Also by Jean Grugel.
POLITICS AND DEVELOPMENT IN THE CARIBBEAN BASIN FRANCO'S SPAIN (with T. Rees)
REGIONALISM ACROSS THE NORTH-SOUTH DIVIDE (edited with W. Hoot)
DEMOCRACY WITHOUT BORDERS (editor)
Democratization
A Critical Introduction
Jean Grugel
i i
50 Democratization
education was much higher for the democratic countries (Lipset 1959). He then used this to suggest causation. Put simply, Lipset claimed to have proved that more telephones, more cars, more consumption - in sum, more capitalism - leads to more democracy. Even scholars broadly sympathetic to modernization's underlying thesis have found themselves forced to reduce these sweeping claims. Diamond (1992), for example, points out that Lipset's own data indicated higher levels of economic development within European non-democracies than in Latin American democracies, suggesting that democracy required the presence of factors other than economic growth. Vanhanen (1990) and Hadenius (1992) have also modified Lipset's claim, from causality to one of correlation. Przeworski and Limongi (1997) suggest that the evidence supports the thesis that democracy survives better in wealthier nations but nol the original proposition that democracy is a simple consequence of economic growth.
Modernization Theory Today
Leftwich (1996) offers the most forceful contemporary restatement of modernization. He applies it, logically enough, only to developing states. He argues that economic development, whether in a democratic political setting or not, will inevitably produce democracy in the long term. As a result he recommends that: 'the West should ... support only those dedicated and determined developmental elites which are seriously bent on promoting economic growth, whether democratic or not. For by helping them to raise the level of economic development it will help them also to establish or consolidate the real internal conditions for lasting democracy' (Leftwich 1996: 329; italics in the orginal).
It is rare, however, for modernization to have survived in such an unre-formed way. Generally, today's modernizadonists do not claim direct causality between capitalism and democracy. Diamond has been particularly influential in updating modernization theory. He has picked up the 1960s concern with mass participation and political culture in newly 'modernized' states and emphasizes, in particular, the role of political culture and a dynamic civil society for democratization. In the process, he has shifted modernization away from a discussion on the causes of democracy towards a focus on consolidation. He argues that long-term democratic consolidation must encompass a shift in political culture (Diamond 1999). Emphasizing the role of civil society and civic freedoms means, in fact, that he is pessimistic about the chances for sustained democratization in much of the developing world because civil society
Theories of Democratization 51
is frequently poorly articulated and weak, and the political culture of institutions and electoral regimes is 'shallow, exclusive, unaccountable and abusive of individual and group rights' (Diamond 1996: 34).
Diamond's work is of considerable academic importance. But this new version of modernization theory is just as profoundly embedded in Western policy-making circles. For Cammack (1994), its success is due to the fact that it addresses a major issue in global politics, namely the difficulties of governance in an era of mass participation, and, furthermore, it does so from the perspective of conservative global elites. Not surprisingly, then, modernization remains the vision behind a number of democracy-promoting initiatives, especially those emerging from US governmental circles. In sum, modernization theory retains vitality and influence through its ability to identify the apparent link between capitalism and democracy. However it is unable to explain why trends to democratization are so often contradictory and partial.
Historical Sociology
Historical sociology is a kind of 'macrohistory* in which history is 'the instrument by which structures are discovered invisible to the unaided eye' (Collins 1999: 1). It is because of this emphasis on structures that the approach is sometimes termed 'structuralism'. 'Historical sociology' and 'structuralism' are often used interchangeably in democratization studies. An important strand of historical sociology has been the search to identify different trajectories of slate development or paths to modernity, through, for example, war or revolution (Skocpol 1979; Tilly 1990). The historical/sociological approach to democratization has two particular intellectual origins. In part, it arose out of a reaction to the excessively society-based accounts of political change implicit in behav-iouralism in the 1960s, and offers instead a state-centred view. It is therefore part of the intellectual labour of 'bringing the state back in' to politics (Evans , Rueschmeyer and Skocpol 1985). It also drew explicitly on a critique of the short-termism and causal simplicity of modernization as an explanation of democratization (Rueschmeyer, Stephens and Stephens 1992). It is, inevitably, a much more diffuse approach to democratization than modernization theory, with a primary interest in explaining, not predicting, outcomes.
