Physician and health economist at Penn.
I study opportunity, hope, and health @oppforhealthlab.bsky.social.
Personal website: https://sites.google.com/site/atheendar/
My #book on #stuckness now exists as a listed forthcoming, preorderable title!
www.harpercollins.com/products/stu...
Coming out end of June 2027.
Working ToC attached.
More details to follow!
@oppforhealthlab.bsky.social
Fascinating ppr by Clemens et al on synthetic control methods.
They look at the impact of various choices (time period, outcome, donor pool) on estimates of impacts of California's $20 fast food min wage on employment.
www.nber.org/papers/w35931#econsky#medsky
Second, we explore in a piece on @statnews.com about what this result and the broader evidence means.
I've come to the uncomfortable realization that the largest public health turnaround in decades may have little to do with what I thought was important.
www.statnews.com/2026/10/23/o...
“There are big questions about why opioid overdose mortality rates are falling nationally,” says Dr. @mlbarnett.bsky.social. “A lot of people think that increased use of the overdose reversal agent #naloxone is a big reason. This study changes the narrative and says we have to look to other causes.”
I'm not so sure about this claim.
Yes estimates are not "statistically significant" (esp with Bonferoni).
But all of point estimates negative. Also scaling them by potential change in tylenol use during recall allows for large negative effects (i.e., less tylenol = less ADHD, autism)!
The 1982 Tylenol recall, which likely reduced acetaminophen use in pregnant women and fetal exposure, was not associated with changes in autism or ADHD among affected cohorts, from Christopher M. Worsham, Charles Bray, Joshua Rothman, Brian Kwan, and Anupam Jena www.nber.org/papers/w35873
I think link between tylenol use during in utero period and autism is unlikely. But this natural experiment doesn't prove that.
Natural experiments can be useful, but lack of power and external validity are real impediments to their usefulness
Let me know if I am reading this incorrectly!
Hi #medsky#econsky! I have an interesting problem.
RCT: both units and timing of entry into study randomized.
Need a staggered adoption estimator.
Problem: cluster S.E.s in new estimators are over-conservative. See www.nber.org/papers/w24963
How've you dealt w this? Has welfare implications!
🏆 Congratulations to LDI’s Scott Halpern and Paula Chatterjee on receiving awards from @acpimphysicians.bsky.social! Recognized for their contributions to behavioral medicine & internal medicine, their work advances research and policy to improve health care.
Read more: https://bit.ly/4i79WEZ
The best #doctors know how to quantify, manage, and deal with #uncertainty.
For the average or bad doc, AI can help with uncertainty -- but only to a point. AI trained on all the data we give it (whatever ends up being measured, with whatever spin and bias we throw in there). It has blind spots.
My daughter Natalia Emanuel and her co-author Emma Harrington have spent years studying how remote work impacts productivity and creativity. I am proud that her new book “In Person” comes out today, making the evidence-based case for why proximity matters at work.
Would love to see an #economics#Nobel that recognizes one of these questions:
"Do we really know what we know and how do we deal with related uncertainty?"
"What are the drivers of well-being?"
Rooting for Chuck Manski, Janet Currie tomorrow.
"They use optical sensors to detect blood pulsing through your veins, and machine-learning models to analyze the data to estimate blood pressure. Like sleep staging, it’s an inference, not a direct measurement."
www.wsj.com/health/welln...
The simple truth is that the president's personal obsession with attacking @drlisadcook.bsky.social is a filthy lie.
There must be not just protection for her job, but accountability for those perverting and corrupting our justice system.
All Americans lose when our institutions are poisoned.
BREAKING Politico:
Trump has just established a committee of inquiry to investigate whether Federal Reserve Gov. Lisa Cook made "false statements in connection with one or more mortgage instruments" — his latest effort to remove her from the central bank.
The administrations obsession with punishing Lisa Cook is absurd and frightening.
It's not even about winning in this case. It's about creating fear and exerting control.
Horrible, horrible look for America.
www.nytimes.com/2026/10/09/b...
Experiments in Uganda reveal that employers and workers prefer to share income through paid work rather than cash, even when the work produces nothing, suggesting that some overstaffing reflects social support rather than inefficiency.
🆕 Work over cash: Why employers and workers prefer jobs to handouts
Today on VoxDev, Elisa Macchi (@brown.edu) & @jeremiastalder.bsky.social (University of St. Gallen) outline evidence on Uganda: voxdev.org/topic/labour...
Summary below with results from 3-year study in Kenya assessing whether savings interventions can strengthen economic prospects for the future and also reduce engagement in behaviors that increase HIV/STI risk. Our results are extremely promising!
