Research report: Institutions, law, and resilience: the political economy of tourism-sector reform in Uzbekistan’s integration into the global economy

Uzbekistan’s international tourist arrivals rose from 2.0 million in 2016 to 6.75 million in 2019 and 8.18 million in 2024, changes that coincided with a package of currency, visa and tourism-law reforms. A comparative study of Uzbekistan and seven post-Soviet economies found that arrivals during 2017–2019 exceeded the trajectory predicted by a synthetic control by approximately 46%. The authors describe this as a large estimate with limited statistical certainty, saying the evidence is consistent with institutional reform contributing to growth and resilience but does not establish causation.

The Brief Research Report examined reforms enacted between 2016 and 2019: the som was allowed to float in September 2017, an e-visa system and expanded visa-free access followed in 2018, and a national tourism law and development concept took effect in 2019. The authors used international tourism as an indicator of regulatory change because the sector responds to administrative barriers encountered by foreign visitors, including currency exchange, entry procedures and legal protections. Tourism is also highly exposed to external disruption, allowing the researchers to examine whether Uzbekistan’s reformed sector recovered from the COVID-19 shock as quickly as, or faster than, regional comparators that had not reformed to the same degree.

To place the reforms in a broader regional context, the researchers assembled a 2010–2024 panel covering Uzbekistan, Kazakhstan, Kyrgyzstan, Tajikistan, Georgia, Armenia, Azerbaijan and Moldova. They combined four analytical approaches: interrupted time-series modeling, synthetic control, difference-in-differences with wild-cluster bootstrap inference, and fixed-effects panel regression. The synthetic-control design constructed a weighted “synthetic Uzbekistan” from Azerbaijan, Kyrgyzstan, Tajikistan, Georgia and Armenia, reproducing the country’s pre-reform arrival trajectory before comparing it with the observed path after 2017.

The other models accounted for common trends, macroeconomic conditions, country-specific differences and a change in the statistical definition of arrivals after 2020. The synthetic control reproduced Uzbekistan’s pre-reform pattern closely, with a pre-period root mean squared prediction error of 0.04, but fell progressively behind the observed series after the reforms began. Between 2017 and 2019, before the pandemic and the arrival-counting change, Uzbekistan averaged 0.38 log points above this counterfactual, corresponding to approximately 46% more arrivals.

When each of the eight countries was treated in turn as though it had experienced the reform, Uzbekistan had the largest divergence, ranking first of eight. However, with only eight possible donor-pool rankings, the smallest attainable permutation p-value was about 0.13, which is below conventional statistical significance. The authors therefore characterize the result as a substantively large effect with limited statistical certainty rather than a statistically significant effect.

Other estimates pointed in the same direction without resolving that uncertainty. The interrupted time-series analysis identified a positive post-2017 level shift of 0.145 log points, but the estimate was imprecise, with a p-value of 0.57, because the model contained only 15 annual observations and arrivals were already increasing before the reforms. A difference-in-differences model comparing Uzbekistan with Kazakhstan, Kyrgyzstan and Tajikistan implied a reform-associated premium of roughly 18%, yet its wild-cluster bootstrap p-value was 0.75.

That design included only four clusters, making conventional cluster-robust inference unreliable. An artificial break assigned to 2014 produced a small, insignificant interaction, consistent with broadly parallel pre-treatment trends, but the placebo test did not establish the size of the original reform estimate. All six Worldwide Governance Indicators improved over the study period and were positively correlated with log arrivals, with correlations ranging from 0.61 to 0.78.

Regulatory quality had a correlation of 0.74, government effectiveness 0.78, rule of law 0.75 and control of corruption 0.73. Voice and accountability had a correlation of 0.71, while political stability had the lowest correlation at 0.61. In the fixed-effects analysis, a one-standard-deviation improvement in regulatory quality corresponded to approximately 16% higher arrivals, although the coefficient lost precision once country fixed effects, a common trend and the measurement-break control were included.

