Sunday, May 11, 2008

Abramowicz on FutureMAP

Michael Abramowicz, a member of John McCain's Justice Advisory Committee, wrote a law review article in the University of Chicago Law Review titled: Information Markets, Administrative Decisionmaking, and Predictive Cost-Benefit Analysis, which discussed issues surrounding FutureMAP, a project of the Defense Advanced Research Projects Agency (DARPA), which was an experiment to determine whether information markets could improve Defense Department decisionmaking.

The article noted:

The technology underlying FutureMAP could have been used before the war to aggregate intelligence agents' predictions about whether weapons of mass destruction would be found. One argument for such a market might have been that it would be better at aggregating information than government officials, but the stronger argument is that it would have provided the public an objective indication of what governmental officials really believed, without release of the underlying intelligence information.

IPBiz notes that the "patent quality" argument in patent reform is analogous to the "weapons of mass destruction" business in the Iraq thinking. At the end of the day, judged by the actions of the Coalition for Patent Fairness, the patent quality issue wasn't really there. But the IT folks couldn't tell you that in advance. The law professors couldn't either (note Lemley's flip flop
on the numbers of Robert Clarke!)

The conclusion of Abramowicz on FutureMap:

The tempest over FutureMAP may have prevented that project from producing useful experiments on the workings of information markets, but it did provide an important experiment about the willingness of the public to allow information markets a role in policymaking. That experiment revealed that the public is not yet ready for information markets, and this Article has suggested some reasons to support the public's caution. Insufficient experimentation has yet been done to establish the degree of information markets' susceptibility to attempts at manipulation, and some of the more adventurous information market designs, such as conditional markets, may be too noisy to be reliable. If such problems can be overcome, however, the objectivity of information markets might allow their predictions to be a useful input into the administrative process. The normative predictions of predictive cost-benefit analysis may be particularly useful, and less [page 1020] likely to excite public angst than direct predictions of variables such as terrorism deaths. Public acceptance will surely be slow, but further academic experimentation could either identify unsolvable problems or produce an academic consensus in favor of some uses of information markets that eventually might lead to gentle introduction of information markets into the policymaking process.

See 71 U. Chi. L. Rev. 933 (2004)

See also

More recently Michael Abramowicz wrote THE DANGER OF UNDERDEVELOPED PATENT PROSPECTS, 92 Cornell L. Rev. 1065 (2007), while a visiting professor at the University of Chicago Law School. Therein he thanks John Duffy (another McCain advisor) for his valuable comments and discussion.

Within, he writes:

The most obvious, though perhaps not the most commercially significant, form of unpatentable development activity is scientific testing. n146 As evidenced by statutes such as the Orphan Drug Act, the need to encourage scientific testing of inventions is particularly acute for pharmaceuticals. n147 However, a wide range of inventions might benefit from further testing to determine how well the invention will perform in the marketplace. If tests reveal that an invention will not perform well, the results will help the patentee make an informed decision about whether to fully commercialize it. In addition, the less patent term remaining, the more incentive patentees have to commercialize products without testing them and thus without finding solutions to potential problems.

IPBiz: scientific testing is NOT a major need in determining how well an invention will perform in the marketplace.

Abramowicz also writes:

Unsurprisingly, courts have been skeptical of efforts to obtain broad patent protection for scientific studies. For example, the Federal Circuit recently upheld a district court's invalidation of a patent for growing and eating cruciferous sprouts, such as broccoli, to reduce the level of carcinogenic substances and the risk of cancer. n152 The patentee argued that it had discovered a new method for "selecting the particular seeds that will germinate as sprouts rich in glucosinolates." n153 The court, however, held that, "while [the patentee] may have recognized something about sprouts that was not known before, [the patentee's] claims do not describe a new method." n154 With its claim rejected, the patentee had no access to additional patent protection. As a result of the Federal Circuit's decision, patentees may not have the incentive to research known but unproven scientific hypotheses about public domain products, such as cruciferous sprouts, or even about their own products. Given the difficulty that courts would face in determining the extent to which a particu-lar experiment reduced scientific uncertainty, the Federal Circuit's decision is likely a good one, but its effect may re-quire the patent system to provide parties with nonpatent innovation incentives, such as governmental scientific grants, to encourage research on patented or once-patented inventions.

He concludes:

Although intellectual property commentators have long discussed the optimal length of the patent term, they have as-sumed that the existing term is long enough to give patentees sufficient incentives to commercialize their patented in-ventions.
Two new developments, one empirical and one theoretical, make such assumptions problematic.

First, patents are increasingly granted on embryonic inventions, particularly in the field of genomics. [IPBiz: !] Such inventions may require study for a period longer than the patent term. Second, economic models of intellectual property suggest that inventors may acquire patents for their option value at a relatively early stage of development. By the time a patent option would be worth exercising if most of the patent term remained, so little time may remain that commercialization would not be feasible.

The patent system includes some mechanisms that significantly limit the underdevelopment problem. First, by imposing substantial requirements for patenting, the patent system reduces the proportion of expenses that need to be made after patenting. The patent underdevelopment concern thus provides a counterweight to the prospect theory of patents, which argues that patents should be granted early. Second, improvement patents can effectively extend the patent term. However, they may not be available for some types of development activity, including scientific experimentation, commercial testing and marketing, and research producing subpatentable inventions. [page 1121] As a result, patent extensions, though probably welfare decreasing for the vast majority of inventions, might enhance social welfare for a small subset.

Auctions for patent term extensions provide a possible solution to the patent underdevelopment problem. Under this approach, a patentee would be able to call for such an extension before the end of the patent term. The patentee, however, would need to offer a bid substantially exceeding that of the runner-up to win the extension, and a patentee who calls for, but does not win, an auction would pay a fine. This approach will result in patent extensions only where the value of continuity of patent ownership is relatively high. These will be cases in which the patentee would like to make investments during the remainder of the original patent term that would pay off primarily in the extension period and, thus, where the degree of patent underdevelopment would otherwise be high.

IPBiz notes the story of Chester Carlson and xerography. Carlson's basic patent ran out long before the Xerox 914 went to market, but he had so many improvements no one competed. Carlson didn't need a patent term extension.

***Further on Carlson (posted on patenthawk)-->

Recall that roughly nine out of ten issued patents don't make any royalties. Further, even when there is a ground-breaking innovation, such as Chester Carlson's, the patent can expire before royalties are made. And, the best and brightest minds at IBM (and a thinktank they hired), did not recognize an equity worth owning in Xerox.


Post a Comment

<< Home