Dana Scott

@danascott.bsky.social

PhD student at Yale Economics, on the JM in 2025/26. Mostly labor. Spare time: cat 1 road bike racer + 2:44 marathoner. dana-scott.com

Suppose I have an instrument Z for some endogenous variable X. Z does a much better job predicting X for some X values than others. E.g. major better predicts occupation for nurses than for management consultants. What problems does this introduce? Is there a go-to reference for this issue?

Hello #EconSky 📉📈 I'm Dana! I'm a 4th-year economics PhD student at Yale. I study labor markets and dabble elsewhere: public, household finance, (aspirationally) applied econometrics. I race marathons and bikes in my spare time. Excited to see what is happening over here!