A nation dividing? Some questions of interpretation

摘要

Dorling and Rees's (2003) analysis of intercensal social polarisation is welcome in many respects and I suspect I am not alone in admiring the technical skill involved in such work, not to mention the speed with which the paper was produced! The purpose of this comment is to suggest that this paper should be the starting point for a debate about polarisation, for several reasons. First, the majority of the indicators actually demonstrate reduced polarisation for the 1990s. Analysis of the evidence presented for the 1991 - 2001 period, in tables 1 - 5, shows that 60 indicators indicate a reduced level of polarisation and 51 an increase. This is not the impression one would get from reading the paper, which emphasises the growth of polarisation in the 1990s [Britain "divided further still, in the 1990s" (page 1288); "continued to polarise in other ways and as a whole" (page 1289)]. This directs attention to the choice of indicators for further discussion. The authors argue that they incorporated every indicator which was capable of measurement, and if so the balance of evidence indicates that the majority of indicators imply that polarisation is declining. There is also some double counting [as the authors acknowledge (page 1308)]: among the groups becoming more polarised are those working over 30 hours per week, those in professional occupations, and those with a degree, and a fair number of people must fall into all three categories. The picture of increased polarisation for the 1991 - 2001 period is therefore much more mixed than the paper suggests.