Jonathan Horowitz
@jonathanhorowi1.bsky.social
Sociology sometimes. Goal is to be the most boring poster on the internet. I can't believe I have to say this, but these views are mine alone.
SUVs are an example of "positional goods." If you have an SUV, you can see out over all the other cars. Unless everyone else has an SUV, in which case you need a bigger SUV. The good provides a literal position, as long as no one else is aiming for the same physical location.
![4-panel vertical comic. (1) 100 Years Ago [two people standing next to bicycle with small car nearby] PERSON 1: It’s too dangerous riding a bike with these cars around. I should get a car, too. (2) 50 Years Ago [two people between smaller car and bigger car] PERSON 2 with short hair: Small cars are less safe in collisions with larger vehicles, so I should get a bigger one. (3) Today [two people between big car and even bigger car] PERSON 1: Everyone has huge SUVs now. If I don’t get the biggest one, I’m putting my family at risk. (4) Soon [two people next to large armored car with spiked clubs attached] PERSON 2: If I don’t install more whirling spike clubs, I’ll be destroyed by all the other drivers who...](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:cz73r7iyiqn26upot4jtjdhk/bafkreiby6can6k76yzfmv2czxs4pu2l7zie7bqbrpswi7h3ifmiadkuqvu@jpeg)
November 14, 2025 at 10:45 PM
SUVs are an example of "positional goods." If you have an SUV, you can see out over all the other cars. Unless everyone else has an SUV, in which case you need a bigger SUV. The good provides a literal position, as long as no one else is aiming for the same physical location.
Reposted by Jonathan Horowitz
Something absolutely perfect about Larry Summers riffing about the scourge of woke cancel culture being unfair to predatory men in a friendly email to his pal JEFFREY EPSTEIN.
November 13, 2025 at 12:25 AM
Something absolutely perfect about Larry Summers riffing about the scourge of woke cancel culture being unfair to predatory men in a friendly email to his pal JEFFREY EPSTEIN.
I've been struggling to articulate this point, but this does it well. If we care about omitted variable bias (or collider bias, or model fit, or theory...) then it makes no sense to run a regression on every possible set of variables.
Billions of regressions ≠ robustness.
My new PNAS Letter shows how unjustified models can drown out justified ones — and why thoughtful model selection matters👇
www.pnas.org/doi/10.1073/...
My new PNAS Letter shows how unjustified models can drown out justified ones — and why thoughtful model selection matters👇
www.pnas.org/doi/10.1073/...
November 11, 2025 at 7:12 PM
I've been struggling to articulate this point, but this does it well. If we care about omitted variable bias (or collider bias, or model fit, or theory...) then it makes no sense to run a regression on every possible set of variables.
It's really hard to take a big research university seriously that doesn't have a statistics program. The content is foundational to at least a dozen different other academic programs in Arts & Sciences alone. They're basically running up the white flag.
November 11, 2025 at 4:18 PM
It's really hard to take a big research university seriously that doesn't have a statistics program. The content is foundational to at least a dozen different other academic programs in Arts & Sciences alone. They're basically running up the white flag.
Reposted by Jonathan Horowitz
wait until you see what happens after the 3rd vote
November 8, 2025 at 3:30 PM
wait until you see what happens after the 3rd vote
I am not sure I have seen *any* movie more than seven times. But if I did, it was this one because my HS friend owned it and put it on whenever I came over.
November 2, 2025 at 10:19 PM
I am not sure I have seen *any* movie more than seven times. But if I did, it was this one because my HS friend owned it and put it on whenever I came over.
What we need is boosters for our sociology departments.
with the firing of Hugh Freeze at Auburn, US universities now owe nearly $200 million to ex-football coaches
the amount US universities are paying non-working football coaches will soon top $140 million, if you thought we were gonna cure cancer or something any time soon
November 2, 2025 at 10:10 PM
What we need is boosters for our sociology departments.
Went down a rabbit hole and found this embarrassment. This is not "AI Fluency," it is outsourcing the work of being a student.
Don't talk to me about teaching "AI Fluency" if everything on the list involves putting words into a chatbot and reading the result.
ai.osu.edu/learn-ai
Don't talk to me about teaching "AI Fluency" if everything on the list involves putting words into a chatbot and reading the result.
ai.osu.edu/learn-ai
Learn with AI | Artificial Intelligence at Ohio State
ai.osu.edu
October 24, 2025 at 3:34 PM
Went down a rabbit hole and found this embarrassment. This is not "AI Fluency," it is outsourcing the work of being a student.
Don't talk to me about teaching "AI Fluency" if everything on the list involves putting words into a chatbot and reading the result.
ai.osu.edu/learn-ai
Don't talk to me about teaching "AI Fluency" if everything on the list involves putting words into a chatbot and reading the result.
ai.osu.edu/learn-ai
Reposted by Jonathan Horowitz
Academia rewards originality, as long as it looks exactly like what came before.
