[The second post in our online symposium about recent California admissions decisions is from our own Brad Wendel.]
Two Stephen Glasses appeared before a California State Bar Court hearing judge. One was a serial liar – a fabulist, to appropriate the title of his roman à clef – who made up dozens of articles for The New Republic and other magazines out of whole cloth, in what the Newseum in Washington, D.C. called one of the worst examples of misconduct in the history of journalism. The other Stephen Glass was a young person trying to cope with intense pressure from his family to achieve professional success, who embarked on a pattern of deception that led to his disgrace, and has slowly but steadily turned his life around with the help of therapists, friends, and a new career aspiration as a lawyer. Which of these characters is the real Stephen Glass, and what can we infer about what that person will do in the future? The trouble is, we have no idea. Given the complex interaction between character and situational factors, at this point in time we can do no better than guesswork if we try to predict whether admitting Glass to practice law is likely to result in harm to clients.
One of the central findings of behavioral psychology is that situational forces are much more significant determinants of behavior than personality or character. The Milgram Experiments on obedience to authority and the Stanford Prison Experiments famously showed that ordinary people will do terrible things given the right group dynamics and social forces. The explanation of why people do bad things is often not that they are bad people but that they are ordinary people in situations that are productive of wrongdoing. However, another well documented feature of human psychology is the fundamental attribution error (FAE) – we tend to attribute the explanation of wrongdoing to character traits or dispositions, not features of the situation. Asked to explain the Milgram results, people will often say the subjects must have been sadists. The same effect can be observed outside the laboratory. Ask people for an explanation of the decision to launch the space shuttle Challenger, the collapse of Enron, the Abu Ghraib abuses, or the failure of the ratings agencies or the risky financial transactions leading up to the 2007 financial crisis, and you will probably hear an explanation in terms of the greed, dishonesty, or cruelty of key players – the “bad apples” account. It turns out, however, that the vast majority of participants are not bad apples, but are ordinary people whose ethical decision-making is subtly influenced by group dynamics such as in-group favoritism, pluralistic ignorance, induction effects that evaluate conduct in terms of previous similar actions, and subtle influences on the way people construe unfamiliar or ambiguous circumstances.
This is not to deny that people react to situations differently. In his book, Eat What You Kill, Mitt Regan gives an explanation of the decision of a bankruptcy partner at a major New York law firm to falsify a disclosure of the firm’s representation of another party in a Chapter 11 proceeding. His story is rich and nuanced, but a reader may ask (as I did in a review of the book) why it was only John Gellene who lied to the court. Other lawyers, including a litigation partner who urged disclosure, did the right thing in the case. Social psychologists do not deny that people have personality traits. The claim, rather, is that “people [do not] typically have highly general personality traits that effect behavior manifesting a high degree of cross-situational consistency.” John M. Doris, Lack of Character: Personality and Moral Behavior (2002), p. 39. The determinants of behavior include both character traits and situational factors, but people commit the FAE when they overestimate the predictive value of character traits. We tend to have a significantly higher degree of confidence than is warranted in our attribution of dispositions (e.g. saying Stephen Glass is a liar) and our predictive judgments (e.g. estimating that it is likely that Glass will commit dishonest acts in the future). It is difficult to overcome the tendency to explain behavior in trait terms, leading to the FAE, because it appears to be a product of unconscious coding and confirmation bias.
John Gellene might have been more predisposed than other lawyers to falsify the document, but before he started working on the case, there would have been no way to know with any significant degree of reliability. Ex post we feel confident in our judgment that “John Gellene is dishonest” is the best explanation of the act of falsifying the document. Research shows, however, that this is nothing more than hindsight bias – that is, the tendency to significantly overestimate the ex ante likelihood of an event. (See Jeffrey J. Rachlinski, A Positive Psychological Theory of Judging in Hindsight, 65 U. Chi. L. Rev. 571 (1998) for the details.) As a torts teacher, I am constantly reminding students not to assume that just because an accident occurred, it was the result of the defendant’s failure to use reasonable care. Humans are unfortunately just not very good at making predictive judgments about risk in general, and when that deficiency is combined with the FAE, the result is gross overconfidence in the reliability of our judgments about when a person’s past acts are predictive of future behavior.
