Reconciling experimental incoherence with real-world coherence in punitive damages.

• Author: Eisenberg, Theodore

Few complaints about a legal system resonate louder than charges of incoherence. (1) A system that fails to treat similarly situated parties equally cannot be squared with fundamental notions of fairness and justice. (2) Incoherence unintentionally converts law into a lottery. Even charges of bias are less pernicious in some ways, because the targets of bias can organize and voice their opposition. Incoherence has no constituency and yet also no organized opposition.

Recent psychological research findings of incoherence pose a dilemma for law. Cognitive psychologists suggest that human judgment is, by nature, incoherent. (3) Judgments made in an individual case often do not comport with judgments made in the aggregate. In particular, Predictably Incoherent Judgments ("PIJ") provides convincing experimental evidence that people lack the basic cognitive skills necessary to translate qualitative moral judgments into quantitative numeric scales. (4) To the extent that a legal system relies on individuals to accomplish such translations, it risks incorporating incoherence into its judgments.

Concern that the legal system requires people to make judgments that are apt to be unreliable is common enough, but the ubiquitousness of the type of judgments that PIJ identifies raises the stakes by orders of magnitude. Converting moral judgments into quantitative scales, such as dollars or years in prison, is a ubiquitous task for law. No legal system, whether it operates through the courts, administrative agencies, or acts of Congress, can function without being able to make such conversions effectively. With a few exceptions, (5) the law is expressed through dollars or years of imprisonment. If damage awards and prison sentences are "predictably incoherent," then our legal system (indeed, every legal system) might be so pervasively unfair as to be indefensible.

The psychological theories that PIJ relies on are consistent with decades of research in cognitive psychology in general, and are supported in particular by dozens of studies on the issue of how people translate qualitative scales into quantitative ones. (6) Furthermore, its experimental translation of this research into the legal domain is also solid. The conclusion that people incoherently translate the moral into the numeric within the context of legal decisions is therefore hard to reject.

Therein lies the paradox. The legal system produces hundreds of quantitative judgments each day. If we take PIJ's results seriously, and accept its characterization of the situation as scandalous, (7) should we not be observing a legal system on the verge of collapse? Any civil society would find a torrent of incoherent legal judgments to be unfair, unjust, unpredictable, and unmanageable. People would not be able to plan their lives, business would not be able to function, and government itself might be at risk. In the United States, however, a country that produces as many legal judgments per capita as most, (8) the legal system functions reasonably, if not perfectly. Most people generally understand the legal consequences of such behaviors as driving too fast, buying a house, or purchasing illegal narcotics. Intuitively, the system seems to work well, despite the well-founded concerns in PIJ.

More important than intuition, however, is that the research on quantitative judgments demonstrates substantial consistency in judgments. The results of lawsuits appear to be quite rational. In civil lawsuits, the worst-behaved defendants who caused the most harm are the most likely to lose and pay the most damages. (9) A recent review of the literature concludes that "numerous field studies support the notion that jury damage awards are at least moderately related to the seriousness of the plaintiff's injury," and that other "field studies have found that punitive damage award amounts are strongly related to the size of the compensatory damages awarded." (10) Most notably, given PIJ's focus, awards of punitive damages, an area that should be riddled with incoherence, produce remarkably consistent judgments. (11)

Although psychological research as applied to law is sometimes charged with being unrealistic and inapplicable to the real world, (12) we do not suggest that the answer to this paradox lies largely in methodological difficulties with the application of cognitive psychology beyond the laboratory. We believe that PIJ's experimental results should be taken seriously. Rather, the answer lies in the legal system's ability to reduce and manage the limitations of human judgment. If human judgment produces incoherence, it is likely that legal systems (like other professional systems such as medicine and engineering) (13) identify ways to reduce the incoherence of such judgments or the effects of such incoherence. In other contexts, the law has managed to adapt to the cognitive biases of decisionmakers. (14) Nevertheless, such adaptations might be improved upon in some ways. The question for reform-minded psychologists and the consumers of psychology, then, is when and where costly incoherence persists in a form in which it can be effectively remedied.

Identifying such circumstances requires more than theorizing and more than conducting further laboratory-oriented experiments; it requires a serious look at the real world of cases. Determining when and in what ways the cognitive sources of incoherent individual judgment produce incoherent legal judgments requires developing the implications of the psychological research for observed cases, and testing whether these implications hold true.

This Essay begins that testing process. Part I reviews the available data on punitive damages because, according to PIJ, such awards should be a prime area of incoherence. Despite this prediction, the available data from many sources, rather than demonstrating incoherence, reveal a fairly sensible pattern of results. Part II presents new research, using existing data, that nevertheless supports the influence of the cognitive sources of incoherence that PIJ identifies. So a largely coherent real-world system coexists with some empirical support for PIJ's experimentally detected incoherence. This suggests that real-world incoherence is not the scandalous plague PIJ fears but that the system should still strive to reduce its more modest degree of actual incoherence. Part III tries to sort through the sources of coherence and incoherence to reconcile PIJ's experimental findings with its modest degree of support in real-world data. Part IV explores the implications of PIJ's findings about incoherence.

1. INCOHERENCE NOT OBSERVED: THE PUZZLING PREDICTABILITY OF PUNITIVE AWARDS AND OTHER LEGAL DECISIONMAKING

   If incoherence distorts the determination of punitive damages awards, then the real world ought to contain substantial evidence of the need for reform. The primary evidence should be either a crazy pattern of awards or many anecdotal awards that are plainly unjust. PIJ seems to simply assume that such evidence exists. It refers to the "notorious unpredictability of punitive damages awards," a proposition for which it cites only the authors' own experimental evidence. (15) Yet, despite many well-funded articles favoring punitive damages reform, (16) researchers have not identified either a crazy pattern of awards or a substantial series of actual punitive damage awards that constitute a shocking pattern of incoherence or unfairness.

   The limited focus of punitive damages reform advocates on experiments and theory is telling, because those who do look at actual cases are impressed by the scarcity of punitive awards and their surprisingly coherent pattern. Juries rarely award punitive damages (17) and generally confine punitive damage awards to cases involving intentional misconduct; punitive damage awards in products liability and medical malpractice cases are exceptionally rare. (18) Judges behave similarly. (19) When either juries or judges do award punitive damages, they do so in similar ways that relate strongly to compensatory awards. (20)

   Those who study actual cases searching for large numbers of unjust anecdotal awards come away more impressed by the justifiability of the awards. (21) Even alleged hotbeds of punitive damages abuse, when subjected to social science scrutiny, turn out to be more sober than shocking. For example, Alabama juries, occasional poster-children for punitive-damages reform, award punitive damages in financial injury cases at rates similar to those of other states, and their punitive damage awards are generally smaller than those in other states. (22)

   Yet the deeply ingrained belief in the craziness of punitive awards is so profound that each time a study reveals the overall sober pattern of awards the news media treat it as a major new discovery. (23) In many respects, the public's apparent failure to appreciate the coherence of punitive damage awards resembles a classic example of the availability heuristic. (24) The availability heuristic refers to the tendency for people to base their understanding of the likelihood of uncertain events on the ease with which they can call examples of the events to mind. (25) As a consequence of this heuristic, people overestimate the occurrence of events that are reported in the media or that are particularly vivid and memorable. (26) Furthermore, "availability entrepreneurs" try to draw attention to such events so as to move public opinion in particular directions. (27) In the case of punitive damages, the media, perhaps with a nudge by availability entrepreneurs, reports "outrageous" damage awards, such as the infamous but largely explicable $2.9 million damages award against McDonald's for coffee that was too hot. (28) Left unreported are the mass of sensible awards (not to mention the inevitable appeals and settlements that bring high awards into line with those of other cases). This bias might even affect the managers...