Prizes and Patents: Using Market Signals to Provide Incentives for Innovations

pay and improve welfare. Thus, welfare maximization implies that the voluntary participation of
the threshold type θ* is binding and the prize for the threshold type T (E^t) = 0. Since the patent
length is constant and π (E) is increasing, then the incentive compatibility constraint implies that
T (θ) = 0, ∀E Q.E.D. ■

Notice that the interim-efficient mechanism yields the same allocations as the mechanism in
Nordhaus (1969) described above. Here, however, we allow for the possibility of prizes in addition
to patents. Incentive compatibility and welfare maximization imply that it is optimal not to use
prizes or taxes but to use patents only. Hence, the result that only patents are used does not follow
by assumption but rather by the need to provide incentives for innovation.

We now formally define a full information efficient mechanism.

Definition 2. A mechanism is ex post efficient (or full information efficient) if it maximizes
the social surplus (3) subject to no money pump assumption (6) and the voluntary participation
constraint (5).

It is immediate that the ex post efficient mechanism has no deadweight loss. Specifically, the
ex post efficient mechanism has the planner recommending the innovator to innovate if S (E) ≥ K.
Note that the ex post efficient mechanism can be implemented by a variety of prizes. Specifically,
any prize that satisfies K ≤ T (E) if δ (E) = 1 implements the ex post efficient outcomes.

5. Market signals, prizes, and patents without manipulation

Consider a version of the economy in which private agents other than the innovator receive
signals about the quality of the good innovated. One can imagine a variety of schemes that elicit the
information that other agents — or more generally, the markets — possess. Two specific schemes
gained significant recent attention both theoretically and in policymaking circles.

Specifically, suppose that in addition to the innovator, another private agent, called a com-
petitor, observes the value of the innovated good E after it was innovated. In this environment, the
planner can allow the length of the patent, and the prize∕transfers depend on information revealed
by the competitor about the quality of the good.

We first show how to implement the ex-post efficient allocation using the signals of the
competitor. We then describe a mechanism that can implement such allocation uniquely.

10

<<   8   9   10   11   12   13   14   >>

More intriguing information