The Evolution of Governance Structures in a Polycentric System

The Tiebout competition model is often criticized for its unrealistic assumptions. We develop an imperfect Tiebout competition model in which households have no information about other jurisdictions (moving decisions are blind), and local jurisdictions operate as revenue-maximizing Leviathans. We show that, even under such harsh assumptions, jurisdictions will not increase taxes without also increasing the quality of their public services. The model also opens the door for understanding various possible vicious spirals, e.g. as a result of factor prices shocks, co-production problems, and income-based sorting. We also show that, in general, the model does not lead to the calculus of consent optimum, which opens the possibility for conceptualizing which constitutional rules would tend to constrain the system to evolve towards optimum levels of centralization or decentralization across all public issues. Last but not least, because the model involves positive moving costs, we explain the origins of voice, as an alternative to exit. When people do not engage in exit, either because they still hold up hope that public services will improve or because they do not have the resources to move, they increase their involvement in other activities such as voice and co-production.