Towards a Historical Analysis of Elites in Latin America Paper to be presented at the 21st World Congress of Political Science Santiago de Chile – July 12 to 16, 2009 Dr. Cristóbal Rovira Kaltwasser ([email protected])1 Social Science Research Center Berlin (WZB)

Abstract: Where lies the biggest potential of elite research within the realm of political science? To answer this question, the following paper returns to the founders of elite theory in order to illustrate their awareness of the importance of the idea that »history matters«. The reason behind this is that by nature processes of elite circulation and elite formation are long-lasting. In consequence, only by investigating elites through a historical perspective it is possible to comprehend their scope of action in relation to the social structures in which they exist. Additionally, we will insist on the need of a theoretical framework that helps us identifying when and why elites do have more or less power. For this purpose, we will draw on comparative historical analysis and authors like David Collier, Ruth Berins Collier and James Mahoney who differentiate the analytic components of path-dependent explanations of historical processes. Based on this theoretical discussion, the paper examines the role of elites in Latin America, with special focus on the developments of Argentina, Brazil, Chile and Mexico from independence until the economic crisis of 1982. Although these countries underwent many common processes (e.g. political independence from the Spanish and Portuguese Empire or the Great Depression), we will demonstrate that they followed different path-dependent processes depending on the actions undertaken by elites, which operated based on their degree of horizontal integration and their capacity to co-opt rival groups into a cartel. In the light of above, we will also argue that major processes of elite circulation throughout the history of these countries took place. Not only the Mexican Revolution and the emergence of Peronism in Argentina, but also the military coup that toppled Allende in Chile and the »Estado Novo« in Brazil are all examples of established elites being replaced by counter-elites aspiring to build a new model of society.

1 For helpful comments and advice, I am indebted with Matthias Bohlender, Sofia Donoso, Klaus Eder, Anna Menge, Wolfgang Merkel and Herfried Münkler.

1

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

INTRODUCTION Elites do not stay in power forever. This idea is one of the key – if not the most important – theses of classical elite theory. In fact, Gaetano Mosca (1858-1941), Vilfredo Pareto (1848-1923) and Robert Michels (1876-1936) all characterized elite circulation as a constant feature due to the arise of new social forces struggling for the representation of the ideas and interests of the masses. However, the formation and renewal of elites do not happen overnight. It is a matter of what Paul Pierson (2004) refers to as long-term processes, i.e. developments in which the time horizon of both outcomes and causes are slow-moving. From this angle, to explore the potential of classical elite theory it is necessary to take long periods of time into account in order to detect critical junctures triggering processes of elite circulation. Moreover, through the analysis of long time frames it is possible to identify variation in the level of elite’s horizontal integration, i.e. the scope of effectiveness of existent formal and/or informal institutions in permitting both communication and negotiation between different and usually antagonistic minorities of power. Following this thesis it is possible to revise one common myth about the history of Latin America: the weakness of elite circulation and hence the persistence of the same people in power2. To confront this, we will follow the theoretical strand of comparative historical analysis (Mahoney/Rueschemeyer 2003), particularly the approach outlined by James Mahoney (2001b) in his study on regime change in Central America and the seminal work of Ruth Berins Collier and David Collier (2002) in relation to the formation of the political arena in Latin America. In a similar way, we will revise the history of Argentina, Brazil, Chile and Argentina from independence until the great economic crisis of 1982 and stress the crucial role of the degree of horizontal integration of the elites in the constitution of different path-dependent processes. In this sense, based on a »long durée« historical approach, we will demonstrate that there is much more elite circulation than normally recognised in Latin America. The article is structured as follows. In the first section (1) we review the classical elite theory in order to show that its distinctiveness lies in the very idea that it is necessary to take long periods of time into account to comprehend the causes and outcomes of the struggle for leadership between elites and counter-elites. The next section (2) provides a link between classical elite theory and comparative historical analysis, which is mainly inspired by the analytical framework developed by Mahoney (2001b) and Collier and Collier (2002) to study Latin America. After this theoretical discussion, the following section (3) deals with the political independence of Latin America and the consequent definition of territorial states during the 19th century, considering this process as elite-driven and thus dependent on the level of elites’ horizontal integration. Section (4) traces the response by elites in relation to the downfall of the oligarchic regime and the configuration of different path-dependent developments in Latin America during the 20th century. Finally (5), we offer a short conclusion, in which the main findings of the article are summarized and where we stress the convenience of combining elite theory and comparative historical analysis. 1. CLASSICAL ELITE THEORY AND ITS PLEA FOR LONG-TERM ANALYSIS Mosca, Pareto and Michels defined history as an open process rather than a teleological one. From their point of view, scholars should research on factors and forces that in different epochs 2 A good example of this common myth is the best-seller »The Opens Veins of Latin America« written by the Uruguayan essayist Eduardo Galeano. In effect, this book is characterized by the development of a romantic view, in which Latin America is the victim of long-lasting and corrupted elites that due to their association with the center of the capitalist system impede the realization of the revolutionary project, which is predestined to Latin America (Halperin Donghi 2006: 68-69). It is not a coincidence that precisely Hugo Chávez – the current President of Venezuela that is implementing a development with far-reaching consequences, which among other things involves a process of vast elite circulation – used the last Summit of the Organization of the American States (OAS) celebrated on April 18, 2009 in Trinidad and Tobago as an opportunity to give Eduardo Galeano’s famous book to President Barack Obama.

2

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

and societies determine changes in elite composition and functioning (Zeitlin 1994: 224-226). Accordingly, distinct elite configurations, which are decisive for the formation of particular trajectories in time depends on the historical context and the kind of social order under analysis. Therefore, the authors put central emphasis on long periods of time (Parry 2005: 50-54), both regarding the idea of elite circulation and the thesis of the struggle for leadership. The main argument of the founders of elite theory was that power hierarchies are present in all societies and that the very existence of elites is inevitable3 (Burnham 1949). Therefore, against the predominant ideas of the 19th and early 20th centuries, they did not conceive that Europe was experiencing a particular historical period due to the decline of the nobility and the emergence of non-aristocratic leaders. In this sense, every society – regardless if its organization is authoritarian or democratic – should be seen as an entity, in which only a few rule and the majority is ruled but where, at the same time, also processes of power transformation and social change take place. Without doubts, Mosca, Pareto and Michels developed a theory of action. In its center lies the continuous historical interplay between established elites and counter-elites4 that rise because their representation of new societal ideas and interests. From this angle, there is a permanent and inexorable struggle between the establishment and emerging minorities of power, which induce circulation processes through different mechanisms (e.g. co-optation, decadence, revolution, war, etc.). How this struggle is negotiated depends primarily on the horizontal integration of elites, i.e. the existence of formal and/or informal mechanisms that permit both dialog and coordination between different and usually antagonistic minorities of power (Hoffmann-Lange 1992; Rebenstorf 1997; Sauer 2000: 69-81). However, the level of the horizontal integration of elites is in conflict with the relation between elites and masses: the more linked the elites are, the less can they represent the different ideas and interests of the population, and therefore, this is accompanied by a lost of legitimacy. Said in another way, the more the elites differentiate themselves, the more capable they are to represent the existing ideas and interests within society. In this regard, there is a trade-off between gaining legitimacy and achieving coordination (Münkler, Straßenberger, Bohlender 2006: 16; Kaina 2002: 38). Nevertheless, it is wrong to assume that the founders of elite theory were proposing a naïve theory of action (i.e. a purely voluntarist approach), in which elite-agency is the only factor determining the development of society. Quite the contrary, they were trying to build a theory that not only could tackle the dilemma between agency and structure but also take history into account – or to say it in the words of Fernand Braudel (2002), to work with a »longue-durée« perspective. Mosca, Pareto and Michels actually recognized that processes of elite circulation and elite formation were long-lasting; something that is well demonstrated in their recurrent analysis of the historical development of the Roman Empire and the city-states of Northern Italy. Moreover, they argued that the rise and fall of the elites is related with structural transformations of society, which pave the way for a new arrangement that reflects the existent balance of power (Bottomore 1966: 48-68). Since all societies experience modifications in their economical, cultural, military and political structures throughout time, it is logic to expect that these modificaConcerning the inevitability of elites, it is worth noting that this thesis should be interpreted neither as conservative nor as authoritarian but rather as very realistic. Even though Mosca, Pareto and Michels had a problematic relationship with Italian fascism and sometimes advanced ambiguous opinions about democracy, they employed the elite concept to emphasize that in every society will always be tiny minorities who accrue greater power and privilege than anyone else. This issue was recently properly pointed out by John Higley and Michael Burton (2006: 6): »Far from stipulating the form that elites take, the inevitability contention stood simply as an orienting proposition for theory and research: in all modern societies, political power is necessarily concentrated in a few persons, but the configurations these persons form vary importantly among societies and within them over time«. 4 Although Mosca and Michels also used the concept of counter-elites, it was mostly employed by Pareto (1923). For sure this is related with his economical thought, and accordingly to this, the idea that social processes always tend to equilibrium (Bach 2004). From the point of view of classical elite theory, societies can change and even experience revolutions, but in the long run established elites will appear and be contested by counter-elites. However, this thesis is not a relict. One example of its contemporary use is the study of political transitions and its focus on the bargaining processes between established elites supporting authoritarianism and counter-elites defending democracy (Burton/Higley/Gunther 1992; Karl 1991; Merkel/Puhle 1999; O’Donnell/Schmitter 1986). 3

3

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

tions pose a challenge for the established elites. They surely try to stay in power, but in certain circumstances other groups can perform socially requested tasks better, so that counter-elites gain influence, which sooner or later and through different means become part of the established elites5. In sum, if we follow the original intention of the founders of elite theory, it is crucial to analyse long historical periods and search for structural developments triggering the rise of counterelites. For this reason, the mentioned authors did not refer to a ruling class like Marx, but rather to a contest between elites and counter-elites that fight for the guidance of society. In this context, a central thesis of classical elite theory is the historical and never-ending struggle for leadership. In other words, Mosca, Pareto and Michels observe that in every social order there is a permanent conflict between different minorities of power, which intend to steer society. Therefore, the action of the masses can take place only through the agency of a select group of leaders. We can affirm, hence, that the concept of leadership – and not the notion of ruling6 – is the main analytical category of classical elite theory. Accordingly, elites are actors who are struggling to shape society and at the same time, are determined by both structural transformations and contingent processes that limit their room of manoeuvre. Moreover, the centrality of the concept of leadership is expressed by Mosca’s, Pareto’s and Michels’ continuous criticism of one of the predominant schools of thought of the 19th century, namely the »great man theory«, formulated amongst others by Thomas Carlyle (1996). According to this theoretical stream, history is driven by the actions of outstanding leaders, so that to understand social processes it is necessary to study the biography of such high influential individuals and the motives behind the decisions that they take. But classical elite theory disapproved the very idea that great men or women had decisive historical impact. The reason for this is that Mosca, Pareto and Michels conceived leadership as a collective and not as an individual issue (Paris 2003). They emphasized the study of leaders and their behaviour, origin and structure in order to understand the ideas and interests that are part of their societies. However, structural transformations are taking place permanently and once these reach a certain threshold, the rise of

If we were obligated to summarize this approach focused on the historical replacement of elites, there is no better example than the following formulation of Gaetano Mosca (1966: 64–65): »What we see is that as soon as there is a shift in the balance of political forces – when, that is, a need is felt that capacities different from the old should assert themselves in the management of the state, when old capacities, therefore, lose some of their importance or changes in their distribution occur – then the manner in which the manner in which the governing class is constituted also changes. If a new source of wealth develops in a society, if the practical importance of knowledge grows, if an old religion declines or a new one is born, if a new current of ideas spreads, then, simultaneously, far-reaching dislocations occur in the governing class. […] Political classes decline inevitably when they cease to find scope for the capacities through which they rose to power, when they can no longer render the social services which they once rendered, or when their talents and the services they render lose in importance in the social environment in which they live«. 6 If we take for example the translation of the main book of Gaetano Mosca, which in Italian was originally labelled as ›Elementi di scienza politica‹ and later was published as ›La classe politica‹, it is worthwhile to point out that its translation into English was made under the title ›The ruling class‹, whilst its translation into German was made under the title ›Die herrschende Klasse‹. In this regard, both labelling are erroneous since they presume that Mosca applied the same analytical categories of Marx, and consequently they obscure the novelty of classical elite theory. In this sense, it is not a coincidence that later an author like Charles Wright Mills (1956) criticised the notion of ruling class in his work on the formation of a power elite. In fact, he used the notion of elite to state the existence of different and in part autonomous sources of social power, which under certain historical conditions could lead to the configuration of a cohesive cartel of elites, i.e. a power elite. In this regard, the following passage is very clarifying: »Ruling class is an economic term; rule a political one. The phrase, ruling class thus contains the theory that an economic class rules politically. […] Specifically, the phrase ruling class in its common political connotations does not allow enough autonomy to the political order and its agents, and it says nothing about the military as such. […] We hold that such a simple view of economic determinism’ must be elaborated by political determinism and military determinism; that the higher agents of each of the three domains now often have a noticeable degree of autonomy; and that only in the often intricate ways of coalition do they make up and carry through the most important decisions. Those are the major reasons we prefer ›power elite‹ to ›ruling class‹ as a characterizing phrase for the higher circles when we consider them in terms of power« (Wright Mills 1956: 277) 5

