“It Was My Church”

The story begins in 1989, in a small Alaskan fishing community named Cordova, in Prince William Sound. Linden O’Toole had just moved into town, and she was immediately struck by the natural beauty. “It was my church,” O’Toole tells me, before describing the stunning snow-capped mountains, rushing rivers, and endless forests surrounding her new home.

It did not take long for O’Toole to fall in love. She fell in love with the environment, with fishing, and with a fisherman. They bought into the local industry, purchasing a commercial fishing permit for a whopping $300,000 — about three times the value of their home. However, that very year, on March 24, 1989, the Exxon Valdez ran aground, spilling approximately eleven million gallons of oil into the sound. O’Toole and her young family lost almost everything, eventually leaving the industry and the community entirely.

Over thirty-five years later, when I ask O’Toole about March 24, 1989, she immediately begins to cry. In the days after the spill, she worked in Cordova’s city hall collecting testimony from townspeople. “I remember one man who described watching a sea otter scratch its eyes out, because of the oil in its eyes,” recounts O’Toole.

“Everybody’s got a Shangri-la in their hearts,” she tells me, “please think of yours when you contemplate what has happened to ours.” For O’Toole, Exxon needed to be held responsible for what they did to this Shangri-la.

O’Toole, and about thirty-four thousand other Alaskans, became plaintiffs in a landmark case against Exxon. They sought punitive damages to cover the devastating economic and environmental carnage of the spill. They won.

In 1994, a jury awarded an unprecedented $5 billion in punitive damages. However, fourteen years later, the US Supreme Court would cut down that amount to approximately $500 million (a 90 percent reduction).

O’Toole and her lawyers lost, in part, because they lost a debate about the nature of rationality. A team of Exxon-funded scholars, including Kahneman, conducted extensive research into jury decision-making. This work purported to show that jurors were incapable of working through these sorts of cases.

12 Angry Alaskans

“It was just overwhelming, and to tell you the truth, it brought me to tears,” Janette Garrison said. Garrison was an elementary school janitor who became one of the twelve jurors in the 1994 case. “We’re not experts. How on earth did they think we’re going to deal with this?” she asks, recounting how monumental the task of calculating damages felt to her at the time.

Garrison was typical of these twelve jurors. They held working-class jobs with little prestige, e.g., janitors, McDonald’s cashiers, flight attendants, factory workers, and so on. A few were unemployed.

Yet Garrison and the other jurors did expertly reason their way through the case. In many instances, they even seemed to grasp the matter better than the experts. That’s what David Lebedoff, an attorney and author of a book about the case, told me. “The jurors in this case seemed to be trying very hard,” Lebedoff said. “And this jury was smarter than the lawyers.”

To give one example, Garrison and other jurors felt that the lawyers did not adequately address certain facts that were most salient to the case. They had an elegant solution to these types of problems: they staged an elaborate role-play. Different members of the jury played each side, and they essentially retried the entire case in the jury deliberation room.

This was not the only case of extraordinary 12 Angry Men–esque efforts to reach the truth. In arguments, Exxon’s lawyers claimed that the company cleaned up oil from the spill. The immediate cleanup was controversial, but it would not be unreasonable to imagine that — by this point — Exxon was telling the truth. After all, the trial was held five years after the spill.

But Garrison was not so credulous. She asked the judge if the jury could go on a field trip to see the beaches with their own eyes. The judge agreed.

Garrison brought her yellow rubber kitchen gloves, determined to flip rocks and look for oil. Her yellow gloves turned black. “I was angry,” Garrison says, “to tell me you’ve cleaned it up and it’s gone? No, it’s not gone!”

Soon after, Judge H. Russel Holland read the verdict to the packed audience. “My memory is that the courtroom just erupted when I read $5 billion,” he later recounted in an interview.

However, Exxon would challenge this particular jury — as well as the soundness of juries in general.

“The General Principle of Cognitive Engineering”

In legal appeals, the company frequently cited academic studies they had funded into jury deliberation. Many of these were compiled into an edited collection called Punitive Damages: How Juries Decide, published in 2002. The lead editor was Cass Sunstein, with contributions from a number of esteemed psychologists, behavioral economists, and legal scholars, including Kahneman, Reid Hastie, David A. Schkade, W. Kip Viscusi, and John W. Payne.

In short, the book argued that jurors lack the experience, expertise, and attention spans to reason through complex punitive-damage cases. In part, this is because a number of cognitive biases and heuristics seem to cloud their decision-making, and because they seem to default to certain — in the eyes of the authors, illegitimate — “folk” notions of blame. The authors’ method generally was to convene mock juries, and they sometimes statistically aggregated individuals into “synthetic juries.” These experimental juries make seemingly incoherent and inconsistent assessments.

To these authors, it would be best if juries did not decide these matters, since this “violates the general principle of cognitive engineering, that we should match human capabilities to the demands of the task.” Their “principles of cognitive engineering” lead them to conclude that an expert administrative system would be preferable to a jury of ordinary citizens.

