A curious thing happens when people are asked to divide a sum of money between them: What they think is a fair division depends on factors that economists assume shouldn't matter at all.
A cornerstone of economic theory is that rational agents are self-interested, that is, they seek to maximize benefits to themselves. Yet more than a decade of research in experimental economics, experimental psychology, and anthropology has challenged the legitimacy of the “rational agent” model as a descriptive model of human behavior.
People do not always seek outcomes that maximize benefits to themselves. Instead, our choices appear to be motivated by concerns about fairness. But even more striking is the discovery that what counts as fair in an economic transaction depends on how we think we measure up to the person on the other side of the transaction.
The most dramatic demonstrations of the status effect come from studies based on Dictator and Ultimatum games. In the Dictator game, one party is given full decision-making authority to keep or share a sum of money. In the Ultimatum game, one party is given authority to propose how the money should be divided (proposer) and the other party (responder) must decide to accept or reject the proposed split. If the proposal is rejected, neither party gets any money.
Rational agents should always choose the selfish choice in each game. In Dictator, they should keep all the money for themselves. In Ultimatum, proposers should offer the minimum amount possible, and responders should accept whatever is offered (because something is better than nothing.) But several decades of research on these games show that people do not behave as "rational" agents. Instead, the modal response in Dictator is to freely give away about 15–35 percent of the resource to partners; in Ultimatum, proposers typically offer 40–50 percent, and recipients routinely reject offers of less than 20 percent.
In order to accommodate these results, theories have been proposed in which individuals are assumed to enter economic transactions with standards of fairness, and their decisions are guided by distance from the expected normative fair division.
But what counts as fair?
One would think that a fair division between two people should always be a 50-50 split. But that turns out not to be the case. Surprisingly, what counts as fair depends on many factors, the most influential of which is perceived relative status. The needle on the fairness dial moves closer to eith exploitation or generosity depending on where you and your transaction partner stand in the proverbial food chain. It also matters a lot how people believe they ended up on a particular rung of the ladder.
In studies where status is arbitrarily conferred (e.g., coin toss, experimenter assignment), people tend to behave fairly toward others (as described above). But in studies where people competed in contests as simple as trivia tests, top performers behaved exploitatively toward lower ranking individuals. They offered people lower down the food chain less when they were in charge of splitting the money, and demanded more when they were responders. Even more surprising, lower ranking individuals were willing to accept less and to offer more to higher-ranking performers.
In other words, both sides believed winners were superior and therefore deserved more in a subsequent transaction.
On the other side of the continuum are studies that report "irrational" generosity. When status reflects prestige (the ability to confer benefits through greater skill or expertise), people behave generously toward those perceived as lower ranking, and higher status individuals are held to higher standards by lower status individuals, a phenomenon referred to as noblesse oblige (e.g., Fiddick & Cummins, 2007; Fiddick, Cummins, Janicki, Lee, et al., 2013).
For example, my colleagues and I investigated noblesse oblige in a seven-country, cross-cultural study that involved asking people how willing they were to continue a simple carpool arrangement: I'll drive if you pay for the gas. They were given ledgers that showed carpooling and payment history. Some partners were paid 100 percent of the time. Others were less reliable, paying 75 percent, 50 percent, or even as little as 25 percent of the time. The catch was that participants were asked to adopt the perspective of an employer driving an employee or vice versa.
We found that those who adopted a boss perspective were far more willing to continue the arrangement despite significant non-compliance on the part of their employee car pool partner, They were also more likely to feel they had been treated fairly even when the employee didn't keep up their end of the bargain all the time, felt less animosity towards their cheating partners, and believed they got the better deal because they felt they bore less cost and received higher value from the arrangement.
You might think that this noblesse oblige could be attributed to the assumption that the boss made more than the employee. But we found the same pattern even when the employee was described as making more than the boss due to sales bonuses.
Another way to put this is that prestige status has social utility (confers satisfaction), which shifts the balance of cost/benefit ratios.
It is important to note that the same results were observed despite marked differences in social values among the countries studied. Countries like Japan and Singapore which score high on power distance, a measure of hierarchical social organization, (Japan and Singapore) were as likely to exhibit noblesse oblige as those that score low on this measure of social hierarchy (Australia, Canada, Germany, U.K., and U.S.).
Is this yet another example of irrational and flawed human decision-making? Perhaps not. The "irrationality" with which people make economic decisions may instead imply that we enter strategic economic games with a wider and longer view than can be accommodated by rational self-interest. From this vantage point, rational self-interest appears myopic, ignoring the possible long term repercussions of ignoring fairness.
For more information about me,