Meritocracy’s unfair

💡

This post was originally published on the Sense of Fairness blog (original post here)

Meritocracy is unfair — at least believing in it too much can trap us in unfairness.

Like many British people I have recently been reading a good deal of history. My main era of choice has been the British revolution, usually known as the civil war, or wars, and mostly I have leaned on the brilliant Christopher Hill, still some years on regarded as the greatest historian of that period. Most recently I have read God’s Englishman, on Oliver Cromwell*.

Most odd to the modern ear is the certainty felt by Cromwell that he was led by divine providence. In this understanding of the world, military and political success, which Cromwell enjoyed in abundance, were merely a reflection of God’s will. There was no room for chance; success was itself a sign that he had been chosen and that his power was rightful. As he said to the 1655 parliament: “What are all our histories and other traditions of nations in former times but God manifesting himself that he hath shaken and tumbled down and trampled upon everything that he hath not planted?” Because Cromwell flourished where he was planted, God’s providence was with him.

Hill’s cynicism about Cromwell’s perspective is subtle, noting the tendency only to engage in battle when the roundheads had an overwhelming numerical advantage, and a somewhat equivalent tendency in his political manoeuvrings to act only when it was clear which way the tides of broad opinion were flowing. Cromwell himself would perhaps have seen both of these as merely waiting until the signs of God’s providence were made plain to him. I’m sure that many Irish might be rather more cynical about Cromwell’s view of divine providence, given the horrors in that land that his understanding of his religion led him to believe were wholly appropriate. Nevertheless, Cromwell felt assured by his perceived successes that providence was with him.

This argument that personal success is the result of being chosen, and that it is rightful because of that fact, seems odd to the modern ear. It seems odd, but the modern brain tends to fall into a remarkably similar fallacy. That is where meritocracy comes in: we tell ourselves that we live in meritocratic societies and therefore that we deserve all that comes to us. And through that same route, unfairness comes in: both because our meritocracies are flawed, and because there’s plenty of dumb luck that drives success, not just the skill and hard work that we like to imagine, particularly when the success is our own.

And by falling into undue faith in meritocracy, we become more unfair in our actions and outlook.

In part, this is down to how humans understand fairness. Various studies have shown that people will tolerate more inequality when it seems justified by merit. Perhaps most striking is the 2019 paper Cutthroat capitalism versus cuddly socialism: Are Americans more meritocratic and efficiency-seeking than Scandinavians? from a team at Norway’s Centre for Experimental Research on Fairness, Inequality and Rationality.

In this large study, so-called ‘spectators’ were given the chance to reallocate rewards between two ‘workers’ to whom pay had initially been allocated wholly unequally (one of the two receiving nothing at all). Playing to national stereotypes, the most popular choice among the US participants was to leave the distribution as initially awarded, whilst a majority of the Norwegians equalised the reward. The core of the study though was to assess how responses differed between situations first where the spectators were told the initial allocation of the reward was based wholly on luck and second where the basis of the reward was said to be differentials in productivity (which was undefined and unquantified) between the workers.

Perhaps unsurprisingly where the source of the inequality was merit (in this sense of greater productivity), spectators from both countries were willing to tolerate more inequality in reward, than if the basis was simply luck. Though many Norwegians still favoured equalising the rewards of the workers, the merit basis led to more being willing to tolerate inequality; more starkly, 80% of the participants from the US allowed unequal distribution of rewards where there was merit involved (though many did not leave the allocation quite as unequal as the starting position), while more than half of them equalised rewards where luck was the driver. As the paper concludes:

“We show that the source of inequality is essential for understanding inequality acceptance in both the United States and Norway; in all subgroups of our samples, we find that the introduction of a difference in productivity as the source of inequality significantly increases inequality acceptance.”

Similar findings come in Merit and Justice: An Experimental Analysis of Attitude to Inequality. Again here there was a comparison between attitudes to earnings based on skill and on luck. A group of individuals played a series of rounds of one sort of game (either skilled or lucky) and then had the chance to fine earnings from the other players. The group then played the other sort of game and had a second opportunity to remove earnings from others. There was a greater tendency to try to ‘rectify’ earnings where the game was one of pure chance, and some greater willingness to tolerate inequality arising from skill.

The paper discusses its findings in the context of a broader philosophical understanding of justice and fairness, and concludes that “Individuals attach merit to an outcome when it is due to skill, and do not when it is due to luck. Thus, the concepts of moral desert and justice are deeply connected, and one needs the other for a proper definition.”

A further paper concludes something similar: “the perception of fairness is sensitive to the sense of entitlement, an important contextual cue that constitutes distributive justice”. This conclusion arises from The Flexible Fairness: Equality, Earned Entitlement, and Self-Interest. The study framed dictator and ultimatum game decision-making in the context of performance in another game, supposedly one of skill. Framing the ultimatum game with entitlement from merit (or the absence of such entitlement) led to different offers and different acceptances. In essence, those who felt entitled by apparent merit were less generous in their offers, and most starkly those who felt lower entitlement because of lesser perceived contributions were much more willing to accept even very uneven ultimatum game offers.