4

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

counter-elites results inevitable. Thus, the latter are collective actors who search for a new balance of power and fight for the shaping of the political order. In light of the above, our ideal-typical definition of the elite concept is the following: elites are a tiny minority of established actors that are positioned in the higher positions of authority and power of the different sources of social power of a given society and whose recruitment operates under specific criteria. Further, they develop cooperative and conflictive relations between them, with the counter-elites and with the population in order to achieve and to practice the direction of society. By sources of social power, we understand the existence of diverse »fields« or »systems« (e.g. economy, culture, politics, etc.), which are not always completely differentiated and autonomous from each other, but rather more or less connected and promiscuous7. By cooperative and conflictive relations, we mean that elites always maintain a contingent linkage with both the counter-elites and the population, since no established elite can stay forever in power. This definition captures both that the elite concept implies the very idea of continuous struggle for leadership and that it must be understood with a historical perspective in order to grasp the contingent character of the established elites. In addition, by highlighting that shifts in the sources of social power may provoke processes of elite circulation, this definition suggests an initial and broad idea about how to tackle the dilemma between agency and structure. 2) FROM CLASSICAL ELITE THEORY TO COMPARATIVE HISTORICAL ANALYSIS 2.1. What is comparative historical analysis? Although the founders of elite theory emphasize the importance of taking history into account, they did not develop a systematic approach about how to do it. This research gap is actually one of the main questions left open by the work of Mosca, Pareto and Michels. Moreover, the empirical operationalization of the elite concept in the social sciences mostly derived in a discussion about qualitative (e.g. interviews) and quantitative (e.g. surveys) measures, in which the historical dimension of the research plays a marginal role. By saying this, we are not neglecting the progress made in the empirical research on elites. We are only stating that this progress was achieved to a great extent at the cost of using the elite concept for the analysis of long periods of time. In other words, within current scholarship on elites there is a vast number of national and international empirical studies, yet there are less efforts of developing a clear theoretical framework through which understand how the elite’s scope of influence varies throughout time (Parry 2005: 105-108). Of course, to acknowledge the existence of this research gap is not to suggest that there is an absolute ignorance on the topic. Indeed, there are several interesting sociological and political works that specifically address the historical analysis of elites (Burton/Higley/Gunther 1992; Dogan 2003; Dogan/Higley 1998; Higley/Burton 1989, 2006, Knight 1998a). However, we lack a common and distinctive heuristic assumption regarding the room of manoeuvre of elites within time frames. How can we identify the influence of elites 7 In contemporary social theory there is a broad consensus regarding the existence of different sources of social power, but there is no agreement on the number or how to refer to them. So, for example, Pierre Bourdieu speaks about »fields of power« and Niklas Luhmann talks about »functional systems«. However, none of them gives a plausible explanation concerning, neither the definitive number of sources of social power, nor the criteria to determine when (autonomous) sources of social power rise or fall (Kneer 2004: 35). In this sense, we adopt here the perspective of Michael Mann (1986, 1993), according to which there are four sources of social power; ideological, economical, military and political power, the so called IEMP-Model. The approach of Michael Mann has two main virtues. Firstly, since it assumes that, depending of the historical context and region, the sources of social power can be more or less differentiated, it is flawed to imagine that any of them can be absolutely autonomous or »autopoietic« in the sense of Luhmann. Quiet the contrary, the sources of social power are promiscuous. Secondly, in contrast to the majority of contemporary social theorists, Michael Mann stresses the importance of the military as an own source of social power which is highly influential in every social order; even in those where the Weberian definition of state operates (Joas and Knöbl 2004: 678-679; 2008: 278-283).

5

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

within a historical sequence? As we will argue in this section, comparative historical analysis provides an answer to this question. Since it beginnings, social sciences paid special attention to both historical reasoning and historical comparison. In fact, this mode of investigation can be traced in many founders of modern sociology and political science, as for example in the work of Karl Marx and Max Weber as well as in the analysis of Alexis de Tocqueville and Karl Polanyi (Steinmo 2008: 122). Yet, during the 20th century many approaches of social research arose and comparative historical analysis lost ground. Instead, several new paradigms and methods flourished that favoreded atemporal studies of contemporary social problems, voting behaviour, or processes of individual status attainment (Skocpol 2003: 407). Even though this way of thinking is still highly present in contemporary social sciences, at least since the end of 20th century we are witnessing a true revival of comparative historical analysis. This can be defined as a long-standing intellectual project, which is characterized by three main features (Mahoney/Rueschemeyer 2003: 10-15): a) Causal analysis. The research is engaged in the explanation and identification of causal configurations, which have significant consequences. For instance, in the famous work of Barrington Moore (1966) on the social conditions for the genesis of democracy and dictatorship, or Theda Skocpol’s (1979) pioneering study on the factors that determine the rise of social revolutions. For the development of causal explanations in comparative historical analysis, a key issue is the study of both structural transformations and intended and unintended effects of agency, albeit there is no privileged method for doing this. However, the study goes beyond the hermeneutic approach, given that there is an explicit search for causes that helps to explain the formation of particular trajectories through time. b) Historical sequences. One of the main premises of the research is that social events are not static incidents, but rather processes within a long period of time. Emphasis is placed on historical sequences since the outcomes of one event depend not only on its durability, but also on its appearance in relation to other events and developments (Tilly 1984: 144-147). Once historical sequences are taken into account, it is possible to explain why certain developments crystallise differently in some contexts, as in the case of the formation of the social question in Latin America in comparison to Europe analysed in this paper. c) Comparative perspective. Similar and different cases must be contrasted in the analysis in order to develop a plausible explanation, in which theoretical assumptions are tested and the characterisation of an absolute »Sonderweg« is avoided. In this regard, a good example is the study of Michael Mann (2004) on fascism as a particular totalitarian regime that emerged not only in Germany, but also in Austria, Hungary, Italy and Spain. Due to the emphasis on comparative research, an interplay between theory and history is favoreded since the first can be examined and rectified trough the second, and as result, plausible explicative narratives can be developed (Tilly 2008). With this purpose, this research agenda usually works with a small set of cases, which are studied in detail with the aim of extracting lessons from past events and its influence both on following and current developments. 2.2. Comparative historical analysis of Latin America After this brief excurse, one question needs to be answered: why is the approach of comparative historical analysis so relevant for our analysis of the Latin American elites? The response is that within this approach exists a body of literature with the aim of studying political processes in Latin America, which provides a solution to tackle the dilemma between agency and structure, i.e. to understand the factual room of manoeuvre of elites throughout time. On the whole, it is possible to identify three major studies8. First, the ground-breaking work of Ruth Berins Collier Although these works are related with historical institutionalism, the latter should not be seen as an equivalent to comparative historical analysis. In fact, there are plenty of investigations within historical institutionalism that are focussed only in one land or one policy field. In this sense, the comparative dimension does not necessarily represent a defining characteristic of historical institutionalism. As James Mahoney and Dieter Rueschemeyer (2003: 11) rightly 8

6

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

and David Collier (2002), in which the formation of a (modern) political arena in Latin America is viewed as the result of the emergence of the labor movement vis-à-vis its diverse incorporation into the party system and the state. Through a detailed historical explanation of the political development of Latin America, the authors understand similarities and differences in the national trajectories as product of the reactions and measures taken to confront the social question. Secondly, the study by Dietrich Rueschemeyer, Evelyne Huber Stephens and John D. Stephens (1992) where the impact of capitalist economy on democracy is analysed, challenging the in social sciences well-established thesis of a positive correlation between economic growth and increasing democratization. Concerning Latin America, the authors show that the rise and fall of democracy depended not so much on economic expansion and the resulting development of the labor movement, but rather on the ambivalent role of the middle classes, i.e. if they preferred to make an alliance with the proletariat or with the upper classes. Finally, James Mahoney (2001b) identifies different patterns of formation of political regimes in Central America, centring the analysis on the reforms implemented in the final decades of the 19th century by liberal sectors in order to modernise the state and the agriculture. These modernization processes triggered the formation of different political regimes that in turn, varied in the 20th century depending on the actions of local and international (mostly USA) actors. Without doubts, the three quoted studies offer key arguments and ideas to think about both the historical role of elites in Latin America and the link between theory and comparative historical analysis. Yet, they also suggest a particular analytical framework, which helps us to define the factual room of manoeuvre of elites in time (Mahoney 2001a: 114-115). Indeed, it distinguishes the components for the analysis of political processes in time and, by doing this, it offers a solution for the dilemma between agency and structure. The following figure presents this analytical framework. As we will posit, applying this framework to elites, they are neither wholly independent actors, nor absolutely determined by social structures. Rather, elites are agents that, depending on the moment within a historical sequence in which they operate, have a bigger or lower scope of action. Components of the analytical framework developed for the realisation of comparative historical analysis in Latin America Path-dependent Development

Critical Juncture A

B

C

Cleavage Antecedent Conditions

Historical Legacy Long Time Axis Source: Figure based on Collier and Collier (2002: 30) and Mahoney (2001a: 113, 115)

point out: »[…] all comparative historical works fit comfortably within the field of historical institutionalism, but historical institutionalist works that are not explicitly engaged in systematic comparison do not fall within the field of comparative historical analysis. Hence, we choose to treat comparative historical analysis as one branch within these larger traditions. Our intention is not to stipulate artificial boundaries that prevent the exchange of ideas among scholars working in closely related fields. Rather, we merely seek to recognize a particular kind of research that is treated as its own distinctive scholarly approach by both its practitioners and its critics«. From this vantage point, our research should be seen as an example of comparative historical analysis, albeit we share some concepts with the historical institutionalist framework . 7

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

As the figure illustrates, the point of departure of the analysis is the historical context that defines the background and preconditions affecting the political process to be analysed. This contextualisation is important since the starting point partly determines the configuration of different developments of one common political process (e.g. democratization, state-formation, nationbuilding, etc.). From this angle, the universal use of a general theory is problematic since distinct historical contexts make common political processes crystallise in diverse manners. This is not the least the case in relation to Latin America where certain preconditions that European and North American social sciences assume in their theories to explain processes occurred in their region, not necessarily are present (Centeno and López-Alves 2001). In this regard, later in this paper we will demonstrate that in Latin America the formation of the state was quite different than in Europe, given that in the case of the former the established elites had an easy access to economic resources trough the export of commodities and no warlike expansion of the state was undertaken, as in the case of the latter (Centeno 2002, 2003). Accordingly, the interpenetration between state and society took place in a later historical period in Latin America and was less significant than in Europe and the United States. However, it is not an easy task to identify the starting point of the analysis. Certain degree of arbitrariness can be founded in the selection of the event that sets up the starting point of the research. But without criteria for identifying a point of departure the investigator easily falls into what Mahoney (2000: 527) denominates the trap of infinite regress, i.e. perpetual regression back in time in order to locate temporally prior causal events. In this regard, it is important to look into the historical context in order to identify the configuration of cleavages9. The latter are the result of the formation and establishment of societal structures, which shape both a certain mode of resource allocation and the scope of action for social agents. So, for example, the expansion of capitalism that in Europe implied a growing conflict between capital and labor, crystallised in the configuration of a clear cleavage that shape the political arena and the identity of the political parties (Rueschemeyer/Huber Stephens/Stephens 1992: 46-47). It is worth noting that the cleavages can exist for longer periods of time and therefore define a space, in which opportunities and limits for agency are determined (Collier/Collier 2002: 3031). Thus, the configuration of cleavages does not automatically involve an explosion of societal antagonisms and tensions, since cleavages also contribute to the generation of a negotiation arena between elites and counter-elites. For instance, it is possible to argue that the effective establishment of a democratic order depends on the occurrence of an elite-settlement (Higley/Burton 1989, 2006; Burton/Gunther/Higley 1992), i.e. a kind of pact between elites and counter-elites by which they suddenly and deliberately agree on a new interaction mode, whereby old conflicts do not disappear but can be debated and solved. Nevertheless, those elite arrangements occur only infrequently in history and their factual success relies not only on a collective learning process but also on contingent factors. Beyond the possible long-lasting existence of cleavages there are certain points in time, in which acute contests between social groups take place and a reorganisation of the power asymmetries is triggered. These points in time can be characterized as critical junctures. Following the work of Capoccia and Kelemen (2007: 348), these are defined as »relatively short periods of time during which there is a substantially heightened probability that agent’s choices will affect the outcome of interest«. According to this definition, critical junctures are difficult to foresee, their outcomes can only be partially estimated and, in consequence, they have an extreme contingent nature. In this sense, we are speaking about exceptional points in time, in which agents not only gain a great scope of action but also take decisions that have a profound historical legacy. Hence, 9 The concept of »cleavage« must be distinguished from the one developed in the seminal work of Seymour Martin Lipset and Stein Rokkan (1967), where the constitution of distinct voting blocs is defined by the crystallization of fault lines that separates the population of the European society into advocates and adversaries on certain issues. From this angle, political parties are seen as representatives of specific social groups and party families are defined on the basis of certain socio-demographic characteristics of their electorate. Even though we sympathize with this concept of »cleavage«, we use it in a broader sense to characterize the constitution of fault lines that separates elites and counter-elites, which in turn are representing ideas and interest of large groups of the society.

