¹ The best-known proponents of this argument are Furtado, Celso, Análise do Modelo Brasileiro [Analysis of the Brazilian Model] (Rio: Civilização Brasileira, 1972)Google Scholar, and Marini, Rui M., Subdesarollo y revolutión [Underdevelopment and revolution] (Mexico City: Siglo Vientiuno, 1969)Google Scholar; Dialéctica de la dependencia [Dialectics of dependency] (Mexico City: Nueva Era, 1973).Google Scholar Also see Erickson, Kenneth P. and Peppe, Patrick V., “Dependent Capitalist Development, U.S. Foreign Policy, and Repression of the Working Class in Chile and Brazil,” Latin American Perspectives 3 (No. 3, 1976), 19–44.CrossRefGoogle Scholar

² O'Donnell, Guillermo A., Modernization and Bureaucratic-Authoritarianism: Studies in South American Politics (Berkeley: Institute of International Studies, University of California, 1973)Google Scholar; O'Donnell, , “Corporatism and the Question of the State,” in Malloy, James M., ed., Authoritarianism and Corporatism in Latin America (Pittsburgh: University of Pittsburgh Press, 1977), 47–87Google Scholar; O'Donnell, , “Reflections on the Patterns of Change in the Bureaucratic-Authoritarian State,” Latin American Research Review 13 (No. 1, 1978), 3–38Google Scholar; O'Donnell, , “Tensions in the Bureaucratic-Authoritarian State and the Question of Democracy,” in Collier, David, ed., The New Authoritarianism in Latin America (Princeton: Princeton University Press, 1979), 285–318.Google Scholar

³ Morley, Samuel A. and Smith, Gordon W., “The Effects of Changes in the Distribution of Income on Labor, Foreign Investment and Growth in Brazil,” in Stepan, Alfred, ed., Authoritarian Brazil (New Haven: Yale University Press, 1973), 119–41.Google Scholar

⁴ For data on the expenditures of workers on consumer durables, see José Serra, “Three Mistaken Theses Regarding the Connection between Industrialization and Authoritarian Regimes,” in Collier (fn. 2), 108.

⁵ See Albert O. Hirschman, “The Turn to Authoritarianism in Latin America and the Search for Its Economic Determinants,” ibid., 81.

⁶ For ample evidence supporting this point, see Serra (fn. 4), 117–27. For an extensive and helpful critique and elaboration of O'Donnell's argument, see Robert Kaufman, “Industrial Change and Authoritarian Rule in Latin America: A Concrete Review of the BureaucraticAuthoritarian Model,” in Collier (fn. 2), 165–254.

⁷ See Serra (fn. 4), 127–28; Hirschman (fn. 5), 76.

⁸ Ibid., 80–81; Serra (fn. 4), 110–11.

⁹ For different (though related) variations of this argument, see Thomas E. Skidmore, “Politics and Economic Policy Making in Authoritarian Brazil, 1937–71,” in Stepan (fn. 3), 3–46; Skidmore, “The Politics of Economic Stabilization in Postwar Latin America,” in Malloy (fn. 2), 149–90; Hirschman (fn. 5); Kaufman (fn. 6); Wallerstein, Michael, “The Collapse of Democracy in Brazil,” Latin American Research Review 15 (No. 1, 1980), 3–40Google Scholar; Sheahan, John, “Market-Oriented Economic Policies and Political Repression in Latin America,” Economic Development and Cultural Change 28 (No. 2, 1980), 264–89.CrossRefGoogle Scholar For recent empirical studies of the links between authoritarianism, repression, stabilization, and economic growth, see Pion-Berlin, David, “Political Repression and Economic Doctrines,” Comparative Political Studies 16 (April 1983), 37–66CrossRefGoogle Scholar, and Cohen, Youssef, “The Impact of Bureaucratic-Authoritarian Rule on Economic Growth,” Comparative Political Studies 18 (April 1985), 123–36.CrossRefGoogle Scholar

¹⁰ For greater detail on the functioning of this mechanism, see Hirschman, Albert O., “The Political Economy of Import-Substituting Industrialization in Latin America,” in Hirschman, , A Bias for Hope (New Haven: Yale University Press, 1971), 85–123, at 117.Google Scholar

¹¹ Hirschman (fn. 5), 73–74.

