The Capacity–Realization Gap: Scarce Complements and the Microeconomics of AI Productive Capacity

This paper develops the Capacity–Realization Gap framework for understanding why extraordinary growth in AI productive capacity may coexist with comparatively modest observed effects on GDP, productivity, welfare, and fiscal outcomes. Building on recent AI-economy measurement research, the paper proposes a microeconomic model in which AI capacity is constrained by scarce complementary factors including human capability, organizational capital, trust, adoption, and institutional integration. The article reviews adjacent literatures, introduces formal bottleneck and elasticity formulations, and discusses implications for AI measurement, AI satellite accounts, and future empirical research.

Keywords 

artificial intelligence

AI economics

productivity

economic measurement

AI GDP

productive capacity

organizational capital

technology adoption

innovation economics

microeconomics

economic growth

AI policy