I have a new report out today on our annual survey of the freelance economy. This year we are able to look at how freelancers are faring under COVID, and we learn a lot about the diversity & flexibility of freelancing. upwork.com/i/freelance-fo…
To start, it’s useful to characterize what has happened to the level of freelancing overall. What we found was that after covid a lot of people started freelancing for the first time. But we also had a lot of people pause freelancing.
One story that has been pushed is that the rush of new freelancers has crowded out previous freelancers, but looking at the data don’t think this is largely the case. New freelancers and paused freelancers are very different. Here are the differences:
New freelancers are largely:
More full time
More remote
Younger
More educated
In harder hit occupations
More likely to be male, urban, and caregivers
In contrast, paused freelancers are:
Freelancing on the side
Less remote
Less educated
In less hard hit occupations
For example, the top occupations for new freelancers are computers/mathematics and finance/business operations, for paused freelancers it’s education and construction. New freelancers in programming aren’t crowding out freelancers in ed.
Decline of side-gigs and rise of full-time freelancers is consistent with other data. In the CPS, self-employment share of employed is up while multiple jobholding is down. Non-employer business registrations are also up.
Another factor driving the demand for some freelancers is that they have always been disproportionately remote, and there is obviously a major need for remote work now. Even before COVID, one in four freelancers was entirely remote.
However, we can also show that is was not *just* labor demand that increased, but labor supply also. This is clear in Upwork data, where freelancer registrations actually picked up before client registrations.
Uncertainty also affects supply & demand. When businesses are uncertain and have to respond to quickly changing conditions, freelancers help them be agile. For freelancers, lots of clients means diversifying when many are at risk of closing
Finally, flexibility is very valuable. For clients, they need to rapidly respond to business challenges. On Upwork, we see demand for customer services and a sustained increase in demand for ecommerce development and web and mobile design
• • •
Missing some Tweet in this thread? You can try to
force a refresh
It's time to talk about labor market utilization indicators.
Domash and Summers wrote that U3 was the better measure of labor market slack than prime age employment rate, and thus wage growth was going to remain high bc labor markets were very tight...
They wrote "the unemployment rate is more important than the prime-age employment ratio in predicting wage inflation". And they argued labor supply is unlikely to come back. Thus there was reason to be pessimistic wage growth would be reduced except by demand reduction alone...
The alternative view was the prime epop was the best indicator of labor market slack in equilibrium, and that the wage pressure we experienced was at least in some substantial portion a departure from that equilibrium. Contra Summers, this suggested room for labor supply to help
I honestly don't think its very mysterious why the public isnt' thrilled with the economy yet. Pandemic economy was a chaotic high inflation mess & people haven't moved on yet. Please tell me the economic rule that says as soon as the problem is fading people must be thrilled
Its very hard to find a house, and prices haven't actually fallen except for a handful of things. The world feels more expensive. And by the way, people can tell that stimulus played a role. Just give it time and hope we keep making further progress, which we should.
Normal people know that there was a ton of money pushed out in the pandemic that made people less willing to work for some time, and this lack of workers created goods & input shortages, problems that have taken a long time to dig out of even as people have come back to work.
@JHWeissmann The two key components are the sacrifice ratio and NAIRU. Now the sacrifice ratio tells you how much extra U you need to reduce inflation by 1pp. Since U hasn't gone up, ignoring month to month volatility, this can't be the incorrect assumption.
Folks like @balajis seem to imagine that to make methodological criticisms of the BLS necessitates the absurd claim of fraud, and that there is some deference issue causing blindness. I wrote a 100+ page dissertation criticizing BLS methods. scholarshare.temple.edu/handle/20.500.…
Institutions can be corrupt but it’s helpful to have basic institutional awareness before lobbing such accusations. If you think the BLS is corrupt by the fed, for example, you literally don’t understand how they function
If you think the BLS is well described by “the western financial system”, you don’t understand. Responsible people do homework before slinging accusations like this
So the big gap between payroll and household is that self employment declined by 369k. Wage and salary employment in the household survey went up 129k.
Since Feb '23, self-employment unincorporated is down 607k. So we're seeing a lot of return to payrolls from self-employment. Now thats a small sample, so caveat emptor, but if so that is less aggregate labor impact than the payrolls would otherwise suggest.
Will have to wait until Fred updates, but some of this has been offset by growth in incorporated self-employment.
Incorporated SE is up compared to pre-pandemic, while unincorporated is back down to pre-pandemic levels.
@Noahpinion I do agree that in some circles the importance of regulation was down-weighted by empirical results from RegData showing little relationship between regulation and outcomes. But I think this too needs updating...
@Noahpinion Goldschlag and Tabarrok didn't find any relationship between reg data and dynamism, and I think that did have some influence on people's priors here academic.oup.com/economicpolicy…