7.8 Problems, politics, and processes: public health sciences and policy in developed countries
Oxford Textbook of Public Health
Problems, politics, and processes: public health sciences and policy in developed countries
Translating knowledge into action
Science and policy
The role of the public sector
The case of public health
Articulating a framework
Defining the ‘problem’
Definition and importance
Managing power relations
The distribution of power
Issue definition and agenda setting
Negotiating the policy process
The institutional framework
Implementation and delivery
Three scenarios for knowledge in action
The stature and significance of the public health sciences derive almost entirely from their application. In the same way that the knowledge and expertise of the clinician make sense only in the service of patient care, so the claims of public health carry weight in proportion to their direct and effective application in policy and practice. Public health is necessarily an applied discipline. The mandate is clear; in keeping with other developed countries, it has been estimated that nine preventable conditions are responsible for over half the deaths in the United States, while less than 5 per cent of health spending is formally devoted to prevention (Atwood et al. 1997). The potential for preventive action is obviously substantial.
Despite its importance to public health, however, this link between action and knowledge has received relatively limited attention in the discipline (Brownson et al. 1997). Essentially, the primary interface between the public health sciences and policy has no clear signposts and lacks a plausible and authoritative academic rationale. This gap in the academic framework of public health partly reflects the predominantly technical ethos of pubic health, particularly given the centrality of disciplines like epidemiology and biostatistics. Just as important, however, is the highly problematic nature of the knowledge–policy link in the public arena. The parallel with the clinician—knowledge in the service of care—breaks down at this point. It is not as easy to translate public health knowledge into policy as it usually is to apply clinical knowledge in the practice setting. Action in the public arena inevitably raises ethical, technical, and political issues of a complexity and intractability that are not usually faced in clinical medicine (de Leeuw 1993).
The focus of this chapter is on identifying the key parameters of the link between the public health sciences and policy, formulating a pertinent framework for addressing this all-important connection, and suggesting a series of guidelines for future academic work and public health practice in this area. While many examples will be drawn from the areas of disease prevention and health promotion, it should be noted that structural change involving the social and economic determinants of health remain part of the public health brief as envisaged here.
For the purposes of this chapter public policy in health refers to the set of principles that guide the decision-making of governments in the health sector. The field of interest extends beyond the health-care system and includes non-governmental as well as state actors. The principal focus, furthermore, is on the policy system—the overall institutional pattern within which policies are made—and, within that system, on the policy-making process (Spasoff 1999).
Translating knowledge into action
The problematic nature of the knowledge–action link can be discussed at a number of levels. In one sense this is a quite generic issue about the relationship between science and policy with a bearing upon the work of a wide range of scientists whose research might be used to inform policy (Jasanoff 1990). A second set of issues—again of a more generic kind—are those of special relevance to disciplines with a public policy brief. These concern the role of government, and the ethics and justification for administrative intervention (Gillroy and Wade 1992). Finally, there are the group of issues that are more specific to the public health sciences dealing, firstly, with their disciplinary base and, secondly, with their translation into policy and effective implementation in practice (Beaglehole 1990).
Science and policy
The traditional concept of the role of the scientist has been as of a ‘dispassionate creator of knowledge’. Under this view science is value free, its findings are of universal application, and scientists are expected to retain a position of objectivity, keeping their personal prejudices at a distance from their work and avoiding public controversy. To an important extent these prescriptions for personal behaviour serve to insulate scientific work from extraneous dispute, thus focusing argument and controversy on matters of common scientific discourse to which agreed methodologies and data can be brought to bear. This is the conventional view of the laboratory sciences.
Beyond this, however, these prescriptions for the scientific role are quite understandable given the very real uncertainty there is about the interpretation of scientific findings once they come to be applied in the public arena. In the paradigmatic scientific research setting, findings are generated under carefully controlled—frequently experimental laboratory—conditions. These conditions are not replicated in the real world where they are applied. Nevertheless, it is in these ‘real world’ circumstances that the carefully nurtured aura of value neutrality is breached because at this point scientific application touches upon the frequently competing interests of different groups in the community (Greenberg 1992).
There are two broad institutional and personal responses to this conundrum at the policy interface; Jasanoff (1990) has referred to them as research science and regulatory science respectively. According to the first, and most common, perspective, scientists retain a conventional interpretation of their role, concentrate on the generation rather than the application of knowledge, and allocate to their citizen role any broader responsibilities in the public arena (Rothman and Poole 1985). This does not mean that their work may not have major consequences for public well being, nor does it mean that their perspective as scientists does not influence their worldview. It is just that the two domains are effectively kept separate (Stallones 1982).
For a growing number of scientists, however, their work is not restricted to the research setting and the generation of knowledge. These scientists are required—as part of their work role—to interpret and apply scientific knowledge in judicial and regulatory environments. Under these conditions the norms of science are brought hard up against the public reality of controversy, of popular and judicial review, and of political and policy imperatives. While many of the core issues remain technical in character, they are debated in a relatively politicized and contentious environment (Fletcher 1997). Nevertheless, institutional arrangements of varying effectiveness have been developed to manage the interface between science and its application to regulatory issues. There are lessons to be drawn here for public health, which features as one of the central institutional strands in the emergence of regulatory science (for example, in environmental protection, occupational safety and health, and vigilance over pharmaceuticals).
The role of the public sector
An important stimulus to the growth of the regulatory function to which science is increasingly contributing has been the burgeoning of the public sector in all the advanced economies. This expansion has been both qualitative—for example, the extension into new regulatory arenas such as environmental protection—and quantitative, with the steady growth in traditional areas of concern such as health, education, and superannuation. This has occurred despite concerted attempts across the developed world by governments of all political colours to reduce taxes and to contain the size of the state sector.
