Lawsuits without Injuries?

December 17, 2003

Unedited transcript prepared from a tape recording

2:45 p.m.	Registration
3:00	Moderator:	Richard A. Epstein, University of Chicago
	Panelists:	Michael Kelly, University of San Diego
		Brian Anderson, O’Melveny & Myers LLP
		Moin Yahya, University of Alberta
		Jonathan Klick, AEI
4:30	Adjournment

Proceedings:
MR. ANDERSON: [In progress.] I also had this naive view that at least in federal court, and in most state courts, in order to have standing to bring a lawsuit, one has to have been injured in a material way.

Well, I'm 20 years older than I was then, and I have defended hundreds of class actions over the course of my legal career, and therefore have come to conclude that my quaint notions of lawsuits being designed to remedy actual injuries does not square with reality.

My experience, to the contrary, has been that the large majority of class actions that are brought these days are brought either by people, named plaintiffs, who have not themselves suffered any injury from the conduct complained of and most likely never will.

And, second, that the classes on whose behalf these lawsuits are brought contain people, the vast majority of whom have not experienced any injury from the conduct complained of and probably never will.

Now, my job, as a defense lawyer on this panel, as I perceived it, was to take the hard-line position that we ought to enforce the injury requirement strictly--no exceptions whatsoever. But with four academics surrounding me, I thought I should try to bend towards at least their notions of intellectual honesty, and so let me admit a few theoretical caveats to that principle that I think most people would agree with.

There are some theoretical exceptions to the principle of present injury. One type of theoretical damages that these lawsuits often are based on is fear of future injury. I, the plaintiff, am entitled to money today because I have a substantial fear that at some point in the future I am going to be injured or I, the plaintiff, am entitled to injunctive relief today because I am in substantial risk, I have a substantial risk of incurring an injury in the future, and an injunction by the court might prevent that or minimize it.

Now, I think most reasonable people would agree that in certain circumstances, that theory makes sense if there is a very high risk that you are going to be injured in the future. You take a drug, and 90 percent of the people who take the drug are going to die, and there is something that can be done to prevent those injuries; you know, a medical monitoring program, early medicinal care. I think most reasonable people would say, well, in that instance, it doesn't seem absurd to bring the lawsuit now and seek some sort of injunctive relief.

The second exception to the actual injury requirement, the second kind of theoretical damage theory that we will be talking about today is the diminution in value theory or sometimes called the benefit of the bargain theory. And the theory there is I bought a product or I paid for a service having certain expectations about that product and service, and so I paid a certain amount of money having those expectations. Now, I own the product, and I find out that it is not what I thought I was getting. It is less useful to me, and therefore it is worth less, in some abstract sense, than I paid for it.

Now, again, I can see some situations where that might be appropriate. An example that immediately comes to mind is the diamond wedding ring that is not made out of diamonds. I go to a jewelry store, and I buy a very nice wedding ring for my fiance, I pay a thousand dollars for it because the person says this is the highest-grade diamond. I take it home, I give it to my wife, six months later an appraiser says this is not a diamond ring, this is a cubic zirconium ring, and it is worth maybe $100, at the most.

Now, I have not incurred any diminution in value in a tangible sense. Certainly, I'm not planning to sell that ring to anybody and get maybe $90 for a ring that I would think I could sell for $900. And in a real sense, the utility is not diminished because, you know, on somebody's finger a cubic zirconium ring looks pretty much like a diamond ring. But still I think reasonable people would agree that I'm entitled to a substantial portion of my purchase price back because I really thought I was getting a diamond ring. That facet was the central part of the reason why I chose this product, as opposed to another, and I grossly overpaid for it.

So let's have some common ground that there may be instances in which an appropriate departure from the no injury requirement is appropriate. But what I have observed, in my 15 years of defending class actions, is that those narrow exceptions to the traditional injury requirement have been greatly expanded over the years to justify a great many class actions that, again, are brought by people who have not been injured, probably will not be injured, on behalf of proposed classes of people, the vast majority of whom have not been injured, probably won't be injured.

Think of the different kinds of class actions that are brought these days, and you will see this pattern at work. Take consumer products, which is what I spend a lot of time litigating. Remember the laptop class actions of a few years ago? There were a number of laptops that had a bug in them, such that if you performed this very complicated mathematical calculation, the number would come out wrong. And so class actions were filed against the laptop manufacturers on behalf of everybody who bought the laptop saying the laptops are defective.

Well, yes, maybe they were, but the reality was that the named plaintiffs had never tried to do one of these calculations on this laptop. So they had never been harmed by this bug. And the chances that the class members would ever perform this calculation and ever be damaged by this bug were infinitesimal.

Nevertheless, the lawsuits were brought in courts where defendants do not stand a very high chance of persuading the judge to dismiss the case, and so several laptop manufacturers entered into very rich settlements in order to get rid of those lawsuits.

Consumer services are another major class action arena. A service provider such as an HMO is sued for some aspect of its business practices. In the case of the HMO, it is the manner in which medical services are reimbursed--how much is paid for certain medical services, what hoops you have to go through in order to get reimbursed.

And the consumer class actions brought against the HMO industry down in Florida were predicated on the theory that this was wrongful conduct. The only problem was that the named plaintiffs in those cases, and the vast majority of the class members, had never tried to get the types of medical services that were allegedly being wrongfully withheld or underfinanced. They were never going to be injured by the policy that they were attacking. Yet the lawsuit went forward.

Consumer fraud is another arena for class actions, and I like to call these the "gocha" class actions. Companies make mistakes. They make labeling errors on their products. They may get the specifications of the product wrong on the package or, in these complicated financial transactional forms, you know, credit cards, loans, cell phones, these sorts of things, you know, sometimes errors creep into these forms. You see a class action.

The service provider, the manufacturer lied about some feature of the product, and so you get a class action by everybody seeking some money for this misrepresentation. But when you look into it, you find that invariably the named plaintiff really didn't care about the specification that was inadvertently misstated or about the feature of the service that was misstated, and again the vast majority of the class members will never be adversely affected by the error. It is the proverbial tree in the forest that nobody hears, but still you get a class action out of it.

And then, finally, the mass tort ground. Medical devices, pharmaceutical devices, lead paint, asbestos, you name it, products that "X" percent of the time will cause somebody to incur a substantial injury, but it is a very small risk in the scheme of things--less than 1 percent--but still you see the lawsuit by everybody who's been exposed to the lead paint or the asbestos or taken the pill or used the medical device for money damages today based on the fear that they may be among the less than 1 percent that are injured in the future.

I would like, in my remaining time, to identify five major problems with allowing these kinds of lawsuits to go forward.

The first is the lawyers' lawsuit problem. Rigidly enforcing the injury requirement is a terrific gatekeeper to separate the wheat from the chaff. To require somebody who has been injured or to limit the courthouse doors to people who have really been injured and really care about their issue is a great way to separate the frivolous lawsuits from the real ones.

But a watered-down application of the injury requirement allows plaintiffs' class action lawyers to invent the lawsuits themselves, and then find somebody in whose name to sue, whether they have been injured or not, and sue on behalf of the proposed class. And so it clogs the courts, and it ends up imposing a litigation tax upon the defendants for conduct that real people really don't care about.

The second problem with these kinds of lawsuits is the windfall award problem. Again, if these kinds of lawsuits are successful, they end up paying money to large numbers of people who have not been injured and never will be injured. Take the fear damages issue. I may be afraid of being injured in the future, but if I use the product for its useful life and it performs effectively, why am I entitled to any money? And if I do get an award because somebody brought a lawsuit on my behalf, that is a windfall award.

The same thing with the diminution in value damages. Yes, maybe in the abstract, if I buy a laptop computer for a thousand dollars and it has a bug that may potentially occur at some point in the future, in some abstract sense, all right, maybe it's only worth $950. But if I use that laptop until it becomes obsolete, and I never experience the bug, why should the manufacturer being paid me and the 99 percent of the other people who won't experience the bug $50? I got what I paid for. I'm not entitled to more.

