Abstract <p>The article examines the features of manifestation of spatial spillover effects in the Russian economy and their relationship with the economic activity level and degree of its difference in regions. The study is based on analysis of spatial determinacy of the distribution dynamics of regional indicators using methods based on the theory of discrete-time Markov processes with a continuous state space. It is shown that not only a low economic activity level, but also a strong difference in its level is an obstacle to positive spatial externalities: regions that significantly lag behind their neighbors do not demonstrate accelerated development and increase the lag both from their neighbors and from the average for the entire population. Conversely, a small degree of difference between neighboring regions and an above-average economic activity level in them contribute to spatial spillover effects, the convergence of economic activity levels and, as a consequence, positive spatial autocorrelation of economic activity levels in regions. In this case, the greatest acceleration is achieved by regions that catch up with their neighbors, while the leaders somewhat slow the pace of development. Thus, spatial autocorrelation, which reflects the degree of similarity between neighboring regions, can be a consequence of spatial concentration or dispersion of economic activity, as well as the cause that intensifies these processes.</p>