With stocks tumbling, it's time to think about how our brains shape the economy.
Note: This article was excerpted from Shermer's recent book, The Mind of the Market. For a more specific take by Shermer on very recent financial troubles, click here.
"There’s an old English proverb that says, ‘It is an equal failing to trust everyone and to trust no one.’ ”
So begins Paul Zak, a professor of economics at Claremont Graduate University, who is taking the study of economic behavior down to the molecular level. His search for the neurochemistry of trust and trade has brought him to oxytocin, a hormone synthesized in the hypothalamus and secreted into the blood by the pituitary. In women, oxytocin stimulates birth contractions, lactation, and maternal bonding with a nursing infant. In both women and men, it increases during sex and surges at orgasm, playing a role in pair bonding, an evolutionary adaptation for long-term care of helpless infants.
“We know that trust is a very strong predictor of national prosperity, but I want to know what makes two people trust one another,” Zak explains as we sit down in his Center for Neuroeconomics Studies nestled in the bedroom community of Claremont, California.
Zak is the oxytocin man. It says so right on his license plate. Tall and handsome with square shoulders and the physique of someone who works out regularly, Zak’s firm grip and warm smile exude, well, trust. Trained in traditional economics, in the mid-1990s his research led him to connect trust to economic growth.
A 2001 study on trust in forty-two countries, for example, asked people in their native language, “Generally speaking, would you say that most people can be trusted, or that you cannot be too careful in dealing with people?” The results were as diverse as they were striking. At the low end of the trust scale, only 3 percent of those surveyed in Brazil and 5 percent in Peru believe that their fellow citizens are trustworthy, compared to 65 percent of Norwegians and 60 percent of Swedes who trust one another. Falling in the middle of the scale were the United States, at 36 percent, and the United Kingdom, at 44 percent. The rankings remain essentially unchanged even when they are controlled for income. Trust is high in the countries of Scandinavia and East Asia but low in the countries of South America, Africa, and especially in the former Communist bloc.
“The simple correlation between national rates of investment (gross investment per Gross Domestic Product) and trust is strongly positive,” Zak continues. “When trust is low, investment lags. The same positive correlation holds for GDP growth and trust.”
Economic mechanics drive the relationship between trust and prosperity. “Trust facilitates transactions by reducing the number of contingencies that must be considered when ‘doing a deal.’ A deal sealed with a handshake between principals can only occur in a high-trust situation. Let the lawyers work out the details—we have a deal,” Zak offers. “Conversely, when trust is low, negotiations are protracted, and therefore more costly. When transaction costs are higher, fewer transactions occur and investment and economic growth are lower. Trust is among the most powerful stimulants for investment and economic growth that economists have discovered. In seeking to understand why some countries are poor and others are rich, it is, therefore, crucial to understand the foundation for interpersonal trust.”
Perhaps this is why countries that have higher rates of generalized trust show higher rates of return on national stock markets. From these and other studies, Zak realized that in order for a nation to achieve prosperity it is vital to maximize positive social interactions among its members in order to increase trust.
The list of positive social interactions identified by Zak’s research will surprise no one living in a liberal democracy with relatively free markets: protection of civil liberties, freedom of the press, freedom of association, freedom of travel (good roads and reliable infrastructure), freedom of communication (working phone systems), mass education, a reliable banking system, a sound currency, and especially the freedom to trade. He even found a connection between a clean environment and trust, whereby people in countries with polluted environments show higher levels of estrogen antagonists, which lower their levels of oxytocin—and thus their feelings of trust.
Zak went so far as to compute the differences in living standards that trust can effect, whereby “a 15 percent increase in the proportion of people in a country who think others are trustworthy raises income per person by 1 percent per year for every year thereafter.” For example, increasing levels of trust in the United States from its present 36 percent to 51 percent would raise the average income for every man, woman, and child in the country by $400 per year, or $30,000 over a lifetime. Trust has fiscal benefits.
