Paper of the week (POW) this week goes to "Designing Difference in Difference Studies: Best Practices for Public Health Policy Research" by Coady Wing and coauthors published in the Annual Review of Public Health.
It contains a lot of useful reminders on best practices for empirical strategies that rely on difference-in-difference estimators.
Very helpful! Used it this morning in exploring some preliminary results.
"Persistent Political Engagement: Social Interactions and the Dynamics of Protest Movements " by Leonardo Brusztyn and numerous coauthors came up while doing some research this week, and is certainly now the paper of the week (POW).
A copy of the abstract, the thrilling bits of which are in bold:
"We test whether participation in one protest within a political movement increases subsequent protest attendance, and why. To identify an effect of protest participation, we randomly, indirectly incentivize Hong Kong university students into participation in an antiauthoritarian protest. To identify the effects of social interactions, we randomize the intensity of this treatment across major-cohort cells. We find that experimentally-induced protest participation is significantly associated with protest attendance one year later, though political beliefs and preferences are unaffected. Persistent political engagement is greatest among individuals in the cells with highest treatment intensity, suggesting that social interactions sustained persistent political engagement."
A working paper version that is publicly available through Google may be found below.
So the first tasty bits of economists starting to become more apparent in the Harvard case appears on page 50.
It gets spicy QUICK.
LINK TO PAGE 50
So the case is an argument about the appropriate econometric specification.
Or if you're a fan of the lingo in the profession, it's an argument about the preferred specification.
Why do different sides have different preferred specifications?
Well that's on account of greater philosophical concerns regarding econometrics and identification. Suffice it to say that, in practice, results can change wildly depending on the controls used in a specification. Which controls are "good" and which controls are "bad" is clearly a point of contention in this particular case. And in just about any empirical research, at the frontier there is an argument about the appropriate specification to use.
I need to read the case more thoroughly and read the expert witness report by David Card before I comment on the preferred specifications, but it seems to be as to whether or not to condition on personal rating and status in particular advantaged groups like athletes, legacies, children of faculty, etc.
Click here for a link to the expert witness report filed on behalf of Harvard by David Card (labor economist)
Click here for a link to the Judge's Opinion hosted on my Google Drive.
It's always refreshing to see that tools you teaching to students are regularly used in their daily lives.
Perhaps the best and most recent example of applications of econometrics is in the case filed against Harvard for discrimination against Asian-American students.
Any empirical economist is going to have a mega complex regarding his quantitative skillset, by which he means his prowess in econometric analysis. It is our comparative advantage over fields like sociology or other liberal arts. Anything they say you can't measure, you probably can, and we probably are arguing over it at this very moment.
Econometrics has the wonderful ability to quantify just about any social science phenomena one is interested in, so naturally this lends to understanding whether, statistically, there is evidence of segregation.
I plan to write more on this soon, but my favorite passage so far is on p. 34
"Mr. Hansen's models could lead a casual observer to conclude that race plays a
significantly larger role in Harvard s admissions process than it actually does... The models
incorporate far fewer variables than those prepared by the parties economic experts for this
litigation and omit many variables that are important to the admissions process. Compare [PX12
at 33 ] , with [ PD38 at 26 ]. Even Mr. Hansen's most complete model almost certainly suffers
from considerable omitted variable bias in light of the likely correlation between race and
important variables that Mr. Hansen did not include. Most notably, his models contain no
controls for socioeconomic and family circumstances that correlate with race and also affect
admissions decisions. See [PX12 at 33] . Given these deficiencies in the models, they are
entitled to little weight for the purpose of determining whether Harvard discriminates against
Asian American applicants, particularly given the availability of the experts far more
comprehensive models and the testimony offered by fact witnesses in this case. See Oct 19 Tr.
19: 19– 20:8... Hansen's models do suggest, consistent with other evidence, that Asian
Americans applicants excel in academic metrics; that tips for legacies and recruited athletes
result in more white students being admitted; that a projection of Harvard' s class based only on
the profile ratings, academic metrics, and athlete and legacy statuses is incomplete and results in
a projected class that is vastly less racially diverse than the one Harvard achieves; and that ,
absent any consideration of race,Harvard s classes would have drastically fewer African
American and Hispanic students."
I'm super pleased to announce that my article with a dear friend and colleague Alex McQuoid has been published in Applied Economics!
We have a few free copies that can be accessed if you click here. At some point this link will no longer provide free copies, a pre-print version can be found here on ResearchGate and embedded in the post below.
This blog is a therapeutic outlet for me to write about life on the tenure track in economics.