The Adoption-Substitution Gap: Administrative Evidence from Swedish E-Bike Subsidies

Policymakers increasingly use green subsidies to advance Sustainable Development Goals (SDGs) related to climate, mobility, and health. These programs assume a causal chain: subsidies drive technology adoption, which induces behavioral substitution that then generates sustainability impact. We test this mechanism in the context of a nationwide e-bike subsidy program in Sweden, focusing on the adoption-substitution gap—the friction between acquiring green technology and displacing high-carbon behaviors. Linking administrative data on all 90,000 recipients to vehicle registry and insurance records, we find that while the subsidy successfully doubled e-bike sales with complete price pass-through, behavioral substitution was minimal. Objective vehicle-kilometers traveled (VKT) changed marginally, leading to negligible emissions reductions and health gains. Notably, survey-based substitution estimates within the same population substantially overstate realized reductions, suggesting a reporting bias in stated-preference data. Our findings highlight the necessity of targeting green subsidies toward populations whose behavior is most elastic to ensure cost-effective climate and mobility outcomes.