Structuralists are interested in how the changing relationship between the state, understood in the Weberian sense of 'a human community that (successfully) claims the monopoly of the legitimate use of physical
32 Democratization
Theories of Democratization 53
force wilhin a given territory' (Skocpo! 1985: 7; italics in the original), and classes shapes the political system. As such, they admit an important role for collective actors. They are agreed that democracies do not come into being overnight; nor does democracy happen simply because some people (individuals, groups, or classes) will it into existence. Structuralists trace the transformation of the state through class conflict over time, in order to explain how democracy - which they see as state transformation - has sometimes emerged. Structuralism also contains elements of apolitical economy of democratization in that it emphasizes how changes in the economy - for example the expansion of production for the market - lead to social or class conflict, although economic change is not, on its own, regarded as determining political outcomes. Unlike the wave approach of modernization theory, historical sociology identifies factors that are distinctive to particular cases.
Barrington Moore's (1966) major study of political change constitutes a significant milestone for historical/sociological understanding of democratization. His comparative analysis of eight 'big' countries, Britain, France, the US, Germany, Russia, Japan, China and India, through the nineteenth century and into the twentieth, identified the different historical trajectories that each had travelled to reach modernity. Fascism and Communism simply constituted a different version of modernity, resulting from very different sets of relationships between collective actors and states. For Moore, outcomes depended on the interactions between three important classes - the peasantry, the landed upper class and the bourgeoisie. Essentially, democracy occurred when
• the 'peasant question' was solved by the gradual elimination of peasant agriculture and the rise of opportunities for transforming the peasantry into urban workers through the expansion of towns and industrial employment opportunities; and
• the landed class was defeated and transformed in its struggles for control of the state by the rising bourgeoisie.
This latter was crucial in determining whether democracy or a form of dictatorship emerged (see Box 3.1).
Moore's work concentrated on the emergence of the first democracies. It was modified by Rueschmeyer, Stephens and Stephens (1992) in the light of later history and the expansion of the number of parliamentary and stable democracies. They describe their theoretical framework as part of the 'new comparative political economy' (Rueschmeyer, Stephens and Stephens 1992: 40). By this, they mean that they view the
Box 3.1 Routes to Modernity
Barrington Moore identified three routes to modernity: a bourgeois revolution, leading ta capitalism and democracy; revolution from above, leading to industrialization and fascism; and revolution from below, leading lo communism.
The Bourgeois Revolution
Moore's understanding of the bourgeois path was based on the historical experiences of Britain, France and the US. He conceptualized the transformation of the pre-modern state into a democracy as a result of two stages:
> the reduction in the overall size of the peasantry and an end to its organic dependence on the landed class; and
• a realignment of upper-class interests around the dominance of commercial and industrial interests.
Revolution from Above
This is the path of conservative modernization. Exemplified most clearly by Germany and Japan, it combines the development of capitalism in agriculture and industry, alongside state-directed change. In both Germany and Japan, the tension between economic modernization and attempts to prevent social change led to the rise of militarism and ultimately to fascism. Revolution from above was a result of:
• the survival dí a large small and middle peasantry, despite the rise of the market;
• the emergence of commercially-minded landed classes; and
• the development of a centralized and strong state.
Revolution from Below
This is the path to modernity through communism and peasant revolution. It occurred in the twentieth century in Russia and China. Commercial agriculture failed to emerge in either country, although there were attempts at modernization and a significant increase in labour repression in the nineteenth century. The social institutions of the peasantry survived intact into the modern era at the same time as the strength of royal bureaucracies prevented the emergence of commercially-minded landed classes. Revolution from below thus depended on:
• the survival, numerically and culturally, of the peasantry;
• a weak landed class; and
• an absolutist state.
