FORECASTING TAX REVENUES IN UZBEKISTAN USING MACROECONOMIC INDICATORS: AN INTEGRATED ECONOMETRIC AND TIME SERIES APPROACH

Abstract

This study provides a comprehensive econometric and time series analysis of tax revenue dynamics in Uzbekistan, with the objective of identifying key macroeconomic determinants and generating medium-term forecasts. Using annual data spanning the period 2010–2024, the research employs a multiple linear regression framework combined with ARIMA-based forecasting techniques to model the relationship between tax revenues and core macroeconomic indicators, including Gross Domestic Product (GDP), imports, and retail trade turnover. The empirical results indicate that the model exhibits strong explanatory power, with an R-squared value of 0.8993, suggesting that a substantial proportion of variation in tax revenues can be attributed to fluctuations in these variables. However, individual coefficient estimates reveal varying levels of statistical significance, indicating the presence of multicollinearity and potential structural complexities within the Uzbek economy. To address temporal dependencies and improve predictive accuracy, ARIMA(1,1,1) models are applied to each explanatory variable, and the resulting forecasts are used to estimate future tax revenues for the period 2025–2029. The findings suggest a generally upward trend in tax revenues, driven primarily by expansion in domestic consumption and trade activity, although signs of medium-term stabilization and fluctuation are also observed. The study contributes to the existing literature by integrating structural econometric modeling with stochastic time series forecasting in the context of a transition economy, while also providing practical insights for fiscal policy planning and revenue management in Uzbekistan