Abstract
The rapid expansion of the social sciences in post-war America produced a new approach to research on the theory and practice of democracy. Some of the main themes of this approach were borrowed from early sociological critiques of democracy developed by a group of European social scientists who were later called ‘elite theorists’ or ‘Machiavellians’. This article outlines the set of theoretical motifs found in the works of Gaetano Mosca, Vilfredo Pareto, and Robert Michels that became a foundation for the study of democracy in American post-war social science. Writing in response to the perceived problems of social democracy at the time, Mosca, Pareto, and Michels each identified the goals and ideals of mass popular sovereignty as ill conceived and dangerous based on ‘social facts’ derived from empirical observation. These ‘facts’ appeared in later studies of democracy as naturalized or self-evident foundational propositions. Joseph A. Schumpeter’s famous critique of democracy’s classical ideals is one of the most important examples of a theory built on the ‘facts’ produced by the early critics. This article therefore presents an analysis of the role these conclusions played in Schumpeter’s theory, which characterized democracy as a series of mechanisms designed to mediate and control, rather than give full expression to, popular sovereignty.
We deem it sufficient for our purposes here to demonstrate that the assumption that the elected official is the mouthpiece of the majority of his electors is as a rule not consistent with the facts; and we believe that this can be proved by facts of ordinary experience and by certain practical observations that anyone can make on the manner in which elections are conducted.
At the close of nineteenth-century Europe, the rise of democracy and the push for universal suffrage appeared both as the natural products of the ‘age of reason’ and as a looming threat to the established social order. Inspired by the ‘ideals of 1789’ the working classes started to organize political parties to challenge traditional arrangements of authority and bring forward serious questions about the existing foundations of social and economic life. This ‘era of democratization’ coincided with what historian Eric Hobsbawm called ‘the golden age of a new political sociology’ (2002 [1987]: 88), and in most cases the leading figures of this new discipline were either ambivalent towards or outright critical of democratic movements. This article returns to some of the early critiques that explicitly attempted to counter the popular ideals of democracy with ‘facts’ about the reality of society and politics. Three key figures stand out as the most systematic critics of democracy in this period: Robert Michels, Gaetano Mosca, and Vilfredo Pareto. These critics, often referred to as either ‘elite theorists’ (Bachrach, 1967; Bottomore, 1964; Nye, 1977) or ‘Machiavellians’ (Burnham, 1943), each based their assessments of democracy on the study of fundamental empirical ‘facts’ that could be generalized to any political society. While their pessimistic conclusions about democracy may not resonate today, the fundamental ‘problem’ that shaped their critiques still continues to dominate the contemporary study of democracy. This basic ‘problem’ is the tension between democratic ideals and the practical realities of politics.
In the following, I argue that an explicit appeal to science allowed the early critics of democracy to produce theoretical systems that effectively separated ‘social facts’ from what was then seen as the unrealistic or ideological foundations of social democracy. Unlike its more specific usage in the Durkheimian sense, here the term ‘social fact’ refers to a more general category of supposedly universal laws that are, on one hand, produced by the empirical study of historical cases and, on the other hand, proved by the ‘ordinary experiences’ and ‘practical observations’ mentioned by Gaetano Mosca in the introductory quotation. While none of these critics explicitly used the term ‘social facts’, the evidence that they produced against democracy allowed them to argue that democratic ideals misunderstood universal laws – or ‘the facts’ – of social organization and leadership.
For many post-war social scientists as well, similar empirical observations continued to represent a large gap between the classical doctrine of democracy and the ‘reality’ of democratic political systems. The economist Joseph A. Schumpeter famously outlined this incongruity in his explicit rejection of the classical doctrine in favour of an alternative doctrine based on the formal competition for leadership (Schumpeter, 2008 [1942]). While this apparent focus on leadership has led some to categorize Schumpeter’s work on democracy as an extension of ‘elite theory’ (Bottomore, 1964; Held, 1987; Medearis, 2001), this article makes clear that both Schumpeter and the elite theorists were less concerned with leadership as such, and more concerned with social democracy and the rise of workers’ parties as the central political problems of the day. More than simply identifying the connections between early critiques of democracy and post-war social science, I argue that this representation of democracy as a ‘problem’ was an important innovation in the marriage of modern science and politics. After a brief introduction of the intellectual and social context in which Mosca, Pareto, and Michels developed their theoretical systems, the discussion turns to an analysis of each theorist’s application of scientific methodology to the ‘problem’ of democracy. Following this analysis, I present an account of how Schumpeter incorporated the central themes and conclusions posed by these critics into his own prominent and influential reappraisal of the theory and practice of democracy.
Whereas democracy in the first decades of twentieth-century social science was identified more narrowly with the rise of the masses through the expansion of workers’ parties, democracy now refers to a proliferation of normative models (Held, 1987) that support a wide variety of sometimes conflicting values, ideals, and practices. As such, the proposition that democracy is the best possible form of government has come to be accepted as practically irrefutable, even as the concept has been emptied of any universally accepted normative content. The purpose of this article, however, is not to propose normative content to fill or unify the concept of democracy, but to reconstruct and examine a particular tradition in the social sciences that aimed to thin out or empty the concept of democracy of ‘undesirable’ or ‘impractical’ normative elements. This tradition helped to reinvent the concept of democracy as a series of institutional mechanisms narrowly designed to produce a stable government through formal representation.
Fusing science and politics
In the second half of the nineteenth century, the rise of workers’ parties and the demand for an expanded political franchise threatened to shake up established European political orders. Where the scale of industrial production was expanding most rapidly, the dynamics of social and political relations were also quickly shifting, and it is in this context that mass-based parties and associations built on workers’ interests became important political actors (Hobsbawm, 2002 [1987]). To this degree, democratic politics were largely coterminous with socialist politics. It is not surprising, then, that much of the intellectual debate about democracy focused on an unambiguously socialist version of democratic politics. Mosca, Pareto, and Michels, for their part, each had some first-hand experience with social democratic politics and their critiques were important contributions to the anti-socialist side of this debate.
While the political positions of the ‘Machiavellians’ come through clearly in their major works, the reason that these works merit attention today is because they represent some of the first attempts to apply newly developing methods of social science to debates that had previously been characterized by arguments for or against political ideals. Of course, a number of socialist intellectuals during this period worked to demonstrate the proper role of democracy in socialist politics scientifically, 1 but the aforementioned critics are unique in that they attempted to use scientific methodologies to develop explicitly apolitical theories of politics. This approach was taken up by later defenders of ‘elite theory’ who distinguished between analytic and normative ‘elitism’ (see, for example, Bennett, 1978; Femia, 2001; Lenski, 1980) in order to retain the ‘scientific’ insights of the former without supporting the anti-democratic sentiment of the latter. The following discussion, however, differs from the standard critique of elite theory that concludes that the anti-democratic normative implications are inescapable (Bachrach, 1967; Pateman, 1970). The present argument instead critically analyses the ‘social facts’ produced by elite theorists as assemblages of analytical and normative claims. To unpack the methodological and theoretical content of Mosca’s, Pareto’s, and Michels’ respective critiques of democracy, it is therefore important to begin by outlining the intellectual context in which they worked. The social psychological theory of Gustave Le Bon best illustrates this context.
