Average Man

Beyond the Average Man
Roger Koenker

Slides for a talk at the 2007 FEMES meeting in pdf.



Abstract: Quetelet's conception of the average man, by introducing systematic measure of a wide variety of socio-economic quantities, marks the beginning of economics as a social science. Galton's introduction of correlation and regression 50 years later underscored the central role of the average man for empirical analysis. But the idealization of the average man, or his modern avatar the representative agent, neglects much that is essential; diversity of tastes, beliefs, information and endowments is crucial to many aspects of economic analysis. Heterogeneity of treatment effects is essential to many policy evaluation problems. Better econometric methods accounting for this heterogeneity are gradually emerging; four examples from recent work on quantile regression are briefly described in the course of the talk.