What explains this growth? In particular, is there any theoretical rationale that can guide us in a normative or prescriptive sense in judging the appropriateness of state activity? Since much of this activity calls upon the professional and academic contribution of various disciplines in the natural and ‘policy’ sciences, how is it best to judge the boundaries for work of this kind?
In simple explanatory terms the growth of the public sector can be accounted for by resort to a number of theories, including demographic change, democratic values and political competition, the rise of social democracy, bureaucratic expansionism, and warfare, among others. Such theories may help to account for the variable size of the public sector in different countries and over time, for example. On their own, however, they do not provide an analytical framework within which to view the role of the state when judging the appropriateness of administrative intervention (for instance, for any specific episode of policy-making).
A central analytical insight in this context is to see public policy as principally, if not exclusively, concerned with collective action problems; according to this perspective administrative intervention of this kind—that is, public policy—is concerned with finding ways of synthesizing a ‘public interest’ from individual wants and needs. In other words, the role of public policy is to co-ordinate potentially disparate individuals towards co-operative solutions for problems that they face jointly but are unable to solve separately (Gillroy and Wade 1992).
The key economic concept that applies here is that of ‘market failure’. This means that market conditions for a good or service are such that they fail under normal circumstances to reach a level that is socially desirable. There are two principal circumstances or mechanisms at play—externalities and public goods. In the case of the first, the actions of individuals far from being solely of private significance—such as a simple purchase in the market—bear in an important way upon the collective welfare of others. Examples in the instance of public health might be passive smoking, infectious disease, or environmental pollution. In each of these cases there is a potentially adverse effect on the well being of others that results from the private actions of an individual. Although the minimization of these adverse effects—such as epidemic infection or air pollution—is likely to be the desired outcome for a great majority of citizens, their uncoordinated efforts cannot achieve this, and any private market is likely to underprovide such services. The role of public policy is to overcome this ‘collective action’ problem by regulation and by other forms of administrative intervention.
In the case of public goods, market circumstances are such that the provision of certain kinds of goods and services is both ‘non-excludable’ (difficult or impossible to exclude people from their use) and non-rival (use by one person does not restrict use by another). Thus, a health promotion campaign cannot exclude people nor can it charge people for ‘using’ it. Furthermore, use by one person (that is, exposure to the campaign) does not restrict use by another, nor does it limit the potential benefit that might be had by all from such an initiative. Under these circumstances it is difficult for the private market—which relies on charging individuals for services received—to provide such goods.
The fact that negative externalities exist, or that certain services—so-called public goods—will not be provided under normal market circumstances, does not necessarily mean that administrative intervention will follow. Nor does it dictate the level and extent of such intervention. These are determined by wider social values and by the balance of political forces. Nevertheless, such analytical criteria indicate areas where public policy might best be directed—so-called collective action problems—if it is going to make the maximum and most effective difference to achieving desired social outcomes (Lane 1993).
The case of public health
In considering the relationship between the public health sciences and policy there are two principal matters at issue. Firstly, are the disciplines typically used in the production of public health knowledge such that the results of such work have a high probability of being translated into feasible and effective public policy? Secondly, what special attempts have to be made to ensure that once usable public health knowledge is available, it is shepherded through the policy process and effectively implemented?
Despite the long involvement of public health researchers and practitioners in social action and despite their reliance on at least a rudimentary level of knowledge about social and political context, the active participation of the social and policy sciences in public health work is of fairly recent origin (Mechanic 1995). The core public health sciences have been of a more technical kind, principally epidemiology and biostatistics, together with related clinical and laboratory disciplines. Such a narrow disciplinary base has, however, come to be seen as unsustainable—for two reasons. Firstly, it is now increasingly appreciated that the traditional interventions of disease prevention need to be adequately contextualized if they are to be delivered effectively (Schmid et al. 1995). The effectiveness, as opposed to the efficacy, of such interventions is not something that can just be left to simple improvisation in the field. Secondly, with the growing salience of the chronic and non-communicable diseases in the developed countries, issues of health promotion are now much more important, and these have brought with them an irreducibly and strongly behavioural dimension to the public health sciences (O’Neill and Pederson 1992).
On both these grounds, therefore, if the true value of public health knowledge and expertise is to be realized, then the way in which public health researchers and practitioners conceptualize problems has to be informed and shaped by the social and behavioural sciences (Snider and Satcher 1997). Thus, to take an example in the area of disease prevention, it might be perfectly understandable within a conventional view of science to advocate, for instance, a particular immunization schedule for the prevention of childhood infectious diseases on the grounds of its technical superiority. But from the point of view of effective delivery, it is important to take account also of ‘contextual’ matters, such as the adequacy of information and recall systems, ease of practitioner use, and patient acceptability. For public health practice, what counts is going to be the level of uptake, particularly among hard-to-reach groups. These features, while considered contextual within a conventional view of science, become part of the disciplinary core for an applied enterprise like the public health sciences (Holtgrave et al. 1997).
Similarly, but much more starkly, with issues of health promotion, smoking behaviour, alcohol consumption, and contraceptive use, for example, are all culturally sanctioned activities that have to be placed in their broader social and behavioural context if practitioners and policy-makers are to have any chance of advancing the cause of health-promoting change. Furthermore, moving into the context of the wider institutional environment—for example, related industries and cultural systems—the canvas of public health is a broad one (Milio 1988). Indeed, there is a strong argument for saying that social structural and wider sociopolitical forces need to be taken into account if effective and sustained public health advances are to be achieved at a societal level (McKinlay 1993).
