TCS Daily

Patent Reform Hits the Hill

By James V. DeLong - June 21, 2005 12:00 AM

Most people pay little attention to the patent system, so it rarely generates heat on Capitol Hill. It is not that people are unaware. Everyone knows that patents exist, but mostly the system hums along in the background, part of the unnoticed infrastructure of our lives, like flipping a switch and having the electricity flow. Preoccupation belongs only to the in-groups of patent lawyers, inventors, and the corporate officials responsible for managing intellectual property.

Sometimes, though, concern breaks the bounds of this narrow circle and grabs broader attention.


This happened in the late 19th Century, when patent issues consumed the attention of both courts and Congress. The Supreme Court argument in The Telephone Cases (1888) stretched over 12 days and resulted in an opinion that takes up a full volume of the U.S. Reports. And see Steven Usselman, Regulating Railroad Innovation: Business, Technology, and Politics in America (1840-1920) for a description of the long and politically charged battles over patents during the latter part of the century.


And it is happening now, as intellectual property, including patents, is regarded, justly, as an important component of wealth and national prosperity. Proposals for change are working their way through Congress, with a good head of steam, and H.R. 2527, a reform bill, introduced last week by Lamar Smith, Chair of the IP subcommittee of the House Judiciary Committee, is given an excellent chance of passage. Smith intends to move to markup before the end of this month.


On the Senate side, Orrin Hatch's IP subcommittee has also held hearings, covering the same set of contentions and with many of the same witnesses (Heaven forfend that anything on the Hill be said only once). Hatch and Senator Leahy also intend to produce a bill, but they are silent on how much -- if at all -- it will track H.R. 2527.


The echo of the 19th Century is intriguing, because pronounced similarities exist. Now, as then, explosive technological, financial, and institutional change has created scope for creativity. Now, as then, the fundamental criteria of novelty, utility, and non-obviousness -- which have governed patents since the Venetians articulated them in the 13th Century -- present puzzles. In an era of continuing change, what is truly inventive and what is within the ready grasp of anyone skilled in the art (only some grasp it a little more quickly than others)? Now, as then, complaints of overreaching and fears of gridlock are heard; and now, as then, patents are seen as hugely important to corporate value, which means that real money is at stake.


Given the combination of concerns, not only the in-groups but a broad circle of concerned corporate executives are anxious for Congress to do something, and the chances are good that they will have their way, at least on the proposals that are not terribly controversial, which is actually most of them.


Avoiding the technicalities -- and a major reason the patent system gets limited attention from the public is the mind-numbing opacity of its language and details -- the proposed changes break down roughly into three major categories:


(1) Reducing the transaction costs of the system, especially litigation;

(2) The "patent quality" issue;

(3) The enforcement of patent rights -- should injunctions be used, or should the infringer just pay a toll?


This division is only rough, because there are some strays designed to fix quite specific issues; and some reform proposals touch on more than one category, but it holds up pretty well.


All the proposed changes are important, with the first two probably somewhat more important than the third, largely because the third one is unlikely to go anywhere. But in many ways the third one is the most interesting, and controversial, because it raises fundamental questions about the appropriate industrial structure for performing innovative activities.


Transaction Costs


In several areas, current doctrine governing patent litigation requires or permits extensive inquiry into subjective states of mind, which guarantees boundless opportunities for intensive and expensive discovery proceedings and greatly increases the costs and uncertainties of litigation.


The most prominent example is that the U.S. holds that the first person to invent something gets the patent, even if someone else files an application first. Changing this to a first-to-file system would be a reform twofer. Not only would it get rid of a doctrine that leads to expensive parsings of unanswerable questions -- when, during the long process of creation, has something now been invented? -- for no serious gain, but it would bring the U.S. into accord with the rest of the world.


Over the years, small inventors have opposed this change, for reasons that are difficult to fathom, since small guys lose these battles as often as they win them, and they are disproportionately burdened by uncertainty and by the expense of litigation. But the stars now seem in alignment, finally, to change this, granting the patent to the "first inventor to file."


