Matias López Interdisciplinary Network for Studies of Social Inequality
(NIED), at the Federal University of Rio de Janeiro, Brazil
abstract
This article reviews contemporary elite theory in political sociology and political science. Elite theory is based on the assumption that elite behavior has a causal relationship with general patterns of state-society relations. The article presents classical concepts of elite theory, such as elite inevitability and elite circulation, while privileging contemporary challenges and trends in elite theory. The discussion addresses elite origins of democracy and elite origins of the welfare state, as well as elites/non-elites interdependence.
Overview
This article reviews contemporary elite theory in political sociology and political science. The concept of 'elites' is based on the notion that every society holds a ruling minority, a group that controls and disputes the most important power sources. Not only do elites dispute power (reaching different levels of conflict and violence), but new elites also enter the game through different mechanisms of elite recruitment.
Thus, the core of elite theory relies in explaining elite behavior, elite interaction, elite transformation and, ultimately, the connection between those instances and state outcomes.
The link between elite behavior and regime change led a number of authors to almost naturalize elite theory as 'regime theory' (e.g. Dogan and Higley, 1998; Higley and Burton, 2006). Yet elite theory is not restricted to the study of regime change and regime support. Although studies do tend to focus on the political realm, elite researchers often explore power relations inside the market and civil society as well. An important focus within elite research is the emergence of welfare states, a political phenomenon directly related with market regulation and the state's capacity to penetrate society (e.g. de Swaan, 1988; Reis and Moore, 2005; Verba and Orren, 1985; Verba et al., 1987). Elite theory also approaches the question of social development and modernization in both Western and non-Western contexts (e.g. Cardoso, 1964; de Swaan et al., 2000; Lipset and Solari, 1967; López, 2013; Reis and Moore, 2005).
Elite theory is deep-rooted in classical sociology, especially that of Weber (2005 [1922]), Pareto (1935), Mosca (1939) and Michels (2009 [1915]).
These authors are usually labeled as 'classical elitists'.
Beyond its strong roots in classical sociology, elite theory developed into a vibrant theoretical field, intersecting other theories, such as rational choice theory and political culture theory.
Most elite theory reviews (e.g. de Hollanda, 2011; Khan, 2012) focus on the work of classical elitists, considered to be the founding fathers of the elitist school. Classical authors are important references and I will briefly review their work as well, yet my intention is to privilege contemporary elite theory and research, highlighting empirical findings and current research challenges.
Classical and current elite theorists share the ambition of explaining state outcomes through elite behavior.
If political sociology is concerned with the relationship between state and society (see Sociopedia.isa article on political sociology by Botelho, 2011), elite theory is based on the assumption that elite action has a causal effect on such a relationship.
Thus, regime types, regime change, liberalization, stateness, secularization and many other political phenomena fit within the scope of elite theory.
Currently, there is an important effort to unify/synthesize elite theory (e.g. Field and Higley, 2012; Higley and Burton, 2006). This effort is often abstract This article reviews contemporary elite theory in political sociology and political science. Elite theory is based on the assumption that elite behavior has a causal relationship with general patterns of state-society relations. The article presents classical concepts of elite theory, such as elite inevitability and elite circulation, while privileging contemporary challenges and trends in elite theory. The discussion addresses elite origins of democracy and elite origins of the welfare state, as well as elites/non-elites interdependence.
keywords democracy ◆ elites ◆ elite theory ◆ new elitism ◆ welfare state Elite theory Matias López Interdisciplinary Network for Studies of Social Inequality (NIED), at the Federal University of Rio de Janeiro, Brazil 2 López Elite theory associated with the idea of a new elite paradigm, or simply new elitism. However, several other authors develop elite research (i.e. research that uses elite behavior to explain political outcomes) without labeling their work as 'elitism'. For this reason I will mostly refer to elite theory, including authors that identify with elitism and others that make no mention of it, while doing relevant elite research.
Elite theory is often treated as contrasting social class theory and other structural approaches. Critics claim that the assumption that elite behavior holds a causal effect on state outcomes suggests a voluntaristic argument, overlooking structure (Cammack, 1990; Collier, 1999). In this article I argue that this is a caricature of elite theory. Being deep-rooted in classical sociology, elite theory is very much concerned with structures, especially authority structures.
