Independent expenditures are designed to influence the preferences and behavior of the electorate. Like candidate expenditures, the money is generally spent trying to persuade voters to act a certain way on election day Because of the continuous nature of the spending, the sophisticated techniques of influence it buys, and the absence of accountability, there is good reason to believe these expenditures are important determinants of the vote. Expenditures for the incumbent and against the challenger should help the incumbent, and expenditures for the challenger and against the incumbent should help the challenger. Using individual-level survey data, we develop a model of Senate vote choice that considers these independent expenditures as both exogenous and endogenous to the process. We then estimate this model in the 1984, 1988, 1990, 1992, and 1996 elections. We find that independent expenditures can significantly affect vote choice, especially when modeled as endogenous to the process. These effects depend, however, on the electoral context, the type of independent spending, and the type of candidate. In general, our results seem roughly to conform to the conventionally accepted account of the 20-year history of independent expenditures in U.S. elections.
Since the late 1970s, several politically motivated groups have launched "independent" electoral campaigns in support of, or in opposition to, candidates running for federal office. These campaigns are fully financed and coordinated by the groups, and thus are exempt from federal campaign finance laws limiting the amount of money groups can contribute to political campaigns. The history of these political expenditures is curious: initially independent expenditures seemed to be a potent political weapon indeed, but the early successes were never repeated. Spending on independent campaigns, nonetheless, continued to increase during the 1980s and 1990s, although somewhat erratically. Also, the character of the groups that chose to launch campaigns changed-the fiercely ideological conservative groups that made up over half of all independent spending in the early 1980s have been replaced by the more pragmatic and bipartisan trade associations and single-issue PACs. Finally, independent expenditures that at first seemed to be the perfect tool for negative campaigning, became an outlet of groups' advocacy more than its expressed opposition.
Such an erratic history makes the nature of independent spending difficult to grasp. Nonetheless, campaign finance reformers see independent expenditures as a "loophole" in the system of campaign finance regulation that must be closed (Novak and Cobb 1987). Others believe that independent expenditures contribute much needed political expression to the election process, and are therefore beneficial (Buckley 1981). Purging the election process of independent expenditures makes little sense if we know nothing of how they affect the electoral process. We must, therefore, understand the consequences independent campaigns have on the choices voters make before we undertake either endorsing or reforming the system.
The purpose of this article is to clarify the relationship between independent spending and the behavior of voters. We hypothesize that independent expenditures significantly affect voter preferences. Several observations support the hypothesis. First, interest groups, and PACs specifically, have spent and continue to spend a great deal of money conducting independent campaigns. This continued dedication implies that some return is being perceived by those who seek to influence elections. It is possible that interest groups do not always see winning elections as the only return on independent campaign activity Several PAC officials have indicated that simply creating obstacles to a politically hostile incumbent's reelection attempts is sometimes the objective in their independent campaigns. Rather than expect to unseat some incumbents, PACs sometimes intend to "send a message" to other candidates that the PAC is willing to wage a large and expensive campaign against those who cross them (Cohen 1986). …