Male and female telephone interviewers are compared on both administrative efficiency and data quality, using data from 24 replications of an attitudinal survey on personal and national economic prospects. The 40 male interviewers used over the two-year period are found to exhibit higher turnover rates and, because of that, lower response rates and higher training costs than the 80 female interviewers. However, there are no real differences on the total per minute interview costs by gender, in missing data rates, or on response distributions for factual questions. There does appear to be a systematic tendency for male interviewers to obtain more optimistic reports from respondents regarding their economic outlook. Multivariate models are constructed that attempt to explain these results and speculations are offered about causes of the impact of interviewer gender on response formation.