8

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

critical junctures are also labelled as »formative moments« (Peters/Pierre/King 2005: 1276), i.e. the room of manoeuvre for social actors is temporary expanded, so that counter-elites gain influence to pressure (and eventually replace) the established elites. As a result, the realization of reforms turns out to be easier. 2.3. Path-dependent developments Although critical junctures represent relatively short periods of time, facilitating both social transformations and a rearrangement of the power asymmetries, their outcome might contribute – under some circumstances – to the formation of a new and stable trajectory. In general, it is possible to state, that the longer the critical juncture, the higher the chance that obstacles emerge which later difficult social change (Pierson 2004: 92-102). The explication for this is that once institutional arrangements are set up, the possibility to change turns out to be very low, even if the implied actors do not profit from this situation. Besides this general statement, the concept of critical juncture is used in comparative historical analysis to highlight crucial points in time triggering the formation of an institutional dynamic, which usually is described as a path-dependent development10. By saying this, three main features for the comparative historical analysis are pointed out: a) First of all, the sequence of the events is very influential in the outcome of the processes that will be analysed. If in an initial phase of a historical sequence some decisions are taken and certain preconditions are present, the actors have key resources and vast room of manoeuvre to realize new projects or to accomplish reforms (Pierson/Skocpol 2002: 700). Further, there are some incidents within a process that turn out to be irreversible and, in consequence, the study of the sequence of the events is very important for delivering an accurate explication (Tilly 1984: 35). For instance, as we will see in the following section, in those Latin American territories where elites, due to the political emancipation from Spain and Portugal, did not agree on how to govern the newly formed countries – e.g. by a federal or unitary state –, political instability rose, which in turn, facilitated the rise of caudillos and provoked massive territorial losses. b) The second characteristic is the importance of contingency. This implies that little incidents and coincidences may have decisive and enduring consequences (Goldstone 1998: 834). This is particularly important in the beginning of a historical sequence, since this phase is characterized by a relative permissiveness to make institutional reforms, so that the occurrence of eventualities can have a big impact on the later historical development. Thus, the very idea of social change as a coherent and linear process is refuted, which at the same times signifies that the analysis should have the capacity to incorporate and explain events that are not necessarily anticipated by the theory (Mahoney 2000: 513). c) By using the concept of path-dependent development a final and crucial characteristic is highlighted, namely, the institutional self-reinforcing dynamics. This notion draws the attention to the establishment of institutional arrangements that reinforce themselves, albeit the fact that the involved actors or the majority of the population not necessarily agree on the result 10 Even though the concept of path-dependency has its origin within economics – particularly after the seminal work of the Nobel Prize laureate Douglass North (1990) – it has had a growing influence in political science and sociology (Beyer 2005). Within these disciplines, the concept is often used to criticize the rational choice theory. Since the social world is very prone to path-dependent developments, neither economical optimum nor collective learning is easy to achieve. Moreover, the existence of path-dependent developments can be recognized only by taking long historical periods into consideration, which permit to consider the not intended consequences of agency and the institutional resistance to change. In this regard it is useful to quote the following definition of James Mahoney (2000: 507-508): »[…] path dependence characterizes specifically those historical sequences in which contingent events set into motion institutional patterns or event chains that have deterministic properties. The identification of path dependence therefore involves both tracing a given outcome back to a particular set of historical events, and showing how these events are themselves contingent occurrences that cannot be explained on the basis of prior historical conditions«.

9

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

(Thelen/Steinmo 1992: 8). The reason behind this is that the more the agents act within a certain institutional dynamic, the more increases the probability of its persistency. Put slightly different: as soon as an institutional dynamic is formed, its modification becomes by and by more cost-intensive. So, path-dependent developments refer to a very common occurrence in social and political processes: once formal or informal institutional schemes come into effect, a fertile soil for self-reinforcing dynamics is created, which promotes the status quo and consequently hinders the future change (Peters 1999: 69-70). As for explaining political and sociological phenomena, the notion of path-dependency is indeed very helpful, but there is a certain risk of using it as a kind of catch-all concept. After all, it evolved from economical theory and the debate on technological developments, so its application in the context of political science and sociology should be made with some reserve. This concerns principally the question of social change, because behind the concept of path-dependency we can notice a kind of »implicit conservatism« (Beyer 2005) due to the excessive emphasis on the idea of punctuated equilibrium. In other words, it is implied that the emergence of a pathdependent process signifies a deterministic no-changeability, which can be challenged only by an external shock (Steinmo 2008: 129). In contrast to this idea, we draw on the concept of pathdependency primarily to emphasise that historical developments resulting from a particular critical juncture produce a historical legacy which determines a pattern of agency, like for example an institutionalized regime or an unstable political dynamic. In the latter case, the »self-destruction« of the legacy may be predictable from the path-dependent development, though the length of time before this occurs may vary greatly and is influenced by other factors as well (Collier/Collier 2002: 34). Nonetheless, there is no legacy that last forever, since there are always actors challenging the established order and critical junctures may arise triggering processes of social change. In sum, the presented analytical framework offers a solution to tackle the dilemma between agency and structure: the factual room of manoeuvre of the elites depends on the moment within a historical sequence in which they are operating. Whereas the scope of action of the elites is very vast at the critical junctures, it is strongly diminished once a path-dependent development is put into motion and a particular historical legacy is formed. In this situation, the structures gain power over the actors, so that the struggles between elites and counter-elites do not necessary have an evident and immediate impact for the transformations of the established order11. 3) THE INDEPENDENCE OF LATIN AMERICA AND THE ELITE’S QUEST FOR A NEW ORDER 3.1. Antecedent conditions and cleavage between Peninsulares and Creoles Regarding Latin America’s post-colonial history, the importance of the elites can be traced back to the processes of political independence. These were indeed struggles for elite circulation, which found an end mostly through war and violence. This section will analyse this process through the lens of the previously revised analytical framework. The antecedent conditions in this regard are the reforms of the Bourbon dynasty during the second half of the 18th century, which attempted to increase the control of the Spanish crone over their colonies. This was translated into rising taxes, growing monopolisation of trade as well as the inclusion of new barriers to the exercise of power of the Creoles (Halperin Donghi 2001: 78–79). Furthermore, the prohibition of the Jesuits in the year 1773 produced social discontent in vast circles of the Latin American colonies, since this religious order was an important channel for social mobility for the Creoles. For example, it is not an accident that in Mexico the most significant leaders of the independence

11 By saying this, we are not affirming that the elites are in certain periods of time merely puppets of social structures, since they usually have decisive influence in slow-moving causal processes, like cumulative causes, threshold effects and causal chains (Pierson 2004: 82-90).

10

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

against Spain were Jesuits priests – Miguel Hidalgo and José Maria Morelos –, who experienced the discrimination of the Spanish born people with particular vehemence (Krauze 2004c: 51-94). During the 18th century, the resistance to imperial reforms grew steadily since it implied more and bigger obstacles for the native minorities, – who by and by were expanding their presence and influence –, to access political power. In other words: what was a rational decision for the Spanish Empire represented a frontal attack to the interests of the local decision-makers (Bushnell 1987). The Creoles gained little recognition and were excluded from the most important positions of power, albeit they were becoming more influential both with regards to the economic development and the political stability of the colonies (Smith 1992: 32). Gradually they became a counter-elite, which the established elites neither saw as a challenge nor could co-opt successfully (Anderson 1996: 65). This cleavage between Peninsulares and Creoles became evident in the case of Mexico, where at the end of the 18th century the national population of six million inhabitants had the following ethnic composition: 60 percent natives, 22 percent Mestizos, 17,8 percent Creoles and 0,8 percent Spanish-born population i.e. Peninsulares (Anna 1987: 52). Although the cleavage between established elites born in the center and the counter-elites born in the periphery was getting stronger in the 18th century, independence was not considered in general as an option. The American and French revolutions evidently had some echo in Latin America, but at the same time the upheavals of José Gabriel Tupac Amaru in Peru and of the black population in Haiti were seen as a warning of what political emancipation from the center could entail (Lynch 1987: 36- 37). Therefore, the Latin American independence was neither wellplanned nor long-premeditated. It was rather a contingent incident triggered by the Napoleonic invasion of the Iberian Peninsula in 1808 and the unseating of King Charles IV (Halperin Donghi 2001: 83-92). This situation provoked a vacuum of power in the colonies, since the questioning of the kings’ authority implied a crisis of political legitimacy and power. From that moment, the Creoles had a singular opportunity to discuss their position in the social structure and so began a struggle between factions in favor either of the Spanish Empire or of political emancipation. In sum, this struggle and the consequent outcome within Latin America was enabled due to the emergence of a critical juncture, which started with the Napoleonic invasion of the Iberian Peninsula in 1808 and ended with the declaration of independence of the colonies12. Contrary to many European movements of the 19th century, in which the workers, the rural population and the petite bourgeoisie were key actors (Eder 1985; Mann 1993; Tilly 1990), in Latin America the independence wars were very restrictive in terms of popular participation (Centeno 2002: 47-52). The Latin American movements for political emancipation were struggles between elites for the quest of a new order and indeed independence initiated a new era, which is usually characterized by historians as the »era of the caudillos« (Carmagnani 2004: 173-174). The decline of the imperial elites implied a change of power in favor of new groups, which due to military means were able to control the colonial territories (Medina Echavarría 1964: 32). Therefore, Alan Rouquié (1984: 62-74) rightly points out that the dissolution of the Spanish Empire yielded a fertile soil for the rise of warlords. But their power was not the same all over the continent: when the elites were capable of maintaining a high level of horizontal integration, the emergence of the caudillos was highly limited and there were no territorial loss. In this regard, we can reconstruct two different path-dependent developments: whilst countries like Argentina and Mexico It is worth noting that the political emancipation of Latin America is also related with a double failure of the Spanish crown. On the one hand, the Spanish elites hindered basic modernization processes within the Latin American colonies, so that the accumulation of capital for the own industry was not promoted and in consequence the Spanish Empire developed its economical potential only to a limited extent (Münkler 2005: 106-107). On the other hand, the Spanish elites did not consider that the rhetoric of the American and French revolutions could help to the formation of counter-elites in the colonies vis-à-vis the idea of »Americanismo«, i.e. the thought that all Spanish American colonies had a common identity and destiny, which should be crystallized in the future configuration of one great nation (Hobsbawm 1999a: 117). Both failures impeded the imperial elites to guide a moderated transition process and thus contributed to a very quick disintegration of the colonial territories. But as we will see later in the paper, the behaviour of the Portuguese Empire was quite different and accordingly Brazil took a special trajectory. 12

11

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

initiated a long era characterized by the caudillos triggering violent conflicts that facilitated the shrinking of the colonial territory, countries like Brazil and Chile experienced a relatively strong horizontal integration between the elites, which favoreded the expansion of the inherited territorial boundaries. Before we explain both path-dependent developments – susceptibility to crisis and political stability – in more detail, next section makes a short account of the national trajectories. 3.2. Reaction to the critical juncture and national trajectories by the formation of a territorial state Argentina arose from the Viceroyalty of the Rio de la Plata, which was a late product of the Bourbon reforms that divided the territory in few dependencies and two centerss of power: Buenos Aires and Montevideo. Nonetheless, the Viceroyalty was poorly interconnected and additionally, both ports had little capacity to exercise their power (Scheina 2003a: 41-42). This situation was aggravated by the political emancipation from Spain for two reasons. On the one hand, the elites of Buenos Aires and Montevideo sustained a rivalry that was fuelled by the actions of international powers and ended with the formation of the Republic of Uruguay as a separate country in 1828 (Rock 2006: 71-79). On the other hand, the caudillos started to build their own (para)military forces to accomplish local interests (Halperin Donghi 2002). In effect, the critical juncture that started with the Napoleonic invasion of the Iberian Peninsula facilitated the rise of Creole elites, which won leadership thanks to an increasing militarisation and ruralisation of power (Romero 2001: 176-196). These elites associated themselves with warlords that had ascended due to the exploitation of large-scale landed property, and together, they opposed any effort of building either a common territory to govern or a central state (Lynch 1992: 132-159). According to the seminal analysis of Tulio Halperin Donghi (2005), it was only with the rise of a post-revolutionary elite generation against the power of the caudillos that a federal pact could emerge, and with this, the constitution of a territorial state. The latter began to take shape with the end of the regime of Rosas and the ascension of General Roca in the year 1880. Not until then could Argentina start to consolidate a stable oligarchic regime and follow a path marked by increasing political integration and economic growth (Cortés Conde 1993). In contrast to the other Latin American emancipation processes, Mexico underwent not only a revolution against the Peninsulares, but also a sort of civil war with a significant popular participation (Anna 1987). The insurgent priests Hidalgo and Morelos turned into military caudillos and, due to their religious authority, had the capacity to mobilise a great part of the indigenous population in the declaration of independence (Bernecker 2007: 123). For both the Creole and the Spanish-born elites, this emancipation process involved a serious risk of upheaval of the Indian population. Therefore, both elites fought united against Hidalgo and Morelos, and tried to rebuild the old order under the leadership of Agustín de Iturbide. However, the revolutionary process had an effect in the whole territory. Caudillos emerged in different parts of the country, characterised by a loose identification with the conservative or liberal camp and a strong compromise with regional elites (Bazant 1991). Further, this conflictive scenario was fostered by the continuous military incursions of France, Spain and USA in the Mexican territory, making the constitution of a central state practically impossible. A turning point in this process was given by the rise of Bénito Juarez in 1858 and Porfirio Díaz in 1884. Under their leadership, the structure of an oligarchic regime was build, which promoted not only the control of the regional caudillos but also the interplay between different elites within the country (Katz 1991: 80-88). Chile, on the other hand, had belonged to the Viceroyalty of Peru under the status of a »General Captaincy« during the colonial period. The country had natural frontiers in the West (the Pacific Ocean) and in the East (the Andes mountain range). The north had practically no population and in the south, the Bio-Bio River corresponded to a consolidated border line which until 1850 was successfully defended by the Mapuche clan (Collier/Sater 2004: 18-22). This was, in comparison to Argentina or Mexico, an uncomplicated geopolitical scenario. Moreover, one 12