¹² See Wallerstein (fn. 9), 29.

¹³ This argument is slightly different from the others in that the authoritarian regime is thought to be needed to repress not only the lower classes, but also a variety of other social groups, including the bourgeoisie. In assuming a greater autonomy of the military, it fits quite well the broader conjecture that military regimes emerged in the third world in order to increase the autonomy of their respective states, an autonomy that would be necessary to their development. For this argument, see Horowitz, Irving L. and Trimberger, Ellen K., “State Power and Military Nationalism in Latin America,” Comparative Politics 8 (No. 2, 1976). 223–44CrossRefGoogle Scholar; Trimberger, Ellen K., Revolution from Above (New Brunswick, NJ: Transaction Books, 1978).Google Scholar

¹⁴ For the policies adopted by the military regime, and the behavior of the economy after 1964, see Skidmore (fn. 9, Stepan); Albert Fishlow, “Some Reflections on Post-1964 Brazilian Economic Policy,” in Stepan (fn. 3), 69–118; Bacha, Edmar, “Issues and Evidence on Recent Brazilian Economic Growth,” World Development 5 (No. 1, 1977), 47–68CrossRefGoogle Scholar; Sheahan (fn. 9); Serra (fn. 4); Baer, Werner, “The Brazilian Boom of 1968–72: An Explanation and Interpretation,” World Development 1 (August, 1973), 1–15CrossRefGoogle Scholar; Baer, , “Brazil: Political Determinants of Development,” in Wesson, Robert, ed., Politics, Policies and Economic Development in Latin America (Stanford, CA: Hoover Institution Press, 1984), 53–73.Google Scholar

¹⁵ Baer, Werner and Kerstenetzky, Isaac, “The Brazilian Economy in the Sixties,” in Roett, Riordan, ed., Brazil in the Sixties (Nashville, TN: Vanderbilt University Press, 1972)Google Scholar, Table 8.

¹⁶ Wallerstein draws on Fishlow (fn. 14), as well as on Hirschman's conjectures on inflation, to argue that perhaps the slowdown of growth triggered high levels of inflation as a way to appease the severe social conflict that it generated. For Hirschman's recent version of his earlier sociological theory of inflation, see his “The Social and Political Matrix of Inflation: Elaborations on the Latin American Experience,” in Hirschman, , Essays in Trespassing: Economics to Politics and Beyond (Cambridge: Cambridge University Press, 1980).Google Scholar For Wallerstein's argument, see Wallerstein (fn. 9), 25–28.

¹⁷ Baer, , “Comment,” Latin American Research Review 15 (No. 1, 1980), 42Google Scholar; Baer, , A Industrialkaçao e o Desenvolvimento Econômico do Brasil [Industrialization and Economic Development in Brazil] (Rio: FGV, 1983)Google Scholar, chap. 9. Baer does not elaborate much on the political turmoil. The elaboration that follows is based on Skidmore, Thomas E., Politics in Brazil (Oxford: Oxford University Press, 1967)Google Scholar, chaps. 7 and 8. For another source supporting the arguments of this section, see Flynn, Peter, Brazil: A Political Analysis (Boulder, CO: Westview Press, 1983).Google Scholar

¹⁸ For an account of this episode, which foreshadowed what was to happen after Quadros's resignation, see Skidmore (fn. 17), 112–15.

