Effects of Incentives in the U.S. Consumer Expenditures Quarterly Survey

David E. McGrath


Response rates to the Consumer Expenditures Quarterly Survey have fallen from about 86 percent in 1990 to 76 percent by 2004. To combat the falling rates, the Bureau of Labor Statistics introduced an incentives experiment beginning in November, 2005. The goal was to increase response rates by offering respondents an unconditional, pre-paid monetary incentive. To achieve this goal, we mailed debit-card incentives along with the survey's advance letter, prior to contacting the potential survey respondent. The experimental design contrasts a control group receiving $0 with groups that receive either $20 or $40 debit cards. In this paper, we show the design of the incentive experiment, discuss some field issues that arose during implementation, and provide preliminary results that focus on the effects of the incentives on response rates.