Problems over using free school meal statistics - a key indicator of families living in poverty - are thought to be holding up the consultation paper on dealing fairly with schools that take children from different home backgrounds.
Education and Employment Secretary Gillian Shephard said this week that it is too simplistic to link poor performance with deprivation.
Department for Education and Employment officials have given initial advice that schools should be set reasonable targets, based on the proportion of children eligible for free school meals.
The analysis of last year's maths results for 11-year-olds by the School Curriculum and Assessment Authority show a 20 percentage point difference in the median score between schools that have fewer than 10 per cent of children on free school meals, compared with schools which have more than 35 per cent on free school meals.
Last November Mrs Shephard announced the Government's scheme to devise national "benchmarks", setting out the standard groups of schools with similar characteristics are expected to achieve. Under the scheme every school must inform parents of their benchmark and set individual targets for improvement.
The Association of Metropolitan Authoritie s is claiming ministers have got cold feet over the idea of benchmarking schools.
Lord Henley, the junior education minister, admitted in the Lords that the consultation paper had been held up. However, he denied a suggestion from Lord Morris of Castle Morris, leader of Labour's education team in the Lords, that Number Ten had intervened to prevent the document's publication.
Policymakers drawing up guidance to schools on setting realistic improvement targets have been hampered by the absence of value-added information. This will be available from next year when the scores achieved by 11-year-olds can be matched against their performance at seven.The Tories have pledged that targets for individual schools will be in the manifesto and are likely to be embarrassed by anything which publicises that schools with high proportion s of free school meals children will not be expected to do as well as pupils in more favoured areas.
It is understood that the consultation on benchmarking will be postponed until after the election, although SCAA is continuing to hold conferences on benchmarking.
SCAA officials say the timetable has slipped because it is consulting on a range of associated issues. Most academics accept that more accurate benchmarks can be set when there is information from the seven-year-old tests that relates to the 11-year-olds being tested.
The DFEE said Lord Henley had made clear in the Lords that the consultation was going through, but there were very difficult issues to be resolved on how to compare schools on a like-for-like basis.