¹⁹ In other words, it was a classic crisis of business confidence that was largely responsible for the decline in growth. See Lindblom, Charles E., Politics and Markets (New York: Basic Books, 1977), 170–88Google Scholar. For information on business opposition to Goulart, see Skidmore (fn. 17), 225, 244–52; Dreifuss, René A., 1964: A Conqmsta do Estado [1964: The Conquest of the State] (Petrópolis: Vozes, 1981).Google Scholar Also of interest is the concise history of the relation between the state and industrial elites in Boschi, Renato R., Elites Industrials e Democracia [Industrial Elites and Democracy] (Rio: Graal, 1979)Google Scholar, and Diniz, Eli and Boschi, Renato R., Empresanado Nacional e Estado no Brasil [National Entrepreneurs and the State in Brazil] (Rio: Forense-Universitaria, 1978).Google Scholar

²⁰ Stepan, Alfred, The Military m Politics (Princeton: Princeton University Press, 1971), 139–41.Google Scholar

²¹ Hirschman (fn. 5), 79.

²² Skidmore (fn. 17), 248–52, 267.

²³ Ibid., 273–76.

²⁴ Ibid., 266–67.

²⁵ In 1961, the military ministers who attempted to block Goulart's accession to the presidency were openly opposed by the commander of the Third Army from Rio Grande do Sul, General Machado Lopes. This split later became a major deterrent against an attempt to overthrow Goulart. See Stepan, Alfred, “Political Leadership and Regime Breakdown: Brazil,” in Linz, Juan J. and Stepan, Alfred, eds., The Breakdown of Democratic Regimes (Baltimore: The Johns Hopkins University Press, 1978), 121–22.Google Scholar

²⁶ For a concise account of Goulart's change in strategy beginning in mid-March of 1964, see ibid., 123–25. Also see Skidmore (fn. 17), 284–302.

²⁷ The usual argument concerning Congress was that it acted as a conservative body against the progressive policies of the Executive. Wanderley G. dos Santos has challenged this thesis, arguing that the Congress was fragmented and polarized, and suffering from decisional paralysis rather than conservatism. See his “The Calculus of Conflict: Impasse in Brazilian Politics and the Crisis of 1964,” Ph.D. diss. (Stanford University, 1979).

²⁸ When the naval minister tried to quell the mutiny, Goulart dismissed him instead of backing him, and allowed the labor leadership, which was clearly associated with the Communist Party, to participate in the choice of the new minister. That was too much even for the nationalist supporters of Goulart within the army. See Skidmore (fn. 17), 297, and Stepan (fn. 25), 130–31.

²⁹ The officers who led the coup would never have intervened merely to stabilize the economy. They clearly stated that they would only intervene if Goulart overstepped the boundaries of constitutionality. Their main motive was to stop Goulart and the left from using organized labor to overturn the constitutional structure—as the central document of the anti-Goulart conspiracy, the memorandum circulated by Castelo Branco on March 20, made clear. See Skidmore (fn. 17), 295. For greater detail on Castelo Branco's and other leaders' motives for staging the coup, see Schneider, Ronald M., The Political System of Brazil: Emergence of a Modernizing Authoritarian Regime (New York: Columbia University Press, 1971)Google Scholar, and Viana Filho, Luis, O Governo Castelo Branco [Castelo Branco's Government] (Rio: José Olympio, 1975).Google Scholar For a detailed study on the political attitudes of Brazilian elites, see McDonough, Peter J., Power and Ideology in Brazil (Princeton: Princeton University Press, 1981).CrossRefGoogle Scholar

