Monday, October 11, 2021

Nobel prize in economics awarded for innovation in causal inference from observational data (Michael Simkovic)

Three renowned labor economists, David Card (Berkeley), Joshua Angrist (MIT) and Guido Imbens (Stanford Business School) shared the Nobel prize in economics for their pioneering work using observational (i.e., non-experimental) data for causal inference.  This work facilitated empirical analysis of the effects of various legal and public policy changes, which are enacted in the real world and not under laboratory conditions.  Many scholars in law & economics and empirical legal studies built on their work and relied on the techniques the prize-winners developed.  

Card is famous for a series of difference-in-differences analyses across state borders that showed that moderate increases in minimum wage often don't lead to unemployment, as had been previously believed based on economic theory and simplifying assumptions.  Card's work was met with substantial skepticism, and conflicting claims from other empiricists, but he eventually changed the conventional wisdom among economists--a triumph of empiricism over theory and of science over ideology.  Card is a co-editor of the Handbook of Labor Economics.

Angrist and Imbens developed new ways to identify Local Average Treatment Effects, such as the use of instrumental variables. Angrist is also a co-author of Mostly Harmless Econometricsa text that is widely used to train economists, law professors with an empirical bent, and other researchers.  Imbens' methodological work is taught heavily in an empirical studies workshop run by Bernard Black at Northwestern and the late Matt McCubbins at Duke.  Imbens is also the co-author of a popular book on empirical methods, Causal Inference for Statistics, Social, and Biomedical Sciences.  

Black & McCubbin's workshop--which I highly recommend--is intended to help law professors and other researchers learn to engage in more sophisticated empirical analysis.

In widely cited work, Angrist found strong evidence that military service--specifically in Vietnam--adversely affected subsequent earnings.  Imbens and Angrist have also found strong evidence that education substantially increases subsequent earnings, using changes and variation in compulsory schooling laws. The causal relationship between education and earnings is now widely accepted among labor economists and other empiricists.

Guest Blogger: Michael Simkovic | Permalink