No it’s not a gig economy
Despite the voluble testimony of pundits and bar companions, the world of work is not one of Uber drivers and temp workers. In fact, the share of U.S. employment accounted for by contingent and “alternative” arrangements is lower now than it was in 2005 and 1995.
That testimony is derived from several original sources. For example, a much ballyhooed 2014 study commissioned by the Freelancers Union—which is not a materially disinterested party—reported that a third of workers are freelancers. The claim of a 2016 paper by Lawrence Katz and Alan Krueger that “all of the net employment growth in the U.S. economy from 2005 to 2015 appears to have occurred in alternative work arrangements” was widely quoted and quickly became folk wisdom. That paper was based on an online survey conducted by the RAND Corporation The survey was small—fewer than 4,000 respondents—and its sample wasn’t very representative of the overall population, a flaw the authors corrected through vigorous statistical handiwork.
Data released this morning by the Bureau of Labor Statistics should put an end to this chatter. According to a special edition of their Current Population Survey, a monthly poll of 60,000 households conducted jointly with the Census Bureau, just 3.8% of workers were classed as contingent in May 2017, meaning they don’t expect their job to last a year. That’s down from 4.1% in 2005 and 4.9% in 1995. (Reports from the years before 2017 are here.) Tighter definitions show smaller shares, but also down from earlier years. In 2017, 96.2% of workers were noncontingent, compared with 95.1% 22 years earlier.
The share of workers in “alternative” arrangements was 10.1%. Of those, 6.9% were independent contractors, 1.7% were on-call, and 1.5% were employed by either temp or contract firms. That means that 89.9% of the workforce has a “traditional” job, down 0.2 point from 1995.
There’s less of a racial pattern to contingency than one might guess: 3.7% of white workers don’t expect their jobs to last, compared to 4.0% of black workers; 4.9% of Asian, and 5.1% of “Hispanic/Latino.” All these shares are down from 1995. Nor is there a vast gender disparity: 3.9% of women, vs. 3.8% of men are contingent.
And not all independent contractors are freelancers hanging on by a thread: 39% are in managerial or professional occupations, slightly less than their share of the overall workforce. These would include self-employed doctors or consultants. Reflecting that, independent contractors are more likely to be white and male than nonwhite non-men. Other forms of alternative arrangements show surprisingly little variation by race and gender; nonwhites are more likely to be temp workers, but there’s no gender gap at all. Almost all demographic groups show little change from 1995, and most of those changes are downward.
Of course, not all contingent workers are consultants or contract programmers. Full-time contingent workers earn 77% as much as noncontingent workers; contingent part-timers earn 89% as much as noncontingent part-timers.. Almost three-quarters—73%—have some kind of employer-provided health insurance, compared to 84% of noncontingents. All in all, 55% of contingent workers would like a traditional job.
“Alternative” workers are better off. Full-time independent contractors make 96% as much as noncontingent full-timers; contract workers (heavily used in tech) make 22% more. Temp workers—0.9% of the workforce—are much worse off, however, making 41% less than the traditionally employed. About three-quarters of independent contractors have employer-provided health insurance, but only two-thirds of temp workers do.
None of this is to argue that the world of work is a delight. But we should be clear about what the problems are. Precarity isn’t the major problem in the American labor market. It’s that wages are stagnant or worse, benefits are eroding, and much labor is dull, alienating, pointless, and sometimes dangerous. Many people with normal, full-time jobs have a hard time making ends meet, and most households have little or no savings to fall back on in a crisis. Emphasizing precarity only makes workers feel even more powerless than they are.
While I rarely comment on other blogs, I feel it is necessary to do so to Doug’s current commentary on the BLS contingent-alternative survey just released.
I certainly agree with Doug that US workers not in what’s called ‘precarious’ employment are being exploited increasingly severely, but that fact is not a justification for arguing that addressing those in precarious employment is a kind of distraction from the former, as Doug seems to suggest.
Nor do I think that just because the latest May 2017 BLS supplement survey is not that different from the previously most recent 2005 survey that it shows contingent-alternative work–which is almost always accompanied by lower pay, benefits, and working conditions–is not a critical issue. If non-contingent labor is being screwed more with every passing year, then contingent is being even more screwed. If American workers are being increasingly exploited (meaning wages stagnating, benefits taken away, privatized, or cost thereof burden shifting, employment security even more tenuous,etc.) then workers in precarious jobs are super-exploited (wages even less, benefits virtually non-existent for many, etc.)
There are serious problems with the BLS supplement survey on contingency to which Doug refers. One should not simply take the BLS ‘at face value’. What’s behind that ‘appearance’ is important. That’s not to say there’s a conspiracy by government to cook the numbers to reduce the magnitude of the precarious jobs growth problem. It’s all in the definitions, assumptions (overt and hidden) that underlay the BLS report.