Institutional quality, visa liberalization, currency reform and the broader growth of arrivals moved together, preventing the panel model from separating their effects. The association was strongest for long-haul visitors from outside the Commonwealth of Independent States, who were less likely than regional travelers to be affected by movements within the region. Non-CIS arrivals increased about 6.5-fold, from 0.15 million in 2016 to 0.98 million in 2024, and their correlation with regulatory quality was 0.80.

The authors also found that voice and accountability improved from an extremely low base but remained the lowest governance dimension in level, reaching –1.0 in 2024. They interpret the pattern as consistent with administrative and regulatory sources of institutional credibility rather than democratization being the binding constraint on tourism. This interpretation remains descriptive because the study cannot isolate which component of the reform package produced the observed changes.

The study separately examined resilience by indexing arrivals to their 2019 level. International arrivals fell 78% in 2020, then recovered to within 2% of the 2019 level by 2023 and approximately 21% above it by 2024. Uzbekistan’s rebound closely resembled those of Kazakhstan and Tajikistan and was stronger than Georgia’s, which remained below its 2019 total in 2024.

The evidence shows that Uzbekistan recovered at least as quickly as several regional peers. It does not show that Uzbekistan was uniquely resilient, and the recovery cannot be attributed specifically to the reforms. Important measurement and design limitations constrain the interpretation.

Data for 2010–2019 measure World Bank and UN Tourism overnight visitors, while the 2020–2024 national border-entry series covers individuals arriving from abroad, a broader category dominated by short visits. The researchers included an explicit break indicator and re-estimated key models using only consistently measured pre-pandemic data; the qualitative results were unchanged. In the pre-2020 fixed-effects model, the regulatory-quality coefficient was 0.44 with a cluster-robust p-value of 0.40, while log GDP was positive and significant at 0.50 with a p-value of 0.02.

Approximately 80% of arrivals came from neighboring CIS countries, and the majority of visits in 2023 were for family purposes, so regional mobility complicates interpretation. The non-CIS analysis reduced but did not eliminate this issue because long-haul arrivals remained small in absolute terms. Several 2024 values were also provisional, and the Worldwide Governance Indicators methodology was revised in 2025.

A further complication is the possibility that currency depreciation helped tourism independently of legal credibility. The 2017 float roughly halved the som’s dollar value, potentially making Uzbekistan cheaper for foreign visitors. The study’s GDP and growth controls do not capture that relative-price effect, and the non-CIS analysis cannot separate cheaper travel from greater confidence in the country’s institutions.

The authors note, however, that the depreciation was a one-time level change, while arrivals continued accelerating through 2018 and 2019 after implementation of the e-visa and tourism-law measures. They also point out that currency convertibility was itself a legal-institutional commitment rather than a purely macroeconomic change. Their overall conclusion is deliberately restrained: legal codification, currency convertibility and simpler entry procedures coincided with a large and statistically uncertain tourism gap, followed by a recovery comparable to neighboring countries.

The bundled reforms, limited donor pool, changing measurement and low statistical power prevent a causal conclusion.

Subject of Research: Political economy of institutional reform and resilience in Uzbekistan’s tourism sector

Article Title: Institutions, law, and resilience: the political economy of tourism-sector reform in Uzbekistan’s integration into the global economy

Article References: Urakova, D. B., Usmanova, A. B., Khamidov, O., Ergashev, J. Y., Murodova, D. J., & Ruziyeva, G. F. (2026). Institutions, law, and resilience: the political economy of tourism-sector reform in Uzbekistan’s integration into the global economy. Frontiers in Political Science, 8, Article 1919183. https://doi.org/10.3389/fpos.2026.1919183

Image Credits: AI Generated

DOI: 10.3389/fpos.2026.1919183

Keywords: Uzbekistan, tourism arrivals, institutional reform, legal reform, economic resilience, synthetic control, regulatory quality, Central Asia, COVID-19 recovery, political economy, Institutions, resilience