Abstract:
The dependence of the emergent characteristic – returns to scale, calculated on the basis of production functions in the regions of the Russian Federation, on various indicators characterizing the regions, including macroeconomic indicators, characteristics of economic reproduction processes and socio-economic institutions, geographical indicators, etc. was investigated. The calculations were carried out using the method of optimal valid partitioning, which implies verification of patterns in the data using permutation tests. As a result of the analysis, a statistically significant relationship with returns to scale was revealed for a significant percentage of indicators or their combinations, which indicates both the real existence of the interdependence of these indicators with the returns to scale, and the sufficient accuracy of estimates of the parameters of the Cobb–Douglas model for the time series involved in the analysis.
Keywords:data mining, optimal valid partitioning, pattern verification, permutation tests, production functions, returns to scale, regional economics, economic reproduction.