³⁰ An alternative, equivalent way of saying this is that, with the “exhaustion,” or crisis, of the easy phase of import-substituting industrialization, the populist class coalition fell apart. The populist class compromise was possible because of the exuberant growth of the easy phase; with the decline in growth, the lower classes had to be excluded both politically and economically. This is how O'Donnell, Erickson, and Furtado see it, for instance: see O'Donnell (fn. 2), 55–69; Erickson, Kenneth P., The Brazilian Corporative State and Working-Class Politics (Berkeley: University of California Press, 1977), 10Google Scholar; and Furtado, Celso, “Political Obstacles to the Economic Development of Brazil,” in Veliz, Claudio, Obstacles to Change in Latin America (London: Oxford University Press, 1965).Google Scholar Still another way of expressing this thesis is Stepan's; he maintains that the crisis leading to the coup was due to rising demands from the masses when the rate of growth was declining; or, as he puts it, to an increase in the loads of the “distributive capability” of the Brazilian system while its “extractive capability” was decreasing. Stepan (fn. 20), 134–40. Thus, this thesis is shared even by explanations like Stepan's, which give almost equal weight to economic and political factors. The problem with Stepan's otherwise excellent book is that it still gives too much weight to economic factors and fails to show the important causal connection between the rise in demands and the decline in growth. The same is true of Stepan's most recent book, The State and Society: Peru in Comparative Perspective (Princeton: Princeton University Press, 1978). On pp. 79–80Google Scholar, he states that both the economic and the political crises were important in bringing about the military coup of 1964, but he never really discusses the causal connection between the two crises, or even ranks them in terms of importance.

³¹ See Erickson (fn. 30), 99; Weffort, Francisco C., “Sindicatos e Política” [Unions and Politics], Ph.D. diss. (University of São Paulo, 1972).Google Scholar

³² On the attitudes of workers, see de Souza, Amaury, “The Nature of Corporative Representation,” Ph.D. diss. (Massachusetts Institute of Technology, 1978)Google Scholar; McDonough, Peter, “Repression and Representation in Brazil,” Comparative Politics 14 (October 1982)Google Scholar; Cohen, Youssef, “The Benevolent Leviathan: Political Consciousness among Urban Workers under State Corporatism,” American Political Science Review 76 (March 1982)Google Scholar; Rodrigues, Leôncio M., Industrialização e Atitudes Operárias [Industrialization and the Attitudes of Workers] (São Paulo: Brasiliense, 1970).Google Scholar This literature suggests that the radical reforms were only one possible way of articulating the demands from the lower classes.

³³ Obviously, even the narrower demands were already the result of the organization and mobilization of the workers; even these were susceptible to control by the leadership. See Erickson (fn. 30).

³⁴ Economic explanations of military regimes tend to make the same mistake as their counterpart where the explanation of revolutions is concerned. They fail to take into account factors concerning the political mobilization of mass grievances. Charles Tilly has criticized such explanations in depth; see Tilly, , “Does Modernization Breed Revolution?” Comparative Politics 5 (No. 4, 1973), 425–27CrossRefGoogle Scholar, and “Revolutions and Collective Violence,” in Greenstein, Fred I. and Polsby, Nelson W., eds., Handbook of Political Science Vol. III (Reading, MA: Addison-Wesley, 1975).Google Scholar For a recent review of resource mobilization theory, see Jenkins, Craig J., “Resource Mobilization Theory and the Study of Social Movements,” Annual Review of Sociology 9 (1983), 527–53.CrossRefGoogle Scholar Also relevant is the compelling analysis of how the organization of the state is related to social revolution; see Skocpol, Theda, States and Social Revolutions (Cambridge: Cambridge University Press, 1979).CrossRefGoogle Scholar

³⁵ do Carmo Campello de Souza, Maria, in Estado e Partidos Políticos no Brasil [The State and Political Parties in Brazil] (São Paulo: Alfa-Omega, 1976)Google Scholar, has argued that the party system and Brazilian democracy could have survived the socioeconomic changes of the fifties and early sixties.

³⁶ For details on the major groupings of the left as well as on the major inter-union organizations that played an important role in the process of polarization, see Skidmore (fn. 17), 223–28, 276–84, and Harding, Timothy F., “The Political History of Organized Labor in Brazil,” Ph.D. diss. (Stanford University, 1973)Google Scholar, chaps. 9 and 10.

³⁷ According to Erickson, labor leaders had relatively little control over the rank and file. They appeared more powerful than they really were because they took advantage of special circumstances—such as the deteriorating economic conditions of the early sixties, which were quite independent from their organizational control—to call strikes and demonstrations, which they then used for their own political purposes. Goulart finally swung to the left because he believed that he could not afford to lose the support of the left and its labor clientele. See Erickson (fn. 30), 99, 148.