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

factor helped for a relative easy transition from the colonial to the independent order, namely the fact that Peninsulares and Creoles were strongly related through blood relationships (JocelynHolt 1999: 73-98). In fact, after political independence from Spain, Chile experienced only a brief period of the caudillos between 1823 and 1829. Following this, an order leaded by conservative elites was consolidated, promoting economic growth and political stability. This rather unhindered trajectory found a short breakup with the emergence of two short civil wars (1851/1859 and 1891), which can be understood as the consequence of the struggles driven by new elites with particular interests in the copper and saltpetre mining (Zeitlin 1984). In spite of this, the history of the Chilean 19th century was a less conflictive development compared to the Argentinean and Mexican ones: struggles between regional elites did not play a key role, the rise of the caudillos was a short phenomenon and the elites maintained a high level of horizontal integration, mainly due to the existence of a closed marriage market (Vicuña 2001). Finally, the Brazilian case is very interesting since it illustrates to what extent the action of the elites of the Portuguese Empire brought about an own trajectory, which is quite different from the rapid and usually instable formation of republics within Hispanic America (Carvalho 1980). It is worth noting that Brazil compassed not only a huge and very rich territory, but also a very social and economic heterogeneous one. So how was it possible to maintain a unity and avoid fragmentation within such a vast territory? To grasp this issue, it is necessary to take into account the following six characteristics of the Portuguese Empire and the Brazilian colony in comparison to the Spanish ones (Bethell 1993: 161-162; Smith 1992: 31): a) The agrarian oligarchies were a late development in the Brazilian colony, so that the rise of caudillos was more difficult than in most Hispanic America. b) The Portuguese colonial system was less repressive than the Spanish one and thus the Creoles had more possibilities to obtain some power positions. c) Both the Portuguese and the Brazilian elites were trained at the University of Coimbra, which facilitated an increasing horizontal integration. d) Brazil was a slave society and the fear of a popular rebellion acted as a unifying factor between Creoles and Peninsulares. e) The commerce monopoly of Portugal over Brazil was less strong than the one of Spain over their colonies, and thus, the discontent of the local elites in relation to the central elites was lower. f) The reforms of the Portuguese Empire from the second half of the 18th century onwards opened some channels of social mobility for the Creoles. Besides these characteristics, the Portuguese Empire followed a different strategy than the Spanish one in relation to the Napoleonic invasion of the Iberian Peninsula: the Portuguese royal house took the decision to go into exile in Brazil. Dom João emigrated the year 1807 and settled the year 1808 in Rio de Janeiro with a court of approximately 15000 persons (Bernecker 2000: 127). So, the status of the former dependent colony changed factually into an equal integral part of the Portuguese Empire. After the return of Dom João to Portugal in the year 1821, his son Pedro was declared emperor of the constitutional monarchy of Brazil. He abdicated ten years later giving the crown to his son Pedro II. The latter governed until 1889, and due to a strategic co-optation strategy, he achieved a significant level of political stability and a relatively high control of the military throughout his regency (Bethell/Carvalho 1993). Additionally, the nobility acquired new mechanisms of recruitment through the maintenance of control of large extensions of land, and by following a successful administrative career in the dependencies of the territory. This way, more and more Creoles came into power and the opinion of regional elites met response in the central state. So, Brazil experienced a relatively easy transition from being a colony to forming an empire of its own13. At this point, it is important to mention that the notion of »Brazilian Empire« is a historical category rather than a political concept. Indeed, the notion of »Brazilian Empire« refers to an epoch that goes from the declaration of independence from Portugal (1822) until the fall of the monarchy (1889). Noteworthy is that during this period

13

13

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

3.3. Path-dependent developments in the 19th century and their historical legacy: the oligarchic regime After this brief historical overview of the four cases included in our analysis, the next step in the unfolding of our argument is to build a generalizing perspective that grasps the two pathdependent developments mentioned in relation to the countries under analysis. As was illustrated by the case-studies, the consolidation of a new order went hand in hand with the level of horizontal integration of the existing elites. Thus, what we refer to as susceptibility to crisis was an expression of a lack of agreement within the elites in the establishment of a new government. In this respect, we posit that the rise of the caudillos was related with a failure of the elites: the lack of agreement amongst the Creole minorities, and thus, the incapacity to build a new order gave rise to a fertile soil for the emergence of autodidactic warlords, who acted as defenders of local ideas and interests. On the other hand, political stability was the result of agreement between the elites concerning the establishment of a political independent regime, which permitted social mobility to (some) Creoles and at the same time maintained the core structures of the colonial order. Regarding this, a main issue was the development of a system of elite co-optation vis-à-vis (a certain degree of) control of the military forces. Two path-dependent developments in the Latin American decolonisation processes and the oligarchic regime as their historical legacy Susceptibility to crisis

Argentina

Political stability

Mexico

Brazil

Struggle between elites Short civil war, precarious from Buenos Aires and restitution of the old order and Montevideo, and emergence of regional forces emergence of regional represented and commanded by forces represented and caudillos, which benefited from commanded by international conflicts with caudillos France, Spain and USA

Weak horizontal integration between the elites and longlasting struggle for the consolidation of an own territorial order

Settlement of the Portuguese royal court in Brazil and practise of a successful policy of elite co-optation in favor of the Creoles

Chile Short duration of the emancipation wars and little appearance of both regional forces and caudillos

Substantial horizontal integration between the elites and rapid consolidation of an own territorial order

Formation of oligarchic regime characterized by a cartel of elites, which is able to expand the state in their own favor due to the export of commodities and international lending

In the cases of Argentina and Mexico we can identify a path-dependent development of susceptibility to crisis, which, characterized by a low level of horizontal integration of the elites, made the process of consolidation of a territorial state extremely difficult. In both countries, caudillos appeared in the course of the political emancipation and could expand their authority and power thanks to the absence of both a central government and an elite agreement. Unlike generexisted an »nobility without an aristocracy«: there were nobles thanks to titles the emperor might conferred to people, but these titles were not to be passed on, so they did not permit a blood aristocracy (Fausto 1999: 80). 14

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

als of a professional army, the caudillos did not have a national reference but rather a regional constituency. Once they were in power, they governed based on the support of local and private interests and not on behalf of a genuine reason of state (Lynch 1992: 411-424). The rise of the caudillos illustrates that through the use of violence, opportunities for advancement for the Creoles were created. Due to the independence, indeed, a new generation of elites came into power, which exercised a particular style of leadership and hegemony-construction based on precarious alliances between local elites (Carmagnani 2004: 177). In this context, the duty of the caudillos14 was to recruit different persons, raise their status and link them to their clientele (Rock 2006: 101). Further, the caudillos maintained contact between them, which provided the base of the political system. However, it was very difficult to find a balance between centripetal and centrifugal forces, therefore no stable political equilibrium was supported and the definition of a territory to govern was not easy to make. The path-dependent development that we have referred to as susceptibility to crisis can be well perceived in the loss of territory and difficult demarcation of an area to govern. The Viceroyalty of the Rio de la Plata contained approximately 5 million square kilometres, whereas today the Republic of Argentina compasses approximately 2,8 million square kilometres (Fausto/Devoto 2004: 29-30). Moreover, Mexico lost between the declaration of independence in 1810 and the year 1853 more than 2 million square kilometres (Scheina 2003a: 181). In both cases, regional forces won power due to the independence and the Creole elites did not reach an effective agreement concerning the establishment of a new government, giving rise to secessionist movements. The latter were fostered by the pressure of international actors and the emergence of caudillos. Federalism and centralism came into being as the principal conflict between the elites, since they were fighting in order to define a common territory to govern (Centeno 2002: 107). By contrast, the cases of Brazil and Chile experienced a smooth transition from colonial order to the consolidation of a territorial state. They represent what we have called the pathdependent development of political stability. In neither country a long or dramatic era of caudillos took place since the elites – in spite of their conflicts – had the capacity to build a new order and defend the inherited territories. Moreover, both countries expanded their colonial boundaries after the independence, process in which the elites played a key role (Knight 1992b: 19). In the case of Brazil, there were several secession movements, but a sophisticated system of elite cooptation with the support of the Emperor and his so called »moderating power« took form, which enabled the different regional elites to gradually get involved in the government (Fausto 1999: 77-78). Brazil lost some land because of the independence of Uruguay, but at the same time, it gained territory in the north and the west15. In the case of Chile, the lack of powerful regional elites favoreded the formation of a central state, which was implemented by a closed and well interconnected elite. At the end of the 19th century, the country went into war with Bolivia and Peru with the intention of annexing domains in the north that were mineral abundant (Collier/Sater 2004: 127). Further, the established elites implemented a strategy of colonization of the southern regions of the country. As a result of this, Chile expanded his boundaries far beyond the »uti possidetis« principle (Beyhaut/Beyhaut 2000: 103). Moreover, the Brazilian and Chilean historical developments after independence are similar in relation to the experience of a growing conflict between conservatism and liberalism. Since in these cases the question about the territory to govern was solved without great problems, the issue about federalism and centralism was not very decisive. Rather, worldviews regarding the organization of society were much more influential. In Chile, the conservatives dominated after Independence, but from the second half of the 19th century onwards the liberals began to win authority and power, in particular due to the weight of the freemasonry and the foundation of In the light of this explication, the etymology of the word » caudillo« is very illuminative since it emanates from the Latin »caput«, which means head or summit. 15 In total, the country gained an extension of 600.000 square kilometres in comparison to the inherited territory, i.e. an extension similar to the one of contemporary France (Zoller 2000: 223). 14

15

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

secular institutions like the »Universidad de Chile« (Krebs 2002: 250). In Brazil, in the beginning of the regency of Pedro II the struggles between federalism and centralism were very strong, but by and by the conflict between conservatism and liberalism gained strength. Whereas the conservative camp was embodied by an alliance between imperial bureaucrats, traditional businessmen and old great land-owners, the liberal camp was represented by an alliance between urban professionals, domestic-market-oriented businessmen and recent great land-owners (Carvalho 1980: 155-176). The conflict between conservatism and liberalism was reinforced by the issue of slavery, since it involved a contest of ideas and interests in which nothing less than the model of society to build was at stake (Graham 1993: 125; Viotti da Costa 1993: 190). Map of Latin America around 1785

In sum, we can trace two path-dependent developments that were formed throughout the independence processes and that shaped the behaviour of the elites in relation to the formation of a new socio-political order. But what became the historical legacy of these developments? Despite the differences between the analysed cases, at the end of the 19th century all of them eventually ended in a common pattern: a relatively stable political system denominated oligarchic regime (Bourricard 1969; Graciarena 1967). However, there was a big variation in reference to the size of the territory to govern and the efforts put to control the caudillos. Albeit this difference, Argentina, Brazil, Chile and Mexico experimented at the end of the 19th century a political and economical situation much more stable than at the beginning of independence (Halperin Donghi 2001: 280). Concerning this matter, two factors were highly influential (Oszlak 1981: 152). On the one hand, increasing control over the regional forces and a relatively pacification of the conflicts with both the neighbour countries and international actors facilitated the centralization of power. On the other hand, the gradual formation of a new generation of Creole elites without direct re16

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

lationship with the independence processes and with a relatively high level of horizontal integration provided a fertile soil for economic and political progress. Map of Latin America around 1900

