rodneywhitecenter.wharton.upenn.edu/2025-summer-...

12 more days to apply to our summer school in structural estimation! It's a hands-on course in which you learn how to do SMM and then work on a problem set with your classmates. Open to PhD students in finance, accounting, econ. And faculty.

Me me me!!!

I was today old when I learned that there are two camps of people who use LaTeX. Those who tolerate a gazillion errors as long as the pdf looks OK, and those who like zero errors.

HOW CAN PEOPLE LIVE WITH A GAZILLION LATEX ERRORS It's just not that hard to learn enough LaTeX to make it work right.

Greatest email from a student:

"Now I get why you taught us about good replication practices - they are really important."

Luke Taylor, Stephen Terry, and I are doing it again. This will be the 5th structural summer school in finance. We do not just teach papers. We teach techniques, and then we have students work on a problem set. Please get your students to apply.

rodneywhitecenter.wharton.upenn.edu/2025-summer-...

So the bottom line is that the original Holmstrom intuition works, but only in small markets when the stakes are relatively low. Otherwise, heterogeneity in costs across competitors matters more for their strategies.

so airlines don't want to enter when demand is high because this good state of the world gets filtered out, and they are less likely to exit when demand is low because they are not punished for this common shock.

RPE has no effect on entry-exit strategies in large markets. Demand is so high that all the airlines always want to enter. In medium-sized markets, removing RPE helps weak competitors, who are no longer compared to their more efficient rivals. In small markets, RPE filters out common shocks,

So we estimated a dynamic game in which airlines compete with one another under the assumption that their contracts contain RPE incentives. (I know. Not an optimal contract. But it is what we see in the data, so our estimation is just characterizing an observed equilibrium with observed contracts.

Not heavily researched. Why? No data. Measuring how much an executive's compensation depends on competitors' performance by looking at proxy statements is a mess. There's information there, but it does not provide clean measures. So it's a casualty of lampost economics.

Well, this got accepted at the RFS a while back, but I've been avoiding social media, so I didn't post about it. My coauthor, Li He, was the brains here. We are looking at the interaction between relative performance evaluation and competition.

papers.ssrn.com/sol3/papers....

A new reason to hate Overleaf. Its grammar checker doesn't know when to use "due to" and "because of."

Work avoidance behavior for reviewing PhD applications.

Just finished teaching the first class of my Econ undergrad course (which is basically an investments course). Love these kiddos. Smart. Good attitudes. So much more fun than in my previous life as a b school prof.

I have now posted 14 decks of slides covering a lot of DiD materials.

I also want to add a few more decks:

- Triple Differences
- DiD with Continuous Treatment
- Instrumented DiD
- Synthetic DiD and related extensions
- Heterogeneity analysis in DiD

I also want to add more applications to these!

Please don't just dump important parts of papers into internet appendices to make them appear shorter. Make a coherent, self-contained essay, and then figure out what is ancillary and what is not. Only the ancillary stuff goes in an appendix.

Referees hate LONG papers.

Rest in peace to Submariner LT James Earl Carter Jr, USN (Ret). Carter, the only U.S. President to graduate from the U.S. Naval Academy, leaves a legacy of service. The USS Jimmy Carter (SSN-23), a Seawolf-class fast-attack submarine, carries his name in honor of his dedication to our nation.

Burgers and slaw for Xmas dinner. And beer.

Just way too many mistakes. It's kinda impressive. This is for coding in anything but python. Even other open-source languages.

I mostly use Claude, so this is why the change in GPT was a surprise.

I haven't used GPT in a while because I've mostly been using Claude.

Haven't used chatGPT in a while. Its answers have gotten VERY wordy. There must be a LOT of overly verbose garbage on whatever training data it has used.

#EconSky Price theory à la UCLA taught by Jack Hirshleifer in 1972. His course outline with reading list has been transcribed and posted at Economics in the Rear-view Mirror. www.irwincollier.com/ucla-price-t...

The Puritans. Edward Hopper.