Dan Bricklin's Web Site: www.bricklin.com
Thoughts on Patent Litigation in 2006
Observations and thoughts following Dan Bricklin's participation as an expert witness in a patent trial in U.S. District Court for the Eastern District of Texas.
In September 2006 I testified as an expert witness in a patent case in U.S. District Court in Marshall, Texas. During the course of that trial I had a lot of time to sit and think about what I was seeing. This essay is a combination of those thoughts (most of which were formed before the jury reached their verdict), as well as things I learned from various lawyers during the course of working on this case and from statements that I heard from the judge. They also reflect my prior background as an MBA, entrepreneur, software developer, defendant and plaintiff in lawsuits, expert witness, and criminal case juror.
These thoughts are general and do not necessarily apply to any particular case in which I was or am involved. I write them here at the request of others (including lawyers and people involved in public policy) who are interested in my thoughts in response to participating in the trial. The opinions here are for discussion purposes and to raise ideas that may or may not be helpful in better understanding patent litigation and how we might be able to improve the process; they are not a full airing of the issues. Of course, they do not include or reveal any confidential or privileged information to which I may have had access during any of my work. Finally, I assume that the reader is familiar with the area of patents and some of the public policy discussions around them.
Patent cases are complex. The charge to the jury can run an hour in length just to state the type of issues to be decided and give a concise description of the elements of the laws to be applied. In addition, the area of the patent is probably not one in which the members of the jury have prior expertise, and they need to become familiar with what patents are in general, patent law, what the particular patent says, the background, facts, prior art, claims, construction of terms determined by the court, and more. "Teaching" these many aspects to the jury has to be done without the benefit of direct feedback from the jurors to those explaining things to them to see what they understand, so that teaching may have to be especially comprehensive and detailed. The interpretation and meaning of patent claims can be quite exacting, taking up much time to cover what on first read may seem to be a simple phrase.
All of this and more leads to the conclusion that it is often not possible to cover all of the relevant facts and issues and to get a jury of normal individuals to learn the field and understand those facts and issues themselves to determine the "right" answer to patent validity and infringement in a trial of practical length. Even in a theoretically perfect world, it could take weeks or months depending upon the patent at issue. There is a reason that patent examiners have education and background requirements, are given tools and training for doing investigations, have back and forth correspondence with the applicants, and often take a long time measured in months or years to do their determination of validity. These are not simple things that anybody can do without proper preparation. Therefore, a trial by jury of limited duration is always a compromise of some sort. A question, of course, is whether, from society's viewpoint, such trials are nonetheless sufficient, appropriate, and the way in which we wish to resolve these complex questions.
We can contrast this situation with common criminal, contract dispute, and even copyright trials. In many of those cases, the jury is asked to make more ordinary or commonsense determinations such as who is telling the truth, who had motivation or intent, who did something improper, who had access, etc. Deep issues of technological equivalence or the state of the art in a particular fast-advancing technology at a particular point in time are uncommon in such situations as far as I know.
Our jury and legal systems have developed specializing in such common situations. They work quite well.
In a normal trial (not one involving patents), there are usually, though not always, two parties. (In a criminal case, one side is the state.) The goal of the trial, stemming from the design of both the law and the process, is to come to the "right" result in the eyes of society given the facts of the case. The actual impact of a particular case is usually relatively localized. There is little ripple effect within society other than to the individuals (persons or companies) involved. Depending upon who the individuals are, there may be a wider effect (such as if it affects the individual's future relationship with the rest of society, as in antitrust or class-action cases). There may also be a ripple effect through legal precedent. The precedent, though, is usually legal, not factual, and there can be multiple, opposing precedents until a long process resolves things in higher courts. Precedent is not absolute.
In a patent case, there are also usually two parties. Just as in the other types of cases, the facts may include other issues about the parties and their relationship. One party may be aggrieved in some way by the other, one may be of a class that society often wants to protect or wants to moderate, etc. In a normal trial, while everyone is equal before the law, these circumstances are often brought to bear by juries in weighing the testimony to make sure that the "right" and "fair" result, in the eyes of society, comes about in the particular case before them. The system is set up so that the same process used in other trials to get the "right" and "fair" result in the eyes of society for the two parties occurs.
