Evidence-based public health

At some basic level, evidence-based decision-making is a good thing. Even more so if the alternative is decisions based on habit, anecdote or outmoded training and education. Over the past decade or so, there has been a movement to strengthen EBPH decision-making.2 This movement is part of a larger and not unrelated emphasis on evidence-based medicine that began in the 1990s if not before.3 And in some countries, notably the UK, the interest in EBPH was almost certainly aided and abetted by the emphasis by the government of Prime Minister Tony Blair on evidence-based policy making.4 The emphasis on evidence by the Blair government, while most certainly uneven it its application sparked a considerable amount of ongoing research and reflection designed to better understand the role of evidence in policy making, including policy making in public health.5

At the same time, a great deal of work has been done and is ongoing on how best to make evidence available to public health decisionmakers. There is an enormous amount of research being done on patterns of knowledge translation, knowledge transfer and knowledge mobilisation (the term of art varies by country, field and researcher).6 ,7 This movement has given rise to the emphasis on plain language summaries, knowledge brokers and closer partnerships between researchers and decisionmakers.

But this emphasis on EBPH quickly gives rise to a paradox. Notwithstanding the emphasis on evidence, there are regular claims that governments are making health-related decisions that ignore or at least downplay the best available science. Consider the following examples. It is generally agreed that increasing the numbers of people vaccinated against a given infectious disease will create so-called ‘herd immunity’ making the vaccination programme all that more effective. Yet in a number of communities in the USA and elsewhere there are widespread exemptions from ‘mandatory’ requirements for childhood immunisation for people who do not want their children to be immunised.8

In the UK, in the face of significant morbidity and mortality associated with alcohol abuse, the government refuses to institute minimum alcohol pricing even though there is strong evidence showing the price affects consumption among those who are inclined to abuse alcohol.9 In Canada the federal government threatened to close the country's only safe injection site despite the evidence that the facility saved lives.10 In a number of Organization for Economic Cooperation and Development countries local municipal councils have decided to stop water fluoridation despite a (but by no means universal) scientific consensus that the practice has a powerful impact on dental health with few if any associated risks to health.11 In each of these cases there would appear to have been a disconnect between scientific evidence and public policy.

Why? Why do governments make decisions that, on the face of it, do not appear to reflect the best available scientific evidence? One straightforward answer is that the scientific evidence that might inform a policy decision is itself contested. Sometimes this scientific disagreement is the result of manipulation by powerful actors.12 In many cases, however, it is because there is, on the face of it, a genuine disagreement over what the science says. For example, notwithstanding the strong public health advice that reducing dietary sodium is critical in reducing the incidence of stroke, there is an ongoing, and at times rather harsh debate between experts in the field.13–15 Similarly, there are those who would challenge, on scientific grounds, the link between water fluoridation and the reduced incidence of dental cavities among children.16

Another popular answer is that ‘politics’ or ‘ideology’ crowds out evidence. On this view, the alternative to evidence-based decision-making in public health, or anywhere else for that matter, is one where decisions are made on the basis of the political or ideological convictions of the government of the day.17–19 This was a common concern raised about decision-making by the government of George W Bush and indeed is a common explanation for decision-making by right-of-centre governments that, on the face of it, do not reflect the best available science.20 ,21 This is also a common explanation for the position taken by centre-left groups who reject the scientific consensus on the safety of such things as vaccinations or fluoridated water.

Challenging the evidence-based premise

There is, however, an alternative answer to the question of why evidence-based policy is so elusive. The fact that governments sometimes make decisions that are at best inconsistent with the best available scientific evidence might be explained by the fact that the whole notion of evidence-based policy making is often based on a linear, direct, causal relationship between evidence and policy.22 There is, and arguably should be, some sort of a relationship between scientific evidence and public health decision-making. However, it is less obvious that the link is, can be (or even should be), a direct one. In other words, just as critiques of evidence-based medicine have tried to suggest a more complex relationship between evidence and clinical practice,23 there is a similarly complex relationship in public health. To better understand this complexity, there is a small but growing trend to try and understand some of the big policy decisions in public health using the tools and insights of political science and policy studies. This is because theories of policy change from political science offer a more nuanced account of the policy-making process and offer a more sophisticated account of the role that evidence does and does not play in policy making.

One excellent example of this is a recent book by Katherine E Smith, Beyond evidence-based policy in public health: the interplay of ideas.1 The book begins with a useful introduction to the rise of evidence-based policy and provides an overview of the strengths and weaknesses of the various ways of understanding the relationships between research and policy. Early on Smith offers an overview of the ‘fluctuating fortunes’ of the theory and practice of evidence-based policy. In doing so, she provides a provocative critique of the oft-repeated ‘two-communities’ conception of the relationship between those who produce evidence and those who ostensibly are expected to use it to make decisions. On this account the challenge is to overcome institutional and cultural ‘gaps’ between ‘researchers’ and ‘policymakers’. The preferred response is a varying mix of knowledge translation practices which will be familiar to anyone interested in the relationship between evidence and policy (eg, provide clear, concise summaries, develop collaborative relationships, etc). Smith shrewdly observes that this definition of the problem and the preferred solutions assumes, among other things, that researchers are able to directly respond to the questions and concerns of policymakers. It is also assumed that both sides share a common conception of, not just the implications of research findings but the issues that are worthy of research attention. She also pointedly asks whether greater use of research in policy making is always a good thing given what we now know about how scientific research is shaped by powerful economic interests. And for those looking for a more systematic account of the role of evidence in policy making, the last part of the first chapter offers a succinct and pithy overview of some of the different political science theories of policy making and policy change and how each understands the role of research evidence. This summary shows how far political science has moved beyond a simple linear, causal model of the relationship between research and policy.

