Joshua Gans
joshgans.bsky.social
Joshua Gans
@joshgans.bsky.social
Professor at University of Toronto
Anyhow, here is the paper. It doesn't involve any fancy techniques. Indeed, it could have easily been written in the 1950s. www.nber.org/papers/w34669
Price Discrimination with Costless Resale
Founded in 1920, the NBER is a private, non-profit, non-partisan organization dedicated to conducting economic research and to disseminating research findings among academics, public policy makers, an...
www.nber.org
January 12, 2026 at 1:18 PM
Is this just a theoretical curiosity? It certainly is that. But I think that a range of different prices based on income-related factors does arise in things like car sales and pharmaceuticals (especially internationally).
January 12, 2026 at 1:18 PM
The result generalises and has interesting welfare properties. While high income effects allow for price discrimination, this is not the only preference-related assumptions that can lead to price discrimination.
January 12, 2026 at 1:18 PM
What is key is that the monopolist can practice market power at an individual customer level by restricting purchases to each individual (whereas the usual market power assumption is that it is a restriction across the market).
January 12, 2026 at 1:18 PM
Subsidising the poor makes them "wealthier," which increases their attachment to the good, which raises their reservation price for resale. The monopolist exploits this by discounting to low-income buyers until their reservation price equals what high-income buyers must pay.
January 12, 2026 at 1:18 PM
... the discount leaves them with higher disposable income, and since the good is normal, they now value it more—raising the minimum price they'd accept to part with it.
January 12, 2026 at 1:18 PM
Assume there are high and low income types, and the monopoly is selling a normal good. Then, as the monopoly discounts to low-income types, if they sell to high-types, they are foregoing more because ...
January 12, 2026 at 1:18 PM
Why? Rather than assuming quasi-linear preferences as we have done in IO for decades, I assume demand comes from utility over all goods maximised subject to a budget constraint. You know, Samuelson type micro. This allows, importantly, for things like income effects.
January 12, 2026 at 1:18 PM