It may or may not be bad. But before we talk about it, we need to understand that many shades of gig work exist, and they might have different regulatory needs (13/13)

My concern is that our lack of precise language around the nuances of gig work will mean that people will cite this paper (on basis of the title and the abstract) as evidence that all regulation for gig workers is bad (12/)

Of course, surplus can still exist. But it is likely to be much lower if your rates are set by an app that can pretty precisely estimate your reservation wages and is somewhat free from the fairness norms around wage discrimination that would exist in a traditional job (11/)

Uber/Doordash etc are different though. The paper describes one source of surplus as buyers' choice being based on worker characteristics. This "rent" does not exist for Uber drivers - the app just matches you with someone (10/)

What is the issue then - as I said above, most calls for regulation or employee status for gig workers are not talking about these kind of platforms. If the rates are set by the workers themselves, they are basically entrepreneurs just using a platform for finding consumers (9/)

So, what does the paper show - in a market with sellers posting bids (Walrasian auctioneer, anyone?) and buyers having preferences over seller characteristics, there is some surplus that is reduced by taxes. Pretty non-controversial 101 economics - cool to see empirically! (8/)

You would expect such surplus to exist. You would also expect a tax to cause some kind of surplus loss - again, the authors show this with a counterfactual 10 percent tax. Workers are worse off after the regulation, even if all the tax revenue were redistributed (7/)

Coming back to the paper - they don't mention the platform, but it is likely something like UpWork - there are buyers and jobs and bids. The paper builds an interesting model to estimate buyer and consumer surplus from these transactions. And there is substantial surplus (6/)

Now let's consider the case of something like Uber/Doordash. Here, the platform is more powerful - they both decide the rates and match buyers to workers. When people are talking about regulation, it is mostly this they have in mind (but it all gets lumped under "gig work") (5/)

Next, you could also consider something like MTurk - buyers post jobs and a rate. Workers choose whether to work at that rate or not. The platform is again an intermediary only (4/)

Consider a platform like Upwork - buyers post jobs, workers look at these jobs and post job-specific bids, then the buyer decides the match. The platform is just an intermediary. Something like TaskRabbit is not much different - only the bids are more general (3/)

This is a problem with the concept of the "gig economy" - how we define it. Is any freelance work gig work? Let's say we only want to focus on app-based platforms - what most people mean. But that hides variation, both in terms of price-setting and the matching process (2/)