The connection between social interactions and trust can also be seen in the laboratory, when subjects participate in experiments utilizing the Prisoner’s Dilemma game paradigm. In the scenario, each of two subjects pretends to be a prisoner arrested for a crime. Each is independently made the same offer, and both assume the other has been presented the same deal. Imagine that you are the subject and these are your options: (1) If you and your partner both cooperate then you each get one year in jail; (2) If you confess that both you and your partner committed the crime, then you go free and your partner gets three years; (3) If your partner confesses and you don’t, then you receive the three-year penalty while he goes free; (4) If you both confess then you each get two years. If you defect on your partner and confess, then you will either get zero or two years, depending on what he does. If you cooperate and stay quiet, you either get one or three years, again depending on his response. The logical choice, then, is to defect. Of course, your partner is likely going to make the same calculation as you, which means he too will defect, guaranteeing that you will receive a two-year punishment. However, you also then come to the conclustion that he is probably computing this same strategy, and hope that as a result he will realize that you should both cooperate. But then, if that is his line of reasoning, perhaps he’ll defect in hopes that you cooperate assuming that he cooperates, getting him off the hook while you get hit with the three-year penalty. This is why it is called a dilemma.
When the game is played once, both people typically defect. When the game is played over a fixed number of rounds, defection is the norm, because each person realizes the other will defect on the last round, and thus is spurred to defect in the second-to-the-last round, which spurs the other player to defect a round earlier, and so forth, until both players are defecting from the first round. However, when the game is played in something more similar to the real world with the game played over an unknown number of rounds—as would be the case in a negotiation process, where the number of steps is typically unknown—both players keep track of what the other has been doing round by round, and cooperation prevails.
Over and over it has been shown that the best strategy in a series of Prisoner’s Dilemma games without a known end point is “tit-for-tat” with no initial defection. That is, the most self-benefiting thing to do in the long run is to begin by trusting and cooperating and then do whatever your partner does. In such a contest, the winning strategy is called “Firm but Fair,” and calls for you to cooperate with cooperators, to cooperate after a mutual defection, to quit playing with constant defectors, and to defect with partners who always cooperate (otherwise known as suckers).
Even more realistic simulations include the “Many Person Dilemma,” in which one player interacts with several others, and in conditions where subjects are allowed to accumulate experience with the other players thereby giving them an opportunity to establish trust. Finally, the Prisoner’s Dilemma protocol has been applied to the real world of business negotiations, marital disputes, and cold war strategies. It turns out that in both computer simulations and in the real-world, cooperation is by far the dominant strategy.
In a related experiment on cooperation, nine subjects were each given $5. If five or more of the nine cooperated by donating their $5 to a general pot, all nine would receive $10. Although it pays to cooperate (you get $10 instead of $5), it pays even more to defect ($15 instead of $5), as long as at least five other people cooperate. The results were mixed, with many groups of nine subjects failing to achieve the critical mass of five cooperators, because there was no cooperation. Then the experimenters added a step: members of some groups were given the opportunity to discuss their strategy options before playing. Those groups that interacted before playing averaged eight cooperators, and 100 percent of these groups earned cooperative bonuses. By sharp contrast, those groups that did not interact before playing earned bonuses only 60 percent of the time. In a similar experiment on social dilemmas, psychologist Robyn Dawes found that groups given the opportunity to communicate face to face were more likely to cooperate than those who were not. “It is not just the successful group that prevails,” Dawes concluded, “but the individuals who have a propensity to form such groups.”
Where in the brain are these dilemmas resolved? Employing the Prisoner’s Dilemma protocol while scanning thirty-six subjects’ brains using an functional magnetic resonance imaging (fMRI), James Rilling and his colleagues at Emory University found that in cooperators the brain areas that lit up were the same regions activated in response to such stimuli as desserts, money, cocaine, and beautiful faces. Specifically, the neurons most responsive were those rich in dopamine located in the anteroventral striatum in the middle of the brain—the so-called “pleasure center.” Tellingly, cooperative subjects reported increased feelings of trust toward and camaraderie with like-minded partners.
