This year's ranking of schools shows only slight variations among the schools that do extremely well in exams and among those that do badly. Five of the 10 high-performing comprehensives were in the top 10 three years ago, when the first tables appeared.
Successful schools, though nominally comprehensive, are over-subscribed and therefore have to introduce some form of selection. The Liverpool Bluecoat school, which is the one of the two top-performing comprehensives, takes its pupils from a wide catchment area.
A number of schools scraping the bottom have managed to improve their results. Stratford School, the grant-maintained school in New-ham, east London, judged recently by the Office for Standards in Education to be failing to provide an adequate education, has improved its results from 4 per cent of fifth-formers gaining five good GCSEs (grades A*-C) to 11 per cent.
The national average for the proportion of fifth-formers gaining five or more higher grade GCSEs has gone up from 38.1 per cent in 1992 to 43.3 per cent in 1994. The average point score for A-level students has gone up from 15 in 1992 to 16.7 this year.
Within the general improvement in results, some schools have managed striking rises in the proportion of fifth-formers getting five GCSEs at grades A* to C. Boston Spa comprehensive in Leeds has seen its proportion of higher-grade passes go up from 46 per cent in 1992 to 60 per cent this year.
The league tables appear set to become a national institution - the Labour party this week announced they would not be scrapped with a change of Government - but the teacher unions remain opposed. Doug McAvoy, general secretary of the National Union of Teachers, complains that they are a predictable sham, reflecting primarily socio-economic background.
The Government is sensitive to such criticism and has asked for advice on producing value-added tables from the Schools Curriculum and Assessment Authority. The sub-committee examining the options is likely to advocate using a measure of progress that allows schools to be compared against their previous results. The Scottish tables already give figures over three years.
Included in this year's tables are figures for the length of weekly teaching time in schools, though OFSTED has already suggested there is little correlation between hours and exam results.
The problem of rating schools according to their success in vocational exams has yet to be solved. The tables only give the proportion of pupils gaining qualifications, but they are not directly comparable with GCSE results.
Schools may be reluctant to direct pupils towards vocational courses if the results do not give a measure of the school's performance on such courses.
The next stage in the development of league tables is likely to be the publication of the test results of 11-year-olds. The Government has yet to give a date - the tests are due to be national next year. League tables of primary schools will generate even greater opposition from the teacher unions than the secondary tables.