Journal of Democracy

The goal of political scientists who study Latin America is not just to understand the origins and nature of the region’s new democracies but also to determine ways in which they can be strengthened and deepened. This is much more than a matter of individual preference or disciplinary ideology. It is widely accepted that liberal democracy comes closer than any other political order to satisfying the political demands, especially the demands for self-determination and self-realization, that people in modern societies typically assert. Consequently, liberal democracy should be more stable, more consonant with economic growth and development (owing to its reliance on private initiative), and less likely to result in aggressive interstate relations than any other form of government. It is also widely accepted, however, that Joseph Schumpeter’s reduction of liberal democracy to fair and competitive elections is conceptually inadequate and that a viable liberal democracy has identifiable correlates in the structure and practices of the economy and society.¹

All Latin American democracies succeeded authoritarian forms of government without the accompaniment of major changes in socio-economic structures; at most there was some reordering of relations between certain social strata and the state as a result of “marketizing” reforms. Hence, these countries must contend with strata and interest groups—the military, elements of the business sector, parts of the political or administrative elite—that supported a previous authoritarian regime and still wield enough power to make democratic governance difficult or impossible, if they choose to do so. Moreover, in some [End Page 138] instances the military retained enough authority during the transition to impose undemocratic features on the new regime (as in Chile, Brazil, and Peru), while in others, the price of liberalization was a “pact” placing the core interests and values of elite strata at least temporarily beyond the bounds of democratic political contestation.² This history has alerted us to the error of treating every successor to authoritarianism as a full-fledged liberal democracy. Lately, the discipline has taken up the notion of “democratic consolidation” as a way of avoiding that error. Defined comprehensively to include the economic and social “correlates of democracy” to which I alluded above, “democratic consolidation” is the state of political affairs obtaining when the transition from authoritarian rule has been completed. This idea has received its most thorough elaboration in a recent book by Juan Linz and Alfred Stepan.³

Wisely, Linz and Stepan avoid any suggestion that liberal democracies, once “consolidated,” are impervious to overthrow or decay (or, conversely, to improvement), since liberal-democratic political orders everywhere are in constant flux, owing to the rapid pace of economic, social, and technological change in modern societies. Nevertheless, I remain unconvinced that they have satisfactorily addressed the concern expressed in the opening sentence of this essay. For if the perfection of existing (perhaps merely nominal) liberal democracies is our goal, and if such perfection entails the refinement of a complex web of institutions, relationships, and social practices, we cannot limit ourselves to a counsel of “full speed ahead on all fronts at once.” The interest groups, political activists, and aid donors who look to political scientists for guidance expect specific advice regarding points and modes of intervention, not just general encouragement. What is more, some elements of this web are related to others in a nonlinear or even an inverse fashion—e.g., strengthening civil society may, in the short run, produce a deterioration of equally important state capabilities⁴—while others, such as a penchant for “civic” behavior and a willingness to compromise, may not be susceptible to any conscious intervention. On balance, Linz and Stepan are far better at determining which countries rank as “consolidated” democracies than at explaining how they got that way,⁵ especially if the explanation we seek is to be grounded in general theoretical considerations applicable to cases besides the ones they discuss. For these reasons I fear that the idea of “consolidation” threatens to become a sort of disciplinary “seal of approval,” even though this is not Linz and Stepan’s intent.⁶

In this essay, therefore, I propose that we abandon the classification...