54 Democratization
Theories of Democratization 55
political system of a particular country in relationship to broader questions of social power. Their work draws on a synthesis of scholarship from classical sociology and Marxism. It was ground-breaking because of its stress on the impact of what they termed 'three power structures': relative class power, the role of the state and the impact of transnational power structures.
They draw from Marxism a view that social class and class conflict constitute the starting point for an analysis of power and the state. They add to Moore's three-class schema, with its emphasis on rural change, a discussion of other subordinate classes, and of the urban working class in particular. But their emphasis on class divisions and class struggle is modified by a recognition of the role of the state and the role of the state system. In particular, they argue that whilst states have a special dependence upon capitalists under capitalism this has not always prevented working-class organizations from reforming the state. In other words, they see democratization as the imposition of reforms on a capitalist state, not as an automatic outcome from the development of capitalist relations of production. Without successful and self-conscious reformist strategies on the part of the subordinated classes, capitalist states will, in fact, almost inevitably be authoritarian (see Box 3.2). Furthermore, they suggest that a third dimension influences the nature of the state: the transnational context. This is particularly so in the case of the under-
Box 3.2   The Role of the Working Class in Democratization
Rueschmeyer, Stephens and Stephens (1992) assert the central importance of urban working classes for democratization:
The organized working class appeared everywhere as a key actor in the development of full democracy ... In most cases, organized workers played an important role in the development of restricted democracy as well. The Latin American ... working class played a lesser role in the historical events there: the relative weakness of the working class certainly has contributed to the infrequence of full democracy in the region and to the instability of democracy where it did emerge. ... In all regions, however, pressure from the organized working class alone was insufficient to bring about the introduction of democracy; the working class needed allies. ... Democracy could only be established if (1) landlords were an insignificant force, or (2) they were not dependent on a large supply of cheap labor, or (3) they did not control the state. (Rueschmeyer, Stephens and Stephens 1992: 270)
developed and dependent countries. In any analysis of democratization struggles,, therefore, the role of geo-political factors will be important. l-Towever, they remain agnostic as to 'the overall relationship between democracy and political/economic dependence in transnational relations' (Rueschmeyer, Stephens and Stephens 1992: 73). In other words, it is unclear whether external dependence supports or hinders democratization.
Evaluating Historical Sociology
The strength of historical sociology is that it is richly grounded and explanatory; and that it provides the possibility of comparison across time as well as across countries or regions. However, historical and structural approaches have been the subject of a number of criticisms. Historical sociology has become largely unfashionable, like all structuralist explanations of social change. Structuralism has, in general, fallen foul of the rediscovery of individual agency and volition in politics, of the questioning of Marxian class analysis and of the post-modern suggestion that power is too diffuse a concept to be understood in any static way; it is, instead, located in changing and fluid relationships. The major critiques of structuralism have therefore been both ontological and epistemological: its view of the world is too simple or simply wrong. AsPrzeworski (1991: 96) putit: 'in this formulation the outcome is uniquely determined by conditions, and history goes on without anyone ever doing anything'. In fact, however, historical sociology does recognize a considerable role for agency in processes of political transformation. The main agents of change are classes, or even the state. But thise notion of collective action is not sufficient to satisfy critics who only accept individuals as agents.
Other, more empirically-based, criticisms have also been forthcoming. Structuralism, with its emphasis on long-term historical change, seemed unable to account for the onset of sudden democratization in societies such as East and Central Europe and the countries of the ex-Soviet Union, where there was apparently little evidence of class agitation or struggle for democracy, except shortly before the collapse of authoritarianism. It was logical, therefore, especially in the light of the rise of agency-based theories of political behaviour through the 1980s, that dissatisfaction with structuralism would lead directly to a new agency-centred paradigm of democratization. Before analyzing the rise of 'transition studies', however, we examine the utility of historical sociology for analysis of contemporary deraocratizations.
