Moving beyond individual choice in policies to reduce health inequalities: the integration of dynamic with individual explanations

Abstract Background A strong focus on individual choice and behaviour informs interventions designed to reduce health inequalities in the UK. We review evidence for wider mechanisms from a range of disciplines, demonstrate that they are not yet impacting on programmes, and argue for their systematic inclusion in policy and research. Methods We identified potential mechanisms relevant to health inequalities and their amelioration from different disciplines and analysed six policy documents published between 1976 and 2010 using Bacchi’s ‘What’s the problem represented to be?’ framework for policy analysis. Results We found substantial evidence of supra-individualistic and relational mechanisms relevant to health inequalities from sociology, history, biology, neuroscience, philosophy and psychology. Policy documents sometimes expressed these mechanisms in policy rhetoric but rarely in policy recommendations, which continue to focus on individual behaviour. Discussion Current evidence points to the potential of systematically applying broader thinking about causal mechanisms, beyond individual choice and responsibility, to the design, implementation and evaluation of policies to reduce health inequalities. We provide a set of questions designed to enable critique of policy discussions and programmes to ensure that these wider mechanisms are considered.


Introduction
Current UK policies designed to reduce health inequalities through preventing non-communicable disease are based largely on an individualistic epistemology. The dominant epistemic assumption (i.e. the assumption about what constitutes admissible evidence to guide action) in the implementation of public health programmes is that human behaviour is a major determinant of health, and that behaviour is largely a matter of individual choice; individuals are therefore responsible for their own health and for making health-related behaviour changes. Reliance on such simple, linear causal explanations and of proximal risk factors in the causes of disease has consistently led to a focus on changing individual behaviour. [1][2][3][4][5][6] We argue for a greater consideration of how more complex relational and dynamic factors, beyond just the individual, impact on health. A number of disciplines including sociology, 7 history, 8 biology, 9 neuroscience, 10 philosophy 11 and psychology 12 have produced evidence of potential mechanisms. The existence of interactions with place and social context, power, economics, institutional relationships and biology, over time and across generations have been widely cited. 2,[13][14][15][16][17][18] Developments in biology include epigenetics 9 and the neuropsychology of executive functioning; 12 and both are highly relevant to understanding how health inequalities are sustained. These remain largely outside the purview of current policy interest.
While some of this evidence may sometimes find its way into policy statements it is seldom realized in interventions. 3,19,20 The six documents chosen for analysis in the study demonstrate that the main assumptions underpinning interventions to address health inequalities have not changed significantly since the mid-1970s, and remain focussed on individual responsibility. The policy interventions put forward reflect this partial view of the totality of the evidence about the generation and amelioration of health inequalities and the social patterning of disease. We highlight this constancy and propose initial steps to change the dominant view with an alternative approach.
We argue that the adoption of a dynamic, relational epistemology is fundamental to a fuller realization of causal mechanisms and for the design and implementation of more effective interventions to address health inequalities. The success of tobacco control in the UK stems from its base in the dynamic and relational aspects of practices linked to tobacco consumption. This includes, as well as individual choices, both the role of markets and advertising, and explicit action to counteract 'dirty tricks' of the tobacco industry over time. By dynamic we refer to changes over time, and by relational we mean the need to consider the relationships involved in the practice of smoking which go beyond the individual just physically smoking a cigarette.
Three themes are central to this approach: 'power', in exploring the relations between groups and between groups and institutions in society; 'history', in investigating how relations change or are sustained over time; and the dynamic 'relationship between the biological and the social'. We explore these themes below.
We first identify evidence of the importance of relational and dynamic factors in understanding the problem of health inequalities. Second, we examine policy approaches over several decades concerning health inequalities, demonstrating the repeated focus on individual behaviour change. Finally, we develop a set of questions which highlight this broader perspective, to sensitize policy-makers, service developers and researchers when developing policies to address health inequalities.
Moving towards more dynamic thinking: power, history and the relationship between the biological and the social Power Health inequalities are seldom described in policy documents in terms of power relations and competition for scarce resources between classes, genders and ethnic groups, or the mechanistic consequences of these power relations on biology. The idea of 'empowering' the individual to make healthier choices is a central tenet of most policy proposals, but discussions of the impact of power dynamics between groups and institutions are overlooked, underplayed or ignored. The interactions and intersections between different groups are essential to the production of social structures and forms of inequality. Power is a central part of these interactions, but individualistic forms of analysis are poor at elucidating such dynamics. The focus on individual lifestyles, for example, precludes any meaningful analysis of the power dynamics in which individuals are involved and that, in turn, influence the types of choices they make.
Power relations are intrinsic to social life. 21 At their most basic, power relations affect people's access to resources, including health services, and determine the lived experience of discrimination, disadvantage, bullying, harassment and social exclusion. They influence educational and employment opportunities, including the type of work people are able to do and their contracts of employment. The associations between these things and their links to health inequalities have been known for decades 22 but the mechanisms of their direct impacts on people's lives have not featured in government policy documents. Of course, attempts to promote equality have been part of government policy but these have tended to focus on protected characteristics rather than on the nature of the damaging or health protecting nature of the relations themselves. 23 Following Sen we should consider the degree to which health policy and its implementation enhance the capacity for health, for example. 24 The historical perspective Likewise, history and a longer time perspective are largely absent from policy discussions of health inequalities. Insofar as history appears it is in terms of a progressive narrative about public health advances since the 19th century. [25][26][27] There is a form of institutional memory loss evident in policy where similar ideas are rediscovered and recycled over a relatively short period of time, 20 and policies demonstrate historical amnesia about the processes which have led to contemporary health inequalities. This leads to, for example, the assumption that health inequalities can be explained by current societal conditions alone and ignores endemic social arrangements which persist over decades and longer. It leads to overoptimistic expectations of performance of likely effective programmes, with rapid payback being expected and public health initiatives facing cuts if they do not produce results swiftly. An historical perspective in policy and research informs a clearer understanding of the longer-term intertwining causes, construction and maintenance of health inequalities. According to Raadschelders, all too often 'History … is regarded as a 'past' that can be recorded for its own sake but has little relevance to contemporary challenges. This view of history is the product of a diminished and anemic sense of time, resulting from organizing the past as a series of events that inexorably lead up to the present in a linear fashion'. 28 We argue that bringing a historical perspective provides a richer understanding of causation that recognizes the 'layered' nature of how the past leaves its mark on the ways health inequalities are reproduced over time. This perspective helps to illuminate the tools available to policymakers and the interventions which could have a lasting impact. [29][30] Social epidemiology has long shown recurring patterns of health disadvantage geographically lasting generations in many towns and cities in the UK. Notwithstanding policy efforts, these patterns have remained stubbornly fixed. Generation on generation, the consequences of the historical reproduction patterns of health differences remain significantly unchanged. It is no longer infectious disease that causes premature mortality in the poorer parts of the country, but non-communicable disease. Absolute death rates are lower than the 19th century but relative differences remain. Meanwhile policy is fixated on simple behavioural solutions not on the reasons why the structural differences are so intractable. 31 The biological and the social Descriptions of the role of biology and its interplay with social factors in producing population patterns of disease, was associated originally with Engel 32,33 and later with Barker. [34][35][36][37] The relationship across generations between the environment and biological moderation of gene expression (epigenetics) is now providing new evidence of interacting social and biological mechanisms underlying the transgenerational transmission of inequalities in health, 2,38 leading to a re-consideration of crossgenerational effects on poor health and amelioration or exacerbation through social conditions. It suggests a reemphasis on programmes to support maternal, foetal and child health as a key strategy in breaking the cycle of inequality in health.

