Policy

An existing Journal of Epidemiology and Community Health glossary describes policy as ‘a guide to action to change what would otherwise occur, a decision about amounts and allocations of resources’.3 This reflects the common usage of ‘policy’ to refer to the content of official statements and other documents. However, ‘policy’ can also refer to a context or broader direction (eg, ‘free-market policies’) or a process, involving multiple stages (including implementation).4

Politics

As another Journal of Epidemiology and Community Health glossary notes, politics can be defined in multiple ways.5 Within this glossary, we take ‘politics’ to mean partisan political competition, which is how it is often used within political science.

Normative and instrumental policymaking

Values and public opinion are legitimate aspects of democratic policymaking.6 Value-based, ‘normative’ policy decisions can be distinguished from ‘instrumental’ policy decisions concerning the best means to achieve certain ends.7 The latter more closely reflects notions of evidence-based policy-making. Yet values clearly play a central role in public health. Even the commonly recognised public health goals of improving health and reducing health inequalities can be in tension with one another and deciding which to prioritise is a normative decision. Moreover, the distinction between normative and instrumental policy-making is rarely clear-cut. For example, a decision to increase taxes on tobacco (or to pay people to quit smoking) may be informed by evidence of effectiveness but also involves a normative decision about whether policymakers should intervene in this manner.

Policymaking and implementation

Simply put, ‘policymaking’ involves the construction and/or implementation of specific policies. One of the most popular accounts of policymaking posits that it involves a number of linked stages.8 While the number and description of stages vary between models, they commonly include: problem identification; agenda-setting; consideration of potential actions; implementation of agreed action; and evaluation. These linked stages are sometimes considered to form a ‘policy cycle’, with evaluation potentially leading to a re-consideration of the problem.9

While the notion of policy stages offers a useful heuristic device,4 this idealistic account of policymaking is generally accepted not to reflect the messy reality, in which multiple ideas, interests, actors and values interact in a non-linear fashion and in which some ‘stages’ may be ignored (eg, evaluation).10 It also fails to acknowledge the ongoing nature of policymaking which can lead (eg, in implementation) to policies emerging in ways that differ considerably from the intentions of the original author(s).11

The stages heuristic has been further challenged by the multi-level governance literature which highlights that ‘policymaking’ increasingly takes place within a variety of interacting levels.12 Processes of globalisation, regionalisation and decentralisation are increasing the number of relevant decision-making fora.4 For example, health policies implemented in European member states may be informed by international treaties and agreements, European Union policies, national and local policies. There may be multiple entry points for influencing decisions at each of these ‘levels’ (eg, European Union policies are informed by officials at the Commission, elected Members of the European Parliament, member state representatives on the Council, and various committees and less formal groups).13 In addition, policymaking is usually an ongoing, interactive process, rather than a single decision.4

The situation is further complicated for public health (as compared with health care) policy because it involves addressing broader determinants of health which operate across multiple policy sectors, many of which have aims and values that compete with those of public health.

Policy inertia (historical institutionalism and path dependency)

Some of the most commonly used theories about policymaking focus on explaining why policies are resistant to change. Historical institutionalism has been particularly influential in this regard.14–17 It posits that policy outcomes can only be understood by considering the historical and institutional context in which decisions are made. This includes ‘the formal rules of political arenas, channels of communication, language codes, (and) the logics of strategic situations’, which all act as ‘filters that selectively favour particular interpretations either of the goals toward which political actors strive or of the best means to achieve these ends.’16 For example, the division of policy organisations into departments and sub-units with particular, demarcated foci represents the institutionalisation of past decisions and often promotes ‘policy silos’ that facilitate policy activity in narrow areas while preventing the development of cross-cutting policies.18 Such theories have been used to explain the partial and fractured influence of health inequalities research on UK policies,19 and the contrasting approaches to healthcare funding in the US and Canada.20

‘Path dependency’ is a related, though somewhat simpler, concept which has in many ways been subsumed by institutionalist theories. Originating in economics,21 the defining feature of ‘path dependency’ is the notion that previous policy decisions limit the possibilities for future decisions. This process has been famously illustrated through an account demonstrating that the common QWERTY format for typewriters and keyboards can only be understood by studying the development of the first commercial typewriters.21 This same basic idea has been employed to help explain the development of healthcare systems.22

Theories in this category usefully draw attention to the importance of temporality and history in understanding policy processes and outcomes.23 Neither ‘path dependency’ nor ‘historical institutionalism’ suggest particular policy outcomes are inevitable. Rather, both imply that it becomes increasingly difficult to change the overall direction of policy trajectories once previous decisions become embedded in institutional structures and discourses.24 Such theories can help explain why public health research may struggle to consistently shift policy debates, particularly when challenging the status quo.25 However, they do little to explain how and why policy change does occur,15 ,26 or, therefore, what role the public health community might play in transformative moments.

