Clint Harris

Economist | Scientist @UWMadison | @PurdueKrannert PhD | I study labor and education using applied econometrics.

[Clint Harris @clintenttotreat]

@alz_zyd_ I didn't decide to become an Economist.

I realized I was one.

[Clint Harris @clintenttotreat]

You can substitute St. Germaine for vermouth in a negroni.

Good luck everyone!

[Clint Harris @clintenttotreat]

@MatthewGNagler I tried to but I couldn't.

[Clint Harris @clintenttotreat]

I started writing papers about monotonicity, and now I can't stop.

[Clint Harris @clintenttotreat]

@_alice_evans Very angry and uses the word "devilish" feels like Luther to me.

[Clint Harris @clintenttotreat]

If anyone wants, I have a way to correct the election for nonresponse bias by comparing early voters to late ones.

An Election Selection Correction, if you will.

New version forthcoming but not by Monday https://t.co/2Q30gZSfKP

[Clint Harris @clintenttotreat]

@besttrousers @surlygopher Did she switch from the left side of the street to the right following the suggestion of a man who was standing on the shore of the street?

[Clint Harris @clintenttotreat]

@lakens Imagine being a Cardiologist who does heart transplants all day and then you hear that the Oncology Department is leeching people.

That's what it feels like to see this in an academic paper.

[Clint Harris @clintenttotreat]

What if the voters herd toward the polls?

I believe I am the first to study this question.

[Clint Harris @clintenttotreat]

@jt_kerwin Congrats! 🎊

[Orgül Öztürk @orgul_ozturk]

Southern Economic Association Meeting are just over three weeks away and it is time to give you a heads up about all the cool CSWEP events happening at the conference. Mark your calendars and join us!

[Clint Harris @clintenttotreat]

@cblatts Inference is fine, same as doing doing 2 studies, lookings the first to guide hypotheses for the 2nd.

I personally would prefer the index be chosen based on something other than ATEs (e.g. ex post construct validity checks), but that's not an econometric concern.

[Clint Harris @clintenttotreat]

@clhubes @maiab The same people who peel potatoes and deseed bell peppers.

Unserious.

[Clint Harris @clintenttotreat]

@MatthewGNagler If you pool 2 different treatment arms x1 and x2 with t stats of a and b, respectively, the t-stat on the pooled treatment z=x1+x2 will be approximately sqrt(a^2+b^2).

[Clint Harris @clintenttotreat]

@trevorsgallen Feasibility aside, this process doesn't seem like demand estimation. It feels too reduced form, but what do I know?

Practically, IDK assumptions under which the predictive model gives good guesses for B for the pricing model.

I'd be interested to learn what these are 🙂.

[Clint Harris @clintenttotreat]

What do you think Kroger's facial recognition technology for price discrimination has under the hood?

1. Actual demand estimation
2. Big dumb predictive model

[Congresswoman Rashida Tlaib @RepRashida]

Families are struggling to put food on the table. I sent a letter to @Kroger about their decision to roll out surge pricing using facial recognition technology. Facial recognition technology is often discriminatory and shouldn't be used in grocery stores to price gouge residents.

[Clint Harris @clintenttotreat]

@trevorsgallen I think no.

I can set prices for customer i as
Pi = Xi*B

for high-dimensional X, randomly trying *arbitrarily many* values of B.

I can estimate the reduced form effect of the policy defined by some value of B.

This is good enough for profit maximization if I try *all* B.

[Clint Harris @clintenttotreat]

@Andrew___Baker My youngest daughter is 1.5 years old and it is her favorite show.

[Clint Harris @clintenttotreat]

@Lois_Miller Well done!

[Clint Harris @clintenttotreat]

@arthur_spirling Agree with you here, for sure.