How The Litigation Lottery Kills Shareholder Value

Any event that wiped out $684 billion in shareholder wealth would be described by economists as disastrous. Congress would immediately order hearings, dragging the offending parties before the TV cameras and rushing to offer legislation to rectify the problem. Yet $684 billion is the amount shareholders lose every year as a result of America’s out-of-control […]

Any event that wiped out $684 billion in shareholder wealth would be described by economists as disastrous. Congress would immediately order hearings, dragging the offending parties before the TV cameras and rushing to offer legislation to rectify the problem.

Yet $684 billion is the amount shareholders lose every year as a result of America’s out-of-control legal system – a figure released in a study by the Pacific Research Institute (PRI).

PRI’s Dr. Lawrence McQuillan examined prior economic “event studies” and determined that the median loss of market value due to a lawsuit was $3.86 million (in 2006 dollars). He then estimated that approximately 177,000 tort cases are filed against publicly traded companies in a given year – generating a yearly loss of $684 billion in shareholder wealth.

With the lawsuit industry booming, that might actually be a conservative estimate. Recent blockbuster cases – such as the litigation threat following Merck’s decision to pull Vioxx – wiped out $25 billion in shareholder value in a single day.

Of course, trial lawyers want us to think that only CEOs and Wall Street tycoons feel the heat from litigation-induced stock plunges. But the most recent figures from the Investment Company Institute tell us that nearly 55 million Americans own mutual funds – 60 percent of which earn less than $100,000 per year. Today, over 50 percent of Americans own stock, compared to just 17 percent in 1980.

The democratization of the equity markets over the past 25 years has extended stock ownership well onto Main Street – from families saving for their children’s education in popular 529 plans to middle class workers socking away retirement funds in their IRAs and 401Ks. Maybe Congress should order hearings after all – and make personal injury lawyers answer for abusing our legal system to pick the pockets of America’s investor class.

Steve Hantler


  • Thank you so much for showing us the forest for the trees. Money has to come from somewhere, in law suits the money comes from the investors or the users of the products.

    All to often I hear that the plaintiff’s want to send a message by astronomical settlement amounts. Society might better be served by sending a message through well though out litigation rather than society at large having to pay.

  • In the long term, on bogus suits, that “loss” usually recovers (that is, the value goes down… then comes right back up to where it had been again).

    The problem is the gaming of the system this allows… sell short, announce massive lawsuit, cover short, drop lawsuit. Lather, rinse, repeat.

  • 1) I think you are getting median and average mixed up. You can’t multiply the median loss of market value by the number of tort cases to get total cost. The median leaves out the blockbuster cases which, as you note, have a huge effect on the total (a hundred million dollars case and a $25 billion dollar case have the same effect on the median but very different effects on the total). Curiously, this mix-up is usually made by anti-reformers trying to play down the costs of the legal system.
    2) The study distinguished between excessive and regular costs of litigation. The figure we should be discussing here is the $198 billion claimed in excessive litigation and not the total $684 billion in costs.
    3) For the effect on typical Americans, you need to know what proportion of Americans own stock AND what proportion of the stock market these smaller investors make up. Surely some of the excessive costs are borne by Wall Street fat cats. The number you need for your claim is what proportion of the excessive costs are borne by smaller investors. You do not provide it.
    4) Walter, Ted, David, anyone–when “bizarro overlawyered” or any other trials lawyers group misquotes a study and can’t tell the difference between a median and an average you, with very good reason, pounce on them. Should Hantler get a free pass just because he’s saying something that most readers of this site agree with?

  • And from an already Overlawyer state comes an inability to say enough is enough!

    UC regents give go-ahead to law school
    From LA Times Staff and Wire Reports
    July 20, 2007

    The UC Board of Regents on Thursday affirmed its decision to build a law school at UC Irvine despite objections from the California Post-Secondary Education Commission that the state doesn’t need another public law school.

    The commission maintained that the state did not need more lawyers, that the cost was not justified and that the law school would duplicate other services.

    The regents, meeting at UC Santa Barbara, unanimously voted without debate to override the commission’s objections and proceed with construction of the law school at UC Irvine.

    It will be the first public law school to open in the state since 1965 and will be only the second in Southern California.

    UC now operates law schools at UCLA, UC Berkeley, UC Davis and the Hastings Law School in San Francisco.

