Assessing the Impact of Campaign Spending in Governors' Races

Abstract

In this article, I explore the influence of candidate expenditures on gubernatorial election outcomes (a heretofore largely ignored topic). I find that spending by gubernatorial contestants dramatically influences the outcomes of these races, matching or even out-performing previously studied predictors of gubernatorial election outcomes. I then deal with two issues. First, I explore possible explanations of candidate expenditures in these contests and find that, among other things, out-party candidates are especially sensitive to spending limits, raising the normative concern that these limits insulate in-party candidates from competition. Second, I explore the possible endogeneity of candidate spending using a two-stage least squares, instrumental variables model. I find that once we account for this endogeneity, the relatively vulnerable position of out-\party candidates becomes more acute. These findings expand our understanding of governors' races to include spending effects and emphasize the importance of aggregate-level factors in these contests. According to Jesse Unruh, money is the mother's of politics. Nowhere is this statement more true than in the electoral process. Money has the potential to dramatically condition the flow of information in campaigns, and clearly has the potential to influence election outcomes. Not surprisingly, then, students of electoral behavior have focused on the impact of spending in a variety of electoral contexts, particularly congressional elections (Glantz, Abramowitz, and Burkhart 1976; Jacobson 1976, 1978, 1980, 1987, 1990; Copeland and Patterson 1978; Silberman 1978; Abramowitz and Segal 1986; Abramowitz 1988; Green and Krasno 1988; Thomas 1989; Green and Krasno 1990; Abramowitz 1991; Kenny and McBurnett 1992; Kenny and McBurnett 1994; Krasno, Green, and Cowden 1994; Epstein and Zemsky 1995; Box-Steffensmeier 1996; Gerber 1998), but also state legislative contests (Breaux and Gierzynski 1991; Gierzynski and Breaux 1991; Hogan 1997; Thompson and Moncrief 1998; Hogan 1999), and statewide initiatives and referenda (Owens and Wade 1986; Hadwiger 1992; Bowler and Donovan 1998). Virtually overlooked in these studies of money in various electoral arenas are elections for state governor.1 This oversight is unfortunate for a variety of reasons. To begin, given the importance attributed to campaign resources by politicians, pundits, as well as political scientists, one would hope that our scholarly assessment of this mother's milk comes from as broad an understanding of American elections as possible.2 For those scholars interested more specifically in gubernatorial elections, past inattention to campaign expenditures may amount to a serious mis-specification of previous models of election outcomes in these races. Elections for public office are, as pointed out by Key (1966), interactions between the public and potential elected officials. Nevertheless, for the most part, studies of gubernatorial elections have focused primarily on the behavior of voters in these races, to the exclusion of candidate actions and behavior (Turett 1971; Piereson 1975; Bibby 1983; Cohen 1983; Eismeier 1983; Kenney 1983; Kenney and Rice 1984; Peltzman 1987; Chubb 1988; Tompkins 1988; Simon 1989; Stein 1990; Simon, Ostrom, and Marra 1991; Kone and Winters 1993; Atkeson and Partin 1995; Leyden and Borrelli 1995; Partin 1995; Svoboda 1995; Atkeson and Partin 1998; Carsey and Wright 1998; Lowry, Alt, and Ferree 1998).3 While much work certainly remains to be done in understanding individual-level voting behavior in gubernatorial races, a largely unfilled field of inquiry deals with the actions and activity of candidates in these contests. In comparison with the congressional elections literature, for example, we have little understanding of strategic considerations by gubernatorial candidates (Jacobson and Kernell 1981; Bianco 1984; Born 1986; Wilcox 1987; Krasno and Green 1988; Fowler and McClure 1989; Jacobson 1989; Squire 1991; Kazee 1994; Krasno 1994; Lublin 1994; Epstein and Zemsky 1995; McCurley and Mondak 1995; Box-Steffensmeier 1996; Adams and Squire 1997). …