A "Rough Justice" Argument for Increased Civil Damages

-- Posted by Neil H. Buchanan

In my Dorf on Law post on Tuesday, I used the example of the "punitive multiple" to illustrate a problem that can arise in a system that uses individualized legal proceedings to address (at least in part) broader social problems.  The basic logical conflict illustrated by the Michael Brown case in Ferguson, Missouri is that most people with even a semblance of awareness know that "cop on black" violence is a systemic problem, whereas we also know that there will be situations in which such deadly force was, as the police say, "a good shoot."  Even when a case is correctly not pursued, the individualized justice fails to address the larger social injustice that persists.

The punitive multiple (which awards larger damages to individual plaintiffs, because of the "larger guilt" of the defendants) becomes relevant here because it shows just how odd it is to try to use a judicial proceeding to deal with a broader problem.  It also, however, shows how we could rely on the predictable frequency with which a systemic problem occurs as a way to try to improve the way courts handle such cases.

Specifically, we know that a large number of meritorious legal claims are never going to be pursued because, despite the mythology to the contrary, Americans generally do not like to sue people.  One study of which I am aware showed that something like 90% of actionable medical malpractice claims in New York were never pursued.  (Also, only a fraction of those that are pursued resulted in a settlement or a win in court, even though we are limiting ourselves only to cases of genuine malpractice.)  The numbers are even worse for other types of claims, for example product liability injuries, because the damages are frequently too small for individual claimants to pursue, compared to legal costs.  And companies know it.  There are certainly some "nuisance claims" and other types of suits that are part of the landscape, but the numbers on those simply do not add up to enough to change the overall picture.

As I pointed out on Tuesday, the punitive multiple concept tries to achieve corrective justice by estimating the percentage of the time that repeat offenders escape justice, and then multiplying the damages in the cases that the defendants do lose such that they end up paying for the full amount of damage that they inflict.  Thus, in my example in that post, a company that inflicts damages in increments of $20,000 on one thousand victims "should" pay $20 million.  If we limit successful plaintiffs to $20,000, however, every case that the company wrongly wins (or that is never pursued in the first place) is a free pass, and the company has an incentive to inflict excessive damages on society, safe in the knowledge that it will never have to compensate most of its victims.  (Many readers will have noticed that I have avoided using the word "efficiency" here, even though that is the word that many scholars would use to describe our goal here.  It is beyond the scope of this post to explain why economic efficiency is an incoherent concept, but interested readers might find this post from two years ago helpful, in which I liken the concept of efficiency to a "phantasm.")

One topic that arose in the comments on Tuesday's post was the comparison between using the civil justice system to change tortfeasors' behavior and using plain-vanilla regulation via administrative agencies to achieve the same result.  Actually, the civil justice approach comes in two versions, either via state actors (usually attorneys general) suing on behalf of the people, or individual victims suing on their own behalf.  We thus have a three-pronged approach to creating appropriate disincentives to companies that engage in potentially injurious behavior: (1) Have the state regulate it directly, (2) Have the state sue for damages, or (3) Have private actors validate their own claims in court.

In theory, of course, we should only need one of the three approaches.  Depending on how many simplifying assumptions one is willing to make, any of the three should do the trick.  In the real world, however, each has its own shortcomings.  A few years ago, a friend and I had a heated argument about whether private actions are a valid or desirable way to pursue the goal of improved public health, specifically in the context of childhood obesity.  At the time, controversy was raging over a private lawsuit that had been brought on behalf of some obese teenagers in New York City.  Because the plaintiffs were underage, the usual "assumed risk" defense was off the table.  Even so, my friend insisted that the best way to deal with childhood obesity was through direct regulation, not private lawsuits.

My friend was sincere, and he was definitely not playing a game of three-card monte, in which sophists argue against any of the three possibilities by saying that the other two are available -- but then attack each the other two by saying that the others are available.  That is a common move by apologists for business, but one can believe in good faith that the best way to create the rules that will lead to a desired social outcome is through regulation rather than other means.  The problem is that doing so still presumes that direct regulation will work, which further presumes that it will be allowed to work.  It is not, after all, as if someone woke up one morning and said, "Gee, there's a childhood obesity epidemic.  I think the first-best way to attack that problem is with private lawsuits."  After years of inaction by regulators (hampered, of course, by hostile legislatures that de-fund or de-authorize active regulators), and after repeated refusals by state AG's to act, the private lawsuit route starts to make sense.

To be clear, state action (via regulation and/or lawsuit) would alleviate a concern that I discussed on Tuesday, which is that the victim who happens to win her lawsuit could be over-compensated.  Even if the damages in a private lawsuit appropriately create the desired (dis)incentives from the tortfeasors' side of the story, using a punitive multiple guarantees that some "lucky" victims will be paid as surrogates for unlucky victims, who will get nothing.  Regulation allows the state to make the wrongdoers pay, and to compensate all victims.  Public lawsuits are a little balkier, but they can achieve the same results.  But again, those outcomes are only possible if the political system allows them to happen, which is far from guaranteed.

And, to be very clear, this absolutely does mean that companies would view any of the three approaches as something that they can build into their financial projections.  I have no problem with making such damages "merely the cost of doing business," if those damages are large enough to align the companies' incentives with society's goals.  Back in the 1970's, Ford Motor Company was found to have decided not to fix exploding fuel tanks in the Pinto line because it would cost something like $11 per car to do so.  Although I am hardly a neoliberal, I certainly believe that the only sensible way to deal with such situations is to make the cost of doing business much higher, through a combination of the three approaches described above.

Which brings us back to punitive multiples.  As one of the commenters on Tuesday's post pointed out, the computation of a punitive multiple is actually quite difficult, such that doing it with any precision sets up odd incentive effects.  (In economics jargon, the punitive multiple has an endogeneity problem.)  Even so, the rough justice of multiple-damage statutes need not be based on a perfect computation of the correct multiple.  As I noted on Tuesday, antitrust damages are automatically trebled by statute.  I doubt that anyone thinks that Congress was convinced that exactly one-third of meritorious cases win every year, so that multiplying winning damages by three is perfectly efficient.

What I think people should believe, however, is that potential litigants need an extra incentive to start down the expensive and treacherous road that is private civil litigation.  Which means that the punitive multiple could be adopted more generally, to recognize that the other two methods of regulation are imperfect.  And, as an added bonus, regulators have the ability to take into account how well the private and public lawsuit prongs of the regulatory regime are working, and to adjust the direct regulatory rules appropriately.

In short, the purely theoretical appeal of punitive multiples fades rather quickly, once one begins to consider how difficult and inaccurate the computations of such multiples must inevitably be.  That, however, does not mean that we must give up on using private civil litigation as part of a regulatory regime.  Indeed, even if we were to legislate a low punitive multiple, this would still be an improvement over the current system, in which the multiple is implicitly set at 1.  Multiplying damages is a departure from individualized justice (for plaintiffs), but we already allow the system to depart from that ideal.  A further departure would be a very good idea.