Originally presented at the November 1998 COPAFS conference, this paper summarizes the research literature on the measurement and abatement of unit nonresponse to establishment surveys, reviews the establishment survey programs at member agencies, gives a general overview of response rate calculation issues, analyzes the trends in establishment nonresponse in 12 federal government surveys, and discusses methods for reducing and adjusting for nonresponse. It also summarizes the current research efforts related to establishment nonresponse conducted at member agencies.

This paper documents the work of the IGEN, which was charged with examining issues associated with unit nonresponse in Federal establishment surveys, sharing and disseminating methodological research findings and survey design practices related to the measurement and abatement of nonresponse, and fostering future collaborative research on establishment nonresponse within the federal statistical community. This paper describes the three research programs IGEN has proposed for adoption by agencies. The paper also notes other efforts at collaboration, and lists difficulties in promoting such interagency collaboration.

This paper reviews the similarities and differences among the response rates calculated across a number of establishment surveys conducted by U.S. statistical agencies, and discusses the extent to which there is standardization in the calculation and publication of such rates and the reasons why this is so. It also explores the possibilities, advantages, and disadvantages of fostering greater coordination and standardization across IGEN-member agencies in this regard.

This paper summarizes agency information on response rates and research into nonresponse that IGEN has collected to date. The paper present trends in response rates for selected surveys and provides examples of how agencies calculate and present response rates. It also discusses various definitions of respondents and survey eligible units which are used to derive the numbers included in response rate calculations. The intention of this paper is to bring together information about unit response from across a wide spectrum of Federal agencies and establishment surveys. We hope that this will facilitate previously unavailable comparisons across surveys and establishment populations as well as promote the exchange of research ideas (both to IGEN members and others) as the IGEN member agencies work to increase response rates in establishment surveys.