October 24, 2025 at 10:27 AM
Academia rewards originality, as long as it looks exactly like what came before.
I had been looking for the right way to put this, but this sums it up well. It's currently harder to come up with threats to inference than to run a billion regressions, but the former is also a lot more valuable.
When there is a random way to do something, there is a less random way that is better but requires more thought. In this case, regression models that make no sense don't belong in a multiverse analysis. An inferential regression without a causal justification is like an opinion without reasons.
October 23, 2025 at 4:48 PM
I had been looking for the right way to put this, but this sums it up well. It's currently harder to come up with threats to inference than to run a billion regressions, but the former is also a lot more valuable.
Reposted by Jonathan Horowitz
When there is a random way to do something, there is a less random way that is better but requires more thought. In this case, regression models that make no sense don't belong in a multiverse analysis. An inferential regression without a causal justification is like an opinion without reasons.
October 23, 2025 at 4:34 PM
When there is a random way to do something, there is a less random way that is better but requires more thought. In this case, regression models that make no sense don't belong in a multiverse analysis. An inferential regression without a causal justification is like an opinion without reasons.
I am reacting badly to every part of this. Using proxy variables to solve an unsolveable mathematical problem? Trying to automate your way out of making hard theoretical choices? Calling period effects "time"?!
October 22, 2025 at 9:45 PM
I am reacting badly to every part of this. Using proxy variables to solve an unsolveable mathematical problem? Trying to automate your way out of making hard theoretical choices? Calling period effects "time"?!
Reposted by Jonathan Horowitz
Reposted by Jonathan Horowitz
FSU is hiring 2 APs in spatial data science! Open to any social science discipline, including sociology. This makes at least 5 faculty jobs at FSU this year open to Soc PhDs. Consider applying if you do spatial research! (ad says Geography, but it's open field)
jobs.omni.fsu.edu/psc/sprdhr_e...
jobs.omni.fsu.edu/psc/sprdhr_e...
Welcome | PS LINKS
jobs.omni.fsu.edu
October 22, 2025 at 2:46 PM
FSU is hiring 2 APs in spatial data science! Open to any social science discipline, including sociology. This makes at least 5 faculty jobs at FSU this year open to Soc PhDs. Consider applying if you do spatial research! (ad says Geography, but it's open field)
jobs.omni.fsu.edu/psc/sprdhr_e...
jobs.omni.fsu.edu/psc/sprdhr_e...
Reposted by Jonathan Horowitz
What attending a conference out of your field feels like
Them: everything’s a social network, even this conference!
Me: haha...
Them: someone should do a network analysis of network analysis
Me (thinking): surely my coffee isn’t a network
Them: take this cup, for example
Them: everything’s a social network, even this conference!
Me: haha...
Them: someone should do a network analysis of network analysis
Me (thinking): surely my coffee isn’t a network
Them: take this cup, for example
October 22, 2025 at 2:46 PM
What attending a conference out of your field feels like
Them: everything’s a social network, even this conference!
Me: haha...
Them: someone should do a network analysis of network analysis
Me (thinking): surely my coffee isn’t a network
Them: take this cup, for example
Them: everything’s a social network, even this conference!
Me: haha...
Them: someone should do a network analysis of network analysis
Me (thinking): surely my coffee isn’t a network
Them: take this cup, for example
I'm seeing a lot of people freaking out about "Sociology from six PhD students to zero", but what it means is sociology has six PhD slots over the next two years and they've decided to put them all in one year.
Wow. Harvard nuking its PhD programs
- Science PhD admissions reduced by more than 75%
- Arts & Humanities reduced by about 60%
- Social Sciences by 50–70%
- History by 60%
- Biology by 75%
- The German department will lose all PhD seats
- Sociology from six PhD students to zero
- Science PhD admissions reduced by more than 75%
- Arts & Humanities reduced by about 60%
- Social Sciences by 50–70%
- History by 60%
- Biology by 75%
- The German department will lose all PhD seats
- Sociology from six PhD students to zero
Harvard FAS Cuts Ph.D. Seats By More Than Half Across Next Two Admissions Cycles | News | The Harvard Crimson
The Faculty of Arts and Sciences slashed the number of Ph.D. student admissions slots for the Science division by more than 75 percent and for the Arts & Humanities division by about 60 percent for th...
www.thecrimson.com
October 21, 2025 at 8:51 PM
I'm seeing a lot of people freaking out about "Sociology from six PhD students to zero", but what it means is sociology has six PhD slots over the next two years and they've decided to put them all in one year.
