Sign in

Jason Kerwin

@jkerwin.bsky.social
2.7K followers 1.2K following 386 posts

Development Economist @ UW. Hawaiʻi born, Stanford alum, Michigan PhD. Views my own. My website: jasonkerwin.com

PostsRepliesMedia
Jason Kerwin @jkerwin.bsky.social · 04/09/2026
Obviously not our original idea, but solid empirical evidence on this is surprisingly rare, which likely contributes to the whole “cueing” disaster
020
Jason Kerwin @jkerwin.bsky.social · 04/09/2026
The specific skills that cumulate are decoding (phonics) skills, and they unlock actual word recognition!
020
Jason Kerwin @jkerwin.bsky.social · 03/09/2026
Also tagging my coauthors @eander462.bsky.social, @simongraffy.bsky.social, and Monica Lambon-Quayefio.
010
Jason Kerwin @jkerwin.bsky.social · 03/09/2026
You can find out more about the intervention (Inspiring Reading, which is part of the broader TFLI program) here: inspiringteachers.org/faq/ This work is a collaboration with the amazing people at @InspirngTeachrs: inspiringteachers.org
inspiringteachers.org
Inspiring Teachers | FAQ
Answers to a range of common questions in relation to Inspiring Teachers coaching programs and Fellowships.
110
Jason Kerwin @jkerwin.bsky.social · 03/09/2026
You can read the full paper here: documents1.worldbank.org/curated/en/0... These are just our grade-1 impacts—we will have more results from the program soon! And we are continuing to study its expansion.
documents1.worldbank.org
120
Jason Kerwin @jkerwin.bsky.social · 03/09/2026
One of our key findings is that skills cumulate, with progress on basic skills making it possible to learn more-advanced ones. This means that a combination of structured pedagogy and targeted instruction (most-famously known as TaRL) will work better than either alone.
230
Jason Kerwin @jkerwin.bsky.social · 03/09/2026
How do kids actually learn to read? We tackle this question in a new paper that studies a program that tripled (!!) reading gains in Ghanaian 1st-grade classrooms. Its impacts match a model of cumulative reading ability, which we will assess with future waves of data.
23311
Reposted by Jason Kerwin
Max Woolf @minimaxir.bsky.social · 07/05/2026
Reddit post: In "The Odyssey" (2026), Odysseus speaks with a Boston accent, despite being from Ithaca, which is in New York.
135103971876
Jason Kerwin @jkerwin.bsky.social · 24/04/2026
Full announcement: www.nber.org/news/nber-ap... Program from the Spring meeting: www.nber.org/conferences/...
nber.org
021
Jason Kerwin @jkerwin.bsky.social · 24/04/2026
Thrilled to be joining @nber.org as a Research Associate in the Economics of Education! I got to present our new work on reading in Ghana at the spring program meeting yesterday—amazing audience & questions as always.
3240
Jason Kerwin @jkerwin.bsky.social · 24/04/2026
My understanding is that these are reporting guidelines rather than standards for analysis plans
100
Jason Kerwin @jkerwin.bsky.social · 27/03/2026
The AEA trial registry is basically uniformly used in economics and has a fixed set of fields. The analysis plans people write are a free for all.
110
Jason Kerwin @jkerwin.bsky.social · 27/03/2026
Did you learn to speak French when you lived there?
100
Jason Kerwin @jkerwin.bsky.social · 24/03/2026
In development econ we pre-specify the sample size in the trial registration form
100
Jason Kerwin @jkerwin.bsky.social · 24/03/2026
Mostly obstinacy :) I can now understand French fairly well and this helps to maintain that skill. It’s also how I learned French; my only other investments in the language have been marrying a native speaker (who speaks it sometimes) and brief trips to Haiti and France.
110
Jason Kerwin @jkerwin.bsky.social · 24/03/2026
My Duolingo streak is now a decade old!
1160
Jason Kerwin @jkerwin.bsky.social · 24/03/2026
Do people file pre-analysis plans for this sort of study? Where do they post them? The AEA trial registry only accepts randomized experiments.
200
Jason Kerwin @jkerwin.bsky.social · 24/03/2026
This is an important issue that I’ve never seen addressed in an actual PAP, even the very long-winded ones
110
Jason Kerwin @jkerwin.bsky.social · 23/03/2026
I’m clear on the conceptual issue here, but have you seen this happen? If you aren’t comfortable posting publicly about a study with this issue, can you DM one to me?
000
Jason Kerwin @jkerwin.bsky.social · 23/03/2026
That all sounds like content for a paper rather than a document about the plan for analyzing the data
100
Jason Kerwin @jkerwin.bsky.social · 23/03/2026
What do you mean by conceptualization? And how would that lead to incorrect p-values?
100
Jason Kerwin @jkerwin.bsky.social · 23/03/2026
Can you give an example of an RCT where this was an issue?
200
Jason Kerwin @jkerwin.bsky.social · 22/03/2026
I think it's all costly signaling, but your view may be right
131
Jason Kerwin @jkerwin.bsky.social · 22/03/2026
This is great lol Did you have AI do it?
110
Jason Kerwin @jkerwin.bsky.social · 22/03/2026