  • What an interesting concept. Lawsuits produce a loss of market value and, by using nonsensical calculations, we are told they add up to a yearly loss of $684 billion in shareholder wealth.

    Have you calculated the stock market losses caused by reports of lower than expected corporate earnings, negative reactions to the testimony of Federal Reserve Chairmen, adverse FDA pronouncements regarding drug approvals, layoff announcements, revelations of corporate fraud, trade deficits, housing slumps, banking irregularities, drops in consumer confidence, or rising gas prices? I suspect that if we could just keep news of these events from getting to the investing public the Dow Jones Industrial Average would have passed 20,000 ten years ago.

    I know it is hard to accept, but we are trying to have an open society in the U.S. That includes our Stock Market. Welcome to America where we have a civil justice system, a free press and public stock exchanges.

  • Drew: I’ve met Steve Hantler several times and have seen him provide unique insights into the issues of liability reform, have read lengthy articles he has written where he shows grasp of subtleties of the issues and challenges, and know that he has war stories from his job that would qualify for “Best of” tales on this site. See, for example, his talk at AEI.

    So that he has instead used his guest-blogging here to rehash soundbites from the PRI study is certainly disappointing and a misreading of his audience.

    I think there’s potentially a bigger problem than the median vs. mean issue with the $684 billion figure. Since Larry McQuillan has been commenting here this week, I hope we’ll get his insights on these questions.

    1) PRI gets their $3.86M figure from Karpoff/Lott (1999), which looked at a sample of 351 cases requesting punitive damages. What about the Karpoff/Lott sample makes one believe that it is a representative sample that one can extrapolate the median case from their sample to the mean tort case filed in court?

    2) Karpoff/Lott attribute some of the loss in shareholder value to reputational costs. They are effectively arguing that the marketplace already provides disincentives to engage in conduct that might result in punitive damages, reducing the argument for the need for punitive damages in the legal system. Why include the entirety of this reputational cost as a cost of the tort system? The cost comes from the accusation of bad conduct. Now, it is certainly the case that some lawsuits imposing reputational costs do so baselessly (the sudden-acceleration cases re an obvious example), but others reflect actual corporate wrongdoing and the reputational costs from that rather than anything the tort system is doing. It’s not immediately clear to me how a quantitative study can suss out that difference, but the answer doesn’t appear to be that 100% is attributable to excesses of the tort system.

    3) Pedantically, the $25 billion loss of market value to Merck mentioned by Steve reflected not a single lawsuit, but the prospect of tens of thousands of lawsuits, as well as the lost expectation of a future revenue stream of billions of dollars from Vioxx sales, as well as the expectation of stricter regulation, as well as the reputational costs alluded to by Karpoff and Lott.

  • Nice comment Ted. Your candidness is refreshing. It does nothing but good things for your credibility that you address arguments on their merits and not by which side you’re on. Your point in #3 is quite right–I did not mean to imply that it was all one $25B suit.

  • Mr. Frank,

    It has been a while since I have seen one of your surveys of Vioxx litigation. They are very interesting to me at least, and I suspect to other visitors to

    I am disheartened by the recent settlement in Los Angeles. It is now established that the daycare prosecutions of a couple decades ago resulted from suggestion, obsession and general hysteria. The priest litigation suffers from the same influences. The money comes from Catholic parishioners, who had nothing to do with the alleged improbable crimes.

    And then the plaintiff lawyers took 40%.

  • Mr. Hantler accurately quotes my study’s findings on the loss of stockholder wealth due to tort lawsuits (see “Jackpot Justice” pages 36-38). The “median/mean” point raised by Mr. Drytellar can be clarified with a short elaboration.

    After reading the scholarly literature on lawsuits and stockholder loss, I concluded that the most reliable study on this topic is a 1999 study by Jonathan M. Karpoff and John R. Lott Jr. They found, after controlling for other factors, that the median loss in the market value of equity due to a lawsuit was $2.9 million. I purposely used the median loss, rather than the mean or average loss, because I wanted a conservative number to use as a forecast for 2006 (the $2.9 million was converted into inflation-adjusted 2006 dollars, resulting in the $3.86 million number cited by Mr. Hantler).

