Estimating net lottery revenues for states

Based on 1980–95 data, this paper estimates and forecasts net lottery revenues for states with and without lotteries. This study indicates that a multi-state estimation is improved when a time-series and cross-section technique is used. Forecasting results are also improved when unequal time series in the data and less-than-full first years of operating lottery are controlled. For states without lotteries, the time-series and cross-section estimation indicates that only two of 14 states without a lottery would have generated net lottery revenue of more than $100 million. The number increases to five of 14 in the cross-section estimation.