Nevertheless, the development of the oligarchic regime in itself was made possible because of the expansion of the state vis-à-vis its appropriation by the established elites. Here lies in fact the particularity of the oligarchic regime: via export of commodities and international lending, the established elites had the means to expand and shape the state, but without making any significant modification of the relationship between elites and masses (Cardoso/Faletto 1976: 91-93; Carmagnani 1984: 142-143). Thus, we can observe that at the end of the 19th century a rentseeking cartel of elites took form in Argentina, Brazil, Chile and Mexico. In these countries, the so called »extraction-coercion cycle« which was so decisive for the constitution of an autonomous state in Western Europe and USA did not come into effect (Mann 1988; Reinhard 2007; Tilly 1985, 1990). The elites of the oligarchic regime did neither have the means nor the interest to expand the state based on war-making (Centeno 2003). As Terry Linn Karl (1997: 61) rightly points out: »They did not need to subdue and appropriate the labor of an entire population or penetrate inaccessible rural zones in order to control indigenous peoples. Nor were they forced to collect taxes beyond the export sector, which might have helped them to develop more extensive extractive capacities. Instead, their earnings depended on a combination of coercion and negotiations with local elites in the capital city and later on the concessions or royalties they were able to win«. In consequence, in countries like Argentina, Brazil, Chile and Mexico an oligarchic order was formed at the end of the 19th century, in which the establish elites maintained a high level of horizontal integration and did neither have the interest nor the capacity to promote an interpenetra17

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

tion between state and society. Therefore, apart from the patronage structures they maintained, no strong links with the majority of the population were built. As long as the economic structure was principally based on export of commodities and the society was preferentially agrarian, there was no great challenge for the further development of the oligarchic regime and its corresponding cartel of elites. But as we will see in the following section, this situation changed dramatically at the beginning of the 20th century due to rise of the social question and of counter-elites demanding a democratization of the political order. 4. THE DOWNFALL OF THE OLIGARCHIC REGIME AND THE RESPONSE OF THE ELITES 4.1. Antecedent conditions and cleavage between oligarchic elites and counter-elites originating from the middle classes The oligarchic regime can be described as the historical legacy of the path-dependent developments that were triggered by the independence of Spain and Portugal. This regime in turn, defined a set of antecedent conditions for later developments and in consequence it is important to bring to mind its three main features: the implementation of a liberal economic model based on the export of commodities, the expansion and formation of a state in the benefit of the established elites, and the promotion of elites’ horizontal integration via the development of common institutions (Carmagnani 1984: 121-128; Balmori/Voss/Wortman 1984: 43; Whitehead 1998: 384). Moreover, by virtue of the weakness of the bourgeoisie and the maintenance of the structures of the colonial order, the elites of the oligarchic regime promoted an economic modernization and sustained at the same time a very restricted political rule, in which the greatest part of the population was not integrated (Cavarozzi 1996: 22). Notwithstanding, due to their closure the cartel of elites that in each country was in charge of the oligarchic order did not have the ability to manage long-lasting processes of social change that were happening, which implied that sooner or later the power of counter-elites would rise (Germani 2003: 98). Without trying to exhaust the list of factors fostering the downfall of the oligarchic regime, we can mention three slow-moving causal processes that indubitable were highly influential: a) Professionalization of the military. Towards the end of the 19th century, countries like Argentina, Brazil, Chile and Mexico initiated a process of military modernization with the assistance of France and Prussia (Nunn 1986). The idea was to build official academies for the armed forces and to import modern weapons in order to have proper protection of the territorial state against both internal and external challenges. In theory, the process of modernization of the military was meant to promote a growing differentiation and professionalization of the armed forces (Nunn 1997: 32). However, in practice it provoked a gradual formation of a new generation of military elites which came from the middle classes and accordingly did not have strong contacts with the elites of the oligarchic order (Loveman 1999: 69; Rouquié 1984: 108-113). Over the years, this new generation of military elites began to condemn the established elites, to form closed clubs promoting particular ideas and to show growing interest in the construction of the nation. In this sense, their political outbreak against the oligarchic regime was only a question of time (Rouquié 1998: 145). b) Transformation of the economy. As we discussed before, albeit the differences between Argentina, Brazil, Chile and Mexico, at the end of the 19th century all these countries presented an oligarchic regime which was characterized – among other features – by a relatively successful economic policy (Glade 2002: 52-53). The countries followed a liberal economic model in order to enhance the exchange between center and periphery. As a consequence, a slow but decisive differentiation of the economy was triggered and hence new actors started to gain power (Carmagnani 2004: 227-237). Moreover, the continuous extension of the railway system opened new opportunities not only for the export economy, but also for the mobility of 18

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

the population, their migration from the rural to the urban world and their political organization (Beyhaut/Beyhaut 2000: 59-60). c) Demographic trends. Without doubts, growing immigration from abroad had a decisive impact in the social structure of Argentina, Brazil and Chile (Merrick 1998: 3). Even though Latin America was still a predominantly rural region in terms of where its population lived in the 1930s, an increasing urbanization took place from the beginning of the 20th century onwards (de Oliveira/Roberts 1998: 243). This was accompanied by the start of a slow decline of the landed elites and the gradual emergence of a class society, in which the importance of urban actors rose (Carmagnani 2004: 245-273). As Gino Germani (1971: 301-305) rightly points out, already before the Great Depression the Latin American societies were experiencing a slow but decisive transformation of the social structure, which involved the emergence of masses that later were mobilized by counter-elites, putting pressure on the downfall of the oligarchic regime. If we take these three slow-moving causal processes into account – professionalization of the military, transformation of the economy and demographic trends –, we can conclude, that at the beginning of the 20th century, the oligarchic regime was being challenged and a central paradox was being born. On the one hand, the formation of a cartel of elites was very effective for the political stability within the recently formed territorial states, albeit this entailed little space for the participation of the masses. On the other hand, the formation of a cartel of elites encouraged modernization processes that in the long run activated causal forces facilitating the emergence of counter-elites. Thus, the success of the oligarchic regime paved the way for its own collapse. The cartels of elites in power in the countries under analysis were neither interested nor capable to enhance broader reforms in order to change the political regime and promote a democratization of society. This process culminated with the formation of a cleavage between oligarchic elites and counter-elites coming from the middle classes, which was usually associated with the urban world and maintained contact with large population groups through institutions like labor unions or new political parties (Graciarena 1967: 45-56; Smith 1992: 38-39). 4.2. Reaction to the critical juncture and national trajectories followed by the downfall of the oligarchic regime As we have previously discussed, the sole emergence of a clear cleavage does neither implicate an automatic transformation of the social order nor an augment of the room of manoeuvre of the elites. So how was a generalized decay of the oligarchic regime in Latin America possible? To answer this question it is necessary to point out that the Great Depression represented a critical juncture, that is to say, a historical moment in which agent’s scope of action is broadened, and the struggles between elites and counter-elites become more evident. As the counter-elites conceived the expansion of the state as the main solution for the crisis, the power of the oligarchic regime was gradually undermined. Additionally, the dramatic modification of the international scenario enabled the expansion of authority and power of those groups that were fighting against the established elites of the oligarchic order (Carmagnani 1984: 176-178). Even though the Great Depression caused a struggle for leadership within Latin America, the factual range of elite renewal varied from country to country. For example, the Mexican Revolution led to a major elite circulation even before the Great Depression (Smith 1979), whilst the established elites in Argentina utilized the economic crisis to make an alliance with the military with the aim of preserving the oligarchic regime, facilitating in this way the later appearance of the populist experiment of Perón (Murmis/Portantiero 2004). Between these extremes of elite renewal in Mexico and of elite closure in Argentina, Chile and Brazil experienced a moderate elite circulation from the 1930s onwards, where an amalgamation between old and new groups took place (Correa et al. 2001: 113-152; Fausto 1999: 190-236). In order to understand these trajecto-

19

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

ries and their effects for the formation of different path-dependent developments in the 20th century, in what follows we make a short explanation of each country case. Considering the economical and political development of Argentina at the beginning of the 20th century16, it is not obvious why this country experienced a prolongation rather than a collapse of the oligarchic regime due to the Great Depression. Historians actually refer to the period between 1930 and 1943 as the »infamous decade«, since during this time the established elites continued in power by reason of massive electoral fraud, systematic corruption of the civil service and an opportunistic approximation to the military (Romero 2006: 67-95). However, the putsch of 1930 did not imply a simple continuation of the economical laisser-faire, but rather a gradual industrialization promoted by the state which entailed a pact between agrarian-oriented and industrial-oriented elites (Murmis/Portantiero 2004). Among the main losers of this new policy were both the working class and the growing urban masses, since little economic redistribution was bolstered and there was barely space for political participation. In consequence, the established elites lacked a hegemonic position and therefore sought to collaborate with the military to stay in power. By doing this, they started a dangerous game, in which the armed forces were seen more and more as a political actor (Potash 1981: 405). Within the army, the so called »Grupo de Oficiales Unidos« (GOU) arose and in consonance with the European fascist movements of that time it attacked the established elites, confronted the communist ideology and endorsed a vigorous nationalist discourse. This group was composed mainly of colonels with a professional instruction coming from the middle classes and thus represented ideas and interest of a broader part of the population (Rock 1993: 222-223). One of main leaders of the group was Colonel Juan Domingo Perón. Particularly after the putsch of 1943 he gained more and more influence, because he was in charge of the relations with the labor unions and from this position he began to cultivate a privileged bond with the masses. In sum, in Argentina the reaction of the established elites to the Great Depression turned out to be counterproductive for their own interests. In fact, in 1946 Perón won the presidential election. His appearance signified a massive elite circulation in favor of new groups that were fighting against the oligarchic regime and that originated mainly from the middle classes and the military (de Imaz 1965; Smith 1974). The new regime represented a populist experience mobilizing the masses and promoting a kind of passionate politics, in which there was no ground for a stable hegemony and ended generating a deep split not only within the elites but also within the Argentinean society (Germani 1973; Horowicz 1985; Martucelli/Svampa 1997). Since different social forces were supporting the continuous intervention of the military, after the putsch against Perón in 1955 and until 1982 there was a continuous back and forth of authoritarian and democratic regimes (Sidicaro 2004). Hence, the breakdown of the oligarchic regime in Argentine went hand in hand with the rise of a political game marked by the radicalization of the friend-foe schema (Schmitt 1932) and the impossibility to form a stable hegemony. Indeed the military itself split and operated as one actor more of the political game (Collier/Collier 2002: 736). In the case of Brazil, with the arrival of the Great Depression the military intervened in order to set a new political course up and to back the leadership of Getúlio Vargas. His rise was possible due to an alliance between the elites of the northern states and the elites of the increasingly powerful state of Rio Grande do Sul, since both disagreed with the oligarchic regime and the »café com leite« arrangement, i.e. the rule of the landed gentries of both São Paulo dominated by the coffee interests and Minas Gerais dominated by the dairy interests (Camargo 1993: 319). 16 At that time, the Argentinean society was experiencing a golden period, which can be noticed by highlighting two features. First: as measured by the per capita income, Argentina was – next to USA, Canada, New Zealand, England, Switzerland and Australia – one the richest countries of the world (Fausto/Devoto 2004: 152). Second: in terms of democratization, Argentina was also one of the leading countries of the world. Accordingly to the declaration of the »Sáenz Peña« law in 1912 all men that were older than 18 years were obligated to go to the polls. Even though this law did not produce automatically a mass democracy – among other things, because neither women nor new immigrants could vote and electoral fraud was highly extended (Smith 1974: 10-11) –, it is worth noting that the Argentinean political system was much more open than in the neighbouring countries and in the rest of the world (Hobsbawm 1999b: 96).

20

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

Once in power, Vargas initiated a gradual process of elite renewal, implemented policies in favor of industrialization and declared the formation of the »Estado Novo«, a new political order that to some extent was similar to the fascist and corporatist experiments of that time in Europe (Fausto 1999: 198-202). Vargas experienced, however, a tragic end, since he committed suicide in 1954 due to a growing dispute with the armed forces. The subsequent governments continued the thus far followed policies: import substitution industrialization financed by the state, corporative arrangements (particularly due to state control of the labor unions), focus on the economic development of the country, and promotion of a partial expansion of both political and social rights (Carvalho 2005: 126-153). Without doubts, after the Great Depression the military played an important role in Brazil, but it acted mostly in line with the constitution and nearly all major disputes between elites and counter-elites could be solved through compromises (Fausto/Devoto 2004: 371). Nonetheless, this partial elite-settlement regarding both the development model of the country and the way to deal with conflicts lost ground, primarily because of the communist challenge and its contentious nature as a result of the growing intensity of the Cold War (Dezalay/Garth 2002: 34-37). The hegemony of the governments after Vargas became more and more precarious. On the one hand, there was a growing struggle for leadership within the elites and especially with a new generation of military elites trained under the U.S. armed forces (Carvalho 1981). On the other hand, the masses put pressure on the established elites with the intention of gaining more political power and obtaining more economic support from the state (Weffort 2003: 75-89). As we will see later, the outcome of this process was not favorableable for the democratic development. As Ruth Berins Collier and David Collier (2002: 169-195) point out in their seminal work, the development of Chile is similar to the one of Brazil since in both countries the Great Depression fostered the formation of a new order in which established elites and counter-elites could arrange a new modus vivendi vis-à-vis a certain level of hegemony. The military intervened in Chile in 1927 under command of Colonel Carlos Ibañez del Campo, promoting policies both for the benefit of the middle classes and to confront the escalation of the social question (Correa et al. 2001: 100-110). Moreover, the outbreak of the Great Depression implied an affirmation for those counter-elites fighting against the oligarchic regime and in particular for a new generation of military officers trained by Prussian instructors, which were very critical about the parliamentary system (Loveman 1999: 82-85). Albeit this manifestation of the armed forces, the particularity of the Chilean development after the Great Depression lies in the maintenance of a civil control of the military and the practice of a gradualist policy, i.e. the establishment of an agreement under which a step by step elite renewal was supported, implying at the same time that those actors located at both the right and left extremes were excluded (Angell 1998: 94-95). In this sense, it is possible to state that in Chile a partial elite-settlement took form after the second administration of Arturo Alessandri starting in 1925, guaranteeing a stable political development for four decades. A good example of this is that until the putsch of 1973 both the constitution and the realisation of free elections were respected in Chile (Hartlyn/Valenzuela 1998: 24). Behind this partial elite-settlement two pillars were decisive. On the one hand, the political system was characterized by an extended use of patronage – particularly in the rural world – and a gradual expansion of the electorate (Drake 1993: 91-93). On the other hand, the elites of the oligarchic regime tolerated a slow rise of new elites from the middle classes and also promoted their co-optation through marriage and state appointments (Correa 2004: 93-100; Correa et al. 2001: 188-193). This partial elite-settlement came, however, to an end not only due to the influence of the Cold War and the intervention of the USA, but also and especially because of the increasing disagreement within the elites (Bermeo 2003: 138-176). In fact, the election of Allende in 1970 represents a sort of chronicle of a death foretold: since he obtained 44 percent of the votes and the political system was presidential and not parliamentarian, Allende governed without having a sustainable coalition which could guarantee the democratic order and stop the polarisation of society (Valenzuela 1989).

