Modeling of Economic Growth of the Russian Federation Regions with Spatial Effects

The spatial method for modeling the rates of economic growth assumes that economic growth depends not only on socially significant indicators and characteristics of the region, but also on the rates of economic growth of its neighbors. The main question of this study is who can be considered as neighbors. In this study, it was proposed to take into account spatial effects not only between regions that are geographically close to each other, but also regions that are similar in sectoral structure. For these purposes, an endogenous weighting matrix W was constructed. Three specifications of spatial autoregression were evaluated: 1) a model that includes a spatial lag built on the geographical proximity of regions, and a spatial lag built on the proximity of regions in the industry structure; 2) a model that includes only the spatial lag built on the geographical proximity of the regions; 3) models without spatial effects. The evaluation results showed that the coefficients estimates are biased in models that do not take into account the spatial lag, built on the proximity of the regions in the sectoral structure. Thus, the economic growth of the regions of the Russian Federation is influenced by the economic growth of the regions not only geographically closely located, but also similar in sectoral structure.