Every year in Singapore, one percent of pupils in the third year of primary school bring home an envelope headed ‘On government service’. Inside is an invitation to the city-states Gifted Education Programme. To receive the overture, pupils must ace tests in maths, English and general ability. If their parents accept the offer, the children are taught a special curriculum.

Singapores approach is emblematic of the traditional form of ‘gifted education’, one that uses intelligence tests with strict thresholds to identify children with seemingly innate ability. Yet in many countries, it is being overhauled in two main ways. The first is that educationists are using a broader range of methods to identify highly intelligent children, especially from poor households. The second is an increasing focus on fostering the attitudes and personality traits found in successful people in an array of disciplines — including those who did not ace intelligence tests.

IQ tests have attracted furious criticism. Speaking for the sceptics, Christopher Hitchens, a journalist, argued that: “There is… an unusually high and consistent correlation between the stupidity of a given person and (his) propensity to be impressed by the measurement of IQ. Like any assessment, IQ tests aren’t perfect”. But as Stuart Ritchie of the University of Edinburgh points out in ‘Intelligence’, researchers in cognitive science agree that general intelligence — not book-learning but the ability to reason, plan, solve problems, think abstractly and so on — is an identifiable and important attribute which can be measured by IQ tests.

Just how important is suggested by the Study of Mathematically Precocious Youth (SMPY), founded in 1971. Julian Stanley, then a psychologist at Johns Hopkins University recruited 5,000 precocious children over 25 years, each of whom had intelligence-test scores in early adolescence high enough to gain entry to university.

Research into how these children did in adulthood has emerged over the past two decades. Of the SMPY participants who scored among the top 0.5 percent for their age-group in maths and verbal tests, 30 percent went on to earn a doctorate, versus 1 percent of Americans as a whole. These children are also much more likely to have high incomes and to file patents.

Officials often cite the SMPY as inspiration for the creation in 2014 of two specialist maths schools in England. Based on the Kolmogorov School in Moscow, these schools accept only pupils who excel in maths at exams at age 16. In January, the government said it wanted to open more as part of its industrial strategy, a plan to boost Britain’s woeful productivity growth. Linking gifted education to economic growth may horrify some people. But it has long seemed like common sense in countries without many natural resources, such as Singapore.

In America, 48 out of 50 states have programmes for brainy children, but in the decade before 2013, 24 redefined them, typically ditching the ‘gifted’ label in favour of ‘high-ability. Today, no state relies on a single IQ score to select students. In his book Ungifted, Scott Barry Kaufman of the University of Pennsylvania calls this “a huge change from just 20 years ago”. European countries have seen similar shifts.

A broader approach to gifted education ensures that more children reach their potential. But the evidence suggests that, so long as they are open to everyone, IQ tests still have a vital role to play. To find lost Einsteins, you have to look for them.