Clustering Russian Regions by Innovative Outputs Using a Multi Indicator Approach

In this research, we aim to discover empirically the similarity of Russian regions in terms of innovation activity forming several groups of them. Design/methodology/approach: We are using clustering technique, namely hierarchical clustering and map Russian regions based on the cluster memberships, which provides measure of similarity between all regions of Russia. We employ the region-level dataset of 83 Russian regions for the period from 2010 to 2015. Findings: We find overall strong stratification of the Russian regions, with most of the Russian regions falling into clusters with low innovation outputs. We also find that over the time cluster memberships do not change for low-innovative regions, while for more innovative ones they do. It may imply that regions with higher innovative outputs increase them significantly faster than other regions. Research/practical implications: Our research contributes to better understanding of Russian regional development and has implications for regional policy decisions. We argue that stimulating innovations in leading regions may be beneficial for development of the neighboring regions as well, due to spillover effects. We understand the limitations of the research and propose to investigate further the innovation inputs, as it may provide insight on why regions are performing the way they are. Originality/value: This paper uses multi indicator approach to assess similarities in innovative output of the Russian regions. Our results provide an original insight of the current state of the innovations and their geographical distribution in Russia. This research may be of a practical use to the managerial decisions, connected to the location choice, as well as for adjusting current innovative policy in Russia.