The third problem with the no injury class actions is the loss-spreading problem. We all learned in law school about the loss-spreading doctrine, and that is that embedded in the price of every good or service is a little amount of money that is designed to ensure that the small percentage of purchasers who are actually harmed by some defect in the product or some mistake in providing the service can get compensated for their actual out-of-pocket losses stemming from that defect or from that mistake.

But if you entertain a system in which everybody who buys a product for, let's say, a thousand dollars, brings a lawsuit, and then gets $100 back because there is a risk of future failure or there is some abstract diminution in value, you're not creating this little pool of money to compensate the people who are really injured. All you are doing in the end is returning 10 percent of the purchase price to everybody who bought the product, skimming some percentage of that off for attorneys' fees, and increasing the price of that product by 10 percent for everybody who buys it in the future.

So you have made it harder for the defendant manufacturer or service provider to have a pool of money to compensate real victims, and you have raised the price of the product or service to such an extent that it exceeds what most rational people would think is the fair value for that product or service.

The fourth problem with no injury class actions is what I call the judges and juries over their heads problem. Courts are good at resolving real disputes between real people. You get a plaintiff, you get a defendant, they each tell their story, a jury decides who wins. If they side with the plaintiff, they award damages. Often, the defendant can afford to incur that judgment, even if it's wrong. In the whole scheme of things, it's not that big a disaster. And if the plaintiff wins on a particular issue many, many times, maybe the defendant decides, you know, we need to change our product or we need to change our business practices. That all makes sense, and that's a good use of the courts.

But courts are not very good at getting into these abstract public policy decisions about how should a product be defined, what level of risk is acceptable in our society, how should HMOs be run, how should these rights be allocated? Those are not things that courts are good at. Why? Because in a lawsuit, the plaintiff's lawyer controls what the lawsuit looks like. They take, in the case of these no injury cases--take a product defect allegation--they will take one arguably undesirable feature of a product, and they will make that the entire focal point of the lawsuit, to the exclusion of everything else about the product.

And so if you design a product with certain advantages and disadvantages, and costs and benefits, these lawsuits will take the arguably undesirable aspect of that product, and that is all you will hear about in the litigation, and there will be no consideration given of the fact that there are countervailing benefits to that design, and perhaps the cost is lower because of accepting the risk of that design.

None of those factors get into the litigation. It is a very myopic way to litigate a very complicated product design issue or service-providing issue, and it also lets lawyers--trial lawyers--grandstand, particularly if you don't have a rigorous injury requirement. Let me give you an example involving antilock brakes, which was the subject of the Brio v. General Motors [ph] case, which was the first federal appellate decision that banned no-injury class actions in the product defect realm.

The plaintiffs in that case attacked antilock brakes on the grounds that when you are in high school, you learned that when you're supposed to stop the car quickly, you pump the brakes so you don't skid, but that's not how you use antilock brakes. You're supposed to apply constant pressure. And so the allegation was that antilock brakes are defective because some people are going to use them incorrectly.

Now, imagine this case. The plaintiff stands up, the plaintiff's lawyer stands up. He doesn't have an actual injured victim. It's a theoretical injury, diminution of value, fear of future injury. He puts in a few anecdotal stories of people who died because they didn't apply their antilock brakes. He waves a few internal memos from GM around some engineer said, "Well, you know, this is a downside of antilock brakes," and waves the memo from the dissenting engineer who didn't want to go to antilock brakes, and he gives a great closing argument, and what does the jury do? Awards everybody who owns a car with antilock brakes some amount of money.

And then the lawyer asks the judge to recall all cars with antilock brakes and replace them with standard brakes.

Now, go down the hallway of the courthouse a few steps, Courtroom B, another class action brought on behalf of everybody who owns cars with traditional brakes. The allegation is that traditional brakes are defective because they're not as good at avoiding skids. Same plaintiff's lawyer, brings in other anecdotes, people who died because they didn't have antilock brakes, brings in the memos from the engineers, "You know, we really should go to antilock brakes. They are more safe." He gives a great closing argument. That jury awards everybody who owns cars with standard brakes some amount of money, and the judge orders the standard brakes be recalled and replaced with antilock brakes.

This is no way to design a car, and it is an illustration of what is wrong with allowing juries to get into these sorts of things.

The final problem with these kinds of lawsuits, and maybe I'll return to it in the question and answer session, is that it usurps government agency authority. Plaintiffs' lawyers like to say you need these kinds of lawsuits in order to hold companies accountable for their products or their services. That's not true. There are myriad federal and state regulatory agencies that have far more expertise, far more access to information, are far more impartial and are much better able to make these kinds of complicated decisions in the public interest than are private class action lawyers who, let's face it, ultimately are looking to generate a big attorney's fee.

So I look forward to talking about these further.

[Applause.]

MR. EPSTEIN: [Off microphone.] On that happy note, Michael, you will give some [inaudible] commentaries [inaudible].

MR. KELLY: Well, I hope.

[Audio break.]

MR. KELLY: Yes, I think I'm the panelist designated to be most sympathetic to remedy without injury, although I'm reminded of Yale Kamisar telling me in law school that if he could frame the question, he would take either side and win. If you treat this as lawsuit without injury, then of course it's pointless.

The problem is, of course, that it's really a narrow definition of injury. Is there a real injury that isn't a physical injury to an individual? I teach contracts, and the answer is, yes, all of the time, right? If you paid a thousand dollars for a computer that was worth $950, you have been injured to the tune of $50. That's a real injury.

The question is, then, should it be compensated? And in contract we would treat the misrepresentation, if you will, that the computer would perform as a standard computer would, as a warrantee that it would do so, and as a warrantee, the measure of damages would be, very simply, the value of the machine as warranted minus the value of the machine as delivered. And, thus, if the difference is $50, end of story.

That would be true, by the way, even if the machine with the defect was worth a thousand dollars, right? You paid a thousand dollars, you got a machine worth a thousand dollars, but you were promised a machine that was worth $1,050. You were promised a profit. You were deprived of the profit. End of story.

Now, there's some question as to whether that should be applied in tort, if you're suing not for the breach of contract, the breach of warrantee, but for the tort, and the answer throughout much of the nation is why should it be different in contract than in tort? Right? It's the same misrepresentation. Whether you plead it as contract action or plead it as a tort of fraud, the statement is the same. The loss, in terms of the loss of the benefit you were promised, if you will, or at least led to believe you would receive is the same.

Why should we have tort of fraud be compensated at a lower level than we would compensate the breach of contract, which doesn't require any culpability at all? There doesn't seem to be much justification for that, let alone much way to divide a statement made into those that are properly treated as warranties and those that are properly treated as fraud. There's no particular way to distinguish those, on any formal basis, and so it really does become a choice.

Once you admit this, once you accept this, I take it we're not dealing really with recovery without injury. There's injury. it's financial injury, and the question might be whether to apply the economic loss doctrine so that tort doesn't extend into what is primarily a contract realm of economic losses. Of course, in cases of fraud and misrepresentation, the economic loss doctrine is never applied because economic loss is the traditional loss that is involved in fraud cases.

Now, theoretically, it's perfectly sound to say tort and contract ought to be different. In contract, we put you in the position you would--in both--we put you in the position you would have been in if the wrong had not occurred. In contract, the wrong is the failure to deliver the promised good.

But in tort, the wrong is making the false statement. Where would the plaintiff have been if the truth had been told? Well, then he would have not bought the machine at all. So it's only the amount that he, what, the difference between the amount he paid and the amount he would have paid-- the out-of-pocket rule--that would govern in tort. Perfectly sensible theoretically, except for the fact that it's hard to distinguish the two.