21

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

Finally, the national trajectory of Mexico is very interesting and different to the one of the other countries analysed, since it is marked by an elite-settlement and the establishment of a solid hegemony. To understand this, it is important to consider that before the Great Depression Mexico was confronted with a double warlike experience: the Mexican Revolution between 1910 and 1920 as well as the Cristero War between 1926 and 1929. Both experiences dislocated the oligarchic regime prior to the Great Depression and paved the way for the formation of a new power constellation vis-à-vis a new generation of elites. The Mexican Revolution, initiated in 1910 as a result of Francisco Madero’s opposition to the continuous re-election of Porfirio Díaz, counted on the armed resistance of liberal circles of the northern states of the country coming from the middle classes (Meyer 1991: 201). However, plebeian forces took an active part in the Revolution since its beginnings, particularly under command of Emiliano Zapata and Francisco Villa17. The Mexican Revolution was indeed a civil war with a very complicated course, in which not only local forces, but also international actors played a key role (Tobler 2007: 251-252). Concerning the Cristero War, its origins lies in a long-lasting dispute between the Mexican State and the Catholic Church and also on a negligent strategy of President Plutarco Elías Calles, who carried out the conflict with the catholic authorities to an extreme. Hence the Church mobilized their constituency against the state and a new civil war arose, so that the Mexican society was again deep disunited (Meyer 2007). After three year of asymmetrical warfare and several diplomatic intermediation trials, a pact between the Catholic Church and the Mexican state was made, in which an amnesty was declared and a new modus vivendi between both entities was defined (Camp 1998: 49). But why are the Mexican Revolution and the Cristero War significant for our analysis? Both had a fundamental impact on the further development of the country and not only because of the deterioration of the economy and the death toll – approximately one million inhabitants (Scheina 2003b: 12) –, but also and foremost due to its consequence in relation to coping with the pressure towards democratization. As a result of both warlike experiences, a great part of the elites of the oligarchic regime disappeared and new elites came into power (Camp 1992, 2006; Krauze 2004a). For this partially new generation of leaders, the main challenge was twofold. First, the establishment of a political rule capable of impeding the continuous re-election of the President. Second, the formation of a regime, in which different forces of the country could have voice and vote. For this purpose, the configuration of the »Partido Revolucionario Institutional« (PRI) was crucial, given that it embodied the constitution of an institutional platform of corporative nature, in which not only political participation was materialized, but also the achievement of a civil control of the military was enabled (Centeno 1994: 51-52; Meyer 2000: 900; Smith 1991: 321). In this sense, it is possible to state that the rise of this party symbolises a successful elite-settlement (Gannon 1992: Knight 1992a, 1998b). The party actually facilitated a systematic co-optation of potential counter-elites and was central for the establishment of a durable and solid hegemony lasting without big troubles until the declaration of bankruptcy of the Mexican state in 1982. Accordingly, the Mexican trajectory after the Great Depression is characterized by a very stable political game, in which political disputes due to the Cold War and the pressure towards democratization was much better mediated then in the other examined cases (Hernández 2002: 22).

As the historian Enrique Krauze (2004a) rightly reveals, the rise of Emiliano Zapata and Francisco Villa represents a mobilisation of broader population groups, which until then were excluded and through the revolution enforced their claims of social inequality, especially concerning the access to land property. Both generals were indeed caudillos, which were far beyond the liberal ideas of other revolutionary leaders and waged a guerrilla war. In this sense, Eric Hobsbawm (2007) correctly notices that Emiliano Zapata and Francisco Villa should be seen as bandits, which were not trying to enrich themselves, but rather to fight for the rights of the poor. In fact, the rise of both leaders represented a major challenge for the established elites of the oligarchic order. In this regard, it is not surprising that a myth arose around their figures, which was very useful for the later legitimatization of the »Partido Revolucionario Institucional« (Florescano 2005: 168-197). 17

22

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

4.3. Path-dependent developments in the 20th century and their historical legacies: bureaucratic-authoritarian and semi-authoritarian regimes Having revised the national trajectories that followed the downfall of the oligarchic regime in the countries under analysis, this section summarizes the path-dependent developments and their historical legacies. In the case of Argentina, the Great Depression implied a reorganisation of the political forces, which culminated in the consolidation of the established elites of the oligarchic regime and thus favoreded the later appearance of Peronism. This populist experiment provoked a deep division within Argentinean society and exacerbated the conflict between elites and counter-elites. In this sense, a path-dependent development of elite disunity and precarious hegemony took form. The pressure towards democratization implied therefore a major challenge, since there was no stable institutional pattern capable to offer neither an effective nor a gradual solution to the lower-class incorporation into the political arena. This was a fertile soil for the continuous intervention of the military, so that a back and forth of democratic and authoritarian regimes occurred. In fact, the rise of Peronism involved a dramatic radicalization of the friend and foe schema, impeding the establishment of both a party system able to represent the different opinions within society and common rules of the game for the interplay between elites and counter-elites. Increasing political polarisation and no capacity for conflict regulation were the corollary of a path-dependent development, which leaded to a blind alley. The cases of Chile and Brazil exemplify another pattern of how the elites did handle the pressure towards democratization. Albeit the Great Depression precipitated the downfall of the oligarchic regime, the established elites were still very powerful – due in part to its clientelistic control of the countryside – and thus the counter-elites opted to make a coalition with them. As a consequence, the process of elite renewal was slow and the state was reshaped in order to impose a new political rule characterized by a compromise between traditional oligarchy and the newer reformist middle sectors. Accordingly, in the cases of Brazil and Chile a path-dependent development of partial elite-settlement and gradual erosion of the hegemony was undertaken. The distinctiveness of this development lies in the protection of the economic interests of the oligarchy, which in conjunction to the introduction of modernization processes with the aid of the state fostered a slow opening of the political system. However, with the time this opening came to a dead end, since the capacity to rule trough consent and mediation had structural limits, particularly due to the increasing political fragmentation and polarisation. The erosion of the political center and the intensification of class antagonisms were further exacerbated by the impact of the Cold War and the increasing problems of the import substitution industrialization, which altogether undermined the hegemony achieved through agreement between the established elites and counter-elites. The two revised patterns – elite disunity and precarious hegemony on the one hand, and partial elite-settlement and gradual erosion of the hegemony on the other – set into motion a particular development that ultimately culminated in the same historical legacy, namely what Guillermo O’Donnell (1972) called a »bureaucratic-authoritarian regime«. All these countries actually experienced the appearance of brutal military dictatorships: Brazil in the year 1964 with the rise of General Castelo Branco, Chile in the year 1973 with the putsch of Pinochet against Allende and Argentina in the year 1976 with the seizure of power of the military junta commanded by Videla. To understand this general trend, it is important to notice the particularity of the so-called bureaucratic-authoritarian regimes. Whilst former governments implemented an economic policy based on the state and tried to maintain a certain level of political openness, the new military regimes relied on the introduction of economical neoliberalism and did not have any interest in sustaining neither political participation nor mass mobilisation (Cardoso 1979: 35-36). Moreover, these regimes counted with the support of the U.S. government and its »doctrine of national security«; an ideology that was inculcated to the Latin American armed forces by the U.S. military, which propagated the idea of proliferation of internal foes due to the expansion of communism and the extension of the Cold War (Loveman 1999: 165-194; Rouquié 1984: 134171). In this regard, it is possible to state that the emergence of bureaucratic-authoritarian re23

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

gimes in Argentina, Brazil and Chile represents a crisis of hegemony: »[…] intra-elite struggles and the weakness of the civil society caused civilian groups to appeal to factions of the military for intervention on their behalf. Dissident elite sectors and the emerging middle classes in many cases appealed for military support in their efforts to gain a share of political power, and the ruling groups frequently relied on the military to squash such challenges« (Rueschemeyer/Huber Stephens/Stephens 1992: 230). An entirely different development prevailed in Mexico. By contrast to our previous examples, this country experienced an abrupt elite renewal as a consequence of two civil wars – the Mexican Revolution and the Cristero War –, which took place before the Great Depression. As a consequence of this particular historical circumstance and some further events (e.g. the formation of the »Partido Nacional Revolucionario« and later the »Partido Revolucionario Institucional«) Mexico is the only case in our analysis, in which the downfall of the oligarchic regime and the pressure towards democratization led to the configuration of a stable political game. The latter was characterized by two features: the military was kept under civil control and the governments in power could maintain a relatively high level of legitimacy. Therefore, we characterize the Mexican path-dependent development as elite-settlement and solid hegemony. By saying this, we are neither glorifying the Mexican trajectory after the Great Depression nor its democratic development. Indeed, the long hegemony of the »Partido Revolucionario Institucional« (PRI) is a classical example of hybrid-regime which – in contemporary political jargon – can be labelled as an electoral autocracy (Diamond 2002), i.e. a regime of a hegemonic party system, in which a relatively institutionalized ruling party monopolizes the political arena, using coercion, patronage, media control, and other means to deny formally legal opposition parties any real chance of competing for power. Nevertheless, this party was a decisive instance to avoid the appearance of the kind of military dictatorships that emerged in Argentina, Brazil and Chile. Three path-dependent developments in the Latin American downfall of the oligarchic regime and their respective historical legacy Elite disunity and precarious hegemony

Argentina Initial perpetuation of the oligarchic regime and later emergence of Peronism, which impeded the formation of a durable hegemony and fostered a continuous intervention of the military until the Falkland War in 1982

Partial elite-settlement and gradual erosion of the hegemony

Brazil Rise of Vargas, formation of the »Estado Novo« and gradual erosion of the achieved hegemony due to the growing intensity of the Cold War and unsolved disputes within the elites, which led to the putsch against Goulart in 1964

Chile An increasing concern about the social question culminated in a new elite arrangement, in which counter-elites of the middle classes were coopted. Therefore a certain level of hegemony was achieved at least until the contested election of Allende in 1970

Bureaucratic-Authoritarian Regime

Elite-settlement and solid hegemony

Mexico After the Mexican Revolution an elitesettlement was achieved with the aid of the »Partido Revolucionario Institucional« (PRI). Consequently, a solid hegemony took form and the military was kept under civil control

