Subject. The article reviews the manufacturing industry in Russian regions, calculates the indicators of regional industrial specialization needed for development of econometric models of spatial panel data.
Objectives. The purpose is to create a methodology for analyzing the regional industrial specialization based on econometric tools; to test it, using the case of the manufacturing industry, for determining the type of externalities in the Russian Federation.
Methods. To build econometric models, we use methods of least squares and maximum likelihood. We apply localization ratios to assess regional industrial specialization in terms of the volume of employment, revenue and investment in manufacturing, workforce productivity, etc.
Results. The findings show the clustering of regions by the level of productivity. The localization of manufacturing industry in regions in terms of localization of employment and localization of productivity is negatively related to productivity in the region. This can be explained by the transition of regional economies to the post-industrial mode, where the service sector becomes more important, and by possible over-industrialization and specialization of certain regions in the context of the need to develop related sectors and to build links between them. The presence of direct negative MAR externalities may indicate a need for further research in positive Porter and Jacobs externalities for Russian regions manufacturing industry.
Conclusions. The developed methodology enables to identify and analyze relationships between regional industrial specialization and regional indicators; to specify the type of externalities and determine the existence of indirect and direct effects of industry localization.