That’s easy to do in Stata, you just need to access an invisible, cryptically-named matrix that requires learning a different programming language to view
Paul Goldsmith-Pinkham
@paulgp.com
Yale SOM professor & Bulls fan. I study consumer finance, and econometrics is a big part of my research identity. He/him/his
Second attempt at bread from 'Four Water Salt Yeast' by Ken Forkish, but now with reduced hydration and again a pure sourdough / levain variant. Getting there. #breadsky
The unsung hero of public EV-charger buildout is [checks notes] Wal-Mart?
Walmart Leapfrogs EV Charging Competitors With Rapid Buildout
In a matter of months, the retailer has become a major player in the business of fueling electric cars.
bloomberg.com
Paul Goldsmith-Pinkham, Associate Professor of Finance at the Yale School of Management, presented "Using AI in Household Finance Research: A Practical Guide" at the Household Finance Conference during the 2026 Summer Institute. www.youtube.com/watch?v=c5g6...
I'll be at New Researchers Conference 2026 this weekend chatting about some recent work on estimating peer effects in noisy networks! arxiv.org/abs/2605.03204
New paper with @floswald.bsky.social @lars.vilhuber.com and collaborators on reproducibility with LLMs, including what researchers should document to make LLM-assisted work transparent ✌️ arxiv.org/abs/2607.24372
In other news news.cornell.edu/stories/2026... with @johnabowd.bsky.social and a bunch of other folks incl. @simson.net #privacy
Abowd, Vilhuber win 2026 PET award | Cornell Chronicle
John Abowd, professor emeritus of economics, statistics, and data science at Cornell and Lars Vilhuber, research professor and executive director of ILR’s Labor Dynamics Institute, have received the 2...
news.cornell.edu
Always enjoy reading @bestofecontwitter.bsky.social , but I‘m turning into an old head in the NBA — I could out post the hell out of those clowns.
Nice paper flagging issues that can show up for researchers using LLMs papers.ssrn.com/sol3/papers....
Randomness in large language models: What researchers need to know (and report)
Large language models (LLMs) are increasingly used to generate data for research. Typical use cases are classifications, annotations, information extraction, an
papers.ssrn.com
Everything you've ever wanted to know about AI and research github.com/paulgp/nber_... Video will be posted I believe.
github.com
Paul Goldsmith-Pinkham@paulgp.com · 2w agoI will be presenting in the NBER Household Finance Summer Institute session this Friday at 1pm on "Using AI in Household Finance Research: A Practical Guide." www.nber.org/conferences/...
Recent case, re-run of the authors' code failed on a first-stage regression, blocked code from completing. Hard to initially diagnose, because "qui do regression_55.do" shows no output... 🤦Back and forth with authors, they provide evidence that it works on their system. We send MWE to Stata support👇
Not every open-source project wants more contributors right now, and that's fine. Four shapes maintainers run — walled castle, clique, university, hippie commune — and your `CONTRIBUTING.md` is saying the quiet part out loud. No wrong answer, just an honest one. blog.stdlib.io/do-you-want-...
Do you WANT contributors?
Four ways to run an open-source project—castle, clique, university, commune—and which one you're really running.
blog.stdlib.io
Here's the video of my presentation of this paper at the NBER SI: www.youtube.com/live/405kSya...
Macroeconomics and Productivity, NBER Summer Institute
YouTube video by NBER
youtube.com
Jeffrey Lin@jeffrlin.bsky.social · last mo.On February 14, 1876, Bell & Gray each filed for a telephone patent. Such collisions were once routine — up to 1 in 20 patents triggered an "patent interference." Over the next 150 years, that rate fell by more than 98%. In a new paper, we explain why — and what it might cost.
OUSD(P) staff when Colby gets to the office this morning:
Haydon@haydonmp.bsky.social · 2w agoI can’t imagine a more personally humiliating thing than to be Chotiner’d by Douthat. Astonishing.
Relatedly, my SIEPR colleague @weakinstrument.bsky.social is in the New York Times today, talking about the AI equities boom and what happens if it looks like AI investments will not be profitable for a while: www.nytimes.com/2026/07/22/b...
A very sensible point from @susanathey.bsky.social @guidoimbens.bsky.social & Zoe Ji. #EconSky arxiv.org/pdf/2607.05792
Even in the very depths of it being so over, I found within me an invincible we are so back
I will be presenting in the NBER Household Finance Summer Institute session this Friday at 1pm on "Using AI in Household Finance Research: A Practical Guide." www.nber.org/conferences/...
SI 2026 Household Finance
nber.org
I’m presenting this paper Tuesday 1:40pm at the NBER SI and live on YouTube. Tune in! www.nber.org/conferences/...
SI 2026 Macroeconomics and Productivity
nber.org
Jeffrey Lin@jeffrlin.bsky.social · last mo.On February 14, 1876, Bell & Gray each filed for a telephone patent. Such collisions were once routine — up to 1 in 20 patents triggered an "patent interference." Over the next 150 years, that rate fell by more than 98%. In a new paper, we explain why — and what it might cost.
Open access in QJE!📢 Using newly assembled archival data, we show how the AMA's nationwide campaign against national health insurance, labeling it "socialized medicine" and promoting private insurance, helped steer the U.S. away from universal health care academic.oup.com/qje/article/...
Why Doesn’t the United States Have National Health Insurance? The Political Role of the American Medical Association
Abstract. This article examines how the American Medical Association (AMA) helped shape the development of the U.S. health insurance system in the critical
academic.oup.com
With additional PM2.5 data through 1000Z today, I think we can safely say that this is the most impactful wildfire smoke event in recent history -- around 41.3 million people have been exposed to at least one hour of Hazardous PM2.5 & the number of person-hours spent in Hazardous range is now 368.9m
John Voorheis@johnvoorheis.bsky.social · 3w agoRough live calculations of impact of the severe #wildfire smoke event as of 13:00 UTC (9AM ET): about 24.7 million people have experienced at least one hour of hazardous PM2.5 (above 225 ug/m3) 201.5 million person hours have been spent in the Hazardous range (US numbers only)
New working paper! Regression to the Mean in Place-Based Crime Prevention