Incremental policy change (‘policy learning’)

In contrast to the ‘policy stages’ heuristic, Lindblom argues that policymakers ‘muddle through’ policymaking, considering a small range of policy options they deem feasible and pursuing the option with the greatest stakeholder consensus.27 ,28 Lindblom argues that this is a more accurate description of how policy develops and that it is normatively superior as it enables policymakers to learn from their growing policy experience and adjust to unanticipated negative outcomes.29 Heclo's influential notion of policymaking as a process of ‘collective puzzlement’ and ‘social learning’,30 similarly implies that policymaking is complex and that policy change is likely to be incremental. Both theories emphasise the multifarious and disorientating nature of policymaking but suggest that individual policy actors are nevertheless capable of learning.29 ,30 However, learning can be imperfect and uneven, which means that even if messages/ideas are adopted by individual policy actors, they may not necessarily be institutionalised within organisations.31 From a public health perspective, these theories suggest a potential role for evidence in aiding incremental policy learning. However, this may be limited where evidence itself is limited (eg, where there is a lack of consensus as to what works in policy terms32 ,33).

Significant policy shifts (‘punctuated equilibriums’, ‘policy windows’ and ‘policy paradigms’)

A third set of theories suggest that, while policy normally develops incrementally, significant policy shifts also occasionally occur. Inspired by biological models of evolutionary development, Baumgartner and Jones’ notion of ‘punctuated equilibriums’ posits that systems can quickly shift from one period of relative stability to another.34 They argue these ‘punctuations’ occur when persuasive ideas gain increasing attention, a situation which depends on external (political) factors as well the inherent qualities of an idea. Punctuated equilibrium theory has been used to help explain both the array of recent tobacco control policy initiatives in the UK,35 and the ‘surprising bursts’ of global priority dedicated to tackling malaria, polio and tuberculosis at various times.36

Another frequently cited theory that fits this category is Kingdon's notion of ‘policy streams’,37 which is based on his observation that key policy actors (in the US) were often unable to retrospectively explain why particular policy outcomes had occurred. This led Kingdon to concur with theories stressing the complexity of policymaking.10 However, Kingdon also challenged claims that serendipity was necessarily a key determinant, arguing instead that significant policy change can occur when three ‘policy streams’ (‘policy’, ‘politics’ and ‘problems’) converge. Kingdon's analysis differs from other theories in this category partly due to his emphasis on the role of ‘policy entrepreneurs’ who exploit the ‘policy windows’ that emerge when the three ‘streams’ converge (see below). Exworthy and colleagues employ Kingdon's framework in their account of UK health inequalities policy development.38

A third key theory fitting this category is Hall's notion of ‘policy paradigms’,39 which is informed by Kuhn's theory of scientific revolutions and based on empirical work concerning shifts in economic policy (from Keynesianism to monetarism).40 Like the other theories in this section, Hall suggests that while low-level changes (eg, the means of achieving particular policy goals) are common, occasionally paradigmatic shifts can usher in completely new ways of thinking about an issue. Securing ‘paradigm shifts’ (eg, from a medical to social model of health) is rare and Hall argues is unlikely to occur through gradual policy change/learning, being sociological and political in nature. In other words, paradigm shifts are unlikely to arise as a result of evidence alone because associated shifts in values/ideologies are also needed. The concept of ‘policy paradigms’ has been used to help explain a shift in French approaches to drug abuse from a curative, abstinence-orientated paradigm to a ‘harm reduction’ paradigm in the mid-1990s.41

In sum, these theories challenge the idea that policies only develop through incremental change, each suggesting there are also rarer, more significant shifts. Such theories provide for the possibility that public health might, occasionally, contribute to significant policy change (as well as more gradual learning). Unfortunately their practical value is limited by a lack of agreement about the precise factors contributing to significant policy shifts, although it is worth noting that none suggest evidence alone can achieve significant policy change and all indicate political competition, power struggles and values/ideologies are important. Therefore the public health community may need to expand discussions about ‘what works’ to better incorporate normative dimensions of policy debates, especially if trying to achieve significant policy changes.