Policy
We now turn to an examination of existing policy approaches to addressing health inequalities. We review the extent to which evidence of the mechanisms considered above are informing policy recommendations.
We reviewed six key policy documents published by the UK government or the English Department of Health between 1976 and 2010 25,27,39-42 using Bacchi's 'What's the problem represented to be?' (WPR) framework for policy analysis (Table 1). 43,44 The interrogation of documents using this framework allows for comparisons over time as well as the development of an understanding of individual documents and the assumptions which underpin policy rhetoric. This examination demonstrated a longstanding continued focus on individual behaviour in strategies to reduce health inequalities, and more generally to prevent non-communicable disease (see Table 2 for verbatim quotations from the policy documents 45 ).
Successive documents did show increasing reference to the impact of a wide range of factors on health (social, cultural, economic and environmental). Nevertheless, these were rarely then used for policy and intervention development or for the evaluation of programmes; in both, individualism remained paramount. For example, 'Saving Lives: Our Healthier Nation' argued that 'the causes of ill-health are many: a complex interaction between personal, social, economic and environmental Table 1 Bacchi's 43,44 'What's the problem represented to be?' framework for policy analysis

Questions in the WPR framework
Aim of question 1. What is the 'problem' represented to be in a specific policy? To understand how a phenomenon comes to be understood as a problem in social policy (problematization), including the causes of the problem.