Especially popular among intellectuals who saw the rise of socialist parties as a potential threat, Le Bon’s argument in his famous Psychologies des foules (1895) is emblematic of the role of mass psychology in pre-war critiques of democracy. In the introduction to the 1960 English translation (published as The Crowd), Robert K. Merton suggests that Le Bon was as profoundly influenced by the upheavals of French plebiscites in the latter half of the nineteenth century as he was by the ‘Great Revolution’ (1960: xxvii). What makes The Crowd so interesting for the study of democracy is that Le Bon was centrally concerned with (1) groups of people – or crowds – as a unit of analysis and (2) the relationship that crowds have with their leaders.
Le Bon’s primary and foundational assumption, found in the book’s opening comments, regards the ‘mental unity’ of crowds. Here the author establishes that there is an identifiable mental state that can be attributed to ‘crowds’ and that this mental state is different from and often overtakes the mental state of the individual. Moreover, he suggests that ‘the crowd is always intellectually inferior to the isolated individual’ (1960 [1895]: 33). For Le Bon, crowds are incapable of reasoning and they come to hold opinions not through a logical chain of ideas, but through a culmination of simplified ideas that act on the imagination of the crowd. Therefore, ‘to know the art of impressing the imagination of crowds is to know at the same time the art of governing them’ (1960 [1895]: 71).
This statement reflected the more broadly held political concern that rhetoric, as opposed to reason, was becoming an indispensable tool in the world of mass parties and popular elections. In fact, even though Le Bon’s explicit project was to create a typology of ‘crowds’, throughout the work it is clear that an implicit fear of the newly powerful mass parties shaped many of his conclusions. Ernesto Laclau (2005), who uses The Crowd as a starting point in his study of populism, 2 places this concern in the context of a general approach to mass psychology at the time by suggesting that the irrationality associated with crowds embeds a more specific association between mass phenomena and pathology (Laclau, 2005: 29). This, in part, explains why Le Bon proceeds from a set of general propositions on the nature of crowds to a more specific investigation of how crowds operate on a basis of pathological irrationality.
Where Le Bon discusses the specific problem of democracy, he presents it as an ideal or concept that resonates with crowds precisely because it resists rational or scientific thought. As he suggests:
Words whose sense is the most ill-defined are sometimes those that possess the most influence. Such, for example, are the terms democracy, socialism, equality, liberty, etc., whose meaning is so vague that bulky volumes do not suffice to fix it precisely.
While this general statement still holds a certain amount of truth today, Le Bon’s point was that the manipulation of ‘ill-defined’ ideals such as democracy allowed leaders to exploit the irrationality of crowds. As with the invocation of ‘democracy’ in the plebiscite of Louis Bonaparte, political ideals are characterized by Le Bon as mere tools that vary only according to how effectively leaders can employ them in the manipulation of different ‘types’ of crowds. 3
The distilled proposition implied in Le Bon’s study, that crowds are irrational and always subject to the manipulation of leaders, became a prototype for later scientific ‘facts’ that were generated to demonstrate the impossibility and undesirability of social democracy. The study’s assertion of universal principles about crowd psychology, and its accessible empirical foundations – any rational person could go out and witness the irrationality of crowd behaviour – became accepted implicitly in later attempts to confront social democratic ideals with scientific analysis. In this way, the irrationality of crowds took on the properties of what Bruno Latour has called a scientific ‘black box’ (1987: 2, 131). The metaphor of the black box, defined as a well-established fact that resists ‘opening’ because of ongoing elements that maintain its acceptance in a scientific community, is particularly useful in describing Le Bon’s conclusions because they gave scientific weight to a concern that already existed among some political commentators and scholars of the period. The irrationality of crowds, once it became an established fact, resisted scientific challenge because a broader interest in refuting the ideals of workers’ parties encouraged further work that characterized democracy as a problem. The three critics to whom I now turn are the most prominent examples of scholars who attempted to work on this ‘problem’ scientifically.
Three critics of democracy
The following discussion turns to the key works of Mosca, Pareto, and Michels and focuses on the political conclusions they reached based on their conceptual and methodological approaches. Although each of these critics wrote on a wide variety of topics, they all used empirical evidence to generate social scientific theories that had the effect of refuting particular elements of social democracy. These theories produced facts that eventually became naturalized (black boxes) in post-war studies of democracy. Three theoretical motifs played a prominent role in this fusion of political or normative sentiment with scientific fact and these motifs organize the following discussion. Much like a narrative motif, the theoretical motifs attributed here to Mosca, Pareto, and Michels represent recurring themes that organize or structure each theoretical system. 4 Furthermore, the motifs invoke some of the normative elements of intellectual currents that were particularly influential at the time and therefore contributed to the self-evident nature of each critic’s conclusions.
The three theoretical motifs, which appear in the work of each but are emphasized differently among the three, are: (1) the theory of elites, which is outlined in Mosca’s The Ruling Class (1939 [1896]); (2) anti-positivism, which grounds Pareto’s sociological treatise; and (3) anti-idealism, which characterizes Michels’ famous work on the German Social Democratic Party. The first motif reflects some of the concerns that inspired Le Bon’s study, especially the threat posed by the rapid expansion of workers’ parties in industrial Europe. The second and third motifs reflect a more specific dissatisfaction with positivist and idealist theories of social action at the end of the nineteenth century. These two intellectual (counter-)currents, which Parsons (1968 [1937]) took to indicate a revolution in the theory of social action, were particularly useful to both Pareto and Michels because they broadly aligned their ‘scientific’ arguments against calls for democracy built on ‘outdated’ theories of social action.
These three common motifs of theory and fact production exemplify the core of a tradition built on ‘social facts of democracy’. However, in addition to tracing the generation of facts and theories in the texts themselves, the following will also include at least some intellectual history in order to provide a critical hermeneutics for the texts that present these critiques of democracy. What follows is therefore part intellectual history and part theoretical analysis, presented with the recognition that author, text, and intellectual tradition are all, to some degree, co-constituted.
The theory of elites (Mosca)
Where Le Bon focused primarily on crowds, Mosca was centrally concerned with developing a political science to explain the relationship between rulers and ruled. In basic terms, Mosca’s famous ‘theory of the ruling class’ suggests that there has always, in every society, been a class of people who rule and a class of people who are ruled. This distinction is a product of some ‘real or apparent’ attribute that makes members of the ruling minority ‘more esteemed and influential’ in their own societies (Mosca, 1939 [1896]: 53). 5 In addition to this fundamental relationship of esteem, he also asserted that organized minorities would always have more power than disorganized masses of individuals. 6 As soon as members of a majority organize themselves around a particular idea or purpose, they cease to be part of the majority and become a minority group qualified to compete for leadership. Moreover, although there are many possible ways for a minority group to garner influence vis-à-vis the majority, the important point for Mosca is that this type of relationship universally defines societies. If social change occurs, it is the product of a change in how a group becomes influential and constitutes itself as a ruling minority (Mosca, 1939 [1896]: 53–54).
In the opening chapters of the Elementi di Scienza Politica (1896, translated as The Ruling Class, 1939), Mosca suggests that in order to gain legitimacy, rulers had to apply some sort of political formulation – an ideology, myth, or ‘superstition’ – that justified the ruler–ruled relationship on grounds other than naked material or intellectual force (1939 [1896]: 71). Such a formulation, for him, must necessarily be based on the moral principles that unify a social group, and it is according to the variation of these justifying moral principles in a social system that a typology (or science) of political forms can be constructed (Meisel, 1965: 169).
In political formulations, the moral principles that supported the ruler–ruled relationship were, for Mosca, best understood as social forces that curbed the ‘naturally’ selfish impulses common to all people, including rulers.
7
To him, the democratic call for unmediated popular sovereignty therefore represented a serious threat to the existing social order. If this Rousseauian ideal of popular sovereignty were to express itself in full, the unleashed impulses of the masses and their leaders could undermine the important moderating forces that formalized moral restraint (1939 [1896]: 126).
8
According to Mosca’s theoretical framework, societies differentiated themselves based on the force of their moderating institutions rather than the content of their ideals. The balancing function of moral social forces presented itself either in complex institutions that relied on more ‘civilized’ forms of restraint, or in simple institutions that relied more on repressive forms. ‘Civilized’ institutions such as modern juridical institutions were thus desirable because they were a sophisticated method for controlling the ‘wicked instincts’ of the population and its leaders (Mosca, 1939 [1896]: 127). The distinction between complex and simple institutions of moral restraint further demonstrated the ‘fact’ that
[t]he absolute preponderance of a single political force, the predominance of any over-simplified concept in the organization of the state, the strictly logical application of any single principle in all public law are the essential elements in any type of despotism, whether it be a despotism based upon divine right or a despotism based ostensibly on popular sovereignty.
Mosca’s critique of social democracy or ‘collectivism’ could thus be understood in terms of its tendency toward ‘over-simplification’, and at the root of this tendency was, for him, Rousseau’s argument that the accumulation of wealth was the first and most important inequality (1939 [1896]: 273). In particular, he found that Rousseau’s conclusion – that modern social institutions built on this fundamental inequality corrupted the ‘natural’ morality of humanity – was central to the rise and continued success of social democracy in the nineteenth century. What was unique about social democracy as political formulation during this period was its effective use of the everyday experience of class inequality to transform the ideal of popular sovereignty into a divine or sacred ‘truth’. This truth then classified all other political formulations as systems operating in the interests of the bourgeoisie. For Mosca, social democracy therefore became a doctrine that aimed at a single, simplified principle of ‘absolute justice’ (1939 [1896]: 287).
As a response to the danger posed by this ‘over-simplified’, singular, and metaphysical principle of justice, given expression through the demand for democratic equality, Mosca offered a social science of politics that was built on his recurring foundational assumption of the universality of the ruling class:
Political power never has been, and never will be, founded upon the explicit consent of majorities. It always has been, and it always will be, exercised by organized minorities, which have had, and will have, the means, varying as the times vary, to impose their supremacy on the multitudes.
In many ways, Mosca’s concern with political power places him in the more narrowly political tradition of Machiavelli, but it is his attention to the role of social forces in the development of his theoretical system that makes his work a foundation of the ‘social facts’ tradition in the study of democracy. His ‘ruling class’ thesis enjoyed a long run of popularity, especially in American social science until at least the early 1970s (Nye, 1977). Although the various formulations of ‘elite theory’ often left aside Mosca’s critique of social democracy, it is the combination of this critique, which grew out of a broader distrust of the masses, with the proposal of a ‘scientific’ theory of society that forged his foundational assumption into a ‘fact’ of democracy. Like Le Bon, Mosca saw the unthinking, disorganized masses as a threat to the social order, but unlike Le Bon he turned to historical empiricism to build a typology of political formulations in order to better understand this threat. Consequently the ‘social facts’ associated with Mosca’s contribution to the motif of elite theory challenged the role of popular sovereignty in social democracy as both impossible and undesirable.
The second and third motifs of the ‘social facts’ tradition, anti-positivism and anti-idealism, focus even more specifically on the role of social action in the relationship between rulers and the masses. They also reflect a response to two nineteenth-century intellectual currents that Parsons identified in The Structure of Social Action (1968 [1937]). According to Parsons, the ‘positivistic’ theory of social action 9 and the ‘idealistic’ theory of social action 10 referred to the foundational elements that characterize a social act. 11 Although Parsons saw positivism as a much more dominant current, with utilitarianism being its leading exemplar, the rationalism of positivistic theory was paired with idealism in a way that placed both in contrast to innovations in social theory which emphasized the non-rational and conditional (or structural) factors that make social action meaningful. The study of social action according to the possible configurations of elements that made an abstracted ‘unit act’ meaningful was, for Parsons, key to the contemporary developments in the systematic study of society. Especially in Pareto’s sociological theory, but also in Michels’ work, social action was understood to be dominated by the non-rational, affective, or impulsive elements of that act. Therefore, to these critics any argument for social democracy that relied on reasoned decision making (positivistic theory) or the power of its highest ideals (idealistic theory) relied on an inferior theory of social action. When combined and applied to political problems, the motifs of anti-positivism and anti-idealism constitute a framework that defines political questions in terms of the non-rational and structural elements of social action. Such a framework led both Pareto and Michels to particularly dire conclusions about the prospects for social democracy.
Anti-positivism (Pareto)
Vilfredo Pareto’s critique of democracy incorporates elements of the ‘theory of elites’ motif, but unlike Mosca’s formulation, Pareto’s famous conclusion about the ‘circulation of elites’ is not central to his larger theoretical approach. Moreover, any contemporary discussion of Pareto’s views on democracy must necessarily work through two related and difficult issues: the first is the complexity of his ‘scientific’ project as a whole; and the second is the contemporary filter(s) that colour how his work is read (or, more often, not read). Regarding the latter, three particular characterizations of Pareto stand out: Pareto the economist, Pareto the fascist, and Pareto the mildly interesting relic.
The overwhelming majority of contemporary citations of Pareto’s work focus on his early application of a concept of dynamic equilibrium to the field of economics. Published in 1896, his Cours d’Économie Politique 12 attempted to make sense of historical data on wealth distribution using mathematical equations to demonstrate that an unequal distribution of wealth was a constant feature of economic systems. Today these equations refer to concepts such as ‘Pareto efficiency’ or ‘Pareto optimality’, which, in their highly technical uses, seem to have little to do with democracy except in specific policy application. However, the proposition that wealth inequality was a universal feature of economic systems – intentionally presented as a ‘scientific’ counter to Marx’s critique of political economy – led Pareto to focus on some of the political problems of socialist economics in his later book Les systèmes socialistes (1902). Likewise, the concept of dynamic equilibrium was central to his later sociological work as well as his conclusions about political systems. While the political implications of Pareto’s work serve as a link between his economic and sociological ideas, they have also led to a more damning characterization.
The combination of Pareto’s critique of democracy and his senate appointment by Mussolini less than a year before his death raised enough of a question about the proto-fascist politics inherent in his work that numerous scholars have been careful to debunk his connection to fascism (Femia, 2006; Hughes, 1974; Levy, 1973; Lopreato, 1973; Sica, 1988). 13 While Italian fascists, as well as left political movements such as syndicalism, echoed his pessimism about parliamentary democracy, Pareto’s ‘scientific’ method of historical empiricism explicitly resisted any type of political advocacy. In fact, the emphasis he placed on developing a rigorously ‘scientific’ method for the study of society is probably an even larger barrier to the casual reader. This is especially the case in his massive Trattato di sociologia generale (1916) – published in English as Mind and Society: A Treatise on the General Principles of Sociology (1935) 14 – which proposes terms and concepts that, because of their contemporary obscurity, appear as historical relics rather than useful social scientific tools. Indeed, Pareto does not do his readers any favours by making his arguments using such complex terms, but his near obsession with the application of ‘scientific’ values and techniques was an important example of a broader intellectual atmosphere that questioned the scientific validity of evolutionary theories of human progress through reason, or what Parsons called ‘positivistic theories’. To understand Pareto’s anti-positivist critique of democracy, it is therefore necessary to outline some of the key elements of his theoretical system.