It's also true I think that that provides too little deterrence value. There's very little disincentive to fraud if, when you aren't caught, you get to keep your excess gains, and if you are caught, the most you have to refund is the difference between the price you collected and the fair market value. You never lose as a fraud feasor in that case. You might be forced to live up to what the fair deal would have been.

But if you want to deter fraud, you want to impose some additional sanctions. Now, of course, there are some systemic sanctions such as attorneys' fees. If you can sell it honestly, you don't have to worry about the suit later and thus lose, sink your profits into attorneys' fees. Of course, the attorneys' fees apply to both sides. Attorneys' fees apply to both sides if you eliminate--

PARTICIPANT: [Off microphone.] [Inaudible.]

MR. KELLY: Yeah. --if you eliminate the lawsuit without injury, if you eliminate the benefit of the bargain. If you drive the suits out entirely, you've got an even deeper problem, right?

But the penalty--I shouldn't say the penalty. I don't want to talk about punitive damages. The remedy ought to be crafted to the point that the compensatory damages are a sufficient deterrent.

Now, there's one other thing that I want to keep in mind, and that's really a reliance measure, not an expectation measure, much more attuned to tort, although still in a contract setting, which is where almost all of the examples come from, and that's the lost opportunity.

Everybody was buying a computer that had this defect in it, had other manufacturers in whom they could have bought a computer probably without the defect. They could have, for a thousand dollars, gotten the machine that wouldn't have had the defect. It's not at all clear that if the truth had been told, they would have continued to buy this machine. This manufacturer wouldn't have gotten that thousand. Even if that's the fair price of the machine, it would have gone to somebody else who was providing a machine with the $50 profit, not to this manufacturer, who was providing a machine that didn't have the $50 profit.

That fact seems to drop out of the equation in a lot of these settings. What would the person have done in other settings? Have you deprived them of the opportunity to make a better bargain, a more profitable situation? And by eliminating the economic loss or down-playing its importance, we lose sight of the opportunities the individual had. In fact, we end up undermining the ability of individuals to make their own deals, to make their own best deal.

Now, I don't think that, what, the tort system is designed to protect the ability of individuals to make good deals on their own. In fact, the tort system is, if anything, designed to say we don't care what kind of deals you would make. These are your duties anyway.

But it's nice that when we get into fraud, we eliminate the conflict between contracts and torts. If we can keep them both aimed at the same goal, which is protecting people's ability to make profitable deals, I think we've got a better system overall.

Now, that's the outline of the position, and the question is does that justify all of the ills that are practiced in the name of lawsuit without injury. I won't begin to tell you that it does.

What I will say is that, in most of the cases given--computers, the lead paint, the drugs, medical devices--we are dealing in a realm of contract, where if the contract was made on the basis of a misrepresentation, if the price paid was paid only because the buyer thought he was getting something more than what he was actually getting, then there is, in fact, an injury. Nor do I think that it will make much difference whether you choose benefit of the bargain or out-of-pocket [audio break] would have made minus the value you received or whether you say you're entitled to the price you paid minus the amount that you would have paid had you known the truth. Either way there's going to be an injury.

This is not a way to say the injury requirement hasn't been meet. It's simply a way to say we're going to measure it differently, and the different measurement will make a difference in some cases: Where the price, in fact, is the fair market value of a computer with a defect, there will be no recovery because that measure of damage is smaller compared to the benefit of the bargain, which might get you the extra $50, but it's going to make a difference in a relatively small number of cases and not by throwing out of court at the initial stage cases because they lack injury. That will simply depend upon the evidence of value.

Now, I've exhausted my remarks.

MR. EPSTEIN: You're allowed to stop soon.

MR. KELLY: Okay.

MR. EPSTEIN: And I will take up the time, of course.

MR. KELLY: Fair enough.

MR. EPSTEIN: But in the interest of--Moin, I think you are third on our list, and since we are back on schedule, our 15-minute rule will be there, and Michael will receive a gold star.

MR. YAHYA: Thank you. I'm using the overhead there.

I'm going to focus mostly on products. I know services were mentioned, but I'll stick to products. So let me just start briefly by saying this is essentially what the plaintiffs' attorneys are saying. If we have a product, and there was a demand for this product, and people bought Quantity X, paid the Price B star, what they're saying essentially is that had they known there was a fault, they would have had a lower demand for this product, and therefore would only pay the amount P1--that's the difference in price--and that's the benefit of the bargain that they're seeking.

So we need to figure out how to estimate this demand for the faulty product. That's essentially the issue here.

But what I want to do, first, is go back to sort of the legal foundations for this and argue first that both in law and in economics, there is no recovery. So the first thing is to understand the basic scheme of tort law is tort law is either you have trespass or action on the case, which is essentially intentional torts or negligence.

The first category protects personal fundamental rights, personal security. This is like assault, battery, false imprisonment, those sorts of things. In those cases, we actually don't need damages for recovery. Damages are presumed. If I walk across your lawn, you can stop me. You can sue me for trespass, even though what's the damage? A little bit of grass blades have been trampled. So damages are presumed.

If I lock you in the room for five minutes, I've falsely imprisoned you. What's the harm? Were you going to leave? There's still damages.

So where we're trying to protect personal integrity and intentional security, intentional torts, we presume damages, we allow nominal damages, plus punitives, because we're trying to prevent fundamental acts of violence and trespass.

But in cases of negligence, where these are not intentional actions to harm someone's body or security or property, the law and the action on the case always require damages. And negligent misrepresentation--which I'll get into in a second--is a form of action on the case. It's not a form of trespass. And so, legally speaking, we always need harm in sort of the negligent misrepresentation cases. The purpose of negligence is compensation, deterrence, things like that--not an issue of a personal--no personal right is being vindicated.

There are also equitable remedies. I want to briefly introduce this, so we get a bit of background where the plaintiffs' attorneys are coming from. This is where harm to plaintiff is less than gain to defendant. Example: I pay somebody to attack you, but the hit man is a bad hit man. He shoots, misses. So the harm to you, apparently, is minimal, but you have a million dollars in your bank account now because you took this case.

I can elect the waiver, what's called waiver in tort. I can say, look, I wasn't harmed much, but I want his million dollars back. That's some sort of unjust enrichment to him. I want that as a restitutionary measure.

There's the Great Onyx Caves case, where somebody had been trespassing the caves under their land, and the courts awarded them half the profits, even though there was no evidence that they could have actually used the caves under their land.

So this is, at best, another theory that the plaintiffs' attorneys come from. It's that there is some form of unjust enrichment; that the firm is making money at the expense of the consumers, and they're keeping that extra cash.

Misrepresentation. It's a doctrine that comes, as Professor Kelly talked to us, it's about tort and contract. Some have called it "contorts." It's an action on the case and called that. So it needs actual damages.

The economic loss rule prevents recovery for pure economic loss, except, and I'll give you a brief thing about this, whether it tends to be special relationships. There are damages, you can have benefit of the bargain, out-of-pocket. We just talked about that. So let me skip over this. But let me remind you that benefit of the bargain is fundamentally an economic loss. It's fundamentally an economic loss. You bargained for this, you got something else, and you want the difference. That's a pure economic loss.

What are out-of-pocket? Well, it's the difference between what the consumer paid and the market price. Well, the market price on most of these products either goes up or is the same. So you should get, under the out-of-pocket measure, zero compensation. Let me say that most case law rejects recovery for economic losses under negligent misrepresentation, not fraudulent.

Where there is fraud, we can argue that there was no contract to begin with. This is a tortuous action. But under negligence--and you need allegations, when there's no fraud, most cases will say that there is no recovery. The few states that allow this require very, very close special relationships, some idea of reliance, and then a lot of them will require fraud or there has to be almost gross negligence.

