Maximizing Web Survey Response: Research from the Current Employment Statistics Survey

Nicholas Johnson

Abstract

The Current Employment Statistics (CES) program of the U.S. Bureau of Labor Statistics (BLS) is a monthly multi-modal establishment survey that measures employment, hours and earnings for nonfarm payrolls. Sixteen percent of CES survey reports are completed online through the BLS web system; however, these respondents represent 36 percent of total reported employment for CES. This underscores the need to maintain high response rates for these web reports, which can be a challenge for any self-reporting method. This paper reviews some of the steps taken to maximize response from CES web respondents and then analyzes new empirical research performed using CES web survey response data. Utilizing 40,000 CES web respondents for this research, we will attempt to determine how best to prompt respondents for monthly data entry. Consideration will be given to the timing of e-mail prompts sent to respondents, time zone effects, firm size and industry classification. This review has implications not only for prompting web respondents but we posit that by measuring the response on a self-reported method, we are discovering the revealed preferences of respondents for other data collection modes as well.