Joshua Goodman

@joshua-goodman.com

Professor @ BU Wheelock & Economics | Co-editor @ JHR | AER Board of Editors 2027-29 | White House CEA 2022-23 | www.joshua-goodman.com

@toddrjones.bsky.socialReposted by @joshua-goodman.com

🚨 Public good alert: I put together a list of 100+ free datasets for social science research, spanning a large number of topics.

I hope it's useful for students, researchers, and anyone looking for interesting data to explore.

It's on GitHub - contributions welcome! github.com/toddrjones/d...

@robertkelchen.comReposted by @joshua-goodman.com

The Department of Education's research arm appears to have gotten essentially all its funds out the door in FY26. This must have been tricky to report on given the number of people who didn't want to talk or go on the record with reporters.

@nobelprize.orgReposted by @joshua-goodman.com

BREAKING NEWS
The Royal Swedish Academy of Sciences has decided to award the 2026 Sveriges Riksbank Prize in Economic Sciences in Memory of Alfred Nobel to Robert E. Hall and Michael Woodford “for microeconomic foundations of macroeconomic stabilisation.”

#NobelPrize

Photo, no description
@jakemgrumbach.bsky.socialReposted by @joshua-goodman.com

Economists really need to get off twitter. EconTwitter is now just like “bad instruments are DEI and women ate stupid” interspersed with a few good economists talking about the latest Claude update

Labrador Skeptic & @SaysSimulation X.com This is a perfect example of DEl in the wild. Women gets Harvard PhD, Princeton postdoc, NY Fed job. Impeccable credentials - on paper. Writes post demonstrating that she doesn't under the basics of causality & statistics. There's also a very strong urban class component 1/ Natalia Emanuel * @NataliaHEmanuel • 2d Coffee shops boost innovation. 10% more Starbucks in an area -> 3-4% more inventions. People gather over a cuppa. They share ideas, discuss how to untangle a knotty problem, help se... log (Starbucks + 1) County & Year FEs Local Time-Varying Controls Num. Counties * Years Num. Counties Num. Instruments Adj. r2 Kleibergen-Paap Wald F-stat Mean of Dep. Var. (1) 1.319*** (0.231) No No 61480 3074 200 0.174 3.099 1.260 log(Patents) (2) 0.419*** (0.0584) Yes No 61480 3074 200 -0.0141 1.616 1.260 (3) 0.370*** (0.0718) Yes Yes 37464 2572 200 -0.0253 3.483 1.844 :Demosthenes Twin @DemosthenesT....9h X Women really are retarded. This one has a phd from Harvard and still can't grasp simple concepts. She actually posted this thinking she was clever. A dimwit can see the glaring error of her simple thoughts. (1) log(Pate. (2) cks + 1) 1.319*** 0.419** (0.231) (0.0584 Year FEs No Yes -Varying Controls No No aties * Years 61480 61480 aties 3074 3074 'uments 200 200 0.174 -0.0141 •Paap Wald F-stat 3.099 1.616 o. Var. 1.260 1.260 Natalia Emanuel @ @NataliaHE... • 2d Coffee shops boost innovation. 10% more Starbucks in an area --> 3-4% more inventions. ...Alpha Numer0@@alpha_numer0.thg These are what you call middle-upper class townies. They look great on paper, seem competent when you talk to them, but have no idea how the world works or any real common sense as to how things relate to other things, with no real beliefs of their own to boot. I see these types of women in software all the time - many of them were shoe-horned into their roles by a system that wanted to "promote women" at any cost, and really are not good at this kind of stuff or have any actual interest in improving the products they work on. 173 3114 il 61只企
@tkeskinturk.bsky.socialReposted by @joshua-goodman.com

peter hull comparing TWFE and modern DIDs & it seems TWFE is OK.

people will go "YOU SEE IT DIDN'T MATTER AFTER ALL" but regardless of whether TWFE is ultimately useful or not that all DiD chaos was a net good because it clarified the estimand.

tinyurl.com/twfevsdid

Two-Way Fixed Effects and Modern Difference-in-Differences: An Empirical Comparison Peter Hull, Brown University 8 October 2026 This note compares two-way fixed effects (TWFE) estimates from top general-interest journals in economics with estimates from the group-time difference-in-differences estimator of Callaway and Sant’Anna (2021; CSA) and the imputation estimator of Borusyak, Jaravel, and Spiess (2024; BJS). We use recovered observation-level data and focus on binary treatments with staggered adoption. Sign reversals are relatively rare, and especially so with statistically significant differences.
@mzinshteyn.bsky.socialReposted by @joshua-goodman.com

Among my value-adds to this SAT at UC debate? I calculated the rates of Ds and Fs in every entry-level calculus course at UC Berkeley from 2017 to 2025 to show that while average GPAs stayed THE SAME, the D/F rate jumped by nearly 40% on average (from ~8% to ~11%).

Older posts