How a Yale research team made history by teaching capuchins to spend money ... and discovered that they're just as smart—and stupid—as your financial advisor.
It’s a little bigger than a quarter and about twice as thick, but because it’s made of aluminum, it weighs roughly the same. It’s flat and smooth, except for what seem to be a few tiny bite marks around the perimeter. To you, it might look like a washer without a hole. To Felix, an alpha male capuchin monkey, and his friends at Yale University, it’s money.
“When one of the monkeys grabs a token, he’s going to hold onto it as though he really values it,” explains Laurie Santos, a psychology professor at Yale. “And the other monkeys might try to take it away from him. Just like they would with a piece of food. Just as you might want to do when you see a person flaunting cash.”
During the past seven years, Santos and Yale economist Keith Chen have conducted a series of cutting-edge experiments in which Felix and seven other monkeys trade these discs for food much like we toss a $20 bill to a cashier at Taco Bell. And in doing so, these monkeys became the first nonhumans to use, well, money.
“It sounds like the setup to a bad joke,” says Chen. “A monkey walks into a room and finds a pile of coins, and he’s got to decide how much he wants to spend on apples, how much on oranges, and how much on pineapples.”
But the remarkable thing about the research isn’t that these monkeys have learned to trade objects for food—after all, a schnauzer can be taught to hand over your slippers in exchange for a Milk-Bone. The amazing part, Chen and Santos discovered, is how closely the economic behavior of these capuchins mimics that of human beings in all its glorious irrationality. Viewed in the context of the daisy chain of near-disastrous human failings that brought the world to the verge of fiscal collapse over the past few years, monkeynomics is eye-opening stuff.
So how much of our wild, dangerous economic behavior is hard-wired, and how much of it is learned? And most important, how much of it can be changed? Watching Felix and friends make financial decisions—some extremely smart, others profoundly dumb—provides groundbreaking insight into the roots of our own dysfunctional relationship with money. And why it all may have started 35 million years ago.
* * * * *
What kind of monkey would Santos be? “A bonobo,” she says with a laugh. “They’re kind of a hippie monkey.” With an infectious smile and curls that cascade down her back, the 35-year-old Santos exudes the cool prof vibe of someone who—all things being equal—would really rather be in a dorm, holding court about the meaning of life. “I’m fascinated by human beings, and monkeys are like humans in their purest form,” she says. She’s quick to offer a funny story about how she decided to pursue primate research after seeing a picture of the lush Caribbean island where the fieldwork was being done. But the truth is that her interest began with the idea that monkeys are like human beings without the cultural baggage.
As a Harvard undergraduate, Santos worked with behavioral scientist Marc Hauser and then signed on to do her dissertation based on research in his lab. Her work centered on basic questions of monkey cognition: How high can monkeys count? (To four.) Do they have a good sense of the practical physics of falling objects? (Not especially.)
This body of work earned her a tenure-track position at Yale, where in 2003 she was charged with setting up the school’s Comparative Cognition Lab. Santos chose capuchin monkeys for practical reasons. They’re smaller and easier to care for than chimps, but they’re almost as smart, resourceful, and social. She got 10 capuchins from noted researcher Frans de Waal at Emory University and planned to continue with the monkey cognition research that she had started at Harvard.
Then one day, one of the caretakers who cleaned the capuchin enclosures in the new lab told Santos that her monkeys were “geniuses.” Felix and friends, he explained with amazement, would hand him their discarded orange peels, trying to trade them for food. Maybe the monkeys were trying to make a point.
Around the time that Santos got the lab up and running, Chen was hired at Yale’s business school. Chen had also worked at Hauser’s lab at Harvard, although not directly with Santos. His dissertation included running game theory scenarios with cotton-top tamarins; he designed experiments to see if the monkeys would employ strategic cooperation to get food rewards and found that they were extremely similar to humans in that regard.
Chen and Santos met in the fall of 2003 at a New Haven student hangout called Koffee and hit it off immediately, recognizing their common interest in tracing the roots of fundamental human behaviors in other primates.
Together, they began brainstorming about what they could do with these “genius” monkeys. They tossed around a host of high-concept ideas, including an elaborate game theory simulation. One of Santos’ grad students constructed a Rube Goldberg–like structure that used stainless steel mechanical arms to divide quantities of food for the classic “ultimatum game,” which measures whether a subject will value fairness over maximal profit. “It was a big, complicated machine with a monkey at one end,” Chen recalls. The idea was ditched after the preposterously strong little capuchins kept casually ripping the machine’s steel arms apart.
And then Santos and Chen settled on something simple and elegant—and provocative.
“On a lark, we started investigating whether or not we could introduce them to a basic market economy,” Chen recalls. “I’m not even sure we had a good idea of how it would work. But if we could, I knew there were a dozen experiments that people in the economics world would be interested in.”
