Introduction

Health inequalities between groups within populations defined by place of residence, race, ethnicity or culture, occupation, gender, religion, age, education, income or other measure of socioeconomic position (SEP) are widely observed (Marmot et al, 1978; Townsend and Davidson, 1982; Charlton and White, 1995) and, in many contexts, growing (Adams et al, 2006). Reducing health inequalities has become an important objective of governments worldwide (US Department of Health and Human Services, 2000; Department of Health, 2004). However, evidence for strategies to reduce health inequalities is limited and systematic reviews have failed to offer substantive analyses or contribute to theory (Arblaster et al, 1996; Gunning-Schepers and Gepkens, 1996). Health inequities are inequalities or differences in health between populations or groups within populations that are considered unfair and avoidable (Kawachi et al, 2002; Macinko and Starfield, 2002). The term ‘inequality’ is used throughout this chapter, although it is impossible to make the necessary value judgements to designate observed differences as ‘unfair and avoidable’ in the space available, nor is it central to the arguments presented here.

Contemporary health strategies at international, national and local levels share the twin aims of improving overall health and reducing inequalities in health between groups within the population. Although common sense may suggest that these aims should be achievable in tandem, an intervention that improves the health of a population overall may also increase inequalities in health (Macintyre, 2000; Victora et al, 2000; Mechanic, 2002; Tugwell et al, 2006).

Variations in the provision of, and response to, many interventions according to SEP have been noted and, to date, have generally been described in terms of the ‘inverse care law’ (ICL) (Tudor Hart, 1971). However, the ICL narrowly describes variations in provision of medical care according to need, not variations in effectiveness of a range of interventions according to SEP. In this chapter, we review past attempts to theorise this problem and present evidence suggesting that inequalities can be introduced at all stages of the planning and delivery of interventions that affect health. When such inequalities are patterned according to socioeconomic variables, they may be due to reliance of interventions on voluntary behaviour change and other limitations of intervention design.