Application of nonparametric method for segmentation of regions by type of consumer behavior

The nonparametric method for constructing economic indices allows to test the consistency of observed statistical data with the mathematical Pareto model of a rational representative agent. The article proposes an approach to identify economic agents by the regional principle using a generalized nonparametric method. The approach is applied to Russian trade statistics data for 2012-2024. It is shown that in order to match with the Pareto model, regions must be grouped. 7 clusters of regions were identified based on the similarity of the structure of consumer behavior. It was revealed that the beginning of the pandemic was the reason for the inconsistency of the statistics of the capital group of regions with the model. In addition, the differentiation of the identified groups of regions by spending on food and non-food goods and services was studied.