    I purposely selected the median because, as Mr. Drytellar correctly noted, medians are less sensitive to one-time, year-specific “outliers” – very high or very low effects. A $25-billion award in any given year, for example, might not be representative of what one should expect in future years, so I selected the median value to control for any extreme, disproportionate effects in the past data that might not be representative of the norm going forward. Incidentally, this is yet another example of how “Jackpot Justice” was constructed using the most conservative assumptions.

    The $3.86 million is best thought of as an independent forecast made in 1999, using the best available statistically significant model, of the typical or average loss in 2006. Mr. Drytellar is correct that you cannot multiply the median loss in a given year by the number of lawsuits in the same year to arrive at the total costs for that year. But that is not what I did.

    The median loss in 1999, adjusted for inflation, became the conservative forecast of the typical or average loss in 2006. This average loss was multiplied by 177,166 cases to arrive at the $684 billion annual figure for 2006 – roughly a 4-5 percent loss in market capitalization.

  • But that is not what I did.

    The median loss in 1999, adjusted for inflation, became the conservative forecast of the typical or average loss in 2006. This average loss was multiplied by 177,166 cases to arrive at the $684 billion annual figure for 2006 – roughly a 4-5 percent loss in market capitalization.

    That appears to be EXACTLY what you did… as stated in your own very next sentence.

    Help us out here.

  • Deoxy: I urge you to please re-read my comment very carefully. You are not grasping some important elements such as there are two different years being discussed, 1999 and 2006, and one year gives the median value and one year gives the mean value. The methodology is spelled out; I cannot be more precise. It’s all there.

  • I wrote at length in “Jackpot Justice” (see pages 36-38) about why the Karpoff/Lott study is preferred to other studies, but I will briefly summarize here. Karpoff/Lott, unlike other researchers, looked at a large sample and wide variety of lawsuits from beginning to end – from the initial announcement of the lawsuit to the final resolution. Most studies only look at the initial lawsuit announcement effect; thus, they provide an incomplete picture of the total stock-price effect.

    I also checked to see if looking at lawsuits where the plaintiffs sought punitive awards introduced an initial bias. As discussed in “Jackpot Justice,” the “initial announcement” effect in Karpoff/Lott is strikingly close to the “initial announcement” effect in another leading study that is not restricted to punitive-damage-seeking cases. The market doesn’t seem to initially price in an additional loss for cases in which the plaintiff initially seeks punitive damages, likely because, as others have noted, punitive damages are rare, random, and often reduced on appeal.

    Also, keep in mind that the Karpoff/Lott sample is cases where plaintiffs initially sought punitive awards – the plaintiffs did not necessarily receive punitive awards. There is no reason to believe that the proportion of cases in this sample where the plaintiffs received punitive awards, or the amount received, varies in any statistically significant amount from the national averages.

    Although no sample is perfect – the only perfect sample is the entire population – the Karpoff/Lott sample has many important advantages that make it preferable to other studies. Finally, the Karpoff/Lott stock-price effect is lower than other leading studies, so the estimate of stockholder loss in “Jackpot Justice,” which relies on this study, is conservative.

    To understand why reputational costs should be included in the total loss, one must understand the informational origins of stock-price changes. New public information moves stock prices. The Karpoff/Lott study examined, from beginning to end, the effect of new relevant information released during the course of a lawsuit on a company’s stock price. Perhaps this information would have become public without the lawsuit, and perhaps the market would have evaluated the information the same way, but this is speculation.

    Therefore, it is appropriate to include reputational stock-price effects in the total effect because the new public information driving the stock-price change emerged during the course of a tort lawsuit, e.g., information on product design, manufacturing processes, or failure rates that could affect perceptions by consumers of a product’s quality or reliability.

    Keep in mind that the information need not be truthful or accurate to move the stock price. The bottom line is whether the information changes the perceptions of market participants. No study that I have seen tries to separate out stock-price changes based on truthful facts or legitimate damages in tort lawsuits, on the one hand, and stock-price changes based on untruthful allegations or meritless/excessive damages, on the other hand. We did not attempt this either in “Jackpot Justice.”

    The lawsuit/stock-price scholarly literature proves that the total cost of tort lawsuits filed against publicly-traded corporations exceeds the sum of damages and attorney fees – something many of us have been saying for a long time. There are secondary, spillover costs associated with these lawsuits, and one such cost is reputational in nature. These reputational costs are real, derive from the lawsuit, and, therefore, should be included in the total stockholders wealth loss due to tort lawsuits.