Can a savings account reduce heavy drinking?
In a randomized trial with 1,500 men in western Kenya, a savings program cut hazardous drinking by 26% and intimate partner violence by 15%.
New NBER working paper 🧵
Fascinating paper @hthirumurthy.bsky.social and team!!!
The possibility of a bright future motivates. The possibility of a dim one does the opposite.
Something I see all the time in the clinic and hospital.
Can a savings account reduce heavy drinking?
In a randomized trial with 1,500 men in western Kenya, a savings program cut hazardous drinking by 26% and intimate partner violence by 15%.
New NBER working paper 🧵
In 2023, we initiated an RCT in low-income Philly neighborhoods examining how neighborhood improvements and household interventions impact #health+ wellbeing.
Today at the #IAPHS conference, Evan Spencer announced how our efforts averted >1000 crimes over 2 years (2000 w/ spillovers). MVPF=♾️.
More to come on cluster-level hospitalizations, ED visits, as well as individual-level survey-based outcomes (where there is a very interesting pattern of attrition that reflects the success of the program more than a nuisance!).
If you are interested in learning more about what we are finding, checking out this recent interview of Aditi Vasan and myself by the NYC Family Policy Project: familypolicynyc.org/2013/07/01/a...
"The housing shortage in the United States not only drives up rents, it is also a significant driver of poverty." The New York Times summarizes my collaboration Pew, which quantifies how high housing costs lead to higher poverty rates for 10 U.S. states. www.nytimes.com/2026/09/29/r...
Nice summary of this paper by @kevinmilligan.bsky.social (w/ Baker and Gruber)
"Even after their children left preschool, mothers who had access to Quebec’s universal childcare program worked and earned more than mothers who did not."
www.nber.org/digest/20261...
Guideline and physician performance in atrial fibrillation depend on how bleeding harms vary across patients; robust treatment rules can perform well across plausible harm assumptions, from @jabaluck, @LeilaEcon, and David C. Chan Jr www.nber.org/papers/w35798
I expect that we'll soon see a full replication layer for empirical econ: every paper from the past 20 years systematically re-run, audited, and documented on a public website.
Even for papers that used proprietary data, the code and methods will be stress-tested in public.
Proud of @oppforhealthlab.bsky.social lab predoc Ritikaa Khanna and @ncsp-penn.bsky.social fellow Michael Wang for leading this interesting paper.
There is a suble hat tip to Heckman-Vytacil in the discussion section :)
In Medicaid, covering obesity-labeled #GLP1 receptor agonists was associated with higher obesity-labeled use, while total glucagon-like peptide-1 receptor agonist use increased less, suggesting possible substitution across label indications. ja.ma/4xPkKxn
Proud of @oppforhealthlab.bsky.social lab predoc Ritikaa Khanna and @ncsp-penn.bsky.social fellow Michael Wang for leading this interesting paper.
There is a suble hat tip to Heckman-Vytacil in the discussion section :)
In Medicaid, covering obesity-labeled #GLP1 receptor agonists was associated with higher obesity-labeled use, while total glucagon-like peptide-1 receptor agonist use increased less, suggesting possible substitution across label indications. ja.ma/4xPkKxn
In Medicaid, covering obesity-labeled #GLP1 receptor agonists was associated with higher obesity-labeled use, while total glucagon-like peptide-1 receptor agonist use increased less, suggesting possible substitution across label indications. ja.ma/4xPkKxn
"If medical schools and universities don't want the literature to be overwhelmed with garbage," Oransky said, "then they should probably stop giving every applicant ... an incentive to publish a lot of crap."
I know we’re focused on looming AI-driven extinction now, but can you spare a thought for what’s going on with labor demand? I have a new paper about that doi.org/10.26509/frb...
👀 Don’t miss the #IAPHS2026 plenary sessions! 🎤
This year’s plenaries bring together researchers and practitioners to explore trust, influence, public-facing communication, research impact, and how population health science reaches the public. iaphs.org/conference/c...
I've been toiling on a book in my spare time for over a year. I wrote it all myself, but worried that my liberal use of em-dashes would make people think my text was AI generated. Luckily, my writing is so idiosyncratic that this happened.
MedPAC has consistently reported overpayments in Medicare Advantage. Health insurance companies want to kneecap how MedPAC reports these overpayments. Now, insurers have allies in Congress: Republicans, John Wilkerson reports. www.statnews.com/2026/09/16/m...
16 years after starting the project, 13 years after first submitting, it lives:
www.journals.uchicago.edu/doi/10.1086/...
It was a good experience.
More on that later.