Spurred by the meteoric rise of platform employment companies (most notably Lyft, Uber, Handy, and Amazon Turk), employment outside of regular hourly or salaried work has been one of the hottest topics in labor in recent years. The conversations have centered on what has been variously called “gig,” “temp,” “contract,” and “freelance” work, among other terms, with little popular consensus about differences in these terms, beyond the fact the jobs are unstable, comparatively low-paid, and rising fast in number. As pertains their prevalence, until recently, these conversations have rested on scattered data from diverse sources, given that there had been no definitive government data on the question since 2005. Last week, however, new data from the Bureau of Labor Statistics Contingent and Alternative Employment Arrangements Supplement (CWS) revealed, contrary to expectations and conventional wisdom, that the government’s estimated percent of the workforce that’s engaged in contingent and alternative work arrangements has not changed in the last decade.

Given how long we’ve been without a comprehensive survey, the new Bureau of Labor Statistics (BLS) data warrants close scrutiny. Does the work types’ general consistency in percentage warrant a recalibration of our fears about the future of the quality of American jobs? If so, how? This commentary will look closely at the survey results, whose data indicate that while the picture may be brighter in some respects than previously thought, the quality of work for many Americans is still far from secure.

Overall Prevalence

The BLS survey separately measures work that is temporary in nature—”contingent”—and work that is structured differently than traditional wage and salary work. Figure 1 shows the latter—alternative work arrangements including but not limited to the gig economy. The BLS data finds an increase in these various arrangements in the early 2000s and a leveling out from 2005 to 2017. In total, workers across the spectrum of “gig,” “temp,” “freelance,” and “contract” work account for just over 10 percent of the total U.S. workforce.

Despite the flat numbers, there are major job quality issues for the millions of Americans who worked in these arrangements, especially for low-wage and minority workers who are concentrated in nonstandard work even if they prefer permanent or regularly scheduled jobs.

Figure 1
Source: Annette Berhardt, “Making Sense of the New Government Data on Contingent Work.

Data Limitations

This BLS survey only accounts for respondents’ main job (the job which one works the most hours) and not those who have a secondary gig or freelance work. Anyone who picks up a job in an alternative arrangement on off-days, nights, and weekends to make up for non-livable income in a first job is not counted. Due to a focus on respondents’ “main job,” the BLS identifies over 10 million workers in alternative arrangements, whereas other surveys, for example from Upwork, a global freelancing platform, have identified upwards of 57 million workers.

Moreover, this data is self-reported and there are discrepancies between this data and tax records from a U.S. Treasury report, which shows a small but steady increase in workers reporting self-employment income. Certainly some workers paid as self-employed workers through an IRS form 1099 should be treated legally as workers, but this discrepancy is another reason that the CWS is not the final word.

In sum, The CWS data don’t tell the whole story for workers in contingent and alternative arrangements.

Where Are the Benefits?

One significant point that the data masks is that workers outside of traditional arrangements mostly miss out on benefits such as health care and retirement plans, putting a greater reliance on the individual or social programs.

Rates of employer-provided health insurance were 41 percent for contract workers, 28 percent for on-call workers, and 13 percent for temporary help agency workers—compared to 53 percent for those in traditional health insurance benefits. However, overall, health insurance coverage has increased among non-traditional workers since 2005 due to the Affordable Care Act. The percentage of temporary agency workers with insurance skyrocketed from 39 to 67 percent from 2005 to 2017, and workers in other alternative arrangements had noticeably increased. These health insurance gains are fragile: new Medicaid work requirements could exclude those nonstandard workers who cannot secure full-time work and federal litigation might eliminate the right to health insurance for individuals with pre-existing conditions.

Retirement savings have not fared as well. Across the board, non-traditional workers are less likely than traditional workers to be eligible for employer-provided retirement plan. For example, only 18 percent of contingent workers were currently enrolled in a retirement plan in May 2017, compared to 43.4 percent of traditional workers. Without conventional savings options or robust government supported programs beyond social security, they further miss out on payroll deductions, auto enrollment, and employer matching; other research has found that only one in three nonstandard workers has a retirement benefit.

Temp Work

Many job quality issues are magnified in temporary help work, including lack of health care and savings, opportunities for advancement, and other fringe benefits. Moreover, temporary help workers earn the lowest ($521 per week) of any of the categories of workers in alternative arrangements

Take, manufacturing for example, which communities looked to historically in many cases as a ladder to the middle class. Last week’s data confirm that employers in the manufacturing sector have particularly escalated the use of temporary help agencies to find workers. The share of temp agency workers in production and material occupations is up from 30 percent in 2005 to 39 percent in 2017, and as shown in Figure 2, the percent of total manufacturing workers employed by temporary help agencies has increased from 2.4 percent to 3.5 percent.

Figure 2

However, the use of temporary workers is undermining this potential. That’s because too often temporary jobs are just that: temporary. A 2010 study by economists David Autor and Susan Houseman analyzing the impact of the Detroit Work First program showed no benefit in wages or employment from temporary help work. While temporary help agencies offer value through job placement, temporary jobs do not often lead to full-time employment or advancement opportunities.

Racial Divide

Racial unemployment and wage differentials felt in the traditional labor market may have spillover effects and drive minority workers to participate in nontraditional work, and this rise leaves more minority workers than white workers susceptible to the poor quality of work described above. CWS data represented in Figures 3 and 4 show that minority representation in contingent and alternative work has increased from 2005 to 2017, by 12 percent and 41 percent respectively for black workers, and 10 percent and 46 percent respectively for Hispanic workers, while representation among whites has declined.

While a growing percentage of workers in alternative arrangements prefer this type of work to traditional work, this is not the case of contingent jobs (only 32.8 percent prefer this type of work). Those working in contingent jobs earn only 77 percent as those in regular jobs. Figure 2 shows larger growth among contingent workers among blacks and Latinos and a decline among whites. In this context, the rise in contingent jobs among black and Latino workers is particularly troubling, and symptom of the polarization of the job market between good jobs and bad.

Figure 4 shows that the racial divide among workers in alternative arrangements is even larger, with both black and Latino workers having a much larger increase. This is an area that deserves close future study to determine whether the increasing concentration of workers of color in these jobs exacerbates inequality.

Figure 3

Figure 4


The growth in contingent and alternative work among black and Latino workers speaks to why the gig economy has generated so much hype in recent years. Alternative and contingent work are a potent symbol of twenty-first century work having less economic security than did work in generations past.

Based on the recent Federal Reserve Survey of Household Economics and Decisionmaking (SHED), 3 in 10 workers’ income varies monthly, 1 in 4 skip necessary medical care because its too costly, and 1 in 4 had a family income less than $25,000.

But the technical definition in last week’s data is just one part of this phenomenon. The increasing use of subcontracting and outsourcing by large firms to poorly paid subcontract workers (described by Professor David Weill as the fissuring of work), and misclassification of workers as independent contractors, are more economically important ways that employers to avoid the responsibility of providing higher pay or better conditions for workers. What is associated with gig work—instability in pay and schedule, lack of health and retirement benefits, and low pay—are ultimately characteristic of our labor market at large.