Money and votes reconsidered: congressional elections, 1972–1982

ConclusionThe evolution of congressional campaign finance over the past decade is characterized by steady growth in campaign spending and other notable changes in campaign finance practices. The main purpose of the research reported here was to determine whether these changes have altered the effects of campaign spending on election outcomes. This required reexamination of auxiliary questions concerning how sending effects operate and can be measured; it also required facing the limits of what analysis of aggregate campaign spending data can tell us about money in elections.Replication of OLS and 2SLS analyses of campaign spending effects produced solidly consistent results over the six congressional elections held between 1972 and 1982. Taken at face value, the evidence is overwhelming that the challenger's level of spending has a strong impact on the vote, whereas that of the incumbent has virtually no impact at all. But the evidence remains open to doubt on two grounds. One is the probable inadequacy of the 2SLS model (or, perhaps, any possible model), leaving the issue of simultaneity bias unsettled; if bias remains, the estimates exaggerate the electoral effects of the challenger's campaign expenditures. The issue cannot be resolved with more of the same data but rather requires data of a different sort—surveys taken over the course of a sample of campaigns would do the trick. Meanwhile, I think it would be foolish to make policy on the assumption that bias is, in fact, substantial; the greatest likelihood remains that restrictions on campaign money will have the general effect of hurting challengers (Jacobson, 1979).The other doubtful, albeit equally stable, finding is that incumbents do not gain votes by spending in campaigns. No incumbent seems to believe it, and there is at least circumstantial evidence (from, for example, the 1982 elections) that their skepticism is quite justified. If so, the problem lies in the limits of what aggregate data of the kind analyzed here can tell us. Few incumbents do not spend heavily when they are strongly challenged. If both candidates spend beyond the point needed to become thoroughly familiar to voters, then the substance of the campaigns, the contents of campaign messages, become the dominant factors. Incumbents may need to spend money in order to change the message as new circumstances arise, to frame new issues that would help them most (or damage them least) to win the battle of establishing the real issues. Like nonincumbents, sitting members may sometimes need to spend beyond a certain threshold to remain competitive; but nearly all of them do so when the necessity arises, so aggregate spending data are largely uninformative. How the money is spent, rather than how much, is what matters.The two most notable developments in congressional campaign finance have opposing implications. National party organizations, particularly the Republican, have assumed an increasingly important role in financing campaigns. Greater central control leads to a more efficient distribution of the party's collective campaign resources, which, among other things, promises to raise the overall level of electoral competition. It also leads to more coordinated campaigning, with greater emphasis on national themes and programs. Thus it should foster more cooperation and loyalty among fellow partisans in Congress. But PACs have also grown in numbers and in financial importance. Because most concentrate on gaining access to incumbents, they do little to make elections more competitive. Because they usually pursue narrowly defined interests, their influence fragments the Congress and weakens party coalitions. In this and other ways, parties and PACs are natural rivals; the most consequential decisions taken in the next round in campaign finance policy will be those which affect the relative importance of parties and PACs.See U.S. Congress (1984), for comments on this and other current campaign finance issues.