STATE AND ALLIANCES IN ARGENTINA, 1956-1976*

In that I study the attempt, begun in to implant and I I the

" A first version of this paper was prepared in Spanish for the Conference on "The State and Economic Development in Latin America", Cambridge University, December 12-16, 1976. 1 For an already published characterization of this type of State see O'Donnell, (1972 and. 2 For a conception which considers any kind of political domination preferable to "political instability", this cannot but seem the consequence of an acute pathology. The inventory of dysfunctional psychological traits of Argentine "masses and elites" has been one of the favourite occupations of influential currents in the social sciences (Kirkpatrick, 1971) and of the apocalyptic elements of the Argentine right. Neither has much been contributed by certain visions of the problem of dependency as a deus ex machina which only has to be invoked to explain everything away. Reflections on the "stalemate", or mutual blockings of political and social forces in Argentina have been more fruitful, above all those which have connected it with the gramscian view of hegemonic crisis (cf. Portantiero, 1973). But beyond the descrihing of the stalemate and outlining some of its consequences, the questions still remains: what are the power relationships that have produced that stalemate? junctures with the historical process in which they have emerged and dissolved. In the book already mentioned I indicate some specific differences between the 1966-1972 Argentine case of "bureaucratic-authoritarianism" and other Latin American cases. Briefly, the principal differences were: (1) the lower threat 3 level prior to the implantation of the BA State; (2) the lesser severity of the repression imposed upon the popular sectors and their political allies; (3) the greater autonomy of the popular sectors (and, within them, of the working class) and of the trade unions, with respect to the State and the dominant classes; (4) the moderate fall of industrial wages and the more pronounced decline in the incomes of a sizeable proportion of the employed middle sectors; (5) the rapid formation of an alliance of the popular sectors and the unions with the domestic bourgeoisie 4 against the new State and, particularly, against its "e££icientist" and internationalizing policies; (6) the conflict between the governmentand, with it, the large bourgeoisie -and the pampas bourgeoisie; and (7) the decisive role of peronismo as the expression and mobilization channel of a heterogeneous constellation of forces in opposition to the BA State. These elements are fundamental in an explanation of the unusual conflicts which arose within the State's institutions and, also, of the social explosions which provoked a collapse unknown so far in the other Latin American BA States. 5 These factors account 3 The "threat" concept refers to the degree to which internal and external dominant classes and sectors considered that the breach of the capitalist parameters and of the society's international alignments was imminent and willingly sought by the leadership of the popular sector. . 4 I define "domestic bourgeoisie" as the fractions of the urban bourgeoisie which control enterprises mostly or totally owned by nationals. The definition excludes, therefore, the local subsidiaries of transnational firms and the agrarian bourgeoisie. Within the latter, the "pampas bourgeoisie" is that which controls the grain-and beef-exporting region of the Argentine pampas. The domestic bourgeoisie must in turn be disaggregated, since it ranges from the urban bourgeoisie's fully national and weakest layers to oligopolistic corporations intimately connectedby diverse mechanisms -with international capital. Making a different distinction, I shall also further on speak of the "large (urban) bourgeoisie", when referring to the set formed by the branches of transnational corporations and by the domestic bourgeoisie's oligopolistic fraction. "Below" the large bourgeoisie, what I shall call the "local" -or simply 'weak' -bourgeoisie is left, formed by capitlists controlling non-oligopolistic firms, smaller and usually less capital-intensive than the large bourgeoisie's. I shall also refer to the Confederaci6n General Econ6mica (CGE) , an organization which throughout invoked the representation of the local bourgeoisie. The "popular sector" refer to the working class and the employed and unionized middle sectors; the Confederaci6n General del Trabajo (CGT) is the national organization of the working class and middle sectors' unions and federations of unions.
for the short term differences between the fate of the BA State in Argentina in the period 1966-1973 and other comparable experiences. But they, in turn, call for an explanation, and this requires a longer historical perspective.

(I) Historical background
In this section I shall point out certain features of Argentina's incorporation into the world capitalist system which gave rise to the country's peculiarity in comparison with the rest of Latin America. 6 These differences continue to bear upon certain characteristics of Argentine capitalism and class structure and also -centrally for our subject -on the power resources and on the political alliances available to the popular sectors.
The following are the most crucial features for our analysis: (1) As in the rest of Latin America, the pace and characteristics of Argentine capitalist expansion were fundamentally determined by the incorporation of some of its regions as exporters of primary products. This allows us to make an initial distinction between those vast regions of Latin America with no direct linkage 7 to the world market (such as the Andean hacienda) and those which were directly linked to such a market as exporters of primary products. Among these the estancia of the Argentine pampas and Uruguay differed substantially from enclaves and plantations, which were the principal form of incorporation elsewhere in the continent. The main differences were: (1.1) The estancia was less labour-intensive than the plantation and the hacienda; (1.2) It was also less capital-and technology-intensive than the plantation and the enclave; (1.3) Largely because of the latter, the control of the principal productive resource (land) was left, in the Argentine pampas and in Uruguay, in the hands of an early domestic agrarian bourgeoisie, whilst the enclave and the plantation usually were directly owned by international capital, and the hacienda was left in the hands of an oligarchy with hardly any capitalist traits; (1.4) This pattern, combined with a high differential rent, endowed the pampean and the Uruguayan bourgeoisie with an important capital accumulation base of their own; (1.5) This bourgeoisie did not escape dependence on European capital via the transport, finance and international marketing of its products; but its base 6 And of Uruguay, to which I shall briefly return. 7 When I speak of direct incorporation or linkage I refer to the role which some regions had as an (exporting) part of the world capitalist system. This of course does not imply that regions not linked to the world capitalist system were not importers of products from the centre, nor subject to the effects of world capitalist expansion, many times through directly incorporated regions. of capital accumulation did foment a significantly wealthier and more diversified urban, commercial and incipient industrial sector, than was to be found in those economies which revolved around the hacienda, the enclave and the plantation. These characteristics are well known 8 but others, to which less attention has been given, stem from them.