“A Swarm of Papers”

Several legal scholars have made stinging critiques of the book’s methodologies, its conclusions, and its normative underpinnings.

Professor Mary Rose, a jury scholar at the University of Texas at Austin, tells me that she was heartbroken upon the book’s release. “I want to be very clear that I don’t think there’s anything wrong with a scholar coming along and saying, we’ve identified these problems . . . [and] we feel that there is an awful lot of variability in these awards,” she says. But Rose tells me that it was unusual to have a topic be captured by such a swarm of papers, and by a set of researchers whose agenda seemed to jump so quickly to “the jury is incompetent,” without considering possible reforms or educational measures to ameliorate potential problems.

Rose’s colleague Neil Vidmar, professor emeritus at Duke University, authored an extensive methodological rebuke of the Exxon-funded research for the Emory Law Journal. The article outlines a number of ways that the experimenters seemed to be biased toward certain results, leading to serious doubts about the generalizability of the particular experiments.

For instance, Vidmar notes one case in which the experimenters’ mock-trial descriptions include five times more lines documenting the plaintiff’s case than the defendant’s. When subjects deliberated based on those descriptions, the researchers observed a “severity shift,” which was taken as evidence of poor group decision-making. Subjects seemed to be wrapped up in a kind of psychological groupthink that pushed them toward harsher awards than they had initially envisioned before entering the group. However, to Vidmar, there is a simpler explanation: the group deliberated toward the more severe decision because the evidence presented suggested that direction.

Rose’s and Vidmar’s research programs, unlike these experiments, involve observation of actual juries, as well as archival research into real punitive awards. These methodologies paint a very different picture of jury rationality. Rose claims that jury decisions are generally not erratic and that their thinking tends to be influenced by reasonable factors, like the strength of evidence presented in a case.

“There isn’t usually an area of research where the experimental results are so at odds with what you see when you do archival studies of actual verdicts,” Rose claims.

“Poisoned the Pool”

Yet despite these methodological concerns, the Exxon-funded research was enormously influential.

Shireen A. Barday reviewed this research for a class project when she was a student at Stanford Law School. Barday, now a partner at Pallas Law, cataloged her findings in a 2008 Stanford Law Review article. The article tracked all legal scholarship published on the subject of punitive damages and jury deliberation between the years 1992 and 2007.

Of the thirty-three total articles she found on the subject, about half were corporate-funded, and most of that half was underwritten by Exxon. Punitive Damages was a collection of many of these Exxon-funded articles that Barday tracked.

In total, Exxon supported a third of the published papers. That is to say, the company underwrote a third of the entire academic knowledge base on this subject, for a span of fifteen years.

Furthermore, the corporate-funded articles were twenty-five times more influential in real court cases. Corporate-funded research was cited in twenty-five separate cases, whereas ordinary university-funded research was only cited in one.

“As I reflect on it now, funded punitive damages research poisoned the pool,” Barday tells me. “It created this idea that the juror is incapable of determining punishment in any rational way.”

Seeing Some Bias, Missing Other Bias

To be clear, there is no evidence showing that Exxon exerted undue pressure on these scholars.

All six authors denied or ignored repeated requests for interviews, including Kahneman, who died while I was doing research for this article. Therefore, it is difficult to know exactly how this research took shape on the inside, and what influence Exxon might have leveraged. One of the coauthors, Emeritus Professor John W. Payne of Duke University, provided me a written statement asserting that he was proud of the book.

However, there is ample scholarly evidence documenting how research funding tends to produce research that is friendly to the funder. Further, much psychological research — which, presumably, Kahneman et al. would be familiar with — suggest that implicit biases may be formed in these kinds of working relationships.

The experience of sociologist William R. Freudenburg vividly illustrates how these biases might form. Freudenburg was approached by Exxon to author research, and Freudenburg kept extensive field notes documenting the experience, later publishing those notes in a little-known academic tell-all. The article anonymizes Exxon, but Freudenburg would later confirm to Alan Zarembo of the Los Angeles Times that Exxon was indeed the subject of the article.

Freudenburg recounts how a candid Exxon lawyer openly admitted the company funds research to use as legal ammunition. “The judges themselves don’t usually read [the studies], but often their clerks will read them . . . and quite a few of the clerks, nowadays, are pretty open to these kinds of arguments. . . . Quite a few of them now come out of a law and economics program,” the lawyer said, according to Freudenburg’s field notes.

Freudenburg agreed to write an article with Exxon support. However, Exxon eventually soured on him, cutting financial support for the article at the draft stage. Freudenburg makes clear that Exxon never pressured him to argue for something he did not believe in. However, he wonders whether Exxon’s kindness might have implicitly influenced his views, pushing him to self-censor or self-edit “simply because I might want to anticipate the preferences of people I was coming to see as friends,” Freudenburg concludes.

What Did Exxon Get for the Money?