Pareto’s purpose in the Trattato was to understand how non-rational, or what he called the ‘non-logical’ elements of society determine much of what was previously assumed to be the ‘logical’ bases of social action. Pareto used the terms ‘logical behaviour’ and ‘non-logical behaviour’ to describe actions in which the subjective end of an act is the same as the objective end (logical behaviour), or actions in which the subjective end is different from the objective end (non-logical behaviour). In a non-logical behaviour, the reasoning of the actor may not be faulty (illogical) but at least one element of the means-end relationship is not objectively observable (Finer, 1966: 34; Pareto, 1963 [1916]: § 151: 78). Political action based on an ideal such as equality, and directed towards an end such as a socialist utopia, was non-logical because the end was unobservable. For Pareto the influence of such ideals on a society was a product not of the rational content of the ideals themselves but of the foundational impulses, sentiments, and instincts that draw people to such ideals. It is because he considered these foundational impulses and sentiments to be the most important elements of the ‘social equilibrium’ – what Parsons narrowly took to mean the social system (Levy, 1973) – that his work is taken to represent the motif of anti-positivism.
Pareto’s methodological approach started with the observation of social action through historical examples and worked inductively towards the classification of logical and non-logical behaviour. After elaborating the principles of this classification scheme in Volume I of the Trattato, he deductively identified the non-logical elements of the social equilibrium that remain constant over time. The two elements Pareto considered most important were residues and derivations. The system of classification that he developed for these two concepts is highly complex, but for the purposes of the present argument, it is more useful to consider briefly how these two elements contributed to a ‘scientific’ critique of democracy.
To understand the role of non-logical elements in Pareto’s theory of society one must return to the concept of equilibrium that appeared in his earlier work on economic systems. As Femia points out, Pareto’s background as a mechanical engineer ‘encouraged him to see society as a system in equilibrium involving the simultaneous variations of mutually dependent variables’ (2006: 49). Such a definition of equilibrium is dynamic rather than static because the mutually dependent variables are constantly changing over time. In Pareto’s system, non-logical social sentiments are akin to the concept of ‘force’ in mechanics (Lopreato, 1973: 458), and the opposition of these underlying forces act, in society, as the primary dependent variables that continuously push (or pull) toward equilibrium. Although individual non-logical behaviour is subject to a person’s psychic state and is difficult to generalize to society, underlying sentiments can be observed in the forms of collective behaviour that express or manifest these social forces. These observable collective manifestations of sentiments are what Pareto calls residues, and these residues make up the important elements ‘that stand in a relation of reciprocal determination with the social equilibrium’ (Pareto, 1963 [1916]: § 1690: 1126).
Producing an explanation that approximates Pareto’s meaning for ‘residue’ is famously difficult, not least because his use of the term was somewhat inconsistent (Femia, 2006: 58). One might think of residues as ‘regular observable patterns of social behaviour’ that have the abstracted quality of Weber’s ideal types, are generated through a method of everyday observation similar to Simmel’s social types, and reflect a Durkheimian reality sui generis (Parsons, 1965: 83) structured by underlying social forces. In a given society, groups of people may try to make sense of residues, or patterns of behaviour, by constructing theories, traditions, ideals, or justifications that explain their meaning. These constructs, which are also non-logical, are what Pareto calls ‘derivations’. An analogy for the two concepts could go as follows: residues are to derivations as universal taboos around killing are to specific laws against murder.
These two non-logical elements of society are somewhat interdependent – learning a law might inculcate a social taboo – but, historically, residues are always prior to the derivations built on them. As Pareto explains:
Even when there is some rough correspondence between derivation and residue, the derivation usually oversteps the terms of the residue and oversteps reality (§ 1772). It indicates an extreme limit of which the residue falls short, and [very] often contains an imaginary element that states a goal far beyond the goal which would be set if it expressed the residue exactly (§ 1869). If, furthermore, the imaginary element expands and evolves, the results are myths, religions, ethical systems, theologies, systems of metaphysics, ideals. That happens more especially when the sentiments corresponding to derivations are intense, and the more readily, the greater the intensity.
If a given ideology or religious system (a derivation) seems to have an inordinate amount of influence, it is simply because it is expressing a particularly strong underlying sentiment. For example, a generalized sentiment of fear and uncertainty brought on by political instability may lead people to find comfort in traditional forms of authority (a residue), which is then rationalized and codified in a particular ideology or religious system.
With regard to political systems, Pareto concludes that since residues are the bases for derivations, such systems should first be studied in the context of historical fluctuations of structured underlying sentiments (residues). As with his observations about the universality of wealth inequality, Pareto also asserts that elites have always held political power. While this basic proposition is more obviously a product of his economic analysis (Femia, 2006: 69), Pareto suggests that the question of how power is distributed among elites relies on the two residues that best express sentiments related to authority and governance: the ‘residue of combinations’ and the ‘residue of group-persistence’. These two classes of residue are important to Pareto’s political conclusions partly because they represent sentiments associated with reason, science and progress (residue of combinations), and tradition (group-persistence). Democratic ideals might appear to flourish if a society puts its faith in progress based on rationality, but such a system will still likely be dominated by elites willing to take advantage of these sentiments and will more closely resemble a ‘demagogic plutocracy’ (Pareto, 1963 [1916]: § 2268: 1613–1615).
While Pareto describes the different types of elites that tend to consolidate power – Burnham (1943: 156) identifies Pareto’s discussion of these ‘lions’ and ‘foxes’ as a key link to his Machiavellianism – this discussion is based primarily on historical ‘evidence’ of broader societal changes. Based on his analysis of a nearly endless stream of historical examples, he concludes that opposing residues are always in a state of ebb and flow, creating a cyclical model of political change that has been constant throughout history. 15 Thus, Pareto’s famous formula of the ‘circulation of elites’ discussed in Volume IV of the Trattato asserts that the leadership of a governing system can not be sustained if it is dominated by only one of the two residues related to authority. For him, a society in which the authority of science and progress is established will always challenge authoritarianism. Likewise, the persistence of traditional authority structures will always challenge democracy. Political change in this framework has less to do with the content of a ruling group’s ideals and more to do with the ability of elite groups of leaders to react to changes in prevailing residues.
Most importantly for the present argument, Pareto’s critique of democracy is a refutation of any political theory built on the evolutionary assumption that progress in human societies based on reason is inevitable or even possible. The ‘facts’ of the social equilibrium are constant over time and are primary to all theories, myths, ideologies, or belief systems. Democracy, as an ideal that expresses the non-logical sentiments of society, could certainly come to define the politics of a given society, but to assume a causal role for the ideal of democracy based on its rationality would be an error. Pareto rejected the possibility that reason determined the spread of democracy, but this conclusion was only part of his larger anti-positivist project. This motif is also found in the work of Michels, but, as will become clear, his case study of German social democracy was much more concerned with the gap between democracy’s ideals and the ‘facts’ of political parties, rather than a general theory of social and political change.