Given this commitment to intervene in the real world in order to improve health outcomes—both at the individual and the societal level—public policy becomes the primary mechanism by which the potential of public health knowledge and skills can be released and translated into effective action (Lomas 1990). Thus an understanding of political context, the policy process, and implementation and delivery is crucial to good public health practice (Williams-Crowe and Aultman 1994). An essential condition for an easy and natural linkage to policy is, as has already been argued, an appropriate conceptualization of public health problems. But while an essential contributory factor, a firm disciplinary grounding of this kind is not on its own a sufficient condition for good policy work. What is also required is an alertness to political context and a sound appreciation of the policy process. These three elements—problem definition, political context, and policy process—are the foundations of the framework to be outlined here. Each requires a substantial contribution from the social sciences, particularly sociology and political science.
Articulating a framework
A number of models have been suggested for framing the knowledge–policy interface in public health. The focus of Lee et al. (1997) for example, was on health policy in the United States. In this context they drew up a framework consisting of three formal stages of policy development—problem identification and agenda setting, policy adoption, and policy implementation—and went on to identify key cultural and institutional factors shaping health policy in the American setting. Richmond and Kotelchuck (1991), by contrast, provided a stronger linkage with the scientific foundations of public health. They saw an essential component of effective health promotion and disease prevention as being the knowledge base, along with political will and social strategy. Brownson et al. (1997) suggested four factors as important to policy development and implementation in public health—identification of health risks and preventive options, intervention development, policy development, and policy enactment and assurance.
These models have a lot in common. They all identify a series of stages or components that appear to be important preconditions for the translation of public health potential into action. In general, they also see both technical factors (for example, knowledge base) and issues of power as being significant determinants of success. Furthermore, for all of the models the policy process itself is seen as central and as encompassing a series of identifiable and distinct stages that need to be traversed if a successful outcome is to be achieved. Finally, the effective implementation and delivery of a policy or a programme are seen as intrinsic to the model—and not outside it—and therefore as essential to success.
These are the core components that will feature at the heart of the analytical framework to be outlined here. Firstly, this chapter will suggest that the essential starting point for any successful translation of public health potential into action is an adequate and appropriate definition of the problem in hand. By this is meant an adequate knowledge base, a professional consensus on both the definition of, and the likely solution to, the problem in hand, an acceptable ethical and philosophical base for intervention, and the appropriate intellectual and conceptual linkages to the policy process. Secondly, an essential precondition for effective public health action is a favourable political climate. It is the very nature of public policy that it requires the mobilization of support and the transfer of resources. This cannot be done without the appropriate political alignment of forces. This is a function of power and the garnering of support. Nor is this a matter of achieving the necessary alignment at just one point in time, for instance, to secure a particular decision. The entire policy process—the third component to the analytical framework—remains hostage to the balance of power, right through to implementation, delivery, and evaluation. Clearly ‘process’ and ‘power’ are intertwined, but this third element to the framework—policy process—can be understood as an analytically distinct dimension in which institutional, managerial, and other instrumental mechanisms are the principal focus.
Defining the ‘problem’
At its simplest level, the issue of problem definition can be stated as one of professional consensus. The issue is whether there is a professional consensus on both the importance of the problem and the best method for its resolution. Thus, the issue of tobacco control has had almost unanimous support among public health professionals as a principal focus of policy attention, at least since the 1970s. The mechanisms of disease causation and its aetiology have been well established and are beyond scientific dispute: it is a major cause of death, it is a discretionary rather than a central life activity, and it appears to be highly preventable. Contrast this with alcohol control. The excessive consumption of alcohol is implicated in a wide range of problem areas. Yet its control has been a point of some controversy among public health professionals, particularly since evidence has revealed some beneficial health effects of moderate consumption for some groups. This lack of consensus has related to the target outcome for policy (alcohol consumption levels, road traffic accident rates, mortality from cirrhosis of the liver), the importance of alcohol vis-à-vis other areas of public health concern, and the mechanisms for intervention (price, availability, consumption control, harm reduction). Furthermore, there have been competing public health, commercial, and social definitions of the significance of alcohol. Besides which, in considering alcohol as a product, while obviously discretionary, it remains intertwined with many core life activities (like eating and socializing), thus making it a more difficult area to isolate and modify through normal policy mechanisms.
Clearly, therefore, problem definition is crucial. In essence, an issue that lacks clarity of definition and professional consensus has little chance of progressing up the policy agenda. Nevertheless the issue of problem definition goes well beyond the simple achievement of scientific consensus. It is the object of this section to canvass these broader dimensions of the issue.
In the conventional view of science, disinterested investigators apply standard and accepted methodologies to data deployed in carefully controlled experimental settings. In the course of this work they generate empirical results that contribute incrementally to the advancement of knowledge. This view of science as a set of practices and procedures is allied to an epistemological perspective that sees the scientific enterprise as one that discovers objective facts and laws of universal application.
Yet, to an important extent, science is not just a set of practices and procedures; it is a living social enterprise. This relates not only to the preconditions for scientific productivity—adequate resources, functioning teams, institutional support, collegial networks, and reward systems—but even to the determination of ‘facts’ and the forging of consensus around scientific findings (Latour 1987). In the case of health research funds, for example, the allocation of resources seems to have more to do with the established hierarchies of science, political imperatives, and commercial potential than with public health significance. Thus the disciplines of public health, particularly the behavioural and social sciences, have generally been the ‘Cinderella’ of health research funding.
Even more fundamental to the scientific enterprise are sociological insights that get to the heart of scientific thinking itself. According to this perspective, scientific thought should be seen as a representation of observed reality rather than as an objective and invariant analytical system (Kuhn 1970). Thus, amendments to such systems may take on more the characteristic of a change in beliefs than just an alteration in analytical categories. This is particularly pertinent to much public health research that has increasingly revolved around risk factor epidemiology, to the detriment of attention to broader social structural and contextual forces (Pearce 1996). These contrasting analytical tendencies represent scientific preferences and belief systems about social reality, and have a long history going back to the early theoretical disputes of mid-nineteenth century public health between contagion and miasma (Tesh 1988).
