Municipal Revenue Forecasting with Time. Series Models: A Florida Case Study

Abstract

Forecasts of revenue from major sources in eight Florida cities suggest that simple time-series models such as the moving average and exponential smoothing may be appropriate for revising revenue forecasts within the fiscal year, a practice that has been shown to help local governments budget closer to the ever-tighter budget restraints they face. However, interviews with local officials and examination of their budget and accounting records confirm prior findings that limited forecasting expertise and data availability in local governments may limit the utilization of these methods. Moreover, the norm of revenue underforecasting, noted in this study and elsewhere, suggests a degree of risk aversion that may preclude experimentation with the models tested in the study. Ironically, this may place local officials at greater risk of misestimating revenue at a time when their revenue streams are becoming less reliant on the stable property tax and increasingly reliant on more volatile, economically sensitive sources such as fees and user charges.