A Jay Holmgren

A Jay Holmgren

@ajh.bsky.social

Associate Professor at UCSF. Digital health, economics, informatics. Michigander in Chicago or San Francisco, depending on the day. PhD Harvard, 2x UMich alum. Go Blue.

1.1K followers286 following407 posts

A Jay Holmgren
@ajh.bsky.social

Students sometimes ask me, "With all the uncertainty, how can we know when to trust natural experiments?" Buddy, they won't even let me observe the counterfactual worlds!

A Jay Holmgren
@ajh.bsky.social

i'm glad there are more pro-social people out there but if you want me to explain the unconditional cash transfers literature to you, you gotta enroll at UCSF. or i guess venmo me a years tuition. otherwise i don't care.

A Jay Holmgren reposted

Stan Oklobdzija
@stano.bsky.social

In case you were curious about why you're poor, this is emblematic of the big brains in charge of California housing policy. This is how stupid people think about affordability.

Max Dubler 🏳️‍🌈@maxdubler.com

This article quotes a sociology professor, a lawyer, and an architect who reject the notion that housing prices are set by supply and demand and do not believe that building more homes reduces rents. It is the housing policy equivalent of platforming climate change deniers.

sacbee.comWould building housing reduce poverty in California? Sacramento-area experts splitA recent study argued that building housing could lower state’s poverty rate, currently the second-highest in the U.S. behind Louisiana.

A Jay Holmgren reposted

Genevieve Kanter
@profgenkanter.bsky.social

Patient and consumer reps on FDA advisory committees are supposed to represent the public, but do they? In our new paper in @healthaffairs.bsky.social Scholar (link in comments), we looked at the orgs that patient and consumer reps were actively involved with and tracked down their pharma donors.

A Jay Holmgren
@ajh.bsky.social

This is why I just automatically delete all of those "you're invited to join the Prestigious Honor Society" emails - I would never be part of a club that would have me as a member!

Max Kozlov@maxkozlov.bsky.social

After our investigation into sham honor societies, many universities quietly deleted posts congratulating their faculty on joining them. Others... are still posting new ones! @stonybrooku.bsky.social news.stonybrook.edu/newsroom/pre...

news.stonybrook.eduStony Brook University Professors Inducted into European Academy of Engineering - SBU NewsStony Brook University professors make up two of 14 members elected internationally STONY BROOK, NY – October 1, 2026 – Two professors from the State University of New York at Stony Brook (Stony Brook...
A Jay Holmgren
@ajh.bsky.social

Even conceptually, they're using data from a decade later - is that a replication, or is that testing whether the effect is durable over time with a different set of people? Social science claims aren't like physics, "migration can change voting patterns" is context-specific!

A Jay Holmgren
@ajh.bsky.social

Deviating from your own pre-registration with no explanation to label the original paper as "Not Replicable" based on an under-powered directionally consistent result in a different context is a stretch.

A Jay Holmgren
@ajh.bsky.social

Their pre-registration says they'll use n=848, and they therefore don't do a power analysis. But then in the actual replication they're down to n=224. They then say "This doesn't replicate" because p=0.19. But its directionally similar and they have one-quarter of the sample!

A Jay Holmgren reposted

lastpositivist.bsky.social
@lastpositivist.bsky.social

It's genuinely fascinating that the US president keeps announcing that he just flat out doesn't know what trade is, he legit doesn't appreciate the concept of exchange and instead views economic activity through the lens of a dragon attempting to maximise its gold hoard to sit upon.

Michael McKee@mckonomy.bsky.social

Well, that took longer than I thought, but he hath spoken:

A Jay Holmgren reposted

Robert Fruchtman
@fruchtose.bsky.social

Replying to @fruchtose.bsky.social

If the Castro District became a landmark district, that means that a big chunk of land close to transit and specifically zoned for buildings up to 160' tall would no longer be eligible for HC-SF. Mandelman is effectively proposing to downzone the Castro District, especially for affordable housing.

