Understanding AI Adoption in Local Government: Capacity Constraints or Political Caution?

Artificial Intelligence (AI) has been widely adopted by governments, but whether and to what extent governments adopt AI varies. A key question is whether variation in AI adoption across local governments stems from differences in capacity or political constraints. Understanding this has implications for the kinds of policies officials may want to adopt to ensure that local governments have access to the technologies that they want and need. This paper combines data from the International City/County Management Association (ICMA) Survey, American Trends Panel (ATP), the Current Population Survey (CPS), and other public data sources, to examine how capacity and political constraints influence local government AI adoption in the United States. Using logistic and Poisson regression models, we find that AI adoption at local governments is associated with citizens’ attitudes toward AI, public employee unionization, local governments’ financial capacity, and infrastructure support. Notably, political factors (citizens’ attitudes and public sector unionization) are equally influential as capacity factors in explaining government AI adoption, and in some cases, more powerful. These findings suggest that variation in AI adoption across local governments results not only from capacity barriers but also from political constraints due to the power of public employees and concerns of the public.