INNOVATION CAPACITY AND REGIONAL INVESTMENT ATTRACTION IN UZBEKISTAN
Keywords:
Regional investment attractiveness; innovative economy; digital capacity; foreign investment; regional development; Uzbekistan JEL classification: F21; O30; R11; R58; C23Abstract
This study examines whether regional innovation-economy capacity is associated with stronger investment attraction across all fourteen first-level territories of Uzbekistan during 2017-2025. A balanced panel of 126 region-year observations is compiled from the National Statistics Committee’s SIAT system. Investment attractiveness is measured through revealed outcomes: fixed investment per capita and foreign investments and loans in fixed assets per capita. The Innovation Economy Capacity Index combines within-year standardized values of information and communication services per capita and gross regional product per capita. The empirical strategy distinguishes persistent between-region advantages from within-region changes using year effects, two-way fixed effects, and a correlated-random-effects (Mundlak) decomposition. Results show a strong and statistically significant between-region association: a one-standard-deviation higher long-run innovation capacity is associated with approximately 63% higher fixed investment per capita and 62% higher foreign investment per capita. Within-region improvement is also positively associated with fixed investment; the two-way fixed-effects coefficient is 0.751 (clustered standard error 0.232), with a wild-cluster-bootstrap p-value of 0.030. The corresponding foreign-investment coefficient is positive but imprecisely estimated. The findings indicate that investment attraction is not generated by isolated digital projects alone; it depends on cumulative productivity, connectivity, business capability, and credible project pipelines. Policy should therefore combine regional innovation infrastructure with investor facilitation, skills, supplier development, and transparent monitoring. The study provides a reproducible regional dataset and code, while explicitly treating 2025 observations as preliminary and avoiding causal claims.
Downloads
Published
Issue
Section
License

This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.






