Government regulations of coastal flood zones have long sought to minimize risk by building coastal seawalls to protect urban centers. The housing market responds to these safety expenditures by building even more behind these new walls, raising expected flood damage. Economists have been skeptical of these government safety efforts because this endogenous response appears to undercut the safety objective of seawalls. In this paper, we argue that society should take a broader welfare perspective of such safety regulations. The optimal regulation should maximize the net value of economic activities that involve risks and take account of the endogenous market response. Perhaps surprisingly, the optimal solution leads to more investment in safety (higher seawalls) but also much higher net welfare. Note that these results apply to other safety regulations as well, for example, increasing the speed of transport versus making cars safer.

We demonstrate this welfare principal using a model of coastal defense, CRESS, applied to six cities in the eastern United States. We start by minimizing the sum of wall costs and residual flood damage assuming no endogenous response. We then predict what endogenous response these walls might encourage. Finally, we optimize the amount of housing in the flood plain given the resulting flood damage. We find the optimal plan increases the height of the seawalls and encourages more building in the flood plain. Expected flood damage increases slightly but welfare (net present value) increases a great deal.