Aaron Erlich

Super interesting! I see you haven't gone back to exams or anything like that. Do you think it's sustainable and that students aren't just using LLMs as a massive crutch anyhow?

Did an access-to-information system in #Mexico work? Did it become more responsive to public requests for information over time, or did portions of the public become more "expert" in their requests, which in turn ensured responses from the system?

The July issue of @worldpolitics.bsky.social is out! Learn about #DemocraticSocialist #PoliticalParty strategies, participatory institutions, #WealthInequality, #migrants and #socialsecurity coordination, and #Kissinger and #SoftPower. muse.jhu.edu/issue/55120 #Inequality #Capitalism

Elder, Am I Right? Age-Group Differences in Social and Political Interactions in Africa Abstract. Age-group differences play an essential role in social interaction across sub-Saharan Africa. However, the social effects of these differences re

I am super excited that this article about how age differences in the survey enumeration process might influence survey results in Sub-Saharan Africa bit.ly/3DUSh3m. Using @afrobarometer.bsky.social data, I show they do. Teaser, they may be more important than coethnicity. Check it out!

Does anyone know the approximate cost savings for journals to outsource their production to India? The quality of the services is just so low... and I truly wonder what the savings are.

Very relicable 😆

It would be cool to compare the false negatives from just using the abstract! But yea, one would think they might under count some stuff. Other elements of precision/recall are also interesting.

I dunno. Maybe it is worth it to distinguish research designs and estimators? I think that methods is growing alongside survey experiments. But I get the point! It's a subset of natural experiment.

I guess we should commend AJPS for making big steps in being more global. How does this differ from this Knutsen paper? www.duo.uio.no/bitstream/ha...

Unexpected Event Study Design?

The # is a bit arbitrary - maybe worse than statistical rules of thumb. But I think you are in this game bc you have multiple good ideas about publishable papers; hopefully, one figures a way to parcel them out steadily. Clumping sucks. I had a COVID backlog and put 10 for review at a time 🔫 🤯.

That's very cool. But the day seems a bit arbitrary hard to estimate, no?

I'm thrilled to see this work of years in print! AJPS just posted the blog version. Key takeaways: we have a more optimistic view of anti-corruption messaging than some recent articles. However, anti-corruption messaging is also by no means a panacea. We should be realistic about its effects!

Thanks @joenoonan.se! I’m glad you found it useful!

Erlich, Aaron, and Jordan Gans-Morse. 2025. “ Can norm-based information campaigns reduce corruption?” American Journal of Political Science 1–20. https://doi.org/10.1111/ajps.12940 Abstract Can norm-based information campaigns reduce corruption? Such campaigns use messaging about how people typically behave (descriptive norms) or ought to behave (injunctive norms). Drawing on survey and lab experiments in Ukraine, we unpack and evaluate the distinct effects of these two types of social norms. Four findings emerge: First, injunctive-norm messaging produces consistent but relatively small and temporary effects. These may serve as moderately effective, low-cost anti-corruption tools but are unlikely to inspire large-scale norm transformations. Second, contrary to recent studies, we find no evidence that either type of norm-based messaging “backfires” by inadvertently encouraging corruption. Third, descriptive-norm messages emphasizing corruption's decline produce relatively large and long-lasting effects—but only among subjects who find messages credible. Fourth, both types of norm-based messaging have a substantially larger effect on younger citizens. These findings have broader implications for messaging campaigns, especially those targeting social problems that, like corruption, require mitigation of collective action dilemmas.

Hot off the Press & Open Access!!!

Jordan Gans-Morse (tw @J_GansMorse / #NUResearch) and Aaron Erlich (@aaronerlich.bsky.social lich.bsky.social / @mcgill.ca ca), "Can norm-based information campaigns reduce corruption?" via AJPS @ajpseditor.bsky.social #polisky doi.org/10.1111/ajps...

That’s awesome. Lemme know if you want me to give any feedback.

Has anybody tried Google Scholar’s new PDF reader?Any good? #polisky

I second this motion!