October 29, 2013
Penn: ‘Endowment Effect’ Not Present In Hunter-Gatherer Societies
Centuries of economic theory have been based on one simple premise: when given a choice between two items, people make the rational decision and select the one they value more. But as with many simple premises, this one has a flaw in that it is demonstrably untrue.
The fields of psychology and behavioral economics have experimentally identified a laundry list of common biases that cause people to act against their own apparent interests. One of these biases — the mere fact of possessing something raises its value to its owner — is known as the “endowment effect.”A new interdisciplinary study from the University of Pennsylvania has delved into whether this bias is truly universal, and whether it might have been present in humanity’s evolutionary past.
The study was led by Coren Apicella, an assistant professor in Penn’s School of Arts and Sciences’ Department of Psychology, and Eduardo Azevedo, an assistant professor in Wharton’s Department of Business Economics and Public Policy. They collaborated with Yale’s Nicholas Christakis and the University of California, San Diego’s James Fowler.
It will be published in the American Economic Review.
A classic endowment effect experiments involves giving participants one of two items, such as a chocolate bar and a mug, and then asking whether they would like to trade for the other. As the starting item is selected at random, there should be a 50 percent chance that participants initially receive the item they like best and thus a 50 percent chance that they will trade.
“What we see, however, is that people trade only about 10 percent of the time,” Azevedo said. “Simply telling someone they own something makes them value it more. That is, the way you ask the question changes what item people prefer, unlike what you would expect from rational economic behavior.”
One problem with these experiments is that they generally involve participants from so-called “WEIRD” — western, educated, industrialized, rich and democratic — societies. Apicella drew on her decade-long study of the Hadza people of Tanzania to provide a new perspective. The Hadza are one of the last hunter-gatherer societies on Earth, living in small, nomadic camps that communally share nearly all their possessions.
“We wanted to examine whether the endowment effect was something that occurs in non-WEIRD societies, since they represent the vast majority of human populations that have ever existed,” Apicella said. “Even if it’s not a perfect window into our past, it’s at least a different perspective than what you get when you study your average college student. The fact that the Hadza remain relatively isolated from Western culture, media and ideals makes them a good group with which to investigate the history and universality of biases like the endowment effect.”
The history of the endowment effect is of particular interest to evolutionary psychologists, as experiments to test its presence in non-human primates, such as orangutans, chimpanzees and gorillas, has been met with mixed results. That some non-human primates exhibit the bias could mean that it was present in the last common ancestor between them and humans, but it could also mean that they learned the behavior by participating in other reward-based studies.
The area of North Tanzania where the Hadza live provided a natural way to further investigate the role of culture in transmitting this bias, as a large lake separates some, but not all, of the camps from a nearby village. People living in the camps on the near side of the lake have much more frequent interactions with tourists and commerce, often buying items from stores in the village, or selling bows and arrows to visitors.
The researchers conducted versions of the endowment effect experiment in several different camps, and compared the results.
In order to avoid bias from items that might be more or less valuable in the environment the Hadza live in, the researchers constructed the experiments so that participants chose between items that had only cosmetic differences. Participants would be given either a package of cookies, with the option to trade it for a different flavor, or given a lighter, with the option to trade it for a different color. They also ensured that the participant knew that the variety he or she received at the start was a random choice, and varied whether the participant got to physically hold the item before given the option to trade it for another variety.
“We wanted to use both food and tools, as experiments with non-human primates show an endowment effect for the former but not the latter,” Apicella said. “However, we saw that it didn’t make a difference whether a person was choosing between cookies or lighters. The difference-maker was their relative level of isolation from modern life.”
“The more isolated Hadza traded about 50 percent of the time — which is what rational people should do,” Azevedo said. People near the village traded about 25 percent of the time, which is much closer to the 10 percent we see with Western students.”
“To make sure this wasn’t a case of the more capitalistic people moving closer to the village, we also asked the people about their social networks,” Azevedo said. “The percentage of people who named someone in a distant camp was very small. Quantitatively, it seems impossible that the difference in endowment effect between two camps could be explained by migration.”
With that potential caveat accounted for, one explanation for the apparent lack of an endowment effect in the more isolated camps is that the bias is a learned behavior that comes with exposure to capitalistic societies. However, an alternative explanation could be that both groups experience the effect, but it is suppressed in the more communal groups by social pressures.
“We need to study this further to see which explanation holds,” Apicella said. “Either way the results suggest that these isolated hunter-gatherers are more rational than the average western consumer when it comes to economic decisions.”
On the Net: