Using distributed lag models to predict regional budget revenues

Abstract

Subject. The article addresses projections of regional budget revenues, using distributed lag models. Objectives. The purpose is to review economic and statistical tools that are suitable for the analysis of relationship between the revenues of the regional budget system and regional macroeconomic predictors. Methods. The study draws on statistical, constructive, economic and mathematical methods of analysis. Results. In models with quantitative variables obtained under the Almon method, the significant predictors in the forecasting of regional budget revenues are determined mainly by the balanced financial result, the consumer price index, which characterizes inflation processes in the region, and the unemployment rate being the key indicator of the labor market. Models with quantitative variables obtained through the Koyck transformation are characterized by a wider range of predictors, the composition of which is determined by the peculiarities of economic situation in regions. The two-year forecast provides the average lag obtained during the evaluation of the models. The exception is the impact of unemployment rate, which is characterized as long-term. Conclusions. To generate forecasts of budget parameters, the results of both the Koyck method and the Almon method should be considered, though the former is more promising.