Endowment effect in chimpanzees can be turned on and off: study

July 5, 2012, Vanderbilt University
This graphic shows the percentage of chimpanzees showing endowment effect in different situations. Credit: Owen Jones, Vanderbilt University

Groundbreaking new research in the field of "evolutionary analysis in law" not only provides additional evidence that chimpanzees share the controversial human psychological trait known as the endowment effect – which in humans has implications for law – but also shows the effect can be turned on or off for single objects, depending on their immediate situational usefulness.

In humans, the endowment effect causes people to consider an item they have just come to possess as higher in value than the maximum price they would have paid to acquire it just a moment before. Economists and lawyers typically assume this will not be the case. And some consider the endowment effect a human-centered fluke, subject to widespread and seemingly unpredictable variation. The origins of the quirk, and satisfying explanations for how it varies, have proved elusive.

In this new research, Vanderbilt University professor Owen Jones, who is one of the nation's few professors of both law and biology, and evolutionary biologist Sarah Brosnan of Georgia State University, developed and tested predictions rooted in evolutionary theory about when and how the endowment effect would appear.

Drawing on prior theoretical work by Jones on the evolution of seemingly "irrational" behaviors, Brosnan and Jones found consistency between and humans with respect to the existence of the endowment effect and, more importantly, showed that variations in the prevalence of the effect could be predicted.

"These results strongly support the idea that the endowment effect reflects the deep influence of evolutionary processes – notably natural selection – on psychological predispositions that affect how brains ascribe value to objects. The endowment effect, and several other seemingly quirky psychological predispositions, may trace to the sharp differences between modern environments and ancestral ones, in which these predispositions may once have minimized losses and maximized gains," said Jones, the New York Alumni Chancellor's Chair in Law, professor of biological sciences and director of the MacArthur Foundation Research Network on Law and Neuroscience.


In humans, the endowment effect can be demonstrated in experiments when preferences between two equally available objects are separately compared to the choices people make when they own one of the objects and have the opportunity to trade it for the other. Typically, far fewer trades take place than economists would expect, given the preferences expressed under the choice condition.

In a prior study Jones, Brosnan and colleagues predicted and found that chimpanzees exhibited an endowment effect; that the effect varied between classes of objects; and that trades among evolutionarily significant objects, like foods, elicited stronger endowment effects than trades among less evolutionarily significant objects, such as toys. That is, chimpanzees were far more likely to keep their less preferred food when they could trade it for a more preferred food than they were to keep their less preferred toy when they could trade it for their more preferred toy.

In the present study, Jones, Brosnan, and colleagues made an even narrower prediction. They predicted that chimpanzee behavior would change for the same non-food item, in this case a tool, depending on whether that tool could be used to immediately obtain food.

"If there is an evolutionary basis to this effect, one would expect manifestations of the effect to vary as a function of immediate usefulness of the object," the authors wrote. "A specific prediction emerges from this hypothesis: the effect should vary as a function of the instrumental value or usefulness of the object."

In this study, the chimpanzees were given a choice of two tools. Each tool could be used to obtain one of two specific foods which could not be obtained without a tool. The tool pairs were tested in three situations:

  • Foods absent and unobtainable
  • Foods visible but unobtainable
  • Foods visible and obtainable
Jones and Brosnan tested the chimpanzees to see how many showed an endowment effect by keeping their less preferred tool when they could have traded for the tool they preferred more in an earlier test.

The results were striking. As the graph above shows, the availability of food generated an instantaneous and dramatic endowment effect in the tools that acquired the foods. The endowment effect turns on powerfully when the tools can actually be used. The endowment effect for food-acquiring tools was non-existent when food was either absent or present but out-of-reach.

Specifically, as soon as chimpanzees could access a food many of them suddenly began keeping whichever tool they were endowed with, regardless of whether they could exchange if for their more preferred tool. Conversely, when the food was either absent or present and unobtainable, there was no endowment effect, and chimpanzees traded tools freely.


This study, like Jones and Brosnan's prior study of chimpanzees, provides further evidence that humans are not the only species to show biases of this sort and that the endowment effect likely evolved prior to the human split from other species.

"This has two implications," said Jones. "First it is clear that we can learn about the development and function of human biases by studying other species. Second, the presence of patterned and predictable endowment effects in close relatives suggests that the effect is not just a quirk, but is likely a robust and widespread feature that evolved in primates to minimize the risks of loss under conditions of uncertainty. This in turn suggests that many specific and patterned cognitive quirks already identified in humans may be connected, in a previously hidden way that evolutionary perspectives can illuminate."

This research can also provide a deeper understanding of where some law-relevant behavior comes from and what factors affect it in predictable ways.

"This could expose previously hidden patterns in the human decision-making architecture—not only with respect to the endowment effect, but also with respect to the entire suite of human biases," said Jones. "And that, in turn, could help the legal system to predict, anticipate, and respond to human behaviors that pose problems for law."

Explore further: Why don't chimpanzees like to barter commodities?

More information: The study, titled "Evolution and the expression of biases: situational value changes the endowment effect in chimpanzees" can be found in the current (March) issue of the journal Evolution and Human Behavior.

Related Stories

Why don't chimpanzees like to barter commodities?

January 30, 2008

For thousands of years, human beings have relied on commodity barter as an essential aspect of their lives. It is the behavior that allows specialized professions, as one individual gives up some of what he has reaped to ...

Chimpanzees use sex tools

May 5, 2010

(PhysOrg.com) -- Many animals are known to use tools, but chimpanzees (our closest living relatives) show the most varied and complex use of tools, and the males in one group of chimps have even been observed using sex tools ...

Exploring reactions to inequality

August 18, 2009

When primates don’t get the same rewards as their peers, they often refuse them. A Georgia State University researcher is exploring why this reaction happens, and how reactions to inequality have evolved in related species, ...

Recommended for you

Packing a genome, step-by-step

January 18, 2018

Genome folding now has a playbook. A new step-by-step account spells out in minute-time resolution how cells rapidly pack long tangles of chromosomes into the tiny, tightly wound bundles needed for cell division. Cells reel ...

First look at pupil size in sleeping mice yields surprises

January 18, 2018

When people are awake, their pupils regularly change in size. Those changes are meaningful, reflecting shifting attention or vigilance, for example. Now, researchers reporting in Current Biology on January 18 have found in ...


Adjust slider to filter visible comments by rank

Display comments: newest first

not rated yet Jul 06, 2012
They're looking at the issue from the wrong perspective.

The real value of the object is what I consider it to be when it is in my possession. The value that I'm willing to pay for the object when I don't have it is less than its real value to me, because if I paid the real value then I wouldn't gain anything, assuming that there's no risk in the transaction.

If there is a risk, then paying the real value is irrational because you end up losing value in the long term. Or in other words: "a bird in the hand is worth two in the bush"

not rated yet Jul 06, 2012
Or, in other words, if the chimp has a tool that can obtain food for it, there's no reason to swap it for another tool that can obtain food just as well - even if there is no risk or cost of effort involved in the transaction, because it makes no difference. If there's even a tiny cost or risk of losing the tool, then swapping makes no sense.

When the food is not present, the tools take on a different meaning and are traded for some other quality, but when the food is brought back you keep what you have. It's like a game of musical chairs. No chair is worse than the other, but you still stick to yours.

If anything, the test just shows the lack of foresight in the chimps because they only "get" that the tool is for food in the acute situation where they can obtain food with the tool, and seem to forget the whole thing when the food is not there.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.