(2) The cereal-, wool-and, later, also beef-exporting economy covered a relatively larger portion of the national territory than the exporting sectors of other countries. Above all, the areas not directly incorporated to the world market, in Argentina, carried much less economic and demographic weight than in the rest of Latin America. Furthermore, in Argentina and Uruguay there was only a very small peasantry subject to precapitalist relations of production such as was the case in much of the continent. The insertion of a much larger proportion of the population in the export economy meant that, from the end of the nineteenth century, Argentina exhibited a significantly greater homogeneity than the rest of Latin America 9 which, in spite of later mishaps, continues to be noticeable. 1 0 (3) Besides the sizeable base of local accumulation due to direct control of 8 Above all since Fernando H. Cardoso's and Enzo Faletto's book, Dependencia y Desarrollo en America Latina, where we find the characterization of the types of exporting economy I have mentioned; an important recent contribution is Albert Hirschman's, who adapts elements of staple theory to his concept of "linkages", widened to include not strictly economic relationships, and from there explores the consequences of the type of export product through which incorporation into the international market took place; unfortunatetly this author does not deal with pampean and Uruguayan products. See Albert Hirschman (1976). 9 With the exception of Uruguay, a case of even greater intranational homogeneity, since practically all its territory and its population were incorporated to the world market in conditions similar to those of the Argentine pampean region. Another exception, though partial and more complicaetd, is that of Chile; here, in the last third of the nineteenth century, the highly homogeneous agrarian economy of the central valley, partially oriented towards the export of foodstuffs, underwent (in contrast with Argentina and Uruguay) a decline, and the mining enclaves of the North emerged. But in contrast with other cases, those enclaves were inserted in a national market and a national State already constituted around the central agrarian region. See Marcelo Cavarozzi (forthcoming). Uruguay's greater intranational homogeneity allowed the earlier and fuller development of a "liberal" and "welfare" State. But for this very reason the problems concealed by the initial bonanza exploded earlier than in Argentina. Besides, the lesser absolute size of the Uruguayan internal market was decisive in interrupting its industrialization much earlier and thus, in recent decades, the relative weight of its working class has been significantly smaller than in Argentina. land, the high productivity of land in international terms until approximately 1930 11 and the small labour requirements of "extensive" farming contributed decisively to the greater diversification and prosperity of the pampas region and its urban centres -compared to the regions dominated by the enclave, the plantations and the hacienda. Suffice it to say that wages in the pampas region and the Argentine urban centres, until approximately the Second World War, were higher than in a sizeable number of European countries,12 whilst those of the rest of Latin America -if and when wage relationships were established -were much lower. Thus, not only was intranational homogeneity higher, but also the Argentine region directly incorporated to the world economy was more diversified and generated a significantly higher income for its popular sector than in the other Latin American countries. This, in turn, has other consequences: (3.1) Around the beginnings of the twentieth century the existence of a fully capitalist and relatively wealthy urban (and, largely, also pampean) consumer market, induced an industrialization which received further stimulus from the import restrictions resulting from the First World War. Argentine industrialization did not "begin" with the world crisis of 1930; it began earlier and proceeded faster than in the rest of Latin America;13 (3.2) An early working class therefore also emerged which developed organizational patterns autonomous both of the State and of the incipient industrial bourgeoisie although it only later entered the political arena. 14 In the absence of a large peasantry providing cheap labour, the strong demand for labour could only favour such an outcome. Because of the specific characteristics of Argentina's insertion in the world capitalist system, its economic growth was powered fundamentally by its civil society and its relationships with the international market. The dynamizing impulse did not depend on the State, as in general tended to happen -and with many difficulties in the other Latin American economies. This point must be developed m greater detail.
In the period between roughly 1870 and 1930 15 the Argentine State had certain features in common with the liberal State of the great world centres: although a more ostensibly fraudulent political democracy, the level of electoral participation was not 10wer,16 and the state machine did not go beyond providing crucial, but limited, general conditions for the functioning of the economy. 1 7 But this State was the creation of the pampas bourgeoisie and its financial and \ commercial appendages, by means of a process which also entailed the making of that bourgeoisie, and of the system it dominated, in a marginal yet integral corner of the world capitalist market. To clarify this statement we must resort to some comparisons.
The pampas bourgeoisie and its urban branches directly constituted a national State, not the regional State 18 which was the main political power base of the dominant classes in so many Latin American countries. The Argentine national State also eliminated -earlier, and with greater ease and completeness -the autonomy of the regions not directly linked to the world market, largely because those regions weighed much less on the country as a whole than in most other Latin American cases. 19 This implied that to a much smaller extent than in 15 That is, between the great exogenous impulse of the incorporation of the pampas region to the international market and the world crisis which altered the main basis of the system. 16 Cf. AtHio Boron, 1972, Ph.D. thesis in progress, Harvard University, Department of Government. 17 Above all the transport and warehouse-network necessary for the embarcation of the pampas' production, the capture of which by international capital was generously subsidized by the State. If the small technology requirements of direct exploitation of the pampean region permitted domestic control of the land, the much higher requirements of such a network (and later on, of the meatpacking industry) determined a high and early inflow of international capital. 18 I am not concerned here with the details of the respective historical processes. In particular the imposition of the nationalization of Buenos Aires by a coalition of provinces of the interior against the opposition of a sizeable part of the pampas interests was no obstacle, once the vigorous exogenous impulses of the European demand for foodstuffs were felt, to the processes alluded to in the text. 19 Even in a case such as Brazil, characterized by relatively early industrialization and by the great weight of State bureaucratic apparatus inherited from the Imperial period, the subordination of the dominant classes of the North East and the elimination of the barriers interposed by the regional States to the effective functioning of a national market was only completed after 1930; d. Centro Brasileiro de Analise e Planejamento (CEBRAP), "Estado y Sociedad en el Brasil: la planificaci6n regional en la epoca del SU-DENE", Sao Paulo, 1976. It should be borne in mind that I am excluding Chile and Uruguay from these generalizations. the rest of Latin America, the State was an expression of changing powerrelationships between some regions directly incorporated to the world market and others marginated from it.
Thus the pampas bourgeoisie and its urban tentacles held both a central economic position and, through the national State, they constituted a central political position as an internally dominant class burdened by other regions. Furthermore, the shifts in the relative importance of export products took place within the pampean zone and its bourgeoisie 20 and not, as in most other cases, by means of the incorporation of new products from new regions leading to shifting alliances with preexisting locally dominant classes and established segments of international capital.
Nevertheless the pampas bourgeoisie and the national State became the principal channels of the internationalization of both society and economy because of the nature of their insertion in the world market. The "liberal" characteristics of the Argentine State and the strong relative weight of its civil society can only be understood as consequences of the position of the State at the intersection point of the pampas bourgeoisie with international capital -which had deeply penetrated the economy through its control of the financing, transportation and the external marketing of pampas· production. Paradoxically, therefore, this original internationalization of an economically dynamic and internally homogeneous region which included a decisive part of a country with barely any peasants through the local retention of capital accumulation shares enabled a highly internationalized State to become so devastatingly national with respect to the regions marginalized from the pampean system. In contrast, the Andean oligarchy or that of Brazil's North-East could directly and diaphanously control "their" regional State apparatus, while international capital based on enclaves and plantations controlled a. State which was less an emanation from than a graft imposed upon a society which lacked a local bourgeoisie endowed with its own accumulation base. Instead, in Argentina, the existence of such a bourgeoisie arising from the very process of incorporation of the country into the international market generated a situation in which the regional States were of little weight; furthermore, the national State was one of the crucial channels of a rapid and early internationalization, which, due to the weight of the pampas economy, encompassed much more of the country than in other Latin American cases. That is why -not in spite of, but as a very condition of its centrality -the relationship between the pampas bourgeoisie and the State did not exhibit the transparency and immediacy which the regional oligarchies and international capital imposed in a large portion of Latin America's regional and (for a long time, mostly nominal) national States. 21 Although the Argentine liberal State did not survive the crisis of the thirties, the factors summarized above allowed her to recover from the economic impacts of the world crisis with greater speed and ease than most other Latin American countries. The crisis induced a new wave of industrialization through import substitution (helped by a comparatively broad internal market 22 ) and provoked the absorption of a large part of the still available work force from the nonpampas regions, thus reducing their relative weight even further. This is not the place to analyze how this affected the emergence of peronismo but to finally enter into the central theme of the paper.

(II) Dilemmas
I have already mentioned the emergence in Argentina of a popular sector, which included a politically significant working class, endowed with larger economic and organizational. resources than those of the rest of Latin America 23 -which in turn resulted from the combination of larger available economic surpluses and from the negligible pressure exerted on the urban labour-market by an almost non-existent peasantry. If this was an advantage for Argentina's capitalist development, the other side of the coin was that it strengthened its popular sector and that, when the bonanza disappeared and the economic conditions approached zero-sum, there was no sizeable peasantry which could bear a substantial part of the costs of agreements negotiated among the classes located within the fully capitalist region.
