Will Queen

[Will Queen @WillQueen5]

Moved to Somerville at the perfect time

[Mike Connolly @MikeConnollyMA]

The GLX Community Path Extension will finally open to the public tomorrow! Join us on the path at 10 am to celebrate the opening at 112 Central St.

[Will Queen @WillQueen5]

@PanditHitanshu @Northeastern @SasserModestino Excited to see you up in Boston!

[Will Queen @WillQueen5]

@KhoaVuUmn I hope that the leaked Instagram platform that looks like Twitter is real. I'd certainly switch to browsing a text-based Instagram app

[Will Queen @WillQueen5]

@ben_golub My teacher sung it to the tune of Hey Jude, repeating "2a" in dramatic fashion over and over at the end

[John Burn-Murdoch @jburnmurdoch]

NEW: I’m not sure people fully appreciate how dire the US life expectancy / mortality situation has got.

My column: https://t.co/dBIhT9eZLv

And some utterly damning charts.

1) at *every* point on the income distribution, Americans live shorter lives than the English.

[Will Queen @WillQueen5]

@KhoaVuUmn Your advisor reading the table notes

[Daniel Piker @KangarooPhysics]

I picked the solution for 17 because that one looked particularly funny to me, but it's not just this number with an awkward looking solution - the arrangements for most numbers are asymmetrical, the neat ones like 16 are the exception.
https://t.co/uBJwYCs8qn

[Matheus Facure Alves @MatheusFacure]

Here is a very simple python implementation of Doubly-Robust DID, by Sant'Anna and Zhao. It helped me understand the paper, but also showed me how to view DID through the lens of delta outcomes. Hope someone else finds it useful.
https://t.co/AnvN50O747

[Will Queen @WillQueen5]

@KhoaVuUmn Mere m:m now, delete later

[Garth Heutel @GarthHeutel]

"Economics is the least socioeconomically and racially diverse of any major PhD discipline" ...

https://t.co/Qo9psUirTE

[Peter Hull @instrumenthull]

Want to learn the latest in estimating school quality w/ quasi-experimental data? Or the various IV tips & tricks @BlueprintMIT has accumulated over the years?

Check out this new-and-improved handbook chapter with Josh (@metrics52) and Chris (@c_r_walt)!

https://t.co/wvfw2aXtZg

[Jonathan Roth @jondr44]

In math, there are lists of "open problems" -- i.e. Qs that the field thinks are important but no one knows the answer to.

Do we have open problems in econometrics?

Or is the issue that we don't know what Qs we should be asking?

#EconTwitter

[Jeffrey Wooldridge @jmwooldridge]

I also recommend using methods that actually recover LATE (or LATT) when covariates are needed. Our (@TymonSloczynski @sderyauysal) DR LATE (LATT) estimators are easy to compute and doubly robust. Let Doctor LATE cure your identification ills. 😬

https://t.co/f8UhMOae9Q

[Giuseppe Cavaliere @CavaliereGiu]

Hi #EconTwitter!

This paper by @cblandhol (@Princeton), @theJohnBonney (@Stanford), Mogstad and Torgovitsky (@UChi_Economics) shows that 2SLS estimates of LATE 𝐭𝐫𝐞𝐚𝐭𝐦𝐞𝐧𝐭 𝐞𝐟𝐟𝐞𝐜𝐭𝐬 could be invalid if the specification includes covariates.

Highly recommended!

[Will Queen @WillQueen5]

@KhoaVuUmn But what if it's just correlation and not causation? /s

[Will Queen @WillQueen5]

@ryxcommar @TresBienHien "well boss, the AI looks great. It just can't quite figure out linear regression, so we're gonna have to keep us around"

[Will Queen @WillQueen5]

@RespectfulMemes

[Will Queen @WillQueen5]

@bartbing71 @KhoaVuUmn We can only argue about the best/worst way to use propensity scores

[Will Queen @WillQueen5]

@jmwooldridge Yes

[Will Queen @WillQueen5]

@instrumenthull It's pretty ordinary