Post

Conversation

joseph francis
@joefrancis505
The issue is the elasticity of substitution between graduates and non-graduates: 1.6 in Katz and Murphy, compared to 6.1 in Richardson. Therefore, Katz and Murphy imply that funnelling money to Cambridge, Masschusetts will reduce inequality; Richardson implies that it won't.
Quote
Brian Albrecht
@BrianCAlbrecht
Replying to @mr_ki11myselF and @joefrancis505
Digging into it now. It seems interesting. The nonstationary seems minor. I guess it depends on what do you take KM to mean. Is it the exact elasticity?
27.8K
Views
David Watson 🥑
Post your reply

Calvo Fairy
@FairyCalvo
I don’t trust any time series identification of such a parameter, so I’m not saying Katz and Murphy’s 1.6 is particularly credible - but 6.1 seems ludicrously high. Educated and non-educated labor are definitely not that close to being perfect substitutes.
306
Brian Albrecht
@BrianCAlbrecht
Correct. And 6.1 is well outside the norm. I'm not saying its not the best estimate. I'm still trying to figure out how he gets such a different number (not just relative to Katz and Murphy).
615
joseph francis
@joefrancis505
Because he tests for unit roots, finds there are unit roots. Then he tests for cointegration, finds there is no cointegration. So he first-differences the time series, and that gives him 6.1. He's just following textbook econometrics––i.e. what you are supposed to do.
531