I agree but to be clear nobody is doing that right now.
110
Jason Kerwin @jkerwin.bsky.social · 22/03/2026
Is any of the rest of that necessary for the analysis plan? It needs to be in the paper, but the point of the PAP is to tie your hands on the analysis.
100
Jason Kerwin @jkerwin.bsky.social · 22/03/2026
Link to post: jasonkerwin.com/nonparibus/2...
jasonkerwin.com
Your pre-analysis plan should contain the plan for analyzing your data, and nothing else - Jason Kerwin
I am currently reviewing a paper where the authors filed a lengthy analysis plan that does not actually say how they plan to analyze their data. This is an endemic problem in development economics RCT...
000
Jason Kerwin @jkerwin.bsky.social · 22/03/2026
Dear development economists: stop writing a ton of extra stuff in your PAP, and just tell me what analyses you are going to run
5358
Jason Kerwin @jkerwin.bsky.social · 12/03/2026
Link to the full post: blogs.worldbank.org/en/impacteva...
blogs.worldbank.org
Six Questions with Monica Lambon-Quayefio
010
Jason Kerwin @jkerwin.bsky.social · 12/03/2026
Development Impact is featuring my fantastic coauthor Monica Lambon-Quayefio! Amazing to see all the cool stuff she is working on.
151
Reposted by Jason Kerwin
Johannes Haushofer @johanneshaushofer.com · 18/02/2026
Team "Elsevier Highlights Subversion" strikes again! Best one yet! @deankarlan.bsky.social, Monica Lambon-Quayefio, Utsav Manjeer, @christopher-udry.bsky.social
24218
Jason Kerwin @jkerwin.bsky.social · 18/02/2026
Happy birthday to @audreydorelien.bsky.social, the most beautiful & amazing woman I’ve ever met.
2260
Jason Kerwin @jkerwin.bsky.social · 20/12/2025
I don’t totally believe that but they’re clearly better than the alternatives
010
Jason Kerwin @jkerwin.bsky.social · 20/12/2025
Yes, a lower bar for significance is common in A/B tests. I was surprised too! But it makes some sense if the goal is to decide quickly what to invest more in
110
Jason Kerwin @jkerwin.bsky.social · 17/12/2025
Also over on Twitter ("X") we are in day two of a debate about the value of RCTs in education research. Lots of value over at the bad place! That's where the people who disagree with me are.
0102
Jason Kerwin @jkerwin.bsky.social · 17/12/2025
The effects on quality are already about 1/4 as big as those of the main intervention we are studying (results coming pretty soon!). So the potential for improving learning is quite high.
000
Jason Kerwin @jkerwin.bsky.social · 17/12/2025
In an A/B test we found that training principals to coach teachers in their schools improved teaching quality by 24 percent! These results are still tentative—but good enough for our implementing partner to keep running this intervention.
Graph of treatment effects (text on graph shows the numbers)
3141
Reposted by Jason Kerwin
Joshua Goodman @joshua-goodman.com · 11/06/2025
Re-upping my advice about how to write a good title and abstract for an academic paper, appropriately called: "How to Write a Title and Abstract" Feel free to share this thread, which will focus on titles. #EconSky #AcademicSky
1022172
Jason Kerwin @jkerwin.bsky.social · 03/06/2025
A fun side note: the seminar is named in honor of the late Paul Theodore Heyne, who was a long-time lecturer at UW. He was one of the undergraduate advisors for Jeff Smith, who was one of my own Ph.D. advisors. Jeff now holds the Paul T. Heyne Distinguished Chair in Economics at (the other) UW.
000
Jason Kerwin @jkerwin.bsky.social · 03/06/2025
Here's a brief blog post about the talk: jasonkerwin.com/nonparibus/2... And here are the slides: jasonkerwin.com/Slides/Paul_... Comments welcome!
jasonkerwin.com
Why USAID is Great for America and the World - Jason Kerwin
Last Thursday I was invited by UW’s Economics Undergraduate Board to give this quarter’s Paul Heyne Seminar—a public lecture about economics aimed at a general audience. My talk was about why USAID is...
200
Jason Kerwin @jkerwin.bsky.social · 03/06/2025
Last Thursday I gave this quarter's Paul Heyne Seminar. My talk was about why USAID is great for America and also the world, why we should keep it, and how we can make it even better. A key point is that it is the best brand in the industry, by far.
1252
Jason Kerwin @jkerwin.bsky.social · 07/05/2025
The highly-touted flowchart revolution didn’t really pan out. Turns out that if you run a garbage can regression with a DAG you’re still just doing a garbage can regression.
010
Jason Kerwin @jkerwin.bsky.social · 07/05/2025
Hey remember when everyone in econ was doing flowcharts (“directed acyclic graphs”) for like three weeks? That was fun.
150
Reposted by Jason Kerwin
Esteban J. Quiñones, Ph.D. @estebanjq.bsky.social · 23/04/2025
What happened to USAID is now slated to happen to the Millenium Challenge Corporation (MCC): www.mcc.gov What I am hearing is consistent with these screenshots! 😤😤😤 The dismantling of US Foreign Aid continues…
SCOOP!!!   DOGE terminates all Millennium Challenge Corporation programs, sources tell me 