At this point, Chen was already something of a curiosity—the only economist in the world who did research on monkeys. “It’s totally bizarre,” he admits. “But I always worked on what I thought was most interesting.” And what was most interesting was seeing if capuchin monkeys could be taught to spend money.
* * * * *
So in the spring of 2004, after months of constructing the methodology and training the capuchins in the basics of token trading, Santos and Chen began their work. The Monkey Market was open for business.
(Image credit: Laurie Santos)
Physically, the Monkey Market is a smaller enclosure attached to the capuchins’ larger communal home. It’s where the monkeys go to trade for treats. A video of one of these early experiments shows that when Felix, the group’s alpha male, entered, he received a “wallet” with 12 of those round aluminum tokens. Two student researchers, one wearing a pink T-shirt, the other blue, stood on either side of that 3-foot cubic enclosure, each holding a different tray of food. The premise at this stage was pretty basic: Felix could swap his tokens for food with either of the two researchers. He didn’t seem to care much about the students. But he did care profoundly about what the researchers would sell him in exchange for that little metal token.
Felix and the others were cautious, observant shoppers. As the video shows, Felix would head first to the researcher holding out pieces of orange, examining them carefully; before leaving, he stopped to smell them. He went to the other researcher and did exactly the same thing—looking, sniffing, shopping. He then headed back to the first researcher and handed over a token to complete the transaction. Oranges, please.
“When you watch it, it looks like they’re contemplating, thinking about what they’re going to buy,” says Santos. What separates these capuchins from the scores of animals who have been trained to perform complex behaviors in exchange for food is the option presented by that second researcher.
“The critical aspect of money is that it’s fungible. It represents a choice,” explains Chen. “A coin is fundamentally different than, say, pressing a lever.” Santos and Chen had not only achieved their preliminary goal, they had made history: The monkeys were using cash. The capuchins were now operating in a sphere where humans had been dwelling alone.
What next? Although Felix’s intense deliberations were fascinating to watch, they were really beside the point. According to economists, one single factor defines rational behavior in a consumer market: attention to price. Most old-school economics, Chen explains, relies on the bedrock principle that participants in a market will maximize value whenever possible. Could the capuchins become rational consumers?
The researchers began messing with the pricing in Monkey Market. The base currency was still one token for one fruit, but the amount of food and how it was delivered would now vary from day to day. Santos’ researchers began presenting the monkeys with two equally appealing options—one would offer a Jell-O cube, the other an apple slice. Then, like Walmart on Black Friday, they would spontaneously slash the price of the apple slices—two slices for a single token! Act now!—while the price of Jell-O remained the same.
The monkeys, like any smart bargain hunters, flocked to the lower-priced item. Or, in econ-speak, they reacted to a compensated price shift. “That’s the critical hallmark,” says Chen. “When the cost and benefits change, do my decisions change?” When he examined the data, Chen found, to his delight and relief, that they most certainly did. The capuchins had proven not only to be consumers but also rational ones. Quantitatively and qualitatively, their behavior matched that of humans.
And not always in good ways. “One of the things we never saw in the Monkey Market was savings—just like with our own species. They always just spent all their cash at once,” says Santos. “The other thing, amazingly, was spontaneous evidence of larceny. They would rip off the tokens from each other and us at every opportunity.” Clearly the monkeys were screwing up in some of the same ways as people. But how far off track would they go? Santos and Chen decided to think big and introduce some of the same problems into the Monkey Market that have bedeviled centuries of humans.
* * * * *
Up to that point, the monkeys had been adhering to traditional laws of economics that rely on rational behavior. But a relatively new school of economics called prospect theory, led by maverick Nobel Prize–winning economist Daniel Kahneman, was challenging these tenets, positing that human economic behavior is often irrational. “We never thought this kind of behavior was learned,” says Kahneman, 77, who began developing his theories in the 1970s without having taken so much as a single economics course. “It was always clear to me that it’s biological.” But would the monkeys prove or disprove his paradigm-shifting theory? (Kahneman was aware of Santos and Chen’s research, but didn’t participate in it.)
Prospect theory argues that economic decision making is, like Einsteinian physics, relative. The theory contends that humans make economic decisions not in absolute terms, the way a computer might, but relative to some specific reference point—and that causes them to make mistakes. Most of us are risk averse; we’ll do almost anything to avoid a loss. And we treat losses very differently than gains. It’s why investors defy logic by selling off the winners in their portfolio instead of dumping the losers. And why homeowners in a housing slump will let their banks foreclose before they drop the price of their houses.
“We were already seeing deliberative decision making in our monkeys that went beyond what scientists had seen in animals before,” Chen explains. “So we just thought, Why not raise the stakes? Why don’t we investigate whether they’ll make the same mistakes that humans make?”
Simply put: Were the monkeys smart enough to act dumb?