Let me give you a couple of examples. There's a case from Tennessee where these farmers used a product that was advertised as designed to protect against frost damage. They used it on their tomatoes. The tomatoes all got damaged. Sued. The Supreme Court of Tennessee said economic loss, bar recovery.

There's a case where a helicopter crashed, actual crash, because the service manual had represented that the service life of the helicopter was so many hours or so many days. They followed what the manual said. Helicopter crashed. Sixth Circuit said economic loss, bar recovery.

The Supreme Court of Georgia in a case where they allowed a negligent misrepresentation to recover for economic losses said you must actually prove with certainty the economic losses, and they must manifest, they must actually appear before you can go ahead and sue.

So I think the requirements for negligent misrepresentation, to allow benefit of the bargain, there's very few cases. Those cases that I did find that allowed benefit of the bargain on negligent misrepresentations, when you actually read the facts, they sound much stronger like fraud or gross negligence, willful indifference, those sort of facts. So I would argue that the case law does not support any recovery for a negligent misrepresentation.

Given that we're dealing with products, we should look at an objective measure of damages, because these are consumer products. Now, let me move to what would the consumer pay. Let's get back to this graph.

I'm going to argue that a firm--this is the point that was made earlier--that a firm charged an insurance premium--the loss spreading-- that reflects the expected damage for the product. Example: There's one chance that it will cause $100 worth of harm, which the firm will charge a dollar extra as an insurance premium. So the price is always the cost plus the insurance premium.

What would the consumer pay? Well, if the consumer knows that there's going to be compensation, then the consumer would also pay the price, plus this insurance premium, so that they are protected in case of an accident. The consumer, therefore, will always pay the same amount, regardless of whether they know that the product is good or bad.

Let me give you a simple model. A firm charges price, cost plus cost of care. The firm wants to minimize this. They ultimately charge the price down here, which is the cost, plus the cost of care, plus the expected damage. They charge that insurance premium.

Now, if the product is good, the consumer will maximize their utility, subject to a certain budget constraint. You don't have to worry about the math. Essentially, they will essentially have a demand that's given by the price that they face.

Now, assume that the product is faulty, and there is no liability. They cannot recover from the manufacturer. The consumer is essentially facing a price that's equivalent to the price of the product plus the harm that they're going to face. So they will lower their demand. They will demand much less because now they're going to have to insure themselves against the harm.

Well, the point is that if the firm has no liability, yes, then the consumer has to insure themselves. But I thought the whole point is that the firm is the least-cost insurer. That's the whole theory between strict liability. So any insurance that the consumer would purchase would always be greater than any insurance that the firm would give you in a strict liability regime. Therefore, the consumer can always never pay more or will always pay no more to the firm, in a strict liability regime, than they would have to pay themselves if there was no compensation.

Now, let me make it more formal. If there is strict liability and the consumer has paid, then there's a state of the world where they will lose some money, but then, because there is a strict liability regime, they get it back if they sue. So then what is their demand? Their demand returns to the original demand, which is what they would have paid had the product been not faulty.

Now this is sort of a bold statement. We can modify this a little bit and say, well, what if I've got intrinsic utility, they would have bought a little less than what they would receive in compensation? That's an empirical question, then. Now we need to look at the empirics of that.

At this stage we can say there is no unjust enrichment. There is no benefit of the bargain. So whether you take equitable remedies or whether you take negligent misrepresentation, there should be no recovery.

PARTICIPANT: [Off microphone.] [Inaudible.]

MR. YAHYA: Well, that's an action on contract. I'm in a pure tort--

PARTICIPANT: [Off microphone.] [Inaudible.]

MR. YAHYA: Well, if you're delivering a good, and there's an express contract, then we--I am in a pure, you know, compensation for harm, you know.

The way we would actually estimate, we would actually figure out how much this difference that two demand curves are is let's look at those products before we found out about the defect and after we found out about the defect, and we can actually measure, controlling for all of their factors, the benefit of the bargain.

And I would argue that, for example, in Oxycontin, which is this drug that supposedly addicts 1 percent or a few percentage of the people, and now there are these consumers who are not addicted to this pain killer and are suing, I would argue, first of all, there's very few alternatives. This is a drug that was unique. So what is their alternative? They would still buy Oxycontin.

In fact, I was looking up their financial records. The demand for Oxycontin has gone up through the roof. It's gone up eight times. The Purdue Pharma can't keep up with the demand. In fact, if anyone, the consumer who bought it the first time should be compensating Purdue Pharma, not the other way around. The implicit price is much higher.

The other issues I have with this is, you know, one of the conditions--there are like eight conditions for negligent misrepresentation. It's not just that an intentional representation was made concerning a presently existing material fact which was false, et cetera, et cetera, and one is that there has to be justifiable reliance by the consumer.

Is there actual justifiable reliance? Did the consumer buy the laptop thinking that they were going to actually do this extra computational thing? Did they buy Oxycontin and think they were not going to be addicted? You have to prove that. So that is another legal, I would argue, a legal bar to all of these cases.

Let me throw in some just general policy issues. Issues of cost of the product would become prohibitive, that's just a general concern on the side. It's not a legal or economic argument against this.

I would raise some First Amendment issues; you know, speech by the firms. Are we now essentially censuring what the firms are allowed to represent?

There are standing issues, and I would argue that if we abolish standing in state courts, is this a form of judicial takings? I would just bring that up.

So the conclusion is that for actual product liability cases, you need injury. Whether it's a legal theory or an economic theory, you should have to show actual injury--physical injury-and not economic loss before recovery.

Do I have a minute left? Okay.

I know Professor Epstein is here, and Professor Epstein was educated at Oxford. You're an English lawyer. So I can't compete with that, but what I can compete with is get some inspiration from the English series, "Blackadder." So here is Blackadder telling us the difference between intentional torts and negligence actions.

[Videotape played.]

MR. YAHYA: Anyway, that's the--

[Audio break.]

MR. KLICK: Thank you, Richard.

Since Moin handled the sort of ex ante incentives of economics behind these lawsuits without injuries, I thought that I would just handle the sort of static economics behind cases that currently exist.

So what constitutes harm? Well, it sort of depends on who has the property right to the surplus. Let's take the Oxycontin example, where the claim is basically that the purchasers assume that they were buying a drug that was nonaddictive, and had they known that there was a chance of addiction, the price they would have paid would have been significantly lower.

So the question is, if that's their demand curve that doesn't change in a full-information world, the difference is whether do they pay Price 1, where the harm is not known ex ante, or do they pay Price 2, where the harm is known ex ante? And so, depending on which price they pay, the gain either goes to the producer or the gain goes to the consumer and consumer surplus.

What if the consumer has the right to the surplus? Well, presumably the loss isn't just P1 minus P2 times the quantity purchased, which would be the difference in the two prices times how much that they paid. It wouldn't even be recision value. Presumably, they should also include the surplus for units that would have been purchased at the lower price.

But what's the problems with this conclusion? Well, for one, there's a difficulty in determining Price 2. It's not observed. There's also a difficulty in estimating the unobserved portion of the demand curve for each individual or even for the market as a whole. Perhaps at Price 2 the product would never have been produced at all, leading to zero consumer surplus.

If the producer surplus were poured into R&D, innovation might decrease on the margin, and so you might even have some future effects of such a regime as the property right going to the consumer.

But if the producer has the rights to the surplus, well, clearly there's no harm in that case, since the consumer is better off under Price 1 than he would have been under a regime where he purchased no product, and we don't have the same valuation issues arising. But beyond those administrative differences, really the question comes down to where should the property right to the surplus reside? And, in principle, I don't know that there's a clear answer to that. In principle, perhaps that a policy issue and not a legal issue.

So, so much for the distribution. What about the efficiency? Well, if the benefit of the bargain damage theories or these lawsuits without injuries are allowed ex post, there can be no effect on the efficiency of completed transactions. That's tautological. This effect is a pure transfer.

