Ph.D. student @ApEcMN: development, household, crime, dairy markets, and all else tbd. I was told #EconTwitter was a thing, but I needed to find out for myself.
St. Paul, MN
Joined March 2021
- Tweets47
- Following58
- Followers43
- Likes17
Reading the new NBER working papers for this week. There were 4 papers that caught my eye: 1st on COVID unemployment benefits, 2nd on the effect of having immigrant students in schools, 3rd on racial human capital disparities, & 4th on Chinese New Year and elderly suicide.
On identification, they allow for differential pre-treatment labor market trends. This violates parallel trends, but they claim that the actual identifying assumption is that w/o treatment outcomes would have evolved similarly in labor markets with different increases in benefits
This is something I need longer to digest. In any case, this result seems particular to this scenario where jobs were already receiving a high # of applicants.
Were there articles other people found interesting?
Link: nber.org/papers
Ivan Strahof retweeted
If you’re a student and want a DD seminar on @agoodmanbacon and then @pedrohcgs and callaway, shoot me an email. First come first served up to some N tbd. I’ll then do a Google form. It’ll include directions on R code as well as easy simulations. I’m also good at intuition.
Every Monday I get an email with the latest NBER working papers. And in my ideal world I sit down with a cup of tea and read through these papers like I would the morning newspaper. Obviously, that never actually happens. 1/n
Franzoni & Stephan ask what risk-taking in science looks like, focusing on what can be found, if the method will work, the chance the outcome will be found, possible secondary findings, & its significance. As they argue, continuing a line of research doesn't imply risk avoidance
Here are where new working papers are posted and where one can sign up for the email: nber.org/papers?page=1&perPa…
Migrants traveling through cartel territory in Mexico pay to secure a wristband that guarantees their safety. It removes one type of risk of migration, at a cost. Who will provide cheaper wristbands and how many will one need? Hope it's not a trend
theguardian.com/us-news/2021…
Tip #5: Don't run two-way fixed effects in a staggered DiD. Your effects are probably not linearly additive, the variance-weighted average effect might not be what you want, and calculating weights this way is a real pain.
#IvansMetricsTips
Tip #6: The standard DiD estimator is equivalent to the two-way fixed effects estimator. The general staggered (multi-period) DiD estimator is equivalent to a weighted two-way fixed effects regression estimator. web.mit.edu/insong/www/pdf/F…
#IvansMetricsTips
I'm startled that "only 26% [Ph.D. students] find their work to be useful always or most of the time, compared to 70% of economics faculty and 63% of the working age population." Why is there such a gap bet students and faculty and how can we bridge it?
teapaw.org/AEAjournals/status/137…
Forthcoming in the JEL: "Graduate Student Mental Health: Lessons from American Economics Departments" by Valentin Bolotnyy, Matthew Basilico, and Paul Barreira. aeaweb.org/articles?id=10.12…
Tip #4: If you do staggered DiD and use two-way fixed effects, your estimator is a weighted avg. You can recover the weights, and estimate a variance-weighted average treatment effect on the treated: Goodman-Bacon (goodman-bacon.com/pdfs/ddtim…)
#IvansMetricsTips
