#econtwitter is bad because it insanely overrepresents very progressive, always online, Anglophone, applied micro folks, who journos and political staff listen to because they are so available. With student loans today, the problem of this bias is super clear.
The interest pause in 21/22 and the forgiveness cost the same as *10 years" of a fully funded Canadian style parental leave policy in the US. Is there any person of any politics in America who prefers the student loan policy to this?
A white shoe lawyer three years in with a stay at home spouse will get, including the interest pause, 20-40k. Two first year lawyers at median get 40-80k! On equity, efficiency and dynamic incentives it is a joke.
Obvious policy here is Australia style income based repayment. Pay 1% of income on $ above 30k, up to 10% of income. Limit loans to programs/schools with very high non payment conditional on demographics. Problem solved. Much cheaper, better on efficiency, equality and dynamics.
Seriously, public engagement and idea sharing on social media can be great for academics. But particular implementation dominant among high-profile-on-Twitter Econs is so bad it has lowered my opinion of the research of many of the participants whose political bias seems so clear
One other issue - same bias is disaster for many other policy areas beyond student loans. Ask someone at the FTC, especially one who actually cares about policies to reduce monopoly power, about craziness there. There's a difference between actual experts & "experts say that..."
(To clarify, btw, the US *already has this*, though w/o the penalties for ripoff programs. Roughly, 10% of discret. income is max until debt retired or 20 yrs, or 10 yrs for pub service jobs. Hire Mitch Daniels to implement Purdue-style cost-control, and really problem solved.)
• • •
Missing some Tweet in this thread? You can try to
force a refresh
Afraid I am working on a new project in Dakar on sabbatical and am under the weather, so no full AFT post this year about the Nobel. But, a few bagatelles about an obvious prize, and also one that is widely misunderstood (but not by the winners, that's for sure!)
Incredibly, the idea of conditional statements (if then) being equiv to counterfactuals is a 20th c idea. Read some old philosophy and you will be shocked by how loosely people thought about what a counterfactual was. ( far looser on "the effect of X..." More on this in a sec) 2/
In practice it seemed there was nothing we can do for a counterfactual if we just use data. What would have happened if I, Kevin, was offered a job as a Michelin sous chef? In the data, it never happened, by definition. 3/
Perhaps of interest, some back-of-envelope math on Covid vax, "herd immunity", and who will be infected. Incomplete effectiveness, growing protection over time, combo of natural/vaccine protection means "% vaccinated" less useful for comparison over time/across countries 1/18
Assumptions below: 1 dose of Moderna/Pfizer/JNJ 80% protection from infection after 4 weeks, growing linearly, 2 doses of M/P give 95%, 80% of US 16+ will get vaccine (~10% in data only get one dose), infection gives 90% protection, # of infections thus far follow @Youyanggu 2/18
"Effective immunity" here is fraction of R0 infections that don't occur due to natural immunity or vaccine. Assume that kids are equally likely to get Covid and that vaccinated equally likely to have natural immunity as unvaccinated (both conservative assumptions). 3/18
Thread on economics and racism: the term "systemic racism" is misinterpreted. It does mean "we all are secretly racist and better HR training will fix things". It means "group inequality is result of historical and societal processes that are difficult to fix at the margin."
Black-white wage gap has been roughly constant at ~ 20% since early 1970s - see Derenoncourt and Montialoux on how minimum wage law changes led to some convergence in the 1960s: clairemontialoux.com/files/DM2019.p… . Adjusting for educ/location reduces slightly, but general trend is same.
There is a huge black-white (and black-Asian) gap in school achievement and dropout rates. We have well-identified studies showing these gaps can be closed with quality schools. See Angrist et al on KIPP: onlinelibrary.wiley.com/doi/full/10.10… and Dobbie/Fryer on HCZ: aeaweb.org/articles?id=10….