Semi-Authoritarian Regime

24

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

As we have seen in this last section of the paper, the pressure towards democratization was a key issue in Latin America during the 20th century, in particular with the onset of the Great Depression and the downfall of the oligarchic regimes. This situation entailed a far-reaching process of elite renewal. Albeit its scope and velocity varied from country to country, in the four analysed cases – Argentina, Brazil, Chile and Mexico – the oligarchic elites declined and new elites belonging to the middle classes emerged. The best example of this is what happened in Argentina in the course of the coming to power of Perón, where the rise of counter-elites that for long time had been criticising the established order built a new political order based on the mobilisation of a great part of the popular sector. On the other hand, in the case of Mexico the formation of the »Partido Revolucionario Institucional« (PRI) implied not only a mechanism for the resolution of conflicts between different minorities of power, but also a system for the continuous formation and renewal of the elites. In effect, the consolidation of the PRI involved the configuration of what Enrique Krauze (2004b) calls an imperial six-year-old presidential term without re-election, i.e. a political regime characterized by the rise of a powerful president accompanied by its own elites, which after the handover of power were mostly predestined to lose ground, since they had to give place to a new president with his own establishment. Finally, in the cases of Chile and Brazil, the process of elite circulation was gradual and occurred primarily thanks to a co-optation system activated by the elites of the oligarchic order. The emergence of a bureaucratic-authoritarian regime implied, however, not only in Brazil and Chile but also in Argentina, a new impulse towards elite circulation, since great part of the established elites went down and, at the same time, new actors (e.g. entrepreneurs, military officials, technocrats, etc.) made appearance. 5. CONCLUSION This article has applied classical elite theory and comparative historical analysis to study the post-colonial history of Latin America. First of all, we argued that one of the main ideas of the founders of elite theory was that to grasp processes of elite circulation it is necessary to take long periods of time into account. Nevertheless, they did not offer a systematic approach about how to deal with a »long-durée« perspective. To address this research gap, we drew on comparative historical analysis. Special reference was made to the framework elaborated by Mahoney (2001b) and Collier and Collier (2002), which allowed for the identification of different stages within historical sequences – antecedent conditions, cleavage, critical juncture, path-dependent development and historical legacy –, through which we identified the variation of the factual room of manoeuvre of the elites. Based on this theoretical discussion, we analysed the post-colonial history of Argentina, Brazil, Chile and Mexico with the aim of bringing two ideas to the forefront. On the one hand, in Latin America the process of political independence was elite-driven, so that in reference to the 19th century, we can reconstruct path-dependent developments which varied depending on the capacity of the established elites and the counter-elites to enhance their horizontal integration in order to set up a new regime. On the other hand, the response of the elites to the downfall of the oligarchic regime was quite different amongst the countries under analysis, and also during the 20th century it is possible to distinguish diverse path-dependent developments depending on the elites’ level of horizontal integration and their ability to manage the pressure towards democratization. After this brief historical analysis of the Latin American 19th and 20th centuries, one thesis should be clear: it is flawed to assume that the history of this region of the world is characterized by low levels of elite circulation. Rather, if we take long periods of time into consideration, we can claim that Latin America exemplify a continuous process of elite circulation. Although this paper is based on the study of the post-colonial history of Latin America, we are of the opinion that the offered analysis is of interest not only for experts on Latin America, but also for those who are dealing with comparative historical analysis and/or elite theory. For the first ones, we have shown that even though elites are key actors for the promotion or hin25

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

drance of reform processes, their factual room of manoeuvre is related with the moment within a historical sequence in which they are operating, i.e. whilst the scope of action of the elites is broad at the critical junctures, it is strongly diminished once a path-dependent development is put into motion and a particular historical legacy is formed. On the other hand, for scholars interested in elite theory, we have revealed that the greatest heuristic potential of elite theory lies in the examination of long periods of time, since elite circulation is usually a slow-moving causal process. In consequence, if we use the elite concept only to describe the current situation of a society and we do not take its history into account, we may get a somewhat distorted view, in which there is little space to grasp the contingency and dynamics of elite renewal. Here lies in fact the quintessence of Pareto’s famous sentence: history is a graveyard of aristocracies. 6. LITERATURE Anderson, Benedict (1996), Die Erfindung der Nation. Zur Karriere eines folgenreichen Konzepts, Frankfurt a.M. Angell, Alan (1998), »The Left in Latin America since c. 1920«, in: Leslie Bethell (Ed.), Latin America. Politics and Society since 1930, Cambridge. Anna, Timothy (1987), »The Independence of Mexico and Central America«, in: Leslie Bethell (Ed.), The Independence of Latin America, Cambridge. Bach, Maurizio (2004), Jenseits des rationalen Handelns. Zur Soziologie Vilfredo Paretos, Wiesbaden. Balmori, Diana/Voss, Stuart F./Wortman, Miles (1984), Notable Family Networks in Latin America, Chicago/London. Bazant, Jan (1991), »From Independence to Liberal Republic, 1821–1867«, in Leslie Bethell (Ed.), Mexico since Independence, Cambridge. Bernecker, Walther L. (2000), »Kolonie – Monarchie – Republik: Das 19.Jahrhundert«, in: Walther L. Bernecker/Horst Pietschmann/Rüdiger Zoller (Ed.), Eine kleine Geschichte Brasiliens, Frankfurt a.M. — (2007), »Mexiko im 19. Jahrhundert: Zwischen Unabhängigkeit und Revolution«, in: Walther L. Bernecker/Horst Pietschmann/Hans Werner Tobler (Ed..), Eine kleine Geschichte Mexikos, Frankfurt a.M. Bethell, Leslie (1993), »The Independence of Brazil«, in: Leslie Bethell (Ed..), Brazil: Empire and Republic, 1822–1930, Cambridge. Bethell, Leslie/Carvalho, José Murilo de (1993), »Empire: 1822–1850«, in: Leslie Bethell (Ed.), Brazil: Empire and Republic, 1822–1930, Cambridge. Bermeo, Nancy (2003), Ordinary People in Extraordinary Times. The Citizenry and the Breakdown of Democracy, Princeton. Beyer, Jürgen (2005), »Pfädabhängigkeit ist nicht gleich Pfadabhängigkeit! Wider den impliziten Konservatismus eines gängigen Konzepts«, in: Zeitschrift für Soziologie, 34, 1, 5–21. Beyhaut, Gustavo/Beyhaut, Hélène (2000), América Latina III. De la independencia a la guerra mundial, México D.F. Bottomore, Thomas B. (1966), Elite und Gesellschaft. Eine Übersicht über die Entwicklung des Eliteproblems, München. Braudel, Fernand (2002), Las ambiciones de la historia, Barcelona. Burnham, James (1949), Die Machiavellisten. Verteidiger der Freiheit, Zürich. Burton, Michael/Gunther, Richard/Higley, John (1992), »Introduction: Elite Transformations and Democratic Regimes«, in: John Higley/Richard Gunther (Ed.), Elites and Democratic Consolidation in Latin America and Southern Europe, Cambridge. Bushnell, David (1987), »The Independence of Spanish South America«, in: Leslie Bethell (Ed.), The Independence of Latin America, Cambridge. Camargo, Aspásia (1993), »La federación sometida. Nacionalismo desarrollista e inestabilidad democrática«, in: Marcello Carmagnani (Ed.), Federalismos latinoamericanos: México / Brasil / Argentina, México D.F. 26

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

Camp, Roderic Ai (1992), Generals in the Palacio. The Military in Modern Mexico, New York/Oxford. — (1998), Cruce de espadas. Política y religión en México, México D.F. — (2006), Mexico's Mandarins. Crafting a Power Elite for the Twenty-First Century, Berkeley/Los Angeles. Capoccia, Giovanni/Kelemen, Daniel R. (2007), »The Study of Critical Junctures. Theory, Narrative, and Counterfactuals in Historical Institutionalism«, in: World Politics, 59, 3, 341-359. Cardoso, Fernando Henrique (1979), »On the Characterisation of the Bureaucratic-Authoritarian Model«, in: David Collier (Ed.), The New Authoritarianism in Latin America, Princeton. Cardoso, Fernando Henrique/Faletto, Enzo (1976), Abhängigkeit und Entwicklung in Lateinamerika, Frankfurt a.M. Carlyle, Thomas (1996), On Heroes. Hero Worship and the Heroic in History, London. Carmagnani, Marcello (1984), Estado y sociedad en América Latina, 1850–1930, Barcelona. — (2004), El otro occidente. América Latina desde la invasión europea hasta la globalización, México D.F. Carvalho, José Murilo de (1980), A construção da ordem. A elite política imperial, Rio de Janeiro. — (1981), Armed Forces and Politics in Brazil, 1930–1945, Washington D.C. — (2005), Cidadania no Brasil. O longo caminho, Rio de Janeiro. Cavarozzi, Marcelo (1996), El capitalismo político tardío y su crisis en América Latina, Rosario. Centeno, Miguel Ángel (1994), Democracy within Reason: Technocratic Revolution in Mexico, University Park. — (2002), Blood and Debt: War and the Nation-State in Latin America, University Park. — (2003), »Limited Wars and Limited States«, in: Diane E. Davis/Anthony Pereira W. (Ed.), Irregular Armed Forces and their Role in Politics and State Formation, Cambridge. Centeno, Miguel Ángel/López-Alves, Fernando (2001), »Introduction«, in: Miguel Ángel Centeno/Fernando López-Alves (Ed.), The Other Mirror: Grand Theory through the Lens of Latin America, Princeton. Collier, Simon/Sater, William F. (2004), A History of Chile, 1808–2002, Cambridge. Collier, Ruth Berins/Collier, David (2002), Shaping the Political Arena. Critical Junctures, the Labor Movement, and Regime Analysis in Latin America, Notre Dame. Correa, Sofía (2004), Con las riendas en el poder. La derecha chilena en el siglo XX, Santiago de Chile. Correa, Sofía/Figueroa, Consuelo/Jocelyn-Holt, Alfredo/Rolle, Claudio/Vicuña, Manuel (2001), Historia del siglo XX chileno, Santiago de Chile. Cortés Conde, Roberto (1993), »The Growth of the Argentine Economy, c. 1870–1914«, in: Leslie Bethell (Ed.), Argentina since Independence, Cambridge. de Imaz, José Luis (1965), Los que mandan, Buenos Aires. de Oliveira, Orlandina/Roberts, Bryan (1998), »Urban Social Structures in Latin America, 1930– 1990«, in: Leslie Bethell (Ed.), Latin America. Economy and Society since 1930, Cambridge. Dezalay, Yves/Garth, Bryant G. (2002), The Internationalization of Palace Wars: Lawyers, Economists, and the Contest to Transform Latin American States, Chicago. Diamond, Larry (2002), »Elections without Democracy. Thinking About Hybrid Regimes«, in: Journal of Democracy, 13, 2, 21–35. Dogan, Mattei/Higley, John (1998), »Elites, Crisis and Regimes in Comparative Analysis«, in: Mattei Dogan/John Higley (Ed.), Elites, Crisis and the Origins of Regimes, Oxford. Dogan, Mattei (2003), »Is there a Ruling Class in France?«, in: Mattei Dogan (Ed.), Elite Configurations at the Apex of Power, Leiden/Boston. Drake, Paul (1993), »Chile, 1930–1958 «, in: Leslie Bethell (Ed.), Chile since Independence, Cambridge. Fausto, Boris (1999), A Concise History of Brazil, New York. Fausto, Boris/Devoto J., Fernando (2004), Brasil e Argentina. Um ensaio de história comparada (1850– 2002), São Paulo. Florescano, Enrique (2005), Imágenes de la patria a través de los siglos, México D.F.

27

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

Gannon, Renee Rogero (1992), Elites, Political Stability and Democratization in Tweentieth Century Mexico, Doctoral dissertation presented to the Faculty of Graduate School of the University of Texas at Austin. Germani, Gino (1971), Política y sociedad en una época de transición. De la sociedad tradicional a la sociedad de masas, Buenos Aires. — (1973), »Democracia representativa y clases populares«, in: Torcuato di Tella/ Gino Germani/Octavio Ianni (Ed.), Populismo y contradicciones de clase en Latinoamérica, México D.F. — (2003), Autoritarismo, fascismo y populismo nacional, Buenos Aires. Glade, William (2002), »América Latina y la economía internacional, 1870–1914«, in: Leslie Bethell (Ed.), Historia económica de América Latina. Desde la independencia hasta nuestros días, Barcelona. Goldstone, Jack A. (1998), »Initial Conditions, General Laws, Path Dependence, and Explanation in Historical Sociology«, in: American Journal of Sociology, 104, 3, 829–845. Graciarena, Jorge (1967), Poder y clases sociales en el desarrollo de América Latina, Buenos Aires. Graham, Richard (1990), Patronage and Politics in Nineteenth-Century Brazil, Standford. — (1993), »Empire: 1850–1870«, in: Leslie Bethell (Ed.), Brazil: Empire and Republic, 1822–1930, Cambridge. Halperin Donghi, Tulio (2001), Historia contemporánea de América Latina, Madrid. — (2002), »América Latina independiente. Economía y sociedad«, in: Leslie Bethell (Ed.), Historia económica de América Latina. Desde la independencia hasta nuestros días, Barcelona. — (2005), Una nación para el desierto argentino, Buenos Aires. — (2006), »Dos siglos de reflexiones sudamericanas sobre la brecha entre América Latina y Estados Unidos«, in: Francis Fukuyama (Ed.), La brecha entre América Latina y Estados Unidos. Determinantes políticos e institucionales del desarrollo económico, México D.F. Hartlyn, Jonathan/Valenzuela, Arturo (1998), »Democracy in Latin America since 1930«, in: Leslie Bethell (Ed.), Latin America. Politics and Society since 1930, Cambridge. Hernández, Rogelio (2002), »Politische Führung und politische Elite in Mexiko: Von der autoritären Herrschaft zur demokratischen Wende«, in: Wilhelm Hofmeister (Ed.), Gebt mir einen Balkon und das Land ist mein: Politische Führung in Lateinamerika, Frankfurt a.M. Higley, John/Burton, Michael (1989), »The Elite Variable in Democratic Transitions and Breakdowns«, in: American Sociological Review, 54, 1, 17–32. — (2006), Elite Foundations of Liberal Democracy, Lanham. Hobsbawm, Eric (1999a), La era de la revolución, 1789–1848, Buenos Aires. — (1999b), La era del imperio, 1875–1914, Buenos Aires. — (2007), Die Banditen. Räuber als Sozialrebellen, München. Hoffmann-Lange, Ursula (1992), Eliten, Macht und Konflikt in der Bundesrepublik Deutschland, Opladen. Horowicz, Alejandro (1985), Los cuatro peronismos, Buenos Aires. Joas, Hans/Knöbl, Wolfgang (2004), Sozialtheorie. Zwanzig einführende Vorlesungen, Frankfurt a.M. — (2008), Kriegsverdrängung. Ein Problem in der Geschichte der Sozialtheorie, Frankfurt a.M. Jocelyn-Holt, Alfredo (1999), La independencia de Chile: tradición, modernización y mito, Santiago de Chile. Kaina, Viktoria (2002), Elitenvertrauen und Demokratie. Zur Akzeptanz gesellschaftlicher Führungskräfte im vereinten Deutschland, Wiesbaden. Karl, Terry Lynn (1991), »Dilemas de la democratización en América Latina«, in: Carlos Solano Barba/José Luis Barros Horcasitas/Javier Hurtado (Ed.), Transiciones a la democracia en Europa y en América Latina, México D.F. — (1997), The Paradox of Plenty. Oil Booms and Petro-States, Berkeley/Los Angeles. Katz, Friedrich (1991), »The Liberal Republic and the Porfiriato, 1867–1910«, in: Leslie Bethell (Ed.), Mexico since Independence, Cambridge.

