The considerable flexibility ingrained in the U.S. labor market has produced a variety of employment arrangements. The Bureau of Labor Statistics of the U.S. Department of Labor has completed the first special survey on workers in contingent jobs, that is, jobs which are structured to last only a limited period of time. Also collected in the survey was information on the number of workers in several alternative employment arrangements, such as those working for temporary help agencies or as independent contractors or "on call." The survey was conducted as a supplement to the February 1995 Current Population Survey (CPS), a monthly sample survey of about 60,000 households conducted by the Bureau of the Census for the Bureau of Labor Statistics. This analysis examines the initial findings from that survey.

Initial results from the survey show that, in February 1995, between 2.7 and 6.0 million workers—a range of 2.2 to 4.9 percent of total employment—were in contingent jobs. This range spans three alternative estimates which were developed to assess the impact of different assumptions about which factors constitute contingent employment. The narrowest estimate includes only wage and salary workers who had been in their jobs for one year or less and expected their jobs to last for an additional year or less. The middle estimate added the self-employed and independent contractors who were in a similar situation. In the third and broadest estimate, the limitation on how long workers had held their jobs and expected to remain in them was dropped for wage and salary workers; thus, this estimate includes almost any worker who believed his or her job was temporary or not expected to continue (see Figure 18).

The February survey also showed that 8.3 million workers (6.7 percent of the total employed) said they were independent contractors, 2.0 million (1.7 percent) worked "on call," 1.2 million (1.0 percent) worked for temporary help agencies, and 652,000 (0.5 percent) worked for contract firms that provided services to one customer at that customer's worksite. (see Figure 24) . In this analysis, contingent employment is defined separately from these four types of employment arrangements, although an individual's employment arrangement could be both "contingent" and fall into one of the alternative employment categories.

Contingent Employment

The major goal of the February supplement was to provide information on the extent to which workers were in jobs that were structured to be of limited duration. Specifically, contingent workers are being defined here as those individuals who do not perceive themselves as having an explicit or implicit contract for ongoing employment. Prior to the February supplement, data on this characteristic of jobs were lacking. The term contingent work has been used, however, to refer to a variety of work arrangements. These include part-time work, self-employment, employment in the business services industry, and, in fact, almost any work arrangement that might be considered to differ from the commonly perceived norm of a full-time wage and salary job.

BLS Definition

Contingent workers are those who do not have an implicit or explicit contract for ongoing employment. Persons who do not expect to continue in their jobs for personal reasons such as retirement or returning to school are not considered contingent workers, provided that they would have the option of continuing in the job were it not for these personal reasons. Figure 18 provides the three overall estimates developed from the survey. These three estimates ranged from 2.2 percent of total employed at the narrowest level that covers only wage and salary workers, to 2.8 percent when the self-employed were added in, and finally to 4.9 percent of the total when time restrictions were relaxed. In terms of numbers, the contingent work force was fairly sizable, ranging upward to 6 million persons.

Major Characteristics of Contingent Workers

Under all three estimates, contingent workers were more than twice as likely as noncontingent workers (those who are not contingent even under the broadest estimate) to be young, that is, 16 to 24 years of age (see Figure 19). Contingent workers were slightly more likely than noncontingent workers to be women and black. About half of all contingent workers were women, compared to 46 percent of the noncontingent. The proportion of contingent workers who were black was about 14 percent, compared with approximately 10 percent of noncontingent workers. Although part-time workers—persons who usually work less than 35 hours a week—made up a disproportionately large share of contingent workers, the vast majority of part-time workers (about 9 in 10) was not contingent. Among 16 to 24 year-olds, contingent workers were more likely than noncontingent workers to be enrolled in school (see Figure 20). Among young people not enrolled in school, a larger proportion of contingent than noncontingent workers had less than a high school diploma; this pattern was also true for persons ages 25 to 64. Contingent workers were concentrated in the professional; service; administrative support; and operator, fabricator, and laborer occupations (see Figure 21). The services industry alone accounted for more than half of the contingent total but about a third of noncontingent workers. The construction industry also accounted for a relatively large share of contingent workers. This concentration notwithstanding, the proportion of workers within the services industry who were contingent ranged from 3.4 to 7.5 percent. Similarly, only 4.5 to 8.4 percent of construction workers were contingent. The proportion of contingent workers who had health insurance from any source ranged from 57 to 65 percent, depending on the estimate chosen. This was 17 to 25 percentage points lower than the proportion of noncontingent workers with health insurance. Moreover, contingent workers with insurance were substantially less likely to receive it from their employers (See Figure 22). The majority of contingent workers preferred to have permanent rather than temporary jobs; only one-third preferred the contingent arrangement (see Figure 23). Among nonstudents, the proportion of contingent workers who were satisfied with their arrangement dropped by as much as 9 percentage points.

