Best of #econtwitter - Week of February 5, 2023: interesting tweets

Welcome readers old and new to this week’s edition of Best of Econtwitter. Please submit suggestions — very much including your own work! — over email or on Twitter @just_economics.

Interesting discussions

Arpit Gupta @arpitrage

We love to believe there are many specialists who enjoy digging through our complex prose. But consequence of turgid academic writing is we have moved away from text to communicate ideas to presentations. Now on Zoom; papers are rarely read except for final certification.

^might put this at the top of the newsletter once every few months. You’re reading this newsletter instead of PDFs for a reason!

Arpit Gupta @arpitrage

Effectively, our conventions on how to use the most powerful and informationally dense medium have gotten so bad that we have largely moved away from it to a much less efficient medium we actually understand. Which we now struggle to use remotely!

Peter Hull @instrumenthull

Econometrica seems to be experimenting with brief author responses to referee reports, prior to editor decisions. Seems cool!

Wojtek Kopczuk 🇵🇱🇺🇦 @wwwojtekk

If you submit a paper to the Journal of Public Economics, you are encouraged to include previous referee reports (with your response, if you wish) sciencedirect.com/journal/journa…

Harvey Lederman @LedermanHarvey

Proposal: If a paper is rejected, authors should be allowed to attach a note to referees on the next submission.

Peter Hull @instrumenthull

How would you feel if you learned your grad school randomized admissions (perhaps conditionally)?

Public bads

Savannah Ostrowski @savostrowski

This makes me so sad. Thinking about all the educators who are using Twitter's API in their curriculum/projects, as well as OSS maintainers of Twitter libraries (shoutout to tweepy).

Twitter Dev @TwitterDev

Starting February 9, we will no longer support free access to the Twitter API, both v2 and v1.1. A paid basic tier will be available instead 🧵

DrG @drgoldstein

Bad news for academics using Twitter to estimate sentiment! No more word clouds! #econtwit #fintwit #fintwitter. ⁦@lukestein⁩ Twitter to Remove Free API Access in Elon Musk’s Latest Revenue Push - WSJ

wsj.comTwitter to Remove Free API Access in Elon Musk’s Latest Revenue PushElon Musk’s social-media company didn’t say how much it would charge.

Public goods:

Deen Freelon @dfreelon

TWITTER: After 2/9, you'll be able to collect Twitter data using snscrape for Python, which does not require API credentials. It also supports Facebook, Instagram, Mastodon, Reddit, and Telegram.

github.comGitHub - JustAnotherArchivist/snscrape: A social networking service scraper in PythonA social networking service scraper in Python. Contribute to JustAnotherArchivist/snscrape development by creating an account on GitHub.

^“As announced last week, Twitter will eliminate free access to its APIs this Thu (Feb 9). This thread collates alternative sources of social/digital media data + tools for collecting them.”

Commentary

Matt Clancy @mattsclancy

Fields of science… …where papers by teams get the most cites relative to solo papers… …are also fields where teammates are most complementary (in a technical sense; a paper’s cites is more determined by least cited contributor) (Also, original charts in a twitter thread?!) https://t.co/7nqfZ9v5zN

Lukas Freund @_LukasFreund_

So here’s a simple binscatter 📈 of the elasticity of complementarity vs. beta_n based on the WOS data: nice + correlation, with meaningful magnitudes and also statistically significant. https://t.co/UW62eOKBp6

Lukas Freund @_LukasFreund_

A 🧵 on team production in science 🧪 & academia more generally: on benefits & bottlenecks and the joy when discovering an unexpected empirical result supporting a theoretical conjecture.

^you can and should cite Twitter threads that do original research/analysis 😊

Steven Ruggles @HistDem

Back in March, Pew published a paper claiming that the U.S. population in multigenerational households quadrupled since 1971. I was skeptical when I saw that, but I didn’t look into it until now. /1 pewresearch.org/social-trends/…

ChatGPT

Alex Tabarrok @ATabarrok

.@Stata needs to put much more code on the web, e.g. upload all Stata books, if it is to train ChatGPT and stay competitive. blogs.worldbank.org/impactevaluati…

Ingar Haaland @Ingar30

1/ More than 70% of the academic researchers who answered my poll "rarely or never ChatGPT". Perhaps some of you are unaware of how good it is. I use it to assist me with all sorts of tasks, including programming assistance and proofreading. See some actual examples below!

Amy Handlan @AmyHandlan

What are the ethical implications of ChatGPT at University? What does ChatGPT have to say for itself? This was our starting discussion for my class “Machine Learning, Text Analysis and Economics” @Brown_Economics.

Peter John Lambert @pj_lambert

ChatGPT & Text Analysis: The text-analysis landscape in Econ relies a lot on dictionaries, which can be adhoc in some cases. I started a new proj with @dan_chandler on tech adoption, and used AI to formulate priors (results below). @ellliottt @I_Am_NickBloom @StephenEKHansen

AI Breakfast @AiBreakfast

Use ChatGPT on your own files This is going to be big: humata.ai lets you upload a .pdf up to 60 pages long and allows you to ask questions about it in plain English ↓

Fin

Jadrian Wooten @Wootenomics

I occasionally get mail at my office address, but NEVER like what I received today. I had a handwritten 3-page letter from an incarcerated man in Oregon who is learning economics through a mail-based program and wanted to ask me (of all people) questions about economics 😭😭😭

Jonathan Roth @jondr44

@AnyaMarchenko6