An understanding of voting in sub-presidential elections is essential to a full understanding of voting behavior in the United States. This derives most obviously from the fact that most voting is performed at the state and local level, choosing government officials responsible for allocating and administering public resources on a vast scale. Despite the importance of these voting decisions, attention to them has been overshadowed by attention to the presidential contest, and to a lesser extent, to congressional elections. In addition to substantive findings, a contribution of this study is the development of a framework for analyzing voting behavior in sub-presidential elections. Important to the study of sub-presidential contests is the variation in the visibility of the several races and the fact that low visibility is a condition common to many state and local elections. An understanding of voting in such elections has not been provided by the much more extensive research on the highly visible presidential elections. Yet such races are the norm in American politics rather than the exception. The site of this investigation is the 1990 general election in Ohio which features a highly visible governor's race, moderately visible contests for four state executive offices, and two potentially low visibility nonpartisan races for seats on the Ohio Supreme Court. The variation in visibility is important to an understanding of the differences in the sources and kinds of information that are used by voters as information becomes less obtainable. Informational differences are expected to be critical to the understanding of the bases for preferences between candidates. Different models of voting behavior will be tested, including party line, split-ticket, incumbency preference, and coattail voting patterns. The sources and levels of information brought to bear on the voting decisions, and how they change as a function of the campaign, will be examined in relation to variation in the visibility of the different contests. Interviews of 1,200 Ohio citizens will be conducted shortly after Labor Day, respondents will be re-interviewed immediately prior to the election, and all pre-election respondents will be re-interviewed immediately after the election. Respondents will be asked a series of questions regarding evaluations of, attitudes toward, and preferences for the competing candidates. Knowledge of the candidates will be critical information as will the sources of this knowledge, such as mass media and the activities of parties and candidates in communicating with and enlisting the support of voters. A full battery of socio-demographic questions will be asked of each respondent in addition to questions on prior voting behavior, political affiliations and policy-related attitudes.
