We designed the model so that the behavior under the DOE regime is very
easy to understand. The competitor never uses the technology because of the
threat of a patent suit. The inventor invents if the cost of invention w is not too
high, and conditional on invention, refines if the cost of refinement x is not too
high. Figure 1 displays the equilibrium outcomes under both regimes; outcomes
under the DOE regime are listed in italics, and outcomes under the no DOE
regime are listed in bold. The cost of invention is measured along the horizontal
axis and the cost of refinement is measured along the vertical axis. The region
labeled NO INVENTION indicates the inventor does not invent if the costs of
invention plus refinement are high. The region labeled EFFICIENT REFINE-
MENT indicates the inventor efficiently invents and refines when invention and
refinement costs are sufficiently low. We describe refinement as efficient be-
cause it is socially valuable and profitable regardless of strategic considerations.
In the remaining three regions, IMITATION,194 PIONEER INVENTION, and
PREEMPTIVE REFINEMENT, the DOE regime leads to invention but no
refinement because in these three regions the cost of refinement x is greater than
the benefit from refinement M2 - M1. Equilibrium behavior is more complicated
under the regime without the DOE, because the possibility of entry by the
competitor influences the inventor’s behavior. That possibility is missing from
the regime with the DOE because the inventor automatically gains rights over F when she obtains set E. Under the regime without the DOE, the inventor may refine the technology for strategic reasons rather than efficiency. In other words, the inventor may obtain F simply to block the competitor from developing an embodiment in F or to enhance his bargaining position in a cross-licensing negotiation.195 We call this preemptive refinement.
Preemption is not an issue when the competitor does not pose a credible threat of entry. Even without the DOE, the competitor will not develop an embodiment in F if the cost of refinement is too high or the duopoly profit is too low. Precisely, if D y then competition is not credible and equilibrium behavior is the same under either regime. Figure 1 shows the equilibrium outcomes without DOE in the interesting case when D y; thus, the competitor will find imitation profitable if he has the opportunity.
Figure 1 facilitates a policy comparison of the DOE and no DOE regimes. It
shows that equilibrium outcomes are the same in the NO INVENTION and
EFFICIENT REFINEMENT regions. The presence of a potential competitor is
irrelevant for parameter values falling in these regions. In the NO INVENTION
region invention is too costly regardless of whether the DOE is available. In the
EFFICIENT REFINEMENT region refinement is profitable regardless of whether
a competitor exists.
In the other three regions the DOE yields different outcomes from the regime
without the DOE. The DOE induces invention in the region labeled PIONEER
INVENTION, but there is no invention without it. The absence of the DOE
induces refinement in the region labeled PREEMPTIVE REFINEMENT, but
there is no refinement given the DOE. Finally, the absence of DOE allows entry
in the region labeled IMITATION, while there is neither refinement nor imita-
tion under the DOE.
The model yields four main results that have policy significance. We present these results in the following four propositions.
Proposition 1. The DOE does not strengthen patent rights when the
inventor’s cost of refinement is relatively small, or when a potential competi-
tor does not pose a credible threat of entry. The two regimes give the same
results when the inventor’s refinement cost is less than or equal to the benefit
from obtaining both sets of embodiments E and F, i.e., x M2 - M1. This
equivalence arises because the greater scope offered by the DOE is immaterial
when the inventor gains a direct benefit from refinement. Likewise, the two
regimes are equivalent when the potential competitor does not pose a credible
threat of entry because the duopoly profit is small relative to the competitor’s
refinement cost, i.e., D y. In these cases policy should be chosen on the basis
of considerations outside the scope of our model. For example, one might choose no DOE because the DOE creates fuzzier property rights and imposes greater litigation costs.196