Book Review: Bruce Schneier, Liars and Outliers
Trust is all around us. I trust drivers to wait at the red light while I cross the street, my doctor to base his diagnosis on the best available evidence, and my neighbor not to crack me over the head and swipe my wallet and phone as we ride the elevator together. In Liars and Outliers Bruce Schneier attempts to make rigorous the intuition that trust is the foundation of the remarkable degree of cooperation that characterizes successful societies. He makes his case in the context of social contract theory, game theory, behavioral economics, and moral, social, and evolutionary psychology. There is a lot in Liars and Outliers to engage moral philosophers, particularly those interested in the likely consequences of an array of laws, policies, and practices whose goal is to promote trust and cooperation on the one hand and to discourage or deter distrust or defection on the other. The defectors are the liars and outliers of Schneier’s title. The liars are those who take but don’t give, outliers those who challenge a prevailing norm on moral grounds. The most successful societies keep defection to a tolerable minimum, stopping short of the repression that reduces willing cooperation.
Schneier holds that some degree of trust is innate in our species and thus that it must have been adaptive, at least among intimates. Without it cooperation and thus society could never have come about. Cooperation is a positive sum game. We are all better off when we can count on others to cooperate with us both at work and at play. In the best of circumstances a positive feedback loop forms: cooperation builds even greater trust, which leads to still more cooperation, and so on.
Robust cooperation, says Schneier, is achieved through four “societal pressures,” which make it “more attractive and defection less attractive” (p. 67). The pressures are morality, reputation, institutions, and security. The first two are informal. Regarding moral pressure, most of us most of the time refrain from violence, stealing, and cheating, not because we fear reprisal, but because we feel good when we do the right thing and bad when we don’t. Whence the moral sentiment? Natural selection. We evolved as social beings, and those whose actions were animated by moral reasons tended to inspire good will and confidence in others, thus leaving on average more copies of their genes for like behavior than their relatively antisocial peers. The rub is that moral pressure works best with family and close friends, that is, with “people whose intentions we can trust” (p. 84). For a broader circle reputation picks up. Concern for reputation is a far more effective goad to good behavior than moral sentiment alone. Again, natural selection can account for the universal obsession with reputation. Those with good ones tended to do better in the mating market than those suspected of unreliability, dishonesty, or treachery. Individuals who were disposed to cooperate would enhance their reputations for trustworthiness, inducing more people to cooperate with them. In other words I’m more willing to trust you, even if I don’t know you particularly well, because I know that you’re concerned about your good name. A strong reputation would not only make its possessors more eligible mates, and it would also put them in a position to benefit relatives, who share more of their genes on average than non-relatives. By contrast, for most of human history those averse to cooperating faced the possibility of banishment, the hunter-gatherer equivalent of the gallows.
Reputation works best with relatively small groups, “where there are strong social ties among the individuals” (p. 184), say, between 150 to 500 people. However, it can scale somewhat beyond this in some circumstances. Schneier cites studies showing that we are more likely to trust people who look more or less like us or who speak the same dialect. Both are extremely difficult to fake. They are in turn more likely to trust us and thus more willing to cooperate. Persuasively demonstrating commitment can also extend reputation’s range. Schneier offers a historical example. After landing in Mexico, Cortez scuttled his ships to show both his own men and the Aztec leaders that he had come to conquer. This is clearly a “signal” that almost no one would fake and a compelling reason to take the conquistador at his deed even if you didn’t know him from Adam.
However, as agriculture permitted societies to expand, reputational pressure became less effective at inducing cooperation and deterring defection. The larger the society, the easier it is to defect without reputational consequences. In a city I can steal you blind, and neither you nor anyone else might have a clue. With computer networks I can fleece you from halfway around the world. As a result, post-tribal societies have sought a way to formalize reputational pressure through laws, sanctions, and incentives. Where individual trust or concern for reputation is no longer enough to sustain cooperation, trust in government and other institutions fills the gap. I fly without fear of crashing, dine in restaurants unconcerned about poisoning, and take pills for high blood pressure in the faith that I’ll be healthier for it. My trust isn’t in that pilot, this chef, or that doctor, but in the relevant institutions, in the system that produces and regulates them. What effective government in particular does, as Hobbes recognized long ago, is to raise the price of defection and lower the costs of cooperation, ensuring that most defectors are appropriately punished and that cooperators can go peacefully about their business.
The final societal pressure is security, which includes everything from locks and high fences to guards and alarms. Security is preemptive in a way that the previous pressures are not. Locks can stop a thief at the start, even if he isn’t concerned about reputational fallout and couldn’t care less about getting caught. More importantly, security is the most scalable pressure, potentially thwarting all comers. And unlike the other three, it is not based on trust. Actually, security is as old as the hills. Its “systems” predate the first barricade or moat by many hundreds of millions of years. Schneier points out that sexual reproduction, which first appeared some 1.2 billion years ago, might have been an evolutionary response to parasites. Immune systems, which emerged much later, offer a firewall against pathogens. Other security measures favored by natural selection include horns, shells, and noxious tastes and smells. All security, both natural and artificial, is subject to an arms race. The better parasites, predators, and intruders get at penetrating defenses, the more sophisticated the security has to be, which in turn prompts attackers to improve their methods, and so on. Advantage: bad guys. Still, the race goes on. In both the natural and artificial spheres the parties have no choice but to run it.
