Regulation and Frontier Housing Supply

Abstract

Regulation is a major driver of housing supply, yet often difficult to observe directly. We show that frontier cost, the non-land cost of producing housing absent regulation, is identified from prices and quantities alone, without instruments even when quantity is endogenous in a mean regression. Identification requires regulation to enter as a nonnegative wedge with zero in its conditional support. The difference between price and frontier cost yields the regulatory tax. We apply the approach to new multi-floor, multi-family residential construction in Israel. Accounting for random housing quality, we estimate economies of scale at low heights (minimum efficient scale about five floors), nearly constant marginal cost at middle heights, and an elasticity of substitution between land and non-land inputs of about 0.15 to 0.2 at the greatest heights. The estimated mean regulatory tax is 47% of housing prices, with substantial variation across locations, and is positively correlated with centrality, density, and prices. We also construct a lower bound allowing quality to differ systematically over location and time, assuming weak complementarity between quality and demand. In 2017, when prices were highest in our sample and the bound is most informative, we bound the mean regulatory tax between 38% (using a 2km radius) and 53%.