Anti-idealism (Michels)
In Burnham’s study of the Machiavellians, his chapter on Michels opens with the following statement:
When someone writes a book on democracy, we are accustomed to share with him the assumption, as a rule not even mentioned, that democracy is both desirable and possible. The book will sing the praises of democracy. Its ostensible problem will often be ‘how to make democracy work’ – because even the most ardent democrats, when they get down to the concrete, discover that it has not been and is not working quite as well as democratic theory would lead us to expect. A similar approach is made to such goals as peace, employment, justice, and so on. It is assumed that these are desirable and possible. A writer then devotes his energy to stating his personal scheme for securing them, and thus saving mankind from the ills that somehow in the past have always beset it.
Michels’ book on democracy, Political Parties: A Sociological Study of the Oligarchical Tendencies of Modern Democracy (1962 [1911]), does not fall into such ‘idealism’, says Burnham. Rather, it favours the study of ‘real existing’ political forms over abstract political principles. As Burnham rightly suggests, Michels’ Political Parties aimed to critique idealist theories of democracy by examining the ‘practical realities’ of a case that demonstrated the limitations of both the theory and the practice of democracy: the German Social Democratic Party (SPD). The choice of this case was a direct product of central place that SPD politics played in Michels’ life during the first decade of the twentieth century and it reflected many of the frustrations that he had about socialist politics. At the same time, the book (first published in 1911) marked the end of a transition in his life away from political activism towards more scholarly work. This turn towards ‘scientific’ analysis started with two significant changes in Michels’ academic career. The first was his invitation to contribute to the journal Archiv für Sozialwissenschaft und Sozialpolitik, starting in 1906, which eventually led to a regular correspondence with Max Weber (Scaff, 1981), and the second was his appointment in 1907 to a teaching post in Turin, which brought him in to contact with Mosca. 16
There are a number of consistent themes that bridge Michels’ early radical work with his later conservative or fascist ideas (Beetham, 1977a, 1977b; Michels, 1949), and there are some important works by Michels that are underappreciated (for example, his work on patriotism; Kelly, 2003), but the overwhelming majority of contemporary treatments of Michels focus on Political Parties. This focus is somewhat justified considering that the popularity and influence of the book were almost immediate (Linz, 2006: 26) and Michels’ subsequent work was often influential only to the degree that it built on the core themes presented in this text (Linz, 2006: 63). However, within the text itself, there is an uneasy mixture of themes and, as a consequence, Michels has been read as either a critic of bureaucracy, along with Weber, or an ‘elite theorist’ similar to Le Bon, Mosca, and Pareto (Scaff, 1981: 1280–1281). 17 The following focuses primarily on the latter reading.
In the opening chapters of Political Parties, Michels deals first with the ‘intellectual’ definitions of democracy and aristocracy, followed by a discussion of ‘direct’ democracy, or the rule by the masses. For this discussion Michels invokes the two motifs discussed above. On the one hand, for him, rule by the masses without designated representatives is impossible because the complex tasks of government require delegating responsibility to leaders. On the other hand, such an ideal is impractical because crowds and masses always act irrationally. 18 These ‘facts’ are then codified by Michels in the form of a ‘sociological law’ that is both universal and objective: ‘… the law that is an essential characteristic of all human aggregates to constitute cliques and sub-classes is, like every other sociological law, beyond good and evil’ (1962 [1911]: 6). From this starting proposition, Michels sets out to analyse the role of leaders and masses in social organizations.
In Michels’ view, the relationship between groups or organizations and their leaders is often mutually beneficial. One advantage is that the existence of a leader allows members of the organization or group to be interested in other things, namely their everyday lives. A particular worker or party member can therefore support representative democracy because delegation to an elected official allows for the separation of politics from one’s day-to-day life. For Michels, this delegation of responsibility is not a rational choice, but, following Pareto, is a non-rational impulse – produced by any number of sentiments – which can later take on a rational character in its justification. As Michels suggests: ‘… not only do the masses need leaders but they want and love [them]’ (1962 [1911]: 93). The customary ‘rights’ given to offices held by representatives and the array of ‘skills’ that become synonymous with leadership are thus ancillary to the initial non-rational support for a given leader. For Michels, an important part of this relationship is that leaders themselves tend to have qualities, such as the ‘prestige of celebrity’ or a strong force of will, that help to produce the non-rational veneration and honour awarded by the masses (1962 [1911]: 100). 19
It is partially on this point that Michels generates an explicit critique of democracy as an idealist political system. When leaders acquire specialized knowledge and are subsequently seen to be better able to express the will of the people, then democracy ceases to be a means for the expression of ‘the people’s’ interests. Especially common in socialist formulations of democracy, Michels points out that such an arrangement could more accurately be described as democracy for the people, and not democracy by the people. For him, ‘Social democracy is not democracy, but a party fighting to attain to democracy’ (1962 [1911]: 113). At this stage, the shift from autocratic tendencies to outright oligarchy is not far off. The problem with idealism in this situation is that once a democratic leadership has access to all claims of democratic legitimacy, it can suppress the interests of the people for ‘democratic’ purposes. An important historical analogy for Michels is the ‘democratic’ plebiscite of Napoleon III. ‘Bonapartist ideology’ is, for him, the most dangerous product of democratic idealism because it essentially gives tyranny a democratic cover.
In democratic crowds, Bonapartism finds an eminently favourable soil, for it gives the masses the illusion of being master of their masters; moreover, by introducing the practice of delegation it gives this illusion a legal color which is pleasing to those who are struggling for their ‘rights’.
A critique of ‘illusion’ or idealism is therefore the basis of Michels’ general critique of democracy, as well as his specific pessimistic assessment of the prospects for socialist democracy.
This pessimism – or what Burnham (1943) identified as ‘realism’ – was then applied to a structural analysis of SPD leadership, the section of the text in which Weber’s influence is most apparent. One example of the idealism that such an analysis countered was the claim then prevalent that German socialism had been overtaken by bourgeois party members and thereby led astray. Against the rhetoric of German Social Democracy at the time, Michels pointed out that party functionaries recruited from the proletariat, not the bourgeoisie, were themselves a strong conservative force within the party. To demonstrate this assertion, Michels produced an inventory of SPD members in the Parliament from 1903 to 1906 according to their class origin and profession (1962 [1911]: 257). Based on his figures, the percentage of intellectuals and members of the bourgeoisie who were leaders in the party was relatively small. Michels also suggested that members from these ‘elite’ groups were more likely to hold radical or doctrinaire views, while members of the working classes who were elevated socially through promotion within the party were much more likely to abandon working-class values (1962 [1911]: 263). This analysis, to him, confirmed the unfortunate fact that the high ideals of social democracy could not bridge the gap between the interests of the masses and the interests of leaders. Instead, such a gap was the ‘real’ product of the organizational relationships, typified by a process of ‘embourgeoisement’ of proletarian party leaders, which characterized the German Social Democratic Party as an anti-democratic organization (1962 [1911]: 263–264).
