Bad Econ Takes

Good Open-Access Econ Takes: The Journal of Economic Perspectives aeaweb.org/journals/jep The American Economic Association Research Highlight's Podcast aeaweb.org/research/podca… National Bureau of Economic Research Working Papers (3+ free per month) nber.org/papers?page=1&…

Electing a businessman is good for the economy.

"Reviving Mercantilism for the 21st Century" twitter.com/TrumpWarRoom/s…

Context: Controlling for outcome variables isn’t causal inference.

@AlanMCole ^This thread is worth reading.^

11. Catch it now? The way you make money supply work at predicting price levels is by completely removing it from the equation: both multiplying by it and dividing by it, so that it ends up perfectly canceled out.

1. This op-ed annoys me for a couple of reasons. (Though perhaps both of these annoyances reduce to the same point in the end.) wsj.com/articles/powel…

@jbarro The composition of interior counties is likely different than exterior counties which leads to strange slopes and can also mean that the treatment effects may or may not be generalizable. An RD identifies treatment effects from near threshold complier units.

@jbarro Not weighing in on the paper (haven’t read it), but the *slope* in an RD doesn’t confer any information about causal effects unless you impose strict assumptions about functional form. This is particularly true when “centering” on a threshold like this.

This polynomial model appears to say mask mandates become less effective the farther away you get from a county that doesn’t have such a mandate. Which.. what?

Context: Also worth noting… <10 think thank employees =/= “economists say”.

Context: I will be using this example when teaching causal inference to my students.

From the @WSJ