Separating the decisions of lottery expenditures and participation: A truncated Tobit approach

This study uses a sample of lottery players from three states to estimate the probability of lottery play and expenditures on lottery products using a different methodology than is used in previous studies. The authors reject the parameter restrictions implied by the standard Tobit and estimate the decisions of whether. to purchase lottery products and how much players spend using a probit and truncated Tobit, respectively. Overall, an interesting demographic profile of the players, which was not apparent from the results obtained by estimating the standard Tobit, emerged as a result of separating the decisions. The more general model provides new information for lottery administrators interested in designing programs that specifically increase market share versus programs that increase spending for those who already play regularly. This study calculates the expected value of lottery expenditures to estimate the tax incidence associated with different lottery products. Previous studies calculating lottery tax regressivity have used an incorrect formula for the expected value of lottery expenditures. The authors find that the taxes on all lottery games are regressive, with instant games having the greatest regressivity.