Methods of Modeling and Analysis of Employment in Cities, Taking into Account the Spatial Factor

The article presents the results of statistical analysis and modeling of employment in cities of the Russian Federation, taking into account the factor of their territorial location. A balanced system of target and factor indicators of employment in the cities of the Russian Federation developed based on the Rosstat database of municipal indicators is substantiated. Using methods of cluster analysis of spatial data, estimating the values of Moran's index at various distance intervals of cities, the statistical heterogeneity of spatial autocorrelation of employment in cities was proven. Based on the construction of spatial autoregression models, the authors established general and specific factors of mutual influence of employment levels for the entire set of cities and cities located in the zone of socio-economic influence of Moscow. The modeling results were obtained using a SAR-type model – a model with a linear additive specification in which spatial relationships between objects are specified exogenously. Recommendations are given for using other types of spatial autoregression models, which differ in the presence of a connection between the spatial autocorrelation coefficient and individual components of the multivariate regression model. Based on the results of the study presented in the article, conclusions were drawn about the directions of employment regulation in cities, taking into account the factor of their remoteness from other cities and the possibilities for further development of spatial factor modeling of employment at the municipal level.