Quantile Regression for Sports Economics

Abstract

Quantile regression provides sports economists with a powerful research tool. Unlike least squares, it is not tied to restrictive assumptions about the distribution of the error term, which makes it particularly valuable in settings with highly skewed distributions, like sports labor markets. It allows investigators to check for heteroskedasticity and to avoid censored variable bias. Researchers can use it simulate the distribution of incomes or profits, not just their mean values. Still, few sports economists use quantile regression, and, when used, it is frequently misinterpreted. This article provides a user-friendly introduction to quantile regression that will stimulate its use in the sports economics literature.