First of all, the gig economy is excluded by the BLS own admittance (see the Technical note). No Uber, Lyft, Taskrabbit, AirBNB, etc. They may add it later, but not in these numbers. So we’re talking about contingency and alternative work only. So what’s the definition of these terms, and is the BLS’s the best definition? Moreover, all jobs that are gig or contingent or alternative that are second jobs are excluded. But shouldn’t the BLS be estimating ‘jobs’ that are contingent-alternative, etc., and not primary employment only?
Contingency refers to a condition that is not permanent in some way. The BLS defines lack of permanency by referring to time–i.e. hours of work and conditions of employment a year or less. But why that definition? Shouldn’t contingency refer to the existence of a different wage structure, second tiered benefits provisioning, reduced legal rights and other working conditions exclusion, or whatever establishes a group of workers relationship to the employer as ‘second class’? Why just time, why not conditions?
Given the BLS’s actual assumptions and definitions, there are significant problems nonetheless. Here’s just a few:
First, BLS defines ‘temp’ workers as those employed by Temp Agencies. But there are hundreds of thousands, perhaps millions, who are hired direct by employers on a temp basis, not through agencies. The CPS has always ignored temps direct hired. Check out the auto industry where their numbers have been expanding for years.
What about public workers and higher ed teachers? I could not find any verification in the BLS study that they interviewed this sector? Many studies show that 70% of higher ed college teachers are now lecturers. (CHeck out the SEIU study). I suspect they aren’t adequately weighted in the BLS survey if at all. What about, as well, public home health workers, and the growing number of K-12 part timers, especially in charter schools.
And what about the company practice of hiring interns without pay for 3 to 6 months, then let them go and hire another cohort without pay. That’s a growing practice in tech. Aren’t they ‘super-contingent’? One could add the general practice in Tech of requiring skilled tech job candidates to solve a company problem, for which they aren’t paid, and then not hire them. Or the exploitation of young workers in so-called ‘coding academies’, where they do projects for companies in the hope of being hired, and then aren’t.
Another big problem with the BLS survey is it was conducted in May. That’s a big seasonality problem. Other studies. that Doug dismisses, were conducted in October-November. Obviously there would be far more ‘contingent’ workers in retail, wholesale, warehousing, etc. that would show up in November than in May. Remember, BLS findings are ‘statistics’, not raw data. They aren’t actual real numbers but estimates of real numbers (as is all BLS data). Seasonality issues are an important problem in the latest BLS survey.
And what about farm labor. They are certainly contingent. Many are undocumented and are not accurately surveyed (their numbers are plugged in based on assumptions about their numbers and employment). The same could be said for the huge underground economy in the US, now at least 12% of US GDP. Millions of inner city youth are not accurately weighted in CPS surveys in general. The CPS does a phone survey. That survey is biased toward workers who are not transient, who have a landline phone (and only most recently has the BLS been adding cell phones to that phone survey). Inner city youth and undocumented workers do not respond to government phone surveys, if they are even called upon in the first place. These are problems with the BLS-CPS general employment and wage surveys, which they ‘resolve’ by simply assuming an adjustment factor.
The BLS admits it excludes day labor. Does that mean also that the majority of longshore ‘B Men’, casual workers (who fit the BLS definition of contingent) are also not included? And why shouldn’t students working also be considered contingent? It fits the BLS definition. Why exclude that arbitrarily?
In short, there’s a lot of problems with the BLS survey, that in general results in a low balling of the magnitude and growth rate of contingent-alternative work. That low balling is baked into the definitions, assumptions, and methodologies it uses. (And of course the many important occupation categories it excludes). The truth is probably somewhere between the Princeton academics’ and freelancers’ union estimates, and the BLS study. But whatever the numbers, it makes no sense to say that precarious employment is not a growing problem in the US (and elsewhere in the advanced economies). Or that we should ignore it and focus on the ‘real problem’ with noncontingent work. They’re both a problem. We should not ignore the growing exploitation and destruction of noncontingent work; nor should we fall in line with government estimates of the precariate world by simply taking their (BLS) report at ‘face value’.
It’s no service to the US working classes, that have been beaten down in countless ways for more than three decades now, to say that the accelerating capitalist restructuring of labor markets creating gig, contingent, and alternative work (with less pay and benefits) is not a problem. The US government is minimizing the problem. Those who call themselves progressives should not join in.
Dr. Jack Rasmus
Misclassification of workers as independent contractors rather than employees has become rampant, as employers dodge their share of FICA taxes, unemployment insurance, liability insurance, and other expenses. Does this practice play into what the numbers are showing?