Structural public health: reconsidering the ‘social’ and ‘determinants’ in the social determinants of health

The social determinants of health movement and social epidemiology, the science which underpins it, have become flagship fields in public health for their emphasis upon reducing disadvantage and promoting social justice. With the central aim of reducing health inequalities between social groups, proponents of these approaches advocate for significant (usually government) intervention across most sectors of society. These aims are justified on four interlinking grounds. (1) That epidemiological indicators like income, education and occupation level are determinants (or causes) of good and bad health. (2) That shifting the distribution of these indicators will result in accompanying shifts in health status. (3) That modifying the underlying social structures is a more effective and compassionate approach than making individuals responsible for shifting their behaviours on their own. (4) That making these structural changes contributes to social justice. In this thesis I argue that each of these four propositions is false. (1) Indicators like income and education level show strong associations with health but are homogenisations of widely diverse individuals into constructed categories and therefore do not capture any relevant causal mechanisms. (2) Shifting the distribution of these indicators is likely to have wide-ranging consequences yet the imprecision of causal claims frustrates prediction of intervention effects. (3) Regulating individual behaviour through structural change rather than behavioural interventions is not less disciplinary. The archetypal structural intervention – the introduction of universal sanitation and clean water in mid-nineteenth England – was pursued in order to maintain the productivity of a labour force working under harsh factory conditions. (4) Epidemiological evidence provides a poor basis for a normative theory of social justice, lacking the required consistency across different groups, different illnesses and different times. Social epidemiology provides an interesting description of associations between abstract indicators of status and population health. It does not demonstrate, however, that its relationships are causal or its interventions desirable. Considerably more work is required to establish the proper relationship between structure and health and to show that changes to this structure could be effective and ethical. Some small steps are taken in this direction using Bourdieu’s account of structure as habitus to better understand the interaction between structure and health.