The second point to be singled out arises from another peculiarity of this economy: its main export products -cereals and beef -are wage goods, foodstuffs which constitute the main consumption item of the popular sector.

21
Of course, if instead of making these comparisons with other Latin American cases we had made them with Australia and New Zealand, the dimensions Argentina and Uruguay had in common with the other Latin American countries would be more noticeable.
For some comparisons in that direction, see Lucio Geller, op. cit. and Hector Dieguez (1969). 22 The effective market is not so much a function of the total population as of that part of the population subject to capitalist relationships and endowed with a monetary income sufficient for the purchase of mass consumption industrial goods; d. Guillermo O'Donnell, 1972, chapter 1. 23 As always with the exception of Uruguay and the partial -and too complicated to be dealt with here -of Chile.
Let us initially note some general consequences of this peculitarity. Other Latin American primary export products have less influence on the consumption of the popular sector and therefore, on the relative prices of their consumption baskets. Furthermore, the manner in which their price changes influence popular consumption is, in most cases, indirect, generated by mechanisms which are difficult to apprehend; in contrast, a change in the relative prices of foodstuffs is immediately perceivable. In addition, this perception arises in a popular sector with a significantly greater income level (and, presumably, expectations) and organizational autonomy (and therefore greater capacity of resistance) than in the other Latin American cases. We are now in a position to analyze more concrete processes.
The world-crisis of the thirties exogenously depressed the prices of pampean goods. Then the peronista government (1946)(1947)(1948)(1949)(1950)(1951)(1952)(1953)(1954)(1955) offered a foretaste of problems which would explode later. First, (1946)(1947)(1948)(1949)(1950) the State appropriated a substantial part of the proceeds of pampean exports, kept internal foodstuff prices depressed, and thus increased the income of the popular sector and provoked an expanding demand for other goods, especially industrial ones. But this was to generate a balance of payments squeeze, due to the "discouragement" effect of low prices on pampas production and to increasing internal consumption of exportable foodstuffs. Subsequently (1952)(1953)(1954)(1955) agricultural prices improved, whereupon -because of the operation of the inverse joint effect -the balance of payments situation improved. But this in turn generated political troubles due to the regressive redistribution of income it entailed and to the reduction of the domestic demand on which the urban bourgeoisie depended.
Following this, around 1960, a wave of direct foreign investments in industry and services provoked a speedy internationalization of the urban productive structure 2 4 (by means of capitals and activities different from those involved in export activities). Contrary to the "developmentalist" hopes, this new stage resulted in a marked increase in demand for imports, more rapid than the growth-rate of GNP, of exports and of pampas production. 2 I) Faced with this situation the only economically "evident" solution -repeatedly expoundedlaid in a large increase of exports which would have provided the urban productive structure with the imports necessary for "self sustained development". Assuming the capitalist parameters of the situation, this solution entailed, fundamentally, an increase in pampean production (and productivity) and/or a reduction of real wages so as to "free" exportable surpluses of food. But the Cartesian 24 It is impossible to quote herein all the pertinent bibliography. The data and main sources can be found in Pablo Gerchunoff and Juan Llach (1975), and Juan Sourrouille (1976). 25 See, above all, Juan Ayza, Gerard Fichet and Norberto Gonzillez (1976). simplicity of these solutions -which were indeed attempted -ran into political complications which we must now analyze.

(III) Cycles
Several consequences arose from the fact that wage goods were also the main export products. In the first place, it offered an objective basis, which was, besides, sub j ectivel y acknowledged, 2 6 for repeated alliances between a sizeable part of the weaker fractions of the urban bourgeoisie and the popular sector. These alliances were forged around the defense of the internal market against the recessive effects (via the increase of domestic food prices) of every important rise in the price of pampas products. In the second place, the mobilizations of the popular sector in defense of its consumption levels reinforced its capacity for organizational and political action through partial but repeated victories. A third consequence was that the above mentioned alliance again and again provoked and revivified a d~ep horizontal cleavage within the urban bourgeoisie, between its oligopolistic fractions and those weaker ones which found a welcome ally in the popular sectors. Fourth, these same processes determined the appearance of another fundamental intrabourgeois cleavage, by separating the economic interest and political goals of the urban bourgeoisie (including its oligopolistic fractions) from those of the pampean bourgeoisie. These changing alliances lie at the origin of the economic and political cycles on which students of Argentina have fastened their attention. 2 7 The solution of Argentina's balance of payments bottlenecks requires a substantial increase of pampean exports. However, when, around 1960, the demand for imports rose rapidly, those exports did so to a much lesser degree. This was partly the consequence of the increase of internal consumption of exportables and, also, of the slow improvements in the pampean region's production and pro-GRAPH I. 1956-1976. (Wholesale cattle and cereal prices against wholesale non-agricultural prices.) .....

~
-< ~ ductivity.28 Neither the available space nor my knowledge allow for a satisfactory explanation of this failure, but it seems obvious that, assuming the capitalist parameters of the context, a necessary (but not sufficient) condition for rising production of pampean goods lies in "satisfactory" prices for the pampean bourgeoisie. The meaning of "satisfactory" is complex, but it includes at least two further necessary conditions: one its that prices should make feasible the investments necessary to increase the capital density of the pampean region, and with it, its productivity. The second conditionless obvious but more important -is that those prices should be stable and that they should be perceived as such at the microeconomic level. I do not know of studies which exactly determine this, but nothing leads us to suppose that, in the 1956-1976 period, the profit rates of the pampas bourgeoisie were lower than that of the urban bourgeoisie. But however, Graph I clearly shows the vast instability of the main pampean wholesale prices (cereals, linseed and beef), measured in relation to wholesale urban prices. A substantial increase in pampean production (and exports) cannot take place without converting the estancias into a much more capital intensive agribusiness. Discarding explanations based on the economic "irrationality" of the pampas bourgeoisie -which are nothing but an implicit avowal of their authors' ignorance -it seems clear that the answer must be found at the level of the parameters which govern micro economic decisions. Those parameters do not spring from some economic "necessity" but from political struggles and from the pendular movements of the State, which in turn result from the specificities of a class structure whose origins I have summarized in the previous Sections. This is what we must analyze.