56 Democratization Historical Sociology Today
An analysis of structures has particularly fed into the debate surrounding the consolidation of democracy. The idea that struggle and confrontation after the immediate transition can be symptoms of democratization comes from structuralism. It is therefore an appropriate tool for the analysis of post-transition systems. Structural analysis recognizes the fundamental class antagonisms of capitalist societies and suggests that conflict is a 'normal' part of democracy. It provides the researcher with the tools to question the idea that democratization requires the subordination of sectional and class interests for democracy to take root. By linking democracy with conflict, structuralism sees confrontation as a normal part of the pattern of the emerging democratic order. A structuralist perspective, with its emphasis on history, conflict, class and the state can also contribute to explanations of partial or incomplete democratizations. Gariorowski and Power (1998) argue that the chances for democratic consolidation are affected by 'development-related socio-economic factors, the contagion effect of democratic neighbors and high inflation'. They therefore draw the conclusion that all explanations of democratization should be placed within a broader structural perspective in order to fully understand the process of political change. Finally, structuralism is important in contextualizing and situating the debate about democratization. It allows for the identification of global structures that condition and shape the environment in which democratization takes place and points to the importance of 'underlying economic conditions and social forces' in democratization (Haggard and Kaufman 1997).
We return to the salience of historical sociology for understanding contemporary democratizations below.
Transition Studies
The transition approach, or, as it is sometimes termed, the agency approach, sees democracy as created by conscious, committed actors, providing that they possess a degree of luck and show a willingness to compromise. Democracy is not, therefore, a question of waiting for economic conditions to mature or the political struggles unleashed by economic change to be won. The divide between agency-centred scholars, on the one hand, and structuralists and modernization theorists, on the other, turns on the roles of actors, structure, culture and class relations
Theories of Democratization   51
in democratization and regime change. The transition school argues that both modernizationists and structuralists see the economy, history and development as overdetermining political outcomes.
For structuralists and the modernization school, democracy is an exceptional outcome which has occurred in only a few areas of the globe. It cannot be reproduced in countries where either the required levels of development are absent or where the class or social structure is unfavourable to it. By contrast, the attractions of the transition approach lie precisely in the fact that it questions these rather pessimistic assumptions. Agency perspectives suggest that democracy can be created independent of the structural context. The optimism of transitology accounts in large measure for its success, politically and academically, for this seemed to be precisely what was happening at the end of the 1980s. By implication, therefore, the transition approach presumes that the chances for spreading democracy in the contemporary world order are good. It hypothesizes successful outcomes for democracy if elites can learn the 'right' way to proceed.
The intellectual starting point for transition approaches is Rustow's (1970) critique of modernization. Rustow argues that the flaw of modernization theory is that it mistakes the 'functional' features of mature democracies - what makes them flourish - for 'genetic' causes of new democracies - what brings them into being. In contrast, he suggests tiiat the only condition for democracy is a unified national state: 'the vast majority of citizens in a democracy-to-be must have no doubt or mental reservations as to which political community they belong to' (Rustow 1970: 350). He then hypothesizes that the creation of democracy is a dynamic process in the context of 'a prolonged and inconclusive political struggle' (Rustow 1970: 352), which passes through three stages -a preparatory phase, a decision phase, in which the choices and negotiations of 'a small circle of leaders' play a particularly crucial role (Rustow 1970: 356), and a habituation phase in which citizens and leaders fully adapt to the new system. These stages were later transformed into liberalization, transition and consolidation.
In 1986, Schmitter, O'Donnell and Whitehead edited the seminal tran-sitologist analysis of democratization, Transitions from Authoritarian Rule which became the key reference for transition studies. It marked the beginning of a massive literature which focused on the processes of democratization by examining the interactions, pacts and bargains struck between authoritarian leaders and the democratic opposition. These deals led to a 'transition', a kind of half-way house between authoritarianism and consolidated democracy, in which the institutional rules are laid down for the practice of democracy. Successful transitions, it was
38 Democratization
emphasized, depend upon agreements between elites, including generally the outgoing authoritarian leaders. In none of their cases did democracy appear predetermined by the structural situation in which the struggles take place and pacts are made. They concluded that skilful leadership, aided by luck, was the key to outcomes which lead to the establishment of democratic procedures for government.
