Variable Cluster Analysis of Employment in the Cities of the Russian Federation

Purpose of the study . The purpose of the study is to analyze the socio-economic factors of employment in the urban districts of the Russian Federation. Within the framework of the study, a cluster analysis of cities is carried out in two variants: (1) based on factor indexes of employment and (2) using both performance and factor indexes of employment. In both cases, a statistical justification is provided for taking into account the lags of the delayed impact of factor indexes on performance ones. The use of this variable approach in clustering allows us to identify statistically homogeneous groups of cities, respectively, without taking into account and taking into account the individual effects of the factor influence on employment and, on this basis, to justify the indexes of employment regulation in the cities of the Russian Federation. Materials and methods . To achieve this goal, a system of indexes was developed based on the data of municipal statistics provided by the Federal Plan of Statistical Work, including quantitative and qualitative assessments of the level of socio-economic factors of employment. The study conducted a cluster analysis using the hierarchical method (Ward’s method) and the iterative method (k-means). Results . As a result of the cluster analysis, statistically homogeneous subgroups of cities were identified that differ in the level of employment, administrative composition, set and nature of the influence of factor indexes on employment. Based on a comparative analysis of two clustering options, it was proven that identifying significant employment factors at the municipal level (by urban municipalities) is possible with clustering using both performance and factor indexes, as well as taking into account the time lags of their relationship. Clustering of cities only by factor indexes does not provide for taking into account the individual characteristics of interdependence of the studied indexes in cities, which leads to rough statistical estimates and the concealment of latent interrelations, the consideration of which is necessary for regulating employment. Conclusion . The study revealed that statistically homogeneous clusters of cities have specific administrative composition, which predetermines the characteristics of multifactor regression models of employment, the parameters of which are recommended for use in developing measures to regulate employment and the socio-economic development of cities.