Millions of workers do not have standard work arrangements—permanent jobs with a traditional employer-employee relationship. Rather, they are in temporary, contract, or other forms of non-standard employment arrangements in which they may not receive employer-provided retirement and health benefits, or have safeguards such as job-protected leave under the Family Medical Leave Act, even if they have a traditional employer-employee relationship. These non-standard arrangements are sometimes referred to as “contingent” work. To collect information about contingent workers, the Department of Labor’s Bureau of Labor Statistics (BLS) has previously supplemented its monthly Current Population Survey (CPS) with a survey on contingent work, known as the Contingent Work Supplement (CWS). While the CWS is a comprehensive source of information on contingent workers, BLS has not conducted this supplement since 2005.
In the aftermath of the recent recession more workers may have become contingent workers with potentially limited access to work-provided health insurance and retirement benefits, as well as coverage under key workforce protection laws. In light of these developments we were asked to examine issues related to the contingent workforce. This report examines what is known about (1) the size of the contingent workforce, (2) the characteristics and employment experiences of contingent versus standard workers, and (3) any differences in earnings, benefits, and measures of poverty between contingent and standard workers.
To assess the size of the contingent workforce, we analyzed population counts of contingent workers identified in various national survey data sources, such as the CWS, CPS, the General Social Survey (GSS), and the Survey of Income and Program Participation (SIPP). Using these national data sources, we analyzed the data to compare population counts both over time and based on various definitions, as applicable. We also analyzed and compared distributions of various self-reported worker and job characteristics, such as demographics and family income, and job security, benefits, and safety. We conducted regression analysis using CPS data, controlling for various external factors, to determine how various measures of earnings and retirement plan participation compared between contingent and other workers. We also compared the distributions of health insurance coverage and measures of poverty (e.g. family income levels) between contingent and other workers.
These data sources were available for varying timeframes over the last two decades and identified types of contingent workers or workers in alternative work arrangements, based on various definitions. While the CWS has been a comprehensive source of information about contingent workers, it has not been administered in 10 years (since 2005). Other surveys offer additional insight about this workforce, but may be less statistically robust or collect less detailed information about the many alternative employment arrangements researchers have suggested could be part of the contingent workforce. For example, some surveys have smaller samples or ask less detailed questions about why workers hold contingent jobs.