In countries from India to Sweden, policymakers are increasingly turning to performance-related pay for teachers as a means of driving up educational attainment.

So far, governments have largely shied away from setting pay purely on the basis of exam performance, preferring to rely on a wider range of factors. But a landmark survey published today reveals that in England more than half of teachers actually want their pay rises to be dependent on their students’ results.

The research, commissioned by social mobility charity the Sutton Trust, shows that the most popular criteria for deciding pay progression are assessments by senior colleagues (backed by 60 per cent) or by headteachers (54 per cent).

However, 53 per cent of the 1,100 teachers surveyed said their pay should be decided by “considering the progress and results of pupils they currently teach” - more than the 47 per cent who said pay should be based on length of service.

Perhaps surprisingly, one in 10 teachers took the view that pupils’ evaluation of their performance in the classroom should also be factored into pay decisions, rising to 14 per cent among secondary staff.

A Department for Education spokesman said the survey showed “strong support from teachers across England for our plans to let schools pay good teachers more”. The reforms, he added, which would allow schools to “reward the most effective teachers who get the best out of their pupils”.

Performance pay for teachers, which will be introduced in England in September, has proved to be highly controversial and has been fiercely opposed by classroom unions. The reforms, designed to prevent teachers from receiving automatic incremental pay rises and to give schools the flexibility to pay their best teachers more, have sparked national and regional strikes. The NUT has threatened to hold a further national strike over pay, pensions and working conditions on 10 July.

But although the Department for Education has argued that tying teachers’ pay to classroom performance will improve educational standards, it has held back from advising schools to directly peg pay rises to exam performance, instead allowing institutions to choose their own criteria.

Right-leaning thinktank Policy Exchange has also urged schools to base pay decisions on a “balanced scorecard” of measures.

Sutton Trust chair Sir Peter Lampl said the results of the survey revealed a “positive response by a majority of teachers to performance-related pay, based on senior staff assessment and pupil progress”.

“Sutton Trust research has shown evidence from the UK and the US that there is a significant correlation between teacher evaluations and exam results,” Sir Peter said. “However, the evidence also suggests that schools should rely on a combination of approaches to gain a fuller picture of teacher effectiveness, and that teachers should be assessed on their cumulative performance over several years.”

Among the teaching profession, the reaction to performance-related pay has been mixed. A Policy Exchange poll last year found that 89 per cent of teachers thought “quality of teaching should be a major driver in pay and progression”. In another survey by the NUT, 88 per cent of teachers said they would prefer a national pay structure to individual decisions being made at school level. And in March, just 3 per cent of school leaders surveyed by TES and the Association of School and College Leaders said that their staff would receive more money in their pay packets as a result of the reforms.

But Jonathan Simons, head of education at Policy Exchange, believes the teaching profession is warming to the idea. “Teachers tend to react more positively than their alleged representatives,” he said. “If it’s a fair system, it’s reasonable for teachers to be paid on the basis of [student] performance.

“If it’s a choice between a decision being made by a third party or empirical data, I can see why teachers would support this. However, there’s evidence which suggests students make different progress in a top set or a bottom set, or in different subjects. I like it as a theory - it’s a good model if you have reliable standardised data. If not, we think a balanced scorecard is the best method.”

Russell Hobby, general secretary of the NAHT headteachers’ union, said concern among teachers was largely based on “fear of the unknown” and worries about the reliability of test data. Reservations remained about whether the impact of an individual teacher could be isolated from exam results, he added.

NUT deputy general secretary Kevin Courtney said the survey “gives yet more proof that performance-related pay measures are not welcomed by teachers”, adding: “There is no evidence worldwide that linking teacher pay to performance leads to improved education standards.”

The suggestion that performance pay is gaining in popularity was also rejected by Martin Freeman, director of economic strategy at the ATL teaching union. “Our members uniformly hate it,” he said. “Not a single member has told me they don’t mind this system.”