An econometric assessment of how science parks, technology parks and innovation technology platforms shape the volume of high-tech production across 25 regions of Ukraine — using a Tobit model built for censored data.
Read the study →The article provides a rare empirical assessment of innovation-ecosystem effectiveness in a country undergoing extreme external shocks and post-war transformation.
Amidst the new challenges facing Ukraine's economy, the problem of finding and implementing effective tools for structural modernization and the transition from a resource-based to a high-tech economic model has become a pressing issue. A key stage of this modernization is the formation of regional innovation ecosystems. According to the theoretical approach, innovation ecosystems are not merely a collection of separate structural elements, but a dynamic system of interconnected economic and social actors that, through cooperation networks, foster the creation of innovations, attract investment and enhance economic competitiveness.
The purpose of this study is to assess the impact of the infrastructural elements of the innovation ecosystem on the volume of high-tech production in the regions of Ukraine, which will allow for the identification of the most effective tools for stimulating innovation activity. An econometric analysis was conducted using regional statistical data and employing a Tobit model to account for the specific nature of the data. The results confirmed the research hypothesis that the considered infrastructure objects have a positive and statistically significant impact.
The analysis revealed that science parks demonstrated the largest positive effect, confirming their key role in generating innovations, while innovation technology platforms and technology parks also have a substantial, albeit somewhat smaller, impact on the regional development of the high-tech sector. Thus, the study empirically proves that the greatest returns are generated by those infrastructure objects that create and support links between the key actors of the ecosystem, namely science institutions and business. This confirms that synergistic effects within the ecosystem arise precisely at the intersection of fundamental knowledge and its practical commercialization. The obtained conclusions can serve as a basis for formulating priority measures of state policy for stimulating innovation development in the context of ensuring the post-war recovery of Ukraine's economy.
In the context of global bifurcations, the structural transformation of the economic system becomes particularly pressing. A resource-based export model has inherent vulnerabilities that, in the long run, limit growth and increase dependence on external market conditions.
A resource-based export model has inherent vulnerabilities that, in the long run, limit the potential for economic growth and increase dependence on external market conditions.
A key component of the transformation is forming a knowledge-based economy, which involves the prioritized development of high-tech and science-intensive sectors and the development of human capital.
Despite innovation-driven development being a priority, there are few empirical studies assessing ecosystem effectiveness — and the most influential factors stimulating high-tech production at the regional level remain unidentified.
The issue of identifying and substantiating effective public-policy instruments to stimulate innovation activity at the regional level acquires special theoretical and practical significance.
The concept of the "ecosystem" was borrowed from biology, where it was first introduced into scientific discourse by the British botanist A. Tansley in 1935. This study adopts the definition proposed by Granstrand and Holgersson.
"An evolutionary set of actors, activities and artefacts, as well as the institutions and relations — including complementary and substitute relations — that are important for the innovative performance of an actor or a population of actors."
— Granstrand & Holgersson (2020)
The organizations and individuals engaged in innovation activity.
R&D, financing, product development, commercialization, education and policymaking.
Technologies, products, services, patents, software and scientific publications.
Formal "game rules" — laws, regulations, IP rights — and informal cultural norms and business practices.
Cooperation and complementarity, but also competition and substitution between actors.
These infrastructural entities create the physical and virtual spaces for interaction, commercialization and innovation — and are the independent variables tested in the model.
Specially organized territories bringing together innovative companies, universities and research institutes. They typically have closer ties to universities and research institutions.
SP · academia-linkedMore oriented towards industrial centres and integration with corporate R&D departments, promoting a culture of innovation and enhancing business competitiveness.
TP · industry-linkedPredominantly digital environments that serve as virtual infrastructure, eliminating geographical and organizational barriers among globally distributed participants.
ITP · digitalThe analysis draws on statistical data from 25 administrative-territorial units of Ukraine for 2023. The temporarily occupied territories of Crimea and Sevastopol are excluded due to the absence of official statistics for the period.
Share of each region in Ukraine's total volume of realized high-tech production, 2023 (%). Kyiv city alone accounts for almost two-thirds of national output.
Distribution of the 57 innovation-infrastructure objects across five geographical macroregions of Ukraine, 2023 (%).
Number of objects by type for the regions that host any infrastructure. Kyiv city dominates with 12 science parks and 7 technology parks.
| Region | Output_high, thsd. UAH | Science parks | Technology parks | Innovation tech platforms |
|---|
High-tech production is absent in 5 of the 25 units. This data structure makes ordinary least squares on a logarithmized dependent variable inappropriate — it would drop those observations and create sample-selection bias.
To obtain consistent and unbiased estimates, a Tobit model was chosen. Based on the maximum-likelihood method, it is theoretically sound for analysing dependent variables censored at a certain level, allowing simultaneous assessment of a factor's impact on both the probability of observing a positive production volume and on its expected level. All estimations and diagnostic tests were performed using the Gretl software package.
Variance inflation factors were all well below the critical threshold of 10 (maximum VIF < 2.0).
Residuals deviated from normality and heteroskedasticity is likely in cross-sectional data, so robust QML standard errors were used.
Re-estimating the model without Kyiv city (the largest residual) left all key variables with the same sign and significance.
The likelihood-ratio test confirms overall significance, with a Chi-squared p-value below 0.001.
All three elements of the innovation ecosystem have a positive and statistically significant impact on the volume of high-tech production — but their strength differs markedly.
The most powerful driver. Underscores the critical link between R&D activities, academia and the real sector for generating innovative products.
A strong impact, confirming their effectiveness as platforms for commercializing technology and supporting innovative projects.
A significant positive effect, completing the picture and confirming their role in stimulating high-tech production.
Average increase in expected regional high-tech production (UAH) from one additional object of each type, ceteris paribus.
| Dependent variable: Output_high* | Coefficient | Std. error | z | p-value |
|---|---|---|---|---|
| const | −2.02959e+06 | 688673 | −2.947 | 0.0032 |
| ITP 1% | 866672 | 315839 | 2.744 | 0.0061 |
| SP 0.1% | 2.61639e+06 | 291735 | 8.968 | <0.0001 |
| TP 5% | 735932 | 293820 | 2.505 | 0.0123 |
| Chi-squared(3) | 126.4402 | p-value | 3.16e-27 |
| Log-likelihood | −326.8490 | Akaike criterion | 663.6980 |
| Schwarz criterion | 669.7923 | Hannan-Quinn criterion | 665.3883 |
Science parks are identified as the most influential element of the ecosystem. This implies that state regional development policy should focus not merely on creating infrastructure facilities, but on ensuring their close ties with scientific and educational centres.
The close "relationships" between the key "actors" — scientific institutions and business — cultivated within science parks are crucial for creating innovative new products. This underscores the importance of prioritizing state support for science-oriented infrastructure that transfers knowledge and technology from academia to the real economy.
The study uses cross-sectional data for a single year (2023). It cannot control for time-invariant unobserved regional characteristics (e.g. entrepreneurial culture), nor for time-varying factors such as foreign direct investment or specific market conditions.
A promising avenue is the transition to panel-data analysis for more reliable causal estimates. Future work could also assess the socio-economic impact of infrastructure on labour markets and demography, and the role of the ecosystem in enhancing national resilience and defence capabilities.
Innovation policy should be strategically directed towards cultivating deep, collaborative linkages rather than merely increasing the number of isolated infrastructural objects — critical for the efficient allocation of limited resources and sustainable, innovation-driven modernization.