The latest set of performance tables have given rise to fears that lower-achieving schools may be concentrating on improving their ranking at the expense of their weakest pupils.
Over the past two years, the proportion of children leaving school without even a single grade G at GCSE has risen from 7 per cent to 8.1 per cent. Pressure on schools to improve their share of GCSEs grades A* to C may be greatest where their results are below the national average.
Those local authorities that achieve below the national average of 43. 5 per cent of fifth-formers achieving five A* to Cs tend also to have the highest proportion of children who do not gain any qualification. However, one exception is Kingston upon Thames, the top borough in terms of pupils scoring higher-grade GCSEs. Nearly 8 per cent of fifth-formers leave without any GCSEs, a slight increase on the previous year.
Schools with the highest proportion of pupils without any qualifications are in the inner cities. More than half the fifth-form at one Manchester school, Moston Brook high, left without a GCSE and there are another 24 schools where a third or more pupils failed to get even a G grade at GCSE.
Overall, the proportion of 15-year-olds gaining five or more higher-grade GCSEs shows a slight increase from 43.3 per cent last year to 43.5 per cent this year. The proportion gaining five A to Gs has also marginally improved.
The number of top comprehensives - those in which more than 60 per cent of pupils achieved at least five higher-grades GSCEs has increased since last year by 20 per cent - while in 107 schools fewer than 10 per cent of pupils achieve that benchmark.
Accusations that there is a deepening divide between the best and worst schools are likely to irritate ministers who see the tables as a successful public relations exercise. The publication of the results of more than a million pupils is the biggest public information effort undertaken by any government department.
Truancy in schools has remained roughly constant, though the figure for authorised absence shows a drop from 11.9 per cent to 8 per cent. Schools with high levels of truancy tend also to be those with the poorest exam results.
There is little change in the relative performance of schools and local education authorities. The high-performing schools tend to be oversubscribed and are often able to select their pupils either by academic tests if they are grammars or, more covertly, through interviews with parents.
Kingston upon Thames tops the table with 55.5 per cent of pupils achieving five or more higher-grade GCSEs. It has been in the top five since the tables were first published. The other authorities that vie for the top place are Sutton, Barnet, Harrow and the Isles of Scilly (which has only one secondary school).
This year, Islington comes out worst with only 17.4 per cent of fifth-formers gaining the benchmark of five A* to C grades. However, this London borough has always featured in the bottom 10. The authorities clustered at the bottom tend to be small and in deprived areas. The exception is Manchester, one of the largest education authorities, where only 22.5 per cent of fifth-formers managed five or more higher-grade GCSEs.
The Government has tried to be fairer to schools this year by publishing alongside exam results the number of children in each school who have special needs, but have not been statemented. Also for the first time, the tables allow schools to show the proportion of Year 11 pupils - not just the 15-year-olds - who achieved five or more higher-grade GCSEs.
Until now, older candidates have not had their results recorded and that has upset an independent-school lobby. The new category records the results of older pupils who may have taken longer to prepare for the exam because they may be wrestling with English as a foreign language.
David Cleland of the Society of Headmasters and Headmistresses of Independent Schools believes the new category is fairer, but he regards the move as a concession rather than an acceptance that the system is wrong. He believes results should be based on the teaching group and not the age group.
However, only a minority of schools have bothered to submit an entry based on the teaching group and those ranking schools are likely to continue to use the number of 15-year-olds.
Gillian Shephard, the Education and Employment Secretary, confirmed this week that she has commissioned work on creating valued-added league tables, which will show the schools which are the most effective, given their intake of pupils. The intention is to use the national tests for 14-year-olds as a base for calculating improvements in pupils' exam performance.
However, it is unlikely that national value-added tables could be prepared before 1997. Pilot schemes are to be undertaken next year.