People may be more honest than psychologists thought

“IN THE state of nature, profit is the measure of right,” wrote Thomas Hobbes, a philosopher with a dim view of human nature. Hobbes’s fellow-thinkers have spent centuries pondering whether humans tend to be self-serving or are more inclined to straight-dealing. (Obviously, people exhibit both kinds of behaviour. The question is which comes most easily.)

So far, most experiments have tended to favour the first idea—that humans are dishonest by default when it serves their self-interest. In one study led by Shaul Shalvi at the University of Amsterdam, for example, participants were told to roll a die secretly three times and write down the results of the first roll. They would then receive 10 times that number in Israeli New Shekels. The researchers found that people who were asked to report their die roll within 20 seconds tended to report higher numbers than those who were given no time limit (though both groups reported higher numbers, on average, than would be expected if they were being truthful).

Now, though, the waters have been muddied by a new study published on arXiv, an online preprint site, by Valerio Capraro of the University of Middlesex. Dr Capraro argues that previous studies have allowed people to ponder in advance how they can best maximise their gains. That means such studies did not properly test how the participants respond under pressure. His study, which presents participants with details of their task just before they perform it, finds that people may be naturally truthful after all.

Dr Capraro recruited his volunteers from Amazon’s Mechanical Turk, an online marketplace where people can be employed to do small jobs. Dr Capraro’s volunteers were either “senders” or “receivers” in each experiment. In the first experiment, 312 senders were told they would be randomly chosen to be in either group 1 or group 2. They then had the choice of telling a receiver which of the two groups they belong to truthfully (in which case they each got $0.10) or untruthfully, in which case the sender got $0.20 and the receiver got just $0.09. (The receivers were not told the rules of the game, so could not know whether the senders lied or not).

In a second “control” experiment, senders were not assigned to groups. That meant they did not have to decide whether or not to lie. Instead, they were simply allowed to choose between giving the receiver $0.10, and getting the same amount themselves (the altruistic option), or sending $0.09 and pocketing $0.20 (the selfish option). In both experiments, about half the senders were told to make their decision in under 5 seconds and the other half asked to think for at least 30 seconds before choosing between the options.

In the control experiment, where lying was not an option, around 25% of 372 participants ignored their self-interest and plumped for the altruistic option, whether or not they had to decide quickly. That established a baseline level of altruism against which to compare the results from the first experiment, in which lying was the route to a bigger payoff. Under those conditions, 56% of participants told the truth when under time pressure, compared with only 44% of those who had time to think. Dr Capraro’s experiments, then, suggest honesty is the more instinctive response. Deliberation, on the other hand, seems to promote more selfish behaviour.

To help ensure his results are robust, Dr Capraro plans to repeat his experiment in a laboratory, under more tightly controlled conditions. If his results hold up, they would suggest a more optimistic view of human behaviour than Hobbes would have dared hope for.