Probabilistic models of estimation of employment in the regions of the Russian federation

the present study is devoted to the development and application of probabilistic discrete choice models for assessing the employment level in the constituent entities of the Russian Federation. The relevance of the work is due to the significant regional heterogeneity of the Russian labor market and the problem of labor resource imbalances, which hinders the effective use of the country’s potential. The aim of the study is to quantitatively estimate the probability that the regional employment level will exceed the all-Russian benchmark value (63.675%) based on key socio-economic indicators. As the primary analytical tool, a logit model (binary logistic regression), belonging to the class of discrete choice models, was used. Modeling was conducted on panel average-monthly data for the period 2014-2024, obtained from official sources (Rosstat, EMISS). The original dataset included twelve macroeconomic indicators, such as average monthly wages, cost of the consumer basket, volume of household deposits, industrial production index, crime rate, housing construction, and others. The resulting binary variable was formed by comparing the actual employment level with the benchmark. Model parameters were estimated by the maximum likelihood method using the Gretl statistical package, with stepwise exclusion of statistically insignificant variables (p-value > 0.05).

logit model, discrete choice, employment, labour market, macroeconomic indicators, regions of the Russian Federation