Some other patent doctrines have similar effects of raising both transaction costs and uncertainty for little, if any, benefit for anyone but the lawyers. One of the most important is the contortions surrounding "willful" infringement, a topic taken up in the next section because it also has significant effects on patent quality. A bunch of these doctrines are likely to be significantly modified, and good riddance.


Patent Quality


The in-group is virtually unanimous that in recent years too many patents have been granted that should not have been, in that if the examiners had only found the proper literature references or existing patents, these patents would have failed the criteria of novelty, utility, and non-obviousness.


The reasons for the failure are several. Some of it is resources. The number of patent applications keeps rising, and the US Patent and Trademark Office has a hard time keeping up. It has not had enough examiners, in part because Congress keeps using patent fees as a source of cash to give pork to worthy constituents. But money is not a panacea, because it is difficult to find, train, and keep enough good examiners anyway, especially in the exotic pastures of biotech and computer science.


Some problems are caused by lack of imagination. For example, an applicant might take a practice well known in the brick-and-mortar world and transfer it to the Internet, and the examiner, finding no prior art exactly on point, will sign off. ("Prior art" means that someone recorded the gist of the idea somewhere, in a previous patent or in published literature.) Thus the delighted applicant might end up with a patent on, say, a Dutch Auction conducted over the Internet.


Other quality problems derive from the structure of the patenting process. Only the applicant is involved, not competitors, customers, or the public, so it is up to the examiner to play devil's advocate. Much can slip by, especially when short-handedness puts examiners under pressure to keep that timber rolling, and when PTO sees its constituency as consisting of those who seek patents and not those who use them.


One reform proposal is to end the fee diversion. The entire patent-dependent community can agree that it needs more money, and this is a rare case in which beneficiaries of government action are perfectly willing to pay for it.


In addition, there are extensive proposals to create mechanisms for reviewing patents after the initial grant. At that point, competitors and customers of the patent holder, who should, after all, know the field well, may come up with prior art that escaped the attention of the examiner.


These changes are all to the good, but it is a mistake to expect too much from post-grant review. Contesting a patent creates many benefits, and the contester gets only a fraction of them. The rest accrue to others, including competitors. So no firm is likely to make the effort unless its interests are strongly and immediately threatened. Most patents, even bad ones, will remain unchallenged, and will wind up undergoing rigorous outside vetting only when someone has a serious financial interest. This usually occurs only when the patent turns out to be valuable and infringement, either actual or potential, is a reality.


This problem of incentives triggers one of the most contentious features of the Smith bill -- the "second window." The bill would allow an attack on a patent not only within nine months after the patent grant, but within six months after an alleged infringer receives notice that the patent holder thinks it is infringing. On this one, business is badly split, with those who see themselves as primarily patent producers opposed the second window and advocating the need for certainty, and those who see themselves as more likely to be the defendants in infringement suits wanting the opportunity to start over once the patent holder starts to threaten.


Related to the quality issue are proposals to change the damages awarded for infringement. An infringer can be assessed the patent holder's lost profits, or, if this is difficult to judge, a reasonable royalty. But if infringement is "willful" -- which can mean only that the infringer knew of the existence of a patent, even if it did not believe the patent was valid or covered its activity -- it can get clipped for three times these amounts as a punishment.


This risk is perceived as unjust, especially when people have doubts about the quality of many existing patents. It also has some practical consequences, in that it creates incentives for firms to tell their researchers not to look at the files of existing patents. Just invent and toss stuff into the patent office, leaving it to the examiners to determine if time should have been spent on the idea.


This is an absurd outcome. A major purpose of the whole patent system is to create bodies of patents that innovators can look at so they can determine how they should or should not spend their time. The for-heaven's-sake-don't-look-at-the-patent-files philosophy also burdens the examiners, who should be able to rely on applicants to find and cite the relevant prior art.


The bill would limit the application of the willfulness doctrine to the most egregious cases.


