Association for Public Policy Analysis & Management

Recent trends in housing development toward new-urbanism, where residential areas offer their own private amenities, community services, and common space may disrupt the efficiency of public goods markets as individuals weigh these housing development amenities against city-provided public amenities. The Tiebout sorting model predicts that the efficient provision of local tax and spending bundles via market forces arises when individuals “vote with their feet”, that is when they migrate to municipalities that offer the most attractive bundles. However, the presence of competing club goods may disrupt the market by incentivizing consumers to locate in areas with less desirable tax and spending bundles as they substitute developer-provided club goods for city-provided public goods.

We examine this question using a conjoint experiment, a methodology that allows us to make strong causal inferences in multi-attribute choice problems. Specifically, we run a series of experiments to explore how subjects make choices among hypothetical apartment homes located in cities with varying public and club good attributes. Conjoint experiments allow us to vary both apartment community amenities and city amenities independently for comparative testing of whether private or public amenities are more influential in shaping residency decisions. By varying multiple types of city amenities, we can also determine whether certain types of public amenities are more likely to be eclipsed by planned community amenities. Finally, we consider whether personal characteristics of housing consumers moderate the importance of publicly provided amenities versus community amenities.

The results of our analysis are consequential for understanding how the individual consumer values public goods versus club goods. Our results have implications for understanding whether residential enclaves might compete with or change the incentives for migration that influence the efficient provision of city-provided public goods. If we find that club goods are as influential on housing choices as city-provided public goods, the results would suggest that the market for public goods may produce an inefficient tax and goods bundle for those living in the same local governance area but outside the club communities. From the perspective of developers, our results speak to whether developer-provided club goods can substitute for city-provided amenities, thereby opening opportunities for developers to purchase cheap land tracts in areas lacking in city amenities in exchange for cheaper taxes and land costs.