Furthermore, elite theorists have incorporated important features of social class theory into elite theory (e.g. de Swaan 1988, 2005; Domhoff, 2009 [1967]; Mills, 1956; Reis and Moore, 2005).
However, the idea that elite theory is opposed to social class theory does not comes out of nowhere. It comes from the classical debate over the causes of political regimes in pre-World War II sociology and was later on reinforced by important names in Marxism, such as Poulantzas (apud Codato and Perissinotto, 2009). Nevertheless, some key Marxist authors such as Mills (1956) and Domhoff (2009 [1967]) have made extensive use of elite theory.
Currently, elite literature presents concerns over both elite action's effect on structure and structural constraints for elite action. Since elite behavior is particularly hard to measure, elite research has engaged in sophisticated and clever methodological tools, ranging from elite surveys to political ethnography.
Classical elitism and the law of the small number The core of classical elitism is the notion of elite inevitability. Classical elitists used the inevitability of elite rule as a premise to counter argue with political liberalism and Marxism. They claimed that both democracy (as the government of the people or demos) and socialism (as a classless regime) were impossible outcomes because society is necessarily elite driven. In the elitist view, elites could only be substituted by another set of elites, meaning that the majority is necessarily ruled by a minority. This notion was expressed in Pareto's (1935) law of elite circulation, in Mosca's (1939) notion of political class and in Michels' (2009 [1915]) iron law of oligarchy.
The principle or law of elite circulation holds that elites alternate in power as a result of either peaceful or violent competition. In Pareto's (1935) terms, history is (and could only be) nothing but a 'cemetery of elites'. The notion of political class, which is currently less frequent in elite studies, defines elites as a class of political rulers, in opposition to a mass of followers.
Finally, the iron law of oligarchy synthesizes the notion of elite inevitability formulating it as a natural law.
Many consider Weber (2005 [1922]) to be a founding father of elitism as well, due to domination theory (Weber, 2005 [1922]: 695). Indeed Weber's concepts of power and domination, as well as his theoretical work on political parties and the related affirmation that social classes are not necessarily social actors (Weber, 2005 [1922]: 682), are fundamental pillars of contemporary elite theory. As I will argue later on, current elite theory often tends to be Weberian.
It is curious to notice that the two most influential authors in classical elitism, Pareto and Weber, were not primarily concerned with elite theory.
Weber's main theoretical interests are well known, yet Pareto is often reduced to elite theory, while he actually devoted most of his work to a general theory of social behavior. Pareto's general theory was followed closely by key authors in sociology, such as Parsons and Aron. Also, Pareto's work reaches beyond sociology and political science, being as well known (or even more known) for his work in economics.
Currently, some authors have recovered much of Pareto's contribution to political sociology and political science (see the works of Femina, 2006; Finer, 1968; Higley and Pakulski, 2012; Sartori, 1987).
Another interesting fact is that the first reception of elitism outside of Europe was not overall positive.
During the first half of the 20th century, and even afterwards, many scholars considered elitism to be anti-democratic in essence because of notions such as the law of the small number. Indeed, Pareto, Mosca and Michels were not good examples of democrats of their time, but there is nothing anti-democratic about elite theory per se, unless one considers realism and democracy as incompatible notions (Sartori, 1987). In current political sociology and political science, one of the main purposes of elite theory relies on finding the causes of democratic rule, thus admitting democracy and, implicitly, prizing it.
The concept of elites In the classical elitist literature, elites were often (but not only) defined through capacity, personality and skill. For instance, Pareto (1935) distinguished elites between those who resembled the lion (domination by force) and those who resembled the fox (domination by persuasion and skill), a typology that resembles Machiavelli's political philosophy. Mosca (1939) 3 López Elite theory made mention of material conditions, but also of the intellectual and moral superiority of elites.
Current elite theory defines 'elites' as actors controlling resources, occupying key positions and relating through power networks (Yamokoski and Dubrow, 2008). Thus, the state-of-the-art concept of elites is more closely related to the Weberian notion of power, understood as the capability of implementing one's will, even against the will of others (Weber, 2005 [1922]: 696). Power can be achieved through material and/or symbolic resources. Consequently, elites can be defined as those in possession of those resources (Reis and Moore, 2005).
