In a recent paper, Majumder and Chakravarty (1990) propose a four-parameter model which they find provides a better fit to some income data than the lognormal, gamma, Singh-Maddala, Dagum, and generalized beta of the second kind (GB2) distributions. This note (1) demonstrates that the model proposed by Majumder and Chakravarty is a reparameterization of the GB2 and (2) reconciles the corresponding contradictory empirical findings reported by Majumder and Chakravarty.
