The Attribution of Government’s Responsibility in Latin America: A Comparative Study of Economic Voting across Institutional Contexts

MARÍA CELESTE RATTO

Seattle

Congreso; 2011 American Political Science Association (APSA) Annual Meeting; 2011

American Political Science Association (APSA)

Numerous economic reforms were implemented in many Latin American countries during the 1990s. The consequences of these reforms were: increase in unemployment, poverty and stagnation of the social budget. In face of these negative results, the emergence of strong opposition from the citizenship and the punishment at the polls of those politicians that had implemented such reforms was predicted. Oddly enough, many of the presidents who implemented the reforms were reelected. What are the factors that account for the paradoxical electoral behavior of the voters? This paper argues that institutional variables condition the process of accountability. As Powell and Whitten (1993) established, some political factors tends to affect the clarity in the process of responsibilities attributed to the government for economic outcomes. Political factors shape the electoral consequences of economic performance. I studied how different institutional designs would affect the attribution of government‟s responsibility for economic outcomes in Latin American countries during the 90‟s. I tested this through a comparative design divided in two stages. The first was based on testing alternative hypotheses found in the literature about the impact of institutional design on clarity of responsibility in 16 Latin American countries between 1996 and 2004. I split the sample by context (unified vs. divided government; bicameral opposition; % of representative seats of presidential party; effective number of parties (Laakso y Taagepera 1979); re-election according to the law; presidents who effectively ran for re-election; volatility vs. stability of voting; change vs. stability of electoral rules for president elections; change vs. stability of electoral rules for representatives; and change vs. stability of party system) and I measured and compare the impact of the economic vote across those institutional features. In the second stage, the research focused on the use of the multilevel statistical technique in order to compare the performance of several contextual variables over economic perceptions at the same time.