The job was simple. A brown capuchin monkey at the Yerkes primate center in Atlanta was handed a small granite rock, and had 60 seconds to put it back into the open palm of the experimenter, Sarah Brosnan. Then she got food. The five females in the study had at least two years of practice at this kind of trade.
The pay was usually a quarter of a cucumber slice. Offered a choice, every monkey picked a grape over cucumber at least 90% of the time. But on its own, cucumber was fine. When two monkeys sat side by side behind a mesh partition and both were paid in cucumber, Emory’s report on the study said, they completed 95% of their trades.
Then the partner started getting grapes for the same rock.
Completed trades fell to 60%, according to Emory’s report. When the partner was simply handed a grape without trading at all, they fell to 20%. The paper describes refusals that were often active: monkeys tossed the rock, or the cucumber, out of the test chamber. They were giving up food they would eat under almost any other circumstances.
Brosnan and Frans de Waal published the result in Nature in September 2003, under the title “Monkeys reject unequal pay.” Years later, in a TEDx talk, de Waal played a video of a monkey tasting her first piece of cucumber, watching her neighbour get a grape, and then testing her rock against the wall. “So this is basically the Wall Street protest that you see here,” he told the audience.
The paper had a fourth condition, though. In the “food control,” no partner was present. The monkey watched a grape being placed where the partner usually sat, then traded for cucumber. Completed trades: 55%, by Emory’s count, not far from the 60% with a partner eating grapes.
In 2004 the psychologist Clive Wynne argued in Nature that this was the simpler explanation: the monkeys were refusing cucumber because they could see something better was available, with or without anyone to compare themselves to. Brosnan and de Waal replied that refusals rose over a session when a partner ate the grapes and fell when a grape merely sat there. Other labs pushed further. A 2006 study of eight capuchins, in four pairs, found that monkeys watching a partner get grapes rarely rejected cucumber, while monkeys that had themselves just had grapes rejected it more; the authors proposed frustration as the explanation. Brosnan and de Waal answered that the study had no trading task, so it was not a replication. A team testing great apes concluded that seeing another get better food raised the apes’ own expectations, and called that more plausible than inequity aversion. A 2009 capuchin test that kept each monkey in one role found neither effect.
De Waal’s talk was recorded in November 2011, with monkeys new to the task. As of October 2026, TED’s site counts more than 5.9 million views. The clip shows a grape going to a partner. In the 2003 paper, a grape placed in an empty seat drew nearly as many refusals.