The influence of social context is very much more apparent once attention is moved away from the cloistered environment of the research setting and towards the arena of application—that is, using Jasanoff’s (1990) distinction, from research science to regulatory science. This terrain is clearly a contested one. Science, the state, and a range of interest groups vie to define key public health ‘problems’, interpreting evidence, weighing alternatives, and attempting to shape the agenda (Spector and Kitsuse 1987). For example, the tobacco industry has been particularly active in trying to influence the debate on passive smoking effects, in much the same way as the pharmaceutical industry has across a wide range of scientific arenas. The stakes are high and the traditional mechanisms of peer review and of open, dispassionate debate are not necessarily a match for successfully arbitrating deeply entrenched ideological differences, whether these derive from political, scientific, or commercial preferences (Greenberg 1992).
An important development in generating professional consensus has been the emergence of the evidence-based medicine movement and the growing use of reviews, meta-analyses, and consensus conferences (Lomas 1991). These devices, together with more deliberate strategies for achieving professional consensus, such as pilot studies, formal hearings, tribunals, and commissions of inquiry, increase opportunities for canvassing alternative evidence and interpretations. A particularly interesting development has been the emergence of formally designated centres for the assessment of health technology across a number of countries. Criteria for the assessment of health technology have been developed and the field has become rapidly professionalized, with an emerging consensus on procedures and key findings.
The development of the evidence-based medicine movement has done much to systematize the search for agreed criteria of evidence (Muir Gray 1997). To the traditional criteria for establishing causal linkage have been added an adjudication on levels of evidence from published studies, and a sophisticated discussion about the methodological strengths and weaknesses of different research designs. It is these more conventional hallmarks of scientific work, together with agreement on their theoretical underpinnings and wider significance, that provide the preconditions for professional consensus on any given public health problem.
Definition and importance
One useful outcome of the ‘sociology of science’ approach outlined in the previous section is the highlighting of the importance of certain core, definitional activities in making sense of a problem area for the professional community of interest. Thus, ‘cot death’ or sudden infant death syndrome, framed a potentially esoteric issue—unexplained infant deaths—in such a way that it became more meaningful within policy, clinical, scientific, and even popular discourse. In a similar fashion, the framing of public health questions in terms of risk behaviours sets the context for a specific group of scientific, and consequential policy, activities. For example, alcohol abuse considered at the clinical level leads to a focus on personality and biological predisposition; at the population level, by contrast, issues of availability and consumption patterns come into focus (Jeffery 1989). A similar argument applies with a preference for certain methodological techniques and research designs emphasizing conventional criteria of rigour; in these instances, approaches that use qualitative research techniques and that try to understand the perspective of key stakeholders may be overlooked (O’Neill and Pederson 1992).
Other broader definitional processes may be at work in the emergence of key public health problems to policy prominence. Feminists have argued, for example, that issues of importance of special interest to women, for instance, domestic violence and menstruation, are not given high priority, a point that may be reiterated for other areas that are of interest to less powerful groups, such as minority ethnic groups and the elderly. Thus, there are social parameters to problem selection and definition that may influence the course of scientific progress and the availability of relevant information for policy formulation (Mechanic 1993).
The likelihood that an issue will be defined as an important and relevant one has much to do with the wider policy process. Political salience, public visibility, personal immediacy, perception of threat—all these are likely to contribute to an issue surmounting the threshold of policy relevance (Rochefort and Cobb 1993). But what about criteria within the professional community—is there agreement at this level on the relative importance of public health issues?
Impact on the heath of the community is clearly a key criterion of importance. Methodologies such as the burden of disease have been significant in establishing a sense of priority among broad competing areas of public health concern (Murray and Lopez 1996). Within these broader categories, the population attributable risk is a marker of significance (Northridge 1995). Where the proportion of a population exposed to a particular risk is high (for instance, a fifth) and where the relative impact of that risk is also high (for instance, a factor of 3) then the argument for public health action in the case of a relatively common disease or disorder can be said to be a strong one.
The scale of the health problem at issue is clearly a precondition for action. Another essential criterion is modifiability or tractability. Although this dimension has not been quantified in a standardized way across problem areas, for the purposes of assessing public health importance it has to be addressed. A judgement has to be made as to whether there are established and tractable mechanisms for intervention that are effective. Other criteria that need to be considered are those of cost-effectiveness and impact on the existing shape of health inequalities. There are a range of methods for prioritizing health problems, risk factors, and interventions in order to select preferred strategies (Spasoff 1999).
To summarize, in order to reach the desired threshold of importance as a basis for professional consensus, a public health issue has to meet a number of criteria. It has to be seen to have a major health impact, the key risk factors should be highly modifiable, its cost should be within the bounds of other comparable interventions, and the proposed action should either reduce, or at the very least not exacerbate, health inequalities. Any such initiative should also meet ethical criteria and norms of social acceptability.
In attempting to transcend the knowledge–action gap an essential feature of successful problem definition is the identification of clear links to the policy process. The conceptualization of a public health issue should be such that the indications for policy intervention are reasonably clear. On its own, for example, the ground-breaking work of Doll and Hill (1950; 1964) in identifying the link between smoking and lung cancer was not sufficient to direct attention to a series of feasible policy options. Nor was it intended to do so. This was core scientific work of immense humanitarian significance, and yet the translation of that essential technical breakthrough into action on any substantial scale had to await further analysis that addressed the contextual and institutional determinants of smoking behaviour. Neither moral exhortation nor simple prohibition were going to be either feasible or effective policy instruments. Thus a conceptualization of the smoking–lung cancer link that brought in wider policy-relevant variables was a crucial precursor to successful policy-making and eventual action (Davis 1994).
