How Come the Gig Economy's Impact Was So Overestimated?

Analysts extrapolated from the convergence of technology and the slow recovery from the 2008-2010 recession. The full employment economy has demonstrated that, as Janis Joplin sang, 'freedom's just another word for nothing left to lose. JL


Josh Zumbrun reports in the Wall Street Journal:

The gig economy appeared swollen because the labor market earlier this decade was so weak for so long in the aftermath of the recession. Rather than heralding a permanent shift in the relationship of Americans to employers, a lot of gig-economy activity was odd jobs people took to make ends meet. As the economy returned to normal, they returned to more familiar work arrangements.
Two leading experts on the “gig economy” now say their estimates of its impact were too high, skewed by spotty data and the recession of a decade ago.
Alan Krueger of Princeton University and Lawrence Katz of Harvard sifted through new evidence to explain how, in a 2015 survey, they overestimated how people cobbling together a living from odd jobs, especially via apps like Uber, would upend traditional work arrangements.
Earlier this decade, many researchers and journalists fretted that the gig economy was taking over the way people work. When the Labor Department finally studied the question in detailed research released last summer, they concluded the gig economy had scarcely changed the U.S. labor market.

Messrs. Krueger and Katz now explain how the thesis itself was flawed, in a working paper to be released this week.
First, the gig economy appeared swollen largely because the labor market earlier this decade was so weak for so long in the aftermath of the recession. Rather than heralding a permanent shift in the relationship of Americans to employers, a lot of gig-economy activity was odd jobs that people took up to make ends meet. As the economy returned to normal, they returned to more familiar work arrangements.
Second, Messrs. Krueger and Katz conclude, the surveys used to measure alternative work arrangements remain riddled with flaws, and the Labor Department does a poor job of accounting for people with multiple jobs.
These explanations, however, were hard to spot because there was so little quality data about the gig economy’s size. Since 2005, the Labor Department had repeatedly sought, but been denied, funding for a survey that examined contingent and alternative workers, such as independent contractors, on-call workers, temporary-agency workers or workers provided by contract firms. Many jobs in the gig economy were a subset of these jobs.
By 2015, as venture capital was pouring into Uber with valuations suggesting the company was worth more than General Motors Co. or Ford Motor Co. , dozens if not hundreds of companies were trying to apply the model to other industries. Yet no good data had been compiled in a decade.
Messrs. Krueger and Katz attempted to fill the gap with their 2015 survey, replicating the old Labor Department survey. They found a significant jump in the share of workers who were in alternative work arrangements, as much as 5 percentage points, though they were among the first to conclude that the online aspect of this was mostly Uber.
Now, they say, even their earlier estimates were likely too high.
“After sifting through the new evidence,” Mr. Krueger said, “Larry Katz and I now conclude that there was a modest rise in the share of the workforce in nontraditional jobs over the last decade—probably on the order of one to 2 percentage points, instead of the five percentage point rise we originally reported.”
Comparisons between 2005 and 2015 had a hard time accounting for the difference in economic health. In 2005, wages and overall gross domestic product were growing around 3% a year—the housing bubble meant this growth would be fleeting—and about 63% of U.S. adults had jobs. The housing bust and recession that began in 2007 wrecked all of that, and by 2015, the economy had been on a slow mend. Wages and GDP growth were stagnant at 2%, and employment as a share of the population was about 4 percentage points lower.
This created a fertile environment for people to experiment with gig work. But it didn’t represent a permanent shift, as the Labor Department’s 2017 survey made clear.
Messrs. Krueger and Katz also conclude surveys of the topic have flaws. By studying individual survey responses, they found that when people answer questions about their spouses, they are much less likely to mention secondary jobs than when they answer questions about themselves.
Their own survey hadn’t asked additional questions about spouses, whereas the Labor Department does, leading to a significant discrepancy.
“These factors distorted our initial comparison, and highlight the importance of using consistent methods to measure labor force trends over time,” Mr. Katz said.