This definition does not appear to be that distant from the Marxist notion of ruling class, which is based on the possession of the means of production.
However, the concept of elite power is not based on economic assumptions, it is based instead on political assumptions. Also, elite theory usually conceives the elite as a smaller group than social class.
In addition, elites can emerge among dissident movements and even from dominated classes or groups (Dogan and Higley, 1998). As Higley and Burton (2006: 7) argue, elites are persons occupying the top of powerful organizations and movements, thus capable of affecting political outcomes both substantially and regularly.
The current concept of elites is often criticized, because the notion of elites emerging from movements and organizations may lead to a vast array of elite sources, contradicting the notion of elites as a small group (Cammack, 1990). This is an important theoretical problem, because elite theory acknowledges that elites may come from anywhere, as long as they find the necessary tools to exercise power. Thus, elites may come from the state and the corporate worlds, but also from guerilla, unions, the media, NGOs, any kind of social movement and so on.
However, elite theory limits elites to a necessary minority. Therefore, provided with the necessary tools, anyone but not everyone could eventually become a member of the elite.
Elite theory does not provide an objective boundary to separate powerful from non-powerful institutions or movements, a limitation that reinforces criticisms. Nevertheless, Higley et al. (1990) argue that this problem is not limited to elite theory. Key concepts in sociology and political science have imprecise empirical definitions, such as 'state', 'market', 'social class' and so on. According to these authors, the concept of elites is actually more specific and detailed than the ones just mentioned.
Elite sectors Elites dispute power, meaning that they may find more antagonism among themselves than among the lower classes. For instance, a workers' strike, from elite theory's perspective, would imply a conflict between union leaders (labor elites) and corporate elites. Even a socialist revolution would be seeing as a case of elite circulation, where elites are overturned by other elites.
The example of a workers' strike brings up the subject of elite sectors. Elites in complex societies are not a homogeneous group, nor share the same amount of power. According to the definition of elites, power comes from different sources, meaning that there are different types of elites emerging from different organizations and movements.
Political elites probably constitute the most researched elite sector, and often in the literature we find the term 'elite' being used as a synonym for 'political elite'. Nonetheless, there is abundant literature on business elites (e.g. Boschi and Diniz, 2004; Cardoso, 1964; Carrol and Sapinski, 2010; Heemskerk and Fennema, 2009), military elites (e.g.
Janowitz, 1964; Mills, 1956), media elites (e.g.
Davis, 2003; López, 2012), state administrative elites (e.g. de Swaan, 1988; Putnam, 1977), religious elites (e.g. Olson and Carrol, 1992; Wald, 1992), among other elite sectors.
Each sector has a specific dynamic of elite recruitment.
Elite size and sub-divisions may vary in different social contexts, from tribal organizations to the multi-state organizations such as the European Union. The mechanisms of elite recruitment within a single elite sector may change over time as well.
The dynamics of elite recruitment and of elite-toelite relationships constitute a key aspect of elite research.
Elite research methods Methods in elite research often rely on an operationalization of the concept of elites based on institutions of different types, such as political parties, governmental agencies, private enterprises, unions, social movements and so on. The position method (Hoffmann-Lange, 2007) is the main sampling strategy associated with an institutional frame of elites. It aims first at powerful institutions and then to the main positions within those institutions. Several elite surveys follow this sampling strategy in order to achieve a significant sample size.
Other operationalizations of the concept of elites rely on reputation, thus targeting those perceived as leaders by others. Another method of elite sampling targets decision makers within a circumscribed political event, for instance the passing of a bill in congress or the making of a given policy. Both methods are very effective in micro-level decision making studies.
Overall, elite research methods are broad. The 4 López Elite theory most popular among elite researchers are elite surveys (often using the position method), network analysis, comparative historical analysis, political ethnography and in-depth interviews. Throughout this review I will provide several examples and results concerning those methods.
Empirical evidence and assessments of empirical research As noted, American scholars had previously rejected Pareto's and Mosca's work because of their supposed anti-democratic quality. Those who embraced elite theory in the US often referred to it as realism (Grynszpan, 1999). Nevertheless, in the post-World War II period, elite theory migrated to the United States and authors such as Mills (1956), Domhoff (2009 [1967]), Putnam (1976) and Dahl (1958, 1961, 1971) delved into the elite question in democratic societies, bringing democracy to the fore as the main research interest in elite theory.