As this example suggests, the translation of knowledge into action may require additional analytical and empirical work in order to bridge the conceptual gap between core technical disciplines, such as epidemiology and risk assessment, on the one hand and the policy arena on the other (Levenstein 1996). This is an argument not for altering or diminishing the contribution of basic technical work, but for augmenting it with closely allied behavioural and social science perspectives. In this sense this is both a traditional call for multidisciplinary research, as well as a plea for academic teamwork within the public health sciences. What is needed is a careful sequencing of academic work such that, once the key findings of the more technical disciplines and the basic sciences are established, these results can be rapidly relayed to those in health promotion and policy development (Hersey et al. 1996). This happens too infrequently.
In some areas of technical work the links into effective delivery are already well established. This is the well-worn path of regulatory science, as in environmental health and occupational health and safety. In these areas technical findings are incorporated into established regulatory systems in a relatively straightforward and uncontentious manner. It is the expansion of these regulatory frameworks into new fields that sanction individual behaviours, such as smoking, alcohol consumption, and contraceptive use, where serious controversy is likely to arise. An important role can be played in these areas by the ‘expert committee’ helping to negotiate the science–policy interface (Berridge and Stanton 1999).
There are two approaches to capturing greater policy sensitivity in the conventional public health framework. The model that prevails in disease prevention initiatives and conventional regulatory science is one of augmentation; that is, adding to an existing scientific endeavour analytical perspectives that are closer to the policy process (Hinman 1997). These are arenas in which a more traditional model of disease is dominant and the function here is a bridging one between science and policy. Health promotion, by contrast, provides an alternative paradigm—a more multidimensional model of health—and in this perspective the behavioural and social science factors are quite central to problem definition (Snider and Satcher 1997). In these fields an alternative to the multidisciplinary teamwork model is one in which the approach is interdisciplinary, and co-operation across disciplines occurs at the conceptual and policy level. To an important extent science is policy under this approach.
Regardless of the particular model, the message is that the translation of public health knowledge into effective action requires a conceptualization of the behavioural and social sciences that can establish the key links to the policy process (Rutten 1995). In particular, as Link and Phelan (1995) argue, it is important to identify what puts people ‘at risk of risks’, and this requires the analysis of social conditions as fundamental causes of disease (the epidemiological expression of this insight is presented by Rose 1985).
Managing power relations
Perhaps the least understood dimension to the interface between scientific knowledge on the one hand and policy intervention on the other is that of power. For the public health practitioner there is almost something distasteful, even underhand, about acknowledging the role that power and political position might play in achieving what appear to be self-evidently laudable and highly noble public objectives (such as saving lives and preventing illness and disease). Yet, this is the reality of the field of public policy; support must be mobilized, resources transferred, regulations instituted, programmes established, and administrative interventions sustained and carried through to successful completion (Williams-Crowe and Aultman 1994). At every point, issues of power, position, and politics are crucial.
One fundamental weakness for public health is that it lacks a powerful, united, and consistent political constituency. This is partly because some of the greatest achievements of public health are nearly invisible; death, disease, and injury averted are hard to relate to and quantify (Remington 1990). Furthermore, there are no major, cohesive and articulate constituencies that recognize and openly acknowledge the benefit they derive from public health initiatives (apart from those they employ), and there are many groups whose interests may be affronted by energetic public health action. Therefore, on political grounds alone, policy initiatives in public health are difficult to mount and sustain.
This being the case, an understanding of the political context for public health action is an essential precondition to policy formation. Tobacco control is an instance where a professional consensus on the case for intervention is not in question and where a relatively sophisticated and policy-relevant conceptualization of the issue has been developed. Yet, formidable obstacles to progress in this area remain because of the powerful constituency against regulatory and policy change that is nurtured by the tobacco companies. The issue of political feasibility has to be addressed right across the spectrum of public health action.
The distribution of power
One of the central concerns of political theory in the area of policy formulation and implementation is the configuration of power. If power is defined as the ability to achieve one’s goals even against the opposition of others, then the distribution of power is crucial to any understanding of the policy process. While such an understanding may have a rationale in political theory, it will—just as importantly for the present discussion—have a basis in empirical case studies and an operational significance for policy-making in public health (Lewis and Considine 1999).
At the risk of oversimplification, there are broadly two theories on the configuration of power: (a) pluralism, which stresses the dispersion of power, and (b) the structural interest model, which emphasizes the relative concentration of power. Behind these two models of the distribution of power lie entire theoretical systems and views of the way in which the world works. The pluralist theory derives from a perspective on society and the distribution of power that is benign and optimistic. According to this model, society is relatively egalitarian and the threshold for entry into the policy arena is low. Thus, almost any group can enter the policy process, and those groups that do enter the policy arena on any given issue are more or less evenly balanced in power and resources. The classic expression of this model remains Dahl’s Who Governs? Democracy and Power in an American City (Dahl 1961) and the theory perhaps best exemplifies the role of interest groups in the relatively fluid system of government in the United States, particularly at the level of community politics.
The structural interest model, by contrast, posits a relative concentration of power in the hands either of a few key groups or of an élite (Duckett 1984). According to this view of power the real decisions and interactions take place at a relatively rarified level in the power hierarchy and frequently at times and places that are largely invisible to the public. While multiple interest groups may be present and apparent, the real work of politics and strategic choice comes down to a few key players with the power to influence the final outcome. In its most public form this model is most consistent with the corporatist decision-making processes of relatively structured societies with highly institutionalized systems of power-broking, frequently organized around the state and its bureaucracy (Lembruch and Schmitter 1982). Examples here would be some European countries and Asian countries such as Japan. Another version of this theory sees such processes as being less formally recognized and institutionalized, and more the outcome of the positional power of key agencies that routinely exercise influence in decision-making. An early outline of this view in the health field is Alford’s Health Care Politics: Ideological and Interest Group Barriers to Reform (Alford 1975).