The conversion of the pampean estancia into a capital and technology intensive agribusiness 29 entails making investment decisions regarding a rather lengthy time horizon. The instability of pampean relative prices, the historical memory of it and, above all, the difficulty of making correct forecasts of future price instability30 have prevented those decisions. Thus, the once dynamic pampas bourgeoisie (even in international terms, during the period prior to 1930) has become less and less so as we approach the present times. And this, fundamentally, 28 On this subject the principal source is Carlos Diaz Alejandro's important book (1970) where the slow growth of the physical quantum of these exports and the spectacular lag of pampean productivity with respect to its principal competitors in the world market can be noticed. Also see Richard Mallon and Juan Sourrouille,op. cit. 29 I hope it is clear that I am speaking at the class level. That is, the change towards an agribusiness would surely displace more than a few individuals who at present constitute the pampas bourgeoisie. 30 The pampas bourgeoisie's demands and declarations of, at least, the last twenty years, are a repeated complaint because it does not receive profitable nor stable prices. because relative prices made it micro economically rational to maintain the "extensive" exploitation of the land. 31 In the short term the increase of relative internal prices of the pampean production entails an almost equivalent net loss for the whole of the urban sector. The income redistribution and the recessive effect on the urban economy which -ceteris paribus _32 this entails, enlarges the export surpluses (via their immediate effect on the internal consumption of exportables) and might induce a medium term increase of pampas production by satisfying the necessary condition of "satisfactory" prices. This would not be too onerous for the oligopolistic fractions of the urban bourgeoisie. They have an objective interest in increasing the balance of payments surplus because of their high import coefficient. 33 The recessions and redistributions of income which usually accompany the increases in food prices are less harmful for these oligopolistic fractions than for the weaker ones. In effect, the former have economic resources and preferential access to internal credit 34 which enable them to shoulder the burden of the recession and, indeed, to further capital concentration in their benefit at the same time. 3 5 Besides, the urban bourgeoisie's oligopolistic fractions address a greater part of their production and supply of services to the relatively higher income strata whose income is less affected, absolutely and proportionately, by rises in food prices. Although this generalization would require qualification in a more refined analysis, it provides the objective basis for a long-term alliance between the large urban bourgeoisie and the pampas bourgeoisie, which could guide the "modernization" of Argentine capitalism simultaneously through 31 Cf. the microeconomic studies quoted in the works I mention below. The issue is however more complicated ,as arises from the controversy which took place in Desarrollo Economico between Guillermo Flischman (1970 and and Oscar Braun (1970 and and Juan Carlos Martinez et at. (1976). Of these important exchanges the central point for our analysis is that the differential rent which the pampean region still enjoys and, especially, the great fluctuations of the whole economy and the high (and erratic) inflation rate, determined that the purchase of pampas land should be an excellent speculative investment -and a defense against the effects of inflation -for the urban and agrarian capital surpluses. This concurs to reinforce the microeconomic rationality of maintaining the region's extensive exploitation. But, and this is what might be added to those authors from the perspective in which our analysis places itself, the subject they discuss seems analytically a consequence (although along time it renourishes them) of the economic and political factors I analyze here.
. 32 Further on I shall complicate this matter introducing other factors. 33 Not only is the coefficient high but it grows with an elasticity greater than 1.0 with increases in its production level; d. Juan Ayza et at.,op. cit. 34 For data on this point cf. esp. Fundacion de Investigaciones Economicas de America Latina (FIEL) , 1971, andMario Brodersohn, 1972. 35 On this point and others closely connected with it, Guillermo O'Donnell and Delfina Link (1973). capital concentration in the urban sector and the development of capital-intensive agribusiness in the countryside. However, at least until 1976, this alliance only lasted for short periods quickly to dissolve in situations which recurrently placed those two superior fractions of the Argentine bourgeoisie in politically divergent camps. Why this deviation from economic "logic"? Fundamentally because this alliance has been faced again and again by another -basically constituted by the popular sectors and the weak fractions of the urban bourgeoisie -which, in spite of its economic subordination, has been able to prevent it from holding together beyond the short term. In the Latin American context this has been one of Argentina's (and, with its own characteristics, Uruguay's) originalities, which can only be understood in terms of the historical perspectives summarized in the previous sections. But we still have to introduce new elements to our analysis.
Which processes posed those dilemmas and conflicts? The periods of low internal prices of foodstuffs and stable foreign exchange rates have been, and not by chance, those of highest growth rates and -till the approach of the end ofrhe cycle -of lowest inflation rates. 3 6 But they have also led to balance of payments crises which generated the implantation of controls, above all on internal prices and on foreign exchange movements, that were not enough to stave off those crises. Once the balance of payments crisis was unleashed, it was dealt with by means of devaluations which (with the exception of the 1967-69 period) implied a correlative increase of the internal price of exportables. These devaluations formed part of stabilization programmes which deepened the recessive and redistributionary effects of the devaluation by means of restriction of money supply, reduction of the budgetary deficit, wage freezes and increases of the real interest rate, tending, on the one hand, to consolidate the transfer of income to the exporting sector and, on the other hand, to adjust the internal level of economic activity to the balance of payments restrictions.
The impacts were not only recessive and redistributive but also inflationary ("stagflation" is no novelty in Argentina), through the rise of domestic food prices caused by the growth of their export value, and the rising price of imported goods and of credit -at times which, on the other hand, wages and salaries were kept frozen or systematically lagging and recession increased unemployment. In the short term (and, as we shall see, in these processes there was never more than the short term), the income transfer towards the exporting sector did not induce an increase of pampean production 37 but the stabilization 36 Cf. the pertinent data in Brodersohn, 1974. 37 Actually, the price elasticity of the pampean production is nil or slightly negative in the short term. This is because for cattle "an increase in relative prices reduces supply and increases the stocks. Besides, an increase in the cattle stock implies a greater use of land due to the rigidity in the supply of land... Therefore, an increase in the relative programmes were instrumental in easing the balance of payments squeeze. True, such successes were due to very different factors from those proclaimed in official speeches, in the "recommendations" of the International Monetary Fund and in the exultant statements of the organizations of the pampas bourgeoisie. They were achieved not by an increase of exportable production hut by recession, which diminished the demand for imports and increased the exportable surplus, especially of foodstuffs. But this generated resistances among the many penalized by these policies, at the same time that the resulting easing of the balance of payments made economic reactivation policies possible. Consequently, the liquidity increase, the relaxing of controls on the fiscal deficit, the availability of foreign exchange, the growth in employment and the salary increases ended the downward phase of the cycle and inaugurated the upward phase. But the latter led into a new balance of payments crisis, 3 8 after which further devaluation, and the consequent stabilization programme, opened up another downward phase ... 39

(IV) Pendulums
In each phase of the cycle the large bourgeoisie has played on the winner's side. I have already pointed out that the recessions provoked by stabilization programmes have, at the very least, not damaged that fraction: at the same time, as a direct appendage of (or intimately linked to) international capital, it is the large bourgeoisie which better perceives -and fears most -the costs of international insolvency. 4 0 This fraction has the most direct interest in an prices for beef also negatively affects the production of cereals since to the lesser supply of beef is to be added the smaller area for cultivation" (Mario Brodersohn, 1974, p. 28). 38 In contrast with what I noted above concerning exports, the income-elasticity of imports is extremely high. It was estimated in around 2.6 for the 1947-1967 period (cf. Carlos Diaz Alejandro, 1966, p. 356); for the period after 1966, Juan Ayza et at., op.cit., p. 13, with a different methodology, estimate an elasticity of 1.8. One piece of information which indicates how through internal consumption this pincer movement closes on the balance of payments in the upward phase of the cycle is that the wage earners' marginal propensity to consume exportable goods (foodsmffs, drinks and tobacco) is of .36 and that of the nonwage earners is .16; Carlos Diaz Alejandro (1974) Cap. IV. 39 This is the briefest of summaries of the principal theme of the works quoted in footnote, to which I must refer. A useful presentation of the economic mechanisms operating in the upwards and downwards phases of the cycles -which unifortunately I came to know with this work substantially finished -is Marcelo Diamand's (1976). improvement of the balance of payments. 41 Furthermore, the free international movement of capital enhances the privileged position in an ever more narrow domestic credit market, of this most internationalized (and therefore internally dominant) fraction, while at the same time reopening the "normal" channels for the transfer of capital accumulation towards the centre of the system, 4 2 of which it is a more intrinsic part than any other. In the final stretch of the upward phase of the cycle, these factors turn the large bourgeoisie into an ally of the pampas bourgeoisie (and of the whole of the exporting sector) in the clamour for the devaluation and deflationary policies which launch the downward phase. Thus, faced with the unleashing of the balance of payments crisis, the large bourgeoisie swings towards the objective interests of the pampas bourgeoisie, favouring and supporting stabilization programmes which transfer a mass of resources toward the latter, mostly at the expense of the urban sector.