Mills emphasized the existence of a power elite within the upper class, arguing that democratic institutions in the US played a smaller role in decision making when compared to interest groups deep rooted in politics, the corporate world and the military.
Domhoff (2009 [1967]), instead, presented the American political agenda as dominated by corporations, arguing that decision makers in the US are few and strongly interconnected.
On the other hand, Dahl (1971) pictured elites as pluralistic and argued that democracy is the result of a complex set of elite bargaining, dispositions toward conflict or agreement, and political culture. Dahl's democratic theory and typological work relied heavily on elite theory's assumptions, turning him into a key author in current elite research. His work was also more ambitious than Mill's and Domhoff 's, who are more restricted to American politics. Putnam (1976) also achieved a wider scope, presenting diffuse elite power in contemporary democracies. The author highlighted how decision making tended to spread into complex sets of bureaucratic organization within the state, fractionalizing power through a wider and more technical set of elites.
During the Cold War the division between a liberal democratic bloc in the West and a communist totalitarian one in the East stimulated some elite research (e.g. Hoffmann-Lange, 1971; Hoffmann- Lange et al., 1980; Matthews, 2011 [1978]). Yet dozens of non-democratic regimes (some of them Western allies) fell outside the scope of the West-East divide, including military dictatorships in South America, traditional monarchical regimes in Asia and sultanistic regimes in the Middle East. In these contexts, elite typologies regarding different political regimes also gained importance. In that regard, although they are not strictly elite studies, the works of Linz (2000), Stepan (1973, 1978) and Linz and Stepan (2011) constitute good examples of theorization on totalitarian elites and authoritarian elites.
Unequivocal elite theory gained new strength during the processes of democratization in Southern and Eastern countries in the late 20th century. In the context of the third wave of democratization, to use Huntington's term (1991), sociologists and political scientists throughout the world were eager to understand what made democracy possible in former authoritarian countries. The collapse of real socialism in Eastern Europe and the almost sudden democratization of traditionally authoritarian countries in Southern Europe and Latin America challenged former interpretations based on rigid notions of structural constraints and cultural determinism.
As facts developed, the role of elites in constructing and negotiating political regimes in these recent and often unstable democracies became a meaningful research interest.
The new elite paradigm The theoretical effort to model regime change and democratization led to a new formal understanding of elite theory, often referred to as the new elite paradigm or simply as new elitism. The main references in this regard are the works of Higley and several coauthors (Best and Higley, 2010; Burton and Higley, 1987; Dogan and Higley, 1998; Field and Higley, 2012 [1981]; Higley and Burton, 1989, 2006; Higley et al., 1990). The American Sociological Review articles 'Elite settlements' (Burton and Higley, 1987) and 'The elite variable in democratic transitions and breakdowns' (Higley and Burton, 1989) marked this new theoretical venture and a number of critical and enthusiastic publications followed in their wake.
Encouraged by the context of democratization, the new elite paradigm relates to Huntington's theory on democratic waves. New elitists use Huntington's 'waves' to build typologies of elite interaction at different points in history. Alongside Huntington's influence, the new elite paradigm is much more Weberian in scope, incorporating a longterm comparative perspective and highlighting elite behavior according to broader historical scenarios of authority transformation.
After several publications, the Higley and Burton (2006) model provided a well-established typology of elite dynamics based on the capacity of elites to share interests and act cohesively, despite inter-elite disputes. Higley and Burton (2006) open their book 5 López Elite theory Elite Foundations of Liberal Democracy referring to recent failed attempts to democratize countries using force, pointing to the example of Iraq. They claim that such efforts fail because the bases of previous non-democratic rule persist as elites remain disunited.
Among the examples of democratization based on elite settlements provided by the authors are the end of apartheid in South Africa, some Latin American cases and others in Europe. The new elitist literature is usually based on comparative historical analysis, although there was some use of network analysis among the first empirical studies (Higley et al., 1991).