These findings make sense in an evolutionary model, where informal and non-codified rules of conduct developed within small bands of hunter-gatherers because knowing all the other players in the game leads to the evolution of cooperation. The psychological impulse to form relationships and alliances is the deeper cause that lies beneath the moral sentiment of trust, and trade is an effective medium that allows people to create trusting relationships with and form attachments to other trustworthy people. In other words, it is not enough to fake being a cooperator, because over time and with experience, deceivers are usually flushed out. You actually have to believe you are a cooperator, and there is no surer way to believe you are a cooperator than to actually be one, believe it, and mean it. As Yogi Berra counseled, “Always go to other people’s funerals; otherwise, they won’t go to yours.”
From behavior to brains to blood, the molecular basis of trust begins at the most fundamental level of human relations. In her book, Why We Love, anthropologist Helen Fisher makes a distinction between lust and love. Lust, she says, is enhanced by dopamine, the neurohormone produced by the hypothalamus that is associated with reward and pleasure. Fisher shows that it also triggers the release of testosterone, the hormone that drives sexual desire, and thus dopamine is implicated in yet another vital component of human relations—the yearning desire for another. Love, by contrast, is the emotion of attachment reinforced by oxytocin.
Here we return to the connection between trust, trade, and oxytocin, and Paul Zak’s theory that there is a direct connection between the three. “Oxytocin and testosterone are two branches of this trust-distrust system, and as we go through the world we are constantly balancing these levels of trust and distrust,” Zak explained as he recounted some of the experiments he and his team have been running in his center.
In one experiment, Zak found that oxytocin increases “when a person observes that someone wants to trust him or her.” It turns out that with the exception of the 2 percent of the population who are sociopathic, when someone trusts, it triggers the release of oxytocin. Although this is the case for both men and women, Zak has found that “when women get a boost of oxytocin they are more likely to reciprocate than men are. Men are more sensitive to violations of social norms, and when the norms are broken men get a disproportional rise in testosterone. Women don’t like being distrusted, but they don’t have this heated and angry response that men get, which is related to testosterone.”
Oxytocin is deeply involved in the attachment process, whereas testosterone is intensely caught up in the enforcement of the social norms (which may help to explain why far more men go into such professions as the military and law enforcement). In exchange games, the more subjects are behaving in trusting ways, the more money they exchange and the higher the levels of oxytocin that are released by the brain. When Zak asks them why are they giving up so much money, subjects say such things as “it just feels right.” The moral emotion drives behavior, even if the moral calculation beneath the emotion is invisible.
Skeptics might reasonably ask whether oxytocin is the cause or the effect of trust. “To control for that,” Zak says, “we set up an experimental condition where instead of subjects freely choosing to trust someone, we had them randomly pull out a ping pong ball from a box that would determine how much money was given or received, and in those cases there was no significant change in oxytocin levels.” To find out if cooperating and trust lead to the release of oxytocin or if increased levels of oxytocin lead to more cooperation and trust, Zak infused oxytocin into subjects’ brains through a nose spray that is quickly absorbed by the body and discovered that it causes them to act more cooperatively.
Zak has surmounted a number of additional lines of evidence to support his thesis. Trust and happiness: People who trust and are trustworthy report being happier. Trust and touch: We all know how good it feels to be touched by someone, so Zak ran an experiment in which he gave subjects in an exchange game a massage by a licensed massage therapist, which led those who received the massage and received a signal of trust to offer up to 250 percent more than subjects who did not receive the massage and trust signal. Trust and smell: Oxytocin may also be mediated by smell, Zak suggests, noting that there are oxytocin receptors on the olfactory bulb and citing an experiment in which the smell of a mother’s own newborn baby triggers the release of oxytocin and incredibly strong feelings of attachment. Trust and neglect: Animals that are abused or neglected shortly after birth show a loss of regions of the brain that have oxytocin receptors, and those animals become withdrawn, socially inappropriate, and depressed.
The implications of Zak’s research are profound. “Oxytocin is the social glue of society. It is what keeps us together as a civilization. If we didn’t have something in our head that indicated who we should trust and who we should not, civilization wouldn’t work. We’re social animals and we need a trust detector in our heads.”
