Anup Malani
@anup_malani
Followers
8K
Following
72K
Media
486
Statuses
10K
Chief Economist, Centers for Medicare & Medicaid Services. On leave: Professor, U. Chicago. Views expressed here are not intended to reflect those of CMS.
Chicago, IL
Joined February 2013
Generalize to history more broadly. How does history address selection in inference? Are sources weighted by representativeness? What’s the go-to methods book on this?
One claim that I find really forced is the claim some people make that we know more about the Romans than the Romans knew about themselves. Its like, suppose that 2000 years in the future historians had to construct a model for understanding the western world in the 20th and
1
1
3
Does any (top) PhD economics program teach methods for business economics? Ie making informative projections for business or policy? As the fraction of students going to firms rather than academia rises, seem useful. Also, is suspect devoting some of our statistical energy to
0
0
9
MODERATOR: Are you willing to commit to NOT raise the sales tax? MIKIE SHERRILL: I'm not going to commit to anything right now. On Nov. 4, vote NO on Mikie Sherrill. ❌
48
104
330
A hook for importing life history theory into economics! Stacy Rosenbaum & I have a paper in progress using this trade-off to motivate test for predictive adaptive response models in medicine & bio anthropology. I’m using a similar method in a paper arguing low fertility is a
"Our findings indicate that having an extra child is linked to a decrease in women’s longevity by approximately 5 months." "Men, on the other hand, experience smaller and insignificant reductions in longevity of about 3.3 months."
0
0
2
Each possession has a statistical value. Better starting yard line improves that statistical value. (Note that turnover could yield worse than average starting yard line). Estimate this to estimate d(score/poss)/d(turn). These may vary by team.
0
0
1
A game has x possessions per team. Causing a turnover increases the number of possession by some amount. Estimate this to get d(poss)/d(turn).
1
0
1
INTRODUCING: Part 2 of our Stablecoin Payments Report! Co-written with @CastleIslandVC and @dragonfly_xyz, with data now through August 2025. Stablecoin payments hit $10.2B monthly, up 70% since February and 137% since Aug '24, now at a $122B annual rate. Here’s how the market
19
73
231
Basic model: score = possessions x score/possession d(log score) / d(turnover) = d(log poss)/d(turn) + d(log(score/poss))/d(turn) Divide by LHS to convert to percent: %d score = poss share x %d poss + score-per share x %d score-per
1
0
0
Causing NFL turnovers have 2 benefits: an extra possession and potentially better field position. Has anyone looked at the relative benefit to each component? We should be able to approximate it…
1
0
1
Last Thursday, I had an interesting dinner with two prominent economists. Soon, the conversation turned to Joel Mokyr’s work and its relationship with @danwwang’s recent book, Breakneck: China’s Quest to Engineer the Future. Mokyr has long argued that Europe’s Republic of
46
178
603
I have a new paper with Luis Rayo on a key, simple question: will AI end careers as we know them? Link below. We all experience AIs usefulness every day: AI writing code, drafting legal memos, and analyzing spreadsheets. AI can already do many of the tasks that young people
31
204
831
A few days ago, I emphasized the importance of the “intellectual Turing test.” It’s an old idea inspired by Turing’s original test for artificial intelligence. Claim: You only truly understand an intellectual position if you can articulate and defend it so convincingly that a
42
87
492
The puzzle is not the baby boom, but the birth dearth between the world wars. Fertility was in decline since the 1800s The aberration is not the boom—which follows long term trend—but the interwar collapse of births, some places to below replacement.
Quote tweet after quote tweet, reply after reply, telling me I’m so stupid for finding it mysterious that America would have a baby boom after 1945. Except… The baby boom started in ~1935. So, as I was saying: Nobody understands the baby boom!
4
1
13
This is a good example for anyone arguing / teaching that federalism fosters experimentation. Here we have successful being policy in one state being copied by another.
I appreciate this @WSJ editorial. But it proves a point I keep trying to make: we are going to spend the next 5 years explaining why the Southern Surge states made large gains in reading, and alllll those states that passed shallow literacy legislation made lesser gains or no
1
1
3
🚨 Nice to finally see this paper out in AEJ:Macro! @AEAjournals 🧵 We study the impact of the abrupt end of extended unemployment benefits in the U.S. in Dec 2013. The reform provides a natural experiment to measure how benefit duration affects jobs, wages, and vacancies.
3
14
112
1/ Economic outcomes often depend on the distribution of some maximum value (e.g. the highest valuation, best idea, or lowest cost). If the average number of options is large, do such outcomes change when some agents have more options than others? A thread 🧵about this paper.👇
2
49
269
What would really help is if econometrics/stats scholars come up with a metric of best possible answer given data constraints.
1
1
10
Economists address only those questions where the answer meets a certain quality constraint (eg credibility). But the real world requires answers to questions even if answers are not perfect. To increase demand for economists, we need to make answering important questions the
US academic economics market continuing to pull away from the COVID market this week. I created a little tracker for JOE here for those who want to play around with the data. Updates weekly: https://t.co/BPxmqthsNu
3
8
95
@JesusFerna7026 has the right approach. University skunk works to get out in front of the change that is coming. Change or be replaced.
1
0
2
And may be unsustainable if AI and robots also put pressure on demand for labor that requires a traditional college degree or certification.
1
1
1