New elitists propose a typology of elite configuration: elites could be (a) disunited, (b) consensually united or (c) ideologically united. Consensual unity is associated with stable democracies, while elite disunity is associated with unstable democracies and authoritarian regimes. Ideological unity relates to totalitarian regimes, adopting Linz's (2000) typology, where elites legitimize and submit to a highly centralized ideological command.
The new elitist model points to several cases in history where elites migrated from one type to the other (mainly from elite disunity to consensual unity), and how elite configuration is an important cause of political regimes. Along with elite types, two important concepts are presented by the model: elite settlements and elite convergence. New elitists argue that most democratization processes in recent history, and several more in the past, are a result of elite settlements. An elite settlement is a rare event in the history of national communities, in which previously confronting elites choose to negotiate a new political order, thus recognizing each other as legitimate political actors (Higley and Burdon, 2006; Higley and Gunther, 1992). The result of an elite settlement is some sort of democratic rule, be it a full democracy or an electoral or pseudo-democracy.
Elite convergence is defined as a phenomenon that often follows elite settlements. It denotes a process in which, in an unstable democracy, politically organized anti-system elites abandon radical opposition and adopt a coalition strategy in order to amplify their chances of electoral gain. Thus, previously radical elites accept the rules of the game, converging with governing elites in the legitimation of democracy and the rule of law. This process of elite convergence is often associated with the transition from an unstable democracy to a consolidated democracy where no elite group significantly challenges the regime.
In sum, the new elitist model argues that sustainable democracies are the result of elite consensus. In order to demonstrate this thesis, Higley and Burton (2006: 45) point to the early settlements and elite convergences in the West, especially in England (elite union since 1689), the Netherlands (elite union since 1813) and the United States (elite union since 1789), reinforcing their causal argument in light of the classical debate on Western exceptionality.
Simultaneously, the authors argue that most European countries presented sets of disunited elites, thus politically unstable, up until the 20th century.
New elitists also call attention to many aspects of political development that influence elite unity and disunity, such as former colonial rule, economic opportunities, war, political violence, ideological movements, the masses' animosity and so on. Higley and Gunther (1992) argued that elite settlements are often a reaction to social or political crisis, where non-elites (or 'the masses') threaten elite positions or elite interests. Higley and Burton (2006) later on strengthened the importance of non-elites in elite theory with the notion of interdependence. Yet the notion of elites/non-elites interdependence was also developed by de Swaan (1988) and empirically addressed by several other authors (e.g. Clarke and Sison, 2003; Hossein, 2005; Hossein and Moore, 2005; López, 2013; Reis, 2011; Reis and Moore, 2005) who do not always refer to the theoretical model of new elitism.
Since there are now many democracies in the world, the new elite paradigm ends up proposing an interesting research challenge: to distinguish the democracies with consensually united elites (i.e.
those that will last) from those democracies where elites remain in disunion (i.e. future non-democracies or unstable democracies).
Nevertheless, the new elite paradigm has not (so far) provided an efficient universal measurement of elite unity/disunity. Although it provides a plausible understanding of democratization and democratic erosion, it does not provide a potent tool to predict them. Higley and Burton realize that contingency plays a major role in their model, and elite configurations are analyzed on a case-by-case basis.
Despite criticisms, the new elite paradigm is highly influential in elite theory and enthusiasm for it sometimes gives the impression that new elitism implies current elite theory in its totality. In that regard, contributions to elite theory based on inputs from political culture theory (e.g. Hoffmann-Lange, 2010; Lijphart, 1969; Verba and Orren, 1985; Verba et al., 1987) or rational choice modeling (e.g. de Swaan, 1988, 2005) are often understated.
Further assessments on democracy and welfare As previously acknowledged, elite research is primarily dedicated to the elucidation of state outcomes 6 López Elite theory through elite behavior. In this section I will explore two topics in elite research: elite origins of democracy (beyond new elitism) and elite origins of the welfare state. Of the two, the topic of democracy is definitely predominant in elite literature. Yet the topic of welfare is a perfect example of how elite theory can explain state transformations, beyond regime change. At the end of this section I will also briefly make note of other perspectives and topics on elite research, namely gender, ethnic relations and microsociological perspectives.
