Sign in

Jesse Rothstein

@jrothst.bsky.social
21K followers 1.1K following 301 posts

Economist, public policy wonk, professor at UC Berkeley, faculty director of California Policy Lab & Center for Studies in Higher Education

PostsRepliesMedia
Reposted by Jesse Rothstein
National Education Policy Center @nepc.bsky.social · 17/09/2026
Higher ed spends more money on educating students from affluent families. Two researchers tried, but failed, to find a justifiable reason why. @jrothst.bsky.social bit.ly/4AoTcl5
nepc.colorado.edu
It’s Hard to Justify This Higher Ed Spending Gap. So Why Does It Persist?
Higher ed spends more money on educating students from affluent families. Two researchers tried, but failed, to find a justifiable reason why.
163
Jesse Rothstein @jrothst.bsky.social · 17/09/2026
A belated thank you for this. Definitely interesting and useful.
010
Reposted by Jesse Rothstein
Laura K Gee @laurakgee.bsky.social · 16/09/2026
Submissions are open for SOLE! I’m on the scientific committee and I can’t wait to see all the cool papers that apply. www.sole-jole.org/upcoming-mee...
075
Jesse Rothstein @jrothst.bsky.social · 11/09/2026
These two campuses (out of nine) admit more or less any UC-eligible students, so are not really affected by the SAT free policy. That's why they are excluded. Here are admissions shares of URMs with those two campuses included - the expanded y-axis obscures what is going on elsewhere.
URM share of admissions by UC campus, including UCM and UCR.
223
Jesse Rothstein @jrothst.bsky.social · 11/09/2026
www.chronicle.com/article/a-re...
chronicle.com
A Researcher Adds What’s Been Missing From the UC Testing Debate: Nuance
A new brief by a prominent scholar argues that some criticisms of the University of California’s test-free policy are predicated on assertions that simply aren’t true.
3269
Reposted by Jesse Rothstein
Stephen Wild @stephenjwild.bsky.social · 09/09/2026
THANK YOU FOR YOUR ATTENTION TO THIS MATTER
The College Board, publisher of the SAT, has been claiming for decades that high
school GPAs were increasingly unhelpful for admissions due to rampant grade inflation.
But, while the existence of grade inflation is not disputed, the evidence in support of the
claim that high school GPAs have been so inflated as to be meaningless is surprisingly
thin.

There are several reasons to doubt the claim, at least as it applies to the UC. First,
validity studies (discussed above) – even the STTF report, despite the methodological
limitations noted above – continue to show that high school GPAs are strongly
predictive of success in college. Second, a commonly observed pattern is that grade
inflation is more rampant at upper-middle-class public and private high schools, and less prevalent at schools serving more disadvantaged students. This pattern would help the high school GPA’s predictive validity, since students from the former schools are generally better prepared for college. Third, policies that guarantee admission to applicants ranked in the tops of their classes by GPAs have led to increases in diversity of admitted applicants, and evidence indicates that the students so admitted benefitted, in terms of their college and career outcomes, from their admission, relative to the outcomes they would have achieved had these policies not been in place (Black, Denning, Rothstein 2023, Bleemer 2024). This result would be unlikely if GPAs, and by extension class ranks, were meaningless.
0112
Jesse Rothstein @jrothst.bsky.social · 09/09/2026
The University of California is in the midst of what looks to be a year-long debate about the use of SAT scores in undergraduate admissions. I wrote a FAQ that attempts to collect the facts that are relevant to this debate, and to dispel some common misconceptions: escholarship.org/uc/item/6kx2...
escholarship.org
13617
Jesse Rothstein @jrothst.bsky.social · 09/09/2026
As far as I know, the campus completely failed to deliver on that promise! It is a real shame - I think it would have been a real attraction (not to mention a safe place for him to park his bike).
011
Jesse Rothstein @jrothst.bsky.social · 08/09/2026
I'm afraid I'm mostly to blame for this one!
110
Jesse Rothstein @jrothst.bsky.social · 08/09/2026
Similar social interactions could bias twins-based estimates of heritability of education, employment, earnings, even IQ. This bias may be part of “missing heritability” – the tendency for twins studies to indicate greater heritability than those based on direct measurement of genes.
