The House of Representatives approved an amendment from Rep. Scott Tipton (R-CO) to the FY2013 Energy and Water appropriations bill (H.R. 5325) on June 6, 2012, to prohibit federal agencies covered by the legislation from funding surveys in which money is included or provided for the benefit of the respondent. While Rep. Tipton was motivated by a specific controversial federal survey, he offered a bill later in 2012, "the Survey Savings Accountability Act" (H.R. 6569), which would have forbidden respondent incentives in all federally conducted or funded surveys.
Incentives are an important tool in the research toolkit
To achieve a representative sample of survey participants, some surveys must provide incentives that attract, retain, and compensate respondents for their time and effort. Incentives range from a small cash token of appreciation to reimbursement for child care or transportation necessary for participation.
Careful experimentation has shown that respondent incentives can improve a survey’s quality and efficiency. In particular, incentives can:
- Improve the response rate: Incentives have been found to significantly increase the proportion of the sample that is willing to participate, ensuring a large enough sample to draw generalizations.
- Improve response from hard-to-reach groups. Incentives can get the attention of individuals who would normally not participate in the survey, significantly improving the survey’s measurement quality. In particular, they can help improve representativeness in survey response from low-population rural areas -- essential to the inclusion of rural voices in federal decisions.
- Increase efficiency. Incentives reduce costly and time-consuming non-response follow-up, which can include dozens of call-backs and in-person visits, and may even reduce the study’s cost.
An estimated 25 percent of federal government surveys use respondent incentives, although the Office of Management and Budget (OMB) usually turns down requests to use incentives, and incentives are forbidden for use in the government’s customer satisfaction research. Incentives offered as part of a grant-funded research study are approved (or not) by various Institutional Review Boards (IRBs).
Long-running federal surveys use respondent incentives
- The National Health and Nutrition Examination Survey (NHANES) is unique in that it combines interviews and physical examinations. NHANES has used incentives since it began in the 1960’s.
- The Bureau of Labor Statistics’ National Longitudinal Surveys, running since the 1960’s.
- The National Survey on Drug Use and Health (NSDUH) provides national and state-level data on tobacco, alcohol, and illegal drug use, prescription drug abuse, and mental health in the U.S.
Rep. Tipton’s efforts would hinder survey and opinion research across all agencies under the legislation’s rubric and establish a dangerous precedent for all other federally conducted and supported research – research that helps the U.S. understand important policy issues and where we are headed. These surveys must be statistically valid and representative. When applied appropriately to certain kinds of surveys, incentives may significantly increase the accuracy, and thus the value, of federally conducted and sponsored surveys. Therefore, Congress must oppose any attempt to limit the appropriate use of respondent incentives in federally conducted and funded research.