Going forward, market prices will capture the insurance premium for this type of harm, as shown in Moin's presentation. On the margin, some products will cease to be produced, depending on the elasticities, which will lower consumer welfare, except under some nonstandard utility functions that explicitly include distributional issues.

More importantly, and when you talk to some of the trial lawyers, which I did trying to get them to join this panel, one of the things that they are particularly interested in is they say, well, what about the incentive effects on fraud, misrepresentation and disclosure? Well, of course, increasing the penalty for fraud or misrepresentation or even for a mistake, depending how you want to define these things, will obviously lower the incidence of those things.

However, the unharmed users are inframarginal users; that is, their use is unaffected by the changes in disclosure under full information. So the efficient level of disclosure will be determined independently of these users.

Now, the efficient level of disclosure will hinge on individuals suffering actual harm from using a product they would not have used under full information. In principle, economically, it doesn't really matter where you provide the loss to create these incentives, but somehow it seems a bit more natural to provide these disclosure incentives to the cases where harm actually occurs, for lots of information reasons, but it's not necessarily the case. In principle, again, it's just a decision that should probably be made on empirical grounds on what is administratively cheapest.

That's it.

MR. EPSTEIN: [Off microphone.] I think that there is a peculiar disconnect in this panel, a peculiar disconnect in this panel between what was modeled and what was presented, and let me say it in the following fashion:

I think Brian starts to put cases forward which is not concerned with the question of who is going to bear the risk with respect to a certain class of defects. He did mention the diamond case, which I think is closest to Michael's model, but the truth about the matter is, if the benefit of the bargain has not been supplied and you count as injury the loss of that benefit, this is nothing which is not available in the king's courts in 1500, and that's not what we're worried about.

What you're worried about in these cases I think are a couple of features. One of them is the kind of cases in which you have successfully supplied the very benefit that you have promised, a product has then been recalled, and the plaintiffs will ask, in effect, for a refund of the purchase price for the product that worked, rather than for the refund of a purchase price of a product that failed.

This often happens in the world of drugs; for example, with Resolin, which is a case now undergoing. It's on the market, lasts for two or three years, gets pulled for a whole variety of reasons, and the claim is that all of the individuals who took the drug did not suffer any adverse side effects and, in fact, benefitted from its medicinal properties are entitled to some kind of recovery.

The explanation that is sometimes given is that if they had known the risks, they would have bought a cheaper product, but that then requires you to make some implicit judgment that the cheaper product would have provided the same set of benefits as the one that's there. And I think that Michael would agree that if the substitute product, in fact, has a risk return ratio which is rather different from the one that had been sold, you can't simply assume that the difference is pure surplus which has to go back. You have to ask, at the very least--

[Tape change.]

MR. EPSTEIN: --offsets the loss in question. That's one set of things.

The second problem I think that one worries about in these cases is the coordination of damage action for the actual harm that's been suffered with, in fact, the refund with respect to the anticipated harm which may or may not have been realized.

And a case I think which illustrates that is the situation that we had with respect to the tobacco litigation against Philip Morris in Illinois, my home state. These were cases of low tar and nicotine cigarettes, and the theory, in effect, was as follows: That when you reduce the amount of tar and nicotine, what you will do is create an off-setting incentives on individuals to puff more deeply and to smoke greater numbers of cigarettes, so that all you do is you get poisoned in a different fashion rather than not getting poisoned at all, and therefore the argument is that the product was not the product that you wanted.

Now, at this point, if you actually do a diminution of value situation, it's clear that the diminution of value is not going to equal the purchase price of the product in question, and yet that's the measure that was used by the judge when he calculated the cool $10 billion in damages in this particular case instead of trying to figure out what the quantification was.

So let's assume that you do have a risk and it's 1 percent. At this particular point, the issue is do you get 1 percent for everybody now, which is one possible alternative? Do you, in effect, wait until one fraction of this group gets a form of cancer which is attributable to excess puffing on light cigarettes and then award them the full damages? Do you do neither or do you do both? And, of course, since we've miscalculated the basic damages in this case, we might as well continue to compound the felony, and in the Illinois case, there was a specific reservation of rights with respect to any injured problems, saying we're just giving them money back in effect for the insurance premium, and then when it's over, we'll allow you to recover both ways. So that, in effect, what happens is you pay for future injuries associated with the products twice, one in the form of reduction of the price in question and, secondly, in terms of compensation that you get with respect to the future kinds of injuries.

Now, it is not all courts that do this. The Seventh Circuit, which is not in the same state as Illinois because it's the federal system, there is the very well-known Firestone/Bridgestone tire case, which Frank Easterbrook has written a number of opinions about. And I think one of the things that Easterbrook said, I think with some degree of force, is you could play it one way or the other way. You cannot play it both ways.

If you are going to talk about this as a risk of taking additional, the loss is the risk of additional harm, then, if you get that, it's the duty of the plaintiffs to take those proceeds and to insure over against the particular risk. You cannot give both measure of damages in these particular cases.

And if that's the sort of range of cases that one is considering under these circumstances, it seems to me that it's not going to solve the question to ask the or to make the observation, which I think is implicit in the presentations of both Jonathan and Moin, that you have a world in which, if you knew who was going to bear the residual risk, the prices would adjust such that there would be differences between the two parties in a zero-transactions-cost world, and the risk would be assigned to that party which could insure most cheaply, in a situation where incentive effects don't matter, in a positive-transaction-cost world.

And I think, for the purposes of discussion, in many ways, one wants to try to focus on the cases that are distinctive. Not all of the cases that were talked about by Brian, I don't want to talk about the diamond case, but I do think that the double brake case is a classic illustration of the problem or another variation on this is you now give somebody--everybody, let us assume--a certain amount of cash by the virtue of the fact that they either have brakes that don't lock or do lock or anti- or nonanti-, and then the question is, after an injury occurs, you proceed to allow the defect cases going forward.

Should we allow the sequence to include one, to include both? If [audio break] framework. So, with that, let's just open it up, and if anybody has a question, I shall not moderate in my usual moderate fashion.

Kate has a microphone. That doesn't mean she's going to ask the first question. It doesn't mean that I should talk indefinitely. It means that Sam should be given the microphone.

PARTICIPANT: Thank you.

Several of you folks got into the question of what consumers would do under full disclosure. And the question I have is to what extent it even makes sense to talk about a consumer having full disclosure. Because for any of these trade-off design choices, it seems to me you could set that consumer down with one advocate and talk him out of purchasing the product entirely and then sit him down in another room with another advocate for the other side and talk him into buying anything whatsoever.

Does it really make sense to talk about what that consumer would have done, what price he or she would have paid if they had known a certain set of facts, when really that set of facts, the possible sets of facts, are simply endless?

MR. EPSTEIN: Yes. Brian, do you want to--

MR. ANDERSON: Yes, let me. I defend a lot of product defect cases, and the problem with allowing no injury cases, where the plaintiff really doesn't matter, and the lawyer is defining the case is that it creates a lot of opportunities for 20/20 hindsight. The lawyer posits a plaintiff who bought a laptop, and the ability of that laptop to do this complicated mathematical calculation suddenly was the raison d'etre for the whole transaction, at least that's how the complaint is framed. And this purchaser, and everybody who bought the laptop, is supposedly very disappointed to now learn that this laptop will not do this calculation.

And had they known, had the manufacturer put a sticker on the laptop saying, "Warning. This laptop cannot perform the following calculation with 100-percent accuracy," well, then, scads of people would not have bought the laptop or they would have paid less for it.

The problem is that a laptop or any other product has thousands and thousands of component parts, thousands and thousands of features. Every product has advantages and disadvantages. It does some things well. It does other things not well. Some parts will break earlier than the industry average. Some parts will now function longer or last longer than the industry average, and so it is artificial to take one feature of a product in isolation and build a whole lawsuit around it and pretend that that is what was driving the transaction and to ignore all of the countervailing considerations that different buyers would have had when they bought their product.