(Image credit: Laurie Santos)
Armed with cutting-edge economic theory, a handful of tokens, and a bin full of fruit, Santos and Chen introduced the concept of risk to the Monkey Market. In a series of three interrelated experiments designed carefully to mirror economic models, the monkeys chose between risky sellers and safe sellers. The first scenario represented a simple choice for the monkeys: Seller A would consistently deliver one piece of apple; Seller B would sometimes deliver one, and sometimes add one and deliver two. Seller B represented a no-brainer gamble, or what economists call stochastic dominance.
And the monkeys immediately grasped the significance of the scenario. They chose Seller B 87 percent of the time.
The second experiment presented a bigger challenge: Seller A would show the monkeys only one piece of apple, but add an extra piece half the time. Seller B, on the other hand, would show the monkeys two apple pieces, but half the time would hand one over and take one back.
Despite the fact that they were conditioned to trade with Seller B from the first experiment, the monkeys quickly reversed course and showed a strong 71 percent preference for Seller A. The data suggested that the two scenarios felt very different to the monkeys, just as they might to a human. But do the math: Each seller represented a 50/50 chance of ending up with two apple pieces. A computer would value each of the sellers equally. And yet the monkeys greatly preferred dealing with generous Seller A, who sometimes added a piece of apple, than stingy Seller B, who sometimes took an apple away. Fear of loss dictated their thinking. Their decision making wasn’t absolute; it was relative.
In the third experiment, the researchers reversed the options, changing from a bonus scenario to a loss scenario.
Seller A would show one apple piece and hand it over, while risky Seller B would show two but always take away one and deliver one. Despite the fact that both sellers gave the same payout—one apple piece—the monkeys strongly preferred Seller A.
Santos and Chen had hit a home run. When taken together, the results of the second and third experiments suggest that capuchins show an overwhelming loss aversion. Just like us.
Chen explains that the data set for the monkeys—which revealed a 2.7 to 1 risk preference in the loss model compared to the bonus model—was completely indistinguishable from what you might find in a trial using human subjects. “It’s a little spooky,” says Venkat Lakshminarayanan, a grad student in the lab.
“Sometimes I’d look at the numbers and forget that they’re monkeys,” Chen adds.
In the fall of 2008, when the housing bubble burst, and some of the world’s biggest financial institutions went straight to hell, Santos and Chen turned again to the monkeys. There were more tests of prospect theory risk behavior, and more confirmation of the evolutionary underpinnings behind the crazy—and yes, wildly irrational—behavior that led to the current recession.
Does this kinship between the capuchins and us have a limit? Chen and Santos seem to have found it. In humans, knowing the price of a costly item makes it more desirable—call it the Château Lafite Effect. Not so for the monkeys. A yet-to-be published study from 2010 showed that, for Felix and friends, raising the price did nothing to boost the appeal of a particular type of food. Finding the end as well of the beginnings of our kinship with the capuchins not only validated the group’s research, it placed a bookend on a groundbreaking body of work.
* * * * *
So what did Santos and Chen really learn after seven years of intense study? “Whatever mechanism in the brain that’s driving these biases is one and the same in capuchin monkeys and in us,” says Santos. “That means these strategies are 35 million years old.”
Moreover, the work with the Monkey Market has helped bolster a growing trend toward viewing economics as a more complex and nuanced science—one in which emotion plays as big a part as cold, hard logic. “The losers are going to fight harder than the potential gainers are,” explains Kahneman. “That asymmetry is really, really strong. It’s why there’s inertia against change. And reducing misery is more important than increasing happiness.”
Some economists have begun to create real-world scenarios that take our innate biases into account. Chen cites the Save More Tomorrow program devised by University of Chicago economist Richard Thaler, in which the defaults for a 401(k) plan at a midsize firm were adjusted in accordance with prospect theory to maximize savings. “They’re framing savings not as a loss of income but as a smaller gain,” says Chen. The results were impressive: Employees enrolled in the plan tripled their savings rate from 3.5 percent to 11.6 percent in just two years.
And, even as the architect of work that shows how inherently flawed (even stupid) humans are when it comes to all things monetary, the ever-optimistic Santos still sees a positive side.
“The problem of modern economics is that it really does assume that we’re homo economicus,” she says. “And we’re not. We make mistakes. So there’s going to be a disconnect when we set up structures that assume we’re going to behave rationally, and we know that we won’t.” She pauses, collecting her thoughts on the couch in her sunny Yale office, which has a “Beware of Monkeys” sign on the wall. “That’s really the message of the work. We’re not doomed. We’re even smarter than the monkeys. We just have to admit that we’re not perfectly rational.”
Bonus: Frans de Waal's Capuchin Fairness Experiment
The article above, written by Allen St. John, is reprinted with permission from the September-October 2011 issue of mental_floss magazine. Get a subscription to mental_floss and never miss an issue! Be sure to visit mental_floss' website and blog for more fun stuff!