Schneier devotes much of the second half of the book to real world societal dilemmas. These include why corporations are more likely to defect from established norms than individuals, why technology means that defection can cause far more harm than in the past, and tragedies of the commons, like collapsed fisheries. This review will focus on one of these, counter-terrorism in the US. He continues in Liars and Outliers his case against the “excesses” (p. 212) of common counter-terrorism measures that he first offered in his well-received 2003 book, Beyond Fear. All kinds of security eventually bump up against the law of diminishing returns: buying the $500 lock will almost certainly not make you ten times safer than the $50 dollar job. Doubling the US security budget won’t reduce the odds of terrorist attack by half. Nor will slashing that spending by 50 % make Americans twice as vulnerable. In fact, there is little evidence that the ballpark $1 trillion that the United States spent to prevent terrorism between 2001 and 2011 has “increased our security proportionately” (p. 135). Schneier adduces full body scanners: There is no reason to believe that this expensive technology has made our airplanes less vulnerable to terrorism. And even if it has, terrorists will simply take their business elsewhere. Perversely, if they bomb a shopping mall instead, killing as many people as they would have had they blown up a plane, the US Transportation Security Administration (TSA) will count that as a win, since its chief mission is preventing air terrorism. It’s true that the TSA was established for the public good. The trouble is that it, like any institution, governmental or otherwise, has some interests that are at odds with the common good. Aside from its official mission, the agency is also committed to surviving and flourishing, even if this means exaggerating the threat of terrorism, conducting intrusive and worthless checks on air travelers, and inadvertently shunting terrorism elsewhere, as in the mall bombing example. It is all but inconceivable that the agency would suggest “returning to pre-9/11 security levels in airports” (p. 201), since such a policy would imply elimination of the agency itself. This is an instance of the principal-agent problem. The agent, in this case the TSA, inevitably has different priorities than the principal, here the government and ultimately, in a democracy, the people. Think of how the US Nuclear Regulatory Commission would likely react to a serious proposal to shut down all nuclear power plants in the country.
The previous paragraph implies that there can be too much security. Raising the cost of defection will mean fewer defectors–up to a point. Schneier’s immune system analogy applies here. A healthy immune system is superb at dispatching the kinds of pathogens encountered by our ancestors but not every varmint that comes along. Similarly, shrewd terrorists can always be one step ahead of security planners. In addition there will always be “irrational actors” (p. 136) who will defect regardless of the costs. As Schneier put it in Beyond Fear, “We have no idea what Al Qaeda’s next plot is, but I can guarantee that it won’t raise an alarm” (p. 253). In the zero-sum game of budgeting, money that the government is spending on terrorism prevention is money that is not spending on cancer research, better public transportation, or larger philosophy departments. Then there are the social costs of diminished privacy, liberty, individualism, time, and convenience–costs that security advocates are wont to ignore or downplay. Although Schneier concedes that the position would be political suicide, he thinks that elected officials ought to propose drastic cuts in security budgets while acknowledging a slight, if unknown, increased terror threat. Unfortunately, policy makers are up against human psychology. We exaggerate risks that are spectacular, rare, random, beyond our control, or much talked about. The 9/11 attacks were nothing if not all of these. Seen in context, however, the loss of life on that day was relatively small compared with, say, the number of highway deaths in the United States in any given recent year. Surely those deaths are just as bad and largely preventable in a way that 9/11 was not. Why no war on traffic fatalities? Psychology again: People tend to downplay risks that are pedestrian, common, taken willingly, or not talked about. The result of vastly overestimating the risk of terrorism, Schneier complains, is that we end up spending far too much on counter-terrorism. And as he suggests in his earlier book, the problem with underestimating common risks is that we don’t devote nearly enough to reducing them.
Before ending this review a few critical remarks are in order. First, Schneier offers some interesting and relevant discussions of evolution in his early chapters. He shows a well-placed faith in accounting for psychological universals like trust in terms of natural selection. But in numerous passages he assumes that evolution proceeds by group selection, that at least some adaptations are favored for the sake of the species (pp. 18, 19, 28, and 242). Here’s one example: “Antelopes don’t need perfect protection against lions…. Instead, they accept the cost of losing the occasional antelope from the herd and increase their reproductive efficiency to compensate” (p. 118). Orthodoxy denies this. To be fair, Schneier does elsewhere speak of natural selection as operating on traits of individuals and, by implication, their genes, but he might have been more careful generally.
Second, in the context of discussing the tragedy of the commons, Schneier defines a non-zero sum game in the following way: “… wins and losses don’t add up to zero: there are outcomes where everyone loses and loses big” (p. 56), as when everyone overfishes and depletes a fishery. First, this is really a definition only of a negative sum game, since in positive sum games wins and loses also don’t add up to zero. Second and more importantly, there are negative sum games in which not everyone loses. Suppose that I get filthy rich from overfishing before stocks collapse. In that case, obviously, not everyone loses. Also, elsewhere he confuses everyone with nearly everyone. In the course of discussing the societal dilemma that the Catholic Church faced when it long ago recognized that it had many child molesters in its midst, Schneier says, “We are all definitely better off if people don’t sexually molest minors” (p. 164). Not all of us, alas: not the molesters, as long as they don’t get caught.
Finally, in at least five instances (pp. 2, 6, 31, 217, and 242) he passes off tautologies as explanatory or informative. Here are two: “If the number of parasites gets too large, if too many people steal or too many people don’t pay their taxes, society no longer works” (p. 2). And “parasites need society to be there in order to benefit from defecting; and being a parasite is a successful strategy only if you don’t take too many resources from your host” (p. 242). We already knew this. Some indication about where the threshold is would have been nice.
But the above is a sideshow. Liars and Outliers is an excellent book. It is carefully argued and brimming with rich suggestions about the central and often unnoticed role that trust plays in successful societies along with the disastrous results when trust fails. Fans of Schneier will certainly want to read it. Those who don’t know his work shouldn’t miss it.