28

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

Kneer, Wolfgang (2004), »Differenzierung bei Luhmann und Bourdieu. Ein Theorienvergleich«, in: Armin Nassehi/Gerd Nollmann (Ed.), Bourdieu und Luhmann. Ein Theorienvergleich, Frankfurt a.M. Knight, Alan (1992a), »Mexico's Elite Settlement: Conjuncture and Consequences«, in: John Higley/Richard Gunther (Ed.), Elites and Democratic Consolidation in Latin America and Southern Europe, Cambridge. — (1998a), »Historical and Theoretical Considerations«, in: Mattei Dogan/John Higley (Ed.), Elites, Crisis and the Origins of Regimes, Oxford. — (1998b), »Mexico and Latin America in Comparative Perspective«, in: Mattei Dogan/John Higley (Ed.), Elites, Crisis and the Origins of Regimes, Oxford. Knight, Franklin W. (1992b), »The State of Sovereignty and the Sovereignty of States«, in: Alfred Stepan (Ed.), Americas. New Interpretative Essays, New York. Krauze, Enrique (2004a), Biografía del poder. Caudillos de la revolución mexicana (1910–1940), México D.F. — (2004b), La presidencia imperial. Ascenso y caída del sistema político mexicano (1940–1996), México D.F. — (2004c), Siglo de Caudillos. Biografía política de México (1810–1910), México D.F. Krebs, Ricardo (2002), La iglesia de América Latina en el siglo XIX, Santiago de Chile. Lipset M., Seymour/Rokkan, Stein (1967) (Ed.), Party Systems and Voter Alignments: Cross-National Perspectives, New York/London. Loveman, Brian (1999), For la patria, Wilmington. Lynch, John (1987), »The origins of Spanish American Independence«, in: Leslie Bethell (Ed.), The Independence of Latin America, Cambridge. — (1992), Caudillos in Spanish America, 1800–1850, Oxford. Mahoney, James (2000), »Path Dependence in Historical Sociology«, in: Theory and Society, 29, 4, 507–548. — (2001a), »Path-Dependent Explanations of Regime Change: Central America in Comparative Perspective«, in: Studies in Comparative International Development, 36, 1, 111–141. — (2001b), The Legacies of Liberalism. Path Dependence and Political Regimes in Central America, Baltimore. Mahoney, James/Rueschemeyer, Dietrich (2003) (Ed.), Comparative Historical Analysis in the Social Sciences, Cambridge. Mann, Michael (1986), The Sources of Social Power, Volume 1: A History of Power from the Beginning to AD 1760, Cambridge. — (1988), States, War and Capitalism. Studies in Political Sociology, Oxford. — (1993), The Sources of Social Power, Volume 2: The Rise of Classes and Nation States, 1760-1914, Cambridge. — (2004), Fascists, Cambridge. Martucelli, Danilo/Svampa, Maristella (1997), La plaza vacía. Las transformaciones del peronismo, Buenos Aires. Medina Echavarría, José (1964), Consideraciones sociológicas sobre el desarrollo económico, Buenos Aires. Merkel, Wolfgang/Puhle, Hans-Jürgen (1999), Von der Diktatur zur Demokratie. Transformationen, Erfolgsbedingungen, Entwicklungspfade, Opladen. Merrick, Thomas W. (1998), »The Population of Latin America, 1930–1990«, in: Leslie Bethell (Ed.), Latin America. Economy and Society since 1930, Cambridge. Meyer, Jean (1991), »Revolution and Reconstruction in the 1920s«, in: Leslie Bethell (Ed.), Mexico since Independence, Cambridge. — (2007), La Cristiada, México D.F. Meyer, Lorenzo (2000), »De la estabilidad al cambio«, in: Centro de Estudios Históricos (Ed.), Historia general de México, México D.F. Moore, Barrington Jr., (1966), Social Origins of Dictatorship and Democracy. Lord and Peasant Revolution in the Making of the Modern World, Boston. 29

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

Mosca, Gaetano (1966), The Ruling Class, New York. Murmis, Miguel/Portantiero, Juan Carlos (2004), Estudios sobre los orígenes del peronismo, Buenos Aires. Münkler, Herfried (2005), Imperien. Die Logik der Weltherrschaft – vom Alten Rom bis zu den Vereinigten Staaten, Berlin. Münkler, Herfried/Straßenberger, Grit/ Bohlender, Matthias (2006), »Einleitung« in: Matthias Bohlender/Grit Straßenberger/Herfried Münkler (Ed.), Deutschlands Eliten im Wandel, Frankfurt a.M./New York. North, Douglass C. (1990), Institutions, Institutional Change and Economic Performance, Cambridge. Nunn, Frederick M. (1986), Yesterday's Soldiers: European Military Professionalism in South America, 1890–1940, Lincoln/London. — (1997), »An Overview of the European Military Missions in Latin America«, in: Brian Loveman/Thomas M. Davies (Ed.), The Politics of Antipolitics, Wilmington. O'Donnell, Guillermo (1972), Modernización y autoritarismo, Buenos Aires. O'Donnell, Guillermo/Schmitter, Philippe C. (1986), Transitions from Authoritarian Rule. Tentative Conclusions about Uncertain Democracies, Baltimore. Oszlak, Oscar (1981), »The Historical Formation of the State in Latin America: Some Theoretical and Methodological Guidelines for its Study«, in: Latin American Research Review, 16, 2, 3–32. Pareto, Vilfredo (1923), Trattato di sociologia generale, Firenze. Paris, Rainer (2003), »Autorität – Führung – Elite: Eine Abgrenzung«, in: Stefan Hradil/Peter Imbusch (Ed.), Oberschichten – Eliten – herrschende Klassen, Opladen. Parry, Geraint (2005), Political Elites, Essex. Peters, Guy B. (1999), Institutional Theory in Poltical Science: The New Institutionalism, London/New York. Peters, Guy B./Pierre, Jon/King, Desmond S. (2005), »The Politics of Path Dependency: Political Conflict in Historical Institutionalism«, in: The Journal of Politics, 67, 4, 1275–1300. Pierson, Paul/Skocpol, Theda (2002), »Historical Institutionalism in Contemporary Political Science«, in: Ira Katznelson/Helen V. Milner: Political Science: The State of the Discipline, New York. Pierson, Paul (2004), Politics in Time. History, Institutions, and Social Analysis, Princeton. Potash, Robert A. (1981), El ejército y la política en la Argentina 1928–1945, Buenos Aires. Rebenstorf, Hilke (1997), »Karrieren und Integration – Werdegänge und Common Language«, in: Wilhelm Bürklin/Hilke Rebenstorf (Ed.), Eliten in Deutschland. Rekrutierung und Integration, Opladen. Reinhard, Wolfgang (2007), Geschichte des modernen Staates: Von den Anfängen bis zur Gegenwart, München. Rock, David (1993), »Argentina, 1930–1946«, in: Leslie Bethell (Ed.), Argentina since Independence, Cambridge. — (2006), La construcción del Estado y los movimientos políticos en la Argentina, 1860–1916, Buenos Aires. Romero, José Luis (2001), Latinoamérica. Las ciudades y las ideas, Buenos Aires. Romero, Luis Alberto (2006), Breve historia contemporánea de la Argentina, Buenos Aires. Rouquié, Alain (1984), El Estado militar en América Latina, México D.F. — (1998), »The Military in Latin American Politics since 1930«, in: Leslie Bethell (Ed.), Latin America. Politics and Society since 1930, Cambridge. Rueschemeyer, Dietrich/Huber Stephens, Evelyne/Stephens, John D. (1992), Capitalist Development & Democracy, Chicago. Sauer, Martina (2000), Gesellschaftliche Steuerungschancen durch Elitenintegration? Kommunikation und Kooperation bundesdeutscher Funktionsträger vor dem Hintergrund funktionaler Differenzierung, Opladen. Scheina, Robert L. (2003a), Latin American's Wars. Volume 1: The Age of the Caudillo, 1791–1899, Washington D.C.

30

Towards a Historical Analysis of Elites in Latin America / Cristóbal Rovira Kaltwasser

— (2003b), Latin American's Wars. Volume 2: The Age of the Professional Soldier, 1900–2001, Washington D.C. Schmitt, Carl (1932), Der Begriff des Politischen, Berlin. Sidicaro, Raúl (2004), »Coaliciones golpistas y dictaduras militares: el ›proceso‹ en perspectiva comparada«, in: Alfredo Raúl Pucciarelli (Ed.), Empresarios, tecnócratas y militares. La trama corporativa de la última dictadura, Buenos Aires. Skocpol, Theda (1979), States and Social Revolutions. A Comparative Analysis of France, Russia, and China, Cambridge. — (2003), »Doubly Engaged Social Science: The Promise of Comparative Historical Analysis«, in: James Mahoney/Dietrich Rueschemeyer (Ed.), Comparative Historical Analysis in the Social Sciences, Cambridge. Smith, Peter H. (1974), Argentina and the Failure of Democracy. Conflict among Political Elites, 1904– 1955, Wisconsin. — (1979), Labyrinths of Power. Political Recruitment in Tweentieth-Century Mexico, Princeton. — (1991), »Mexico since 1946: Dynamics of an Authoritarian Regime«, in: Leslie Bethell (Ed.), Mexico since Independence, Cambridge. — (1992), »The State and Development in Historical Perspective«, in: Alfred Stepan (Ed.), Americas. New Interpretative Essays, New York. Steinmo, Sven (2008), »Historical Institutionalism«, in: Donatella della Porta/Michael Keating (Ed.), Approaches and Methodologies in the Social Sciences. A Pluralist Perspective, Cambridge. Thelen, Kathleen/Steinmo, Sven (1992), »Historical Institutionalism in Comparative Politics«, in: Sven Steinmo/Kathleen Thelen/Frank Longstreth (Ed.), Structuring Politics. Historical Institutionalism in Comparative Analysis, Cambridge. Tilly, Charles (1984), Big Structures, Large Processes, and Hugh Comparisons, New York. — (1985), »War Making and State Making as Organized Crime«, in: Peter B. Evans/Dietriech Rueschemeyer/Theda Skocpol (Ed.), Bringing the State Back In, Cambridge. — (1990), Coercion, Capital, and European States, AD 990–1990, Oxford. — (2008), Why?, Princeton. Tobler, Hans Werner (2007), »Mexiko im 20. Jahrhundert: Die Revolution und ihre Folgen«, in: Walther L. Bernecker/Horst Pietschmann/Hans Werner Tobler (Ed.), Eine kleine Geschichte Mexikos, Frankfurt a.M. Valenzuela, Arturo (1989), El quiebre de la democracia en Chile, Santiago de Chile. Vicuña, Manuel (2001), La belle époque chilena. Alta sociedad y mujeres de elite en el cambio de siglo, Santiago de Chile. Viotti da Costa, Emília (1993), »Empire: 1850–1870«, in: Leslie Bethell (Ed.), Brazil: Empire and Republic, 1870–1889, Cambridge. Weffort, Francisco (2003), O populismo na política brasileira, Rio de Janeiro. Whitehead, Laurence (1998), »A Note on Citizenship in Latin America since 1930«, in: Leslie Bethell (Ed.), Latin America. Politics and Society since 1930, Cambridge. Wright Mills, Charles (1956), The Power Elite, New York. Zeitlin, Maurice (1984), The Civil Wars in Chile (or the Bourgeois Revolutions that Never Were), New Yersey. Zeitlin, Irving (1994), Ideology and Development of Sociological Theory, New Jersey. Zoller, Rüdiger (2000), »Präsidenten – Diktatoren – Erlöser: Das lange 20. Jahrhundert«, in: Walther L. Bernecker/Horst Pietschmann/Rüdiger Zoller (Ed.), Eine kleine Geschichte Brasiliens, Frankfurt a.M.

31