The pluralist model has a good deal of empirical plausibility in the health sector. Crucially, the sector is heavily populated with interest groups. Thus, almost every health condition has its own voluntary association and mutual support group, many communities are organized to lobby for local services, there are the major charities, and, finally, provider and supplier groups are all well organized and active. More than almost any other sector, health is characterized by a myriad of groups who lobby and act in a very public way, to an important extent because health is publicly funded in most countries. Debates about the allocation of public resources are highly visible and politicized.
Another feature that has until recently underscored the pluralistic nature of the health sector has been the relatively static quality of health system arrangements, at least until the 1980s (Klein 1990). Essentially the central architectural features of most health systems had remained largely unchanged for decades, and change, to the extent that it occurred, took place in an incremental fashion. Thus, in a period of reasonably steady economic growth, increments of service took place at the margins, and frequently these incremental changes resulted from the activities of one group or another advocating for a particular service or policy. This contrasted with the large-scale, systemic, and structural changes more characteristic of health-care policy in the 1980s and 1990s (Ham 1997).
While a pluralist model may best reflect the public face of debate and incremental change in the health sector, the realities of power for major decisions and for the backdrop to the ‘small’ decisions accord far more to the structural interest model. Thus, in major issues of health system change, the medical profession is a key veto group. Similarly, health insurance, medical suppliers, and pharmaceutical interests constitute a formidable force in the politics of health. Thus for major health service issues it would seem that the state, the medical profession, and, in some countries, health insurers and business, are the key players in shaping the course of policy-making (Bergthold 1988).
What of public health issues? Again, there is evidence for both of these perspectives. At the community level, and in the case of conditions for which there is no obvious commercial interest, a pluralistic model would seem to be the relevant one. In many instances groups are trying to achieve incremental change over a reasonably long period of time with uncertain policy pathways and endpoints. Thus, actions to address diabetes or asthma are constrained more by the intractability of the problem—including broader structural determinants—than any direct opposition by affected groups. Similarly, there are some issues where the policy problem is more one of gaining public support than addressing the interests of any well-organized groups (road traffic accidents, for example) (de Leeuw and Polman 1995).
One example of a public health issue where a pluralist model seems particularly appropriate is that of community water fluoridation. The adjustment of fluoride levels in central water supplies has been shown to be a safe and effective public health measure for preventing tooth decay for over 50 years (Ripa 1993). And yet, once the technical health issues have become secondary to wider political considerations of safety and individual rights, adoption of this measure has been slow. Crucial in the decision to fluoridate community water supplies have been the media, the nature of formal decision-making processes, consumer and wider social movements, the role of public agencies and key health professional organizations, and other community variables (Frazier 1984).
Nevertheless, for some of the ‘big’ issues of public health, structural interests are at the heart of the political arena. The major example in this regard is tobacco control, where controls on advertising, price increases, and the creation of smoke-free environments, have all been opposed by the industry. Similarly, the alcohol and food industries have been very active in other areas of public health concerned with diet and alcohol consumption. The public health task is made very much more difficult in these circumstances since a single, powerful opponent can work at a number of levels of the policy process to block public health initiatives.
Issue definition and agenda setting
A central element of the political context to policy is the ability of different groups to shape the definition of key issues and to set the agenda (Kindgon 1995). This occurs within a framework of values and wider economic and social forces that set the environment and provide the openings for policy initiatives (Burris 1997). For example, the United States is unique among the advanced Western democracies in the limited role it accords the state in the funding, organization, and delivery of health care. This represents a particular historical tradition, the values of individualism, and a special combination of economic interests—such as health insurers—and social groups (a large, affluent middle class able to afford private insurance), together with a federal political system characterized by weak party alignment (Lee et al. 1997). This particular set of environmental forces places major constraints on the range and characteristics of possible health reform initiatives that have any chance of being comsidered for the policy agenda in the United States.
Therefore, there are certain larger forces that shape the structure of opportunities and possibilities for policy. Within these parameters, however, issue definition and agenda setting are crucial to determining the initiation of the policy process, and these are to an important extent reflective of, and contributory to, the structure of power. Essentially, groups that are able to get issues defined in a way favourable to their interests, and that can then go on to lodge those issues in a prominent place on the formal policy agenda, are in a strong position to get their own way in the policy-making process.
Again, as in the case of the distribution of power, there are two broad tendencies in the study of issue definition and agenda setting. According to the pluralistic model, advocacy groups initiate issues from outside the formal institutional framework. Such advocates articulate an issue or grievance, identify solutions, and attempt to expand support for their case, culminating, if successful, in their cause entering the formal, institutional agenda (Cobb and Elder 1983). There are many examples of this kind in the health sector, with groups championing local causes or particular services, attempting to establish coalitions of support, and working through the media towards building the momentum for formal recognition of the issue in the policy process. Examples of this are the siting of local facilities, the provision of a new technology or service, or the recognition of some new public health or environmental hazard. In the public health field the building of public momentum behind a tobacco control agenda is a classic example of social mobilization and coalition building (Sato 1999).
An alternative model derives from the structural interest model and emphasizes the élite and relatively closed nature of decision-making in many instances. For certain key power groups the quickest and surest way to achieving a desired position on the institutional agenda with a view to some policy initiative is through relatively private avenues of influence rather than by public campaign. This may be partly because the issues at stake are rather technical and do not have a broad constituency—for example, screening practices in primary care (Florin 1998). In other cases, this route is preferred because it is a way of achieving policy success without attracting public opposition. For example, nearly every provider group in the health field has its own statute governing matters such as qualifications, recruitment, regulation, discipline, and so on.
There is unlikely to be a broad public interest in these matters and many such issues are relatively technical and uncontentious. Yet, the question of the quality of health practice and the methods by which, for instance, the medical profession goes about dealing with patient complaints and professional competence are potentially of wide interest and generate recurrent public unease and dissatisfaction (Stacey 1992). Similarly, powerful commercial lobbies would far rather deal with key issues in private where their influence can be exerted to its maximum and decisions can be shielded from potentially disruptive public scrutiny. In other circumstances, where such issues are already in the public domain—or where governments need to feel the pressure of public opinion—such interest groups are also adept at building and nurturing public support for their position.
