Oder Walverhalten?

Logo of the European Political Science Review, featuring the acronym 'EPSR' in large letters with the full name underneath and the hashtag '#OpenAccess' at the bottom, all set against a dual-tone background.

#OpenAccess from @epsrjournal.bsky.social -

Federalism and citizen preferences: a vignette experiment on policy-making in Germany - https://cup.org/3QDZnyY

- @jos38.bsky.social, Antonios Souris, @cgnguyen.bsky.social & @lenamasch.bsky.social

#FirstView

Under apreciated perk of teaching: Exposing students to 16 year old memes.

Can we run an IV based on rainfall when you had your first methods class?

Saying that "empirical" and "practical" approaches are exclusively quantitative is definitely a choice.

(Yes, regional context notwithstanding)

A lot more in the research note, which is available at the link above.

A special thanks to the James Madison Charitable Trust for funding the project and the data collection. 6/8 (Don't need the other two after all, edit function when?)

Turns out, people basically don't care who adopts as long as the policy works (and is paid for my the federal government).

"Federal adoption" sits on zero. 5/8

We use a vignette experiment to disentangle this cleanly. Varying
• who adopts (Bund or Land)
• who pays (Bund or Land)
• how effective the policy is
• whether it widens or narrows regional gaps

Respondents rated the policy. We looked at how these factors influenced policy evaluations. 4/8

The problem: Observational studies keep finding Germans prefer federal policy-making. But this conflates "the Bund did it" with "they paid for it," "it worked," and "it narrowed regional gaps." Any of those could be what people actually like. A survey can't tell them apart. 3/8

A fun collaboration between federalism scholars, @jos38.bsky.social & Antonios Souris, who bring the institutional questions, and experimentalists @lenamasch.bsky.social & me who bring a design to test them cleanly. A real fun pairing, and hopefully useful for both fields.2/8

Do citizens care who makes policy?

In Germany, some policies can be adopted either by federal or state governments. But do voters have preferences over which actor adopts a policy?

Turns out, they mostly do not.

New paper out @epsrjournal.bsky.social 🧵 1/8 www.cambridge.org/core/journal...

Job Vacancies - Universität Bremen Offene Stellen

I am looking for a postdoc to join my team at the University of Bremen. 5-year-contract, top-up to 100% possible for most of the contract period.

Apply by 04 May 2026.

If you have questions about the position, send me an email!

www.uni-bremen.de/en/universit...

I think he plays black and blue but gets annoyed because he isn't actually smart enough to pilot that.

a man in a suit is carrying a stack of pizza boxes . ALT: a man in a suit is carrying a stack of pizza boxes .

Emerging after a long day of research and student supervision to read the news and bluesky.

It's @professormusgrave.bsky.social o'clock again, alas

Can you say a bit about how the algorithm underneath it all works ?

This is very cool! Do you take feature requests on github or here?

They made it very clear that they refuse to interact with "synthetic writing" or some such, so no, they have not used it.

sweepthestrait.com

As a non-IR person this also clarified something for me. This framework still assumes some kind of larger rationality or end goal. But what happens if the hurt or ability to hurt is not a tool to achieve the goal but the goal itself? E.g. what if the sadism is the point ?

Honestly, this is maybe a great thing for them to see. One of the reasons I always struggle teaching DAGs is that as soon as we actually try to apply it to anything we have covered in class, it breaks down. Tbf, that says as much about our research as it does about DAGs.

psychologists and ingroup/identity folks: any work you can point me to on research that looks at cases where an ingroup lacks a clear outgroup? I'm thinking this may be an intriguing idea for thinking about state identities in the US.

That would be great.

I keep thinking of our cohort mates with military backgrounds, who did fieldwork in war zones while we ran regressions.

I think I can just repost this every couple of months. Seems to be eternally relevant.

The more advanced the method, the longer the code runs. So the longer the code runs, the more advanced I am as a user, and the more fancy my research. That is just science.

I feel like we should have exactly this paper, but for political science.

Dream job alert in eigener Sache: Wir suchen eine Werkstudent:in Datenjournalismus! 📊

⏳20h/Woche, befristet auf 12 Monate

Jetzt bewerben: jobs.spiegelgruppe.de/Vacancies/13...

Gibt es einen Datensatz der Wahlkreise über die Jahre harmonisiert / die Veränderungen der Wahlkreise maschinenlesbar macht? Ich weiss, dass das auch für andere Leute ein Problem war, aber nicht mehr, wie es gelöst wurde.