Digitalization, Government Revenue, and Structural Breaks: An ARDL Cointegration Approach for Uzbekistan
Abstract
Governments around the world see digital tools as key to upgrading public administration. They pour money into e-governance setups, hoping to tighten up fiscal operations. Yet hard data on how digital shifts actually affect government revenue, tax take in particular stay scarce for countries in transition. Uzbekistan, following its comprehensive 2017 reforms including the “Digital Uzbekistan – 2030” strategy and the 2020 Tax Code renewal, provides a unique natural experiment to examine how rapid, state-driven digitalization affects fiscal performance in emerging markets. We draw on quarterly figures for Uzbekistan, running from 2004Q1 through 2024Q4, with 84 observations. We construct a composite digital governance index using Principal Component Analysis from four indicators (e-government index, internet usage rate, mobile stations, fiber optic infrastructure). The Autoregressive Distributed Lag (ARDL) bounds testing methodology is applied to examine cointegration, complemented by structural break analysis (Bai-Perron, Zivot-Andrews tests), Toda-Yamamoto causality testing, and extensive diagnostic procedures including wild bootstrap validation. The ARDL bounds test confirms cointegration with F-statistic of 10.586 and t-statistic of –8.046 (both significant at 1% level). The error correction coefficient of –1.181 indicates rapid equilibrium adjustment. Our research indicates a significant U or J shaped correlation between digitization and government revenue. The research identified an inflection point in this relationship at 4.33 units of the focused digital governance index. Below this barrier, digitization incurs efficiency costs. A structural break study highlights the 2016Q4 as a key reform phase. GDP elasticity (1.345) confirms strong fiscal economic linkages. Tax administration digitalization shows long-term positive effects on revenue collection. Such patterns echo the public sector productivity paradox. Early outlays on digital tax platforms may squeeze budgets before payoffs arrive. Our results match broader studies on digital tax systems, yet provide unique insights for transition economies undertaking rapid institutional transformation. That threshold matters for any transitional country that sequencing digital spending and fiscal forecasts.