But the regressive and recessionary impacts of those measures generated a reaction among the weaker fractions of the urban bourgeoisie and of the popular sector 43 at the same time as the improvement in the foreign exchange position made feasible the economic reactivation measures for which they were clamouring. Faced with this, the large bourgeoisie again and again did what all bourgeoisies do in the absence of a tutelary State which induces them to adopt longer-term strategies: they looked to their short-term economic interests, supported the economic reactivation policies, and thus rode the crest of the wave of economic recovery -from which, we may safely assume, they were able to cnsls cropped up, the imposition of foreign exchange controls and of reStrictIOns to capital transfers abroad became serious hindrances, above all, to the firms more closely connected with the centres of world capitalism. Truly, none of these controls attained their goals, nor did they prevent massive flights of capital, but many of the high ranking staff of large firms (national and transnational ones) which I interviewed in 1971 and 1972 said that for that effect they "had" to act "excessively" beyond the pale of Argentine legislation, with the consequent uneasiness at times when, during the upward phase of the cycles, "demagogues" and "nationalists" with access to State institutions were not lacking.
. 41 In terms of their high import coefficient and demand for foreign exchange, and in spite of their better access to international finance, which allows them to make excellent deals in pre and postdevaluation periods of acute scarcity of foreign exchange. 42 Even within private capital's oligopolistic fraction, the more fully and directly internationalized firmsthe subsidiaries of the transnational corporations -usually are those largest in size (in capital and sales), with the greater growth rate, and the most capital intensive; cf. especially Juan Sourrouille, op. cit. Of course this is not peculiar to Argentina, on Mexico see Fernando Fajnzylber and Trinidad Martinez Farrago (1976) and Carlos Von Dellinger and Leonardo Cava1canti (1975). profit in a privileged manner.44 In so doing this fraction covered a full swing of the pendulum, joining the whole of the urban sector and abandoning the pampas bourgeoisie to a solitary lament because of the deterioration of its relative prices; 4 5 all of which produced the great fluctuations of relative prices observable in Graph 1.
Although this describes the recurrent pendulation of the large bourgeoisie, I have still to explain it. But it must be added that, besides their economic consequences, these displacements had political implications of the greatest importance: they repeatedly broke up that intra-bourgeois cohesion essential for its stahle political domination. More precisely, they fractured the cohesion of its two superior fractions (the urban oligopolistic and the pampas bourgeoisie), whose respective capital accumulation base made them potentially· capable of "modernizing" Argentine capitalism. Another aspect, no less important and which I shall soon deal with, is that those swings not only generated the political space for, but also were to a large extent the consequence of, an alternative alliance which encompassed the weaker fractions of the bourgeoisie and the popular sector.
Let me insist on a crucial point. The alliance of the superior fractions of the bourgeoisie could have borne fruit if it had lasted long enough to induce significant productivity increases in the pampas region. This temporal requisite was violated by the large fluctuations in relative prices. But in their political demands, the pampean bourgeoisie concentrated on the level and not the stability of their prices, thus contributing to the pendulations I have already mentioned. The said productivity increases could have taken place with relatively depressed but stable pampean prices (thus meeting the necessary condition of stability stated above), combined with public policies which would have forced it through by more structural measures. Herein lies the import of the diverse projects designed to tax the difference between the potential and the actual productivity of pampas land. Such an alternative, obviously conflictive with the short-term interests of the pampas bourgeoisie in its present composition, is not against those of the urban sector as a whole (since it does not presuppose a fall in their relative prices), and in the medium-term it could have attained the sought~after increase in pampas production and productivity. However, the attempts to implant such tax on the "potential rent of the land" repeatedly failed. This must be contrasted with what has happened in many other Latin American countries, where the State -impelled by and allied with the large bourgeoisie -has usually been able to impose the "modernization" of agrarian regions and their dominant classes. 4 6 But those agrarian classes were fundamentally regional ones 47 and, although their production might temporarily fall, their contribution to total exports was not comparable to that of the pampas bourgeoisie. That is why other Latin American States have been able to subordinate those classes, and the regional States which they controlled, without simultaneously worsening their balance of payments problems. The case of the pampas bourgeoisie has been very different. I have pointed out its early condition as a national class, even with respect to its linkage with a national State. This meant that the intrabourgeois struggles usually occurred, in contrast to other Latin American cases, in the very heart of a national State which was continuously fractured by them.
Besides, the "discouragement" of the pampas bourgeoisie 48 because of the fall in its prices and the attempts to "restructure it" by means of tax mechanisms, had strong immediate repercussions on the balance of payments -at the same time that, in the upward phase of the economic cycle, the increase in domestic consumption of exportables further diminished the potentially available exports before pampean productivity had undergone any substantial improvement. Thus a balance of payments crisis ensued, whose alleviation by means of devaluations not only turned relative prices against the urban sector but also entailed the expulsion from the governing alliance of the sectors which had impelled the reactivation of the cycle. As long as the stabilization programmes lasted, the immediate interests of the pampas bourgeoisie weighed heavily in the institutional system of the State. And, of course, it opposed any prospect of its own "restructuring", centering the issue on a sharp increase of its prices and thus creating the conditions for a renewal of the cycle... In other words, although 46 Which of course did not prevent these processes from being acutely conflictive. The point is that the capacity of these classes to resist was less than that of the pampas bourgeoisie and that, besides, the cost of such policies -in terms of their short-term impact on the level of internal economic activity and exportswas lower. 47 In the case of the enclaves it obviously was not a matter of modernizing the economy's most capital and technology intensive sector, but of renegotiating with international capital the percentges which could be retained locally. In the cases in which (1) "excessive" pressures were made (reaching or threatening nationalization, above all) and (2) the enclave's product was as important as the pampas production for total exports, (Bolivia and, more recently, Chile) the familiar falls in production and/or prices equivalent in this sphere to the pampas bourgeoisie's recurrent "discouragements unleashed the consequent balance of payments crisis. 48 For the pampas bourgeoisie's insistence on its "discouragement" because of its internal prices and the attempts to "smother it" with taxes, suffice it to consult any of the Sociedad Rural Argentina (SRA)'s and of the Coordinadora de Asociaciones Rurales de Buenos Aires y La Pampa (CARBAP),s collection of documents. it has long lost its posmon as the dynamic vanguard of Argentine capitalism, the pampas bourgeoisie, compared with other Latin American agrarian classes, has retained an unusually central economic and political position. This position was sufficient both to block any attempt to "restructure" it and to use periodic balance of payments crises to bring about massive income transfers in its benefit. Meanwhile, and as a consequence of this, the capital accumulation channels in Argentina entered into recurring short circuits and the State danced to the pendular tune of civil society.