071
Jesse Rothstein @jrothst.bsky.social · 08/09/2026
The tendency to work at the same firm isn’t because twins have genetic attractions to specific firms – it almost certainly means they like to work together. That kind of social interaction, if different for same-sex twins, biases twins estimates of heritability.
Photo of twins in matching UPS uniforms.
141
Jesse Rothstein @jrothst.bsky.social · 08/09/2026
But we show that is misleading: SS twins are much more likely than OS twins to work at the same firm – meaning they have the same firm effect. When we remove the same-firm pairs from the sample, firm effects no longer appear heritable, and heritability of other outcomes shrinks.
132
Jesse Rothstein @jrothst.bsky.social · 08/09/2026
On their face these results point to a genetic component of the variation in employment, earnings, and access to high-paying firms.
110
Jesse Rothstein @jrothst.bsky.social · 08/09/2026
We find all outcomes are more strongly correlated among same-sex twins than among opposite-sex twins. OS twins look like siblings, but SS twins don’t. In the CTD, this indicates that labor market outcomes are moderately heritable – even the pay policy of the firm at which people work.
Figure 3: Between-sibling and between-twin correlations of earnings outcomes.
122
Jesse Rothstein @jrothst.bsky.social · 08/09/2026
Discussion of the validity of the twins design has focused on the “shared environment” assumption, which is strong. We show there’s another problem with it, potentially more serious, when applied to social outcomes.
110
Jesse Rothstein @jrothst.bsky.social · 08/09/2026
The “classical twins design” uses the contrast of identical vs. fraternal twins to measure the “heritability” of outcomes. We extend it to same-sex vs. opposite-sex twins. Heritability = the genetic component of outcomes, the share of variance due to genes rather than environment or chance.
110
Jesse Rothstein @jrothst.bsky.social · 08/09/2026
We use Census Bureau data to study correlations among siblings and twins in labor market outcomes in the United States. Data cover the full population of eight birth cohorts, born 1983-1990. We observe employment and earnings as adults, and decompose earnings into person and firm effects a la AKM.
120
Jesse Rothstein @jrothst.bsky.social · 08/09/2026
New paper alert! “Heritability and social interaction in labor market outcomes: Evidence from the population of twins in the United States,” with David Card and Moises Yi, out today in PNAS. www.pnas.org/doi/10.1073/...
pnas.org
PNAS
Proceedings of the National Academy of Sciences (PNAS), a peer reviewed journal of the National Academy of Sciences (NAS) - an authoritative source of high-impact, original research that broadly spans...
43716
Reposted by Jesse Rothstein
Caitlin Patler @caitlinpatler.bsky.social · 04/09/2026
🚨 Job Alert! 🚨 Chloe East (CU Boulder Econ) & I are hiring a postdoc to work on causal research on impacts of imm policy on immigrant children & families + rapid response descriptive work. Please share & encourage qualified candidates to apply! aprecruit.berkeley.edu/JPF05505
aprecruit.berkeley.edu
PostDoctoral Scholar Employee - Goldman School of Public Policy
University of California, Berkeley is hiring. Apply now!
14246
Reposted by Jesse Rothstein
NBER @nber.org · 28/08/2026
Colleges that high-testing and higher-income students attend spend much more on instruction. Potential economic justifications for this allocation have little empirical support, from Zachary Bleemer and Jesse Rothstein www.nber.org/papers/w35650
024
Jesse Rothstein @jrothst.bsky.social · 27/08/2026
www.insidehighered.com/news/admissi...
insidehighered.com
New Paper Takes Aim at “Meritocratic” Admissions Policies
Wealthy students with high standardized test scores make up the majority of students at the most selective universities, which also spend far more on student instruction than less selective institutio...
1177
Reposted by Jesse Rothstein
Sandy Black @econsandy.bsky.social · 19/08/2026
The Society of Labor Economists is piloting a new mentoring program intended to provide small group mentoring, providing advice on how to improve research and navigate the publishing process. We are seeking both mentors and mentees--apply here! www.sole-jole.org/sole-mentoring
sole-jole.org
SOLE Mentoring
02416
Jesse Rothstein @jrothst.bsky.social · 18/08/2026
I believe this one is canonical. 😉 www.berkeley.edu/wp-content/u...