Elite origins of democracy As seen, one of the main research interests in elite literature concerns the relationship between elites and democratic/non-democratic outcomes. Regime theorists usually include the elite variable in their models of democratization (e.g. Dahl, 1971; Diamond, 1999; Lijphart, 1969; Linz and Stepan, 2011; Sartori, 1987) and their work is often incorporated by elite theorists (e.g. Higley and Burton, 2006; Hoffmann-Lange, 2010).
Although elite researchers agree that elite behavior causes regimes, there is less agreement on what motivates elites. For instance, political culture theory is definitely not at the heart of classical elite theory, and yet several elite researchers explain elite behavior through political culture (e.g. Altmeyer, 1996; Diamond, 1994; Feldman, 2003; Hoffmann- Lange, 2011; Stevens et al., 2006; Verba and Orren, 1985; Verba et al., 1987). Researchers that merge elite theory with political culture theory often make use of considerably large n elite surveys to sustain their arguments (e.g. Hoffmann-Lange, 2010; Stevens et al., 2006; Verba et al., 1987). On the other hand, important authors in political culture literature such as Inglehart and Welzel (2005: 2) implicitly criticize elite theory by arguing that 'democracy is not simply the result of clever elite bargaining'.
Within a less cultural vein, Dahl (1971) argues that elite beliefs and values are key elements to sustain democracy, yet the author highlights that values should not be taken as explicatory variables, but as intervenient variables. The idea that values are important, yet not causal, is significant even among some celebrated authors within the political culture literature, such as Lipset (1981) and Huntington (1996), who claim that democracy is a result of both socioeconomic modernization and cultural settings.
Then again, the new elite paradigm (or new elitism) reframed circulation theory to introduce the notions of elite settlements, elite convergences, elite unity and elite disunity as the determinants of democratization processes worldwide (Field and Higley, 2012 [1981]; Higley and Burton, 2006). In the new elitist perspective, democracy can be the result of a pact between elites, who sometimes perceive more advantages in accepting elite circulation than in promoting a 'zero sum' game. Yet new elitists also acknowledge that democracy may come without elite consensus. In their view, democracies built by disunited elites are unstable democracies, because the source of stability (elite union) is absent.
Overall, the relationship between elites and democracy is heavily researched, using different methods, but also different assumptions about elite behavior. Different research questions are responsible for a great number of differences in the literature, yet it is also true that different conceptions about elite behavior may lead to opposite causal claims within similar research concerns. One example is the expected effect of elite perceptions of threats. Using rational choice modeling, Acemoglu and Robinson (2000) claim that social democratic regimes were the result of elites' threat perception. They argue that elites preferred to include the masses in the political game in order to prevent political and economic threats posed by non-elites. Therefore, threat perception would tend to lead to stronger democracies.
Meanwhile, Stevens et al. (2006) used elite surveys and came to the conclusion that, when threatened, elites tend to adopt authoritarian values. As they assume authoritarian regimes to be based on authoritarian values, threat perception would tend to lead to less democracy.
This type of contrast is revealing of how elite theory absorbs other theories, such as rational choice theory and political culture theory. Yet it is also an important call for dialogue within elite research, especially concerning measurement and competing causes. In this case, if one joins both causal claims, threat perception appears to be a necessary yet insufficient condition for both democratization and authoritarian rule, thus not a good predictor of either. Both arguments neglect the role of inter-elite dynamics, which may elucidate the mechanisms that explain democratic/authoritarian outcomes.
Elite origins of the welfare state Another important concern among elite researchers is the formation of the welfare state in both Western and non-Western countries. Once again, an important part of the elite literature on the topic relies on the notion of political culture. Meanwhile, another set of literature relies on a combination of rational choice modeling and comparative historical analysis.
Verba and Orren (1985) and Verba et al. (1987) are the main authors to defend the importance of elites' political values in the shaping of welfare states.
They argue that the welfare structure found in developed countries is a result of political struggles over 7 López Elite theory the idea of 'equality' (Verba et al., 1987). Based on elite surveys in the US, Sweden and Japan, the authors argue that such values are not homogeneous across countries or even within countries. They also argue that different elite conceptions of equality result in a plurality of welfare state models throughout the world.
Taking a different approach, de Swaan (1988, 2005) argues that the welfare state was a result of elite reactions to negative consequences of poverty.
