Behavior of casino gaming revenue over the business cycle considering alternative measures of “income”

Despite the importance of casino tax revenue to state and local governments in the U.S., few studies have examined the growth and variability of casino gaming revenue over the business cycle. Existing studies use personal income as the measure of the business cycle. Personal income is composed of three main income components: wage income, wealth income, and transfer payment income. We suggest that the use of personal income to assess the growth and variability of casino revenue inherently assumes that all individuals spend all forms of income identically and that the three income components behave identically over time. Or, in other words, personal income does not consider that different demographic groups have income from different sources, as well as the fact that the propensity of an individual to gamble out of each income source may be different. Previous literature on consumption expenditures supports this notion.

Our objective is to estimate the growth and variability of casino gaming revenue in the states of Illinois, Iowa, Indiana, Colorado, Mississippi, South Dakota, Nevada, Connecticut, and Pennsylvania in response to quarterly changes in wage income, wealth income, and transfer payment income. All income data are from the Bureau of Economic Analysis and all casino revenue data are available online from each state’s gaming regulatory agency. The motivation for our approach is as follows: variation in state personal income comes from the variation in each of the three personal income components and the relative size of each component; economic shocks likely affect the variation in each income component differently depending upon the type of shock; and the impact of the variation in each income component on casino gaming depends upon the income source of casino patrons and their propensity to consume out of each income source. For each state, we run dynamic ordinary least squares regressions to obtain estimates for long-run income elasticities (to assess growth) and short-run income elasticities (to assess variability) for each of the three income components.

Our preliminary results suggest that casino revenue behaves quite differently over the business cycle in response to changes in each of the three income components. Specifically, we find that in the majority of cases the income elasticity of demand from transfer payment income is larger than the elasticity from other income components. The results have implications for casino gaming research, public policy toward gaming, and the appropriate structure of casino revenue forecasting models.