The motif of anti-idealism is less apparent in Michels’ Weberian or social scientific discussion of social processes such as embourgeoisement, 20 but it comes through clearly in the way that he positions his empirical case study. The figures he reproduces regarding the ‘origin’ and ‘profession’ of socialist parliamentarians represent a reality that he saw as obscured by party ideology. At the root of the ‘iron law of oligarchy’ is the simple observation that the masses are always susceptible to leaders who may suppress the interests of ‘the people’ while using high-minded ideals to suggest that they are, in fact, protecting these same interests. Although he laments the tragic nature of democratic movements – ‘The democratic currents of history resemble successive waves. They break ever on the same shoal. They are ever renewed. This enduring spectacle is simultaneously encouraging and depressing’ (1962 [1911]: 371) – what makes Michels’ analysis of democracy an important part of the ‘social facts’ tradition in the study of democracy is his technique of confronting the goals of social democracy with a scientific analysis based on the ‘real’ relationships that define social organizations.
In this way Michels builds on the elite theory of Mosca, and on the political application of Pareto’s anti-positivist theory of social action. Although Pareto was the only one of the three who built a comprehensive theoretical system to understand social action in general, the work of each represented a general trend in social theory that rejected positivistic and idealistic formulations of action. Each also attempted a sober and scientific response to the populist energy and the perceived threat of social democrats. In a review of Michels’ Political Parties, Mosca characterized this response as an emerging ‘scientific school:’
In Italy … a scientific school has consolidated itself that does not fight democracy with the usual argument in favour of aristocracy or monarchy, but purely and simply denies the possibility of a true and sincere democratic government. This is a school which, if I may be allowed a neologism, is not anti-democratic but a-democratic… .
The combination and recurrence of anti-positivism, anti-idealism, and the theory of elites as theoretical motifs represent a critique of democracy that aimed at separating the ‘reality’ of social relations from the realm of ‘ideal’ ends. As Mosca’s label ‘a-democratic’ implied, the analysis of any system of politics, including democracy, should not attach moral or ethical content to the system itself. To identify one system as more moral or better able to deal with a specific set of ethical concerns is to obscure the actual functioning of political systems. This premise, which would eventually be called ‘analytical elitism’ (Bennett, 1978; Lenski, 1980), appeared neutral on political ‘ought’ questions, but it was founded on ‘facts’ that were a sedimentary mixture of fear-based mass psychology and selective historical experience. The normative bases of these ‘facts’ are easier to identify now, but the events of the Second World War only further entrenched these ‘social facts’ in the social scientific study of democracy.
Democracy as a method of control (Schumpeter)
Reading the scientific studies of democracy outlined above leaves little ambiguity as to whether these critics found the rise of social democracy to be a serious contemporary problem. As a response to this problem, the three motifs identified above represented a set of fundamental propositions that could ground a ‘realistic’ theory of democracy. These propositions asserted that: (1) political systems remain stable or are subject to change primarily on the basis of ‘non-rational’ factors; (2) these non-rational factors should be studied as ‘real existing’ social relations and not as idealized notions of how social relations ‘ought’ to be; and (3) in all political systems, real existing social relations are defined by the constitution of an elite ruling class that will always tend to consolidate its power in the absence of effective checks on that power.
These three foundational propositions alone do not make a coherent theory of democracy, but they do form an important conceptual groundwork for post-war attempts to study democracy scientifically. However, since social scientists in this period had experienced the violence of authoritarian political movements, many were less apt to adopt, wholesale, theories that were largely anti-democratic. To make the transition from theoretical motifs to established scientific ‘facts’, the foundational propositions above had to be extracted from critiques of democracy and applied to theories that posited democracy as a desirable, or at least a best possible, political system. The economist Joseph A. Schumpeter offered one prominent example of a theory that accepted these basic propositions but avoided the normative dissonance of anti-democratic conclusions.
Democracy reinvented or democracy without the ‘social’
The importance of Schumpeter’s Capitalism, Socialism and Democracy (2008 [1942]) is widely recognized, but the particular innovations attributed to this work vary along disciplinary lines. As an economist, Schumpeter was concerned with the economic mechanisms and institutions of capitalism that drove economic change. From this perspective, Capitalism, Socialism and Democracy was an attempt to understand the rapid economic and social changes that followed the First World War in a way that built on his just completed historical-theoretical treatise on Business Cycles (1939; see Shionoya, 1997; Swedberg, 1991). Like the ‘elite theorists’, Schumpeter primarily worked as a historical empiricist 21 and produced conclusions by inductively tracing trends of observed historical dynamics. Thus, also like the ‘elite theorists’, Schumpeter’s methodological approach produced ‘facts’ by identifying historical processes that were so regular that their naturalism was irrefutable. The elucidation of such facts unites his entire oeuvre.
At the same time, one section of Capitalism, Socialism and Democracy – on ‘Socialism and Democracy’ (Part IV, 2008 [1942]) – bears special consideration because it offered what remains one of the most widely accepted definitions of democracy in US social science (Diamond and Plattner, 1993b; Lipset, 1960; Medearis, 2001; Schmitter and Karl, 1996). For the most part, Schumpeter’s political philosophy, which he outlines in this section, is treated as separate from the other sections of Capitalism, Socialism and Democracy, 22 or even his other major works. 23 Although his definition of democracy proved to be influential, understanding this definition in the context of his broader methodological and theoretical approach to social life brings his conclusions much closer to the work of the ‘elite theorists’ discussed above. Schumpeter accepted many of the foundational assumptions of the ‘elite theorists’, but his argument in Capitalism, Socialism and Democracy went a step further in reformulating, or reinventing, a critical analysis of democratic ‘ideals’ as a positive theory of formal democracy. 24
Schumpeter begins his discussion of democracy by outlining the contemporary relevance of the classical theory of democracy. The theory, for him, invokes an ‘arrangement’ that realizes the ‘common good’ and reflects the will of the people (2008 [1942]: 250). This theory resonates for many as an ideal, but difficulties appear for Schumpeter once he begins to examine the ‘real existing’ circumstances of politics. For example, he questions whether there is such a thing as the ‘common good’, and whether this common good is necessarily a product of a ‘common will’. The classical doctrine of democracy, according to Schumpeter, attempts to assert the existence of these questionable ideals in two variations of the doctrine: the Rousseauian and the utilitarian (2008 [1942]: 250). In addition to claiming that both variations of the classical doctrine produce an idealist theory, Schumpeter goes on to argue that this theory also assumes an unrealistic distribution of reasoning among participating citizens. This principle of equality with regard to rationality in political matters is crucial to the ideal of a government by the people that is able to produce satisfactory political results for the people (2008 [1942]: 256). According to his appraisal of the voting public’s behaviour in most existing democracies, such a theory is clearly inaccurate.
By identifying the classical doctrine’s idealist ‘ends’ as well as its requirements for rationalist or ‘positivistic’ action on the part of common citizens, Schumpeter calls for a theory of democracy that recognizes that ‘the people’ are often ‘fooled, step by step into something they do not really want’ (2008 [1942]: 264). To build such a theory, he reflects on the social psychology of human behaviour in politics, and it is on this point that he explicitly linked his argument to both the popularity of Pareto’s Mind and Society in the social sciences 25 and the ‘uncomfortable truths’ behind Le Bon’s otherwise overstated study of crowds (2008 [1942]: 256–257). Echoing Le Bon’s remarks, he argues that not only is political action defined by non-rational impulses but, even in the most advanced democracies, ‘the typical citizen drops down to a lower level of mental performance as soon as he enters the political field’ (2008 [1942]: 262). For Schumpeter, then, the inability of citizens to act rationally with respect to political issues undermines the possibility of achieving a ‘common good’ through politics. While the classical doctrine had managed to persist based on the attractiveness of its ideals to the masses, the gulf between this doctrine and the ‘facts’ that determine the actual practice of democracy required an alternative theory.