This appears to be a universal trait — apparent merit colours how we perceive fairness. And, sadly, we do not seem to test too closely how significant is the merit factor in the overall result, and do not try hard to test whether merit is really there at all. The global nature of this attitude is shown by further work by the same team as delivered Cutthroat Capitalism, under the banner Fairness Across the World: the Nature and Consequences of Inequality. This is ongoing work based on redistribution experiments like those for the earlier paper, involving 65,000 people in 60 countries. It is not yet published, though headline results were released in a talk last November. I have scraped the following images from a twitter thread by @RobertoIacono83, Roberto Iacono of the Norwegian University of Science and Technology, Managing Editor of the Journal of Income Distribution.

Redistribution across countries is less equal in cases of merit than of luck:

Clearly, our understanding of what is fair is framed by an understanding of merit — even where the extent of merit and what has driven differential performance is unclear. Particularly striking is the variation between countries, the willingness in many countries to tolerate high levels of inequality even when it arises purely from luck. But overall, it is clear that people worldwide generally tolerate much less inequality where it arises from luck than where it arises from merit — the luck chart is set at a much lower level than the merit chart — though people from a few countries (most notably India and China) make little distinction between the two, tolerating remarkably similar levels of inequality whether it arises from luck or merit.

Fairness is not inequality. Our understanding of what is fair is shaped by our understanding of merit — or just an indication that one party merits more than the other. The problem arises because that meritocratic urge is not well calibrated and we seem willing to accept undue inequality on even the weakest evidence of differential merit. We want to believe in meritocracy and so are willing to do so even when it isn’t really there — we are tricked by our faith in fairness into believing that we are more surrounded by meritocracy than we really are. And that leads us towards unfairness.

Studies show that where managers are asked to frame their promotions and bonuses in terms of merit, they are more likely to display bias in favour of men rather than women. Where merit is not so heavily pushed in organisations, ironically managers are more likely to operate fairly between the sexes. Even in hypothetical situations, our bargaining position and more significantly our assumption about what we see as fair is coloured by what role we are asked to perform, meaning that our understanding of merit displays bias. In the Flexible Fairness experiments participants’ vision of merit and what proportion of the overall reward that entitled them to was blurred by their own position, as either the better performer or the worse performer: “earned entitlement is evaluated in a self-serving way”. We recognise merit when it flatters us, but are less swayed by the same measures when they do not portray us in as positive a light.

We like meritocracy and respond more favourably to inequalities that arise from merit. And yet we all know that our supposed meritocracies are not perfect. We know that luck is often a big driver of success — though like Cromwell we want to think that success (particularly our own success) is always righteous rather than just the product of chance. We want to believe in meritocracy, and particularly those of us who enjoy some success want to believe in meritocracy. And that leads us to act less fairly than we might, because we are tempted to see success as evidence of merit, not simply of a combination of factors including luck. We might be better off, and happier, if we were grateful for the chances that have helped us to be successful, than if we believe it is simply on the basis of merit.

As the Cutthroat Capitalism paper states: “The lower acceptance among Scandinavians than among Americans of inequalities reflecting luck may contribute to explain why there is greater support in Scandinavia than in the United States for policies aimed at reducing the accident of birth as a source of inequality.” Remember that fairness is a choice and that the evidence shows that the heritage of Scandinavia was no more equal than any other part of the world. There appears to be nothing inevitable that made Scandinavia different, other than political choices in the second half of the twentieth century. The article I reference in that blog was discussing Sweden but note that the chart shows Norway followed the same trajectory as its neighbour, with inequalities as extreme as anywhere in the years around 1900, and notably low inequality now.

A belief in meritocracy seems to play a significant role in the US view that there is less need to intervene to prevent accidents of birth having broader impacts. This slide from a recent presentation by Harvard University’s Stefanie Stantcheva on Perceptions of Inequality (given at the San Diego meeting of the Allied Social Sciences Association and the Econometric Society earlier this month, as part of a special session dedicated to the IFS Deaton Review into inequalities in the twenty-first century) is very striking in this context. It compares the real likelihood of someone born in the poorest fifth of the population making it to the richest fifth in their lifetime with the perceived probability. In the study, only the US participants thought this was more likely to happen than is in reality the case (and they put the greatest likelihood on that event of all participants even though it is least likely to happen in their country).

People in the US are lulled by the myth of the American dream into believing their nation is more meritocratic than it is in reality — leading them to tolerate more inequality and to accept undue unfairness. It is surely this excess faith in meritocracy that leads to the US stereotypes displayed in a number of the studies discussed here.

We are all lulled by our wish to live in meritocratic societies into a belief that merit is a greater driver of success than it in fact is. Meritocracy may be fairer than the alternatives, but we risk becoming decidedly unfair when we have too much faith that we actually live in wholly effective meritocracies.

I post this blog on the anniversary of Oliver Cromwell’s hanging and beheading. That was in 1661, more than 2 years after his death and burial.