The 1988 survey was conducted by the authors. We drew a random sample of contributions to each of the eleven major presidential nomination candidates in the 1988 election. These candidates were: Babbitt, Bush, Dole, du Pont, Dukakis, Gephardt, Gore, Jackson, Kemp, Robertson, and Simon. The sample was drawn from the lists of contributors that candidates filed with the Federal Election Commission (FEC), and was based on pre-July 1988 contributions that aggregated to over $200 to a single candidate.

We received completed surveys from 1,246 respondents, representing a response rate of 37%. Our estimate of response rate was a conservative one. We did eliminate known invalid addresses (of contributors who had moved without forwarding addresses, or whose forwarding addresses had expired, or whose addresses were incorrectly recorded by the campaign committees to begin with), together with deceased or hospitalized contributors. However, we are confident that the 37% is a lower bound for the response rate because we certainly have not identified all of the inappropriate addresses or deceased or ill respondents.

Although this is not a high response rate, it is not unusual for this type of survey or sample. Two preceding surveys of contributors, one of 1972 presidential contributors and another of 1978 House contributors, produced response rates of approximately 50%. (See Brown, Hedges, and Powell, 1980a, for the 1972 survey and Powell, 1989, for the 1978 survey.) However, responses to all types of surveys (mail, telephone, and in-person) have dropped sharply in recent years, and our results reflect that. Guth and Greene (1991), surveying a similar population, were able to get a 40% response rate with five mailings and an elimination of institutional addresses.