020
Jesse Rothstein @jrothst.bsky.social · 18/08/2026
I do wish @chalkbeat.org had chosen a different cover image, though - our story is not just about Harvard, but about the broad spectrum of higher education. This graph, from @mattbarnum.bsky.social's article, just shows *public* colleges and universities. bsky.app/profile/matt...
273
Reposted by Jesse Rothstein
Matt Barnum @mattbarnum.bsky.social · 18/08/2026
New- American higher education is designed so that the academically and financially rich get richer through a well-funded education. The startling scope of this unequal structure is documented in a new study by @zbleemer.bsky.social and @jrothst.bsky.social. www.chalkbeat.org/2026/08/18/s...
24019
Jesse Rothstein @jrothst.bsky.social · 18/08/2026
A great article by the always-excllent @mattbarnum.bsky.social about my new study with @zbleemer.bsky.social on the "meritocratic consensus" in higher education: www.chalkbeat.org/2026/08/18/s...
chalkbeat.org
How American higher education helps the rich get richer
Higher education’s admissions hierarchy is also a spending hierarchy. The students who arrive with the most advantages often receive the most resources.
11814
Jesse Rothstein @jrothst.bsky.social · 17/08/2026
Either the AI bot will come up with better phrasing, and I can just use that, or it won't. How would you word it better? FWIW, I can't set an AI bot to respond to these automatically - as of now, the university says we can't give AI access to our email due to FERPA concerns.
001
Jesse Rothstein @jrothst.bsky.social · 17/08/2026
For obviously AI ones, I add to the template: "PS. You should know that your email reads like unfiltered AI writing. While I appreciate your apparent interest in my work, I'm not sure whether the interest is yours or the AI's. I don't think this is doing you any favors." Got anything better?
2172
Jesse Rothstein @jrothst.bsky.social · 17/08/2026
I've tried to respond politely to all of the requests I get that are personalized - just a "no, sorry, good luck" template. But the volume is going up - students are clearly setting AI bots to write 1000s of "personalized" emails. Argh.
1336
Reposted by Jesse Rothstein
Jake Grumbach @jakemgrumbach.bsky.social · 13/08/2026
🚨🚨🚨 We're hiring a junior professor in Policy Analysis/Public Administration/Public Management here at the UC Berkeley Goldman School of Public Policy 🚨🚨🚨 aprecruit.berkeley.edu/JPF05497
aprecruit.berkeley.edu
Assistant Professor - General Public Policy and Administration: Policy Analysis, Public Administration and Management - Goldman School of Public Policy
University of California, Berkeley is hiring. Apply now!
09456
Jesse Rothstein @jrothst.bsky.social · 13/08/2026
🧵
087
Reposted by Jesse Rothstein
Berkeley Population Sciences @berkeleypopcenter.bsky.social · 13/08/2026
The UC Berkeley Department of Demography is recruiting a tenure-track Assistant Professor, open field. Applications are due October 15, 2026. See details and applications: aprecruit.berkeley.edu/JPF05488 Please share widely.
aprecruit.berkeley.edu
Assistant Professor - Open Field - Demography
University of California, Berkeley is hiring. Apply now!
12125
Jesse Rothstein @jrothst.bsky.social · 07/08/2026
Somehow the full graph didn't come through. Here it is.
FRED graph of seasonally adjusted and non-seasonally-adjusted employment in local government education, July 2016-July 2026.
2220
Jesse Rothstein @jrothst.bsky.social · 07/08/2026
I always find it helpful to look at (a) levels and (b) non-seasonally adjusted data alongside the seasonal. Here I've done both. You can see why the SA employment looks like it declined, but also why it is important to be cautious in interpretation - the SA algorithm has a *lot* of work to do here.
Seasonally adjusted and non-seasonally-adjusted employment in local government education from July 2016-July 2026. Source is the Bureau of Labor Statistics Current Employment Statistics series; the graph was made by FRED, a tool provided by the St. Louis Fed.
2375
Jesse Rothstein @jrothst.bsky.social · 20/07/2026
For cross-country comparisons, it seems important to take account not just of UI financing but also of general employment protection rules. E.g., many European countries have rules that limit idiosyncratic layoffs, making macro conditions a larger share of unemployment insurance risk.
