Luis Garicano REFERRALS*

This paper studies the matching of opportunities with talent when costly diagnosis confers an informational advantage to the agent undertaking it. When this agent is under-qualified, adverse selection prevents efficient referrals through fixed-price contracts. Spot-market contracts that rely on income sharing can match opportunities with talent but induce a team-production problem which, if severe enough, can prevent the referral of valuable opportunities. Partnership contracts, in which agents agree in advance to the allocation of opportunities and of the revenues they generate, support referrals where the market cannot, but often at the expense of distortions on those opportunities that are not referred. “The peculiar character of the problem of a rational economic order is determined precisely by the fact that the knowledge of the circumstances of which we must make use never exists in a concentrated or integrated form.” Friedrich Hayek (1945:519) A crucial task of economic organization is to match opportunities with talent. Such matching requires that an agent first diagnose the opportunity confronted. The agent may conclude from the diagnosis that some other agent is the right person for the job, in which case he must pass it on (or refer it), potentially losing the rents he could obtain if he dealt with the opportunity himself. For example, a personal injury lawyer may, upon study and diagnosis, determine that the case he is working on requires more trial experience than he has. Which institutions result in efficient referrals when the agent in charge of the diagnosis may or may not be the right person for the job? The study of this problem, which we consider the fundamental ‘referral’ problem, is the object of this paper. An obvious way to allocate opportunities would involve trading them for a fixed price. However, this solution requires that the quality of the opportunity be observable and verifiable. This is unlikely to be the case in these markets. For example, in the injury claims market, a lawyer may determine he should refer his client to another lawyer. This transfer must take place under partial ignorance, as the lawyer receiving the claim cannot be allowed to personally evaluate the truthfulness of the client, the merits of the case, etc., since this would create the risk that he concludes an agreement that would result in the complete loss of the case to the referring lawyer. Clearly, all agents prefer, under these conditions, to keep the most valuable opportunities and to refer the least valuable ones. As we show, this implies that whether informational asymmetries are an obstacle to the efficient allocation depends on the direction of the referral flow. A fixed-price contract exists such that the agent prefers to transfer the worst opportunities and keep the best ones. As long as the diagnosing agent is more skilled at handling the opportunity than the agent receiving it, this is the efficient allocation. In other words, ‘topdown’ diagnosis generates no inefficiency. The opposite happens when diagnosis is ‘bottom-up.’ Here, efficiency requires that a less-skilled agent keep the worst opportunities, and pass on the better ones; but any fixed price that is high enough to motivate the transfer of the best opportunities also leads to the transfer of the least valuable ones. In this case, agents may instead choose to sell a share of the problem’s value.1 While