Surveyed to Death? Another AI-assisted example of clearing out the backlog of orphaned RA projects

Surveyed to Death?
open.substack.com/pub/jasonmfl...

We Selected on Independence—Does AI Make It Obsolete? (part 1) What Will Hiring Committees Look For When Everyone `Has Ideas'

We Selected on Independence—Does AI Make It Obsolete? (part 1)
open.substack.com/pub/jasonmfl...

The Transitivity Fallacy in Academic Publishing Why “better" papers don’t land in "better" (or even the same) journals

The Transitivity Fallacy in Academic Publishing
open.substack.com/pub/jasonmfl...

Portrait of Assistant Professor Héctor Pifarré i Arolas.

Portrait of Professor Jason Fletcher.

Assistant Professor Héctor Pifarré i Arolas and Professor @jasonmfletcher.bsky.social co-authored a new study suggesting longevity gains across all states and regions for those born from 1941-2000, contrasting previous claims of stagnation and divergence across regions.

🔗 Read more: buff.ly/wbmQAR1

I'm co-organizing a free, 2-day Zoom workshop from the NAS & hope you'll consider attending!
📅 April 23-24: Discuss how we can enhance scientific integrity in the social and behavioral sciences
REGISTER: www.nationalacademies.org/projects/DBA...
PROGRAM: www.nationalacademies.org/cdn/material...

Our new paper challenges a bleak picture of U.S. state longevity gaps: universal cohort gains (1941–2000) & a more nuanced regional story than recent estimates.

News: eurekalert.org/news-release...
BMJ Open: bmjopen.bmj.com/content/16/4...

The Beauty of Slow Research Assistants Immediate and comprehensive feedback on all your research ideas can overwhelm

The Beauty of Slow Research Assistants
open.substack.com/pub/jasonmfl...

I think I'd want to pair this with (1) much shorter applications (2) more reviewers per application, but short reviews ("skims") or something that would be time-saving

I think we should do some randomization, but how much time savings would happen with this setup?--PIs would still submit full applications, review panels would do the same work; it sounds like we would just "disregard" the precise ranking and randomize over the 5%-10% or so?

The NIH Grant Lottery: Buy More Tickets, or Pick Better Numbers? Reviewer disagreement, grant-writing strategy, and what NIH’s new submission cap may or may not change

The NIH Grant Lottery: Buy More Tickets, or Pick Better Numbers?
open.substack.com/pub/jasonmfl...

Better Disagreement Through AI How machine mediation might improve—not eliminate—intellectual conflict.

Substack Post: Better Disagreement Through AI

AI and the End of the Dataset Monopolist...in some fields Why cheaper data construction increases the quantity of research in economic history and other areas

AI and the End of the Dataset Monopolist...in some fields
open.substack.com/pub/jasonmfl...

Will AI Kill Economics’ Comparative Advantage among the social sciences? How the credibility revolution might finally spread—and what that means for junior scholars

Will AI Kill Economics’ Comparative Advantage among the social sciences?
open.substack.com/pub/jasonmfl...

The One-Shot Paper Isn’t One Shot Why AI-generated research looks replicable—but isn’t

Substack: The One-Shot Paper Isn’t One Shot

Can AI fix journal reviewing...a little? Why AI-assisted peer review can quietly eliminate one of the most common (and infuriating) reviewer mistakes

Substack: Can AI fix journal reviewing...a little? substack.com/@jasonmfletc...

Interdisciplinary arbitrage and an AI-assisted research paper (paper 2 of n) Another quick peak into a project I can lay to rest (but still report)

Substack: Interdisciplinary arbitrage and an AI-assisted research paper (paper 2 of n)

I am so excited to share that I will be joining the La Follette School of Public Affairs at the University of Wisconsin–Madison as an assistant professor in Fall 2027, after spending the 2026–2027 academic year as a Frank H.T. Rhodes Postdoctoral Fellow at the Cornell Population Center!

The Postdoc Isn’t One Thing Why the type of postdoc you take may matter more than where you take it

Substack: The Postdoc Isn’t One Thing

Owning -All- the Numbers Why every number in your output is your responsibility

Substack: "Owning All the Numbers." A key step in research training.

“When I grew up I did become a scientist, a social scientist. I’m an economist, which allows me to study how we humans coordinate and cooperate and compete with each other, in ways that have made us, on average, live longer and healthier lives.”
#econsky
marketdesigner.blogspot.com/2026/03/ps-2...

From Training RAs to Training Workflows A Hidden Cost of AI-Assisted Research

Substack: From Training RAs to Training Workflows

Ongoing work with @hamidnb.bsky.social on NAFTA and mortality
nber.org/papers/w34840

Which AI? Why researchers are talking past each other more than ever

Substack: Which AI? Why researchers are talking past each other more than ever

In the classic Chains of Affection paper in AJS, Bearman et al found that four-cycles -- two couples switch partners -- were non-existent in a network of high schoolers romantic ties. They also found a giant component of exes exes.

I take their ideas to the full network of Norwegian partnerships.

A Hidden Productivity Gain from AI: Testing, Reporting, and Abandoning Low-Return Ideas Lots of concern abounds in academia (and other places) that Claude will replace us.

Substack: A Hidden Productivity Gain from AI: Testing, Reporting, and Abandoning Low-Return Ideas

Running up that hill...how will AI shape pre and post tenure research productivity? scott cunningham has a poetic ode to being a researcher who loves research and some worries about what AI will do/is doing to people like that (him, me, you).

Substack: Running up that hill...how will AI shape pre and post tenure research productivity?

Saving lives: lifting many boats or patching holes Two papers in the newest issue of the Journal of Health Economics highlight an open question in research on US mortality—how should we consider tradeoffs between investing in early life and investing ...

Substack: Saving lives: lifting many boats or patching holes

Hidden Curriculum: picking journals New paper: “Submissions Related to Editors’ Research Are More Likely To Succeed: Evidence From Economics Journals” shows evidence that (1) editors are less likely to desk reject manuscripts related to...

Substack: Hidden Curriculum: picking journals

Fingers crossed--I think universities are likely to make rules that exclude this option

The Return of the Office Supercomputer For several decades, many universities solved the computing problem.

Substack: The Return of the Office Supercomputer