June 24, 2018
Dean Baker
The Hankyoreh, June 24, 2018
Earlier this month, the Bureau of Labor Statistics (BLS) released its survey of non-standard work. This release was long awaited both because of a long gap with the prior survey and the view that gig economy jobs were coming to dominate the labor market.
The last survey had been done in 2005. The BLS had been fielding the survey every three years, but due to budget cuts, it had cancelled the last three surveys. With the growth of Uber, Task Rabbit, other Internet-based platforms for employment, many people expected that this survey would show a large increase in this sort of short-term contract employment.
That didn’t happen. The data from BLS actually showed a small decline in the percentage of workers who reported that contract employment was their primary form of work, dropping from 7.4 percent in 2005 to 6.9 percent in the new survey. There is enough error in these numbers that we shouldn’t assume there was an actual decline in the share of the workforce doing contract work, but clearly there is no explosion in this type of employment.
In fact, the data seem to indicate there is even some fall in the share of the workforce who have secondary jobs, such as driving an Uber to raise extra money. The percentage of workers reporting that they hold multiple jobs fell from over 6.0 percent two decades ago to less than 5.0 percent today. Clearly more people are driving Ubers, since Uber didn’t exist two decades ago, but it seems that Uber driving is largely substituting for other forms of secondary employment, not hugely expanding it.
The pundits who were convinced that everyone was going to do gig employment clearly got it wrong. The famous New York Times columnist Thomas Friedman has long been mocked for presenting his conversations with cab drivers as expert insights into countries that he has briefly visited. It seems that many pundits commenting on the economy seem to have jumped to the conclusion that their Uber driver was the future of work.
While the gig economy may not be the big labor market story that many believed, there are important changes in the labor market. Perhaps most notably we are seeing a “fissuring” of the work place, as large companies contract out an ever larger portion of their employment. This process has the effect of driving down pay, since the companies doing the contract work are generally smaller and pay less than the parent company.
My colleague at the Center for Economic and Policy Research, Eileen Appelbaum has documented this process in the health care sector, where hospitals have gone from contracting out custodial services and cafeteria work, to contracting out emergency rooms and dialysis facilities. Brandeis University Dean David Weil has documented this trend in the larger economy.
This process of contracting out an ever larger set of services has been yet another way in which wages have been depressed. This process of breaking up companies as a mechanism to depress wages is a much more important labor market development than the gig economy. Now that we have good data putting the gig economy into perspective, perhaps policy debates can focus on bigger problems.
In the same vein, perhaps we will be able to divert attention from the fixation with the idea that job-killing robots are going to leave everyone unemployed. This fear flies in direct contradiction to the economic data. The story of job-killing robots is one of rapid productivity growth in which robots are now doing work that might have taken many hours of human labor.
However, our productivity data show the exact opposite. In the last dozen years in the United States productivity growth has slowed to a crawl, averaging just over 1.0 percent annually. By contrast, productivity growth averaged roughly 3.0 percent in the long Golden Age from 1947 to 1973 and then again from 1995 to 2005.
The slowdown in productivity growth since 2005 was largely unexpected and still not well-explained. Nonetheless it has persisted for too long to be a blip in the data and has occurred pretty much across the world among the wealthy countries. The productivity slowdown is also projected by the I.MF., the OECD, and other forecasters to persist into the indefinite future.
The concerns in a world of slow productivity growth are directly opposite to the concerns in a world where the robots are taking all the jobs. In the job-killing robot world, we are looking for things for people to do and for ways to get income to displaced workers. In the slow productivity growth world, we are worried about budget deficits and how we having enough resources to meet important social goals.
There are reasons why we may not need to be as pessimistic about the future as the projections of slow productivity growth might imply, but at least these projections are grounded in the real world. Those hyping the job-killing robots story are talking about science fiction, not the economy that we see around us.
The new survey from BLS has hopefully killed most of the gig economy myths. Maybe we can now look to kill the job-killing robot myth as well.
