From the Wall Street Journal's The Experts column:
Mark Muro (@markmuro1) is a senior fellow at 
the Brookings Institution and the policy director of the Metropolitan 
Policy Program there.
Does the so-called gig economy of app-based freelancing for platforms
 like Uber or TaskRabbit complement or “cannibalize” more conventional 
payroll work?  Given the sketchiness of the data available, it’s been 
hard to tell.
However, last month, the economist Ian Hathaway and I were able to piece together a plausible look at trends
 in two slivers of the much larger gig economy—the rides and rooms 
industries anchored by Uber and Airbnb, respectively—by looking at data 
from the Census Bureau’s “nonemployer firms” series, which tracks the activities of small freelance businesses that employ no workers.
In doing that, a picture emerges that is by no means definitive but 
that suggests—to my eyes at least—that the spreading gig economy (at 
least in the case of ride-sharing) is, in fact, substituting for some 
payroll employment, or at least depressing its growth.
To be sure, a first reading of the data lends credence to the view held by enthusiasts of the “sharing” economy
 that Uber and Airbnb are serving unmet consumer demand or stimulating 
new demand—and so are creating new work opportunities that complement 
those in existing taxi or hotel companies. You can see that in our data,
 which show that both nationally and in most of the 50 largest U.S. 
cities, payroll employment has actually increased somewhat in those 
industries during the years 2010-2014, even despite the influx of 
nonemployer contractors working for Uber and Airbnb. 
However, payroll growth in the two industries looks surprisingly weak
 by our count.  And I believe that a closer look at the metro-by-metro 
data suggests that platform freelancing was already substituting for 
some payroll employment and cannibalizing it in the 2012-2014 period. 
Payroll employment, after all, had declined in 15 metros by 2014 in 
ride-sharing and 10 in room-sharing, despite a modest economic recovery.
  Especially on the ride-sharing side, where Uber and Lyft offered 
marked quality advantages over many incumbent cab companies, payroll 
employment growth in incumbent firms remained anemic in most markets or 
else turned negative while platform-based freelancing (measured by 
nonemployer firms) took off. In San Jose, nonemployer-firm gig 
employment spiked 145% but payroll employment slumped by 31% as taxi and
 limo companies shed 250 payroll jobs.  In Sacramento gig employment 
jumped 92% but payroll employment declined by 22%. And in Pittsburgh, 
freelancing surged 85% (by 560 drivers) but payroll employment slumped 
5%.
All in all, these early developments raise the possibility that the 
online freelance marketplaces may well gain workers at the expense of 
competing payroll businesses in some industries, particularly where 
incumbents are struggling in weaker markets or fail to respond with 
better service. 
All of this is important because the rise of online temping, 
freelancing and independent contracting has huge implications for the 
circumstances of workers and families in cities....MORE
HT: FT Alphaville's 
Further Reading post.