A gap in both these models is a failure to allow for initiatives that may emanate not from interest groups, large or small, but from within the apparatus of government itself. Governments and bureaucracies, albeit frequently on the prompting of interest groups, initiate issue definition and agenda-setting exercises to try to set the climate for administrative intervention (Robins and Backstrom 1994). To some extent this can be seen as an acknowledgement that an issue is on the formal, institutional agenda. For example, tobacco control legislation may be under consideration after a long public campaign started externally to formal institutional structures, and the public heath bureaucracy may see it as timely to maintain a climate of public support (Sato 1999). In other instances, initiatives may derive from other parts of the bureaucracy, such as the finance ministry, or some official watchdog agency, or even from the executive or legislative arms of government.
Aside from positional power and influence in issue definition and agenda setting, a key dimension that advocacy groups, lobbyists, and governmental and bureaucratic agencies have to consider is the cultural salience and relevance to core values of their problem focus. Issue definition is not merely a function of positional influence, size, resources, and institutional and environmental opportunity—although these all help and are essential preconditions for successful agenda setting. Centrality to core values and culturally salient concerns and symbols are crucial as well. For example, certain social groups and conditions are stigmatized, while others are highly valued. Thus, causes associated with mental health struggle, while those associated with the ailments of children thrive (Nelson 1984).
The resolution of many key public health issues turns on the particular value framework within which they are viewed by the public. Whether it be the fencing of private swimming pools, gun control, or restrictions on smoking and drinking, a primary aspect of the public definition of these issues is whether they are seen as matters of health and safety or controversies impinging on the expression of individual rights (Burris 1997). For advocacy groups, lobbyists, and governmental agencies involved in these controversies, positioning of the issue in relation to these two value complexes becomes crucial and much energy is devoted to presenting the issue in the appropriate light (Chapman and Lupton 1994). Again, certain groups are greatly advantaged because of their material resources (which enables them to run public relations campaigns), their access to the media, or relevant cultural assets (such as positive public image, high social status), and these can become crucial in issue definition and agenda setting. Thus, an important stage in the campaign for greater tobacco control in many countries has been the advent of both the medical profession and key charity groups, such as the cancer societies. This has both lifted the public perception of the cause and its legitimacy, and at the same time detracted from the relative standing of its opponents, at a critical juncture in the development of the issue and its passage onto the formal, institutional agenda.
Apart from the philosophical battle between health and safety definitions on the one hand and individual rights on the other, there have been other polarities in the coloration of public policy issues concerned with individual well being. One important tussle has been that between legal and public health definitions. The issue of HIV and AIDS remains an area in which public health concepts, and a framework of human rights, are in contest with more punitive policies deriving from the law and the criminal justice system (Kirp and Bayer 1992). There are many other areas of fluidity where groups are attempting to negotiate new definitions. Thus, groups with varying disabilities—physical, mental, intellectual, vision—have sought to redefine their condition from that of dependent patients in treatment to a human rights perspective as clients in partnership with service agencies (Scotch 1989). Pharmaceutical companies, by contrast, are moving in the opposite direction (Payer 1992). They are attempting to convert an ever-expanding range of life conditions into treatable ailments, a process called ‘medicalization’ by sociologists (Conrad 1992). More broadly, this touches upon competing clinical and public health definitions of issues, where treatment philosophies compete with prevention in issue definition (for example, treatment for alcohol dependency versus public policy initiatives to reduce alcohol-related harm) (Remington 1990).
The question of issue definition and agenda setting, therefore, relates to power, both in its positional and in its cultural and ideological sense. Issues that are defined by influential groups and that are able to claim salience to core cultural values have a much greater chance of being seen favourably by the public and by those in decision-making positions. This, in turn, gives them a much better chance of entering the formal institutional agenda.
Negotiating the policy process
The question of power relations has been discussed to this point as a characteristic of the environment external to the policy process. As will be evident in the discussion that follows, the negotiation of positional power and influence remains central to the passage of policy. Nevertheless, the emphasis in this section will be on the policy process as an institutional framework in which policies are formulated, decisions made, policy instruments selected, and programmes delivered. Clearly, this is a somewhat formalistic approach, and one that is adopted only for the purposes of exposition. In practice, problem definition, political context, and the policy process are inextricably intertwined, although exerting varying influences at different stages of the policy cycle.
The institutional framework
An analysis of power can tend to emphasize the raw ingredients of group relations and positional influence. Power can be seen as a force that emanates from the brute facts of material resources and privileged social position. Thus the emphasis in the previous section is on advocacy groups, lobbyists, and governmental agencies as actors in a dynamic field of power relations. Such an approach emphasizes the role of actors with distinct interests and capabilities. Yet, the interaction of these agencies, and to some extent their source of power, is shaped by institutions. In other words, the patterns of interaction between agencies are shaped, though not ultimately determined, by relatively established rules and processes that are socially sanctioned and often legally enshrined (Goodin 1996) (for a health application, see Immergut (1992)).
In the case of the policy process there is a clear organizational context through which policy issues are channelled. Principally, this represents the formal structures of government for a health application (Howlett and Ramesh 1995). In the case of the advanced Western democracies this means understanding the respective roles of the different arms of government—executive and legislature—and of the bureaucracy. For example, a very important distinction has to be drawn between the differing policy capacities of the executive and the legislature, respectively. The executive arm of government is in a position to formulate and carry out policy, but it cannot on its own necessarily secure political legitimacy, full financial support (where this is taxpayer funded), or legal sanction (for instance, a change in the law). These are functions of the legislature. Crucially, it is important for the purposes of political symbolism that key policy changes are publicly debated and passed through the legislature.