A deeper issue of quality is raised by a Petition for Certiorari now pending before the Supreme Court in KSR v. Teleflex. The petitioners argue that the standard for determining whether an invention is "nonobvious" has been made too stringent by the lower court -- that is, too many patents are granted that should be denied because the invention is within the ken of someone skilled in the field -- and that the Supreme Court should restore right and justice.


The view of the filers here (including PFF and a group of 24 law professors, both of which filed as amici in support) is that the advocates of post-grant review are going to be chagrined when the new mechanisms produce the result "yep -- that's a patent, all right!" The review processes will not improve things if the standard is not fixed.




The Nuclear Option in patent law is the injunction. Once the validity of a patent is upheld, and the fact of infringement is established, the patent holder can get a court to bar the infringer from using the patent without permission. The statutory language is: "The several courts having jurisdiction of cases under this title may grant injunctions in accordance with the principles of equity to prevent the violation of any right secured by patent, on such terms as the court deems reasonable."

Nothing wrong with that, one might think, but the Federal Circuit, which Congress has designated to hear all patent appeals, instructs the trial courts not to engage in elaborate balancing or other shilly-shallying, barring some pubic health emergency. If a patent is held valid, and infringement is established, an injunction should be granted.


Parts of the business community, spearheaded by Intel, Apple, Oracle, and the Business Software Alliance, want to make injunctions more difficult to obtain, and have persuaded the drafters of H.R. 2527 to insert "In determining equity, the court shall consider the fairness of the remedy in light of all facts and the relevant interests of the parties . . . ." The bill also adds language to encourage district courts to stay the enforcement of any injunction pending appeal of the case.


Other parts of the business community, including several associations of IP owners, the pharmaceutical industry, and the growing number of firms that specialize in creating innovations, are dead opposed. They are reinforced by universities, which are making an increasing effort to profit from their patents.


Given the forces at work, change is unlikely, but fights over this issue could derail other reforms, which would be a pity.


For the proponents of the change, the problem is that they create complex products that touch on dozens of technologies that are affected by hundreds or maybe thousands of patents. In hearings before the Federal Trade Commission in 2002, a lawyer for Intel noted that CPUs are covered by 90,000 patents, held by 10,000 different parties, and that semiconductors generally are subject to 420,000 patents held by 40,000 parties.


Even with maximum good will and considerable sophistication, it is easy to miss one, to err about the scope of some murky grant, or to dismiss a patent as invalid only to have a jury disagree. (Yes, many questions in this most arcane of fields are left to juries, so the tech companies have less than perfect faith in the accuracy of the system. Interestingly, juries tend to favor inventors more than do judges.) Of course, the disincentives to look at the patent files created by the willfulness issue exacerbate things.


The prospect of having a product line halted by injunction because one of thousands of possibly relevant patents was overlooked is a nightmare. If an injunction is a serious possibility, then the patent holder can extract a lot of value from the product developer. The holder is in the position of the hold-out owner of the parcel of land that is indispensable to the construction of a major project, when all the other parcels have been purchased and the excavators are already on the way.


Mechanisms for avoiding hold-outs exist, but not when the problem is discovered only when the concrete is about to be poured.


The tech companies' alternative would be to halt development work until the lawyers could chase down and analyze everything, but this is clearly unworkable. It would make "Internet time" a synonym for "regulatory time" or "litigation time," and for little purpose, since the intrinsic uncertainties are so huge that the lawyers would never produce a solid answer.


The tech companies often simply pass the ball down the field, pray, and stand willing to litigate. They have also developed some practical coping methods, such as filing for as many patents as possible to create bargaining chips. Then, because the big tech companies are all in the same situation, they engage in massive cross-licensing. As an alternative, they play the old cold war game of Mutual Assured Destruction (MAD). If a company plays hardball with its patent portfolio, others will respond in kind, so it is better not to start a war.


Creative things are taking place in the area of standards, where firms with patents on crucial technologies proclaim to all the world a generalized willingness to license on reasonable terms. This allows others to develop first and work out the details later.


