This paper looks at the determinants of regional housing construction using Israeli panel data. We propose a simple model of regional housing markets in which people prefer to live where housing is cheaper and building contractors prefer to build in regions where housing is more expensive. We distinguish between relative profitability, which establishes a relationship between regional construction through regional house prices, and spatial spillovers. The latter may be induced, for example, by scale economies in which building costs in a region are affected by construction in neighboring regions, thereby inducing a spatial lag in building construction. We find that models are miss-specified unless they incorporate these spillovers. More specifically our results point to housing construction being locally cointegrated, and that local price elasticity of supply ranges from 0.45 – 0.65. We also find substitution from neighboring regions and a strong national complementarity effect that offsets neighbor substitution. Finally, we look at policy complementarities between the public and private sectors in construction and find that public sector housing crowds-in private construction when share exceeds 8%.
