Delegating Investment in a Common-Value Project.

I investigate the problem of delegating an investment effort when it is not known in advance which firm is most efficient, or whether the investment should be made at all. The motivating problem is that of commissioning R instead of relying on patent incentives. Firms have different private signals of a project's private (and social) value, and different costs of achieving it. I show that the two allocation problems of (i) making an efficient decision whether to invest, and (ii) delegating the investment to the least-cost firm can simultaneously be solved with no more profit dissipation than a procurement mechanism would require, assuming that the signals of value were known in advance.