“We use ML trained on US death certificates from March 2020 to December 2021 to predict 155,536 unrecognized COVID-19 deaths. This indicates 19% more COVID-19 deaths occurred in US than officially reported.”
The undercounting was concentrated among people of color & low income (read below).
1/n
P.S. Pre-post differences are *not* valid treatment effect estimates. Why? Here's a post by @statsepi.bsky.social: statsepi.substack.com/p/one-simple..., here's a post by me: www.the100.ci/2025/01/22/r... >
🧵on my new paper "Synthetic personas distort the structure of human belief systems" w Roberto Cerina I'm v excited about...
🚨 Do synthetic samples look like human samples?
We compare 28 LLMs to the 2024 General Social Survey (GSS) to find out + develop host of diagnostics...
@agazmararian.bsky.social: Federal leadership can be lost in the "fog of messaging." Gazmararian et al find Biden-Era clean energy investment policies had limited political returns: Voters overwhelmingly attributed new clean energy projects to their governors. news.umich.edu/voters-credi...
The Nazis stopped Jewish doctors from practising, so thousands of them left the country (my great grandfather was one of them). So many left that this was a good natural experiment for estimating the causal effect of losing doctors on infant mortality (& thousands died)
As of 2025, this analysis of privacy policies indicates that every major AI company uses your private conversations to train their models by default. Every prompt, file, photo, personal detail: all of it feeds directly into model training. arxiv.org/pdf/2509.05382
This is consistent with earlier psychometric work that suggests 5-7 is the best response scale options, but good to see that the finding holds up in contemporary research. Also, good to see that labeling scales whether anchored or not has little impact on findings. academic.oup.com/ijpor/articl...
Now published #openaccess: my sole-authored article
@readdemography.bsky.social investigates how circular parental labor migration affects children who experience frequently changing family living arrangements.
New from APPAM: The Journal of Policy Implementation & Evaluation (JPIE)!
JPIE will publish short, peer-reviewed articles with timely, real-world insights on policy design, implementation & evaluation.
Learn more: https://ow.ly/TPNW50YmzVr
me and @cmulhern.bsky.social's paper on school counselors is out in econ of ed review
We find more funding for counselors in California in the 2000s increased grad rates, college enrollment & school climate
in short: counselors improve student outcomes and we should invest more in them!
fixest is an R package for fast and flexible econometric estimation, providing a comprehensive toolkit for applied researchers. The package particularly excels at fixed-effects estimation, supported by a novel fixed-point acceleration algorithm implemented in C++. This algorithm achieves rapid convergence across a broad class of data contexts and further enables estimation of complex models, including those with varying slopes, in a highly efficient manner. Beyond computational speed, fixest provides a unified syntax for a wide variety of models: ordinary least squares, instrumental variables, generalized linear models, maximum likelihood, and difference-in-differences estimators. An expressive formula interface enables multiple estimations, stepwise regressions, and variable interpolation in a single call, while users can make on-the-fly inference adjustments using a variety of built-in robust standard errors. Finally, fixest provides methods for publication-ready regression tables and coefficient plots. Benchmarks against leading alternatives in R, Python, and Julia demonstrate best-in-class performance, and the paper includes many worked examples illustrating the core functionality.
arXiv📈🤖
Fast and user-friendly econometrics estimations: The R package fixest
By Berg\'e, Butts, McDermott
image of a report cover, with a US flag in the background, and the text "The Political Disconnect: Working-Class and Low-Income People on What Politics Means to Them and How They Might be Mobilized”
Today I'm releasing probably the most important scholarly thing I've ever worked on - a report based on talking with 144 people about why they don't vote or only vote regularly, and on what needs to be done to build a democracy that can include everyone.
www.swarthmore.edu/u...
please share!
Deadline extended! Apply for the MCCFAD 2026 Summer Data Immersion Program by Monday, Feb. 16th!
#Umich #AcademicSky #PSID
myumi.ch/mRWPj
Very happy that our review article (with the @annualreviews.bsky.social) on „Austerity and Populism“ is now available as preprint: www.annualreviews.org/content/jour... - with @sattlersthomas.bsky.social
Recently accepted by #QJE, “The Price of Housing in the United States, 1890–2006,” by Lyons (@ronanlyons), Shertzer (@econhist-allday), Gray (@econhistoryorbust), and Agorastos: doi.org/10.1093/qje/...
Now out in Sociological Science
(How) do sociologists use GenAI for their research? Find out in our paper.
Written with @ajalvero.bsky.social @dustinstoltz.com and Marshall Taylor. Thank you to everyone who participated in the survey!!
@zoningwonk.bsky.social
Just out in @wepsocial.bsky.social: how housing wealth shapes whether people feel heard. Together with @madselk.bsky.social and @benansell.bsky.social, I looked at a neglected determinant of political efficacy: homeownership.
Read the #OA paper: doi.org/10.1080/0140...
Quick overview below (1/5)
New research by @imbernomics.bsky.social and colleagues using Texas data dives into different economic returns to college majors by race and gender. There are differences in the share of students going into high-paying fields, as well as earnings within those fields.
It's been a long time coming. After several years of revisions, my book Emotional Filipinos is coming out on April 15th. Please share and retweet! It would really mean a lot to me. Preorder sales available here: ugapress.org/978082037545...
🚨New paper in the International Journal of Sociology of Education where I analyse how growing up in single-parent families influences tertiary education attainment, which mechanisms explain this, their heterogeneity by parental SES and how it has changed over time.
👉🏽 doi.org/10.17583/ris...
"We find that minimum wages substantially reduce intergroup wage inequality at least up to the 20th wage percentile, with no evidence of adverse employment effects."
Is one of your goals for 2026 to write more? One question I get all the time is, "How do you make time to write?" So, here is my unsolicited writing advice for the new year:
This thread perfectly explains why AI has little value for qualitative research.
Yes, large language models can find patterns in qualitative data. But, they're trained on what we already know. So, they won't find anything surprising. And, surprise is qualitative research's primary value-add.
Community college enrollment has been declining since 2009/10. The optimal policy response to this depends on the root of this decline.
I'm thrilled @nber.org today released my working paper with Harvard PhD Joe Winkelmann titled:
"Labor Market Strength and Declining Community College Enrollment"
Segregation was not just a Southern phenomenon, it was a national issue. Explore the spatial history of racial segregation and discover how it connects to our communities today: greenbookproject.osu.edu #GreenBookProject #CommunityMap #EconSky
Clarifying the Diploma Divide: The Growing Importance of Higher Education for Political Identity: https://osf.io/cm2np
Why did that happen, what does it tell us, and what can we do next?
That’s what my forthcoming book, The Wage Standard, is about: how rules, norms, and power in the labor market shape who shares in growth. Coming out March 31. Pre-order here:
www.thewagestandard.com
broke: teach the kids to code because it's employable
woke: teach the kids humanities because it's important to know about art, philosophy and other languages
bespoke: teach the kids to code because it's important to know about art, society, philosophy and other languages
this is why quantitative social scientists, the perfect balance of both worlds, should be in charge of everything.
