How large is a policy greatest improvement when asset markets are incomplete?

Joint with: Mario Tirelli

Research shows that absent all frictions but incomplete asset markets, at almost every equilibrium a Pareto improvement is supported by many types of intervention, financial, monetary, and fiscal. Surprisingly, little is known about the size of these Pareto improvements, or even how to define their ”size.” We provide a measure of the maximal Pareto improvement, as the largest fraction of current resources a society is willing to pay for the improvement. It can evaluate global policy changes, not just local. This measure admits an exact formula in the quasilinear case, an upper bound in the general case, and it obeys the law of diminishing returns. We show that local information already captures the benefits of global policies, thus supporting the literature’s focus on local Pareto improvements despite global ones being greater.

We define and calibrate the insurance deficit in future income, and then estimate the maximal Pareto improvement in the US to be one third of one percent. We justify this calibration by proving a novel correspondence between insurance deficit and equilibrium consumption: equilibrium consumption is the maximum of a social welfare function, whose parameters are the insurance deficit as well as individual weights, extending a classical result of Lange (1942) to incomplete markets.

For more information, contact Felix Kubler.