OPTIMIZING THE TAX BURDEN TO REDUCE THE SHADOW ECONOMY IN UZBEKISTAN: ECONOMETRIC EVIDENCE AND INSTITUTIONAL REFORM

Abstract

This article develops an integrated approach to optimizing the level, composition, and distribution of the tax burden as a strategy for reducing the shadow economy in Uzbekistan. The analysis combines descriptive fiscal statistics, a Multiple Indicators Multiple Causes (MIMIC) model, ordinary least squares (OLS), Scully-type optimization, threshold estimation, and autoregressive distributed lag–error correction (ARDL-ECM) techniques. The estimates indicate that the shadow economy accounted for 38.9% of GDP under the static MIMIC specification and 39.6% under the dynamic specification in 2025. Growth-oriented tax burden estimates range from 17.86% to 18.78% under the OLS and Scully approaches, while the threshold model identifies a value of 22.09%. Based on these results, a policy benchmark of approximately 19-20% of GDP is identified. The analysis also reveals variations in effective tax burdens under simplified taxation regimes for small businesses, opportunities to improve VAT C-efficiency (averaging 42.5% over 2021-2025), and prospects for strengthening risk-based tax collection and proactive tax administration. The findings provide a basis for a sequenced reform agenda, clearly distinguishing between econometric estimates, illustrative scenarios, and policy recommendations.