Brief summary of research findings
Dr. Katherine Smith University of Edinburgh
Aims: This project set out to study the relationship between the mass of research on health inequalities in the UK and the policy responses which emerged in Scotland and England between 1997 and 2007, a period in which official commitments were made in both countries to reducing health inequalities and to employing research evidence in policymaking.
Methods: Review of research literature, documentary analysis of 42 major policy statements, and 61 interviews with academic/other researchers, policymakers, research funders and journalists.
How has health inequalities research influenced policy?
- Very little evidence was found to support the notion that policies to tackle health inequalities have been based on available research in either Scotland or England but the data do suggest that research-based ideas have travelled into policy. Whilst this may seem like a relatively simple distinction, it is crucial because once ideas are separated from the evidence on which they are based, they become far more malleable entities, transforming as they move. The clearest illustration of this process involves psychosocial determinants of health inequalities, a research-based idea frequently visible in policy contexts but largely only in connection to an emphasis on social capital, confidence and stress and not (as much of the research implies) to inequalities in income (or other social or material determinants).
- Some ideas about health and health inequalities have demonstrated a remarkable degree of persistence over the past thirty years (e.g. medicalised, individualised and risk based approaches) whilst others appear to have faltered or splintered as they moved from research into policy (particularly those concerning ‘upstream’ social and economic determinants).
- These different journeys (or fates) can be explained by focusing on 3 idea types ideas:
(i) institutionalised ideas (embedded in the organisation of institutions and accepted as facts);
(ii) charismatic (transformative) ideas (which persuasively challenge institutionalised ideas);
(iii) chameleonic (transformable) ideas (which succeed in influencing policy discourse because they are easily amenable to varying interpretation and, therefore, adaptable to different points of view, yet these constantly changing interpretations limit their influence).
- The data suggest that, for health inequalities in the UK, some ideas (e.g. medical models of health and market-based approaches to policy) had become so extensively institutionalised that the intellectual space from which charismatic ideas might have been expected to emerge has been constrained.
- Hence, in the decade following 1997, the ideas about health inequalities which moved successfully from research into policy were either those which posed no challenge to institutionalised ideas or those with the metamorphic qualities of chameleonic ideas.
How did the relationship between health inequalities academics and policy actors function?
- Most health inequalities researchers wanted to influence policy and many had extensive and/or regular contacts with civil servants.
- Far fewer had regular contacts with broader ‘policy actors’, such as politicians, the media, think tanks or advocacy groups (e.g. major charities and campaigning organisations).
- Most academics felt under a growing pressure to demonstrate engagement with policymakers but they generally interpreted this to mean engagement with civil servants and ministers (and not the broader kinds of policy actors outlined above).
- Some academics were wary about engaging with the media, both as a result of previous negative experiences and recognition that media appearances were not always well regarded by academic colleagues. Despite this, most academics were keen to engage with the media but felt unsure how best to do this for health inequalities related work. Interviewees based in the media also suggested it was difficult to gain sufficient editorial interest in health inequalities.
- The interviews with academics suggest that the dominance of institutionalised and chameleonic ideas is encouraged by the fact that, when promoting their ideas to policy audiences, researchers are extremely conscious of the need to be able to secure future research funding/posts. There was a widespread perception amongst academics that if they were overly critical of government policy, they risked being labelled ‘radical’ and losing policy and funding credibility. Hence, when engaging with policymakers, academics often described deliberately constructing less critical, more ambiguous ‘policy messages’ than when engaging with some academic colleagues.
- In addition, the research highlights that policymaking organisations have a very limited institutional memory (due to the short time-frames within which policymakers are often required to work and rapid staff turnover). This enables the same ideas to be constantly re-cycled. Academics often have little incentive to highlight the recycling of ideas, given the pressures they feel under to generate research income.
- Research tends to influence policy through the translation of ideas, which can (and often do) change as they move between actors.
- Researchers need to recognise this and put more resources into tracking and responding to the evolution of ideas with which they are associated.
