RT @BardHarstad: Last month, I stepped down as an editor after 6.5 years (4 REStud; 2.5 JEEA). I have gathered a few suggestions that can….

RT @TSN_Sports: NICK OLESEN GIVES DENMARK THE LEAD WITH UNDER A MINUTE TO PLAY! 😱🇩🇰. #MensWorlds

Let that sink in: Denmark beat Canada 2-1 in the quarterfinals at the Hockey Worlds!! . Before the game, my family teased me that it would be amazing if Denmark would only be behind by 4 after the 1st. I am heading outside to look for flying pigs 🐖.

One of the fun things about reading both the Danish/European and US news: On one side Trump claiming to hold all the cards, on the other EU: let us wait and let him commit more errors (paraphrasing Napoleon).

RT @I4Replication: A re-analysis of Ciacci's (2024) "Banning the purchase of sex increases cases of rape: Evidence from Sweden" Journal of….

RT @dwarkesh_sp: Unreasonably effective writing advice:. "What are you trying to say here? . Okay, just write that.".

RT @jenniferdoleac: I'm beginning to put together the summer schedule for the ViCE seminar. If you have a paper related to crime or the cri….

It’s hard to overstate how valuable DHS data has been for research across many fields. Its demise marks a huge loss. Personally, I’ve published 3 papers using DHS and have another 4 in the pipeline. An ode to the Demographic and Health Survey Program:.

RT @cblatts: I was skeptical of the $200/mo price tag for ChatGPT DeepResearch, but man is it paying off. The ability to get a cogent, well….

The paper is full of descriptives, and I hope you check it out. The author list is very long, and I will not even attempt to tag everyone, but please do tag yourself on this thread if you're an author. Big thanks to the @SloanFoundation for funding.

Other things do not matter at all—the researcher's background or experience, for example. We also had a peer review element that seemed to have no effect whatsoever on later stages of research!.

Cleaning everyone's data moves the needle, narrowing the effect distribution even as modeling choices continue to differ. This makes sense, as we don't teach or discuss standardized/best data-cleaning practices, so there's a big variance in how we do it.

As we add more restrictions, things get interesting. Forcing a shared research design actually leads to *less* agreement on effect size (although more on sample size). This is partially driven by improper adherence to the specified research design.

We find huge differences in the methods used, including just how big the sample sizes of "eligible people" are. But compared to other many-analyst projects, variation in the actual effects is fairly small - an IQR of just 3.1 pp in the effect size, although there are outliers.

We have researchers estimate the impact of DACA on employment rates for those eligible for DACA. At first they can do it however they like. Second, they have to use a shared research design. Third, it's a shared design AND we clean your data for you.

The Many-Economists project with @nickchk, The Sources of Researchers Variation in Economics, is now on SSRN. We had 146 teams perform the same research task three times, each with less freedom, to explore what drives differences across researchers.

An excellent way to spend a Saturday evening. #RIHlive

My paper with Shamma Alam and Ishraq Ahmed, “Impact of Twin Lockdowns on Hunger, Labor-Market Outcomes, and Household Coping Mechanisms: Evidence from Uganda,” is now in open-access pre-print at the World Bank Economic Review!.