Alternative Employment Arrangements

Another major goal of the February 1995 supplement was to measure the number and characteristics of workers in alternative work arrangements. Workers in the four alternative employment arrangements examined were: Independent contractors, on-call workers, workers paid by temporary help agencies, and workers whose services are provided through contract firms to only one customer at that customer's worksite. The four groups varied a great deal in terms of their demographic and other characteristics. Furthermore, these worker groups—particularly independent contractors and on-call workers—had sharply different characteristics from workers in more "traditional" arrangements.

Characteristics of Workers with Alternative Arrangements

Data on the characteristics of workers with alternative work arrangements are described in the following paragraphs and in Figures 24 through 29. Independent contractors were considerably more likely than workers in traditional arrangements to be men, white, and at least 25 years old; they also were more likely to be out of school and have at least a bachelor's degree. They were somewhat more likely than traditional workers to work part-time and to hold managerial, professional, sales, or precision production jobs. In terms of industries, they were more likely to work in construction, agriculture, and services, and less likely to work in manufacturing or wholesale and retail trade. The demographics of on-call workers (including day laborers) were similar to those of workers in traditional arrangements, but on-call workers were slightly younger. In terms of job characteristics, they were more than three times as likely to be in the construction industry and also more likely to be in the services industry. On-call workers were more likely than traditional workers to work part-time and to be in professional; service; and operator, fabricator, and laborer occupations. Workers paid by temporary help agencies were more likely than workers in traditional arrangements to be women, young, and black and were slightly more likely to be employed part time. They were heavily concentrated in administrative support and operator, fabricator, and laborer occupations. Workers provided by contract firms were disproportionately male. The largest proportion of contract workers was employed in the service industry, although substantial proportions worked in manufacturing and transportation and public utilities. Contract workers were more likely to be in professional specialty or service occupations than were traditional workers.

If I Had My Druthers...

Most on-call workers and workers paid by temporary help agencies would have preferred to be in traditional work arrangements. In stark contrast, more than 4 out of 5 independent contractors preferred their current work arrangements (see Figure 28). As noted above, not all workers in alternative work arrangements met the definition of being contingent and not all contingent workers were in alternative work arrangements. Under the broadest estimate (estimate 3) of contingency, the majority of employees of temporary help agencies were in contingent jobs and a substantial proportion of on-call workers were contingent; however, only about 20 percent of contract workers and 4 percent of independent contractors were contingent (see Figure 29). When asked, 80 percent of those paid by temporary help agencies said that their jobs were temporary or could not last as long as they wished. However, 18 percent of temporary help agency workers had been working for the customer to whom they were assigned for more than a year, implying that temporary help agencies act as intermediaries for some relatively long-term employment relationships.

Implications for Washington

Because these data are for a single point in time, and only for the nation, one must speculate as to their local implications. Though Washington State has a higher than average share of part-time workers, it has a lower than average share of workers in personnel supply services firms. As a result, while the share of contingent workers in Washington may be higher than the national norm, the percent of workers with alternative working arrangements may be measurably lower. Growth rates in the business services sector—where personnel supply services are found—has been well above average in Washington. While the current growth rates locally have surpassed the national average, as of late, the lower starting point, and ensuing small relative share, has yet to be overcome. Single copies of Contingent and Alternative Employment Arrangements, Report 900, are available from BLS, (202) 606-6378 or (202) 606-6373.