.

43. For a summary of the country's development since 1930, see Schmitter (1971), chapter ii, pp. 20-46. This estimate of urbanization is taken from Pedro Pinchas Geiger (1962).

44. Rezende da Silva (1971), p. 245.

45. Rezende da Silva (1971), p. 256.

46. Cf. Table 2.1, Schmitter (1971), p. 23. For detailed figures on industrialization in the twenties and thirties, see Boris Fausto (1970), pp. 19-28.

47. The bibliography on the revolutionary movement of 1930 is quite extensive. On the "Tenentismo" movement, cf. Virgilio de Santa Rosa (1933); R. J. Alexander (1956); J. D. Wirth (1964); Hélio Silva (1968). This list is far from complete.

48. The radical wing of the revolutionary movement was embodied in several efforts of political organization and mobilization, such as the Legion of October, the October 3rd Club and the Revolutionary Legion. For an overview of these movements, particularly the latter, cf. Peter Flynn (1970), pp. 71-106.

49. Boris Fausto's work is a very convincing demonstration of the untenability of the classist interpretations of the 1930 movement, either in terms of a bourgeois or of a middle-class revolution. Theoretically, however, he turns his careful historiographic study into a not. too clear discussion of the "dualist" theories of political development, and it becomes quite difficult to understand the links he establishes between the "dualist" and "classist" interpretations of the Brazilian political process. One example of the mistakes created by the class interpretation of the 1930 movement, quoted by Boris Fausto, is André Gunder Frank, who tries to link Rio Grande's political role in the episode to the presence of European migration and the beginnings of a rudimentary industry in the area. Cf. Boris Fausto (1970); Celina do Amaral Peixoto Moreira Franco, Lucia Lippi Oliveira and Maria Aparecida A. Hime, (1970); André Gunder Frank (1967).

50. Octávio Ianni, (1965) pp. 135-36.

51. Ianni, (1965) p. 138.

52. The theorists of the middle class include Virgilio Santa Rosa (1963), Guerreiro Ramos (1961) and Hélio Jaguaribe (1962). A summary of their theories is given by Boris Fausto, C. Moreira Franco and others.

53. As lanni's example above typifies, these theses are not subject to empirical falsification. Short of the very top and very bottom of the social structure, everyone else is "middle class," a truism which has not much explanatory power. The insistence on "middle class" theories as an explanation of social movements in Latin America, and mainly as an explanation of the military presence in politics, is probably a hangover from a conceptual framework which cannot be freed from the few alternative explanations based on three or four classes and their permutations. "Middle classes," or better still, "middle sectors," is a residual category which can be used when the others obviously cannot. But this type of explanation often comes from something deeper than this conceptual difficulty, as I am trying to show.

54. Azevedo Amaral (1934). Wanderley Guilherme dos Santos has called my attention to the fact that neither Amaral nor his contemporaries in the thirties (Nestor Duarte, Oliveira Vianna, Francisco Campos) show in their work the presence of a direct and strong German influence. In this book, as a matter of fact, the only German author which appears is Spengler, quoted from a Spanish translation. The indirect influence, however, is some- times very obvious. In a given point, Amaral says that "as fases áureas do Estado em todos os tempos e em todos os países têm sido sempre as épocas de predomínio politico dos elementos sedentários recrutados da classe agrícola," (1934), p. 151. It is from the land that the true and best nationality emerges, and the future lies in the dialectic process by which the nation will recover its authentic roots. For a comprehensive bibliography and analysis of Amaral's work see Aspásia Brasileiro Alcântara (1967). For an overview of names and subjects in the history of Brazilian social thought, cf. Wanderley Guilherme dos Santos (1967).

55. To establish linkages between substantive interpretations and normative evaluations is a necessary and important task in the sociology of knowledge, even if it runs the unavoidable risks of all generalizations. Bolívar Lamounier, in an unpublished paper, made a lengthy criticism of the analysis above, saying, for instance, that Raymundo Faoro perceives the Brazilian political system as centralized and authoritarian; he does not have a class interpretation of the political process, although he is ideologically anti-authoritarian. However, he says it is implicit in my discussion above that he should not be so. I am certainly not responsible for all the stretching that can be done with the general links which I am suggesting exist between substantive analysis and ideological preferences. The fact, however, is that an author such as Raymundo Faoro, who is a strong critic of the patrimonialist and tutelary political structure which he studies, does not seem to suggest a meaningful political alternative. Other authors prefer to accept the centralized and hierarchical system of authority as something to be supported and stimulated. What remains to be seen is whether they have really a substantively different approach to the political issues involved, or if they are simply two sides of the same coin.

56. E. E. Schattschneider (1960), chapter i.

57. Benedito Valadares (1966).

58. Valadares (1966), p. 36: "Foi um choque tremendo, pois, além de o estimar muito, ficara desarvorado, sem o chefe ou guia tão necessário aos moços na vida pública. Arthur Bernardes estava do outro lado, Antônio Carlos tinha os seus preferidos. Os novos se engalfinhavam na competição política..."

59. Dean (1969), p. 183.

60. Cf. Dean (1969), pp. 191-92, for a description of a conflict between the Ministry of Labor and the industrialists shortly before the 1932 insurrection. Free from direct industrialist interests, the central government was eventually able to pursue social welfare values which collided with São Paulo's new, exploitative young capitalism.

61. Dean (1969), p. 205.

62. João Neves da Fontoura (1963).