Policymakers

Public health texts commonly refer to policymaking being undertaken by ‘policymakers’ without defining who policymakers are,42 implying they are a clearly identifiable, internally homogenous group.43 Yet, policy organisations are divided into a vast array of groups and sub-groups, only some of which are directly involved in constructing policy statements (others, for example, provide analytical or support services and may not consider themselves ‘policymakers’).19 ,44 What is more, multiple external actors may contribute to policy decisions via cross-sectoral ‘policy networks’ (see below), and important political and epistemological divisions can occur within policy communities.43 Hence, it may not be possible to identify when and where particular decisions were made or who was responsible.4 This is largely because policymaking is a complex, dialogical process, in which the authority of official documents often rests on the very fact that they ‘are not identifiably the work of an individual author’.45 In considering promotion of public health messages, it may therefore be important to acknowledge that ‘policymakers’ are not homogenous and messages may benefit from appropriate tailoring to the differing needs of specific policy audiences.43

Policy networks and advocacy coalitions

Another set of theories focus on the role of diverse sets of actors, or ‘policy networks’, in shaping policy outcomes. Terminology relating to ‘policy networks’ is diverse (and not always consistent).46 The concept of ‘iron triangles’, developed in the US, generally refers to stable relationships that develop between relatively few actors (typically the relevant Congressional Committee, powerful interest groups and bureaucrats).47 ,48 From this perspective, policy decisions are viewed as the outcome of negotiations within these tight-knit networks (from which others are generally excluded).47 Heclo directly challenged the importance of ‘iron triangles’, arguing that policy decisions often result from negotiations within much larger, fluid groups which he terms ‘issue networks’.49 It may be helpful to conceive of the ‘policy networks’ literature as forming a continuum, ranging from tightly defined ‘policy communities’ (such as ‘iron triangles’) at one end, through to broad, unstable ‘issue networks’ at the other.50

Sabatier and Jenkins-Smith's ‘advocacy coalition framework’ (ACF) provides a particularly specific account of ‘policy networks’ (falling somewhere in the middle of the above continuum).51 It suggests diverse groups of actors contribute to networks (eg, journalists, academics and think tanks as well as policymakers and interest groups) but that these networks are relatively stable because they form around core ideas (relating to values and beliefs about causation). The ACF posits that these shared ways of viewing the world (rather than political or economic interests) bind actors together in competing coalitions which seek to influence policy decisions. Members of dominant networks are unlikely to promote radically innovative ideas, given their shared view of the world. Hence, the ACF coheres with theories positing that sustained periods of policy stability are likely. However, it also suggests significant policy change can occur when a particular coalition's ideas are perceived to be so successful that some actors switch between competing coalitions, shifting the balance of power in relation to the ‘core ideas’ driving policy.

From a public health perspective, employing a policy networks approach emphasises the possibility of influencing policy through diverse routes (eg, via journalists, think tanks or non-governmental organisations) as well as by working directly with officials. To understand how to influence policy networks, a better understanding is required of the relationships and actors involved and the multiple entry points.50 More specifically, the ACF suggests public health advocates ought to consider the perspectives/values of dominant networks.52 Where evidence and ideas challenge a dominant policy network's values, success may depend on attracting support to competing (non-dominant) coalitions.

Knowledge brokers and policy entrepreneurs

Some theories highlight the importance of individual actors. For example, Kingdon's account of policymaking (see above) places a great deal of emphasis on ‘policy entrepreneurs’ who work to promote their preferred solutions.37 Accounts of policymaking that emphasise the role of evidence in policy tend to emphasise the importance of ‘knowledge brokers’.53 Potentially, researchers, practitioners, think tanks, advocacy groups, lobbyists and others can all function as ‘policy entrepreneurs’ or ‘knowledge brokers’ (or both). Yet, research examining how individuals work to influence policy is limited.

Ideas, evidence and policy transfer

Researchers and practitioners represent potential sources of inspiration for policy,32 as do policies implemented in other contexts.54 However, there is an increasing consensus that evidence-translation and ‘policy transfer’ are complex processes.31 ,43 This has led some to focus on ideas as the entity that moves between contexts (from research into policy and practice, or from one geographical location to another).31 ,43 Indeed, the past two decades have witnessed a burgeoning interest in the role that ideas play in policy change.14 ,15 ,25 Focusing on ‘ideas’ acknowledges the potential for translation, rather than transfer,25 and can also be used to capture some of the interactions between politics, ethics, values and evidence.55 However, the concept of ‘ideas’ is often poorly defined,56 and has been used to refer to ideologies, frames, norms or ‘paradigms’, explanatory theories and specific policy solutions. Further, because ideas ‘do not leave much of a trail when they shift’,39 it can be extremely difficult to assess whether what appears to be the translation of a particular idea is merely another idea with some similar characteristics.25 However, references to ideas help emphasise that even evidence-informed messages can be continually translated and intertwined with politics, ethics and values.43 ,55 For example, various studies reveal how corporations involved in producing health-damaging products have often represented evidence in highly misleading ways in their efforts to shape the policy environment.57

The concept of ‘framing’ represents another way of thinking about the role of ideas in policy. This involves assessing the frameworks (or narratives) being used to portray particular issues.58 Policy frames can inform beliefs and ideas about particular issues, limiting how actors perceive potential policy options and, relatedly, informing the positions of networks and coalitions. As such, they have been described as a ‘weapon of advocacy’.59