Read on Twitter
Accepted in Econometrica, Redistribution through Markets (with awesome coauthors, @skominers and Piotr Dworczak).
Link: https://stanford.io/3gJQxbT
We have multiple takes on how to present this, with no optimal one. Here's one take, based on Econ 101: (1/10)
Link: https://stanford.io/3gJQxbT
We have multiple takes on how to present this, with no optimal one. Here's one take, based on Econ 101: (1/10)
What the standard argument in favor of market-equilibrium pricing? It maximizes total surplus (or, equivalently total willingness to pay). Hence the famous picture. (2/10)
If you have "price control" you can give more surplus to sellers and less to buyers, but you'll necessarily have a deadweight loss. If you care about sellers more, it's sounds like a good idea, no? Well, an economist would respond: (3/10)
"Separate the problem of efficiency from redistribution: Set the market-clearing price, maximize total surplus. Then there's a way to compensate/pay sellers & buyers such that everyone is better off than what you can do with price control." Hence, the 2nd welfare theorem. (4/10)
What's wrong with this argument? Here is a big one: it's not Incentive-Compatible! If my value is my private info, & I know there will be future compensations coming, I'd behave differently in the market to begin with. 2nd welfare theorem doesn't survive with private info. (5/10)
Thus: You cannot separate the problem of efficiency from redistribution! If you wanna give more surplus to sellers (than their market-equilibrium surplus), you *have to* give up some allocative efficiency. We characterize the IC-constrained Pareto frontier. (6/10)
The main theorem shows that any redistributive goal (any point on the red curve) can be achieved with two simple tools: Price controls (rationing) & lump-sum transfers. (7/10)
Next, we take an applied approach: *if* a society puts more weight on giving 1$ to a poor person than to, for instance, Jeff Bezos, then we can take as input the wealth inequality and produce as output the optimal structure of price controls. (8/10)
A key insight here is: You don't want to have price control if inequality is low. Price controls become optimal precisely when there's substantial inequality, when redistributive value of markets outweigh the allocative inefficiency of price controls. (9/10)
It's an ongoing agenda, so happy to hear feedback/ideas/critics. (link: https://stanford.io/3gJQxbT )
I close by one of my favorite paragraphs of the paper! (10/10)
I close by one of my favorite paragraphs of the paper! (10/10)
