To win election and reelection, presidents, senators, and representatives work hard to interact with constituents, often through personal contact and visits (Shaw 1999a; Fenno 1996; Jacobson 2001; Mayhew 1974; Cohen, Krassa, and Hamman 1991). Many voters, however, may condition their support on more than contact with an elected official. These voters also want that politician to enforce or change government policies. Exerting meaningful influence over government policy is often difficult for one person, even the president, because such influence commonly requires collective effort often by politicians in the same party (Aldrich and Rohde 2005; Cox and McCubbins 1993).
Observers have documented extensively how elected officials from the same party work together to change government policy in Washington (Bond and Fleisher 2000; Edwards and Wood 1999; Kernell 1997; Neustadt 1960). Shared goals of policy influence may lead these politicians to extend their coordination from the policy-making arena to the electoral stage. We know less about this area of coordination. Party leaders may encourage their candidates to campaign on similar issues and to "stay on message" (Herrnson 2000; Jacobs and Shapiro 2000). Congressional party leaders often coordinate fundraising efforts, redistributing resources to vulnerable incumbents and strong challengers (Herrnson 2000, 2005). As the most powerful elected official with the broadest constituency, the president is a national leader in elections. The president's policies may link his party to prominent national trends, thereby extending his coattails down the ticket to other candidates from his party (Brady 1988; Fiorina 2005). The president himself can also work directly with his party's candidates and thereby shape electoral outcomes. In congressional races, the president helps raise money for both individual candidates and the national party committees (Jacobson, Kernell, and Lazarus 2004). This article focuses on another area of the president's direct involvement: visits to the home constituencies of his party's Senate candidates.
Knowledge of the motivations behind these visits and their effects can help us better understand the nature and extent of presidential power. For example, in the 2002 midterm elections, President George W. Bush campaigned vigorously for Republican candidates. He spent 241 days in 43 different states during the 22 months between his first inauguration and Election Day. The election produced historic results: the GOP regained control of the Senate, and the party's six-seat gain in the House marked the first time the Republican party gained House seats in a midterm election while controlling the presidency (Clymer and Rosenbaum 2002).
Bush may have chosen states to visit with the goal of helping Republicans win control of Congress. If his visits actually helped Republican candidates, Bush could convincingly claim credit for the Republican victory that November. The news media's widespread acceptance of this interpretation boosted Bush's subsequent influence with Congress and the American public. (1) However, the election events and outcomes may have other explanations. One possibility is that Bush chose his campaign visits to help his own chances of reelection in 2004. Each visit may have only reinforced the effects of other influences on each race's outcome, and Republican congressional candidates might have won without Bush. These answers would suggest that the link between Bush's efforts and the Republicans' November victory is more coincidental. His credit claiming would be less believable, and his increased influence more fabricated than real.
This article investigates these explanations of presidential visits, using the midterm Senate elections between 1982 and 2002. The first section of the article discusses how the president's conflicting incentives affect his interaction with Congress, particularly in the context of Senate elections. …