Academic journal article
By Larson, Tom; Ong, Paul M.
Journal of Economic Issues , Vol. 29, No. 4
In our original article [Larson and Ong 1994], we examined reasons for changes in involuntary part-time employment and did not directly address the broader topic of changes in the use of contingent labor, so it is certainly true that we did not make a strong case that there has been little change in the use of contingent workers. We feel, nevertheless, that our data and analysis of part-time voluntary and involuntary employment are inconsistent with common claims made regarding the use of contingent workers. We have additional reasons to be skeptical of the contingent labor thesis, which we discuss below. After squarely addressing the contingent labor thesis, we respond to specific points raised in the Comment.
A dramatic increase in the relative size of the contingent labor force could affect income equality [Harrison and Bluestone 1988], alter the full employment level of unemployment,(1) and/or make U.S. firms more competitive in the global economy [Belous 1989a]. Whether the contingent labor force is growing is therefore an issue of some importance and deserves to be further studied and better understood.
The concept of contingent labor is usually not well defined when discussed, but the thrust of the now commonplace argument is that full-time permanent workers are being replaced in the work force by part-time, temporary, or contracted workers. This replacement is seen as chiefly benefitting employers by reducing labor costs through cuts in benefits and wages. For workers, part-time and temporary employment could mean a decline in earnings with more job growth occurring in less secure jobs with low-wages and few benefits. For some, part-time employment could also mean an increase in family earnings as opportunities increase for those who do not want full-time work but wish to supplement family income.
Many commentators assume that a strong case has been made that the contingent labor force is growing. This assumption has become part of textbooks,(2) notwithstanding a weak empirical literature. A critical problem with assessing the size of and changes in the contingent labor force is finding a good yardstick for measuring the contingent labor force. The major problem with most claims is that contingent labor has not been directly measured, and our knowledge of the phenomenon remains, in fact, meager. Only recently has the Bureau of Labor Statistics added questions on the Current Population Survey that will allow identification of contingent workers. Unfortunately, contingent workers will not be regularly counted [Bureau of National Affairs 1994], so controversy regarding trends will continue.
The belief that there is widespread growth in the use of contingent labor is primarily based on observations of increased part-time employment and growth in employment among temporary agencies. Some argue that the self-employed and those employed by business services should also be counted as contingent workers [Belous 1989a]. Tilly  includes multiple-job holders when counting part-time workers. There are empirical problems with adding these categories together in order to estimate the number of contingent workers, and there are empirical problems in doing trend analysis for most of the categories. For example, it is impossible to analyze the trends for moonlighting because data are available only from special counts that do not permit calculation of an annual average (the last counts are for May 1980, 1985, and 1988).
Many reports on the contingent labor force cite Richard Belous, who has claimed that there has been a "dramatic" increase in the contingent labor force and argues that "many employers have altered their basic [human resource] system to reduce their core work force in favor of contingent workers" [Belous 1989b]. Central to his thesis is a purported growth in the use of temporary and part-time workers, the self-employed, and those working in the business service sector.
In Table 1, we report data from Belous [1989a] in columns 1 and 3, for temporary(a), part-time(a), business service, and the self-employed [Belous 1989a, Table 2. …