This paper assesses the efficiency of AI adoption among small enterprises in Türkiye and 30 peer EU and EU-neighborhood countries using data envelopment analysis (DEA) for 2021 and 2023. The results reveal pronounced cross-country heterogeneity and sizable scale inefficiencies: efficiency scores differ markedly between CRS and VRS specifications, while orientation choices matter little. For Türkiye, near-unity VRS scores alongside much lower CRS scores indicate high pure technical efficiency but a persistent scale gap relative to frontier economies. These findings underline the need to prioritize scalable and inclusive AI diffusion among small firms rather than merely expanding generic digital capacity.
Authors
Muhammad Salar Khan
Assistant Professor of Public Policy, Rochester Institute...
Authors
Abdullah Altun
Associate Professor of Economics, Gebze Technical University,...