You talk about fraudulent concealment, you know, if I had known, you could fill boxes, and boxes and boxes with warnings and disclosures about what may go wrong with this laptop, and under the full disclosure world that you often get in these kinds of fraudulent concealment complaints, you would have a scenario in which, you know, you buy a laptop, and you buy a wheelbarrow full of scientific reports about everything that could go wrong with the laptop. And come on, let's get real. When people buy a laptop, they don't really care about these remote risks.

PARTICIPANT: I have a question that has a practical and a theoretical aspect. The theoretical aspect is this--and this was prompted by Richard's remark about the king's court in 15-God-knows-what.

My understanding of the common law is that, at least for fraud, tort, there were always all sorts of pleading requirements and substantive requirements. You had to show reliance, and you had to show that you relied on the representation to your own detriment.

Now, what are the efficiency characteristics of those requirements? Do we still believe those traditional requirements to be so--hang on. If the answer everyone says is, sure, then it seems to me that the difference between the diamond case and all of these other cases is not so much that, oh, they want to have it both ways, but that in the diamond case you actually have somebody who relied to his own detriment, in some manner of speaking, right? And in these other cases, they obviously didn't.

The practical question is this: To what extent are these kind of cases--and this is for Brian--to what extent do you see these cases under sort of really traditional court-made common law and to what extent are they predicated on state consumer fraud or consumer misrepresentation theories? That is to say--

MR. ANDERSON: Usually, they are based on both. A typical product defect class action complaint will contain a common law fraudulent misrepresentation or a fraudulent concealment claim and a statutory Consumer Protection Act claim. Why? Because the complaint will contain a pro forma allegation that the named plaintiff and everybody who bought this laptop relied on the assumption that it would perform perfectly. You know, do people really think that things are perfect any more?

But if we can't prove up that allegation with facts downstream, and of course one pernicious aspect of the class action device is that, if a class is certified, the individual circumstances driving all of these different purchase transactions, who really did rely on this feature and who did not, doesn't come out.

But the alternative theory that is usually pled is one of these Consumer Protect Act theories, such as the California Unfair Competition law, which as a matter of statutory construction, does not require that the plaintiff plead or prove injury or plead or prove individual reliance.

MR. EPSTEIN: Other ways to think about all of this is as follows: One of the reasons why you like reliance and why it works for individual cases is if, in fact, you can prove that the outcome would have been different--purchase or not purchase--with the reliance, then you can show that this is, in fact, the marginal consumer. So, in an individual case, the only person who would come forward in Brian's hypothetical would be the computer geek who's always out there at the extreme, and he will say these are the programs that I use, and this is the one that failed, and all of a sudden you [audio break]. --you might consequential damages, though I think most warranties would start to exclude that, but you do this.

Now, once you get to the class action, essentially what's wrong is not the reliance theory. It's the assumption that everybody is the marginal consumer because you've now announced that the first person in from the gate, your geek, is, in fact, the marginal consumer.

Now, the right way to look at it is two things: It may well be that without this feature, the generic value of the computer is not worth a thousand dollars. It's worth $999.98 to everybody else. So you get the 2-cent plane, and then for the one guy you get the much larger loss. And that means, in effect, you give the benefit of the bargain for those people who are indifferent to the situation--that's 2 cents--and the real loss to the very high-priced user who is going there in the opposite way.

What happens with the class action, when you erode the reliance requirement, in a word, is that everybody becomes the most sophisticated consumer, and so you don't have differential levels of reliance based on this, and therefore you instead of giving 2 cents to everybody, you give a thousand dollars to everybody, and it turns out that it's a mistake.

I don't think there's anything--Michael, you've been reticent thus far--is there anything about what I've said that you would disagree with on the frame of your analysis?

MR. KELLY: Well, the basic answer is, no, though it strikes me that some of the complaints about this are driven by mistakes, right? It's clearly a mistake to say that the benefit of the bargain measure is the entire price of the product or the entire value that the product would have had. It's got to be the price of the cigarettes minus the value received or the benefit you expected to receive from the computer minus what the value that it had as you received it. And so they gave the entire money back, a mistake. They gave the entire thousand dollars back. That was a mistake in judgment. Those are just bad application of the theory.

MR. EPSTEIN: That's right.

MR. KELLY: And so to let the mistakes drive the law, I mean, granted we ought to educate judges a little better perhaps--more conferences like this--but, no, I don't have any objection.

MR. EPSTEIN: It is extremely important, in all of these cases, that you figure out what the status quo ante was, the before and the after, and figure out what the accurate valuation of these things is going to be. And then in some cases I think the appropriate response is not to give 2 cents back to everybody with respect to one of these consumer situations.

There might be some sort of a small regulatory forum which would be the substitute, but then there's another half, and I want to stress what Brian said before to make his point in a slightly different fashion. He says, look, you're getting a computer, and it's got 100 attributes, and they promise something on every one of them. His answer is, in many of these cases, you cannot be as precise as you hoped, and at least in some dimensions you give more than you've promised with respect to the computer in question--you do. I mean you say, in effect, that this thing has got a rating of 34 Hz, and it turns out it's really 35 megahertz. It goes a little bit faster.

What Brian's point or what my point to Brian would be that's a benefit which you get every time you turn on this machine, whether you're a professional or an amateur and so forth. And the query is do you settle off against a 2-cent loss with the exotic use, the benefit that they receive from the, as it were, overpowering of the machine in the universal case? And I think, in effect, the right answer is that you would always allow extra [audio break]--it could be zero--and that the correct remedy is that only the guy who relies on the feature, then, requires a diminution in value.

Most of these cases simply do not take into account overperformance as being an element in the case. Again, Michael, I'm going to push you again to take that situation where you've got two attributes; one which increases the value to every consumer by 1 percent over what has been promised and the other which reduces the value to every consumer by 1 one-hundredth of 1 percent relative to power, do you consider the negative and ignore the positive on the articulation that you gave and the [audio break]? We'll go right down the row and ask the hypothetical.

MR. KELLY: I think you consider the value the machine as delivered, and if the value, as delivered, is greater than the consumer anticipated, then that's what they got.

MR. EPSTEIN: [Off microphone.] They don't [inaudible], but there's no liability.

MR. KELLY: Yes.

MR. EPSTEIN: Let's go down [audio break].

PARTICIPANT: [Off microphone.] [Inaudible.]

MR. EPSTEIN: No, no. That's not the issue. The question is whether it's an offset against the perceived shortfall in the second dimension.

PARTICIPANT: [Off microphone.] No, a lot of [inaudible] are always about what is wrong with the product--

MR. EPSTEIN: But never--

PARTICIPANT: [Off microphone.] What is right about the product [inaudible].

PARTICIPANT: [Off microphone.] [Inaudible.]

MR. EPSTEIN: After today?

[Laughter.]

PARTICIPANT: I have a question for Richard, though. I'm very much in agreement with you, and these were my thoughts. But then how do you separate out the case where deviations are really stochastic. Sometimes you're going to get a better outcome, sometimes you're going to get a worse outcome, and the manufacturer has done his level best to hit the expectation versus those situations where in fact there's some kind of moral hazard at work that's causing--and how does the rule feed in to prevent or not?

MR. EPSTEIN: That's the right question to ask. I mean, in some cases, the way in which you handle it is simply to state at the outset what you're doing is you're picking a bag out of a barrel. And so, for example, when you purchase diamonds in these little packets, some of which are better or worse than the others, the rule is you take what you get unless you can show, A, it's a horrible batch or that it's been raped. And so you handle that by contract in advance. That tends to work in a commercial context between traders, not in ordinary consumer cases, so you get that difference.