A Jay Holmgren
@ajh.bsky.social

infuriating when "orthogonal" got broadly popular, that belongs to Persons Experiencing Doctoral Studies

Jake Anbinder@jakeanbinder.bsky.social

I don't think this is unique to the left or originated in 2020; there is a stunning amount of social science speak that has infiltrated everyday conversation now

A Jay Holmgren
@ajh.bsky.social

I find this weird guilt-by-familial association re: the Lakers deal very strange. Josh Kushner co-founded a company built on the ACA, publicly stated that he is a Democrat, marched in the Women's March and March for Our Lives (and donated $50k to the latter), and gave $250k to a Dem PAC for 2024.

A Jay Holmgren
@ajh.bsky.social

my hot take on the subject du jour is that the faculty experience at bates college and my faculty experience at UCSF, a large urban medical school, seem almost nothing alike.

A Jay Holmgren
@ajh.bsky.social

i know very little about the world of elite SLACs and i'm sure they're filled with very nice old buildings but to me they seem built for and staffed by the already very wealthy and i kind of laughed at the sub-title about "elite academia." the stakes here are almost impossibly low!

A Jay Holmgren reposted

S. Trent Rosenbloom
@trentrosenbloom.bsky.social

Replying to @annualreviews.bsky.social

Excited about our article covering the Future of EHRs (building on a review of the past) www.annualreviews.org/content/jour... @ajh.bsky.social

annualreviews.orgBuilding Toward a Future for Electronic Health Record SystemsHealth information technology tools, including electronic health records, are ubiquitous in healthcare across the United States. Despite the promise and opportunity of these tools, their benefits have...

A Jay Holmgren reposted

AEA Journals
@aeajournals.bsky.social

Forthcoming in the AER: "Difference-in-Differences with a Continuous Treatment" by Brantly Callaway, Andrew Goodman-Bacon, and Pedro H. C. Sant'Anna.

aeaweb.orgDifference-in-Differences with a Continuous Treatment(Forthcoming Article) - This paper analyzes difference-in-differences designs with a continuous treatment. We show that treatment-on-the-treated-type parameters are identified under a parallel trends assumption analogous to the binary treatment case. However, comparing these parameters across treatments is challenging because parallel trends does not rule out selection bias. We discuss alternative, typically stronger, assumptions that eliminate selection bias. We further show that popular two-way fixed effects estimands admit multiple interpretations, depending on the underlying causal building block, all having important limitations as meaningful summaries of treatment effects. Finally, we introduce alternative estimation procedures that avoid these drawbacks and demonstrate them in an empirical application.
A Jay Holmgren
@ajh.bsky.social

Bit of an experimental addition here: I created a companion web app to this paper that lets you interact with the distributions here across different physician cFTEs in each specialty bucket. www.ajholmgren.com/pmar-calcula...

ajholmgren.comPMAR Calculator — A Jay Holmgren, PhD
A Jay Holmgren@ajh.bsky.social

New paper out today in Health Affairs: there has been a lot of attention on the increase in patient-initiated messages since 2020, but how are those messages distributed across physicians and patients? We use national + local EHR data to find out! www.healthaffairs.org/doi/10.1377/...

healthaffairs.orgThe Uneven Distribution Of Patient Portal Messages Across Patients And Physicians | Health Affairs JournalResponding to patient-initiated portal messages is a growing source of work for physicians, yet little is known about how these messages are distributed across patients and physicians. Understanding t...
A Jay Holmgren
@ajh.bsky.social

This is one area where we can get cool improvements from AI in research. It was trivial for me to get the distributional estimates in Stata, and Claude turned it into an interactive web app in minutes. Research can move from standalone PDFs to things you can really interact with!

A Jay Holmgren
@ajh.bsky.social

Increased demand and stagnant supply for care is here. Finding ways to incorporate asynchronous care into the ecosystem is important to free up synchronous visits. That probably requires a better reimbursement mechanism, and we should experiment with new models here.