³⁸ Stepan (fn. 20), chap. 5, has shown that successful military coups have occurred only when the military had strong civilian support. By 1964, the Brazilian military knew that this was the case (see fn. 25).

³⁹ See W. G. dos Santos (fn. 27), 94–97, 190–93, 232–33.

⁴⁰ According to W. G. dos Santos, the deadlock and polarization were produced not so much by a conservative reaction of Congress to the reforms; rather, a decisional paralysis was caused by, and exacerbated, the fragmentation and polarization of the party system. W. G. dos Santos shows that after Quadros's resignation, the system moved toward what Sartori has called “a situation of polarized pluralism.” See fn. 27, chaps. 1 and 2, and Sartori, Giovanni, “European Political Parties: The Case of Polarized Pluralism,” in LaPalombara, Joseph, ed., Political Parties and Political Development (Princeton: Princeton University Press, 1966).Google Scholar Fora similar argument linking polarization and the breakdown of Chilean democracy, see Valenzuela, Arturo, The Breakdown of Democratic Regimes: Chile (Baltimore: The Johns Hopkins University Press, 1978).Google Scholar

⁴¹ W. G. dos Santos (fn. 27), 186–236.

⁴² Stepan (fn. 25), 123.

⁴³ Ibid., 124; Skidmore (fn. 17), 276–84, 294–302.

⁴⁴ This is not to say that the state played an insignificant role in the industrial revolution of Europe. The questioning of the myth of the “laissez faire” state goes back at least to Polanyi, Karl, The Great Transformation (Boston: Beacon Press, 1944).Google Scholar Even though the state no doubt had an important role in early industrialization, it had a far more central one in later industrialization, especially in third-world countries. For the role of the state among late developers, see Gerschenkron, Alexander, “Economic Backwardness in Historical Perspective,” in Gerschenkron, , Economic Backwardness in Historical Perspective (Cambridge: Harvard University Press, 1966), 5–30Google Scholar; Cardoso, Fernando H. and Faletto, Enzo, Dependency and Development (Berkeley: University of California Press, 1979)Google Scholar; Cardoso, Fernando H., Empresário Indusirial e Desenvolvimento Econômico no Brasil [The Industrial Entrepreneur and the Economic Development of Brazil) (São Paulo: Difusão Européia do Livro, 1964)Google Scholar; Evans, Peter, Dependent Development (Princeton: Princeton University Press, 1979)Google Scholar; Trimberger (fn. 13); Hirshman (fn. 10).

⁴⁵ I am obviously excluding cases where democracy is imposed by military conquest or is brought about as a part of a struggle for national independence against a colonial power.

⁴⁶ For a clear statement on some of the ways in which state-dominated societies shape the character of the democracies they engender, see Kohli, Atul, “Democracy and Development,” in Lewis, John P. and Kallab, Valeriana, eds., Development Strategies Reconsidered (New Brunswick, NJ: Transaction Books, 1986), 153–82.Google Scholar A number of other scholars have noted that democracy often comes as a “gift” to third-world nations. See Touraine, Alain, “Industrialisation et Conscience Ouvrière à São Paulo” [Industrialization and Working-Class Consciousness in São Paulo], Sociologie du Travail 3 (No. 4, 1961), 77–95CrossRefGoogle Scholar; Weffort, Francisco C., “State and Mass in Brazil,” in Horowitz, Irving L., ed., Masses in Latin America (New York: Oxford University Press, 1971)Google Scholar; Rustow, Dankwart A., “Transitions to Democracy: Toward a Dynamic Model,” Comparative Politics 2 (April 1970), 337–63.CrossRefGoogle Scholar Rustow makes an explicit connection between the gift-like character of democracy in the third world and the likelihood that it will degenerate into an autocratic regime. He argues that democracies that skip the stage of struggle leading to democracy will be unstable and likely to relapse into authoritarianism. His example is Turkey, which “paid the price in 1960,” when democracy was interrupted by a military coup, “for having its first democratic regime as a free gift from the hands of a dictator.” Ibid., 362. Rustow does not elaborate on this point, however. It would be interesting to find out whether the dynamics of the Turkish case resemble those of the formation of a democracy from above in Brazil.

