Integration of input–output approach into agent-based modeling. Part 1. Methodological principles

Abstract

The article presents a comparison of the Agent-Based Multiregional Input–Output Model (ABMIOM) of the Russian economy and a small-size version of the Optimization Multiregional Input-Output Model (OMIOM) from the per-spective of their possible integration for spatial economic research. These models share many similar features, technically and conceptually. The two models regard the economy as a complex system of interdependent compo-nents; both have the same spatial and sectoral structure, similar production technologies. Both allow to perform a structural analysis of processes taking place in the economy in terms of industries and territories. The former is a simulation model, the latter is a linear optimization model; their major differ-ences relate to the methods of modeling the market and price mechanisms and of representing the space. We propose to use a spatial input-output model based on the real data as the source of disaggregated information at the stage of initialization of agent-based models, which opens opportunities of gradual turning them into an experimentation area for the real economy, rather than an instrument for an analysis of an artificial economy.