Causal Attribution and Economic Voting in American Congressional Elections

Article excerpt

This article examines the ways in which political sophistication conditions economic voting in U.S. congressional elections. At the congressional level, evidence of economic voting has been generally mixed and sometimes contradictory. In our view, much of the inconsistency in existing studies may result from a tendency to overlook significant heterogeneity in voter decisionmaking. Specifically, we argue that an individuals ability to attribute responsibility for economic outcomes to congressional actors is a function of political sophistication. According to our theory, less sophisticated voters tend to focus their attributions of responsibility on the President (the most obvious national political figure), ignoring the influence of Congress on the national economy. More sophisticated individuals, by contrast, are capable of more diffuse attributions. Thus, to the extent that conventional economic voting occurs in congressional elections, it should be confined to the more sophisticated portion of the electorate.

Over the last several decades, scholars studying American elections have established conclusively that economic conditions powerfully shape citizen vote choice for a variety of offices at both the state and national levels (Kramer 1971; Tufte 1975; Hibbs, Rivers, and Vasilatos 1982; Chappell and Keech 1988). There is much less consensus, however, on the question of exactly how individual voters evaluate the economic information that they receive and relate it to the political world. Do people make political choices based on their own personal economic well-being (the "pocketbook" hypothesis-Kramer 1971, 1983), the economic health of the nation at large (the "sociotropic" hypothesis-Kinder and Kiewiet 1979, 1981), or a combination of both (Markus 1988)? This central question has dominated empirical research on individual-level economic voting. We would assert that there can be no single satisfactory answer for the entire electorate. Rather, the nature of an individual's economic voting depends centrally on how he or she attributes causal responsibility for changes in both personal and national economic circumstances, a process that, we argue, is strongly conditioned by one's level of political sophistication.

In this work, we explicitly incorporate political sophistication into existing models of economic voting. Here, we develop our theory of helerogeneous attribution, first proposed elsewhere (Gomez and Wilson 2001), in the domain of congressional elections. Our theory, explained in greater detail below, maintains that an individual's ability to make causal linkages is strongly conditioned by his or her level of political sophistication-less sophisticated voters will tend to make simple, proximal attributions, while sophisticated voters are more willing and able to make distal ones. In a study of economic voting in presidential elections (Gomez and Wilson 2001), our theory leads to a set of counter-intuitive hypotheses that challenge the conventional wisdom on sophistication and economic voting: namely, that politically sophisticated individuals are the ones most likely to vote their pocketbook, while less sophisticated voters tend to rely exclusively on sociotropic assessments. We find strong support for these hypotheses in two different presidential elections, suggesting that sophistication is indeed key to a proper understanding of heterogeneity in economic voting.

In many ways, however, congressional elections provide a richer, more interesting arena in which to explore the dynamics of economic attribution. In the congressional case, the task of assigning responsibility is rather complex. As Mayhew (1974: 53) asks, "Where can credit be found? If there were only one congressman rather than 535, the answer would in principle be simple enough-but there are 535." In attributing responsibility for economic outcomes, voters might plausibly focus on the entire Congress (thus crediting or blaming all incumbents, regardless of party), or alternatively, they could adopt a more partisan approach, giving credit or blame to all candidates of the party that has controlled Congress, or to all candidates of the President's party. …