Fixed effects in unconditional quantile regression

In this article, Nikolai T. Borgen shows that when the number of fixed effects is large, the computational speed is massively increased by using xtreg rather than regress to fit the unconditional quantile regression models 

Abstract 

Unconditional quantile regression has quickly become popular after being introduced by Firpo, Fortin, and Lemieux (2009, Econometrica 77: 953–973) and is easily implemented using the user-written command rifreg by the same authors. However, including high-dimensional fixed effects in rifreg is quite burdensome and sometimes even impossible. In this article, I show that when the number of fixed effects is large, the computational speed is massively increased by using xtreg rather than regress to fit the unconditional quantile regression models. I also introduce the xtrifregcommand, which should be considered a supplement  to rifreg. The xtrifreg command has many of the same features as rifreg but can be used to include a large number of fixed effects, to estimate cluster–robust standard errors, and to estimate cluster–bootstrapped standard errors.

 

Continue reading full article

Emneord: xtrifreg, unconditional quantile regression, fixed effects Av Borgen, Nikolai T
Publisert 8. mai 2017 12:56 - Sist endret 8. mai 2017 12:56