On 28 February last year, the new Labour UK government announced the first ever health inequalities targets for Britain. In the words of the secretary of state for health, Alan Milburn, “these targets represent the most fundamental challenge to the opportunity society we seek to build”. This is clearly a good step forward in the rhetoric of the UK government and should be applauded. As far as we know Blair's cabinet has been the first one in setting such targets and in naming things by their true name: clearly accepting the use of the word inequalities, instead of the euphemistic “variations in health” used by previous UK governments and by the official documents in many other countries.

The commitment of the UK government is, by concentrating their efforts on just two general targets (reducing inequalities in infant mortality and in life expectancy) to obtain a general reduction of 10% in inequalities in health in the British population for the year 2010. Despite these positive aspects, the reading of some papers just published by chance in this issue of the journal leads to three main invitations for reflection on this British initiative. These three points may hinder the ability of the UK targets to obtain fully its proposals, and are: the statistical definitions of the targets; the general focus of the target actions on individuals and not in social processes; and the near to complete lack of an explicit intersectoral approach, as the proposed actions are heavily rooted in the Department of Health's area of influence.

The formulation of the targets comparing the worst group (manual groups) with the whole population is new and extremely unusual in the inequalities in health literature. This decision could result in a reductionistic appraisal of the very concept of health inequalities. Usually health inequalities are measured comparing the best positioned group with the worst one; by comparing with the population averages, inequalities in health will be by far less apparent. The selection of the statistical measures of health inequalities is a not negligible issue, as is shown in this issue of the journal in a paper in the Theory and methods section by T A Blakely and I Kawachi who compared the use of mean and median of income. Also we publish in the Speaker's corner this month a letter from Fran Baum addressed to the politicians of the richest countries, of which Alan Milburn is one. The British targets concentrate the efforts solely around the poor, which are presented, even statistically as deviants from the societal averages. Baum argues that health inequalities are not a problem of the poor but a result of the social processes, including policy decisions, which permits poverty to exist. Perhaps a shift is needed from a focus on the poor to a focus on the processes that generate poverty including policy decisions.

An intersectoral policy, as analysed by L M van Herten and colleagues in the first paper in our Theory and methods section, is a good strategy to make that change. But the British targets fail to identify entry points in policy niches different from the Department of Health, just one of the five actions proposed are clearly out of the area of influence of health services. If we focus on the health indicators of poor people we can perhaps reduce the mortality and morbidity rate in the poor, but attaining this objective is not a guarantee that the number of people exposed to poverty has decreased. This difference between relative and absolute effects could be of dramatic importance in Britain nowadays where the number of poor people in the past two decades has increased.

Additionally, even if these questions are not explicit in the British targets, interventions to reduce inequalities in health can never forget gender inequity—as important as social class ones—and the international dimension of health inequalities. As gender and international issues interact with social class inequalities these objectives could perhaps be specifically added to these targets.