David McKenzie
@dmckenzie
Lead Economist at the World Bank’s Development Research Group Development Impact blogger
Most packages are in Stata, followed by R. Only 1 in 5 submitted reproducibility packages is computationally reproducible on first review – meaning the team is able to run the data & code provided & obtain the exact results reported in the publication - so it's important to have these checks 3/3
Most packages in the repository run in under an hour, so may be useful for teaching. There is a demand for these packages, with nearly 20,000 users downloading a package so far. 2/3
Some journals have at least increased the number they publish as submissions increase, with over 1,000 papers published in these 14 journals in 2025. So there is a lot of new development research. Acceptance rates are in the 6-8% range at many journals
Submission numbers grew rapidly in the last year at several development journals. World Development got 1276 more submissions than in 2024, reaching over 5,000; and JDE got 830 new submissions, reaching over 3,000. Both exceed the AER's 2,114 submissions.
Reminder, last couple of weeks to get your papers in for a special issue of the WBER on migration and development.
Call for papers for a World Bank Economic Review special issue on migration and development: we are doing a special issue that builds on a recent conference, but with submissions open also for non-conference papers. Short papers in the AER insights type format also welcome. Deadline November 30.
Desk rejection rates average 72%, so conditional on getting past the desk reject stage, there is about a 30-40% chance of acceptance at most journals. Some journals now have mean time to a decision conditional on going to referees under 90 days, but a small % don't have decisions in 6 months still.
There has been rapid growth in the number of submissions at many leading development economics journals over the past 9 years, with submissions up more than 50% at World Development, JDE, JDS and WBER. Some, but not all, journals have increased the number of papers published
Happened across biostatistician Keith Goldfeld's research page today (med.nyu.edu/faculty/keit...), & saw this little picture of research topics published by year. This would be cool to do for my own papers (and might be helpful in stopping referee requests for things you haven't worked on for ages)
This figure shows the key result on profits: over 2 years men and women have similar average and quantile treatment effects, with profits growing $61-65/month (for a $750 training). After 7 years, treatment effects on profits for men are $148/month, vs $39/month for women. 2/3
Ok #EconSky, I have two more squares to fill in for this year's job market reference letter Bingo. What else should we be looking for as we read lots of applications?
After seeing discussions of 1/10th of the U.S. budget going to interest payments, I realized that I normally only see debt/GDP ratios compared across countries, and not interest spending/government spending. @ourworldindata.bsky.social to the rescue - U.S. is high until you add developing countries
Call for papers for the 17th annual conference on Migration and Development. This year to be held in Bologna on December 9/10, with submissions due September 20. Option to submit to a special issue of World Development. This is one of my favorite conferences to attend.
While the literature typically models migration as a family decision, we show the family itself frequently changes with migration. Adult migrants to the U.S. are more likely to marry after migrating than before - and so different destinations entail meeting and marrying different spouses. 2/n
Returns to capital investment: Vegas edition! $2.3 billion dollars on massive highest resolution in the world LED screens made for an incredible U2 concert experience at the Sphere. Never seen anything like it.
Efforts to directly get jobseekers on online job platforms have typically not increased employment and occasionally backfires. But government can help make these platforms work better. E.g. trust is a big issue, so a role for background checks, reputation mechanisms; and skill certification.
Surprisingly few surveys ask those with jobs how they found them (more ask the unemployed what they've done to look for jobs, but without knowing what succeeds). We show most jobs found through social networks & direct applying to employers: not through intermediation agencies or online platforms
Available for free at this link for the next 50 days, my paper with Catia Batista on using lab experiments to test classic theories of migration is now out in the Journal of International Economics. authors.elsevier.com/a/1hvpL5330f...
Fun morning visit to the pop-up Museum of Failure to see great product failures like Colgate Beef Lasagne, a hula office chair that is meant to work your abs while you work, the Atari ET game, a cabbage patch doll that chewed kids’ hair, scentovision that would read a story & spray scents & more