Best of #econtwitter - Week of February 21, 2021 [2/2]

Welcome readers old and new to this week’s edition of Best of Econtwitter. Thanks to those sharing suggestions, over email or on Twitter @just_economics.

This is part two of two; part one is here.

Paper summary threads

Tamay @tamaybes

A recent paper about innovation over the long run reveals a very neat snapshot of the composition of inventions over time. Using data on US patents, it identifies the following key waves: nber.org/system/files/w…

Katherine Stapleton @KathAStapleton

Some important new points from @JosephEStiglitz & @akorinek on AI and development: 1. Big difference between looking at impacts of AI within a country & globally. When benefits accrue in one country & costs in another, need winners to compensate losers across borders

NBER @nberpubs

Labor-saving advances in AI may undo gains from globalization of developing countries and pose new challenges for economic development, from @akorinek and @JosephEStiglitz https://t.co/OF835fzbeq https://t.co/TSLmys2nih

C. Kirabo Jackson @KiraboJackson

Nine credible studies looked at what happened to test scores when capital spending increased for K12 public schools. Most found nothing at first. Only two found stat sig impacts after a few years (left). But *on average* looks like a compelling increase over time (right).

Alexander Berger @albrgr

Interesting paper from Greg Clark and Neil Cummins arguing for a null effect and widespread publication bias in the literature on compulsory schooling impacts on health and income lse.ac.uk/Economic-Histo… Wish they had a better income measure than house prices though.

Matthew A. Kraft @MatthewAKraft

We say "Teachers are the largest school-based factor influencing student achievement" What if we're wrong? New @WallaceFdn report by @JasonAGrissom @annaegalite @Dr_ConstanceL shows principals have large direct effects for more kids + indirect effects. wallacefoundation.org/knowledge-cent…

More education threads from this week: technology in education RCT number one, number two; RCT on peer tutoring

Lionel Page @page_eco

Very interesting paper looking at national identity. 👉Alsace-Lorraine regions were contested between 🇫🇷 & 🇩🇪 👉 Both countries repressed regional identities 👉 Regional identities got stronger afterwards by @sirdeh & @KaiGehring1 forthcoming @AEAjournals papers.ssrn.com/sol3/papers.cf…

Matt Graham @Matt__Graham

NEW PAPER "Measuring Misperceptions?" The one-word answer is "no." There is a substantial gap between the beliefs researchers describe and the beliefs surveys measure. Paper: m-graham.com/papers/Graham_… Presenting today (Tu 2/17) at 2 pm eastern: jawspolisci.network [1/n]

Nathan Miller @NateHMiller

How much are markets likely to self-correct after an otherwise anticompetitive merger? I address this in a new working paper with wunderkind Peter Caradonna and frequent coauthor @GloriaSheu: “Mergers, Entry, and Consumer Welfare” papers.ssrn.com/sol3/papers.cf… (1/N)

Joel van der Weele 🇪🇺 @JoelvdWeele

1/ Tobacco & oil firms famously foment doubt about the harmful impact of their actions. In a new pub in @EJ_RES w/ @foerster_manuel we show people do the same: they justify selfishness by "casting doubt” on its impact, reducing others' prosociality: bit.ly/3jYHdnh Thread

Sephorah Mangin @SephorahMangin

My paper "Efficiency in search & matching models: A generalized Hosios condition" is now published in JET. The main result applies to any market with frictional meeting process, but it’s particularly relevant for macroeconomists who study unemployment. authors.elsevier.com/a/1cW7w50waSOD3

Aaron Schwartz @A_Schwa

Using Medicare claims data on heart attack outcomes, I find shrinkage measures create big, hidden, performance "taxes". For hospital heart attack mortality measures, the tax is about 40% on average. For smaller hospitals in the bottom size quintile, the tax is ~80%. (6/11)

Jon M Jachimowicz @jonj

🚨 New paper in press @ JESP 🚨 With ⬆️ racial diversity one might expect ⬆️ interactions btw White and non-White Americans Instead we find that ⬆️ racial diversity prompt White Americans to structure environments to ⬇️ contact with non-White others 1/ psyarxiv.com/yzpr2

Ajay Shenoy @AjaycencyMatrix

Lots of observational studies find that granting land tenure to farmers has big impacts on investment. That's NOT what Heather Huntington and I find in our large randomized controlled trial (forthcoming JDE doi.org/10.1016/j.jdev… ).

doi.orgRedirecting

Peter Hull @instrumenthull

Just posted a new (short) working paper! It’s on the interpretation of "marginal outcome tests" for racial bias, as recently popularized by Arnold et al. (2018) and others bit.ly/2ZbijY4 Here's a quick summary thread 👇

More: power subsidies in Indian agriculture; robustness to misspecification in macro models; bureaucrat allocation

Interesting discussions

Ben Golub @ben_golub

Something I'd like to understand better: This far into the pandemic, I'm surprised not to be seeing big Slacks organized around research topics. What I have in mind is a substitute for conference or department hallway talk - the same in some respects, different in others. 1/

Arpit Gupta @arpitrage

Some good innovation here

Journal of International Economics @JIntlEcon

New submission option! You may now elect to expedite the editorial process at the JIE using our Prior Review Process. Along with your paper, submit the decision letter and referee reports from a top-5 journal by choosing "PRP" in the Article Type Menu. https://t.co/4ZSrlTiwhA

Ben Moll @ben_moll

#EconTwitter hivemind: what are your favorite papers combining “causal” micro estimates (say from DiD or RCT) with a general-equilibrium macro model to answer an interesting macro question? This is for my PhD teaching so the easier to read the better. Thanks in advance!

Anthony Lee Zhang @AnthonyLeeZhang

Ppl in academia maybe don't have a great sense of how big the losses from coordination frictions are. For one example. How many grad students do you think have independently written cleaning code for common public datasets e.g. HMDA, ACS, call reports, etc.?