The Persistent Illusion of Impartiality
It's impossible to guarantee that our decisions are impartial.
Posted June 2, 2010
On a Saturday afternoon in late November, 1951, Princeton and Dartmouth played an aggressive football game. By the end of the game, Princeton's star player, Dick Kazmaier, had been carried from the field suffering concussion and a broken nose, and a Dartmouth player was sidelined with a broken leg. No one denied that the game was unnecessarily rough, but the spectators apportioned the blame differently depending on their allegiances. The Daily Princetonian student newspaper laid the blame "primarily on Dartmouth's doorstep," whereas The Dartmouth emphasized that "most of the roughing penalties were claimed against Princeton." A week later, researchers Albert Hastorf (from Dartmouth) and Hadley Cantril (from Princeton) asked groups of Princeton and Dartmouth students to evaluate video footage of the game. Everyone watched the same footage, but the Princeton students perceived five more infractions against Dartmouth than against Princeton, whereas the Dartmouth students believed both teams committed the same number of infractions. Everyone watched the same game, but their perceptions of the game were tainted by partisanship.
Perhaps it's not surprising that football supporters aren't impartial--their allegiances are clear and the consequences of impartiality are trivial. But Hastorf and Cantril's classic study was one of the first to show that people with pre-existing views perceive the world--even in real time, as it progresses--differently depending on those preconceptions. In contrast to the situation at a football game, the real problems arise when we assume incorrectly that people are impartial, and some of our most venerated institutions run on the assumption that humans are capable of perceiving the world impartially. Justitia, the Roman goddess of Justice, stands blindfolded to show that she considers the merits of each case without bias. Supreme Court justices may not be deities, but we nonetheless demand of them superhuman impartiality. Each time the president nominates a new Supreme Court justice, the citizens who didn't vote for the president claim the nominee is biased while the president's supporters defend the nominee's ability to remain impartial. Now, conservative bloggers and Republican intellectuals are claiming that Elena Kagan, President Obama's latest nominee, is incapable of deciding cases impartially.
The truth is that no human--Supreme Court justice or otherwise--is impartial. If your mission were to design a being incapable of perceiving the world impartially, you couldn't improve much on our current design. Part of the problem is that we have sophisticated methods of hiding our biases. In one study, Mike Norton and his colleagues asked male undergraduates to choose amongst three job applicants who were competing to run a construction company. To help them choose, the experimenters gave a brief description of the candidates' education and work backgrounds. Two of the candidates were clearly superior to the third, and of those two candidates, one had more industry experience and the other was better educated. The students generally felt that education was the more important criterion, so they selected the better educated candidate for the job about 75% of the time. But the researchers threw in a twist for some of the participants, telling them that one of the superior candidates was a man and the other was a woman. When the male candidate was better educated, the students again suggested that education was the more important criterion, and they preferred the male candidate 75% of the time. When the female candidate was better educated, however, only 48% of the participants selected the female candidate, and only 22% suggested that education was the more important selection criterion. The results imply that the male students had always subconsciously preferred the male candidate, and they ranked the relative importance of the selection criteria to support that preference.
This subtle technique, known as casuistic reasoning, is insidious because it's very difficult to detect. The male students who suggested that the male job candidate was better qualified weren't bad people; they weren't overtly choosing an inferior male candidate based on gender biases. Instead, their pre-formed preferences for a male candidate clouded their judgments and led them to prefer a different hierarchy of selection criteria that supported their preferences.
The same problem affects politicians and, more worryingly, Supreme Court justices. Comedy Central's The Daily Show with Jon Stewart has taken on the crusade of ferreting out casuistic reasoning. In one segment, Republican strategist Karl Rove defended Sarah Palin's candidacy by noting that she had been the mayor of the second largest town in Alaska (population 10,256). But when the Democrats considered Virginia governor Tim Kaine for the Vice Presidency, Rove observed that "[Kaine] was mayor of the 105th largest city in America. And again, with all due respect to Richmond, Virginia, it's smaller than Chula Vista, California; Aurora, Colorado; Mesa, or Gilbert, Arizona; North Las Vegas, or Henderson, Nevada. It's not a big town..." Perhaps, but Richmond's population of 200,000 dwarfs Wasilla's population by an order of magnitude. Rove is widely acknowledged as a canny strategist, so it's difficult to believe that he intended to make two blatantly inconsistent comments. More likely, when Rove was motivated to support Palin's candidacy, the criterion "mayor of a moderately sized town" seemed to be a sign of experience. Instead, when Rove considered Governor Kaine's VP candidacy, Richmond's population seemed measly, and serving the town as mayor was no longer a sign of experience. Like Democratic strategists, Karl Rove is necessarily partisan--but that doesn't justify their adoption of shifting standards when interpreting the same facts for different purposes.
In January 2004, Vice President Dick Cheney and Supreme Court Justice Antonin Scalia embarked on a hunting trip. Three weeks earlier, the Supreme Court had agreed to hear a case involving the VP, so Justice Scalia's decision to join the VP prompted the media to question whether he could remain impartial during the trial. In typically florid (and entertaining) style, Justice Scalia refused to recuse himself and lashed out at his detractors: "I do not believe my impartiality can reasonably be questioned. If it is reasonable to think that a Supreme Court Justice can be bought so cheap, the Nation is in deeper trouble than I had imagined. For Pete's sake, if you can't trust your Supreme Court justice more than that, get a life." Justice Scalia's response suggests that he has no overt bias, and perhaps that's true. But no one--not even the most devotedly impartial Supreme Court Justice--is immune from the hidden cognitive foibles that plague every human. Just as Mike Norton's male students unwittingly moved the goalposts to justify hiring an inferior male job applicant, so Supreme Court Justices are liable to emphasize precedents that support their preferred legal conclusions. It's not enough to want to be impartial; our biases are so well hidden that we're destined to be partisan as long as we're motivated to promote one conclusion over its alternatives.