What presuppositions or assumptions underlie this representation of the 'problem'?
To understand the discursive practices surrounding the representation of the problem (archaeology), i.e. what can and cannot be said about a problem.

How has this representation come about?
To understand the history (genealogy) of the development of understanding of a problem.

What is left unproblematic in this problem representation?
To identify silences and highlight explanations which are not discussed and to consider why these views might be excluded from this particular representation of the problem.

What effects are produced by this representation of the 'problem'?
To understand the creation of subjectivities produced by representations of problems in policies: how individuals and population groups are conceptualized.
6. How/where is this representation of the 'problem' produced, disseminated and defended? How has it been (or could it be) questioned, disrupted and replaced?
To identify where this representation of the problem has been reproduced, including in other policy documents.  • Follow a balanced diet with plenty of fruit and vegetables.
• Manage stress by, for example, talking things through and making time to relax.
• If you drink alcohol, do so in moderation.
• Cover up in the sun, and protect children from sunburn.
• Take up cancer screening opportunities.
• Be safe on the roads: follow the Highway Code.
•  There is significant scope to use approaches that harness the latest techniques of behavioural science to do this-nudging people in the right direction rather than banning or significantly restricting their choices. factors' ( Table 2, 3B). Yet the main approach offered to reducing health inequalities focused on ensuring individuals are informed about risks to their health as 'in most cases it is for the individual to decide whether to take the risk' (Table 2, 3E). The responsibility falls to the individual to make the least or less risky decision. Here and elsewhere there is a 'disconnect' between policy rhetoric about causes and recommended remedial actions. Evidence about the wider determinants of health is used to justify interventions on health inequalities, which then paradoxically concentrate on individual behaviour. Rather than addressing the wider determinants themselves, certain groups are viewed as lacking the capacity to negotiate successfully the effects of these determinants, demonstrated by their proclivity for making unhealthy choices. Thus, a number of interventions focus on the provision of opportunities to enable individuals to improve their current social and economic situation through, for example, education, training and work opportunities. 41 A common argument is that by embracing such opportunities people will have more control over their lives, which would result in them being more interested in their health and therefore more likely to make healthy choices. 42 The idea of the individual as a chooser dominates these discussions. 45 This idea shapes the understanding of the problem as one of individuals, in themselves, being incapable of making healthy choices unless steered by interventions which change their attributes as individuals who will then change their behaviours. The result has been described as 'lifestyle drift' 46 in policy measures and 'lifestyle push' from politicians and markets. 20 This focus on the individual effectively neutralizes the effects of social context and airbrushes out of the picture a number of important contextual agents and institutions-specifically the state, markets and industry. So the state and its retreat from interventionism consequent on neo-liberal economic thinking, the role of markets (as a cause of rather than a solution to the problem), and the (incidental) health damaging roles of the food, advertising and alcohol industries are conveniently put to one side. 47 Actions addressing these actors and getting them to change are not central to contemporary policy focus. The alternative view that the State has a duty to enable as far as possible everyone to have a fair opportunity to live a healthy life and that governments should try to remove inequalities that affect disadvantaged groups or individuals, including a duty of proportionate regulation, has not always been a dominant motif in policy documents. 23,48 Following Sen, among others, we would emphasize the importance of addressing how public and private actions and organizations shape the capability of individuals to make positive choices regarding their health. 49 For example, obesity as a significant public health threat and an important cause of health inequalities has become engrained in policy discourse in the last decade and a half. 19 Numerous policy documents and a raft of guidance have been produced on the topic, including the recent 2016 childhood obesity strategy. 19,50,51 Yet success in curbing obesity has been minimal. This is despite compelling evidence that an obesogenic environment is generating the obesity epidemic, 15,50 and that its structure and dynamics should be the target for arresting the epidemic. Policy solutions have persistently focused on proximal determinants, most prominently individual diet and exercise. 52 Similarly with alcohol consumption, the focus remains on individual capability and how to ensure individuals make healthy choices, 4,53 reinforced by statements by ministers past and present. 54 A most important exception to this perspective was the Foresight report on 'Tackling Obesities', commissioned by the Government Office for Science and published in 2007. The report presented evidence demonstrating the social and biological complexity of obesity, and emphasized the need to intervene beyond individuals into processes of governance and decision-making to stem rising obesity and the necessity of evaluation: 'The deceptively simple issue of encouraging physical activity and modifying dietary habits, in reality, raises complex social and economic questions about the need to reshape public policy in food production, food manufacturing, healthcare, retail, education, culture and trade'. 15 The midterm review of the Report at 3 years drew attention to 'stakeholder inertia in adopting more accurately informed perspectives of the roles of the individual and of society (which) can hinder the development of strategies and interventions to manage the current and future obesity epidemic'. Implementation of the report was terminated long before effects could be appropriately evaluated.