- Policymaking organisations lack institutional memory. Addressing this could avoid resources being spent on research that is unlikely to offer much that is significantly new.
- If researchers and funders want to encourage the development of innovative research and ideas, they ought to be at least as concerned with the influence of policy on research (which tends to cause researchers to focus on short-term policy goals) as the other way around.
- If policymakers genuinely want researchers to offer clear policy messages, there needs to be a greater acceptance of the role and utility of challenging ideas.
- If researchers want to promote challenging messages beyond academia, they should consider the potential role of broader policy actors (particular advocacy organisations, think tanks, politicians in opposition and the media) as well as that of civil servants and ministers.
In addition to reviewing the health inequalities research literature, a two-part qualitative methodology was employed, consisting of: (i) the documentary analysis of 42 relevant policy statements (25 from England and 17 from Scotland); and (ii) 61 semi-structured interviews with: academic and other researchers; civil servants; ministers; policy advisors; research funders; public health practitioners; and journalists. The data from both approaches was combined, first to trace the presence of research-based ideas about health inequalities in policy and then to attempt to understand why some of these ideas appeared to have been far more influential in policy than others.
I) Ideas-based policy: Despite official commitments to employing research in policy responses to health inequalities in both Scotland and England during the study period, there are extremely few examples within the data to support the notion that research evidence has informed policy. However, this is not to say that research evidence has not influenced policy at all. Instead, the thesis argues that it has been through the movement of research-based ideas that health inequalities research has influenced policy. Whilst the observation that ideas (or knowledge-claims) are central to understanding the relationship between research and policy is far from new (e.g. Bartley 1988, 1992; Knorr-Cetina 1981; Rein 1980), it is worth re-stating in the context of the discussions about ‘evidence-based policy’ that were taking place during the study period. The crucial point in making such a distinction is that, once ideas become separated from the evidence on which they are based, they are far more malleable entities. Indeed, as Latour (2005) insists, for ideas to move between actors and across boundaries, they must be translated. So, unlike metaphorical batons in a relay race that can be passed from one actor to another, the movement of ideas is more comparable to a complex game of ‘Telephone’. Consequently, whilst many of the well-known theories about health inequalities are identifiable within policy contexts, they have all undergone varying degrees of transformation in their journeys into policy, some far more than others.
II) The varying journeys into policy of research-based ideas about health inequalities: The following six distinct journey types capture the varying ways in which research-based ideas about health inequalities have moved into policy:
(i) ‘Successful journeys’ involve ideas that are visible both in policy rhetoric and proposed interventions and which appear to have changed very little in their path from research into policy. Furthermore, these ideas are applied to policy interventions in ways which are consistent with their theoretical construction in the research literature. Consequently, the ways in which these ideas are articulated in policy contexts is not dissimilar from the ways in which they are described in the research literature (or by academic researchers in interviews). Only one example of a ‘successful journey’ was found and this relates to the importance of intervening in the early years of life.
(ii) ‘Re-contextualised journeys’ involve ideas which seem to have been applied to policy interventions in ways which the research evidence does not necessarily support. For example, whilst research evidence suggests inequalities in healthcare can play an important role in exacerbating health inequalities, there is very little support in research for the policy decision to focus on the health services as a central means of tackling health inequalities. Nevertheless, this body of work appeared to have been reinterpreted (i.e. translated) within policy contexts in ways which suggested it did.
(iii) ‘Partial journeys’ refer to ideas that have not been substantively transformed during their journey into policy. However, these ideas appear to have exerted far more significant influence on policy rhetoric than on related policy interventions. The key example of such a journey related to social and economic determinants of health inequalities.
(iv) ‘Fractured journeys’ are those in which the translation and transformation of ideas is most overt. In these journeys, it is apparent that accounts of the idea (or set of ideas) within policy contexts are substantively different from (and at times even in conflict with) descriptions within the research literature. Consequently, whilst frequent references within policy statements to terms associated with an idea might suggest that it has ‘successfully’ travelled into policy, further analysis reveals the ways in which such ideas are conceptualised within policy contexts differ significantly from the ways in which they are articulated by researchers. The most overt example of this kind of transformation involved psychosocial ideas about health inequalities, which appeared to have travelled into policy in ways which emphasised only the importance of notions of social capital and social support but which did not relate this to the importance of tackling inequalities in socio-economic factors.
