Estimating the effect of land protection policy when regulatory boundaries can change

Abstract

Environmental policies vary across space, and a growing body of empirical research compares land prices across administrative boundaries to estimate the causal effects of local policies. However, this approach can be confounded if the market anticipates the boundaries may change, and the prices of land respond accordingly. This is the case for protection zones for endangered wildlife in the oil sands region of Alberta, Canada. We propose a way to separately identify the effect of local policy and the market’s beliefs that boundaries may change, and we apply this approach to Canadian land prices and wildlife protection zones in Alberta. We find that anticipation matters: market expectations that land will become protected reduces land prices by nearly one-quarter, and empirical analysis that omits anticipation underestimates the cost of regulation by one-third.