Hey #EconTwitter : Does your department offer some type of incentive (e.g. teaching relief) for serving in an editorial role (editor/co-editor) of a top journal (say top field or higher)? If so, do you mind sharing that?

Tom Carpenter, PhD (@tcarpenter.bsky.social) That’s bayesed lol

The academic year begins with bayesian literature...😂 Robert Kubinec

Yes but he's not taking into account the timing of the Neolithic revolution and rainfall patterns during the Roman Empire

A 70% desk reject rate is efficient but also raises ethical concerns given the APSR's prominence. As we often say in political science, who will guard the guardians? We should ask for audits to know if rejects are more likely for certain schools, class of scholars, etc.

UT-Dallas is hiring for an open rank faculty position in IR! We're particularly interested in scholars studying IPE and/or Conflict. Feel free to reach out if you have any questions! jobs.utdallas.edu/postings/27562

Any sufficiently large number of fixed effects is indistinguishable from causal identification.

Stata ordbetareg *sneak peek* 😎

we've got marginsplot working for ordbetareg in Stata... look at that nice bend... *whistle*

I've got an initial release of ordbetareg available for Stata (!!) on Github as a set of .ado files. It has the model + a predict method so you can use margins/marginsplot. Would appreciate it if people could test it/look it over before it goes to SSC :) github.com/saudiwin/ordbe…

Increasingly allergic to the quiet polisci norm that empirical papers should come up with elaborated theories/expectations first even if effectively they're intuitions from the process of data exploration. We/I have been inculcated this is the way to present but often doesnt work

Happy to see this in print. This covers opinion mining with supervised classifiers, NLI classifiers, and LLMs. Consider adding it to your syllabus if you’re teaching text as data!

but this plot is significant at the p<0.001 level so it must be worth it right

scientist trying to fit article into journal formatting requirements before submission

Interesting but who wants to listen to a podcast generated by AI? Could be helpful for making some generic content that then people edit into a podcast, but it seems like this will and just like LLM writing did (i.e., widely despised for real human tasks).

ordbetareg is coming to python...

OLS we are coming for your lunch

Initial version of ordbetareg for Python now available via Github: github.com/saudiwin/ordbe… Implements the model in MLE with scipy and includes predict/plotting functions. No marginaleffects yet b/c Vincent Arel-Bundock didn't know about this model until just now 🤣 please use/test!