This accounts for some of the characteristics of the period beginning in 1966, especially with the economic policies followed between March 1967 and May 1969. The Minister of Economy, Krieger Vasena, transparently carried out the policies of the large bourgeoisie. This entailed, among other things, a large devaluation that for the first time was not to benefit the pampas and exporting sector. On the contrary, the March 1967 devaluation (40 %) was wholly appropriated by the State, which withheld a percen~age of the value of pampas exports equivalent to the devaluation. This fiscal revenue was used in a substantial programme of investments in physical infrastructure and communications. A fixed peso price of pampean production depressed the internal price of pampean foodstuffs, as can be seen in Graph I. It also allowed a rapid reduction of inflation and -in contrast with other cases of bureaucratic-authoritarianismonly a moderate fall in industrial wages. (See Graphs II and III.) But not even then could this situation be maintained and, as can be seen in Graph I, since 1970 pampean prices (especially those of beef) rebounded until they reached a very high level in 1971-1972. Krieger Vasena's was the only clear and sustained attempt by the large bourgeoisie to unilaterally subordinate the pampas bourgeoisie 49 to its own accumulation. But the result was a rupture of the cohesion of the BA State from within and a political and economic collapse impelled from without by other social actors. While this attempt marked the limits of an unilateral enforcement of supremacy by the large bourgeoisie, the history of previous devaluations, by pushing the big bourgeoisie into alliances with the urban sector, had shown that it was impossible to return to the good old times of pampean supremacy.50  : 1956-1976. 1959 1960 1961 1962 1963 1964 1965 Monthly percentage increase in the cost of living index for the city of Buenos Aires.  The same for metal worker.  1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 , 10

(V) The defensive alliance
If the political and economic centrality of pampas bourgeoisie singles out an important difference with respect to other Latin American countries and their agrarian classes, a no less important difference stems from the greater political vulnerability of the weaker (and genuinely national) fractions of the urban bourgeoisie in those countries when faced with the expansion of the large bourgeoisie. The growth of the dominant productive structure, oligopolistic and internationalized, has occurred at the expense of many fractions of national capital, weakening its position vis-a-vis international capital and the State. This has caused complaints and strains, but has not, so far, been translated into serious political challenges to such "development" patterns. No such thing has happened in Argentina. The reason for the local bourgeoisie's comparatively greater political capacity is not to be found so much in itself as in the characteristics of the popular sector and -another aspect of the same -in the country's relative national homogeneity. Elsewhere, a weaker urban sector, less organized and autonomous, deprives the weaker fractions of the Latin American bourgeoisie of the extremely important ally they had in Argentina. This is a crucial point.
Not only is the Argentine popular sector endowed with greater autonomy and organizational capacity than in most other Latin American countries. It also so happens that the medium and long term alliance of the upper fractions of the bourgeoisie depends on the level and stability of the relative prices of the main internal foodstuffs. This gave the popular sector a precise target for its political action, which has blocked the connection of the accumulation circuits of those upper bourgeois fractions. These are necessary but not sufficient conditions for the recurrent ruptures of the latter's alliances. To account for the specificity of the phenomena we are-dealing with we must also see how the popular sector linked itself with the objective interests and the political action of the weaker fractions of the urban bourgeoisie.
These fractions are usually penalized by devaluations and stabilization programmes. Given an alleviation of the balance of payments squeeze, their immediate interest lies in economic reactivation policies which increase employment, liquidity, and credit availability and once again place the State in an expansionary role. This effect also results directly from wage and salary increases; thus, it is not surprising that this bourgeois fraction in control of the most labour intensive enterprises should support those increases when the even greater costs to them of recession are taken into account. The concurrence with the unions in demanding wage increases is, besides, the token that this bourgeoisie hands over to the popular sector to forge the alliance. 51 Such a bourgeoisie -more or less weak and more or less penalized by the expansion of oligopolistic and internationalized capital -exists in other Latin American countries, but only in Argentina has it found a popular ally whose immediate, short term interests are compatible with its own, and which possesses a significant capacity for political action. 5 2 The main organizational supports of this alliance have been the CGE, the GGT and the national leadership of the main unions. Its first, principal, and possibly last expression has been peronismo. It was not the only one, sinceabove all in the periods in which peronismo was proscribed -it was channelled through other parties and in the interior of the State's institutions by diverse, "nationalist", military and civilian groups. Their banner has been the defense of the internal market, in the sense both of raising the level of its activity and of limiting the expansion of international capital in it.
The characteristics of this popular sector and of this local bourgeoisie can not be understood in isolation from each other. It has been their conjunction in the multiplying effect of their alliance which has made it possible to impose, again and again, the satisfaction of their immediate demands. We can now examine the characteristics and principal consequences of this alliance.
(1) The alliance was sporadic but recurrent. It only appe~red in the downward phases of the cycle, when demands for wage and salary increases and for diverse measures to relieve "the suffocation of the small and medium-sized national enterprise"53 concurred in the reactivation of the internal market at the expense of the pampean exporting sector. Once the cycle reactivated the alliance dissolved, partly due to the attempts of the local bourgeoisie and of the unions 51 Since these wage and salary increases impel economic aCtIVIty at the same time that other policies, allowed by the transitory balance of payments relief, increase the employment level, the orthodox warnings that all this feeds inflation matter little -even more so since inflation, with a fixed or systematically lagging exchange rate, accelerates the reversal of the relative prices in favour of the urban sector. 52 In Uruguay the lower level of industrialization, fundamentally due to the smaller internal market, weakened both actors a lot more; the local bourgeoisie has in itself been weaker and in the popular sector the working class has weighed relatively less. In Chile the political expression of the working class through marxist parties (and the inexistence as in Argentina and Uruguay of a direct target in the relative price of foodstuffs) made this alliance more ambiguous and discontinuous. In the remaining countries of the region the greater weakness of the popular sector -due to a greater intranational heterogeneity deprived the local bourgeoisie of that fundamental ally. 53 There are subjects and terms which recur in the CGE's demands and declarations; d. e.g., its Memorias Anuales.
(2) The alliance was defensive. It arose against the offensives of the upper fractions of the bourgeoisie. Its ideology of "nationalist" and "socially just" development overlooked what it was unable to problematize: the deeply oligopolistic and internationalized structure of the capitalism in which its members were the weakest components. It was defensive because in its triumph it could not create an alternative capital accumulation system, all it attained was the transition from the downward phase to the upward phase, in conditions which were deemed to provoke the repetition of the cycle.
(3) But despite its defensive nature and although its victories signified the completion and not a way out of the cycle, this alliance was quite successfuL It scored repeated victories in annulling the stabilization programmes, limiting the domestic expansion of international capital, and launching economic reactivation policies and new "discouragements" for the pampas bourgeoisie. Thus, it is not surprising that the wage series of Graph II should also show an erratic behaviour; its peaks are the result of victorious struggles which soon led to marked falls in wages. The upward movements of wages were accompanied by growth of the GNP and, in general, by higher profit rates for the industrial bourgeoisie as a whole -although, being also subject to the overall cycle, profits also seem to have undergone marked fluctuations. I) 4 The consequence of these processes may be appreciated in a phenomenon as intrinsically political as it is economic: the inflation which, as can be seen in Graph III, is even more remarkable for its fluctuations than for its generally high leveL More basically, the defensive alliance was victorious because it achieved the rupture of the alliance between the two superior fractions of the bourgeoisie. The large bourgeoisie, when the time came to weigh the immediate benefits from a reactivation of the economy against the political abyss entailed by accompanying the pampas bourgeoisie and the exporting sector when the remainder of civil society had joined forces against them, opted to support a new upward phase. Repeatedly the defensive alliance -politically -broke 'from below'the cohesion of the dominant fractions and -economically -blocked the only alliance which in this capitalism could implant a new capital accumulation system and extract the economy from its cycles.
