Tools for assessing and forecasting regional poverty levels using a simulation model: the Republic of Bashkortostan

Under current conditions, the poverty rate has become one of the key indicators of the population’s social well-being. Poverty, in turn, varies considerably across the main sociodemographic groups of households, depending, for example, on household members’ employment, the number of dependents, social status, place of residence, and other factors. A specific feature of poverty in present-day Russia is that it is driven by structural factors, foremost among them the large share of low-paid jobs in the labor market. The study aims to develop effective tools for forecasting poverty rates using an agent-based approach. The article describes an agent-based model for forecasting poverty in the Republic of Bashkortostan. The concept of the simulation model is presented, and the types of agents, their parameters, and behavioral algorithms are discussed. The decision-making process for managing poverty levels is presented as an activity diagram created in Enterprise Architect. The model inputs include the results of sociological studies and population surveys, as well as regional socioeconomic indicators provided by the territorial offices of the Federal State Statistics Service. The model outputs are projected changes in the share of people living in poverty, their income levels, and the number of people with incomes below the minimum wage. The control parameters include benefits for large families, social benefits, targeted assistance for rural residents, the annual wage indexation rate, and the minimum wage. The article presents the results of scenario experiments aimed at reducing poverty in the Republic of Bashkortostan, including scenarios based on best practices from leading regions in the Volga Federal District and the Russian Federation. The article also provides practical recommendations for developing effective poverty alleviation policies at the regional level.