Combining Alternative Perspectives: Internal Trajectories versus External Influences as Explanations of Latin American Politics in the 1940s

Abstract

This article analyzes labor politics and regime change in Latin America in the 1940s, paying central attention to major shifts in political coalitions that include or exclude labor, episodes of mobilization and demobilization, and patterns of regime opening and closing. The analysis considers two different perspectives for explaining changes in domestic politics. The first concerns the impact of international conjunctures, specifically factors surrounding World War II, the international campaign against fascism, and the beginnings of the Cold War. The second focuses on internal trajectories of change set into motion by the different ways countries experienced a major period of reform and state building earlier in this century. The study thus juxtaposes two perspectives: the analysis of historical conjunctures and a path dependent model of change.' The effort to juxtapose these alternative perspectives is motivated in part by a concern with the problem of dealing with multiple explanations in comparative analysis. Causation in the social world is complex, and it is a particularly difficult issue in the more qualitative, sociohistorical tradition of research. Statistical methods explicitly recognize the problem of multicausality and offer a number of techniques for analyzing and distinguishing rival versus complementary, or partial, causation. However, qualitative approaches in the social sciences have had a harder time coming to grips with these issues. Often such approaches present an overdetermined analysis, in which many plausible factors are seen as contributing to a particular outcome; these analyses offer no basis for eliminating some factors and advancing a particular hypothesis. On the other hand, those qualitative approaches that systematically attempt to advance a particular causal claim may acknowledge other explanations as competing theories, but they typically lack a satisfactory way of considering the possibility of complementarity in which more than one explanatory perspective help to account for the outcomes under study. This tendency has inhibited cumulative theory building in the subfields where qualitative and case study methods predominate. Small-N comparison is often seen as a way to bring rigor to the causal claims of case studies.2 Yet the usual approach of systematically comparing cases through the method of agreement or the method of difference does not take us very far with the problem of multicausality and the consideration of complementary causation, as opposed to rival claims. The purpose of the methods of agreement and difference is precisely to allow the analyst to eliminate what are seen as rival hypotheses, rather than to incorporate them as complementary causes. The handling of complementary causes is further complicated by the fact that small-N