Children in an education authority praised by inspectors for its efficiency have done less well in national tests than those in more deprived areas.

This anomaly - and others - has been revealed by a TES analysis of English authorities’ 1998 key stage 2 test results which takes poverty into account.

The analysis, which uses number of children eligible for free school meals as an index of poverty, suggests which authorities are not doing as well as might be expected, and those that are doing better.

The London borough of Newham’s test results - it came second to bottom out of all LEAs - appear to be worse than those in at least 14 authorities with a higher proportion of deprived pupils.

Newham also has a high proportion of pupils with English as a second language, but probably no greater than its more deprived neighbour, Tower Hamlets, which was criticised by the Office for Standards in Education.

In its defence, the borough’s chief inspector, David Lister, suggests Newham has a more diverse ethnic population than Tower Hamlets, and more refugees.

The inspection report on Tower Hamlets, which did marginally better than Newham in The TES analysis, suggested results might have been higher had the authority introduced better systems of support for schools.

Newham’s inspection report commends it for taking the view that “onus for school improvement, particularly so far as the quality of teaching is concerned, rests with the schools”.

Greg Wilkinson, an education consultant working for the Cabinet Office, said it was not surprising that local authority inspection reports might judge success differently because “the relationship between outcomes and processes is not a straightforward one”.

Mr Wilkinson, who formerly inspected local authorities for the Audit Commission, said: “An authority can be doing all the right things but for all sorts of reasons can have a poor level of achievement or get good results even though they have poor processes.”

The scatter graph, above, shows that once poverty is taken into account, the local authorities that appear to be struggling at key stage 2 include a string of new unitary authorities. Similar patterns emerge at key stages 1 and 2.

The councils that are furthest away from the level that might be expected include Sandwell, Walsall, Thurrock, Newham, Peterborough, Milton Keynes, Poole, Bristol, Rotherham and Leicester. The City of Nottingham was the lowest-scoring authority: 8 per cent below Newham and 28 per cent below the national average. But no comparable figures for free meals are available to place it on the graph.

Deprived inner London boroughs that appear to achieve below what might be expected have many pupils with English as a second language.

However, there are exceptions. The London borough of Hammersmith and Fulham has similar levels of black and ethnic pupils as other parts of the inner city and yet it scores well above Lambeth, Hackney and Southwark.

Christine Whatford, incoming president of the Society of Education Officers and director of education at Hammersmith and Fulham, said: “This analysis is a better way of looking at the results than simply raw scores.

“An Ofsted inspection is a more subjective exercise because it is based to some extent on what the schools tell inspectors about what the lea offers them.”

The graph also shows that some local authorities are achieving remarkable results with more advantaged children. The results in the London borough of Richmond are far better than might be expected even though only 12.3 per cent of its primary pupils are eligible for free school meals.

Martin Rogers of Education Network, which promotes the role of local authorities in education, said of The TES analysis: “This produces a picture which is much more valid than crude statistics.

“In order to solve the problem of underachievement we need to discover its root causes. Poverty is not a cause of underachievement, but the things that go hand in hand with poverty are. A range of criteria would be considered: language, free school meals, children’s prior achievement, parental achievement and nursery provision.”

Mr Wilkinson said: “The TES analysis is very useful for highlighting authorities that appear to be close to the national average on their raw scores, but are not doing as well as they should be.”