Zak’s new findings about oxytocin get at something very deep in the evolutionary origins of morality: the role of evil people in society. In my 2003 book, The Science of Good and Evil, I attributed evil to our dual dispositional nature and the fact that in addition to being trusting, cooperative, and altruistic, we are also distrusting, competitive, and selfish, and that the evolutionary forces that led us to be pro-social with our fellow in-group members also led us to be tribal and xenophobic against out-group strangers. Zak pushes the evolutionary model further by looking not just at the potential for evil that resides in all of us, but at the anomaly of evil that lurks within the 2 percent of individuals who are sociopaths. While sociopaths comprise 3 to 4 percent of the male population and less than 1 percent of the female population, they are believed to account for 20 percent of the U.S. prison population and between 33 percent and 80 percent of the chronic criminal offender population. Altogether, these individuals may account for half of all crimes in the United States.
In his social game experiments, Zak typically finds that about two percent of his subjects (he calls them “bastards”) do not respond to oxytocin or other social cues that encourage trust and cooperation. “These individuals have a dysfunction in their oxytocin release and have an identifiable difficulty attaching to others.” But instead of seeing this as nothing more than a biological accident in the wiring of the brain, Zak thinks there may be adaptive evolutionary reasons behind such misfirings. “Bastards are necessary from an evolutionary standpoint because they keep the physiologic balance between appropriate levels of trust and distrust optimally tuned. Without these exceedingly selfish people, humans might have evolved into being unconditional trusters. If so, we would become susceptible to invasion by those who would prey on our perfectly trusting nature.”
As part of his evidence, Zak cites the case of a woman who has a rare genetic disorder that caused her amygdala to calcify and die.
“The amgydala is a primary target for oxytocin in the brain that helps maintain the trust/distrust balance,” he notes, “and she has difficulty reading the trustworthiness of faces, is very impulsive in her decision-making, and is terrible with money. She has normal I.Q., but is often a target for unscrupulous predators who sense her unconditional trust in others and take money from her.”
The Cal Tech neuroscientist Ralph Adolphs examined the woman for other deficiencies and discovered that she is unable to recognize fear in the face of others. When you look at another person’s face, your eyes rapidly dart about, scanning for clues and collecting data about its emotional expression, which, according to the psychologist Paul Ekman and many others, evolved to be universal and thus identifiable. The amgydala-damaged woman, by contrast, stares straight at the face without scanning for details, and thus is unable to make an emotional assessment. During social interactions, most of us maintain a balance between trust (mediated by oxytocin) and fear (mediated by epinephrenie, norephinepherine, and other stress hormones), and we quickly and unconsciously adapt to our environment and the people in it. But when that balance is broken, so too is our trust detector.
Zak’s model fits well with my own that we evolved an innate sense of right and wrong that is expressed through the moral emotions, and that free trade is an integral component to breaking down the normal tribal barriers blocking trust. That being the case, economic transactions can occur with a minimum of top-down interference. A shadow of enforcement—a hint of potential punishment for a norm violation—is all that is needed for most people in most circumstances to grease the wheels of commerce. A law enforcement dummy placed along a stretch of highway, for instance, has been shown to get motorists to slow down, not because they thought it was an actual enforcer but because it reminded them of the law.
Trade makes people more trusting and trustworthy, which makes them more inclined to trade, which increases trust ... creating a self-enforcing cycle of trust, trade, freedom, and prosperity. An additional evolutionary connection between trust and trade can be seen in a two-part experiment on cooperation and cheating conducted by Dan Chiappe and his colleagues at the California State University, Long Beach, in which subjects first categorized individuals as cheaters, cooperators, or neither based on a written description of how they behaved in an exchange involving, for example, borrowing and paying back (or not) money. After reading the description, each subject was then asked to rate how important it was to remember the individuals using a seven-point scale. In the second experiment, participants categorized the individuals on the seven-point scale as before, but then had the opportunity to look at photographs of their faces. They then participated in a face recognition test. The first experiment found that cheaters were rated more important to remember than cooperators—especially when a greater amount of resources was involved—and cooperators were rated more important to remember than those categorized as neutral. The second experiment showed that cheaters were looked at longer and the subjects had a better memory for their faces.