The overall argument of the book is grounded in two case studies—health inequalities and tobacco control—to illustrate the difficulty in actually doing EBPH policy. Smith uses these cases to make a more general argument. She seeks to demonstrate public policy is driven less by scientific evidence and much more by research-based ideas. To do so, she surveys the political science literature on the role of ideas in policy making and then draws on an extensive body of qualitative evidence (ie, 141 interviews and an extensive analysis of documents) to look more closely at the role of ideas and evidence for each of the two case studies. In the case of health inequality, Smith argues that there are six different journeys from ideas to policy ranging from successful through recontextualised, partial, fractured and weak, to non-journeys where research-informed ideas do not appear to have travelled into policy at all (chapter 3). In the case of tobacco, as others have done,24 she argues that, rather than a simple evidence model, policy change is best explained with reference to advocacy coalitions and, more specifically, how the tobacco control coalition successfully deployed a series of policy frames which helped expand support for specific policy and programme initiatives. In doing so, however, she makes the perceptive and provocative point that on a number of occasions tobacco control measures were introduced before the research evidence was available. Note that a similar pattern can be found in other areas of public health notably harm reduction and illegal drugs.25 ,26 Having made the case for the primacy of ideas (rather than evidence), the three subsequent chapters develop a second core argument of the book namely that the relationship between research and policy is, following Rein, one of ‘interplay’ or, as Smith puts it a “continual exchange and translation of ideas” (ref. 1, p.75). She elaborates this to suggest that there are four types of ideas: (1) institutionalised, (2) critical, (3) charismatic, (4) chameleonic, institutionalised ideas which have become ‘unchallengeable’ and embedded in policy and discourse; critical and charismatic ideas, which in different ways may challenge the status quo and usurp institutionalised ideas; and chameleonic ideas, which can simultaneously appeal and be acceptable to a variety of policy actors but also challenge existing policy.

Public health meet political science

Smith's excellent book is part of a larger body of literature that seeks to apply the insights of political science to public health policy making. There is a large political science literature that seeks to understand the process by which public policy is made and to explain policy change be it major or, what is more likely, incremental.27 A small number of political scientists and public health specialists have begun to use this body of theory and research to better understand public health policy. A comprehensive survey is beyond the scope of this essay but useful overviews are available elsewhere.1 ,28–30 However, a few examples will demonstrate the trend.

A prominent theory of policy change is the advocacy coalition framework (ACF) first developed by Sabatier and extended and modified since.31 ,32 So, for example, Eric Breton and his colleagues have used the ACF, a well-known approach to explaining policy change, to better understand tobacco policy in the Canadian province of Québec.21 Similarly, Exworthy uses the ACF and other models of policy making to offer an alternative account of how governments might develop policies and programmes to address the social determinants of health and the associated constraints and opportunities facing policymakers.33

Kingdon's multiple streams approach is another model of the policy-making process that is increasingly used to better understand a range of public health policies. To take but three examples, this approach has been used to better understand a range of municipal-level public health policies in Sweden,34 the implementation of health-in-all policies in South Australia,35 and the efforts to implement a housing first approach to homelessness in Canada.36 However, it is important to note that, in these cases as in others,37 ,38 the authors selectively and partially deploy Kingdon's notion of a policy entrepreneur or the concept of a policy window, without necessarily embracing the other elements of the approach to explaining policy change.39

Other insights from political science and other social sciences have also been applied in public health. For example, the political scientists Donley Studlar and Paul Cairney have performed a considerable amount of research on tobacco control drawing very explicitly on theories of policy change.40–42 While middle-level theories of policy change as developed by Kingdon, Sabatier and others are predominant in contemporary political science, there are other theories of policy making and politics more generally that offer a more critical if not radical perspective on policy making including for public health. So, for example, Dennis Raphael has drawn on work that emphasises social class and the power of the business and corporate sector to better understand the limited impact of our increased understanding of the social determinants of health and the importance of healthy public policy.30 ,43 In a related vein, Carles Muntaner and others have written extensively on the implications of the different kinds of welfare states and different political parties for the basket of policies that states adopt and their concomitant implications for population and public health.44–46

However, it would be a mistake to overemphasise the extent to which the public health enterprise (or at least public health scholars) draw on the insights of political science. Systematic reviews reveal that there remains limited mutual learning between political science and public health. Based on their review, De Leeuw and Breton argue that the body of knowledge developed by political science “has still made little inroads into health promotion policy research” (ref. 25, p.87). Similarly, Oliver et al47 reported that when considering the use of evidence by policymakers few health policy studies provide clear definitions of policy, policymaker or evidence. Bernier and Clavier conclude that public health researchers “still have a naïve, idealistic and narrow view of public policy that is detrimental to understanding public policy as a determinant of health” (ref. 26, p.110).

As an infamous political theorist and revolutionary once asked, what is to be done? How can students of public health policy deploy a more nuanced conception of the role that scientific evidence does, and does not, play in the development of public health policy? As should be clear by now, my main suggestion is that the public health community should systematically engage with the insights of scholars in political science and policy studies. This can take the form of different kinds of research collaboration and training. However, this work is hard to do insofar as political scientists and public health scholars and professionals, while they may share an interest in public health policies as they are related to say, tobacco control or the regulation of dietary sodium, approach these issues in sometimes very different ways. Managing this creative tension requires imagination and flexibility on the part of research funders and public health managers.48 In effect, it requires public health to expand its reach beyond the cast of usual suspects.