Best of #econtwitter - Week of January 30, 2022 [4/3] special metrics edition

Because there were already three parts to last week’s newsletter (part one, two, three), broken out here is a high concentration of good metrics content from last week:

John Bonney @theJohnBonney

Hi there #econtwitter—you may have noticed my new paper with @cblandhol, Magne, and Alex: "When is TSLS Actually LATE?" I am stoked about this paper and wanted to briefly summarize our findings. (1/n) Paper here: papers.ssrn.com/sol3/papers.cf…

John Bonney @theJohnBonney

Our findings? Without additional parametric assumptions, there is precisely one 2SLS specification that is *guaranteed* to be a positively-weighted average of LATEs. Remember that SW specification? The one that no one ever uses? Yep, that’s the only one. Oof. (9/n)

John Bonney @theJohnBonney

Where does this leave us? Well, unless you are using the SW approach (and we both know you aren’t), then under the usual IV assumptions alone, 2SLS is *not* LATE. You must also assume that you have rich covariates and that your 1st stage is monotonicity-correct. (13/n)

Matt Blackwell @matt_blackwell

Kudos to everyone who is figuring out each case, but maybe just make your working hypothesis that if you’re using least squares with anything other than a fully saturated discrete covariates, you may not be getting what you expect.

Average Weighter @agoodmanbacon

Every day I stray farther from the view that controlling for things using a regression is a good idea.

---

Peter Hull @instrumenthull

Ok, so I come bearing good news for ~93% of you: esp. those bootstraping complex models (e.g. w/many FEs) Instead of resampling, which can be seen as reweighting by a random integer W that may be zero, you can reweight by a random non-zero non-integer W

Peter Hull @instrumenthull

Question for economists who do empirical work: When you need to bootstrap your SEs, do you:

^thread, lots of discussion in the replies/QTs, truly the best of econtwitter, including:

Grant McDermott @grant_mcdermott

Okay, here’s my quick attempt at making this a fast, reusable function. gist.github.com/grantmcdermott… I _think_ it’s working, but please kick the tyres (examples at the bottom). If all looks good then I might port to a package.

Peter Hull @instrumenthull

Ok, so I come bearing good news for ~93% of you: esp. those bootstraping complex models (e.g. w/many FEs) Instead of resampling, which can be seen as reweighting by a random integer W that may be zero, you can reweight by a random non-zero non-integer W https://t.co/Rpm1GmomHg

Kevin Chen @jiafengchen42

Seems like today is a good day to talk about why the bootstrap works Here's my intuition---

Paul Goldsmith-Pinkham @paulgp

Here's a nice write-up by Bryan Graham on this method here: bryangraham.github.io/econometrics/d… also Chamberlain and Imbens (2000): nber.org/papers/t0200 (ungated older version here: scholar.harvard.edu/chamberlain/fi…)

Peter Hull @instrumenthull

Ok, so I come bearing good news for ~93% of you: esp. those bootstraping complex models (e.g. w/many FEs) Instead of resampling, which can be seen as reweighting by a random integer W that may be zero, you can reweight by a random non-zero non-integer W https://t.co/Rpm1GmomHg

---

Kevin Chen @jiafengchen42

Excited to share a new preprint with @DRitzwoller on semiparametric estimation (read, *machine learning* estimation) of long-term treatment effects, such as those identified by a latent unconfoundedness or a surrogacy assumption (1/n) arxiv.org/abs/2107.14405

scott cunningham @causalinf

Couple things I'm learning today: 1) Limited information maximum likelihood (LIML) originates in the Cowles Commission work on simultaneous equation modeling and instrumental variable methods in the 1940s and 50s by Haavelmo, Koopmans and others 1/n

Francis DiTraglia @economictricks

I'm coming around to the idea that we need to teach DAGs. The textbook advice is to control for anything that's correlated with both treatment and outcome without being caused by the treatment. This is wrong, but it's very hard to explain why verbally. @causalinf #EconTwitter

Robert Kubinec @rmkubinec

Econometrics is this really confusing mix of "your standard errors could be off asymptotically, use this non-parametric clustering adjustment algorithm" and "it's fine if the data are binary, just use OLS, no one cares about the errors anyway." #polisci #econtwitter