An official website of the United States government
This paper reviews a set of randomized experiments designed to identify causes of nonresponse in establishment surveys. It draws heavily on the experience of the authors with the U.S. Current Employment Statistics (CES) program, a monthly survey of about 400,000 employers. The purpose of the survey is to provide monthly estimates of jobs created in the economy and of changes in payroll levels. The experience of the authors is limited to experimental research to increase the cooperation rate of sample employers. The paper is organized by stages of the survey process. For some of these stages the discussion begins with some theoretical constructs that appear to be relevant to the stage.