(v) ‘Weak journeys’, as the name suggests, refer to ideas which are only just detectable within policy and, hence, appear to have exerted only minimal influence.
(vi) The term ‘non-journeys’ was employed to describe ideas that were clearly visible in the health inequalities research literature but which were not present in the policy data.
Table 1: The varying journeys into policy of research-based ideas about health inequalities
|Journey type||Research-based idea(s) about health inequalities*|
*Categorisations are based on a review of the existing research on health inequalities, which forms Chapter Two of the thesis.
III) Explaining these contrasting journeys
(i) Constructions and understandings of health inequalities
The way in which health inequalities have been constructed and understood as a policy problem has inevitably shaped potential policy responses. As Graham and Kelly (2004) emphasise, a conception of health inequalities as the consequence of ‘health gaps’ resulting from the ‘health disadvantage’ of some groups (rather than as ‘social gradients’ which traverse society), encourages policy responses which focus on trying to achieve health improvement amongst particular groups or areas. This contributes to a blurring of the twin policy aims of ‘improving health’ and ‘reducing health inequalities’, enabling research-based ideas relating to health improvement (particularly lifestyle-behavioural interventions) to be ‘re-contextualised’ within policy as logical responses to health inequalities. This situation appears to have been further exacerbated by the short-term nature of national targets for reducing health inequalities (in both countries) and the decision to place responsibility for meeting these targets with NHS bodies.
(ii) Academic research – a restricted arena for the emergence of charismatic ideas?
The possibilities for undertaking research on health inequalities during the study period were frequently presented positively in comparison to 1979-1997, when Conservative governments were in power. Nevertheless, the data from interviews with health inequalities researchers suggest many felt unable to approach the issue as freely as they might have liked (or as is often assumed in literature concerning the relationship between research and policy) even after 1997. Rather, many reflected that they pitched proposals for, and wrote-up accounts of, research based on their perceptions of what would, and what would not, be deemed credible amongst the following key audiences: other academics; organisations/individuals with the potential to fund research; policymakers; and sometimes (although much less frequently) the media. This demonstrates that it is necessary to think about the relationship between research and policy as an ‘interplay’ (Rein, 1980), rather than as a unidirectional movement of ideas from research into policy, and that the role of research funders in this ‘interplay’ also requires attention. This raises some important questions about the impact of calls for ‘evidence-based policy’ on the production of research-based ideas. Echoing Hammersley’s (2003, 2005) concerns, the data suggest that the promotion of the notion policy ought to be better informed by research may well have contributed to an ‘imaginative squeeze’ in academia. For the flip-side of this approach has been a pressure on researchers to become more attuned to, and informed by, ‘policy needs’, thus reducing the potential for innovative, transformative (or ‘charismatic’) ideas to emerge from research. Combined with an increasing pressure to disseminate academic work to ‘research users’ and the media, this situation helps explain why several of the academic interviewees claimed to have packaged and promoted ideas in ways which made them appear less challenging to policy than they had the potential to be. This process contributes to explaining both the ‘fractured’ and the ‘partial’ journey types that were identified.
(iii) The constraining effects of institutionalised ideas
The interviews with policy-based interviewees suggest a risk-based, medical model of health is deeply institutionalised within the policymaking bodies responsible for health inequalities in Scotland and England. This significantly shapes the potential routes into policy that research-based ideas about health inequalities are able to take, resulting in a situation in which only ‘bounded innovation’ is encouraged. In other words, ideas which can easily be fitted within risk-based, medical models of health are likely to be translated into policy far more easily than those which present (or require) alternative ways of thinking about health. The interview material suggests that this situation may be exacerbated by a lack of any formal interface between research and policy, resulting in a dependence on bi-lateral relationships between individual researchers and policymakers who are operating from specific divisional locations. In this context, it is not surprising that holistic, cross-cutting ideas about health inequalities (such as those relating to socio-economic and psychosocial determinants) have encountered significant barriers in their journeys into policy. Thus, this helps explain both the ‘re-contextualised’ journeys of ideas concerning the role of lifestyle behaviours and health services and the ‘partial’ journey of ideas concerning socio-economic and material determinants of health inequalities.