⁴⁷ The argument that I am about to develop is congruent with Dahl's conjectures concerning the path to democracy that he labels as a “shortcut” to democracy. This path is characterized by a sudden transformation of the old regime into a full-blown democracy. Unlike the European countries that took the more gradual path, countries that take the shortcut will engender democracies that are susceptible to failure. According to Dahl, this is because the shortcut does not give elites the chance to develop a system of mutual guarantees before the great diversity of social groups is incorporated through universal suffrage. Consequently, severe conflict is likely because the groups among which a system of mutual guarantees has to be developed will no longer be small and relatively homogeneous elites; they will now include the leaders of the different groups that were suddenly incorporated through universal suffrage. See Dahl, Robert A., Polyarchy: Participation and Opposition (New Haven: Yale University Press, 1971)Google Scholar, chap. 3.

⁴⁸ My sketch of the links between “state-dominated society,” “democracy from above,” and polarization draws on the pioneering studies of the Brazilian system by Campello de Souza (fn. 35) and W. G. dos Santos (fn. 27).

⁴⁹ Alain Touraine (fn. 46), 87, has called the process by which democracy came to Brazil a “democratisation par voie autoritaire,” For a detailed history of the last years of Vargas's dictatorship and the inauguration of the Brazilian democracy of 1946, see Skidmore (fn. 17), 39–64; also Campello de Souza (fn. 35), 64. For an interesting explanation of the continuity of Brazilian authoritarianism based on Weber's notion of patrimonialism and Faoro's earlier work, see Simon Schwartzman, “Back to Weber: Corporatism and Patrimonialism in the Seventies,” in Malloy (fn. 2), 47–88, and Schwartzman, , Bases do Autoritarismo Brasileiro [The Bases of Brazilian Authoritarianism] (Rio: Campus, 1982).Google Scholar

⁵⁰ For documentation, see Campello de Souza (fn. 35), chap. 5. The elite's willingness to cooperate was not only due to Vargas's power and to strategic considerations. Vargas's ability to perpetuate his power was greatly helped by the fact that Brazilian elites shared a basic distrust in democracy, a distrust that ultimately favored the perpetuation of a state-dominated society. For the roots of distrust in democracy, see dos Santos, W. G., Ordem Burguesa e Liberalismo Político [Bourgeois Order and Political Liberalism] (São Paulo: Duas Cidades, 1978)Google Scholar, and Lamounier, Bolívar, “Formação de um Pensamento Político Autoritário na Primeira República” [Formation of an Authoritarian Political Thought in the First Republic], in Fausto, Boris, ed., O Brasil Republicano [Republican Brazil], Vol. II (Rio: Difel, 1978).Google Scholar

⁵¹ The process by which Vargas extended the power of central government over state oligarchies is described in Campello de Souza (fn. 35), chap. 4. On the corporative labor structure, see Simão, Aziz, Sindicato e Estado [The Sindicato and the State] (São Paulo: Dominus, 1966)Google Scholar; Filho, Evaristo de M., O Problema do Sindicato Unico no Brasil [The Problem of the Non-Competitive Sindicato in Brazil] (São Paulo: Alfa-Omega, 1952)Google Scholar; Rodrigues, Leôncio M., Conflito Industrial e Sindicalismo no Brasil [Industrial Conflict and Unionism in Brazil] (São Paulo: Difusão Européia do Livro, 1966)Google Scholar; Rodrigues, J. Albertino, Sindicato e Desenvolvimento no Brasil [The Sindicato and Development in Brazil] (São Paulo: Difusão Européia do Livro, 1968)Google Scholar; Francisco C. Weffort (fn. 31); Schmitter, Philippe C., Interest Conflict and Political Change in Brazil (Stanford: Stanford University Press, 1968)Google Scholar; Erickson (fn. 30); Mericle, Kenneth S., “Conflict Regulation in the Brazilian Industrial Relations System,” Ph.D. diss. (University of Wisconsin, 1974)Google Scholar; Amaury de Souza (fn. 32); and Vianna, Luiz W., Liberalismo e Sindicato no Brasil [Liberalism and the Sindicato in Brazil] (Rio: Paz e Terra, 1976).Google Scholar