The transition approach thus pioneered a separation of political negotiations from economic circumstances. This was partly due to the insistence on contingency and negotiations and the rejection of functional determinism; but a normative note also crept in, and some transition studies warned would-be democratizers of the dangers of mixing transition with economic redistribution. This was evident in the influential work of Adam Przeworski, who argued:
we cannot avoid the possibility that a transition to democracy can be made only at the cost of leaving economic relations intact, not only the structure of production but even the distribution of income. ... Democracy in the political realm has historically co-existed with exploitation and oppression at the workplace, within the schools, within bureaucracies and within families. (Przeworski 1986: 63)
By 1995, Schmitter was offering the following as advice for would-be democratizers:
to the extent that it is possible, political choices should be give priority over economic ones. Incentives for the restructuring of national political institutions should precede, temporally and functionally, those aimed at reforming national systems of production and distribution. (Schmitter 1995: 33)
Agency-centred theories of democratization have the virtue that they situate the study of democracy within mainstream political science methodologies and epistemologies. By drawing on theories of political action - and by implication abandoning either sociological or historical approaches - transition studies offer a 'political' explanation of democratization. Democratization is seen as a process. For process-oriented scholars, 'choices are caught up in a continuous redefinition of actors' perceptions of preferences and constraints' (Kitschelt 1992: 1028). The task is to trace and explain these processes. Przeworski (1991: 19) pioneered a rational choice explanation of transition processes, which ran parallel with the rise to prominence of rational choice in other areas of the study of politics. He argued that '[w]hat matters for the stability of any regime
Theories of Democratization 59
is not the legitimacy of this particular system of domination but the presence of preferable alternatives' (Przeworski 1986: 51-2).The strategies adopted by key actors are dictated by cost - benefit calculations:
If the expected gains for the opposition (more freedoms, material well-being and political participation) are higher'than the risks (danger to life, imprisonment etc.) then it will continue to press for change. In turn, the regime elite is most likely to split into hard- and soft-liners along the two basic alternatives, either to suppress the opposition or to regain legitimacy by using a strategy of liberalization. Successful transition is most likely when soft-liners ally with the opposition and are transformed in this process into reformers. (Schmitz and Sell 1999: 31-2)
Above all, transition studies emphasize the agency and interactions of elites. They have thus made an important contribution by detailing how elite pacts, formal or informal, or compromises shape new democracies in the first place and contribute to their institutionalization (see Box 3.3). Elite-led democratization is viewed as positive for post-transition stability. But there are also some problems that result from pact-making. Karl (1990: 11) argues that foundational pacts can be a means through which economic elites assure themselves of the 'right' to continue to
. exploit a majority of the population and are therefore essentially 'antidemocratic mechanisms'. Hagopian (1992) is similarly critical: she argues that it has been precisely the behaviour of the 'political class'
: during the Brazilian transition that has prevented democracy taking root, leading to what she terms a 'compromised consolidation'. Przeworksi (1995: 54) recognizes that while pact-making creates stability, it can also lead to the institutionalization of forms of political exclusion. In other words, pacts shape the terms of transition and those terms may not be conducive to democratization in the long term.
