ÁñÁ«ÊÓƵ

David Rose (Letters, May 21) defends the new "nominal" government social-class schema as being the product of a seven-year programme of research involving fellows of the Academy of Social Sciences and the British Academy. It is telling that class still matters enough that such credentials are wielded in defence of the schema, but whether the programme took seven or 70 years is irrelevant.

In the original report proposing the new scheme (Constructing Classes, 1997) Peter Elias and Abigail McKnight argue that "one would expect, therefore, that a social classification will correlate with employee earnings". That correlation was eroded by the new schema, which makes comparisons over time and between people grouped by income near to impossible using census data. Social classifications tend to reflect the status of those involved in the classification. As academic salaries rise only slowly, but the titles academics award to each other multiply, it is hardly surprising that academics should redesign the class schema to downplay the importance of income. Unfortunately, over the period 1994-2003, people in Britain have become more clearly divided by income.

Hidden in the detail of the latest official "households-below-average-income statistics" is the fact that after the cost of their housing, the poorest 20 per cent of people in households where the head or their spouse are unemployed had ?16 a week to live on in 2002-03 compared with ?49 in 1994-95. This can be compared with a rise from ?569 to ?722 a week after housing costs for the 20 per cent best-off people in full-time work.

The report thus highlights one mechanism by which unemployment has been reduced, how some income inequalities have grown, and that the Chancellor's target of reducing child poverty by 25 per cent by 2004 is unlikely to be realised. An archaic if newly constructed measure of social class is of little use if it confuses rich and poor, no matter how many fellows contributed to its creation.

Danny Dorling
Sheffield University