(iv) The limited institutional memory within policymaking organisations
Once an idea has travelled into a policy context, the lack of vertical and horizontal connectivity within policymaking institutions serves to limit its circulation. Furthermore, a lack of institutional memory within policymaking organisations (caused by rapid staff turnover and short-term deadlines) enables ideas that have previously been circulated to re-appear as ‘new’ ideas. The combination of the lack of policy connectivity and institutional memory mean that the translation of an idea between research and policy does not necessarily secure its translation into policy in any broad sense. Instead, the same research-based idea might travel into policy through various different routes, or recurrently over time, potentially being translated (and therefore understood) in a number of contrasting ways and without necessarily having any significant influence on policy outcomes. This helps explain how and why the ‘fractured journeys’ identified might have occurred.
(v) Wider political and social ‘contexts’
Finally, the data demonstrate that many of the interviewees believed the wider social and political ‘contexts’ in which they were situated were relatively hostile to the reduction of health inequalities. More specifically, a significant number of the interviewees said they believed that these ‘contexts’ acted to ‘block’ some of the most widely supported research-based ideas about health inequalities, namely a belief that material and socio-economic deprivation, or inequalities in these factors, are the underlying cause of health inequalities. The word ‘contexts’ has been placed in inverted commas because the interviewees’ descriptions of what these ‘contexts’ involved were vague, encompassing references to global economic processes, financial institutions, the mass media and social and political cultures/beliefs. The data collated for this thesis do not facilitate a detailed exploration of this issue but they do suggest that the way interviewees acted and interacted, based on their perceptions, is likely to have contributed to the ongoing domination of particular ways of thinking about health inequalities. Two perceptions seem particularly important: one is a belief, or acceptance, that a medical approach to health ought to be (or is) accorded higher status than other types of knowledge; the other concerns the centrality of the economy to public policy. Each of these appears to have been institutionalised well beyond the physical organisation of policymaking bodies, shaping research accounts of health inequalities as well as policy responses. Indeed, the extent to which economic discourses were evident within the language employed by interviewees from a wide range of sectors suggests that this way of thinking is so deeply embedded in society that it is almost impossible to avoid using language that reinforces the hegemony of this way of thinking.
IV) A Weberian theoretical framework: Drawing together these various explanations, an ideational theoretical framework was developed, involving three distinct genres of ideas: (i) ‘charismatic ideas’; (ii) ‘institutionalised ideas’; and (iii) ‘chameleonic ideas’. Inspired by the work of Max Weber, the first two genres are antithetical to one another. Institutionalised ideas are those which have been able to move so successfully across boundaries that they have become embedded within the organisation of institutions and the language with which actors communicate their ideas. Once embedded, ‘institutionalised ideas’ move between actors and across boundaries in ways which work to maintain, reinforce and embed their appearance as ‘facts’. This ongoing process affects the movement of ideas that have not yet been institutionalised by encouraging the translation of those ideas which complement (or at least do not overtly challenge) the institutionalised ideas, whilst simultaneously working to ‘block’ ideas which do present a challenge. Accordingly, the findings suggest that ideas about health inequalities which fit within the boundaries of ideas that have already been institutionalised (i.e. the primacy of the national economy and a medical model of health) have found it far easier to travel between actors and across boundaries (experiencing only minimal transformation in this process) than ideas which do not. This process helps ensure that ‘charismatic’ ideas, with the potential to radically challenge institutionalised ideas, are unlikely to emerge. Instead, ideas which might have become charismatic (such as psychosocial determinants of health inequalities) are promoted in ways which emphasise their vagueness and flexibility, thereby facilitating their continuing circulation despite less than hospitable circumstances. These ‘chameleonic ideas’, deliberately imbued with mercurial qualities, constitute the third strand of the theoretical framework.
