Macroeconometric Simulation às à Method îf Regional Studies

The article presents the evolution of regional macroeconometric models. The authors allocate two types of models which differ in the complexity level of approximable interrelations - aggregated and integrated ones. This paper shows the advantages of integrated models over aggregated ones due to a combination of techniques of multivariate econometric and balance analysis. The algorithm of creation of integrated models includes three stages: defining a dynamic kernel, structural kernel and the technical way of their integration. The authors pay attention to the implementation results of the first stage of the algorithm - the creation of dynamic kernel of integrated model. The dynamic kernel includes equations which characterize relations between economic agents of the regional system and which are merged in the four structural units: consumption, output, employment and prices and incomes. On the basis of multivariate cointegration analysis the researchers obtain estimates of the aggregate model of economic dynamics of Khabarovsky Krai. The peculiarity of this model is high accuracy of approximation of regional dynamics. The authors summarize the simulation results in the form of assessments of short-term and long-term elasticity coefficients which reflect the relations between regional macroeconomic indicators.