.....  Because of the emphasis on elites, agency-centred perspectives have
; devoted relatively little time to the analysis of civil society, associational life, social and political struggles and citizenship in the construction of democracy. As a result, the transition perspective takes a rather ambiguous attitude to the role of civil society in democratization. Some agency scholars have seen an active civil society or social activism as unimportant for democratic consolidation. Przeworski (1991) suggested that
-in some cases popular mobilization has been detrimental to democratization since it threatened the interests of powerful elites who then went
-to considerable lengths to close down tentative experiments in political liberalization. This position was modified later, by a recognition of the
60 Democratization
Theories of Democratization 61
Box 3.3   Pact-Making and Democratization
The importance of pact-making has been a dominant theme of transition studies. Pact-making is a way of describing the 'establishment of substantial consensus among elites concerning the rules of the democratic game and the worth of democratic institutions' (Burton, Gunther and Higley 1992: 3). According to O'Donnell and Schmitter (1986: 37), elite pacts are 'an explicit, but not always publicly explicated or justified, agreement between a select set of actors which seeks to define (or better to redefine) rules governing the exercise of power on the basis of mutual guarantees for the "vital interests" of those entering into it.' They claim that elite pacts facilitate 'an institutional breakthrough' and make negotiations over the institutional format of the new democracy possible. The main benefit identified with pact-making is creating a stable environment and limiting uncertainty during the transition.
The literature on pact-making has generally stressed the role of 'the political class': politicians, important party officials, bureaucrats, and office-holders. But Di Palma (1990) argues that accommodating business and labour, as well as the state, is important. The Spanish transition provided an important example for understanding pact-making. The defining moment of the Spanish transition was the establishment of an elite agreement dirough the creation of a new democratic constitution, but the transition was aided by the creation of a tripartite economic agreement, the Moncloa Pacts. The Spanish case is thus frequently taken as the paradigmatic example of transition through pact-making and to some extent its success led to imitative pact-making in Poland, Hungary and Czechoslovakia.
difficulties that weak civil societies pose in new democracies (Przeworski 1995).
Evaluating Transition Studies
Transition perspectives have shed light on the micro-processes of regime breakdown, the opening of transitions and the mechanisms of democratic construction. But they have also been criticized for being overly elitist, excessively empirical and voluntaristic. Furthermore, they have tended to apply theories constructed out of the experiences of Southern Europe and Latin America to regions which are culturally, politically and economically different, such as East and Central Europe, the territories of the ex-Soviet Union, Africa and China.
Remmer (1991) has articulated most clearly the view that transition theory is a 'retreat into voluntarism' or 'barefoot empiricism'. This and
its other problems, it could be argued, stem from its excessively narrow understanding of democracy. Democracy is visualized as a set of procedures for government negotiated by and between political leaders. Thus the transition approach separates democracy from its essential meaning as rule by the people and conceptualizes it principally as the establishment of a set of governing institutions. At the same time, the perspective's elitism consigns the mass of the people to a bystander role in the creation of new regimes. This ignores empirical evidence which points to the role of popular struggles in some transitions as the determining element in unleashing democratization in the first place. It also ignores the importance of civil society in democratization or at best 1 confines it to a purely instrumental role (Baker 1999).
Typically, the transitology literature sets out a straightforwardly insti-lutionalist and electoralist definition of democracy, then quickly passes on to identifying mechanisms of regime change as the more interesting phenomenon. In an important article, Schmitter and Karl (1993) attempt to describe what 'democracy is ... and is not'. Although they recognize that democracy is contingent upon socioeconomic performance and entrenched state structures and policy practices and reject electoralism in favour of arguing that democracy must offer a variety of competidve processes and channels for the expression of interest apart from elections, they ultimately prefer to concentrate on democracy as a set of procedures for creating institutions and the government (Schmitter and Karl 1993). They argue that democracy is too abstract a concept to tie down in any useful way. Instead, they suggest that it makes more sense to establish a 'procedural minimum' for a functioning 'democracy' and work from this.
