The Size Distribution of Income in an Income Tax Revenue Forecasting Model

Abstract

This paper presents a personal income tax revenue forecasting model which incorporates a functional specification of the entire size distribution of income estimated using aggregate data. The model is shown to yield substantially more accetrate forecasts for Minnesota than other revenue forecasting models which do not incorpocate a functional specification of the income distribution. Although the choice of arevenue forecasting model is somewhat subjective, our results indicate that the general approach of modeling the entire income distribution by a functional form is highly desirable in terms of forecast accuracy.