010
Jesse Rothstein @jrothst.bsky.social · 20/07/2026
Because I think seasonality is likely an industry-specific source of variation - likely the dominant part. I'm less worried about idiosyncratic firm choices to take on more risk - by the time the bad outcome from those choices is realized, the firms are probably out of business anyway.
120
Jesse Rothstein @jrothst.bsky.social · 20/07/2026
But there is a systematic between-industry component; absent experience rating, industries with stable employment are subsidizing those without. I'm less worried about having firms bear firm-specific risk than having industries bear industry-specific predictable costs.
120
Jesse Rothstein @jrothst.bsky.social · 20/07/2026
Do you anywhere address seasonality directly, or other similar practices (e.g., auto plants used to do seasonal shutdowns and all workers would go on UI)? Why should other firms subsidize that? Maybe experience rating is bad enough that this is a cost worth bearing, but seems impt to address.
120
Jesse Rothstein @jrothst.bsky.social · 20/07/2026
Interesting. I see the cost of supporting workers during the off season as part of the cost of employment in seasonal industries, and worry that eliminating experience rating (which could be done much better, of course!) would lead firms to business models with more boom+bust employment.
341
Jesse Rothstein @jrothst.bsky.social · 20/07/2026
Don't the higher taxes paid by construction firms reflect higher benefit payments to construction workers, via experience rating? How could one redesign without excluding construction workers from UI benefits or creating large net transfers in UI system from non-construction to construction firms?
1140
Jesse Rothstein @jrothst.bsky.social · 15/07/2026
This gets to what I see as a similar place from a different direction: bsky.app/profile/dsqu...
020
Jesse Rothstein @jrothst.bsky.social · 14/07/2026
I think Scott would argue most of the important knowledge can’t be codified. But maybe he’s wrong - there’s an interpretation where he’s just arguing against doing centralization badly.
230
Jesse Rothstein @jrothst.bsky.social · 14/07/2026
I didn’t! Adding it to the reading list now. I won’t let it linger there as long as this book did…
030
Jesse Rothstein @jrothst.bsky.social · 14/07/2026
Yeah, that’s another of my critiques - it should have been called “Seeing Like A Large Organization.” Probably wouldn’t have sold as well, however.
1200
Jesse Rothstein @jrothst.bsky.social · 14/07/2026
Last thought: One of my criticisms of the book is that Scott emphasizes the costs of centralization/standardization but ignores the benefits. Some jobs might be worth doing only half as well for 1/100 the price. This is a story of AI-driven enshittification, I’m afraid. It seems plausible. 6/6
1422
Jesse Rothstein @jrothst.bsky.social · 14/07/2026
So is metis as important to as many jobs as Scott makes it out to be? I suspect yes. But we are about to find out. If it is, AI will not be able to replace many jobs wholesale – though it may supplement them. If it isn’t, the employment impacts of AI may be as large as many are predicting. 5/6
2262
Jesse Rothstein @jrothst.bsky.social · 14/07/2026
Insofar as metis is important to a job, AI as currently constituted won’t be able to replace it. An AI can learn an instruction manual. But metis is non-codified knowledge – there is no manual. And LLMs’ limited context windows prevent them from building up knowledge from long practice. 4/6
3466
Jesse Rothstein @jrothst.bsky.social · 14/07/2026
Metis is, as he defines it, skill and knowledge earned by long practice that can’t be written down – e.g., how to navigate ships through the particular currents of a particular harbor. 3/6
2313
Jesse Rothstein @jrothst.bsky.social · 14/07/2026
Scott emphasizes the importance of craft or “metis” – “local knowledge” about how to farm in a particular location, what customers will need, or how the parts of an ecosystem interact – and argues that large-scale, centrally-governed systems can’t accommodate the complexity of local knowledge. 2/6
1413
Jesse Rothstein @jrothst.bsky.social · 14/07/2026
I just finished reading “Seeing Like a State.” It argues that “high modernism” (Brasilia, soviet collectivization, Tanzanian forced villagization) failed because it put complex humanity into overly-simple schema. I think it has interesting things to say about AI’s impact on the labor market. 🧵
58618