A problem for society with this process for patents, especially for software patents and probably business process patents, is that the impact is not just on the two parties. A patent found valid by a jury acquires a status even greater than the normal "presumption of validity". It becomes like a super precedent that can be used against other parties. Even if the "right" result occurs with respect to the two parties, with the "good" party prevailing, all future parties, even if they are not party to the suit, whether they are "good" or "bad" or neither, are affected. A jury that feels that damages should be awarded cannot award those damages unless they also hold the patent valid. The place of the standard of "clear and convincing evidence" for proof of invalidity on the continuum between "more likely than not" (51% vs. 49%) and "beyond a reasonable doubt" (something many jurors have a hard time grasping) can be affected by other issues in the case.
If patent cases only affected the parties in the suit this would be less of an issue. But that is not the case. Many patents, especially software patents, have implications far beyond the allegedly infringing party in any given suit and the facts and circumstances of their infringement.
A patent that is valid is a grant of limited-duration monopoly by society in return for teaching something that may later be practiced by anyone. A process that incorrectly lets a patent that is actually invalid be declared valid is taking something that belongs to society as a whole and giving it to a single party alone to exploit. The system is not set up to make sure that patent validity is given its own day in court by itself in a process tuned to the needs of the laws of patent validity. Therefore, we must be sure that the process we do use is careful to ensure that determination of patent validity in that process is given as fair and effective a hearing as possible. Such a determination by a court is part of the checks and balances we have with respect to the issuing of patents and is an argument defending against concerns people may have with the patent issuing process.
Let us look at other aspects of the situation.
Often the two parties to patent litigation are business entities such as corporations. For many businesses, uncertainty can be a major problem. Not knowing what the final outcome of the litigation will be could be as much of a problem as many bad final outcomes themselves. The time from the start of litigation until final resolution can be a cost factor where a shorter resolution is far desirable, even if the odds of success are changed slightly (if at all). In this case, judicial process that speeds up the time until final resolution can be a boon to one or both parties.
Another issue is length of trial. Jurors often have some forms of hardship to serve on juries, with the longer the service the greater the hardship. This is something we accept as society because of the need for juries as an integral part of the process of justice. However, we do not want to take up too much juror time without good cause. Likewise, courts have a limited amount of time to devote to all of their cases and we do not want to unduly overburden the courts without good cause for any one particular case.
An issue that I have heard and read about the U.S. District Courts for the Eastern District of Texas (such as the September 24, 2006, New York Times article "So Small a Town, So Many Patent Suits") is that the process there is skewed towards speedy resolution and short trials. A question is what effect this may have on the "quality" of the results. Another issue sometimes mentioned is the particular juror pool and how the general juror backgrounds affect the results.
As I covered above, no reasonably short trial with a random selection of regular people in any federal venue can fully investigate all of the issues in many patent suits with the jurors all considering all of the issues having the knowledge of one skilled in the art covered by the patent. Also, even if they did, without special training and experience, like that possessed by patent examiners, they may not know how to appropriately weight each of those issues in making a determination.
I wrote a blog post about a talk given by Malcolm Gladwell during the time he was finishing up his book "Blink". Here's an excerpt related to this issue:
"He had a new book coming out. It's about when is there too much information. He compares rapid cognition (e.g., quick first impressions) to having lots of data. Patterns appear much more rapidly than we think... Finally, he says that the bias should be in editing information, and not in adding more information...My take on his claim that less information is better? I think it has merit if understood in the context he presents it. As I see it, people are very bad appliers of weighting factors when evaluating lots of criteria. There ends up being a compression of ranges, and some items are given heavy weight because we have them (or because they were expensive to obtain) and others ignored because we don't have them. Using a simpler method for making decisions, based upon fewer factors that we know are relevant, may work better than having many for which we assign incorrect weightings. In those cases, less is better." [I then discuss and show a visual example of simplified data giving better results.]
The point is that more data, such as presented in a longer trial, does not necessarily result in better decisions. The important thing is to have appropriate factors that can be fairly weighed by the jurors we have to come to a good decision.
Prior to trial, each side has hopefully determined all of their points and counter points on each issue. The image I see is of two armies coming to battle. As in the hypothetical situation where each side would send one, gallant soldier to battle it out by proxy between the two sides, saving lives (the losing army ending up in slavery instead of dead), in court they choose a few key issues to decide the case. Discovery up until then, and the practicalities of the trial length, help narrow the issues to just those "trial issues". (Another analogy, perhaps, could be to a single play each in U.S. football.)
