When class counts

David Rose (Letters, June 25) explains why we need to group people by social class using the example of how such classifications reveal social inequalities in health. But the introduction of the National Statistics Socioeconomic Classification can make it appear as if inequalities have narrowed. For instance, under the Registrar General's classification, life expectancy for men at the social extremes differed by 9.5 years by 1996. David Rose quotes NS-SEC derived statistics that appear to suggest that the difference has fallen to seven years.

The last time the gap was this narrow the Callaghan government was in power. To determine whether under this government social inequalities in health have narrowed to levels akin to those of the 1970s requires a classification that is comparable over a long time. Classes evolve but long-run comparisons have to be made if we are to understand trends and underlying processes.

How can NS-SEC be used to determine whether class matters more to the life chances of a child born in the UK now compared with that of their parents or grandparents?

Danny Dorling
Sheffield University