This article presents an exploratory analysis of the flow of funds raised in American presidential nomination campaigns. Using detailed financial data for all the major party candidates in the 1988 presidential contest, we explicate and test two rival models of fund-raising: A "campaign-driven" model, which ties the fund-raising success to performance in the campaign itself, often in the form of momentum; and an "organization-driven" model, which ties it to the quality of candidates' political bases and fundraising efforts. We find much more support for the latter model. Most candidates' funds are raised before the start of the primary season; the patterns of fund-raising remain largely stable even during the primary season; and variations are largely a function of fund-raising efforts rather than poll standings, campaign events, or even primary/caucus performance. Though specific primaries and caucuses do sometimes affect fund-raising, the impact is typically modest and short-lived. These findings suggest that at least under current campaign conditions, candidates should give priority to building fund-raising organizations rather than relying on campaign performance to provide monetary momentum.
No one denies the importance of money for would-be presidential nominees. Without money, candidates typically are strapped for staff and facilities, ignored by the media, and unable to gain voter attention. Especially since the advent of a primary-based nomination process, presidential hopefuls must acquire adequate and timely funds for mass campaigning.
But there is no clearly established evidence on how this is done. We know a lot about total amounts raised, types of contributors, expenditures, and even the impact of money on nominations and elections (cf. Wilcox 1991; Corrado 1996). But we do not know exactly how money flows into presidential nomination campaigns, the timing involved, or the factors that explain the flow. Extant studies suggest two sharply contrasting possibilities. One, derived from the literature on primary voting and "momentum," holds that fund-raising is bound up with the relative performances of the competing candidates during the campaign itself, and so can be described as "campaign-driven" The other postulates that the flow of funds is determined not by progress of the campaign, but by candidates' pre-existing political bases and organization; it is therefore "organization-driven." Though neither of the approaches is particularly well-developed, each does implicitly incorporate different theories about campaign strategy, the nature of fund-raising, and donor behavior. In this study we seek to explicate these models more fully, match them to existing evidence, and test them with data from the 1988 presidential nomination contests. The results, which rather strongly confirm the organizational basis of fund-raising, have important implications for both the theory and practice of modern elections.
MODEL ONE: CAMPAIGN-DRIVEN FUND-RAISING
The idea that the pattern of candidate fund-raising stems from the progress of the campaign was first clearly expressed by John Aldrich, who pointed out that the most important campaign resources-national popularity, media attention, and money-can all be acquired by strong showings in highly publicized and contested primaries. A successful candidate may receive "ever-increasing flows of money" (1980a: 106), thereby enhancing his chances in future contests. Meanwhile, unsuccessful candidates get caught in a "vicious circle" of declining funds and declining success (1980b: 659). Aldrich thus views the acquisition of money and other resources as highly unstable, especially in multicandidate races (1980a 111-12). Both other political scientists and political practitioners have agreed; Larry Bartels cites Hamilton Jordan's observation that "early primary showings can have a profound and irrevocable effect on succeeding primaries and a campaign's ability to raise funds and recruit workers" (1988: 25). …