Reposted by Jonathan Horowitz
Is childhood exposure to local wealth inequality associated with upward income mobility achieved in adulthood? Yes! Check out my new paper, just published in @natureportfolio.nature.com here: doi.org/10.1038/s414... #EconSky #Sociology #Demography
October 15, 2025 at 12:11 PM
Is childhood exposure to local wealth inequality associated with upward income mobility achieved in adulthood? Yes! Check out my new paper, just published in @natureportfolio.nature.com here: doi.org/10.1038/s414... #EconSky #Sociology #Demography
This is what I think we're missing. There's going to be another big round of automation, and it will be much more useful than whatever general-use LLMs can do.
Most robots will not be humanoid. The future of robots is an automated logistics and distribution system that can bring stuff from a container ship to your house, all with minimal intervention; and the robots involved are often huge and inhuman.
itcanthink.substack.com/p/how-robots...
itcanthink.substack.com/p/how-robots...
How Robots Bring You Your Stuff
Robots in shipping and logistics are basically the opposite of the generalizable robotics that we like to talk about and which capture the imagination
itcanthink.substack.com
October 20, 2025 at 3:59 PM
This is what I think we're missing. There's going to be another big round of automation, and it will be much more useful than whatever general-use LLMs can do.
Reposted by Jonathan Horowitz
I'm starting to explore {simDAG}, a nifty #rstats package that has a nice API for specifying DAGs and then simulating data from them: robindenz1.github.io/simDAG/
Simulate Data from a DAG and Associated Node Information
Simulate complex data from a given directed acyclic graph and information about each individual node. Root nodes are simply sampled from the specified distribution. Child Nodes are simulated according...
robindenz1.github.io
October 13, 2025 at 4:14 PM
I'm starting to explore {simDAG}, a nifty #rstats package that has a nice API for specifying DAGs and then simulating data from them: robindenz1.github.io/simDAG/
Reposted by Jonathan Horowitz
New article out in @sociusjournal.bsky.social.
It shows how closely linked motherhood penalties 🤰📉 and gender inequalities 👨💼💰👩💼 are by studying many local labour markets.
Thread 👇
journals.sagepub.com/doi/10.1177/...
It shows how closely linked motherhood penalties 🤰📉 and gender inequalities 👨💼💰👩💼 are by studying many local labour markets.
Thread 👇
journals.sagepub.com/doi/10.1177/...
October 13, 2025 at 4:37 PM
New article out in @sociusjournal.bsky.social.
It shows how closely linked motherhood penalties 🤰📉 and gender inequalities 👨💼💰👩💼 are by studying many local labour markets.
Thread 👇
journals.sagepub.com/doi/10.1177/...
It shows how closely linked motherhood penalties 🤰📉 and gender inequalities 👨💼💰👩💼 are by studying many local labour markets.
Thread 👇
journals.sagepub.com/doi/10.1177/...
I don't have any clue whether he will win his lawsuit or not because I am not a lawyer, but they're not even doing a good job pretending he violated any policies. The logic they apply here works equally well for writing a nasty review of the latest Taylor Swift album.
I just learned, Thomas Alter, tenured professor, has been terminated effective immediately. Shame on Texas State! We mst build support for Tom. spectrumlocalnews.com/tx/south-tex...
Texas State upholds the firing of associate history professor Thomas Alter
Alter’s termination occurred on Sept. 10, stemming from his remarks at the Revolutionary Socialism Conference on Sept. 7.
spectrumlocalnews.com
October 14, 2025 at 8:03 PM
I don't have any clue whether he will win his lawsuit or not because I am not a lawyer, but they're not even doing a good job pretending he violated any policies. The logic they apply here works equally well for writing a nasty review of the latest Taylor Swift album.
There was a period of my life when I wanted to be just like Hahrie Han. I still kind of do. This is richly deserved.
October 8, 2025 at 7:56 PM
There was a period of my life when I wanted to be just like Hahrie Han. I still kind of do. This is richly deserved.
Reposted by Jonathan Horowitz
working on a new unified theory of american reality i'm calling "everyone is twelve now"
September 8, 2025 at 12:37 AM
working on a new unified theory of american reality i'm calling "everyone is twelve now"
Reposted by Jonathan Horowitz
The guys who automated petard production: OMG! You're never going to guess what we've been hoisted by!
September 9, 2025 at 10:59 AM
The guys who automated petard production: OMG! You're never going to guess what we've been hoisted by!
I like how this post is framed so that never having sex leads you to do things like become more educated, and make you less likely to drink alcohol. Clearly this is all very causal.
A study of over 500,000 Brits and Australians finds that people who never have sex are more educated, less likely to use alcohol and smoke, more nervous, lonelier, and unhappier. Regions with high income inequality had more sexless residents. In PNAS: www.pnas.org/doi/abs/10.1...
October 6, 2025 at 8:23 PM
I like how this post is framed so that never having sex leads you to do things like become more educated, and make you less likely to drink alcohol. Clearly this is all very causal.