Long -term growth sources for sectors of Russian economy

Abstract

Theoretical models suggest stationary structure of sectors. Sometimes this suggestion is hidden (balanced growth). The ratio of variables for 2 sectors is unit root at the most cases (for 14 Russian sectors and 6 variables per sector). The lowest share of stationary ratios is 5/91 for real value added with ADF test (KPSS test for the same variable leads to 38/91 stationary ratios). The cointegration rank differs across sectors in wide ranges (from 1 for trade (G) or government administration (L) till 5 for agriculture (AB)). The dynamic stochastic partial equilibrium (DSPE) model is created. It is model of firms in DSGE-style and description of the rest economy by exogenous rules. The model is estimated for each of 14 sectors. The model includes 5 sources of stochastic trends: TFP; labor supply; investments efficiency; investments prices; prices of intermediate goods. Any 2 sectors significantly differ by key parameters (production function shares, capital depreciation, and demand elasticity). The drift of unit root sources differs across sectors (including sign). Only few pairs of sectors differ insignificantly (3/182 or 8/91 depending on test specification). The variance decomposition of trends (for various variables) is computed. It varies in wide ranges across sectors and variables. Thus, usage of aggregate data in theoretical model leads to loose of large amount of information.