Specification(s) of Nested Logit Models

The nested logit model h s become an important tool for the empirical n lysis of discrete outcomes.There is some confusion about its speci fication of the outcome prob bilities.Two m jor v riants show up in the liter ture.This paper compares both nd finds th t one of them (called random utility maximization nested logit,RUMNL) is preferable in most situations.Since the comm nd nlogit of Stata 7.0 implements the other variant (called non-norm lized nested logit, NNNL),an implementation of RUMNL called nlogitrum is introduced.Numerous examples support nd illustr te the di fferences of both speci fic tions.