Characteristics of households who participate in gambling markets in the US, and the determinants of household expenditure on gambling, are investigated using data from the Consumer Expenditure Survey (CEX). I estimate empirical models of participation in gambling markets and gambling expenditure using Tobit and double hurdle estimators. A likelihood ratio test rejects Tobit in favor of the double hurdle model. The double hurdle model estimates show that key explanatory variables affect participation and expenditure with different signs. Tobit, which is widely used in the literature, forces the signs to be identical, calling into question empirical regularities reported in this literature.
Job Market Prospects of Breast vs. Prostate Cancer Survivors in the US: A Double Hurdle Model of Ethnic Disparities