I get the general reaction to Stephen Glass. I have subscribed to The New Republic since my undergrad days, and felt betrayed when I learned that the articles of his I had enjoyed were fabrications. My off-the-cuff assessment of him would be “sleazeball” or “liar.” I understand this kind of evaluation from a reader comment on Andrew Sullivan's blog:
This is a person who has demonstrated, time and time again, that he is morally and ethically challenged – much more so than a person who has committed petty offenses or who has a drug conviction, in my opinion, but someone who literally cannot be trusted to tell the truth.
(Emphasis added.) The FAE is an ingrained tendency, and it is a deeply counterintuitive claim that past wrongdoing does not reliably support an inference to character traits, the existence of which enables one to make reliable predictions of future behavior. But the same is true of many of the findings of cognitive psychology. I love the story of the “hot hand” study by Tom Gilovich (of the Cornell psychology department), which disproved the folk wisdom of basketball fans that players sometimes tended to get on a hot streak and make a series of field goals or foul shots with unusual success. When hundreds of hours of game films were studied, however, it turned out that the sequence of made and missed shots were within the range of random distribution. But this didn’t satisfy former Boston Celtics head coach Red Auerbach, whose reaction to the study was “Who is this guy? So he makes a study. I couldn’t care less.” As Daniel Kahneman puts it: “The hot hand is a massive and widespread cognitive illusion. . . . The tendency to see patterns in randomness is overwhelming – certainly more impressive than a guy making a study.” Daniel Kahneman, Thinking, Fast and Slow (2011), p. 117.
While the stakes are obviously much higher, an analogous problem is the use of expert testimony to predict future dangerousness for the purposes of capital sentencing. The American Psychiatric Assocation has stated in amicus curiae briefs that psychiatrists should not testify as an expert that a defendant has a long-term likelihood of committing future acts of serious violence, because there is simply no scientifically reliable method for making this prediction. (See, e.g.,) The familiar “reasonable degree of medical certainty” standard for expert testimony accordingly cannot be satisfied. Faced with this evidence, however, prosecutors sound like Red Auerbach responding to the hot hand study. The ABA Journal quotes one district attorney who said, “common sense and 23 years of experience as a lawyer have convinced him that such predictions can reliably be made.” Who needs scientific evidence when you’ve got common sense and 23 years of experience? So some guy made a study – so what?
The California Supreme Court held that Glass had not carried his burden of demonstrating his fitness to practice, in part because of his lack of candor in the admission process, both in California and, earlier, in New York. Any evasiveness, partial disclosure, or game-playing with the process is the kiss of death in the character and fitness evaluation. I always advise students this is not the place to try out Bill Clinton’s techniques for avoiding answering hard questions. One gets the sense, however, that the result would have been the same if Glass had disclosed each and every instance of fictionalizing articles. Some commentators on the case have argued that the California court’s decision is justified by the allocation of the risk of error: On the one hand, false positive – i.e. an erroneous prediction that Glass will offend in the future – will affect only Glass; on the other hand, a false negative may result in harm to clients. Isn’t it better that the risk be borne by the concededly sleazy Glass than by an innocent client? Stated in this form, the argument is a version of the precautionary principle, which is, when in doubt, avoid doing anything that will create a risk of harm. Or, in cases of doubt, better safe than sorry. Problems with the precautionary principle are well known, however. For one thing, it considers only harms on one side of the equation. While courts have repeatedly stated that admission to the bar is a privilege not a right (and thus Glass has no Mathews v. Eldridge-type claim to be judged on the more competent evidence), there still seems to be a moral right on Glass’s part to have his application considered fairly, given the significant investment he has made in his legal training. “In real-world controversies, a failure to regulate will run afoul of the precautionary principle because potential risks are involved. But regulation itself will cause potential risks, and hence run afoul of the precautionary principle too.” Cass R. Sunstein, Risk and Reason: Safety, Law and the Environment (2002), p. 103.