But these mechanisms dont work when the patent holders are not part of the network of big companies, and over half of inventions come from other sources. Innovation has always been somewhat separated from product development and marketing, as Thomas Edison demonstrated 150 years ago, and this trend may be accentuating. Universities are fertile sources of invention, and specialized innovation firms are betting that separating innovation from the downstream processes will work better than integration.


Firms of this type are not subject to retaliatory threats and are not interested in cross-licensing, so the big tech companies have no lever. So what the tech companies want is a limit on injunctions, and court-determined damages that are set in an amount that does not recognize the strategic value of the hold-out position at a late stage in the decision process.


Their goal is strikingly similar to the solution to analogous problems in the field of real property. The techies want the power of eminent domain, the right to take property and then have a court determine just compensation.


This would not be as novel as it sounds, and again the 19th Century is illustrative. Many network-type industries have been granted eminent domain powers by state legislatures, especially those that were building rights-of-way, where the last parcel problem can be acute.


Naturally, the IP owners, the innovation shops, and the universities object.


Their first line of defense is often a ringing affirmation of the sanctity of property rights, sometimes in language that would do credit to a 19th Century Robber Baron. Comparative equities should have nothing to do with it because patents are property and property is sacred. Those of us who have been through the environmental wars reflect wryly that most of these Silicon Valley tech barons and university scientists never met a piece of real property they would not happily seize in the name of endangered species protection, wetlands preservation, or smart growth.


The IP owners have some realpolitik policy arguments that are better, though.


  • The case that injunctions are a serious problem for the tech industry is oddly bloodless. There have been no studies; no serious horror stories. Even accepting the Washington rule that multiple anecdotes equal "data," the argument is skimpy.

  • Not many patents are that crucial. Given notice before a product launch, a developer can usually design around them. This puts a limit on the rapacity of a patent holder.

  • To a large degree, the problems are related to the issues of patent quality and litigation uncertainty. Other parts of the reform proposal address these issues directly, so why mess with the second order issue of injunctions?

  • Injunctions are necessary to force the tech companies to the bargaining table. Without this threat, they would shrug and say "sue me." Patent litigation is expensive, and a large company can wear down almost anyone if it adopts a hard-line stance. Proposals to change the injunction standard do not include any mechanisms by which the tech companies would guarantee their own good faith. They could simply stop negotiating at all, which would turn the whole system into one of compulsory licensing, with license fees determined by the courts.

  • Some of the specific proposals under consideration (though not in the bill language) would favor the big tech companies -- those that integrate invention, development, and marketing -- and disadvantage specialized innovation shops. This meddling in the basic structure of innovative activity would be a big mistake. It may be that innovation should be a function separate from product development and marketing, in the way that designing computer chips is not done by the firms that do the fabrication.


An argument that is not made is that the software companies supporting the dilution of injunctions are acting against their own interests. One of the growing issues in the contest between proprietary software and the FOSS movement (free and open source software) concerns patents. The FOSS people oppose them, ostensibly on philosophical grounds. In fact, though, FOSS as a movement needs access to the standards and technologies being developed by the proprietary software sector. It cannot gain this access under the existing system, in which such technologies are patented and then licensed out, usually in exchange for licenses back, or for "reasonable" royalties.


FOSS cannot give the return licenses or pay the royalties. So FOSSers should like the idea of a system in which they just use the patents, shrug, and let a court determine an appropriate royalty to be paid by a program distributed at no charge.


The IP owners will win this injunction issue, as they should, given the lack of empirical support for the proposition a real problem exists, or that any problems worthy of note will not be addressed by reforms of quality and certainty. But it raises interesting questions, and it will not go away. The techies may come to the next battle better armed.


On the other reforms, opinions seems fairly united that they are good things, except for the "second window," perhaps. There are, of course, devils in the details -- the blog Patently-O commented that the bill "still has something to offend every interest" -- and the deals could come unstuck. When Chairman Smith spoke last week at a National Academy of Sciences conference, he outlined the provisions of his bill and declined to take any questions, explaining that the political balance is so delicate that any mis-statement on his part could derail the train.


So stay tuned.


TCS Daily Archives