She felt there were two possible explanations. One possibility is that our species’ intelligence leads us to build highly complex environments, such as financial markets, that eventually go far beyond our ability to understand them. So, it is not surprising that occasionally it all comes crashing down around our heads.

Alternatively, such errors may not stem from flaws in the design of our environments, but rather within ourselves. In recent years, social scientists have learned that when placed in certain social contexts people tend to make mistakes that are predictable and repeatable. Worse, people seem immune to learning from evidence about the inherent mistakes in those decisions.

To find out if such errors are embedded in our DNA or a product of our complex environments, Santos, who runs the Comparative Cognition Laboratory (CapLab) at Yale University, conducted a series of experiments beginning in 2006 with capuchin monkeys. Intelligent and inquisitive, capuchins are found in Central and South America. Humans last shared a common ancestor with this primate twig of the evolutionary branch some 35 million years ago. Santos wanted to see if you placed the capuchins in the same context as humans, would they make the same financial mistakes?

TEAMING UP WITH economist Keith Chen and graduate student Venkat Lakshminarayanan, her team first taught the monkeys to use metal discs as hard currency to purchase treats such as apple slices or gelatine cubes from humans in the lab. The capuchins quickly became adept at handling and trading the tokens. It was then the team introduced a monkey market.

The capuchins could retreat from their social enclosure into a smaller area, where they could trade tokens for goods with humans. Two distinct salesmen were present: one consistently offered one grape per token, the other, two grapes per token. Not surprisingly, the monkeys showed a preference for the more generous storekeeper.

In variations of the experiment, they also selected salesmen that had better food, paid close attention when there were sales and generally behaved in every manner expected of astute consumers interested in getting full value for their monkey dollars.

“We were surprised at not only how quickly the monkeys become good at swapping tokens, but at how closely they obeyed market principles, such as price theory,” Santos recalls. “In fact, when we looked at the monkey data using economic tools, they basically matched up qualitatively and quantitatively with human data in a real market. You couldn’t tell them apart.”

The critical question was, would monkeys mess up in the same way humans do when confronted by similar financial challenges? In a series of ingenious preference experiments that involved salesmen adding to or subtracting from the amount of food on offer for each token, the team tested if the monkeys’ decisions demonstrated loss aversion and reference dependence.

In one experiment, the salesmen offered the same average payoff of one-and-a-half apple pieces, but in different ways. One salesman began every trade showing two apple pieces. When he was paid, he either delivered the two pieces or removed one, delivering less than the monkeys expected. The other salesman started with one piece, but when paid, either delivered one or added a second piece. The monkeys strongly preferred the second scenario and disliked the first, where one piece of apple was taken away – even though the average payoff was the same.

Throughout the various experiments, Santos and her colleagues showed that the capuchins represented their payoffs relative to arbitrary reference points and appeared to avoid gambles framed as losses. The researchers then went on to examine whether capuchins demonstrate an endowment effect (Thaler, 1980), a phenomenon in which ownership increases an object’s value. Again, the test showed that the monkeys demonstrated a true effect (see Santos and Chen, 2006).

“The endowment effect was less surprising. We had already shown that the capuchins were rational in certain aspects of their markets in the same way that humans are. In some sense it was comforting that the same spots where humans go wrong when making financial decisions may be shared with these monkeys,” Santos says.

WHAT IS NOT quite SO COMFORTING are the implications. The biases that cause such irrational behavior may be biologically ancient, stretching back 35 million years or more, predating the development of money.

“It is a little bit worrying in terms of thinking about how to develop policies that will let us override these biases,” explains Santos. “When natural selection builds in such survival strategies, they tend to be hard to overcome, such as our fear of snakes.”

What is positive is that we are a smart enough species to work out how we want things to run and we can then set up the systems and structures to help us do it. But, she says, first we have to realize our natural failings to overcome them.

“We overcome our limitations in many other fields, but the legacy of classic economics notions like utility maximization and rationality have really hurt us in this respect. It assumes we are all ‘Homo economicus,’ but until we get over this Panglossian view we will not introduce the approaches to help us overcome these biases.”