1. Tomás Molián y Pilar Vergara, “Estado, ideología y políticas económicas en Chile: 1973-1978,” p. 82 in Estudios CIEPLAN 3 (June 1980). A classical text by O'Donnell and a further development of the discussions may be found in The New Authoritarianism in Latin America, David Collier, ed. (Princeton, N.J.: Princeton University Press, 1979). The two most recent books on the Uruguayan political process are Martin Weinstein, Uruguay: The Politics of Failure (Westport, Conn.: Greenwood Press, 1975), and Edy Kaufman, Uruguay in Transition (New Brunswick, N.J.: 1979). For the period before the military takeover, the best article is Carlos Real de Azúa, “Política, poder y partidos en el Uruguay de hoy,” in Luis C. Benvenuto et al., Uruguay hoy (Buenos Aires: Siglo XXI, 1971). A general view of the first four years of military rule can be found in François Lerin and Cristina Torres, “Les transformations institutionnelles de l'Uruguay (1973-1978),” in Problèmes d'Amérique Latine, Nos. 4485-86 (November 1978) Notes et études documentaires, La Documentation Française.

2. OAS, Doc. 1229/80, October 1980.

3. María del Huerto Amarillo, El proceso de militarización del Estado en el Uruguay, (Madrid: Centro de Estudios Constitucionales, August 1981), p. 126ff; Jorge Lanzaro, “La constitución nonata,” in Cuadernos de Marcha 2, 9 (October 1980).

4. Manuel Antonio Garretón, “Procesos políticos en un régimen autoritario: dinámicas de institucionalización y oposición en Chile, 1973-1980.” FLACSO Santiago, Documento de Trabajo No. 104, December 1980.

5. The “Colorado” party is one of the two traditional political parties in Uruguay. Further on, the position taken by its different factions vis-à-vis the referendum will be examined.

6. There are innumerable examples here. For Carlos Martínez Moreno, a prestigious Uruguayan criminologist who lives in Mexico, “as far as the results go, there is no mystery. If the dictatorship does not have enough [votes] with those obtained, it will change them to suit its taste.” “Ni el sí ni el no abren el camino,” Cuadernos de Marcha 2, 9 (October 1980), p. 59. In general, the informed opposition in Montevideo was pessimistic and was surprised by the results of the plebiscite.

7. Juan Mar$iaa Bordaberry, who was elected president in 1971 and was de facto president 1973-76, in Las opciones (Montevideo: Imprenta Rosgal, 1980), p. 44. Bordaberry's thought has close affinities with the “organic” ideal type described by Alfred Stepan in The State and Society: Peru in Comparative Perspective (Princeton, N.J.: Princeton University Press, 1978).

8. Bordaberry, Las opciones, pp. 15, 45.

9. This is a theme that Guillermo O'Donnell developed in an unpublished comparative essay on the economic policies of the military governments of the Southern Cone.

10. In “Economic Policy and Elite Pressures in Uruguay,” American Universities Field Staff Reports 1979, No. 27, South America, Howard Handelman defends a similar viewpoint (p. 12). Divergent opinions can be found in Samuel Lichensztejn, “Le Bloc Financier dominant en Uruguay,” Amérique Latine No. 5, CETRAL (Spring 1981): 72; and in Gerónimo De Sierra, “Uruguay 1973-1980: Éléments pour un bilan des rapports entre politique économique et régime politique,” Amérique Latine No. 7, CETRAL (Fall 1981): 10. On politics and economic policy during this period, see Danilo Astori, Tendencias recientes de la economía uruguaya (Montevideo: Fundación de Cultura Universitaria, 1981); Luis E. González and Jorge Notaro, “Alcances de una política estabilizadora heterodoxa” (Washington, D.C.: Wilson Center, Latin American Working Papers No. 45, 1980; Luis Macadar, “Un nuevo ensayo de reajuste económico: Uruguay 1974-1979” (Montevideo: CINVE, 1980); M. H. J. Finch, A Political Economy of Uruguay since 1870 (London: 1980); and the collection of essays Uruguay: dictadura y realidad nacional (Mexico: ERESU, Unidad de Investigación Latinoamericana, 1981).

11. But patience is not without limits. Since 1980, military concern about the general economic situation has been noticeable, particularly with regard to real wages. The situation has become very difficult in the last few months.

12. The identity of “we” is positive (“shared benefits”) rather than negative (“something bad shared by ‘them’ ”). Popular culture is not in the least chauvinistic.

13. The differences with Chile are obvious and are linked to the far more authoritarian and antidemocratic attitudes of the Chilean elites. The observation concerning the considerable difference in attitude between Chilean and Uruguayan elites with regard to a political opening is Alfred Stepan's.