The alternative theory of democracy proposed by Schumpeter begins with the assertion that the factual limitations of the classical doctrine were already widely accepted at the time (2008 [1942]: 269). In this way, he frames the ‘problem’ of democracy as a technical incoherence between theory and reality, and his solution is thus a simple technical alteration to the theory. Schumpeter concludes that the fundamental role of the people in a democracy is not to express a collective will, but to produce a government through free and fair elections (2008 [1942]: 269). Democracy in this definition is a methodological or procedural question rather than the achievement of a fictional ‘common good’. Such a simplified and ‘realistic’ theory had, for him, a number of explanatory advantages over the classical doctrine. These include an ‘efficient criterion’ for distinguishing democracies; a recognition of the difference between an ‘ideal’ will of the people and the functional will of a simple majority; and an emphasis on the relationship between democracy and individual freedom through the basic principle that any person is theoretically eligible for election (2008 [1942]: 269–272).
Another advantage of a theory of democracy narrowed to the principle of competitive elections, however, is that it recognizes the ‘vital fact of leadership’. Much like the theory of elites, Schumpeter states that ‘collectives act almost exclusively by accepting leadership – this is the dominant mechanism of practically any collective action which is more than a reflex’ (2008 [1942]: 270). There will always be leaders in society, so the simplest way to mitigate the tendency for leaders to consolidate their power is to give the people control over the mechanism that grants them power. This is accomplished through opening leadership positions to competition, and giving voters the right to remove a particular leader through their electoral power.
Schumpeter’s ‘alternative’ theory of democracy clearly invokes the earlier motifs of anti-positivism, anti-idealism, and the theory of elites to first strip down the classical doctrine to its most basic formal elements and then to rebuild democracy as a method that can work in the ‘real world’ of politics. However, the propositions distilled from these motifs take on the weight of facts only in the context of the larger historical project in Capitalism, Socialism and Democracy and his other works. The larger project of Capitalism is essentially an evolutionary argument about the fate of existing capitalist economies (Shionoya, 1997), but in keeping with his historical empiricism, Schumpeter also frames his analysis of democracy in evolutionary terms. To do this he compares the past formulations of bourgeois democracy with the ‘likely’ future problems of a socialist democracy. By reinventing democracy as a method, Schumpeter also effectively bypassed any ‘ideal’ ends in his analysis of whether ‘democratic’ mechanisms were more or less compatible with socialism. Of course, for him such a judgement relied on the historical ‘evidence’ of democracy in previous economic (capitalist) systems. It was from this perspective that he identified a set of baseline conditions for the functioning of a democracy. In brief these conditions were: (1) a sufficient number of high-quality members of a political class; (2) a clear limit to the effective range of political decisions; (3) a well-trained bureaucracy of good standing and tradition; and (4) a sufficient amount of what he calls ‘democratic self-control’ (2008 [1942]: 290–294). Historically, these conditions first appeared with ‘Gladstonian’ liberalism, or what he calls ‘individualist democracy’ (2008 [1942]: 126).
In the theoretical foundations of eighteenth-century bourgeois democracy and nineteenth-century social democracy, Schumpeter found a common acceptance of the rationalist principles outlined by the classical doctrine. However, the bourgeoisie of the late eighteenth century were the first to introduce democratic practices that met each of the above conditions to a sufficient degree, partially because their interests demanded a degree of separation between government and the sphere of economic affairs (2008 [1942]: 297). Moreover, Schumpeter argues that this separation of government and the capitalist economy did not just open the door for democracy, it actually helped to encourage the flourishing of reason associated with the Enlightenment. As he suggests:
The capitalist process rationalizes behaviour and ideas and by so doing chases from our minds, along with metaphysical belief, mystic and romantic ideas of all sorts. Thus it reshapes not only our methods of attaining our ends but also these ultimate ends themselves.
According to this argument, the rationalist thought demanded by capitalist processes had the effect of encouraging the ‘free thinking’ necessary for the development of democracy. At the same time, the claim that capitalism reshapes our ultimate ends is also consistent with the broader theory of social and economic change that unites Capitalism with his earlier works. In his second book on economic theory, The Theory of Economic Development (1934 [1911]), Schumpeter proposed a theory of change in capitalist economies in which entrepreneurs were singled out as the innovative engines that drive progress. 26 According to this argument, factors that were external to the economic system, such as the development of bourgeois democracy, appear as by-products of social changes produced by mechanisms internal to the capitalist economy. Similarly, in his treatise Business Cycles (1939), Schumpeter expands this argument to present a massive historical analysis of innovative ‘boom’ periods as well as corresponding and necessary crisis periods. Schumpeter’s larger concern – in his application of his business cycle theory to the events of the interwar period as well as his prediction in Capitalism that the existing capitalist system probably cannot survive (2008 [1942]: 61) – is that crises such as the depression of the 1930s might induce governments to interfere with the natural processes of capitalism. This interference could then restrict the entrepreneurial innovation that, history shows, will eventually pull the economy out of crisis. From this perspective, Capitalism is a somewhat pessimistic analysis of the possible future of a post-capitalist order that was brought about by the ‘free thinking’ that was itself a product of capitalism’s successful innovations. In this way, Schumpeter shares a number of the concerns of the earlier elite theorists.
These concerns become even clearer when Schumpeter compares the moderate success of eighteenth-century democracy to the possible future of a socialist democracy. He assumed that a socialist society could be ‘advanced’ enough to maintain some of the conditions for democracy, such as a qualified political class and a solid bureaucracy, but his crucial question was whether such a society could possibly exercise ‘democratic self-control’ (2008 [1942]: 298–299) once a government starts down the path of centralized economic management.
Schumpeter’s assertion that a socialist government would have difficulty meeting the necessary condition of ‘democratic self-control’ represented both an implicit rejection of socialism on grounds of ‘historical experience’ and a subtle assertion of the ‘social facts’ of democracy. For him, socialist societies required well intentioned and skilful bureaucrats to somehow manage the economy, the basis for social life, without restricting the population’s democratic freedom to make any number of irrational demands – a tall order for any governing bureaucracy, no matter how skilful. While the ‘automatic restrictions imposed upon the political sphere’ developed by the eighteenth century bourgeoisies acted as a buffer between economic life and the ‘inefficiencies of political procedure’ (2008 [1942]: 299), socialist democracy could easily get bogged down with the irrationalities of political participation.
