Health and Retirement Benefits: Data from Two BLS Surveys
Herz, Diane E., Meisenheimer, Joseph R., II, Weinstein, Harriet G., Monthly Labor Review
Both the household-based Current Population Survey and the establishment-based Employee Benefits Survey have strengths and limitations with respect to collecting information on health and retirement benefits: demographic information is best obtained from household surveys; details of benefit plans are best collected from establishments
Employee benefits are an important aspect f job quality. In assessing the quality of different types of jobs, workers, employers, and researchers often consider benefits along with other characteristics of jobs, such as pay, job security, job safety, and the type of work involved.(1) Many employers are concerned about the cost of benefits, which compose 28 percent of compensation costs for employers in the private sector and State and local governments.(2)Public policymakers also frequently focus on employee benefits. For example, many observers have expressed concern in recent years about the number of Americans who lack health insurance. In response, policymakers have debated whether universal health coverage should be a national goal. Central to that debate are the role employer-provided health insurance plays in the current health care system and what role it might play in any proposed new system. Employer-provided retirement plans also have been the subject of public policy discussions. As the baby-boom generation--he huge cohort of Americans born between 1946 and 1964--approaches retirement age, concern has arisen about whether Social Security and private pension plans can withstand the strain of providing retirement income to so many people.(3)
Clearly, having accurate information on employee benefits is important for workers, employers, and public policymakers.(4) Two BLS surveys provide estimates of participation in employee benefits plans: the Current Population Survey (CPS) and the Employee Benefits Survey (EBS). The is a monthly survey of 50,000 households from which information is obtained on employment, unemployment, demographics, earnings, and more. The CPS is jointly conducted by the Bureau of Labor Statistics and the Bureau of the Census. The EBS obtains data from establishments on the number of participants in a variety of employee benefits plans and the detailed provisions of those plans. The EBS is being incorporated into the National Compensation Survey, which, when fully integrated, will provide measures of occupational earnings, trends in compensation costs, and participation in, and details of, benefit plans.(5)
This article compares information that the CPS and EBS provide on two of the most important categories of benefits: health and retirement plans. According to the CPS, 66 percent of full-time workers in the private sector participated in a health plan provided by their employer in 1995. The EBS indicates that 71 percent of full-time private-sector workers participated in an employer-provided health plan. The gap between the two surveys is greater in regard to participation in retirement plans: the CPS indicates that 49 percent of full-time workers in the private sector participated in an employer-provided retirement plan in 1995; the comparable figure from the EBS is 60 percent.
The material that follows is intended as a guide for researchers, public policymakers, and others to understand the strengths and limitations of CPS and EBS data on employee benefits. Among the topics examined are differences in estimates derived from the two surveys and possible reasons for inconsistencies between them. The types of information that each survey provides also are described.
Data on prevalence of benefits
Although the CPS is a monthly survey, it does not include questions each month on employee benefits. Rather, supplementary questions on benefits have appeared periodically in the CPS since the early 1970s. CPS supplementary surveys on employer-provided benefits were conducted in April 1972, in May of 1979, 1983, and 1988, and in April 1993. …