(4) The alliance was polyclassist, in the sense that it included the popular sector (with a strong working-class component) and various bourgeois fractions. Its repeated successes were based on this conjunction. But, on the other hand, this determined that its orientations were nationalistic 5 5 and capitalist. Its polyclass character, built on the attainment of shared tactical goals, offered a popular base for the demands of the weak bourgeoisie. Thus, this fraction appeared as a "progressive" one which, contrasting with the large bourgeoisie's "efficientist" orientations and the "landholding oligarchy's" archaism, seemed to embody the possibility of a "development with social justice". On the other hand, the popular sector (especially the unions and the working class) gained, through the polyclassist nature of the alliance, access to resources and mass-media which it could not otherwise have had. In particular, the bourgeois respectability which the alliance entailed made more difficult the harsh repression directed at the popular sector elsewhere in Latin America when they have acted in isolation and/or in pursuit of more radical goals. That is why the impact of this alliance stemmed from the multiplying effect of the concurrence of social actors which had their own resource base and which could coincide in very concrete and short-term goals. In other Latin American countries, the absence of these joint conditions has meant that the local bourgeoisie has lacked popular support and the popular sector (weaker, in any case, because of the greater intranational heterogeneity) has not enjoyed the political protection of a bourgeois ally. This, in turn, has made it possible in these cases for the large bourgeoisie to advance overwhelmingly, encountering -of course -protests and conflicts, but not the limits and the pendulations which this peculiar alliance imposed in Argentina.
(5) Locked in capitalist parameters, the principal political channel of the defensive alliance, peronismo, did not transcend those limits. This limitation also arose from the experience of repeated victories and subsequent defeats. The political activation of the popular sector in pursuit of the goals of the defensive alliance, the protection granted by its bourgeois component and the changes in public policies which it attained, led, on the one hand, to a positive reinforcement of that activation and, on the other hand, to the solidifying of the organizational basis -above all the unions -which articulated the popular sector's action. Let us take a closer look at this.
(6) Concerning the historical experience, the alliance was a function of the fresh memory of prior mobilizations which managed to reverse the downward trend of real wages and of economic activity. It was also a function of the low deterrence effect of a repression which tended to cease the moment the State, indicating a shift of the governing alliances, launched a new upward phase of the cycle. This built up the popular sector's capacity for and disposition to political activation, but it also led to an equally repetitive experience of defeat: the periods of low wages and salaries, and of mounting unemployment during 55 Deprived fundamentally from coinciding to defend the domestic market against the internationalized character of the exportrelated activities and of the large bourgeoisie.
which the spokesmen of the defensive alliance were removed from the governing coalition. But, in contrast with the transparent stimulus entailed by rising food prices and falling salaries and wages, the reversal of the cycle took place because of problems (such as balance of payments crises) and through mechanisms (such as devaluations and restrictions in the money supply) more difficult to apprehend in their functioning and impact. The benefits derived by the pampas bourgeoisie and by the exporting sector, and the initial support lent by the large bourgeoisie to each downward reversal of the cycle, fostered the hostility of the popular sector against both fractions and against the internationalization and big business which they embodied. At the same time, the defensive alliance could not abandon its capitalist ideology and goals. Thus, the interpretation of the sequences of such successes and defeats became a mythology of conspiracies of "big interests" which had a magical ability to defeat the "people" and hinder "development". Failure and tension generated in some cases a fascist ideological syndrome and in others a challenging of the capitalist parameters of the situation. But against these centrifugal tendencies a great centripetal force operated: as the CGT, and CGE and peronismo tirelessly repeated, since 1955 they had been prevented from carrying out the kind of "socially just" capitalist development which, "placed on the people's side" and exercising wide control of the State's institutions, the local bourgeoisie and the unions seemed to offer. The feasibility of uniting "the national and the popular" against the "landholding oligarchy" and the "international monopolies", which the short-term coincidences of the defensive alliance seemed to confirm, were expressed in peronismo's unusual appeal and were a decisive element in the great wave which in 1973 returned it to government. A further condition for this was that in the previous period the large bourgeoisie had ignored the limits of its supremacy and had pretended to unilaterally impose it, even on the pampas bourgeoisie. The great social explosions of 1969 and 1970 sealed the defeat of that attempt and, impelled by a great popular activation, forced the political withdrawal of the large bourgeoisie which, in 1973, for the first time lost its place in the governing alliance. Only then could the alternative that the main spokesmen of the defensive alliance claimed they embodied be positively put to the test.
(7) More than cycles we must now speak of spirals, inasmuch as -politically, above all -each swing of the pendulum, with its succession of temporary victories and defeats, sharpened the conflicts from which they derived. The actors were not classes, fractions and organizations which retained unchanged, beyond these" struggles, their "structural" characteristics. Rather, the actors were the political, organizational and ideological expression of classes and fractions created and transformed during and through this pattern of alliances and oppositions. In particular, the popular sector found in the unions and -politi-cally -in peronismo, an organizational, ideological, and political expression which closely corresponded to the limits of the situation. The mobilization behind the defensive alliance's demands, with its precise aims and polyclass framework, obtained frequent and spectacular victories. This explains the peculiar combination of impressive popular activation with an economicism of demands which emphasizedas a token of its alliance with the local bourgeoisieits rejection of any leap beyond capitalism. Precisely this militant economicism, merged with 'the weaker fractions of the bourgeoisie, permitted repeated defensive victories and perpetuated the illusion of an alternative path of capitalist development.
On the other hand, the moments of political victory and reversal at any point in the economic cycle were those in which the temporary victors took the State apparatus by storm, seeking to strengthen institutional positions from which they would fight future struggles when the situation once again turned roundas experience taught them it would. Of course, the unions were no exception to this: the history of the defensive alliance is also that of the extraction from the State of important institutional advantages. These, in turn, reinforced the possibility of renewing the mobilization of the popular sector. The conquest of institutional positions enabled the unions to enclose the popular sector in a dense organizational net, from which they could canalize it again and again toward a militant economicism, towards the polyclass alliance and towards the mirage of the "other" capitalist path which peronismo announced.
(8) These multiplying fusions of the defensive alliance impelled the large bourgeoisie to recurrently abandon the pampas bourgeoisie to a solitary lament because of the falling prices of their products. Such fusions both impelled economic reactivation and opened the political abyss of a wide and active "national and popular" mobilization which, somehow, had to be reabsorbed. By swinging from support of the pampean bourgeoisie to support of a new upward phase of the cycle, the large bourgeoisie optimized its short term economic interests and managed to remain the only stable member of the governing alliance. It did not lose its dominant position, but the peculiar conditions outlined implied that its domination should continuously shift in that pendular motion. At the same time and for the same reasons, the capital accumulation channels entered into repeated shortcircuits. These clues enable us to understand Argentine politics as a less surrealistic phenomenon than its "political instability" and erratic "development" might lead one to believe.
As I hope is clear, if in all this I am speaking of the constitution of the classes, I am also speaking of the State. It is from this perspective -starting from and returning to civil society -that the problem of the State must be approached.

