Monkey inequality

Over at Wired, Jonah Lehrer has written a post in which he looks at a couple of lines of evidence about the innate response of humans to inequality. The first line, based on brain scans, is nicely discussed by Jeff at Cheap Talk.

The second involves an experiment with capuchin monkeys:

A similar lesson emerges from a classic experiment conducted by Franz de Waals and Sarah Brosnan. The primatologists trained brown capuchin monkeys to give them pebbles in exchange for cucumbers. Almost overnight, a capuchin economy developed, with hungry monkeys harvesting small stones. But the marketplace was disrupted when the scientists got mischievous: instead of giving every monkey a cucumber in exchange for pebbles, they started giving some monkeys a tasty grape instead. (Monkeys prefer grapes to cucumbers.) After witnessing this injustice, the monkeys earning cucumbers went on strike. Some started throwing their cucumbers at the scientists; the vast majority just stopped collecting pebbles. The capuchin economy ground to a halt. The monkeys were willing to forfeit cheap food simply to register their anger at the arbitrary pay scale.

This labor unrest among monkeys illuminates our innate sense of fairness. It’s not that the primates demanded equality — some capuchins collected many more pebbles than others, and that never created a problem — it’s that they couldn’t stand when the inequality was a result of injustice.

While this experiment shows that problems emerge following arbitrary outcomes, I would not describe the reaction as demonstrating an “innate sense of fairness”. The revolt was one-sided, and the monkeys who received grapes were not handing them back. It is more like an innate dislike of being on the bottom.

More importantly, I am not convinced that the experiment fully tested whether fair outcomes could generate unrest. As the stone gathering ability of monkeys is not likely to vary much, particularly when compared to the variation in human ability across different professions, the variation in opportunity for the monkeys was negligible. The variation in outcomes was also relatively small.

As a result, if a monkey was hungry, it is an easy task to get some stones. However, if a human needs money, the range of possibly courses of action are large, with the returns widely variable. A cleaner and a brain surgeon earn significantly different rates of return for an hour of effort.

What if the monkeys were paid on the basis that whoever collects the most stones gets the vast majority of the cucumbers or grapes? Or what if effort started to pay compound dividends, meaning that long-term focussed stone gathering delivered much higher returns. The rules might be clear and the victory fair, but I am not sure that the reaction to the significantly different outcomes would be benign.

Lehrer uses this experiment to draw the following conclusion:

When the rich do something to deserve their riches, nobody complains; that’s just the meritocracy at work. But when those at the bottom don’t understand the unequal distribution of wealth — when it seems as if the winners are getting rewarded for no reason — they get furious. They doubt the integrity of the system and become more sensitive to perceived inequities. They start camping out in parks. They reject the very premise of the game.

Even if the belief that nobody complains when the rich deserve their riches was true, do the majority of people understand the basis of the distribution of wealth in today’s economy? Has the economy reached a level of complexity that will always generate a certain level of distrust by those at the bottom?


  1. From Debruil’s work linked above:

    “I examine five motivations that we must take into account when explaining punishment and the response to punishment:

    1. Cooperation: Players might punish to make the provision of public goods more advantageous and free riding more costly.

    2. Envy: Players might punish because they are discontented with having less than others.

    3. Fairness: Players might punish others for having violated norms of fairness. If they considered themselves personally wronged, punishment can then be conceived as a form of revenge.

    4. Equality: Players might punish to promote an more equal outcome or because they have a negative attitude toward players who have more than others.

    5. Unexpected unfairness: Players might punish because others have been unexpectedly unfair.

    In each case, the “might” refers to motivations that have actually been measured/observed in certain contexts (game theory and primatology in particular). Each motivation may occur to varying degrees situationally. In other words, it’s generally not accurate to conduct these discussion in dichotomous/either-or terms.

    1. I agree with your either-or point – and part of the issue is semantics around the word “fair”.

      I just borrowed an electronic version of Debruil’s book from the library and am about to spend two weeks on an island off Sumatra in which I can read it. I’ll post some thoughts on it afterwards.

  2. I remember reading somewhere that another experimented following the same sort of outline was conducted on chimpanzees where in exchange for performing some trick the (two) chimps got a slice of cucumber. After this was established the scientists than gave one chimp a slice of cucumber for doing their trick and the other chimp a grape for doing their trick. The response was the chimp receiving the cucumber flipped out and refused to perform any more tricks and the chimp that did receive the grape also refused to perform any more tricks until the other chimp also got a grape.

    1. And interestingly enough the place where I read about this chimp experiment also compared the results against the same experiment being performed on capuchin monkeys. The result for the monkeys was the same you listed above. The monkey receiving the cucumber refused to do any more work with the monkey receiving the grape not caring whether the other monkey was rewarded ‘equally’.

      The takeaway message I took from the comparison of these two experiments was that as intelligence/complexity of social system increased in a primate species, the more human like their moral system/sense of fairness, seemed to become/appear.

      1. Found it. It was in an interview with Frans de Waal.

        Johnson: Today we are faced with what has widely been termed a “culture of corruption.” In your latest book you point to the abuses on Wall Street in which financiers have willfully defrauded the public and Washington politicians who operate through a revolving door of political favors and corporate kickbacks. Is there something in this research with our evolutionary relatives that can help us change our political culture? For example, you and your colleague Sarah Brosnan discovered something very interesting in your study with chimpanzees and capuchin monkeys concerning economic behavior.

        De Waal: Yes, the first experiment was with capuchin monkeys where we would put two monkeys side by side and we would give them rewards for a very simple task. If you give them the same reward, such as small pieces of cucumber, they’re perfectly happy to do this many times in a row. But if you give one of the two monkeys a grape and the second a cucumber then the second monkey gets mad and refuses to perform the task. We have repeated this with chimpanzees where Sarah found that the one who gets more is also affected and refuses the task unless the other one also gets a grape. With this we’re getting very close to the sense of fairness.

        Johnson: How does this translate to modern human society?

        De Waal: I think the sense of fairness in humans is very strongly developed and that’s why we react so strongly to all the bonuses received by Wall Street executives. We want to know why they deserve these benefits. The anger we have towards Wall Street is probably a very old primate reaction that relates to cooperation. If you are a cooperative animal you need to watch what you get. If you, or even a whole community, invest in something but then a few individuals receive a much larger return, it’s not a good arrangement. If it happens consistently, it’s time to look for an arrangement that is more beneficial. That’s why we’re so sensitive to how rewards are being divided.

        Another blog article about it:

        The actual study itself:

      2. That’s a lot closer to my concept of fairness that in the capuchin monkey case – perhaps Lehrer should have run with that study.

Comments welcome

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s