In 2008, the Supreme Court would reduce Exxon’s punitive damages by 90 percent, saving Exxon $4.5 billion. The case, Exxon Shipping Co. v. Baker, is often cited as one of the ten worst Supreme Court decisions of all time.

In its decision, however, the court appeared to rebuke the Exxon-funded research. In its curious seventeenth footnote, the court writes that they are “aware” of the literature that we have been discussing here, but “because this research was funded in-part by Exxon, we decline to rely on it.”

However, many of the legal scholars I spoke with doubted the sincerity of this footnote. “You’re trying to appear unbiased, argues Barday, “but is it really possible for you to separate what you have learned?” she asks. Barday says that the nature of the court’s reasoning suggests the research played some role, even if they claim to have ignored it. Furthermore, references to Punitive Damages had already appeared in appeals at lower courts, as well as another Supreme Court case.

The case was actually decided on the peculiarities of maritime common law. However, the court’s statements do sometimes parrot the thinking of Punitive Damages. For instance, in the decision, Justice David H. Souter calls the punitive damages “unpredictable” and “eccentric,” and he claims that “a penalty should be reasonably predictable and consistent.”

One of the book’s coauthors agrees that the research played some role in the verdict.

“I am also very proud that the research appears to have had an impact on the US Supreme Court decision involving the Exxon Valdez case, and other court decisions, although I know that the Supreme Court was cautious in citing our work,” Professor Payne wrote in a prepared statement.

After Punitive Damages, most of these scholars never returned to jury-deliberation research. Kahneman would go on to win a Presidential Medal of Freedom, awarded by President Barack Obama. Sunstein, his coauthor and frequent collaborator, would lead Obama’s much-vaunted “nudge” unit.

The Rationality Wars

Kahneman’s program of heuristics and biases research has faced sharp critiques, although these criticisms are seldom heard outside scholarly circles. For instance, in the late 1980s and early ’90s, the psychologist Gerd Gigerenzer squared off against Kahneman and Tversky in a series of acrimonious talks and papers that came to be known as “the Rationality Wars.”

“It is a chance to think about something really fundamental. What is rationality? And also an insight that there are always several possible answers,” says Gigerenzer of the debate.

Gigerenzer argued that Kahneman and Tversky’s dire experimental results simply disappear if you tweak the studies so that experimenters express themselves in more natural ways. Earlier, I suggested Kahneman’s work was a “sleight of hand.” Here this can be read literally: the experiments amount to parlor tricks, where careful experimenters try their best to fool unwitting research subjects.

Gigerenzer’s own account of rationality is more sensitive to a subject’s actual environment and social context, and it considers a pluralistic set of decision-making processes. This involves appreciating cultural and social knowledge, conversational maxims, habits, traditions, and other mental shortcuts (or heuristics) that generally work. The picture that emerges is that of a person and community muddling through shared problems however they can, improvising along the way.

That is not the picture in Kahneman’s research, which tends to posit narrowly defined answers to the questions it gives subjects. Subjects must follow abstract logical rules toward the ultimate goal of utility-maximization. Predictably, they fail.

This is the second way in which the research can be read as a sleight of hand. While the reader is fixated on the clever experiments and our inability to ace them, they miss the normative assumptions that are smuggled in behind.

Market-Friendly Assumptions

Punitive Damages follows in the tradition of the academic field of law and economics, which arose from the Chicago school of economics.

“It is just taken as a given,” says Professor Rose, describing the assumptions of these authors, “that there should be a particular kind of rationality that is helpful to the marketplace.”

Market-friendly assumptions undergird at least three core normative claims in the book; these claims are asserted but never fully defended. First, the rule of law is predicated on predictability. Second, the identity of the defendant (e.g., firm size, locale, etc.) should not matter to how much is assessed in damages. Third, juries should be friendly to corporate cost-benefit analyses, including costing potential legal damages from environmental disasters.

All three claims are controversial. There is no reason in principle that punitive damages must always be predictable; different communities can punish environmental damages differently, if they so choose. Second, it may be reasonable that a larger firm face stiffer punishments, since it takes more to make them feel the pain of a financial penalty. Third and finally, a corporate cost-benefit analysis can look morally objectionable to a community, especially when something of seemingly immeasurable value — i.e., their home, their way of life, etc. — is what is being costed.

When research subjects fail to meet these contentious normative standards, their reasoning is dismissed. However, is this not begging the question? Perhaps the initial definition of rationality is the real issue. On this reading, the research subjects are simply disagreeing with their experimenters.

“The fact that people differ in the level of punishment that they would give in any given situation doesn’t make them irrational; it means they differ in how much punishment they think is required,” argues Barday.

Reflecting back on the jury of which she took part, Garrison believes they did a good job reasoning through the case. She was disappointed to learn that Exxon-funded ideas might have contributed to lowering the $5 billion in punitive damages.

“What they did with their little theories, I didn’t really care,” says Garrison, “they gave this job to the jury, and the jury did what was asked. What makes them more qualified?”