There are many different ways to define success. It is a thing said to console, usually to someone who has just come off badly in a comparison, and it lands as a soft phrase rather than a claim about the world.
Three large studies turn it into a claim, and a sharper one than the comforting version. The definitions of success do not merely differ in what they value. They differ in how they behave. And the one most people inherit by default has a property the others do not: the supply of it is fixed.
The arithmetic first
In 2010, Christopher Boyce, Gordon Brown and Simon Moore published a short paper in Psychological Science. Buried in its discussion is a sentence that does more work than the results.
The Artful Age
A weekly letter on aging well, family across generations, and the creative life after the kids leave home.
“There are fixed amounts of rank in society. Only one individual can be the highest earner. Thus pursuing economic growth, although it remains a key political goal, might not make people any happier.”
If satisfaction with money runs on your position rather than your pile, then everyone getting richer changes nothing, because nobody’s place in the queue moves. Your raise helps you only to the extent that it carries you past somebody, and that somebody loses precisely what you gained.
Whether satisfaction actually runs on position is the empirical question. That is what the studies test.
Britain
Boyce and colleagues used seven waves of the British Household Panel Survey, roughly 12,000 adults answering repeatedly between 1997 and 2004, giving 86,679 observations. Household income was adjusted for regional living costs and household size, and demographics were controlled throughout.
They set two predictors against each other. Absolute income. And relative rank: your ordinal position in the distribution, scaled from zero to one, a measure that knows only how many people earn more than you and how many earn less, and nothing at all about by how much.
Entered separately, both predicted life satisfaction. Entered together, rank dominated and absolute income added no variance whatsoever.
They repeated the analysis with three definitions of the comparison group, in case the result was an artifact of who people were being measured against: region, of which the survey has nineteen; gender crossed with education, giving six groups; and age, in twelve bands. Rank won every time, and neither absolute income nor the reference group’s average income explained anything further.
Worth keeping in proportion. The gap in life satisfaction between the highest and lowest earners in the entire sample came to 0.29 of a standard deviation. Doubling your income was worth about 0.1. These are real effects. They are not large ones.
One further result. Comparison does not run symmetrically. The best-fitting model weighted people earning more than you about one and three-quarter times as heavily as people earning less.
India
One country, one survey, one decade is a hypothesis rather than a fact about people.
In 2025 Tapasya Raj, at the University of Strasbourg, and Srikanth Reddy Umenthala, at the One Health Trust in Bangalore, ran the test on the Longitudinal Ageing Survey of India: nearly 67,000 observations from adults aged forty-five and over, across every state and union territory, controlling for everything from caste and religion to chronic disease and state unemployment.
Rank predicted life satisfaction strongly. A one standard deviation rise in relative rank went with a 0.697 standard deviation rise in reported satisfaction.
But they did not reproduce the British result exactly, and they say so in plain terms: “This finding diverges from previous studies conducted by Boyce et al. (2010) and Macchia et al. (2020), which found the role of absolute income insignificant.” In India, absolute income kept a small but statistically significant coefficient. Rank dominated. It did not eliminate.
They also found something the British data could not show. State-level wealth inequality moderated the effect: where inequality was higher, rank mattered more. The steeper the pyramid, the more your rung determined how you rated your life.
109 countries, with a label first
The third study needs that label before any of its findings. Edika Quispe-Torreblanca, Jan-Emmanuel De Neve and Gordon Brown published it in Nature Communications this year, using Gallup World Poll data from 109 countries and more than 90,000 people. Brown also co-wrote the 2010 British paper. This is one research program extending itself, not an independent check on it.
What it contributes is separation. Inside any single country, absolute income and income rank correlate at about 0.97, which is exactly why single-country studies struggle to tell them apart. Comparing across countries breaks that.
In 80 percent of countries, wellbeing was more strongly associated with income rank within the nation than with absolute income or with relative deprivation. It held across four different wellbeing measures and replicated across survey years.
Then the result that changes the argument’s shape. Rank coefficients were more than three times larger in the most materialistic countries. And in the countries with the highest civic engagement, the association between income rank and wellbeing was roughly 80 percent smaller.
The arithmetic of rank is fixed. How hard it presses on people is apparently not.
What is not being claimed
All three studies are observational, and every set of authors volunteers it.
Boyce and colleagues note that rank might be shaping an underlying experience of wellbeing, or merely shaping how people report it, and that they cannot rule out the second. They add that income rank may simply be standing in for social rank in general.
Raj and Umenthala’s respondents are all forty-five or older and their data are a single cross-sectional wave. And a detail in their tables that they pass over without comment: once rank and state mean income are in the model, the coefficient on household income per capita is not merely small, it is negative.
The Nature Communications authors write that “the data remain observational, and our estimates should not be interpreted as establishing causal effects,” and name omitted variables, reverse causality and measurement error. Their rank measure is computed at national level, which leaves out the local comparison groups people probably actually use, and they point out this could understate rather than overstate the effect. They also note that people systematically underestimate where their own income ranks.
There is a live disagreement in the field, too. A separate and well-known line of work finds that absolute income keeps pushing happiness upward with diminishing returns. The rank studies find absolute income dropping out once position is accounted for. Those are different claims about the same relationship, and they have not been reconciled.
The part that would be wrong to take from this
None of it says money does not matter. It says something narrower, about what money is doing once a person has enough of it to be ranked rather than to be struggling. The cross-national study found the rank association smaller in lower-income countries, not larger. Poverty is not a positional problem, and the Indian paper is working in a country where the top one percent hold over a fifth of the income. Telling someone without enough that their difficulty is really about comparison would be both false and unkind.
Back to the sentence
There are many different ways to define success. The version worth keeping is not that every definition is equally good and you should pick whichever feels kindest. It is that the definitions have different shapes, and the default one is a queue.
Measuring yourself by earnings means entering a ranking where the good positions are fixed in advance and every step up is a step somebody else takes down. That is not a character flaw in anyone who does it. It is a property of the measure itself. And on the evidence from 109 countries, even that property is not destiny, because its grip loosens substantially in places where people are more involved in one another’s lives.
Which replaces a tired question with a better one. Not whether money buys happiness, but what your income is being compared against, and whether you chose that comparison or simply inherited it.