Public Acceptability and Land Value Taxation

This research seeks to use economic experiments to investigate how different property tax institutions affect homeowners’ investment behavior and whether land value taxation (LVT) is acceptable to homeowners and voters. The experiment is based on an economic model with heterogeneous households that was calibrated using a myopic optimal solution to satisfy two conditions that are associated with LVT theory: (1) that LVT would produce the greatest social welfare; and (2) that more households would lose from a change to LVT than would win and, therefore, LVT would be defeated by a voting mechanism. This model became the basis of an induced value experiment for 90 participants in six sessions. Experiment participants made property-investment decisions with interdependency through neighborhood land value capitalization and a revenue-neutral tax return. Using a between-subjects design, participants made decisions under uniform property tax, LVT, or split-rate taxation, and then a voting treatment. Surprisingly, the results of the experiments show that the land tax produced greater welfare than the uniform property tax in only half of the treatments, even though it was designed to produce one percent greater welfare. A statistical analysis of the choices shows that there were systematic over-investment patterns in the LVT treatment and among participants who were induced to have low preferences on improvement. The researchers argue this result is likely due to the positional-good characteristic of housing and the interdependency of land markets, which LVT will tend to enhance. The experiments also show that the participant-voters unexpectedly supported the land tax treatments. Although it is unclear whether the induced-value structure of this experiment—especially the heterogeneity in development preference—matches the real world, the results do suggest that the efficiency and acceptability of the land tax may be more complex than anticipated by economic models using entirely rational and simple profit-maximizing behavioral assumptions.