(VI) Sta,te
The State is not merely a set of institutions. It also includes -fundamentally -the network of "political" domination relationships activated and backed by such institutions in a territorially defined society, which supports and contributes to the reproduction of a society's class organization. In the Argentine case the pendular movements of the large bourgeoisie and the difficulties it has faced in subordinating civil society as a whole are a tangible indication of a continued crisis of the State as a system of political domination. So are the defensive alliance's recurrent and partially victorious fusions. Out of this was born a democratization by default, which resulted from the difficulties in imposing the authoritarian "solution" that seemed to offer a chance of extracting Argentine capitalism from its political and economic spirals.
By "governing alliance" I mean an alliance which imposes, through the institutional system of the State, policies conforming to the orientations and demands of its components. The large bourgeoisie was the stable member of the governing alliance, but each phase was marked by a temporary change of its partners and by an alternation of scarcely consistent capital accumulation circuits. That is why public policies were continuously changing and hardly ever implemented, as the State danced to the tune of the dynamics of civil society.
The State was recurrently razed to the ground by civil society's changing coalitions. At the institutional level, the coalitions were like great tides which for one moment covered everything and when they ebbed, washed away entire segments of the State -segments which would later serve as bastions to piece together a new offensive against the coalition which had just forced the opponents to retreat. The result was a State appartus extensively colonized by civil society. The upper fractions of the bourgeoisie were not the only ones to hang on to it; its weakest fractions and a part of the subordinated classes did the sameanother fundamental difference with respect to other Latin American cases. Civil society's struggles were internalized in the State's institutional system in a way which expressed not only the weight of the bourgeoisie's upper fractions but also the peculiar characteristics of a defensive alliance endowed with a remarkable capacity for partial victory. As a consequence, this colonized State was extraordinarily fragmented, reproducing in its institutions the complex and fast changing relationships of dominant and subordinated classes -and these classes could use those institutions to fuel the spiralling movements of civil society.
Such a State could not "keep at a distance" from the governing alliance's immediate demands and interests, it could only reinforce the cycles and pendulums. It was, quite clearly, a weak State. As a support of social domination, because of the recurrent (and increasing) loosening of such domination implied by the popular sector's mobilizations and the unions' bargaining power. As an institutional sphere, because of its deep colonization and fractionalization. This meant that one possible way out of the cycles -shift towards some sort of State capitalism -was blocked; the fairly stable and consolidated bureucratic apparatus, with non-negligible degrees of freedom vis-a-vis civil society, which would have been a necessary condition for such a solution, was not available. Another great obstacle arose from the fact that in the periods when the large bourgeoisie was in alliance with the pampas bourgeoisie, the stabilization programmes entailed an "antistatist" offensive. This offensive not only aimed at slashing the fiscal deficit but also at dismantling the advances which had taken place in a statist direction during the previous phase, when the defensive alliance had been part of the governing alliance. Those attempts blocked any trend towards State capitalism, by dismantling the institutions which could have impelled it and by dismissing the "technicians" which could have carried it out, substituting them with others who would issue a string of "antistatist" declamations and decisions. In addition, any movement towards State capitalism by the defensive alliance encountered the ambivalence (and, frequently, the opposition) of the governing alliance's permanent member -the large bourgeoisie.
Feasible or not, this possibility was blocked ab initio by the dynamic of civil society. 56 I can be said, then, that in all its levels the Argentine State of the 1956--1976 period was a case of extremely limited autonomy. Its peculiarity was not only that it moved fundamentally in time with the upper fractions of the bourgeoisie but, also, that it expressed the fluctuating political weight of the subordinate classes in their alliance with the weaker fractions of the dominant classes. The alliance's limit -which shows that this must not be mistaken for an equilibrium of forces -arose from the fact that, on the one hand, it had to share the governing alliance with the large bourgeoisie and, on the other hand, that it could only be a defensive alliance.
Could this defensive alliance be a governing alliance on its own, excluding the large bourgeoisie (and, of course, the pampas bourgeoisie)? Only a crude mechanicism could lead us to believe this impossible on the grounds that the defensive alliance contained Argentine capitalism's weakest and least capitalist 56 Even within possibilities which do not pressuppose a change of the capitalist parameters of the simation, tax policies might have cushioned the cycles to an extent which would have modified many of the political aspects we have analyzed. But the ability to extract and reallocate resources by means of fiscal instruments (not only taxes on pampean land) also pressupposes the medium-term stability of those instruments and their implementation and a fairly consolidated bureaucracy which can "ignore" of the immediate pressures of the interests involved. These requisites could hardly be fulfilled in the midst of the pendular motions and the consequent colonization and fractionalization of the State's instimtional system. fractions. In fact, it did happen in 1973, when the defensive alliance attained an extraordinary but pyrrhic victory.

(VII) Provisional epilogue
The experiment initiated in 1966 sought, on the one hand, to rebuild capital accumulation mechanisms which subordinated the whole of society to the large bourgeoisie and, on the other hand, necessarily and correlatively, to implant a system of political domination which, reversing the preceding situation, would aggressively impose itself on civil society. I have mentioned the collapse of that attempt and how this made way, for the first time, for the conquest by the defensive alliance of the State's institutional system independently of the large bourgueoisie. Recent history cannot be written here. But it is necessary to point out that this alliance could only briefly ignore the large bourgeoisie's and the pampas bourgeoisie's economic supremacy; a glance at the data already presented demonstrates how, after a brief truce in 1974, the cyclical fluctuations repeated themselves much more violently. Even before Peron's death, the intrinsically defensive content and limitations of the alliance had been shown beyond question. The old crisis reproduced itself with unusual acuteness and the local bourgeoisie had to abandon ship without even saving its organization from sinking with it. On the other hand, the exacerbation of "union power" could not go beyond repeating, with increased force after the retreat of the local bourgeoisie, the practices which had made it what it was: an aggressive economicism and the search of new institutional advantages -pursued now from the very heart of the State's institutional system. This cumbersome heritage of past victories created threatening gaps between the union leaders and their own class. It also generated conservative reactions which threatened the substantial autonomy which the unions, and the popular sector, had retained throughout this complex process.
Peron's death, a peculiar "palace" irrationality, and a violence which speedily fed on itself, helped to shake the foundations of a society which accelerated the spirals of its crisis, the same happened with a State that too obviously failed to guarantee the reproduction of this capitalism. But beneath those factors was the fact that, when the defensive alliance managed, at last, to be the governing alliance on its own, it ran into its own limitations; the very reasons which had brought about its extraordinary victory precipitated an unprecedented crisis. The promise of a "nationalist" and "socially just" path of capitalist development was subjected to a positive test and the alliance's centrifugal tensions fired off in their opposing directions.
The defensive alliance's great victory led to the paroxysm of the political and economic crisis, to the ebbing away of the nationalist ideology, to the implantation of a new bureaucratic-authoritarian State and to the dissolution or subjection to government control of the main organizations of the popular sector and the local bourgeoisie. With all of which, for the first time, the defensive alliance's political, ideological, and organizational supports have been neutralized. This has enabled the superior fractions of the bourgeoisie to explore the possibilities for a long term reaccommodation on more egalitarian terms -between them -than those prevalent in 1967-1969. The implication and precondition for such a reaccommodation is the dispersal of the defensive alliance. This does not entirely preclude a return of that alliance or of the spirals we have studied. But for such a thing to happen, the local bourgeoisie would have to set itself on a hazardous "road to Damascus" towards a renewed alliance with the popular sectors; and it is not certain that, by then, the popular sectors will still be confined within the ideological and political parameters which cemented the defensive alliance prior to its greatest and most catastrophic victory.