Quick Stata tip ICYMI: use "getcensus" to pull variables from the American Community Survey (ACS). ssc install getcensus *replace XYZ w/ your key global censuskey XYZ *get population by county getcensus B01003, year(2015) sample(5) geography(county) clear Code to graph below

Are you evaluating heterogeneous treatment effects in selected samples? Are you missing an instrumental variable? Are conventional effect bounds (e.g. Lee bounds) too wide? Then look no further: Bounds for heterogeneous treatment effects under sample selection + R package. 🧵 1/8

A helpful paper for thinking about robust estimators for DiD in finance sciencedirect.com/science/articl…

A 🧵 of useful #EconTwitter posts:

So in between student presentations in class, I mouthed to a student that her head scarf was a lovely color. Twice. Before realizing that if you are wearing a mask, people's ability to lip read is dramatically reduced.

For those of us who think in pictures, this is a great teaching resource

I am working on some theory and had an 'ah ha' moment where I realized I could rewrite a f.o.c. and it became a solvable differential equation. So, I feel 10% brilliant and 90% sure that either what I proved is either obvious, wrong, or both obvious and wrong.

What a great thread!

Emergency visits are expensive🚑 We find huge variation in costs by diagnosis, metro area, severity - with many exceeding the threshold that consumers can afford. Cynthia Cox Matthew Rae Dustin Cotliar, MD MPH @KrutikaAmin KFF Peterson Foundation bit.ly/3hx92qa

Is 2023 the year you or someone you love finally decides to learn R? If learning R is one of your New Years Resolutions, you should consider checking out/RT-ing my Introduction to R Programming course jacobjameson.com/Intro%20R NO PROGRAMMING EXPERIENCE REQUIRED!

Are you (or your referees) worried that the parallel trends assumption might not hold exactly? Ashesh Rambachan & Jonathan Roth propose methods for robust inference / sensitivity analysis that bound violations of PT using pre-trends: restud.com/paper/a-more-c…

Get ready for the 'insurance economics' session at #ASSA2023 tomorrow morning! This session organized by #ARIArisk is an opportunity to learn more about the institutional details of insurance. I will be presenting work with #KatieSchool 's Tice Sirmans on health insurance.

As cost sharing continues to rise in commercial markets, a surprising share (~50%) goes unpaid at hospitals. This often results in medical debt for consumers, but affects providers in important ways. w/ Boris Vabson in Health Affairs healthaffairs.org/content/forefr…

*Package Launch* One of my favorite trends in empirical IO is the increasing availability of portable code to estimate structural models – see PyBLP Chris Conlon Jeff Gortmaker, several packages by Jimbo Brand. This is key to make IO tools accessible (also outside IO) [1/5]

At which level should one cluster standard errors? What's good empirical practice? Have you ever heard about placebo regressions in this context? Here is a very useful guide that was recently published in the JoE doi.org/10.1016/j.jeco…. 🧵 with a short summary. #EconTwitter 1/9

First time memer, long time follower. I recommend Khoa Vu as referee.

Synthetic difference-in-differences, in Stata! Check out our article describing the package. Makes nice graphs illustrating time trends and unit weights. Comments welcome! arxiv.org/pdf/2301.11859…

Medicare just released draft details describing how historic drug negotiations will go. What happens if a drugmaker and Medicare can't agree on a proposed price? Up to three potentially very dramatic Zoom meetings! cms.gov/files/document…

Hi #EconTwitter, I feel sure there is a more efficient way to take lincom estimates in Stata (stored in r()), store each statistic as a scaler, and then manually describe the output for esttab, but I haven't figured it out yet. Any help is most welcome!