In the consumer case, I think the way in which you handle it is, if you are a producer, is to do as follows: Is to promise 34 and to build to 35, so that the stochastic stuff means that you've got 99 percent over the line. And without even knowing industry practice, I think that's what's done routinely anyhow; isn't it, Brian?

MR. ANDERSON: Yes.

MR. EPSTEIN: I mean, you don't quite--there's not enough gain in it for you so as to put the median as you're expected. What you do is, if you could get good manufacturing, it means the variance goes down. You give up a half a percent on the stated and you knock out--basically all of the cases become uncertain size of surplus cases, as opposed to surplus or loss cases.

MR. ANDERSON: But here's the risk.

MR. EPSTEIN: There's always a risk.

MR. ANDERSON: A manufacturer designs and sells a car and says, "I'm going to sell you the car, and as part of your purchase price, you will also get the rights to free repairs of any defects that occur during the first 36,000 miles or 3 years of the car's life." In fact, the manufacturer has designed that vehicle with the expectation that it will go 100,000 miles. That is the designer's goal. They want to triple the warrantee period in terms of the way they design that car.

The lawsuit comes along. The named plaintiff's car malfunctioned at 75,000 miles. The manufacturer says, "What are you complaining about? You got far more good service out of this product than you paid for, and you want us now, despite the warrantee that you negotiated, and you could have bought an extended warrantee for more money, you want us to pay for that repair."

They say, "Well, here is an internal design document that says that you intended this component part to last 150,000 miles. It didn't do it. It's defective. You owe me money."

PARTICIPANT: [Off microphone.] [Inaudible.]

MR. EPSTEIN: [Off microphone.] I think the answer to that case has to be just an emphatic [audio break].

The next question--thank you--I want to ask you, Brian, is do you win or lose those cases when you defend them?

MR. ANDERSON: Judges more often than not will not dismiss the case simply if I waive the warrantee in the air and say too bad.

MR. EPSTEIN: So it's basically the summary judgment problem.

MR. ANDERSON: Yes.

MR. EPSTEIN: For the nonlawyers in the room, there are two universes. There's winning a lawsuit, which every defendant regards as a defeat if they have to go to trial, and then there's getting a summary judgment which means if the plaintiff knows the defendant will get it, the plaintiff will not bring the suit in the first place, and so that becomes, in terms of operations, the standard by which you design products and goods. Can you bring yourself a situation in which you can make the summary judgment motion? [Audio break]

MR. YAHYA: I was going to say, if the cases are actually litigated, the cases that I've read, and the ones that go up to the State Supreme Courts, they all dismiss them, especially if you're doing a negligent misrepresentation-type thing, they all get kicked out on the economic loss doctrine. And even though states will say, well, there is an exception for economic losses for negligent misrepresentation, the facts of those cases are egregious. They look almost like intentional misrepresentations.

MR. EPSTEIN: As a general proposition, the law of product liability has made a distinction which is not entirely happy, but seems to be quite firm, which says that if there's disappointed expectations in how a product performs relative to what you would hope, the contract determines the scope for the nature and the timing of the remedy. If it causes damage to some other product or personal injury, then it's a tort system.

I have actually a paper, if any of you are curious, in which I say that this is all quite incoherent and that a freedom of contract should apply to all losses between merchants, regardless of types or sort.

But what Moin said is in fact correct with respect to these issues, and I think in fact you will generally win these cases. So that the rule is, I think, don't go up to 100,000 maybe, but kind of if you have a--let's put it this way. The reason they go up to 100,000 is that they use a variance with respect to an automobile in terms of abuse per mile is infinitely greater than it is computer per click, right?

MR. ANDERSON: Run that by me again.

MR. EPSTEIN: In other words, the variance, the computer, you know, if you keep it for 90 days, some people do this, some people do that application, but it's not as though a guy who makes an intensive use, like driving on a rough road, where it's going to break the chassis.

MR. ANDERSON: Right.

MR. EPSTEIN: And when they build these models, they allow for normal and proper use, but that does not exclude any of the variances that you get from back roads, and front roads, and things of that sort.

MR. ANDERSON: And here's a lurking reason I think we have so many of these lawsuits that, to my mind, seem contrived. And you opened it at the beginning of the session by saying products are always getting better.

Somebody once told me that they bought a 1938 Buick vintage car, and they went out on the Internet and they got the owner's manual for this 1938 car, and the maintenance schedule for this car said bring it in every 500 miles for an oil change, and to change the tires, and to do this and do that and the other thing. That was the expectation 50, 60 years ago.

PARTICIPANT: [Off microphone.] An engine with 40,000 miles [inaudible].

MR. ANDERSON: Right. Today, if a car doesn't work maintenance free for 100,000 miles, it's defective, regardless of what the warrantee says. That is the era in which we live, and the expectations are so high.

PARTICIPANT: [Off microphone.] You do not include [inaudible]--

MR. ANDERSON: Parts, well, even routine maintenance. I mean, I have cases that--

PARTICIPANT: [Off microphone.] [Inaudible] oil changes?

MR. ANDERSON: No, not oil changes, but things that, you know, not long--

PARTICIPANT: [Off microphone.] [Inaudible.]

MR. ANDERSON: Yes, spark plugs. Not long ago, you know, brake pads wearing out, things that 20 years ago people just, well, every 5,000 miles I'm going to have to do this, we don't expect to do that any more. And we have a much lower tolerance for failure in our society any more, for things breaking. And if something does break, if something doesn't work perfectly, if our heightened expectations of perfection are not met, well, it's not because my expectations are too high, it is because the manufacturer must have done something wrong.

MR. EPSTEIN: But look, I mean, in fairness to somebody--I don't know whom, but I think Moin--is that what happens is you guys capitalize on this because when you sell the product, you say the oil changes need to be made only 20- and 40,000 miles, the brake changes only at 50-, and by the way, this business which was common when I was a kid of getting new engines in old cars is completely gone now because engines outlive bodies, and the reason of course is that the engine is relatively insulated from external affairs, and the body is not, and so therefore you can wreck a body by crashing, and it's harder to wreck an engine by crashing, I think, and I think that that's already internalized in the market.

Now, the question I would ask you, if you looked at warrantee--just as a straight business in the car area, Brian--and you ask yourself, when they give you these warranties, now they have 36,000 miles, you said, 36 months, what was the comparable warrantee you got in Hennington v. Bloomfield Motors in 1960?

MR. ANDERSON: Ninety days.

MR. EPSTEIN: What?

MR. ANDERSON: Ninety days.

MR. EPSTEIN: Oh, so it really is.

MR. ANDERSON: Yeah, it was 90 days.

MR. EPSTEIN: And this, of course, look, I mean, I think there's a great study here for somebody, and I'll tell you why. The way in which this market has been organized over the years, everybody always says that there's market failure. People simply do not respond. Competition does not work.

If, in fact, that were correct, you would expect today to see 90-day warranties on cars that have 36-month warranties, in fact. So what's happened is basically the warranties that we have seen have gone up over tenfold in the period of that year. By the standard theories of the product liability judgment, this could not possibly have happened. Why is it that these things start to--and the answer is that there's powerful competition in the warrantee market.

And so, you know, for people who are trying to figure out do markets work, I think it would be an extremely--Judy is writing this down--I think it would be an extremely useful study to actually go compare the warranties to see exactly what they are, and then on the other side to do the thing which is how do you book them on the company side; that is, how much do you actually put into a reserve to handle the warrantee? And my guess is that, if you took the percentage of cost of an automobile for warrantee service in 1960, it would be vastly greater than what it is today.

So the general argument that most of us familiar with it make that markets actually work and improve things, this is a very nice way to sort of run that particular study. And the point is it works. You get these longer warranties notwithstanding that there's no real liability for consequential damages. I mean, companies still basically think this is an important thing to compete about and do so in an effective fashion.

More questions? We've got to get some more questions, otherwise I'm going to have to run on.

Yes?

