It is important to understand that these are all lies. Deans don't tell applicants why they don't get a job that they applied for. The job market doesn't react to PhD defenses with phone calls for interviews. This is all false. These people are liars, and they are lying to you to start a race panic.

Now that I'm off ElonTwitterX, I realized that I had so many good jokes on there lost to the ether. In short: I'm sorry, my political science comedy albums are going to really delayed.

When I first heard the term "LLM" I was terrified it was yet another regression model that I needed to pretend to know about, whew thank god it was not that.

I am the face of Middlebury faculty research! Oh no.

"About 243,000 more children under 5 years have died or will die this year than in 2024" due mostly to DOGE's illegal dismantling of USAID

A picture is worth a 1000 words. The birth dose of hepatitis B vaccines in the USA eliminated perinatally acquired hepatitis B infection from over 20,000 cases a year before the vaccine was introduced to <20 cases a year. The gaslighting by the current members ACIP is obscene, unethical and cruel.

New idea to measure academic impact: how many books of yours were illegally downloaded and used to train LLMs by Anthropic.

Just want to mention two articles of mine since we are talking about the hilariously-named "Credibility Revolution." 1. "Randomized Controlled History?" It explores design-based inference approaches to history and finds them mostly problematic. (www.nowpublishers.com/article/Deta...)

This is an interesting paper. I continue to sense that we as a field have concluded, either you have an identification strategy as defined here, or you don’t have a valid causal claim. And that is really not true.

Just want to mention two articles of mine since we are talking about the hilariously-named "Credibility Revolution." 1. "Randomized Controlled History?" It explores design-based inference approaches to history and finds them mostly problematic. (www.nowpublishers.com/article/Deta...)

Amazing letter in The Times (also an unbelievable play fwiw)

when someone tells me not to use em dashes because AI uses em dashes

www.nytimes.com/2025/12/01/b... Corporations: wait a minute, if everyone is poor, who is going to buy pizza?

FYI: law and science, political science, and sociology all still have DDRIG programs run outside of the NSF but funded by NSF. The rest of these programs unfortunately do not. I run the political science DDRIG.

Public man-made death: The dismantling of USAID “has already caused the deaths of six hundred thousand people, two-thirds of them children.” That’s 1000+ excess dead children every day. Every day. hsph.harvard.edu/news/usaid-s...

This is the type of hater energy that will fuel us in the years ahead.

I have been thinking about writing this paper for over a decade: my definition and conceptualization of "religion." No idea if any of this makes sense. Feedback welcome. Enjoy! t.co/ykdwhEAEsK

this is a metaphor for waking up and logging on

I think the main problem with existing definitions of religion is that they also work for other items like "magic" or "astrology" or "Alcoholics Anonymous." Read my paper that tries to differentiate religion from other "supernaturalisms." papers.ssrn.com/abstract=577...

Study after study shows that using LLMs is bad for cognition, bad for learning, bad for understanding, bad for mental health. So why are our schools and universities still relentlessly pushing them?

I have been thinking about writing this paper for over a decade: my definition and conceptualization of "religion." No idea if any of this makes sense. Feedback welcome. Enjoy! t.co/ykdwhEAEsK

Ouch! Feeling kind of old as one of that “earlier generation” that engaged in “shopworn” debates :) Great review of important recent work on qualitative methods! I would add Fairfield and Charman on formal Bayesian process tracing as one of the key recent contributions.

Check out my review of some exciting new books in qualitative methods!

Today's conundrum: "The proper concepts are needed to formulate a good theory, but we need a good theory to arrive at the proper concepts." - Abraham Kaplan, Conduct of Inquiry.

Are you interested in the study of religion but tired of reading about Christianity? Boy do I have the article for you! Now under review. papers.ssrn.com/sol3/papers....

I hope one day the Wilson Center, where I worked last summer, will be back but I fear it will not. thehill.com/opinion/inte...

🧵 Today's revocation of Harvard's ability to enroll international is awful for many reasons, chief among them the human cost as @jeremywallace.bsky.social notes. It's also an attack on the US economy. But maybe less obvious: it is terrible national security policy. It makes Americans less safe. 1/

Just reposting this, for no reason at all

Some good news for secularism in America: www.nytimes.com/2025/05/22/u...

Truly an incredible post. Calling on economists, political scientists and sociologists working on corruption to please share your favourite corruption papers so we can help the NYT get up to speed on concepts and measurements in this field.

Was trying to do one of those revelatory "Failure CVs" but it basically just looked like my actual CV, I don't understand this exercise?

Happy May Day!

Happy to announce that the Fulbright Program still exists, and that I won a Fulbright Flex to travel to India for fieldwork over the next several years. The program is open for next year too, so think about applying!

One piece of advice about my research pipeline that I've always been given is: have multiple irons in the fire. But I really work best doing one thing at a time, often for weeks or months at a time. So maybe this advice isn't for everyone.

Happy first birthday to my article submission to [will potentially reveal journal name in the future to shame them]! A whole year, my baby is growing up.

New seminar this fall:

I'm not using social media much these days but I'm breaking the hiatus to say RIP David Lynch, one of my favorite directors. A true visionary, there was no one else like him.

New paper on how to conceptualize and measure Hindu religiosity, been working on this one for several years. Feedback welcome! papers.ssrn.com/sol3/papers....

Revised and resubmitted -- an updated version of my paper on studying history like an experiment: "Randomized Controlled History?" papers.ssrn.com/sol3/papers....

Trying to thread that needle of raising my daughters to question male authority -- except for me, right now.

Solo parenting an almost-one year old tonight, thinking of watching Rosemary's Baby with her. Good idea?

I know I'm supposed to be scared of AI, but when my phone has no autocorrect suggestions for the mysterious word "Frday," I feel a bit better.

[During a timeout] Four-year old: "Do you still love me?" Me: "Yes, even when I give you a timeout I still love you." Four-year old: "And even when I punch you, I still love you."

Stunned to learn that "two-way fixed effects" is not a band but some kind of regression thingie.

Masks are off in Miami-Dade. Masks are in the trash can. There are no more masks.

Election forecasters 15 years ago: "Punditry is stupid. We need data-driven scientific analysis!" Election forecasters now: "Our model shows it will be either a slim GOP victory or a slim Dem victory, but a Dem blowout is also possible. My vibes tell me..."

Happy pitchers and catchers report day!