By combining elements of rational choice theory (see Sociopedia.isa article on rational choice theory by Sato, 2013) and comparative historical analysis, de Swaan (1988) argues that elites have historically benefited from poverty, as it provided workers, soldiers, consumers and voters. Yet, throughout the 19th century, with great industrialization and urbanization, West European and North American elites started to experience major negative externalities of poverty, such as epidemics, political threats and violence. In that regard, de Swaan argues that elites could actually prevent those negative outcomes by improving the life conditions of the poor through private assistance and charity. The problem was that some elites (or a great deal of them) tended to act as free riders as they realized that they could benefit from the results of poverty mitigation without sharing its costs. Free riding made mitigation too expensive and risky for those willing to act. That frustration led elites in toward the state, which was the only actor capable of imposing costs and guaranteeing results. In the long run the welfare state was born.
Nonetheless, the development of an empirical research agenda based on that approach led de Swaan himself among other authors to question the notion of elites as moved solely by rational interest (de Swaan et al., 2000; Reis and Moore, 2005). Quite often elites may have found themselves pressured by negative externalities of poverty, such as epidemics or violence, without reacting to them according to the European and American experience. That was the case in South America (López, 2013; Reis, 2000, 2005, 2011), South Africa (Kalanti and Manor, 2005), Malawi (Kalebe-Nyamongo, 2010), Bangladesh (Hossein, 2005; Hossein and Moore, 2005), the Philippines (Clarke and Sison, 2003), among others.
Elite research outside of the developed world led to a critical adaptation of de Swaan's theoretical model, and political values were often reintroduced in order to explain outcome variance between Western and non-Western countries (Clarke and Sison, 2003; Hossein and Moore, 2005; López, 2013; Reis, 2005, 2011). Nonetheless, those authors do not tend to approach values as explicatory variables, as did Verba and Orren (1985) and Verba et al.
(1987). Overall, the elite origin of the welfare state in current egalitarian societies has contributed to the understanding of elite dynamics in highly unequal societies.
A note on gender, ethnic relations and micro-sociological perspectives Although elite positions are predominantly occupied by men, few studies focus on gender relations regarding the elite world (an exception is Vianello and Moore, 2004). The same can be said about elite's racial or ethnic backgrounds, at least in Western and Latin American contexts, with the important exception of elite studies about the South African case (e.g.
Higley and Burton, 2006; Kalanti and Manor, 2005; Kotzé and Toit, 1995; Lieberman, 2001, 2003).
Some authors have committed to elite studies within a micro-sociological perspective. For instance, Yamokoski and Dubrow (2008) use in-depth interviews to approach the mechanisms through which elites build notions of power and influence based on their individual strategies of prestige building. Also using in-depth interviews, Silber (2012) demonstrates how elites can feel frustrated about state action, arguing that feelings of anger help to shape political attitudes.
In a different research tradition, tribal elites in African contexts have continued to be studied by political anthropologists. In a way, those studies strengthen elite theory's ambition of universality and could be further considered by elite researchers in sociology and political science.
Discussion Elite theory's roots in classical sociology still account for many of its core concepts, such as elite circulation and elite inevitability. As I have shown, those concepts are summarized by the law of the small number, attributed to the combined works of Pareto, Mosca and Michels. Max Weber is also highly referenced in elite theory reviews, mainly because of domination theory, but also due to his theoretical work on political parties and his extensive work on political development. It is my conception that current elite theory often benefits from a more Weberian turn.
I have shown that there are significant efforts in rebuilding/unifying elite theory, as well as important contributions from researchers that do not claim to be elitists. The new elite paradigm constitutes the main theoretical venture in current elite literature, presenting a model of regime change based on elite unity/disunity.
Throughout this article I have discussed how 8 López Elite theory contemporary elite theory developed into different research concerns, absorbing contributions from other theories such as political culture theory and rational choice theory. Nonetheless, I also argued that different approaches within similar research concerns may lead to contradictory causal claims.
Contradictory arguments in the elite literature call for dialogue. In this regard, the International Political Science Association's Research Committee on Political Elites must be saluted, and elite research activities within the Research Committee on Political Sociology, shared by the International Sociological Association and the International Political Science Association, should be encouraged.