⁵² On the formation and history of these parties, see Peterson, Phillys J., “Brazilian Political Parties: Formation, Organization and Leadership,” Ph.D. diss. (University of Michigan, 1962).Google Scholar See also Soares, G. A. Dillon, Sociedade e Política no Brasil [Society and Politics in Brazil] (São Paulo: Difusão Européia do Livro, 1973).Google Scholar For a detailed analysis of how Vargas shaped the electoral system, see Campello de Souza (fn. 35), 105–34. For a review of the extensive literature on political parties, see Lamounier, Bolívar and Kinzo, M.D.G., “Partidos Polftícos, Representação e Comportamento Eleitoral no Brasil, 1946–1978” [Political Parties, Representation and Electoral Behavior in Brazil, 1946–1978], Dados (No. 19, 1978).Google Scholar

⁵³ Leff, Nathaniel H., Economic Policy-Making and Development in Brazil, 1947–1964 (New York: Wiley, 1968).Google Scholar Weffort (fn. 46), 146, notes that the democratic regime of 1946 was radically different from the model of democracy known in Western Europe and the United States, in that “all important organizations that mediate between the state and individuals were, in reality, extensions of the state rather than genuinely autonomous organizations.”

⁵⁴ Campello de Souza (fn. 35, pp. 32–33) is the first scholar who has explicitly attributed the weakness and clientelism of the parties and the party system to the fact that they emerged after much power had been concentrated in the state. Since they did not participate in the decision-making process, their function was to purchase support and quiescence through the dispensation of patronage.

⁵⁵ For a detailed history of the relations between the UDN, the military, and the Vargas camp, see de Mesquita Benevides, Maria Victoria, A UDN e o Udenismo [The UDN and Udenism] (Rio: Paz e Terra, 1981).Google Scholar

⁵⁶ Between 1945 and 1964, the UDN (or factions within it) urged the military to intervene on at least six different occasions. In view of the awful performance of what Stepan calls the “moderating pattern,” it is not surprising that the military finally stayed in power in 1964. That this fact may have been instrumental in the breakdown of this pattern (rather than the ideological changes Stepan mentions as the causes of the breakdown), is suggested by Markoff, John and Baretta, Silvio R.D., “Professional Ideology and Military Activism in Brazil,” Comparative Politics 17 (No. 2, 1985), 175–91CrossRefGoogle Scholar, in a recent critique of Stepan (fn. 20).

⁵⁷ For the strategy of the left, see Erickson (fn. 30), Parts II and III. On the Communist: Party, see Chilcote, Ronald H., The Brazilian Communist Party (New York: Oxford University Press, 1974).Google Scholar

⁵⁸ Skidmore (fn. 17), 112–42.

⁵⁹ Ibid., 149–50.

⁶⁰ On the division within the military, see Skidmore (fn. 17), 303–21. As for the dynamics that led to the victory of the hard-liners, see Cardoso, Fernando H., O Modelo Político Brasileiro [The Brazilian Political Model] (São Paulo: Difel, 1979), 77–78.Google Scholar Of course, the further radicalization of the extremes was not the only factor that prevented a return to civilian rule. By 1964, military officers felt they were prepared to rule and had developed a new ideology that could justify their rule. See Stepan (fn. 20), chap. 8. Recently, however, Markoff and Bareta (fn. 56) have criticized Stepan and suggested, among other things, that it was the poor functioning of the “moderating pattern” rather than the ideological changes mentioned by Stepan that led the military to remain in power.