Résumé
We study how spatial connectivity and fragmented ownership affect the management of mobile common-pool resources. We develop a dynamic two-patch model of a mobile renewable resource: it predicts that (i) management efficiency in a given patch declines when the neighboring patch is managed by multiple agents rather than a single owner, and (ii) greater resource mobility amplifies inefficiencies at the global scale, especially under mixed ownership. We test these predictions in a preregistered laboratory experiment with 294 participants by varying mobility rates and ownership structures. Consistent with theory, shared management in one patch reduces efficiency at the local scale (in the adjacent patch) through spillover effects, and higher resource mobility further erodes efficiency at the global scale. Using group extraction trajectories, we identify three robust behavioral patterns: early over-extractors, mid-period preemptive groups, and late preservers. Mixed ownership and higher mobility shift the distribution toward early over-extraction and away from late conservation. Beyond these institutional and ecological effects, cognitive ability is strongly associated with more forward-looking extraction paths and higher payoffs.