|Trends and Challenges for Work in the 21st Century|
Flexible Staffing Arrangements
A Report on Temporary Help, On-Call, Direct-Hire Temporary, Leased, Contract Company, and Independent Contractor Employment in the United States
Susan N. Houseman
The term "contingent work" was coined by Audrey Freedman (1985, p. 35) to describe "conditional and transitory employment arrangements as initiated by a need for labor--usually because a company has an increased demand for a particular service or a product or technology, at a particular place, at a particular time." Since then the term has been used in various ways. Barker and Christensen (1998, p. 1) note that the term contingent employment is "generally thought to include those jobs that are done on temporary, self-employed contract, or involuntary part-time bases." The types of employment arrangements covered in this report would certainly fall under the rubric of contingent employment, as commonly understood in much of the academic and policy literature.
However, beginning with the February 1995 Supplement to the CPS, BLS sought to collect data on the number of "contingent" workers, and, by implication, provide a more precise definition for the term. In an article published prior to the design of the February 1995 Supplement, BLS economists Anne Polivka and Thomas Nardone sharply criticized the common usage of the term contingent worker and discussed how it should be defined and measured (Polivka and Nardone, 1989). In that article, they correctly pointed out that some had expanded the notion of contingent employment to include all part-time and self-employed workers and yet these arrangements often involve long-term and stable employment.4 They proposed that, based on Freedman's original concept of contingent employment, two key criteria be used in classifying a worker as contingent: 1) a low degree of job security, in particular the amount of job security embodied in the employment arrangement and 2) variability in hours worked. However, in designing the Contingent and Alternative Worker Supplement, BLS only incorporated the first criterion into its definition of contingent workers. In addition, the classification is based on workers' perceptions of job security, rather than on the actual contractual nature of the employment relationship. Specifically, in the BLS data, only workers who do not expect their jobs to last for economic reasons are classified as contingent workers.5
Using Freedman's original concept of contingent worker, it is clear that while some researchers have over counted the number of contingent workers, the BLS definition undercounts them. For instance, on-call workers are intrinsically contingent, because their hours of work vary. In fact, Polivka and Nardone (1989) cited on-call workers like substitute teachers as examples of workers who should be classified as contingent. However, this aspect of contingency--hours variability--is not measured in the BLS concept. In 1997, only 28 percent of on-call workers were classified as contingent under the broadest BLS definition of contingency. The problem is not limited to on-call workers. For instance, temporary help agency workers are also intrinsically contingent. While agency temporaries may establish a long term relationship with the staffing agency, their employment, hours, and pay will vary with the availability of assignments. However, only 57 percent of agency temporaries are classified as contingent in the 1997 BLS survey. The surprisingly low percentage of agency temporaries who are classified as contingent in the BLS survey suggests a problem with the way the concept of temporary is being measured in the survey or with respondents' understanding of the questions being asked.
To avoid confusion over the various definitions used for contingent work, I use the term flexible staffing arrangements when referring collectively to the set of employment arrangements covered in this report. Generally, however, I discuss the various employment arrangements separately. As will be shown below, the characteristics of workers in flexible staffing arrangements and the quality of the jobs, as measured by average wages, benefits, and job security, vary considerably by arrangement. Additionally, when considering policy issues, the contractual nature of the employment relationship--not workers' perceptions about their job security--is what is relevant for employment statutes and regulations.