A Jay Holmgren
@ajh.bsky.social

Patient portal payment is a canary in the coal mine. New technology will continue to allow ambulatory care to expand from discrete, episodic visits to more continuous, high-frequency, low-intensity interactions, both as a complement and substitute. Payment policy shouldn't hold back new care models.

A Jay Holmgren
@ajh.bsky.social

Physicians with lower weekly visit volume saw more messages per visit - building on our evidence that EHR and asynchronous care burden has a relatively greater impact on low-cFTE physicians.

A Jay Holmgren
@ajh.bsky.social

All of this matters for payment design: existing CPT codes are clunky and require too much admin time relative to the reimbursement, but fully capitated models are likely to over or under-reimburse many physicians. Automated billing, bulk billing, or partial capitation could be solutions to explore.

A Jay Holmgren
@ajh.bsky.social

On the patient side, similar to pre-COVID studies, we found a small number of patients send the majority of messages - 52.8% of messages were sent by just 5% of patients! Physician receipt of messages is similarly skewed: PCPs have a median of 19/week, but the top quartile averaged 53.3/week!

A Jay Holmgren reposted

Joe Fish
@sadbusdriver.bsky.social

i feel the loss of econ twitter with stuff like the quality of consumer goods discourse the literature on reputation incenvtives is very rich; it covers a lot of interesting stuff (why do bad mechanics not get competed, why do firms harvest reputation). alas, nobody here does those kinds of threads

Joe Fish@sadbusdriver.bsky.social

I'm curious on what the econ lit is on this. intuitively, it seems like a few things could generate the "great price and good-enough quality" equilibrium 1. cost of quality is convex and willingness to pay for it is not 2. information frictions force competition to be mostly on prices, not quality

A Jay Holmgren
@ajh.bsky.social

Just described an unexpected performance meltdown in a video game as “Richie Tenenbaum-esque” to which she just stared at me because I somehow married a woman who doesn’t like Wes Anderson

A Jay Holmgren
@ajh.bsky.social

This is a really gross thing to do given she lost to Daniel Biss, who just happens to be Jewish, and not the AIPAC-backed candidate Laura Fine.

Kat Abughazaleh@katmabu.bsky.social

AIPAC may have bought our election in Illinois, but we're not going to let that happen in Michigan. 5 days left to make them panic. Vote for Abdul on or before August 4th.

A Jay Holmgren
@ajh.bsky.social

I lived in SF for 5 years and basically had no contact with the tech industry outside of the occasional professional interaction and overheard conversations. Cities are big! You can find a lot of sub-cultures!

William B. Fuckley@opinionhaver.bsky.social

it's really pathetic how, for many people, much intra-urban 'politics' in places like the Bay Area seems to boil down to 'I think these people are lame'. then don't hang out with them! no one is forcing you at gunpoint to go to their fucking happy hours.

A Jay Holmgren
@ajh.bsky.social

I was forced to interact with the elderly dog owners of the Castro frequently at the Eureka Valley dog park, though, and to Lucy’s mom: train your crazy dog why did you even get a Malinois bigger than you we are mortal enemies.

A Jay Holmgren
@ajh.bsky.social

One reason I'm skeptical of these big claims is that, if true, we basically subject the western half of many time zones to something much worse. The difference between sunrise times in eastern Maine and west Michigan is 1.5 hours!

Scott McGrath@smcgrath.phd

🧪 Only 12% of Americans want to keep seasonal clock-switching. Permanent daylight saving time sounds great, but circadian scientists warn against it. Later sunrises disrupt the morning light signals our bodies rely on to regulate metabolic health.

nytimes.comShould We Make Daylight Saving Time Permanent? (Gift Article)More light at the day’s end comes at a price: much later winter sunrises across much of the U.S. Scientists have documented consequences for public health and the economy.
A Jay Holmgren
@ajh.bsky.social

my guess based on having read a random sample of observational studies, probably-confounded-geographic-RDDs, etc. is that "stop switching" has some evidence, the direction of switching (permanent standard or daylight time) is whatever. but i'd prefer permanent DST because I'm an adult.