Schumpeter’s moral philosophy
Although Schumpeter draws his conclusions from empirical studies of economic history, his opposition to the classical doctrine of democracy has a strong undertone of liberal moral philosophy and is apparent in his identification of individual freedom with market capitalism. This ‘negative freedom’ (Berlin, 1969) is actually much closer to the concept of freedom formulated by English utilitarianism than is initially implied in Schumpeter’s rejection of classical democracy’s utilitarian foundation. Like the social democratic inheritors of Rousseau, however, the theorists of utilitarian democracy (Bentham and Mill) based their politics on a moral philosophy of social life. To propose his alternative theory of democracy, Schumpeter therefore had to separate this liberal moral philosophy of social life – outlined in his discussion of ‘the civilization of capitalism’ (2008 [1942]: 121–130) – from any set notion of the political. This essentially amounted to reversing the means and ends of democracy and freedom. In the classical doctrine, self-government by the people was an end made possible by the means of a collective will or the aggregation of individual wills. Freedom, collective or individual, was therefore the means to the end goal of self-government. Schumpeter’s alternative theory of democracy stripped down each of the elements of the classical doctrine by labelling the ‘collective will’ as idealist, the utilitarian theory of aggregate individual action in politics as overly rationalist (positivistic), and the ideal of self-government as a historical impossibility. He then reversed the formula by asserting simple negative freedom as the only possible end and produced a theory of government that was most likely to ensure this end. What makes this theory of democracy a reinvention, therefore, is that it alters the role of democracy from that of a tool for the people’s self-expression and a driver of social change, to a conservative force that most appropriately resists social change.
Critics have pointed out that Schumpeter’s portrait of the classical doctrine is essentially a straw man (Pateman, 1970) and reduces a variety of classical ‘models’ (Held, 1987) to a single doctrine. However, as stated above, the doctrine of competitive leadership became a widely accepted definition of democracy and inspired a revival of the ‘elite theory’ in what Bachrach (1967) called ‘the theory of democratic elitism’. Work that built on the popularity of Schumpeter’s definition of democracy as method reiterated that democracy itself had ‘no overriding purpose to promote’ and that it had the important function of letting elites ‘exercise their rightful power effectively’ (Bachrach, 1967: 21). 27 An important reason for this popularity was that it gave a pro-democracy veneer to a set of social scientific propositions developed in theoretical systems critical of Marxist or socialist politics. This stripped down or ‘empty’ definition of democracy – which proposed the formal separation of government from the public sphere – opened a number of research avenues in political science and sociology that could focus on the functioning of a democracy without committing to either capitalist or socialist principles. 28 This apparent neutrality differentiated Schumpeter’s work from other similar arguments that more explicitly set democracy against socialism (see, for example, Hayek, 2007 [1944]).
Mechanisms of control
Schumpeter’s theory of democracy has been categorized as a technocratic or ‘elite’ theory of democracy (Bachrach, 1967; Held, 1987; Medearis, 2001), and indeed when it is compared with critiques of central planning (Hayek, 2007 [1944]) or ‘managerialism’ (Burnham, 1941) written during the same period, his ‘alternative’ theory seems like a logical answer to the tyranny of unchecked government bureaucracy. However, when understood as part of a tradition concerned with looking past democratic ideals to uncover the ‘social facts’ observable in democratic societies, the focus on elites becomes secondary. This is primarily because Schumpeter’s reinvented doctrine of democracy, like the conclusions of the earlier critics, first and foremost aimed at restraining the supposedly non-rational impulses of the masses.
A related concern for the ‘elite theorists’ mentioned above, as well as for Schumpeter, was to formulate an empirically – or ‘scientifically’ – grounded opposition to Marxian theories that made social class relationships a determinant of societal change (Bottomore, 1964: 18). In the Marxian system, the early critics rightly saw a politicization of the everyday experiences of the working classes. However, by constructing theoretical systems that produced ‘facts’ demonstrating the impossibility of social democracy, they were able to separate ‘political’ questions from the non-rational excess of ‘the social’. What Schumpeter’s new doctrine added to the ‘social facts’ tradition was a technical solution that purports to strip away the idealism, or the ‘ends’, of democracy once and for all by reducing it to a mechanism that actively ensures the separation of the social and the political. This separation – which is based on the threat (or promise) of the removal of a leader from office through elections – is functional because ‘the people’ are protected from their leaders, and the leadership is insulated from the excessive and irrational desires of ‘the people’. Moreover, with regard to Schumpeter’s more general theory of economic change, the irrational desires of ‘the people’ clearly conflict with the rational processes of capitalism embodied by the figure of the entrepreneur.
Dismissing this doctrine as just another form of liberal idealism would be an easy conclusion, but such a conclusion would miss the important place Schumpeter’s work had in a larger process that opened the political realm to social scientific analyses based on market-style mechanisms of self-regulation. Schumpeter, in fact, compared voter behaviour to the behaviour of consumers by pointing out that both citizens and consumers often fail to act according to their best interests (2008 [1942]: 257). In politics, as well as in the market, individuals can themselves become ‘experts’ about things that matter to their daily lives, but as a political issue or understanding of a product gets further from immediate experience, the responsibility for comprehensive understanding dissipates. The larger processes of the market or national affairs that do not have the same import to an individual’s daily life are thus very difficult for the average person to comprehend in a rational way. Schumpeter frames this disconnect as a lost ‘sense of reality’ (2008 [1942]: 261) that defines most political discourse on the important issues of the day. More than just separating the masses from particular political issues, their non-rational impulses therefore separate them from a larger reality of political and economic systems.
Conclusion
In the introduction to Walter Lippman’s Public Opinion (1965 [1922]), the author poses the problem of democracy as a gap between the reality of the ‘the world outside’ with ‘the pictures in our heads’. Because the latter is often so far from the reality of the former, Lippmann argued that
… representative government, either in what is ordinarily called politics, or in industry, cannot be worked successfully, no matter what the basis of election, unless there is an independent, expert organization for making the unseen facts intelligible to those who have to make the decisions.
What the motifs of elite theory, anti-positivism, and anti-idealism offered was a way to integrate the complexity of ‘the world outside’ into universal social theories that posited regular patterns of change in political systems. The early critics of democracy called for a scientific intervention that could ‘make intelligible’ the ‘unseen fact’ that society was governed by the non-rational excesses that were only predictable in their unpredictability. With the fact of elite rule and the unreality of the ideal of popular sovereignty, Schumpeter was able to translate this unpredictability into a new technocratic approach to democracy that could be workable as a mechanism to control these excesses rather than a vehicle to express them.
The early critics’ public distrust of democracy – or eventual support of fascism, as in the case of Michels (Beetham, 1977b) – may betray a clear normative position in each of their scholarly works, but simply identifying these normative positions is not the purpose of this article. Guided by motifs that translated a more general current of mistrust regarding workers’ movements into the language of ‘science’, these three critics produced theoretical and methodological innovations that they used to fuse their politically oriented normative critiques with ‘scientific’ analyses. As a product of these innovations, the ‘social facts’ of democracy represented black boxes upon which theoretical systems critical of social democracy were built. Schumpeter further naturalized these black boxes by reinventing democracy as a set of formal-procedural mechanisms that appeared as by-products of economic historical processes.
While critics of elite theory have always argued that democracy must have a normative content, such challenges have been, and continue to be, deflected by appeals to science or an empirically observed historical ‘reality’. Returning to early twentieth-century attempts to deal with the problem of democracy is instructive because the tension between the ideals and the facts of democracy continues to steer debates about the prospects for new democracies or the revitalization of old democracies. The so-called ‘Arab Spring’ and the ‘Occupy Wall Street’ movement are two examples that explicitly test both the ideals and the ‘realities’ of democracy. This tension is at the heart of the many competing definitions of democracy, and any critical appraisal of these conceptualizations could therefore benefit from opening the black boxes that have shaped the modern social science of democracy.