Skip to content

Category: Social Sciences

On “Rationality,” Other Misinterpreted Words, and Cultural Exceptions

A long-time friend and colleague writes (in French, so my own loose translation follows):

Hey,

I was thinking about your post on rationality, a concept whose [economic] definition differs from its popular definition.

Other examples: “structural,” “public good,” or “efficiency.” In the limit, “profit” and “rent.”

Is it the layperson’s job to learn accurate definitions, or the economist’s job to be more precise about their vocabulary?

I think it’s our job to define the terms we use when we engage in public debates, for two reasons. First, because I believe the onus is always on the writer to be understood by his readers. That belief of mine probably stems from studying philosophy in college in a French-speaking university, and from the allergic reaction I got from being exposed to some of the most willingly obfuscating writing ever published (see Derrida, Jacques; or don’t.)

Love It or Logit, or: Man, People *Really* Care About Binary Dependent Variables

Last Monday’s post, in which I ranted a bit about the opposition to estimating linear probability models (LPM) instead of probits and logits, turned out to be very popular. In fact, that post is now in my top three most popular posts ever.

(Credit: xkcd.)
(Credit: xkcd.)

Last Monday morning, when my wife left for work, I told her I was expecting a meager number of page views that day given my choice of post topic. I was wrong: people really care about binary dependent variables.

A Rant on Estimation with Binary Dependent Variables (Technical)

Suppose you are trying to explain some outcome [math]y[/math], where [math]y[/math] is equal to 0 or 1 (e.g., whether someone is a nonsmoker or a smoker). You also have data on a vector of explanatory variables [math]x[/math] (e.g., someone’s age, their gender, their level of education, etc.) and on a treatment variable [math]D[/math], which we will also assume is binary, so that [math]D[/math] is equal to 0 or 1 (e.g., whether someone has attended an information session on the negative effects of smoking).

If you were interested in knowing what the effect of attending the information session on the likelihood that someone is a smoker, i.e., the impact of [math]D[/math] on [math]y[/math] The equation of interest in this case is