Why should this be? Because from an evolutionary economics perspective, cheaters—like the 2 percent of individuals who routinely break the laws of society—keep cooperators on their moral toes. It is therefore vital to be able to discriminate between the cheaters and the cooperators. As Chiappe explained: “Everything else being equal, knowing that a person cooperated may not tell us much about their character. Knowing that they cheated would be more relevant. This is because cheaters have to give the appearance of being trustworthy and thus they may have to cooperate much of the time.”
Indeed, most people most of the time in most circumstances cooperate, so it takes a lot more mental storage space to focus on and remember details about them (such as their faces) compared to the small handful of cheaters, whose social norm violations stand out as the exception and thus must be recalled for future transactions. And as research in memory has consistently demonstrated, we are more likely to recall unusual events than we are common occurrences. The findings also indicate that we remember cooperators more than neutral people. This too makes sense from an evolutionary perspective, since the memory gradient from cheaters to cooperators to neutrals would be adaptive in our management of social relations in all forms of human exchange, including and especially economic trade.
If this theory is true, there should be neural networks associated with cooperation in an exchange task, and sure enough, the George Mason University neuroeconomist Kevin McCabe and his colleagues, in a fMRI brain scan study of subjects who participated in a “trust and reciprocity” game with either another person or a computer, found that players who were more trusting and cooperative showed more activity in the part of the brain known as “Brodmann area 10,” associated with reading others’ intentions, and more activity in the limbic system, associated with the processing of emotions. Tellingly, such activity was not seen in this brain region of players who either played against computers or engaged in non-cooperative games, which makes perfect sense in a social primate species that evolved moral emotions that drive such pro-social behaviors.
Evolutionary economics suggests that instead of thinking of human culture as being cleaved by different national economic institutions and varying individualized personal values, we should think of it as an evolved human nature that gives rise to a set of core institutions and values which vary in details across different cultures.
As a test of this interpretation, over the past quarter century hundreds of experiments in behavioral economics have been conducted on people from dozens of countries around the world, including fifteen small-scale indigenous tribes. Take the Ultimatum Game, where one player proposes how to divide a sum of money and the second player accepts or rejects the proposal. For example, say you are given $100 to split between yourself and your game partner. Whatever division of the money you propose, if your partner accepts it, you are both richer by that amount. How much should you offer? Why not suggest a $90-$10 split? If your game partner is a rational self-interested money-maximizer—as predicted by the standard economic model of Homo economicus—he isn’t going to turn down a free ten bucks, is he? He is. Research shows that proposals that deviate much beyond a $70-$30 split are usually rejected.
Why? Because they aren’t fair. Says who? Says the moral emotion of “reciprocal altruism,” which evolved over the Paleolithic eons to demand fairness on the part of our potential exchange partners. “I’ll scratch your back if you’ll scratch mine” only works if I know you will respond with something approaching parity. The moral sense of fairness is hardwired into our brains and is an emotion shared by all people and primates tested for it. Thousands of experimental trials have consistently revealed a sense of injustice at low-ball offers. This includes peoples in non-Western cultures around the world, including those living close to how our Paleolithic ancestors lived, and although their responses vary more than modern peoples living in market economies do, they still show a strong aversion to unfairness.
Any theory of economics must begin with a sound theory of human nature. Evolutionary economics redefines the borders of our nature, showing just how driven we are by ancient programs designed for another time and another place. But we also evolved the adaptation of adaptability, and it is here where we see how and why humans behave as we do in such social institutions as markets: We cooperate for the same reason we copulate—because it feels good. On a deeper evolutionary level, the reason cooperating feels good is because it is good for us, individually and as a species. Trust and cooperation leads to a viable free market of exchange, and free markets lead to greater trust and cooperation.
Michael Shermer is the founding publisher of Skeptic magazine, the executive director of the Skeptics Society, a monthly columnist for Scientific American, and an adjunct professor of economics at Claremont Graduate University. His latest book is The Mind of the Market, on evolutionary economics.
