This article originally posted at Harvard Business Review
A number of reports in recent weeks have stressed that employment effects of the so-called gig economy—contract workers on software platforms such as Uber and AirBnB—have been overstated. At minimum, these reports indicate, any increase in gig economy employment hasn’t shown up in the aggregate statistics—at least not yet anyway.
But my analysis tells a different story, showing that the impacts can in fact be seen if you look more deeply at the data and in the right places.
By examining key industrial segments (rides and rooms) and geographies (early-adopters in San Francisco), I found a substantial rise in collaborative economy “gigs” between 2009 (when uptake began) and 2013 (the latest year data are available). I also found that more gigs haven’t been accompanied by fewer workers on payroll, at least so far.
The data show that despite the attention the contractor status of these workers has received—including by leading Presidential hopefuls—software platform companies are not the first to take this approach. In fact, the San Francisco taxi and limousine industry shows a clear shift of activity from workers on employer payrolls to those of independent contractors for at least a decade before the first ride was hailed using Uber.
The gig economy doesn’t show up in aggregate data for several reasons. It’s relatively new, early uptake has been concentrated in some cities and not others, and most of the data we have on self-employment only counts someone’s “primary job.”
To see if other analysts, who have looked at broad national aggregates, have missed something in the details, I analyzed just the passenger ground transit (e.g. Uber) and traveler lodging (e.g. AirBnB) industries. Since the data are current through 2013, I only looked at San Francisco—where uptake started the earliest, so results can reasonably be seen in that timeframe. This is a critical point, because even in this leading city, services such as UberX didn’t really take off until 2013.
Like the others, I also analyzed publicly available data produced by the Census Bureau on so-called “nonemployer firms,” or businesses that earn at least $1,000 per year in gross receipts but don’t employ anyone. The substantial majority of these “firms”—86% across all industries to be precise, 91% for those industries studied here—are self-employed, unincorporated sole-proprietors.
To benchmark the non-employer firms, I matched Census Bureau data on employment in San Francisco for the same industries.
Three major findings stand out. First, there are clear growth surges in nonemployer firms in each of the two industries associated with passenger ground transit between 2010 (when Uber launched in San Francisco) and 2013, and in the two industries linked with traveler accommodation from 2009 (the year AirBnB opened). These increases amount to thousands of workers earning a living in some way (either supplemental or in full) because of these platforms. Because of the way income activity is reported, this is almost certainly an undercount—representing a lower bound on activity. This suggests an increase in the number of contractors employed in these industries.
Second, we do not see declines in payroll employment in the same industries during this period. Instead, we actually see increases in all four—particularly in the passenger ground transit sectors.
Caution is needed when interpreting these figures. Nonetheless, such strong employment growth contradicts the idea that Uber drivers are pushing incumbent firms out of business. Instead, it lends credence to the story behind Uber’s founding, and the experience of San Franciscans at the time (myself included)—Uber and other peer-to-peer ride services like Lyft were meeting unmet consumer demand in a city with a massive shortage of taxi services.
Finally, these figures suggest that the trend toward contractors over payrolled employees in the taxi and limousine industry was going on in San Francisco long before Uber’s arrival. If the employment security of drivers is truly the issue, perhaps the debate has to expand beyond concern over platforms like Uber.
Though more work needs to be done, this simple case study shows that gig economy employment is showing up in the official statistics when one knows where to look. Previous reports that analyzed a wider range of data at higher levels of industry and geographic aggregation, though also important, miss these finer points.
Data for 2014, which will be released one year from now, will likely show an acceleration of these trends in San Francisco, and an extension of them to other cities. Future analysis will need to look more closely at the net effects of the gig economy on employment—and on wages—as well as the impact that a broader and potentially harder-to-measure range of platform services, such as Etsy and Thumbtack, are having on the economy and the workforce.
It’s important to remember that these platforms are very new and that good data usually comes with a time lag. We can’t yet analyze the true impact of these platforms. But, as this case study has shown, the effects might be bigger than has been previously thought.