There are a couple of supposed absolute truths that science holds as self-evident. The first is that research is self-correcting: incomplete, irrelevant and incorrect findings are shaved from the scientific record over time, to leave a reproducible and robust foundation for the future. The second is that science operates as a genuine meritocracy. Research and researchers advance on neutral data and objective analysis, so talent emerges alongside the truth.

Nature special: Inequality in science

But articles in this journal and elsewhere have drawn attention to doubts about the self-correcting nature of science and an apparent crisis in reproducibility. And this week we take a swing at the second of those supposed maxims, with a special series of articles that offers an analysis of science and inequality.

The good news is that science is keeping up with modern trends. The bad news is that trend seems to be towards wider inequality, fewer opportunities for those from more disadvantaged backgrounds and a subsequent smaller pool of people and talent for research to draw on. From the United Kingdom and Japan to the United States and India, the story is alarmingly consistent. In many places, careers in science tend to go to the children of families who belong to the higher socio-economic groups.

There are various reasons for this, and many of them are explored in the pages that follow. The problem is complex, but one of its implications is stark. Unlike many other sectors of society and the economy that rightly draw fire for a lack of social mobility, science relies heavily — almost exclusively in some places — on public money. If the research system is soaking up billions of pounds and dollars and yen from taxpayers merely to hand a subsidy to an already privileged sub-section of society — cementing their advantage in the process — then in no way can that system be described as positive for human welfare, however noble its goals.

In a Comment article, Mike Savage calls on researchers to settle their differences over definitions of social class. Those who categorize people according to their occupation are at loggerheads with those who classify according to income, wealth, culture and social ties. Each approach has its uses, he urges, and easing hostilities would allow better analyses of “our unequal, riven societies”. In another Comment piece, Branko Milanovic illustrates the power of the long view: looking at archival data on wages and incomes from as far back as the 1200s, he argues that inequality is cyclical and likely to fall soon.

Access is one thing, but the picture is no more comforting for those who have managed to secure a place on the inside looking out. For in scientific careers, there is a growing gap between the monetary rewards showered on the few at the top and the relatively meagre compensation that trickles down to the rest.

Our biennial Nature survey on salary and job satisfaction, which this year drew close to 6,000 respondents from around the world at all career stages, finds that most scientists — almost two-thirds — are happy with their jobs. But there is a considerable vein of discontent. Many, especially in Europe, are frustrated with the state of scientific research, with their own pay and with the competition for grants. Fewer than half of all European respondents said they’re excited about their future job prospects.

Still, worldwide, more than 60% of those asked in our survey said they’d recommend research as a career. This journal agrees with them. So how can we make science more accessible to all those who would like to get into it? There are echoes here of the ongoing struggle for equality for women scientists and for greater representation of ethic minorities in places such as the United States. And some of the same measures used to rectify inequality in those cases can be copied to stem economically based bias. Indeed, some — such as social inclusion schemes in Brazil — may already be bearing fruit. Active intervention to identify and encourage those being excluded, with the support of institutions and funders, seems to be crucial. Equally important is increased awareness, among those who pay for science and those who control who gets to do research, that the system is riddled with inequality and risks getting worse. This is one problem that absolutely, truthfully — and self-evidently — will not self-correct.