MOVING BEYOND INDIVIDUAL CHOICE IN POLICIES TO REDUCE HEALTH INEQUALITIES
There are a small number of other examples where what we refer to as a relational approach has found its way into the public domain. For example the NICE Public Health Guidelines on Community Engagement, 55 Social and Emotional Wellbeing of Children, 56,57 Healthy Working Conditions, 58 Preventing Harmful Drinking 59 and Preventing Cardiovascular Disease. 60 Not only were these exceptions, but also in the case of the Cardiovascular Disease and Harmful Drinking guidelines they were rejected by Ministers. One of the implications of the focus on individuals and on behaviour change is that it pushes policy interest towards proximal risk factors (which the NICE Guidelines did not) and the role of these in the aetiology of non-communicable disease in particular. However, knowledge about risk and its links to behaviour in causal pathways of disease do not on their own provide any solutions as to how to change those things. For example, knowledge about the dangers of exposure to cigarette smoke or alcohol suggests that reducing exposure would be beneficial, but does not explain how to achieve that. A focus on the dynamics of the systems involved in the relations between industry, markets, advertising, human preferences, group behaviour, as well as the individual actor provides a richer theoretical frame for developing and evaluating integrated programmes to address the problem. This has been the case with the success of tobacco control in the UK. We advocate a similar 'how to' approach with respect to obesity, alcohol and physical activity. 31

Discussion
Main finding of this study We have highlighted extensive cross-disciplinary evidence about the relational nature of health inequalities, and causal mechanisms beyond individual choice and responsibility. We have shown that this evidence rarely follows through to preventive programmes.

What is already known on this topic
Health inequalities persist in the UK driven largely by the social patterning of non-communicable disease. Interventions designed to reduce health inequalities are still primarily shaped by a focus on individual behaviour. Yet, there is substantial evidence of supra-individualistic and relational mechanisms relevant to health inequalities from a range of disciplines including sociology, history, biology, neuroscience, philosophy and psychology. This evidence is not yet applied systematically in policy or action, where it may inform the design and implementation of more effective policies.

What this study adds
A perspective recognizing the complexity of the systems in which public health must practice, including its politics, shifts us away from narrowly focussed linear behaviour change models to a focus on reflexive systems and the power of players in those systems. [61][62][63][64][65] Prioritizing the study of relationships To consider previous attempts to address health inequalities; to highlight that this is a problem with a long history rather than a contemporary issue; and to draw attention to both evidence of effect/non-effect and lack of testing over time. between people and groups of people, how these relationships are sustained and change over time, and acknowledging links between power, time and the biological and social, we suggest will bring into the discourse a wider perspective to complement the existing focus on individual behaviour change.
In order to move towards integrating this thinking into policy considerations, we have developed a set of questions to use in writing and critiquing policy papers, which aim to ensure that proposed interventions to address health inequalities take into account relational and dynamic factors, as well as individual behaviour (Table 3). These questions can be used not only by policy-makers and service developers but also by academic researchers to ensure that relational and dynamic factors are brought to the forefront of policy evaluation.

Limitations of this study
This is not a comprehensive or systematic review of all documents relating to English policy recommendations to address health inequalities.

Conclusion
Current policies dominating efforts to reduce health inequalities through prevention of non-communicable disease target individual behaviour change and have not worked well. We have argued that an individualistic epistemology limits their impact. Programmes predominantly focus on individual behaviour change foregrounding individual choice and responsibility. There exists strong and extensive evidence that interconnecting and interacting factors, beyond the individual, impact on health including place, context, power, economics, institutional relationships and biology, over time and across generations. The existence of such evidence, however, has not been sufficient to garner policy action even when existing strategies have failed to successfully address health inequalities. We propose a list of questions that researchers and policy-makers can use when writing or critiquing policy in order to bring this broader perspective to the forefront of their analysis of the problem. This is one small step in moving from the rhetoric of whole systems interventions to long-term intervention and evaluation and towards broadening the range of approaches and evidence we use to unpack the problems of health inequalities and work towards policies to address them.

Conflicts of interest
Natasha Kriznik is in receipt of funding for health services research from the Wellcome Trust. Mike Kelly is in receipt of grant funding for public health-related research from MRC, ESRC, the Wellcome Trust and NIHR. He also has one consultancy for providing general evidence-based advice on obesity prevention to Slimming World. From 2005 to 2014 he was the Director of the Centre for Public Health at NICE.