____ Bargaining between the patent holder and improver would face a variety of obstacles.
____ First, there would be very substantial difficulties in valuation. It is by no means clear how
____ one goes about valuing an EST of unknown function, let alone uncertain but potentially
____ highly lucrative research using that EST. Disagreement about the value of the patented
____ invention relative to the value of the research project might make it very difficult for the
____ parties to agree on the terms for a license. This disagreement might well be exacerbated
____ by cognitive biases that could lead the EST patent owner to overvalue its asset. In par-
____ ticular, the owner might overestimate the possibility that the EST would lead to the find-
____ ing of a valuable drug therapy and then negotiate based on this irrational estimation.
____ Negotiation would also be hampered by Arrow's information paradox: it might be im-
____ possible for the subsequent researcher to get a license _ex_ante_ without disclosing to the
____ patent holder valuable information about its own research plans. But because this re-
____ search plan would not be protectable as intellectual property, the competitor might fear
____ that the patent holder would appropriate the information for its own use, with no com-
____ pensating benefit to the competitor. Even if these difficulties did not lead to bargaining
____ breakdown, they would create transaction costs that reduced the cooperative surplus to
____ be gained from a license and would thus deter at least some inventors and improvers
____ from negotiating in the first instance. Transaction costs would be compounded by the
____ likelihood that the would-be follow-on improver would likely have to negotiate licenses
____ not simply with one owner of basic research but with many such owners. For example,
____ in order to develop a commercial treatment for a genetic disease (particularly a polygenic
____ disease), it may be necessary to have access to a large number of ESTs and SNPs, each
____ conceivably patented by a different entity. Similarly, in order to screen potential phar-
____ maceutical products for therapeutic effects and side-effects at the pre-clinical testing
____ stage, it may be very useful to have access to a large number of different receptors, each
____ potentially controlled by a different owner. Some of these difficulties might be addressed
____ if the bargaining occurred _ex_post,_ after the follow-on improver had already developed
____ the improvement. In that case, both the initial inventor and the follow-on improver
____ would have greater knowledge of the value of the improvement relative to that of the
____ original invention. Moreover, in the _ex_post_ situation, the possibility of "blocking
____ patents" might provide a way around Arrow's information paradox. Blocking patents arise
____ when a follow-on improver takes a patented product and improves it in a nonobvious
____ way. Although the follow-on improver can then secure a patent on that improvement,
____ the improvement may nonetheless infringe the original patent. In the blocking patent
____ situation, neither the initial patent holder nor the follow-on improver can sell the im-
____ provement without cross-licensing. But, if the improvement is truly significant, such that
____ the initial inventor would want to sell the improvement itself, it would presumably have
____ a financial incentive to cross-license the improver.

Ibid., 125-126. Rebecca Eisenberg makes a similar point: "Michael Heller and I argue
that too many patent rights on 'upstream' discoveries can stifle 'downstream' research
and product development by increasing transaction costs and magnifying the risk of bar-
gaining failures." Rebecca S. Eisenberg, "Bargaining over the Transfer of Proprietary Re-
search Tools," 224. See also Robert P. Merges, "Institutions for Intellectual Property
Transactions: The Case of Patent Pools," in _Expanding_the_Boundaries_of_Intellectual_
_Property,_ Rochelle Cooper Dreyfuss and Diane Leenheer Zimmerman, eds. (Oxford:
Oxford University Press, 2001), 127-128.

[11-96] Bessen and Maskin, "Sequential Innovation, Patents, and Imitation."

[11-97] Steven Levy, "The Great Amazon Patent Debate," _Newsweek_ (March 13, 2000): 74

[[321]]

p320 _ -chap- _ toc-1 _ p321w _ toc-2 _ +chap+ _ p322

v?