PARTICIPANT: I'm not a lawyer, so excuse my ignorance.

MR. EPSTEIN: That's all right.

PARTICIPANT: So I guess this is mainly a question to Mr. Anderson.

Talking to defense lawyers, I understand that, from the defense's perspective, one of the problems with the class action suits is the aggregate costs, the fact that even small amounts per plaintiff aggregate up to very large awards, and the risk of this for a business, even a [?] business like GM, is simply too large to not sort of settle out of court and minimize the risk.

So my question is whether or not you think that this is, in fact, an accurate description, and if it is, what are, in your experience, what are the actual implications here? Are you settling on really low-probability suits just because the potential awards are really large? And can you give some sort of quantitative sense of what's going on here?

That's a hard one, but--

MR. ANDERSON: Sure. Well, a couple of reactions. First, there's two things about class actions that scare defendants. One is what you might call the inexorable math. One million people buy a product, and they're all seeking $1,000, and there is a 10-percent chance of their prevailing. You multiply that all out, and I went to law school, not business school, so I don't know what the number is--

PARTICIPANT: It's $100 million.

MR. ANDERSON: It's a big number, and there aren't many companies that are willing to roll the dice--

PARTICIPANT: [Off microphone.] [Inaudible.]

MR. ANDERSON: A hundred million, big number. And particularly in some of the courts in which these cases are brought, a defendant does not want to take the chance.

The second thing that scares defendants about class action is the due process invasions that they often bring about. In a single-plaintiff lawsuit, one plaintiff, defendant, the plaintiff has to put the facts on the table. And if they haven't relied on whatever it is that they're complaining about, their lawsuit is going to fail, and the defendant has every opportunity in the world to put the plaintiff to its proof and really have robust debates about whether this product is good or not, were you injured or not, do you care or not, were you lied to or not.

But in a class action, plaintiffs take--the perfect plaintiff, you know, the proverbial geek that Professor Epstein was talking about--and if the case is certified for class treatment, if the geek wins, everybody wins, even people who aren't geeks. And so that scares a defendant.

What does that cause them to do? Well, that causes a lot of defendants to class actions to try to settle the cases, which leads to another problem.

PARTICIPANT: More cases.

MR. ANDERSON: Well, more cases--

[Laughter.]

MR. ANDERSON: --and the fact that the defendant, knowing that the vast majority of the claims are frivolous, is not interested in paying anything like a dollar value for those claims, and so they come up with coupons or other kinds of nonmonetary benefits to the class, and of course those kinds of class actions are coming under more and more scrutiny. Because what happens is the plaintiffs' lawyers get all of the money, and it looks like a really fishy deal.

So the other thing you do is you try to litigate the case, but you're taking a big chance if you do.

MR. EPSTEIN: Let me mention, consistent with this, what I see as the problem. Go back to the reliance question. One of the things that happens is, if you have individual reliance in a case, it is very difficult to get the class action because you have to run topologies of each individual plaintiff. And so what the judges will do is they will look at the substantive law which has had this reliance component in it for eons, and all of a sudden they will eliminate that requirement, saying, in effect, that we could presume reliance if there's any differentials.

And so you get a constant effort to change the substantive law in order to satisfy the commonality requirements, and this, in effect, really means that you're not just taking a thousand cases and putting them together, you're taking a thousand cases with an expected value of 10, and when you put them together, each of their values goes to 25 or to 30, so you get the double effect. You get the multiplier effect from aggregation which may be legitimate, and that from redefinition of the substantive law, which strikes me as being much more controversial.

And the test that I always put to people is I say, if you were to prepare to eliminate the reliance requirement in an individual case, then you may be able to do it in a class action, but in an individual case, your sense of justice is going to be, if it didn't make any difference to you, that's going to change the nature of the world. Maybe he'll get something, but it will be a lot less than he would if, in fact, it did make a difference to him. That's what differences mean. They make differences. And if they make a difference in practice to the person, then they ought to make a difference in the law. If they don't in practice, they ought not into the law.

So that the class action dynamic is extremely complicated on these things.

And the last point I want to mention on this is, again--which Brian hinted at and that we all understand--is that one form of ill behavior gives rise to another; that is, they bring worthless suits, we get rid of them by giving them worthless slips. And then it turns out there's actually a real case there with a real valid injury, and you know habits die hard, and you still give slips. And the way in which you compensate for this is the attorneys' fee is now double what it was before, and the slip is no more valuable.

So that you get the real sense that defendants then, in effect, become co-conspirators with plaintiffs' counsel in the reverse auction, and that puts everybody in an enormously difficult kind of position, and it creates a real legislative backlash.

I can assure you the case that comes before the committee is the one in which the practice is most dubious, not the one in which it's most vulnerable. That's the way political selection works.

We have time for one more question.

PARTICIPANT: Just a brief practical question, which is the same question which was just raised.

It's not just reliance that helps you in traditional cases distinguish the individual members of the class, right? It's also they're all injured in different ways. Except how do you defeat a class in these kinds of cases, when they all say what we have in common is we weren't injured? If you're in "Frank Easterbrook Land," you can sort of toy around with the sort of no multi-state class actions. That gives you some implicit class action protection, at least with respect to multi, I mean, national class actions. But what's the argument here? I mean, putting that aside, what's the argument in these no injury cases?

MR. ANDERSON: Right. And that is what makes defending these kinds of cases more difficult. Because the more the plaintiffs' lawyer can water down the elements of the claim and get around [audio break] particular bug because who, in their right mind, would buy a defective laptop in the first place? That's common sense. We don't need to debate that individually.

And if you can come in with a theoretical damage claim; that is, a tautology, you know, fear of future injury, that's got to be worth something. I don't know what the dollar value is right now. It may not have been realized, but I'll bring an economics professor to come in at trial and testify about what that, you know, give us a formula. Then, it becomes very difficult to defend against these class actions because suddenly there's no meat to the lawsuit. It's all abstract. It's all theory, and it seems like something that can be wrapped into one big trial and debated among--

PARTICIPANT: I will say it seems to point us in a different direction, right? The question isn't really lawsuits without injury because there is injury, economic injury here, and in fraud, economic injury is compensable. It's lawsuit without reliance.

MR. EPSTEIN: There's also I think another element in this, which I think people who work in the field know, which is the segmentation question. The truth is, if you actually had some guy who bought this computer and his entire business failed because the decimal point didn't move, he's not going to be a class plaintiff. He's going to bring an individual lawsuit and prove it out.

So that the way in which these things get organized is all of the real cases get brought as individual cases, which means that the guy who is the geek is not the geek with the losses, it's the geek with the machine who has not suffered any particular loss. So that you therefore, systematically, are compensating ex-post losses when they occur in serious cases and ex-ante compensation and everything else, where, in fact, the ex-ante compensation is designed to deal with the guy who actually did bring the lawsuit in those particular cases where it happened.

And it turns not only, by the way, is there a difference in the way these things are brought, but actually there are two different separate parts of the plaintiffs' bar that bring these cases. One of the issues in the asbestos litigation, which is quite real, is the mass tort cases. They're called the plaque classes are brought by one type of class action lawyers, and the serious cancer cases are brought by another. And if you want to get a settlement which is these guys get nothing and those guys get 100 cents on the dollar, you can't keep the plaintiffs' bar with you. So you have to sort of divide the money up between the two pots, which then breaks the deal and puts you right back to where you currently are.

So the question of what becomes an injury is extremely important because really serious injuries never become the template for class actions.

Is there one more question that you had or shall we say that summer's lease hath run all too short a date. But this is a body which I think we shall commend for brevity, and wit, and the soul of discretion is to end on time. It is 4:29. If somebody wants to give a 30-second round of applause, we will be able to run it up to 4:30. Otherwise we could just say thank you and good luck.

[Applause.]

[Whereupon, the proceedings were adjourned.]