By focusing mainly on short-term changes, transitologists fail to examine deep-rooted obstacles to democratization over the long term. When democratizations go wrong it is, by implication, because individuals 'get it wrong'. The transitology approach does not explain adequately why outcomes are different, except by presuming inadequate leadership styles or the adoption of incorrect policies. It does not distinguish between outcomes - they are all 'democratic' in some way once elections are held and authoritarian office holders are forced out - or explain why apparently democratic institutions can operate in non-democratic ways. And finally, it omits to analyze in any depth the roles of culture, development, history or the internationalization of politics in democratization. In sum, it does not pay sufficient attention to structural contexts and constraints. Yet as more authoritarian regimes collapsed in different parts of the globe, the understanding of democracy had to be
62 Democratization
Theories of Democratization 63
stretched, confused and weakened in order to fit regimes that sometimes barely appeared to qualify for the label. At the same time, a number of 'transitions to democracy', for which hope was initially expressed, have ended very far indeed from the democratic ideal, indicating that the 'catch-all' definition of democracy was rather loo loose.
Transition studies offered a general approach to democratization based on an interpretation of experiences of Southern Europe and Latin America. Its relevance elsewhere has been questioned. For Pei (1994: ]) it was possible to 'treat the process of regime transition from communism as identical to the regime transitions from authoritarianism that occurred between the mid-1970s and mid-1980s in southern Europe, Latin America and East Asia', with the proviso that Russia and China experienced a 'dual transition' (author's italics) to democracy and to market capitalism. Bunce (1995a) and Parrott (1997), however, strongly disagree. According to Parrott (1997: 2)
the relevance of the paradigms of democratization ... is far from self-evident. Just as some economists have challenged the applicability of models drawn from non-communist societies to the dilemmas of economic reform in postcommunist states, some political scientists have questioned whether paradigms of democratizations drawn from non-communist countries are relevant to the study of postcommunist political change.
He argues that, in particular, transition theory pays insufficient attention to the problems of ethnicity, the 'legacy of large internal ethnic "dias-poras" and the emergence of ultranationalism in internal ethnic "homelands'" in the East (Parrott 1997: 10). We return to this important issue in Chapter 9.
Transition Studies Today
Transitology is largely responsible for the suggestion that democratization constitutes the most appropriate paradigm through which to analyze the complex process of regime decay and political change in apparently dissimilar countries such as Spain, Portugal, South Africa, Mozambique, Nigeria, El Salvador, Mexico, Turkey, Poland the ex-Soviet Union and China since the 1970s. For, by divesting democracy of its structural context, the transition perspective suggests that democracy can take root outside Western Europe and the US and that the global upheavals of the 1980s and 1990s were, in fact, struggles for democracy. Thus transitol-
ogy is inherently responsible for the global scope of democratization studies.
Transition studies havegenerated an important literature on the state and transition. According to Przeworski (3986: 59), the emergence of democracy does not necessarily signify that all key political actors have become democrats; rather it means that the opposition and the soft-liners in government have persuaded hard-liners that there is more to gain from cooperating with change than from opposing it. Hence a democratic transition is only a 'contingent institutional compromise' (Prezworski 1986: 59). Consequently, the new institutions take on an important role in their own right. The design of the new institutions is paramount for the success of the transition:
if a peaceful transition to democracy is to be possible, the first problem to be solved is how to institutionalize uncertainty without threatening the interests of those who can still reverse the process. The solutions to the democratic compromise consist of institutions. (Przeworski 1986: 60)
Transitology is therefore responsible for the emphasis in contemporary studies on the creation of institutions, the writing of constitutions and the choice of electoral systems.
Furthermore, transition studies have shaped academic perceptions that the micro-politics of democratization are significant. Studies of transition have emphasized agency, negotiation, compromise and the politics of change. They have also emphasized the importance of distinguishing different stages of democratization - liberalization, transition and consolidation. In some of his work, in fact, O'Donnell (1992: 18) goes further and talks of 'two transitions', the first an empirically verifiable transition from authoritarianism, and the second, the creation of a 'political democracy (or polyarchy according to Dahl), which may coexist with varying degrees of democratization in the economic, social and cultural spheres'. The result of this approach has been to allow the disaggregation of research into different moments, with the ultimate aim, according to O'Donnell, of specifying the relationship between political change on the one hand and socio-economic and cultural change on the other. In practice, however, transition research has emphasized political factors and democracy as a set of institutional practices which do not transform social, economic or cultural power relations. Not surprisingly, therefore, transition studies offer a vision of democracy stripped of its revolutionary potential: 'the wider picture that
64 Democratization
emerges is of a near-consensus ... thai actually existing liberal democracy is the only form of democracy on offer' (Baker 1999).
