SPATIAL ECONOMETRIC MODELING OF ECONOMIC GROWTH IN RUSSIAN REGIONS: DO INSTITUTIONS MATTER?

Abstract

As a rule, regions of the same country do not develop independently, but influence each other. When modeling the economic growth of regions, it is necessary to take into account such mutual influence, since, for example, with the growth of one of the regions, neighboring regions can grow as well (i.e. there is cooperation) or, conversely, slow down their growth (due to competition), or not react in any way. Omission of the corresponding variables (spatial factors) can lead to a bias in the estimates of the model parameters. As a rule, spatial econometric models are used to take into account the relevant spatial factors. These models differ from classical regression models by the presence of spatial lags of the dependent and independent variables. The coefficients for these spatial lags are usually assumed to be constant. Thus, the degree of influence of neighboring regions on the chosen one does not depend on the characteristics of this region, which seems to the authors to be an excessively strong assumption in models for heterogeneous countries, such as Russia. The authors weakened this assumption by replacing constant coefficients of spatial lags with linear functions characterizing the quality of institutions and the degree of business activity in the region under consideration. The level of entrepreneurial activity in the region, the index of the region’s provision with banking services, and the investment potential of the region were chosen as such variables. The number of estimated parameters increased insignificantly, but with the help of the modified model the authors were able to test and empirically confirm the hypothesis that the higher the quality of the institutional environment and the degree of business activity of a given region, the more sensitive it is to the influences of neighboring regions.