People allowed substantially more inequality between workers when unequal rewards reflected performance rather than luck, an experiment involving more than 65,000 people across 60 countries found. The result helps explain why two people can see the same income gap and disagree about whether it should be reduced.
The study, Fairness Across the World, was published online as an accepted manuscript in The Quarterly Journal of Economics on September 7. Its authors are Ingvild Almås, Alexander W. Cappelen, Erik Ø. Sørensen and Bertil Tungodden.
They examined both people’s ideas about fairness and their willingness to accept a cost when redistributing money. Here, redistribution means moving part of a reward from the person who received more to the person who received less.
Participants decided what should happen to real rewards
Participants acted as outside decision-makers for pairs of workers recruited through an online labor platform. They could leave unequal earnings in place, reduce the difference or make the workers’ rewards equal. Their decisions affected real payments.
In one version, chance determined which worker received the higher reward. In another, the more productive worker received it. This let the researchers compare reactions to a gap of the same size with a different explanation for how it arose.
A third version introduced a cost to redistribution: some money was lost when earnings were transferred. That tested whether participants would tolerate a larger gap to preserve more of the workers’ combined reward.
The experiment’s design separates questions that are often entangled in public debate. Someone can believe an outcome is unfair while also worrying that changing it will leave less money available overall.
What the 85% figure means
The study reports that implemented inequality, meaning the inequality left after participants’ decisions, was 85% higher when the original difference reflected merit rather than luck.
This is a relative increase in the experiment’s inequality measure. It does not mean that 85% of participants supported unequal pay, or that inequality in the countries studied rose by 85%.
When redistribution carried a 50% efficiency cost, half the money taken from the higher-paid worker was lost during transfer. Implemented inequality increased by 14% compared with the luck condition without that cost. The source of the original inequality produced a much larger difference in participants’ decisions.
The comparison applies to these experimental conditions. It does not measure how a particular tax rate affects work, investment or an economy’s output.
People disagree about fairness and about the facts
The researchers distinguish three broad views. An egalitarian favors equal outcomes regardless of whether the gap began with luck or performance. A meritocrat accepts performance-related differences but favors correcting those caused by luck. A libertarian, in the experiment’s terminology, is more willing to leave either kind of difference in place.
These describe patterns of choices in the study, rather than complete political identities. Views varied within countries as well as between them. Meritocratic preferences were especially prominent in many richer Western societies, while the other views were widespread elsewhere.
People can also share a fairness principle and disagree about its application. Two meritocrats might reach opposite conclusions about redistribution if one believes high incomes mostly reflect achievement and the other believes inherited advantages or chance explain them.
The study found associations between fairness preferences, beliefs and attitudes toward redistribution. It also found links with redistribution through countries’ tax and transfer systems. Those relationships do not establish that the measured attitudes caused governments to adopt particular policies.
Real pay decisions mix several influences
The experiment gave participants a clear explanation for each reward. Actual salaries can reflect performance, qualifications, bargaining, family advantages and chance at the same time.
Separate research on salary negotiation and social class illustrates that complexity: people differed in their willingness to negotiate, and an employer experiment found different reactions to applicants’ requests depending on their class background.
For employers, the fairness study raises a question about whether workers understand and trust the reasons for pay differences. It did not test a company’s bonus policy or show that explaining a pay gap will make employees accept it. Establishing whether a reward genuinely reflects performance remains a separate task from deciding whether performance should be rewarded.
Cover image: Editorial composite using dice and banknotes to illustrate chance and financial rewards. These objects were not used in the study. Photos: Sara and Jonathan Borba / Pexels.