An Alternative Approach: The State, Civil Society and Globalization
The theories of democratization that we have identified so far vary in terms of the importance they allocate to rationality, culture, individual and collective action, economic development, social conflict and transnational factors. They draw on different epistemologies and explanations of social change. Taken together, these studies have generated a rich body of literature and have established the study of democratization as a core area of social sciences. They have shed light on different aspects of regime breakdown, transition and the nature of immediate post-transition politics. In some cases, they also illuminate aspects of why some systems are able to move on towards consolidation and others do not, although their main focus has generally not been on consolidation but on causation.
This was an appropriate focus for research during the first phase of democratization. However, some contemporary experiments in democracy are now more than twenty-five years old. Logically enough, academic interest now centres on the survivability of new democracies and the quality of democracy, not the number of transitions that are taking place. Democratization is a risky enterprise and experiments that begin with transition do not always end in consolidation. Yet why democracy succeeds in some cases and not in others is not always clear. Theory has not yet quite caught up with this changing research agenda. New approaches, that explain what happens after the initial transition as well as during it, are needed.
There is therefore a need to explain democratization holistically. In order to do so, this book draws on a framework that builds on the insights garnered from the historical sociology approach in particular. Structures, in other words, are vitally important for explaining outcomes. But the key contribution made by the transition perspective to the democratization debate - namely that democratization is a dynamic process, shaped by human behaviour and choices - is also centrally important. Actors, whether collective or individual, engage in struggles to transform authoritarian states and to build democracy. But they operate within structured environments. The options that are open to them are crucially shaped by the weight of structures such as the patterns of interaction between the
Theories of Democratization 65
state and society, traditions of organization and mobilization, state capacity and the global order. This is not to say that outcomes are predetermined in any way. Democratization is not inevitable in some countries; nor are attempts at democracy condemned to failure in others. But the chances for democracy are certainly greater in some societies than others. Our framework suggests that the chances for democratization are furthered by economic development, the development of a complex state and the emergence of a strong, working class or other subaltern groups that organize to promote political change. Democratization, in short, requires collective action - that of classes or social movements - more than the agency of particular individuals.
.. Democratization became a global movements at the same time as the new global political economy of marketization and liberalization emerged. This points both to the salience of the transnational context for understanding democratization and the importance of adopting a comparative economy focus similar to that pioneered by Rueschmeyer, Stephens and Stephens (1992) in order to explain outcomes. The shift to more open models of economic development, especially in post-communist and developing countries, has led to increased pressure for political change, has transformed the capabilities and ideology of the state and has created very different opportunities for state-society interaction. These inevitably shape the democratization process. It is important, huwever, not to assume that globalization is inevitably a positive force for democratization. In fact, the evidence that is presented through the case studies points to the ambiguous role of globalization and liberalization in democratization. Like Rueschmeyer, Stephens and Stephens (1992), we are therefore agnostic as to whether increased transnational-tzation leads to democratization and this is reflected in our framework.
Our alternative framework makes use of three key concepts, namely the stale, civil society and globalization or the global order. This framework can be used for the analysis of the problemcttique of consolidation as well as of transition and has the advantage that it incorporates within it a substantive understanding of democracy. The aim is to shed light not only on the onset of democratization but also to explain the very different trajectories, processes and outcomes that are grouped together under the banner of 'democratization'. In reality, of course, these three dimensions, state, civil society and global order, are to some extent overlapping and interactive because they are structures through which power is deployed. Why each of these three dimensions is so central to democratization is explained below.
The state is the embodiment and essence of political power (Mann 1993). The state is central to democratization in a number of crucial