“All MCC programs will be closed. No new programming is expected,” according to an email sent to MCC staffers moments ago 

“The agency will effectively be shut down.”

1/2For the countries that received an exemption and that have active construction the MCC has requested a 4 month wind down (Ivory Coast) and 3 month wind down (Mongolia, Senegal, and Nepal)

Staffing levels will be maintained for statutory requirement which is expected to be bare minimum (eg just a few), according to the email.

DOGE will draft a resolution for the board as soon as next week. No board resistance is expected.

Staff have until next Tuesday to decide to take deferred retirement/fork or RIF. The RIF process timing is unknown but expected to start as soon as next Wednesday.

MCC contracts will start being terminated tomorrow (via dodge)
MCAs in non-exempted countries have already issued stop work notices and will be told to terminate. RCMs have been given green light to communicate closure with host country governments/counterparts today.
33022
Jason Kerwin @jkerwin.bsky.social · 12/04/2025
poast pics
000
Reposted by Jason Kerwin
Audrey Dorelien @audreydorelien.bsky.social · 11/04/2025
The #PAA2025 Karaoke session is no joke! @jkerwin.bsky.social you would have loved this.
161
Reposted by Jason Kerwin
David Evans @daveevansphd.bsky.social · 09/04/2025
A remedial education program in India worked well when teachers were given strict implementation guidelines. It worked just as well when teachers had discretion to adapt. www.nber.org/papers/w33242 by Beg, @annefitz13.bsky.social, @jkerwin.bsky.social, @profalucas.bsky.social, & @kwrahman.com
12412
Reposted by Jason Kerwin
Matthew Yglesias @mattyglesias.bsky.social · 06/04/2025
There are a lot of interesting details for scholars to explore about basically everything, but for practitioners you get the most value out of the basics: — Moderates do better at elections — Free trade makes most people better off — The past was poorer
101579
Reposted by Jason Kerwin
CJ @cjeanmartini.bsky.social · 06/04/2025
One more incredibly creative protester ... #HandsOff #SaveOurDemocracy #LFG
They're eating the checks, they're eating the balances. Sign held by a Bill
271648396
Reposted by Jason Kerwin
VoxDev @voxdev.bsky.social · 02/04/2025
🆕 People think it’s easy to contract HIV. That might not be a good thing. Today on VoxDev, @jkerwin.bsky.social (@uwnews.bsky.social) discusses how providing information on actual (often lower) HIV transmission rates helped curb risky sexual behaviours in Malawi: voxdev.org/topic/health...
voxdev.org
People think it’s easy to contract HIV. That might not be a good thing.
Despite the high HIV prevalence in Malawi, individuals do not seem to adjust their behaviour to avoid infection—this may be due to the perceived transmission risk being so high that people become fata...
363