zach markovich

PhD Student in Political Science/ Statistics at MIT, retweets imply endorsement. My tweets are the opinion of my employer @MITPoliSci

[Arin Dube @arindube]

AI and economics research

To date, I've seen a lot more on how AI can help revolutionize economics research than I've seen how concretely AI made new research happen in economics.

In my own experience, I've had Claude Code markedly increase my productivity by cutting down on coding time. But it has failed to give me a single new economic insight. (I've tried.)

[Ye Zhang @zhangye68]

I’m on the academic job market this fall! I study comparative political economy, state-business relations, social policy, and China. Learn more about my work here: https://t.co/JI2CCfRyn7

[Cambridge University Press - Politics @CUP_PoliSci]

#FirstView from @polanalysis -

Estimating the Local Average Treatment Effect Without the Exclusion Restriction - https://t.co/YOYLlwkppU

- Zachary Markovich

[zach markovich @zachmarkovich]

@DanBischof @jeffreymziegler @seanjwestwood It’s framed more generally than earlier versions (which look directly at fmc’s) but the running example uses one

[zach markovich @zachmarkovich]

@DanBischof @jeffreymziegler @seanjwestwood Thanks for flagging this for me @jeffreymziegler. The latest version of my work on this here https://t.co/N91h0oXUeT

[Kat Hoffman @kat_hoffman_]

fwiw, the idea in this img was super useful for my learning

causal inference requires both identification *and* estimation of an estimand (AKA functional, parameter)

you can (and should!) think of them separately, but you'll want to make good decisions on both sides to succeed

[zach markovich @zachmarkovich]

In social science, I feel like we treat coding errors as a personal failing, but they're inevitable for anyone doing that much programing. We should institutionalize practices (like code reviews/ unit testing) that help people find the mistakes before they're published.

[Bernhard Clemm @BernhardClemm]

I herewith offer free code review for social scientists (R, Python and SQL) - I am serious!

Why? 1/6

[Bernhard Clemm @BernhardClemm]

I herewith offer free code review for social scientists (R, Python and SQL) - I am serious!

Why? 1/6

[Melody Huang @melodyyhuang]

My paper with Sam Pimentel on variance-based sensitivity analysis for weighting estimators is now on ArXiv! (1/6) https://t.co/CAcHRMW6jX

[Taniel @Taniel]

I've been asked what specific local & state elections matter a lot in the wake of Dobbs, & what concretely they'd affect.

With the caveats that the next steps need to involve mobilization, solidarity, & protests that aren't electorally focused, here are some thoughts. ↓

[Liam Allen-Miller @LiamAllenMiller]

@MattMcMuscles reposting my friend's shin godzilla airplane experience because it never gets bad

[Randall Munroe @xkcd]

Using GPT-3 to gain new insight into historical figures

[Ariel White @ArielRWhite]

The JOP blog has a piece about our forthcoming paper (with @zachmarkovich). The title pretty much gives you the punch line, but click through for more details!

[Journal of Politics @[email protected] @The_JOP]

🚨BLOGPOST ALERT🚨
Increasing the minimum wage spurs political participation of affected workers. @zachmarkovich and @ArielRWhite compare voting rates of New York City employees before and after minimum wage increases. Read what they find on our blog⬇️
https://t.co/9iJ0HA8HeA

[John Horton @johnjhorton]

Not your typical mental health professional advice:

[Guillermo Toral @guillermo_toral]

Very important research by former colleagues at @MITPoliSci examining mental health among grad students in top PhD programs. Departments should take these findings seriously and act accordingly.

[zach markovich @zachmarkovich]

My paper at #NeurIPS2021 this week introduces a more nuanced approach which explicitly defines causal effects in terms of the complete bundle of causal variables. If you work with these sorts of complex variables, please check out the preprint at https://t.co/n5lfbph6MO

[zach markovich @zachmarkovich]

Complex traits like democracy, race, or national power typically enter political scientists' models as single variables. Such approaches are inherently reductive and risk understating the magnitude of causal effects because they discard so much of the variation in these traits.

[zach markovich @zachmarkovich]

@amilliff Thank you! Will hopefully have the "camera ready" text up soon

[zach markovich @zachmarkovich]

I have a website now -- everyone needs to click on it so google will figure out it should come first for me https://t.co/iNvBOeEt5H