ECONOMETRIC MODELING OF THE AVERAGE TARIFF RATE FOR FREIGHT RAILWAY TRANSPORTATION UNDER INDUSTRY DIGITALIZATION

Abstract

The paper presents an econometric modeling of the average tariff rate for freight railway transportation based on official statistical data of JSC Russian Railways (RZD) for 2016–2025. A factor space of 24 indicators grouped into six thematic blocks is formed: volume and structure of transportation, rolling stock utilization, technical equipment, personnel motivation, innovative development, and digitalization. A correlation analysis is performed, and the factors with the strongest relationship to the tariff rate are identified. A multiple linear regression model with the coefficient of determination R² = 0.984 is constructed. Diagnostics of the ordinary least squares (OLS) assumptions is conducted: normality of residuals (Jarque–Bera test), homoscedasticity (Breusch–Pagan test), autocorrelation (Durbin–Watson statistic and Breusch–Godfrey test). It is established that the locomotive daily productivity and wage growth rates have the most stable influence on the tariff level. A medium-term forecast of the average tariff rate for 2026–2030 is made under four scenarios: baseline, intensive, conservative, and regulatory-cost shock. The results can be used to justify tariff policy under industry digital transformation.