Revenue forecasting models for small municipalities in Texas

Abstract

The major objective of this dissertation was to investigate if existing revenue forecasting models could be modified to provide useful predictive models for revenues of small Texas municipalities. Specifically, an attempt was made to adapt the existing econometric revenue forecasting models described in this study to two small Texas municipalities. The accomplishment of this objective required three major stages. The initial stag e involved analyzing and collecting available data for both the dependent and independent variables. Although many problems were encountered at this stage, reasonably accurate and consistent data with a sufficient number of observations were obtained for most variables. The second stage involved estimation of the revenue forecasting equations using multiple linear regression. The amount of variation in the dependent variable explained by the final equations was high and although the accuracy of the independent variables that applied to small areas was questioned, local independent variables were significant in most equations. In the final stage, projections for the independent variables were obtained and used to generate forecasts. Projections for the independent variables were generally obtained through relating the independent variables to similar independent variables for which projections were available or through extrapolation of recent trends. Overall forecast accuracy of both models was quite good. The forecasts obtained from the models were generally a significant improvement over the estimates of the cities.