The secondary school that pupils attend makes very little difference to their academic achievement. Their attainment is, instead, largely determined by their family background and prior academic achievement. And this means that test scores are an inadequate way to measure a school's success, according to Stephen Gorard and Emma Smith of Birmingham University.
They surveyed 12,575 15-year-olds in schools in England, Belgium, France, Italy and the Czech Republic, looking at the teenagers' academic records and quizzing them about their aspirations.
Their enquiries reveal that between 80 and 100 per cent of pupils' secondary school achievements can be explained by background factors, such as wealth, class and prior attainment.
"In developed countries, it does not appear to make much difference which school a student attends," the researchers say.
Going to school, they add, is obviously preferable to not going to school at all. But attending a different school within the same system has very little effect.
"Almost all schools are free, compulsory, roughly equal in funding, inspected, with trained staff, widely shared curricula and standardised tests," they say. "There is very little variation left to attribute to the differential impact of schools."
Pupils' background, however, does affect raw test scores. The researchers point out that all "failing" schools in Britain have high levels of pupil poverty. So mixing pupils by class, income or sex has no direct impact on these scores; poor pupils will not get better grades just because they have rich classmates.
However, teenagers' ambitions can be influenced by studying alongside the children of better-educated parents. Those who go to school with a high percentage of pupils who have educated, professional parents are likely to have higher aspirations than pupils whose classmates are mostly from less well-educated families. This applies even if a pupil's own parents are less well-educated.
"Allowing students from professional, educated families to cluster in specific schools will encourage social reproduction," the academics say. "There is a cost in terms of social mobility. Thus ... we conclude that comprehensive and undifferentiated schools are the best."
Test scores are regularly used to measure school effectiveness and success. This has over-emphasised the importance of examinations, at the expense of the other benefits of schooling. For example, the researchers believe that as much as 20 per cent of pupils' academic success is determined by their teachers' and classmates' helpfulness and trustworthiness. Aspiration - and therefore social mix - also plays a role.
"The implications for policy are clear," they say. "To raise the occupational and educational aspirations of the most disadvantaged ... a mixed school intake is desirable.
"A mixed, comprehensive and undifferentiated system of schools is preferable ... to a tracked, selective, faith-based or specialist one."