The key to a fair and just patent trial is to choose trial issues that actually are key disputes at issue and that the jury, given their backgrounds, will be able handle. They must be good proxies for theoretically "perfect" examinations of all the facts. For example, determination of infringement could come down to whether the defendant's product has one required feature of one element of one claim, pretty much ignoring all of the other elements and claims. They may be simple decisions to make, but that can be OK -- many "rules of thumb" are simple yet yield better results than more error-prone detailed analysis. (Of course not all simplifications always work.) A common simplification, often emphasized in all trials, is if a witness can be shown to have lied about one thing you may want to assume that they may have lied about something else more material to the case. The battle over which expert you trust more could end up deciding whose testimony you will give more weight.
Shorter trials, with jurors with particular backgrounds, affect how you choose trial issues. Lawyers who are used to more time to cover more issues or in greater depth could end up choosing issues that they cannot adequately address for their client. Other issues, such as inter-party relationship issues that are facts in the case but not part of the patent claims nor allegedly infringing products could end up being the only ones the jury can adequately address to their satisfaction and inappropriately end up determining the outcome. Worse yet, in using such non-patent issues to decide a case in a way that feels "right" from society's viewpoint, the jurors may be forced to ignore validity issues that they cannot adequately address and have ripple effects throughout the industry.
The short time scale common in Eastern Texas for parts of the discovery phase of the litigation brings up additional issues. Here is a bit of discussion of some of those:
From what I have heard, the parties often get no second chances to find material, make a point or counterpoint, etc. Unlike long, drawn out litigation (such as the software industry has seen in the SCO vs. IBM litigation with the initial complaint in March 2003 and discovery possibly still going on in mid-2006) in these patent cases you may only have a few weeks from when you know exactly what you need to prove (such as when the Court decides upon the meaning of some disputed terms in the patent claims) until when you must produce any material to back up your claims. Only a limited number of back and forths are allowed to file arguments. This encourages very deep preparation for all contingencies in advance before you know what you might need. This can be very expensive for the parties and different than other litigation with which the lawyers or experts may be familiar.
I found the strict timetable akin to testing a patent for validity or infringement when viewed at a very deep level: very unforgiving. This entire, unforgiving process means you need great, careful lawyering, and deep, calm experts. Whatever the issues are that get chosen, or that come up at trial, you must be ready to teach them to the jury to a depth with which they feel comfortable. The lawyers who question the witnesses must also understand those issues and how to reach the jurors. With all of the technologies, legal technicalities (e.g., were the patent filing procedures followed correctly?), and business issues (e.g., is the determination of lost profits based on sound theory?), this is a tall order.
There are, of course, always those other, non-patent issues that come up and probably affect the results in some way (e.g., who is the "bad" party?). I cannot speak to those, and assume that those are normal skills involved in jury trials.
In the case of the trial I was just involved in, the judge, as I saw it from sitting in the back and up on the stand, was quite fair. He clearly showed the jury how he cared for them and valued their task. He was warm to us all and funny, but very strict. Watching him handle our case and part of a personal injury case that fit in the middle, he was what I would want looking over a case that affected me if I did not know which side I would be on. He had good reasons for his decisions, whether or not they were what I would have chosen.
The jury I testified in front of seemed to be as able as any I would expect to find anywhere in the country. (I have been on a jury in Cambridge, Massachusetts.) The technical issues we presented were no more foreign to them than to other non-programmers. Many had personal computers and some had used Lotus 1-2-3, I believe. Some watch cable television court programs and were used to lawyer tricks. This was not the "there must be a confession like on Perry Mason" type of jury we used to hear of. They understood the impact of their actions on the rest of the industry. They asked to see exhibits with relevant details during deliberation.
From all of what I have written above, I can see some areas that could perhaps be worked on to improve the results of patent trials. Here are some ideas for discussion:
There is difficult access to potential prior art in the software industry, especially from the pre-World Wide Web days. The software industry did not grow up with patents having much import (they were rare before the Supreme Court decisions in the early 1980's that opened the gates) so it was not the practice to publish or save what was commonly known or to follow other practices you would expect in an area where patents, and patent litigation, are common. Special consideration needs to be given to the time it takes to get actual copies of relevant prior art after the final claim construction is determined. Just locating a physical copy of something an expert remembers and using the legal process to get access from reluctant keepers of that material may take longer than allowed. Experts are often needed at this time. They must be available on short notice, sometimes for long hours to meet the deadlines. This means that their fee arrangements need to reflect the requirement for an option on their time to be held available.
