THE ECONOMIC EFFECT OF ARTIFICIAL INTELLIGENCE ADOPTION IN RUSSIAN REGIONS: ASSESSING THE IMPACT OF AI ON PRODUCTIVITY AND GROSS VALUE ADDED

Abstract

Expectations about artificial intelligence are usually formulated at the level of the national economy, while decisions on supporting it are increasingly made by regions. Regional statistics, however, do not separate the contribution of artificial intelligence to gross value added, and the assessment of its effect is replaced by counting organizations that use the technology. The aim of the article is to develop an approach to estimating the contribution of artificial intelligence to regional value added growth that distinguishes the channels of its impact. The methodology combines a decomposition of value added growth into three channels — production of AI solutions, accumulation of related capital and total factor productivity growth in user industries — with an analysis of empirical studies of regional heterogeneity of adoption, sectoral effects and barriers. A calculation for three conditional types of regions shows that the contribution of AI to annual value added growth may differ more than sixfold: from 1.05 percentage points in a financial and service center to 0.16 percentage points in an agricultural region. The structure of the contribution also differs: in a larger center more than half of the effect comes from producing AI solutions, while in industrial and agricultural regions most of it comes from productivity growth among users. It is argued that for most regions AI policy should focus on diffusing technologies in their industries of specialization rather than on building their own development sector.