14. Fernando Henrique Cardoso, “The Authoritarian Regime at the Crossroads: The Brazilian Case” (mimeo, 1981), p. 10.

15. This kind of “generals' democracy” is not the least bit disguised. Some of the results can be surprising: during the last round of promotions and appointments, a newspaper in Montevideo came out with the headline “The Navy Made Its Choice; Now the Army,” La Mañana (Montevideo), 19 January 1982. This statement meant that the admirals had reached a determination concerning their new peers and that now the generals would do the same (including the selection of the new primus inter pares) on the given date. Official rules automatically employ the criteria of age and time in rank to determine retirement, which insures that succession will be orderly. The format mentioned in the newspaper article is exactly the same as that which describes the activities of any important collegial group that gathers to collect candidates for key positions: photographs of the electors arriving and leaving, details about the location and other aspects of the gathering, photographs of the elected, their vitae, and so forth. The only difference is that the article lists themes discussed and decisions taken, but gives no details on the actual discussions, even though these may have lasted for more than a day.

16. Finch comments that Bordaberry “was retained as president, possibly, to disguise the

unconstitutional nature of the regime but more probably to inhibit the emergence of presidential ambition within the ranks of the military high command.“ Political Economy, p. 248.

17. A significant question, then, is, Why is Chile ruled by a single military figure while Uruguay is ruled by a military college? Or in other words, What explains the difference at the top of their armed forces? It is likely that structural and contextual variables give rise to “differential probabilities.” But circumstantial factors (and directly, the personalities of the actors involved) must also play an important role. According to some observers of the Uruguayan case, the only candidate who from the beginning perhaps could have become a lasting leader was General Cristi. He was head of the First Military Region (the most important one in the country) during the most intense period of the struggle against the Tupamaros.

18. This is an obvious consequence of size and context. If war is an imaginable recourse (whether more or less remote), the instrument used to wage war becomes far more important than when war is inconceivable.

19. During its entire professional history, the only war has been the War of the Triple Alliance, in which Argentina, Brazil, and Uruguay fought against Paraguay. Quite aside from the role of Uruguayan troops, the political context of this war as well as the consequences that it had for Paraguay made it preferably forgotten. Even today, although the defeat of the Tupamaros was a military success, it is considered a “dirty war.” This is not the kind of conflict that produces heroic legends for future generations.

20. Commentaries published just after the plebiscite generally examine the figures in terms of the second and third hypotheses. See María del Huerto Amarillo, El proceso, pp. 131ff; Carlos Quijano in “Sí o no, siempre no,” Cuadernos de Marcha 2, 9 (October 1980); Luis Rico, “Uruguay, un plébiscite insolite,” Amérique Latine, No. 5, CETRAL (Spring 1981).

21. Índice Gallup de opinión pública, Op. No. 315, January 1981.

22. This is a key point that cannot be developed here for lack of space. From the mid-nineteenth century on, the political life of Uruguay has been dominated by two major traditional parties: the Colorado party and the Nacional (Blanco) party. Until 1971, these two major parties had the support of over 90 percent of the electorate. But these two are catch-all parties, dividing the Uruguayan class structure vertically rather than horizontally. Electoral legislation always allowed them to put forward more than one presidential candidate, and these candidates could gather votes within the same party (the “lema”). The winner, then, is the candidate who gets the most votes (the “sublema”) from the party that gets the most votes (the “lema”), not just the candidate who obtains a plurality of votes. In 1971 the candidate who obtained the largest number of votes was Ferreira Aldunate (440,000 votes), but the winner was Bordaberry (380,000 votes) because the Colorado party got 681,000 votes compared with 669,000 votes for the Nacional (Blanco) party. It is often the case that there is greater ideological distance among rivals from the same party than among “enemies” from different parties. In practice, all major political initiatives were carried out through an agreement between the two major parties and had to overcome opposition within each party. In the elections of 1971, for the first time an independent coalition (the Frente Amplio or Wide Front) obtained 18 percent of the national electorate and 30 percent of the electorate in Montevideo. The Frente Amplio was a version of Chile's Unidad Popular, but it was located nearer the center of the political spectrum. It included the Christian Democrats, dissidents from the traditional parties, and the entire parliamentary left.

23. Estimates based on the 1975 census give somewhat higher values. The percentages vary from 18 percent (Rocha Department) to 24 percent (Artigas Department). Montevideo (and the national average) gets 19 percent. The population of ages 18-26 was the one that could vote in the 1980 plebiscite, but had been too young to vote in the previous national election of 1971.

24. R² adjusted for degrees of freedom is .85. Bivariate relationships between voting for the left and each one of the regressors (controlling the effect of the others) is clearly linear for industrial workers and urbanization, and acceptably linear for the other factors. There is no multi-collinearity (the highest of the auxillary R² has a value of .16) and the residuals do not suggest heteroscedasticity. The data on the 1971 vote come from Julio T. Fabregat, Elecciones uruguayas (Montevideo: Cámara de Senadores, 1972); and the independent variables are from Carlos H. Filgueira, “Indicadores comparativos de los departamentos del Uruguay,” Cuadernos CIESU 13 (1976). They are all calculated on data from the 1963 national census.

25. Expressing all variables as percentages. The data on the plebiscite used here are the results of the final tally of ballots published in Opinar, Montevideo, 22 January 1981.