Seema Jayachandran

We're hiring for a short-term (mid-Sept to mid-Dec) role as a field RA on an RCT about rural healthcare in northern Uganda. Must have fieldwork experience and ideally experience in Uganda/East Africa. Please share with anyone who might be interested.

Holy - and I cannot emphasize this enough - shit.

Also, this guy was a senior project 2025 organizer.

I've been pleasantly surprised by AI tools before (Claude on conceptual/technical points), but this is the first time I've been kind of flabbergasted, in a good way.

I just tried out Refine on a paper of mine and found the comments very helpful.

Coming to the Royal Sonesta for NBER SI never gets old. Even after... three centuries.

Stunning charts show how disproportionately the Roberts Court favors the Trump regime.

"This pattern exemplifies what scholars call 'autocratic legalism'—the systematic weaponization of legal institutions to consolidate power while maintaining the appearance of legal process."

Niche question: Academics doing applied metrics trainings as a consulting gig isn't something I recall from 10+ yrs ago. Is this new or my awareness is new? If it's new, are there more rapid advances now? More awareness of advances b/c of social media? Data science in the private sector who can pay?

I am so conflicted.

Such a stupid timeline right now: Epstein, Epstein, Epstein

But if this breaks the fever, hot damn.

I was explaining to my Ukrainian colleague the phrase ‘There’s no such thing as a free lunch’. She told me the equivalent in Ukrainian is ‘The only free cheese is in the mousetrap’ - which is so much better

I have a new review article w/ Alessandra Voena on women's power in the household in LMICs (just submitted to JEL).

I've written ~6 review articles, and I think this is the best one. We make some useful conceptual points IMO. But you decide! Comments welcome!

seemajayachandran.com/womens_power...

The rhetoric around cash transfers seems to conflate two things:

1. The studies powered and designed to detect GE effects have studied cash transfers. They find important positive spillovers.

2. An advantage of cash transfers over other interventions is that they have these positive spillovers.

Homeland Security DHS account on X: Remember your Homeland's Heritage. New Life in a New Land - Morgan Weistling

The official DHS account is now tweeting about “your homeland’s heritage”

Killing MIT seems dumb to me

There are many annoying defaults in Windows, but one of the worst is to open pdfs in Edge.

I know this is beside the point (which is the NYT bias and ethics) but it's also an inane insinuation given that Mamdani was a faculty kid. He tried to fool the Columbia admissions office, who were familiar with his parents, by checking a box on his application??

All of that parsing of “seriously but not literally” or “literally but not seriously”. Right now, it’s both, even for the crazy alligator shit.

They want the trappings of power (comfort), not using their power (discomfort)

May is still miserable in Chicago. You mean mid-June to mid-Oct.

Ha, don't be impressed by my probably bad time allocation choices.

I have a new review article w/ Alessandra Voena on women's power in the household in LMICs (just submitted to JEL).

I've written ~6 review articles, and I think this is the best one. We make some useful conceptual points IMO. But you decide! Comments welcome!

seemajayachandran.com/womens_power...

Hypothesis: Bullies get cited more, ceteris paribus

If the GOP reconciliation bill passes, ICE gets through FY2029:

- $45 billion for detention, on top of the current annual budget of $3.4 billion
- $14.4 billion for transportation and removal, on top of the current annual budget of $750 million
- $8 billion for hiring/retention
- Billions more.

The meta-analysis point applies not just to components of an index. Michael Kremer made the ex post obvious good point that a study that is not powered to detect the effect of (say) clean water on child mortality can/should measure and report it for use in meta-analyses.

I appreciate the multiple hypothesis testing reason for indices. But think of a meta-analysis that can't use the study b/c disaggregated results aren't shown.

I'd be fine if people just shared this post-publication.

(My use case: an index that combines women's gender attitudes and employment.)

bsky.app/profile/seem...

I'd be fine if people didn't include it in the version submitted to journals and later just have a link on their website with the results for index components. I agree their is a psychic cost of a bloated appendix at the refereeing/editing stage.

When studies report results on indices that combine many vars, they should also show results for each var in the appendix. This is prompted by writing a review article; some studies have an index where only a subset of the vars are on topic for me. Plus the original vars have interpretable units.

This. Is. Insane.
Read this, please.

That not being obsequious to your senior colleagues when you are junior can cost you

Something that Claude can do far better than me is read the documentation for "gender sensitization" programs (or similar part-skills, part-norm-change programs) and state what the distinct main mechanisms are the theory of change in a useful way. Most of those program guides make my head spin.