When it comes to challenging the validity of a patent, some of the issues can be addressed at any time from the start of litigation, such as some aspects of enablement or analysis of prior art disclosed in the patent. Other issues are different. Challenges to a patent's validity based upon prior art (either as anticipation or together with other prior art for obviousness) need to actually have prior art that is identified and that can be presented in court. If procedural timetables preclude reasonably finding relevant prior art that actually exists, they would cut off that entire avenue so it could not even get to be considered in a trial, no matter how long or short the trial itself would be. This could be the case if you need to wait for some event, such as the court determining claim construction, before you even know the specific parameters of the prior art which you need to identify. Prior art that could be reasonably expected to be relevant earlier in the litigation, such as that called for by claim construction agreed to early by both parties, would not fall into this category. However, claim construction determined by the court that is different than that proposed by the parties much earlier could open up whole new criteria for determining relevancy of potential prior art. In that case there should be sufficient time after the claim construction process for the parties to identify, procure, and evaluate such art for there to be a fair trial. Keeping relevant issues out of the trial for scheduling purposes that could not reasonably be determined in advance through extra preparation, especially if they could be the keystones for invalidity (the main issue that affects parties other than those in the litigation), is doing a disservice to society.
The whole courtroom process perhaps could be made more appropriate if the jury actually included at least some people (or all) who were "persons having ordinary skill in the art" at the time of the alleged invention. That would help remove all sorts of barriers to getting to the real issues of the claims, though there would be issues making sure that they only considered the evidence in the case and not hazy memories of potential prior art that was not presented. Getting such people during jury selection would not be easy and may require a wider jury pool than that available to a normal district court. However, the issues being decided have national implications so perhaps such an expansion would be appropriate. The litigation is so expensive that a little more expense for a deeper and more informed (in terms of the specific subject area) consideration of the case should not be a problem.
In any case, I think that the task being asked of the jurors, be they regular people or ones trained in the field, is to them as great as any being asked of the others in the courtroom. It pains me to have high priced lawyers and experts whose fees are disclosed in court and who were chosen because they are comfortable with the technologies and legal standards of the case on one side, and on the other side are jurors struggling with arcane terms and technologies and other soporific goings on and having the weight of having to decide for the entire country on their shoulders yet getting paid something like only $40 a day. I would feel much better if the jurors were paid something like that received (not billed) by a normal attorney in the case, perhaps $75 or more an hour. (Of course, the judge may want to give them a simple quiz at the end to check that they paid attention and jury deliberation pay would be capped.) The fee would be split by the two parties. Given the millions of dollars these cases can cost, this cost would add little to the total, especially to the short trials in Eastern Texas.
Finally, a problem I see with the whole system (as it exists now) relates to companies or law firms that do not care about any particular patent but rather play the odds with multiple suits. If you bring suits multiple times, perhaps heavily weighted to contingency fees, there are easy ways to game this system. You can make sure each case has an emotional aspect, such as little vs. big, or aggrieved party vs. bad party, thereby giving a jury a good reason why society may want to find for one side vs. the other. You then keep the technology discussion simplistic but confusing in the details, make broad claims in an area hopefully unknown to the average juror and make the patent seems more groundbreaking than it is, give a few simple reasons for infringement and lack of prior art so that the jury has something to "hang their hat on", and push for high damages to increase the expected value. (Compare the impact of a longer argument about validity that increases your chance of being paid for damages from 30% to 60% to the impact of a longer argument on the amount of damages that raises the value from $10 million to $50 million. Which, totaled over a few trials, is expected to pay off with more money? It's the damages arguments.) Perhaps by the luck of the draw you will get a jury that will find that the emotional issues resonate and that they really do not understand the technical issues well enough to feel clearly convinced the patent should be invalidated. You do not have to do too good (or expensive) a job at the technical end, just give the jury a reason to go with you if they want. You only need to win once in a while to make that a financially rewarding business. Unfortunately, not only do companies end up needing to mount a very expensive and distracting defense and perhaps be mortally wounded financially no matter what the verdict (they cannot afford to do an inexpensive job and play the odds because they have no cumulative upside, only downside), but society also loses because the patents are not given the examination that they deserve and sometimes acquire enhanced status that they should not have and go on to cause problems elsewhere in the industry. In addition, many smaller companies targeted by such suits cannot afford even the most basic defense, let alone that needed to get a fair trial as I have described above.
I hope the ideas and thoughts presented here are helpful in discussions to make our handling of patent infringement cases the best they can be for society.
-Dan Bricklin, 11 October 2006
© Copyright 1999-2010 by Daniel Bricklin
All Rights Reserved.