CASE STUDIES: LEGISLATIVE INSTITUTIONS
IN BRAZIL AND INDIA
Most institutional theories of corruption posit a specific causal link between national institutions and levels of corruption. In order to test their predictions regarding the corruption implications of these institutions, they conduct panel or cross-country tests at the macro level of institutions without testing the causal mechanism itself directly. Unfortunately this country-level macro-analysis cannot rule out the possibility that the predicted outcomes of corruption may be linked to institutions by a causal mechanism different from the one being proposed. As discussed in the introduction, this has been a common constraint in the study of corruption. Given the illicit nature of corrupt exchanges, scholars have been hampered in this effort by the severe challenges of collecting the data required to directly test various causal mechanisms. In this chapter, I discuss how the research design adopted in this book addresses this problem of micro-level empirical testing of the causal mechanism by adopting a strategy of comparative case analysis and studying two theoretically important cases—Brazil and India. I discuss the logic for selecting these specific cases and present relevant details on their legislative institutions and lawmaking process.
As discussed in chapter 2, table 2.2 summarizes how this research design meets the research goals of this study by combining the analysis of a large number of countries over time with in-depth study of two cases. Data from a time-series cross-section (TSCS) dataset of sixty-four developing democracies from 1984 to 2004 will be analyzed in chapter six to robustly test hypotheses H1b H2b, and H3b which state that each of the three party-focused legislative institutions lead to higher corruption levels. In order to address the challenge of testing the causal mechanism directly, I adopted the strategy of conducting a detailed case analysis of two selected cases. Fearon and Laitin (2008,757) argue that case studies are powerful tools for evaluating whether the causal arguments proposed to explain empirical regularities in large datasets are in fact plausible. They state that “one typically uses additional data about the beliefs, intentions, considerations, and reasoning of the people who made the choices that produced the