Despite theoretical diversity, elite researchers share the assumption that elites are key actors in shaping political and social outcomes. Many criticize elite theory for supposedly overlooking structural constraints and competing causes of political phenomena (Cammack, 1990; Collier, 1999, Inglehart and Welzel, 2005). Furthermore, elite theory has been criticized for working with a vague definition of the term 'elite', consequently with vague distinctions between elites and non-elites (Cammack, 1990).
It is my belief that Higley et al. (1990) are correct in affirming that elite theory does not deny structural constraints of elite action. Furthermore, structural constraints of elite behavior are currently much discussed and measured in elite research. Authors such as Dogan and Higley (1998), de Swaan (1988, 2005), Reis (2011) and Reis and Moore (2005) have pictured elites as reactive actors, often exposed to social threats. They argue that elite behavior can be shaped by external threats, such as epidemics, competing ideologies, social violence and political violence.
In that regard, the interdependence between elites and non-elites (or simply the 'masses') has become a key dimension of elite theory.
Finally, what should we consider as the 'real' advantages of elite theory in contrast to competing theories within political sociology and political science? Elite theory maintains its great advantage of doing macro-theory while operating with a smallgroup/ personal level of conceptualization (Marcus, 1983, apud Higley et al., 1990). This advantage results in a very close relationship between theory and empirical research. In sum, elite theory worries about the state, without reifying the state.
Future directions in elite research As said, the third wave of democratization encouraged a revival of elite research due to elite theory's suitability for the study of rapid political transformations.
Along with this revival, came new theoretical understandings of elite behavior as well as empirical research on the follow-up of regime changes (e.g.
Dogan and Higley, 1998; Higley and Gunther, 1992; López, 2013; Reis, 2011).
Decades after the third wave, we have a significant accumulation of research on the European and Latin American transitions. Nevertheless, democratic stability also represents a great challenge for elite researchers today. In some cases, such as Brazil, and South Africa, elites sustain fairly stable democratic institutions and yet tolerate settings of extreme social inequality, violence and state inefficiency. With no consensus about the mechanisms that would conduct elites toward distributive and inclusive measures, elite research is still far from predicting which paths new democracies are likely to adopt. Will they resemble Western egalitarian democracies? Or is this as much change as we should expect? Western societies have also changed a great deal and we do not know yet how elites will follow along.
Inequality has increased in North America and Europe and immigration is rapidly transforming the social landscape of the most egalitarian Western democracies. Cultural cleavages incentivize radical political elites to enter the democratic game and extreme right-wing leaders have broadened their influence in countries like France, the Netherlands and Switzerland. This phenomenon is being addressed by elite researchers in panels and research committees in the International Sociological Association (ISA), the International Political Science Association (IPSA) and the European Consortium for Political Research (ECPR), so we should expect relevant publications in the near future.
Furthermore, we are now facing rapid regime changes in Islamic societies, whose patterns hardly resemble those of Eastern and Southern transitions.
As former autocratic regimes fall, it is not clear yet what types of elites are likely to prevail in the aftermath of great social upheavals. Egypt, for instance, seems to experience violent elite disunity, with massive conflicts between military elites and political/religious elites. The relationship between elites and non-elites seems to be critical in such junctures.
Last but not least, the world economy is changing and many key global players are either former authoritarian countries, such as Brazil, Russia and South Africa, or current authoritarian countries, essentially China but increasingly Russia once again.
This means that a growing amount of the world's power resources depends on elites that come from different attitudinal and ideological backgrounds in comparison with Western elites. It is noteworthy that the democratic status of successful new democracies is often lower than those of old Western democracies, frequently due to limitations in individual freedoms and the rule of law (at least 9 López Elite theory according to the most renowned indexes, such as Freedom House and the Transformation Index BTI).
Should we expect some of these new democracies to erode? What attitudes should we expect from increasingly powerful non-Western elites? These are important questions ahead for elite research.
Acknowledgments I am grateful to Prof. Graziella Moraes Silva who was a reader of earlier versions of this article and to the entire team of the Interdisciplinary Network for the Study of Social Inequality (NIED). I also thank the editor and two anonymous reviewers for their suggestions, especially 'Reviewer 2', for his/her sharp observations.