Implications for research and policy
- It would be both more honest and possibly more helpful for conversations concerning ‘evidence-based policy’ to be replaced with discussions about ‘ideas-based policy’. Not only does this phrase help focus attention on the centrality of ideas to understanding the relationship between research and policy but it places a spotlight on the characteristics and qualities of different ideas. This should help highlight the influence of ‘institutionalised ideas’ on the emergence and circulation of other ideas; an influence which appears to have been effectively obscured by discussions about ‘evidence-based policy’. The purpose of such a shift would be to enable conversations to take place outside the boundaries of ‘institutionalised ideas’. In other words, a focus on ‘ideas-based policy’ might help facilitate more imaginative research spaces to develop, spaces from which charismatic ideas may be more likely to emerge. So, rather than the kinds of conversations which ‘evidence-based policy’ appears to have encouraged, in which certain ideas are treated as unchangeable facts (and, consequently, others as impossible dreams), ‘ideas-based policy’ might open up the opportunity for more radical and imaginative dialogue. Such a shift does not imply that research is in any way an unnecessary activity that could be replaced by abstract philosophical debates about the merits and deficiencies of particular ideas. Rather, as Weber (1968) argues in Science as a Vocation, the role of science in society is not to tell us precisely what we should do or how we should live, but to provide us with information that makes more meaningful choices possible.
- Given the emphasis that academic interviewees placed on the role of funding opportunities in shaping their research and communication, this is an issue which has so far merited less attention amongst academics, policymakers or researcher funders than is warranted. This is particularly so in light of the growing pressure on academics to ‘disseminate’ their work beyond academia. Such discussions are likely to involve some reflection on the desirability (or otherwise) of different kinds of academic (intellectual) spaces.
- With regards to policy, the findings point both to a need to increase the institutional memory of policymaking organisations and to improve the means by which research-based ideas are circulated within policy. These issues both relate to the way in which policymaking bodies are structured and are, therefore, inevitably difficult to address. However, not doing so is likely to result in the continuing (re)circulation of ideas which fit within existing ways of thinking and, therefore, a failure to encourage, and engage with, alternative ways of thinking. Not only do existing circumstances serve to constrain the potential for innovation within research and policy, they also encourage the expenditure of policy resources on research which is unlikely to offer anything that is significantly new.
Bartley, M. (1988). Unemployment & health 1975-1987: A case study in the relationship between research & policy debate: University of Edinburgh.
Bartley, M. (1992). Authorities and Partisans: Debate on Unemployment and Health. Edinburgh: Edinburgh University Press.
Graham, H. and Kelly, M.P. (2004). Health inequalities: concepts, frameworks and policy. London: HDA.
Hammersley, M. (2003). Social Research Today: Some dilemmas and distinctions. Qualitative Social Work, 2(1): 25-44.
Hammersley, M. (2005). Is the evidence-based practice movement doing more good than harm? Reflections on Iain Chalmers’ case for research-based policy making and practice. Evidence & Policy: A Journal of Research, Debate and Practice, 1(1): 85-100.
Knorr-Cetina, K. (1981). The Manufacture of Knowledge: An essay in the constructivist and contextual nature of science. Oxford: Permagon.
Latour, B. (2005). Reassembling the Social: An Introduction to Actor-Network Theory. Oxford: Oxford University Press.
Rein, M. (1980). Methodology for the study of the interplay between social science and social policy. International Social Science Journal, xxii(2): 361-368.
Weber, M. (1968) Science as a Vocation. In Eisenstadt, S.N. (Ed.), Max Weber – On Charisma and Institution Building. Chicago and London: The University of Chicago Press: 294-309.
 ‘Telephone’ is a game in which one person whispers a message to another, who then whispers it to another, who whispers it to another and so on. The conclusion of the game is marked by a comparison of the eventual message relayed to the final participant compared with the actual message of the first speaker (the point of the game being the extent to which messages are transformed as they are communicated).