Maybe observers aren’t particularly sympathetic to the risk posed to someone like Glass. If you’re a lying sleazeball, you risk not being admitted to the bar – too bad for you. This, of course, depends on our confidence that we can determine that someone is, in fact, a lying sleazeball. What if the truth of the matter is that the real Stephen Glass is the second character described above, who screwed up royally, realized it, and has spent the last ten years trying to make it right. It’s a long, unsteady process, and maybe he didn’t do everything someone else would have done as part of a process of rehabilitation, but do we not believe in the possibility of redemption? It’s at least conceivable that the character of the real Stephen Glass is that of someone who made a big mistake but has since turned his life around. My view of the psychological evidence is that we simply do not know enough about the character of either Stephen Glass – either the serial liar or the rehabilitated person – and its cross-situational stability to justify making a predictive judgment of his future dangerousness. But for those who are less persuaded by Milgram, Darley, Batson, Zimbardo, Nisbett, Ross, and the rest of the social psychologists who believe that situational factors are more important than character as determinants of behavior, what makes you so inclined to believe that Glass isn’t on the right road at this point in his life? Presumably the answer is that he fudged the truth on his New York application. I’ll grant that is a very bad fact indeed. But I still get the feeling the California court would have denied his admission anyway, and that’s troubling for someone who believes in the possibility that anyone can make a new beginning.
If the character and fitness requirement is justified primarily as a prophylactic means of protecting the future clients of a lawyer like Stephen Glass, I think it has to be abandoned. As Deborah Rhode showed in her classic article, the bar tends to articulate a public protection rationale for the character and fitness screening process. As one bar spokesperson rather colorfully put it, the objective is “eliminating the diseased dogs before they inflict their first bite.” Deborah L. Rhode, Moral Character as a Professional Credential, 94 Yale L.J. 491, 509 (1985). (Rhode’s article, pp. 555-59, provides an overview of the behavioral psychology literature summarized above.) The theory for denying Glass’s application for admission is that he is a diseased dog. But the situationalist critique of the FAE shows that we are all potentially diseased dogs. A better approach to regulation would be to aim at mitigating the situational factors that tend to produce unethical behavior. The Gellene case is a good example. Bankruptcy Rule 2014 requires disclosure of any connection with a creditor or other party in interest; it does not employ language similar to that of Model Rule 1.7(a)(2), requiring action only where the concurrent client relationship is a material limitation on the representation. The bankruptcy rule therefore avoids the judgment calls that lawyers must make in evaluating conventional conflicts of interest and makes it less likely that these judgments will be influenced by self-serving cognitive biases. There might have been an argument that the firm’s concurrent representation of a principal in an investment firm that was a creditor in the Bucyrus-Erie restructuring proceeding was not a material limitation on the representation of the debtor (although I think that would be a pretty dubious argument), but there was no argument that there was no connection. By providing less latitude for judgment, the bankruptcy rule is less vulnerable to abuse by lawyers who may feel a great deal of pressure to keep clients, or other lawyers in the firm, happy by not disclosing a concurrent representation.
I understand the symbolic and signaling function of the character and fitness requirement. As one of the comments on Andrew Sullivan’s post noted, a student just beginning law school usually hears at orientation that he or she is entering a profession with high ethical standards, which makes demands above and beyond simply complying with law and expects its practitioners to satisfy demanding requirements of honesty and trustworthiness. Without the character and fitness process, would the law become, or at least seem to be, just another trade or business? (As I’ve written elsewhere, I’m a bit uncomfortable with this tacit dissing of the ethics of businesspeople, not only because it sounds sanctimonious by lawyers, but also because it tends to reinforce the attitude that business ethics is nothing more than the morals of the marketplace + maximizing shareholder value.) I’m all in favor of symbolically reaffirming our profession’s commitment to ethics, but it is more than merely symbolic when someone who has invested three years and probably in excess of $150,000 in tuition and living expenses to become a lawyer is denied admission because of prior acts of dishonesty. If we’re going to deny someone access to a valuable privilege (n.b. not saying it’s a right), we had better be confident in the reliability of our decision-making process.