However, the opportunities for legislatures to play an active role in policy formulation and development are limited. This is partly because in most parliamentary systems there is a close relationship between the political alignment of the executive and the majority of the legislature. It is also partly because of the relatively technical nature of most policy issues in the modern state. Such is the technical nature of many issues that individual legislators cannot be expected to have the resources or the background to command the policy agenda on the myriad of matters that come before them. A combination of the resources of the bureaucracy at the command of the executive, together with strong party allegiances in most parliamentary systems, means that the legislature plays little active role in policy formulation. Nevertheless, individual members can raise issues publicly in the debating chamber and they do scrutinize policy proposals and monitor programme implementation in the committee room.
In presidential systems the separation of powers between the executive and the legislature means that the support of the majority in the house cannot be taken for granted, and hence there is more room for negotiation and bargaining. A further distinction needs also to be drawn between federal and unitary systems of government. The capacity for making consistent and coherent policy is much more difficult in federal systems because of the distinct sets of responsibilities and jurisdictions at national and state levels. This means that there is a frequent requirement for negotiation and considerable opportunity for jurisdictional disputes. If we add to its presidential and federal nature the lack of party discipline in the American system, then consistent, coherent, and long-term policy development would seem to be difficult indeed (Steinmo and Watts 1995).
Leichter (1991) compared these fundamental differences in political structures for the United States and Great Britain across a number of major health promotion issues (smoking, alcohol, road safety, and HIV/AIDS). While the unitary nature of the British political system encouraged clarity and decisiveness in policy-making, it did not necessarily lead either to greater creativity or to the ‘right’ result (as judged from a conventional public health perspective). Indeed, the multiple nature of the American political system meant that, while federal initiatives on health promotion might be blocked at the centre, experimentation with a diversity of interventions could take place at the level of the individual state. While clarity, coherence, and decisiveness might not have been apparent in the American context, creativity, inventiveness, diversity, and local ownership were.
In many respects the key element of the institutional framework for the purposes of policy-making is the bureaucracy. While, symbolically speaking, the legislature is important for imparting public support to policy proposals, and while the executive plays a central role as the embodiment of political leadership, appointed officials largely drive the detail of policy formulation. In principle they are answerable to elected politicians in the executive arm of government, but there is a limit to the extent to which generalist politicians can effectively maintain scrutiny over specialist advisers and bureaucrats supposedly at their command.
This being the case, the structure of the bureaucracy itself becomes a key factor in the policy process. In the case of the health bureaucracy, officials usually retain a host of formal committees and informal advisory groups in order both to nurture links with key constituencies in the sector and to maintain competency and intelligence in a range of policy and technical areas. Beyond this there are important distinctions to be made within the health bureaucracy—functionally between public health and health care, for example, and operationally between policy advice and service delivery. Apart from the health ministry there are other sections of the bureaucracy that are important to health policy-making. For example, the finance ministry will typically set the fiscal parameters for health funding, the transport ministry will be important for certain public health campaigns, education and environmental health for others, and the police and judiciary for still others. Any attempt to negotiate the policy process requires a full understanding of these organizational intricacies. Indeed, most public health initiatives require intersectoral allegiances and partnerships of this kind for their success.
If, as argued above, policy formulation and development does not take place in the legislature—the most public and symbolically important debating forum in the advanced Western democracies—then where does it occur? It is at this point that the institutional and agency approaches to policy meet. There is an institutional framework—part public, part invisible—that sets the context for policy formulation, but these institutional settings have to be populated by active agencies and individuals motivated by the desire to achieve certain policy goals (Howlett and Ramesh 1995).
P>In this connection a new terminology has been developed that combines the organizational apparatus of institutions with the active and instrumental characteristics of advocacy groups, lobbyists, and official agencies. Thus analysts have identified policy networks as being sets of individuals and groups with an interest in a particular policy issue which they wish to advance. Members of networks are in regular interaction on their common issues of interest, although the regularity and coherence may vary greatly. In some areas, where interests are well defined and of long standing, the issue networks that have formed are highly integrated and stable. In some instances such networks are so small, stable, and predictable that they have been termed ‘iron triangles’, referring to their longevity and intractability (Jordan 1986). To take a public health example, such an iron triangle would typically include the public health bureaucracy representing the interests of the state, the industry or interest being regulated (such as the tobacco or alcohol industry), and an advocacy coalition (of health and related consumer groups).
The concept of an ‘iron triangle’ is relevant to the discussion of power and its distribution; this particular form of policy network derives from a structural interest configuration of power. Clearly, power and influence are concentrated in this set of circumstances, and policy debates are promoted in a highly stylized and predictable fashion. Nothing much will change until, and unless, there is some significant shift in the balance of power (for instance, a change of government, or an international event, or a new ally to the coalition). An alternative set of interacting relationships to that based on the structural interest model is the issue network, an association based on a wide range of loosely connected groups. This follows a more pluralistic concept of power distribution. Issue networks are perhaps more common in the health arena, since the health sector exhibits such a wide range of issues around which local and small special interest groups can coalesce.
Policy networks as defined here—ranging from the cohesive and predictable sets of actors to broader and looser associations—provide the framework of allegiances, interest, and ideology that help to define the parameters within which policy is formulated. Debates, public and private, take place among the principal actors, and these debates in turn help to shape the progress of policy (Read 1992). Another important contributor to these debates is the policy community. While policy networks refer to associations of actors in relatively frequent and predictable interaction over key policy areas in which they have strong and potentially conflicting interests, policy communities have their basis in a common area of expertise. The participants in policy communities may have markedly different perceptions of their policy area, but they enjoy a sufficiently common intellectual base as to permit a community of discourse on that policy. Thus there is a constant exchange of information, setting the context for policy formation.
At its most public the policy community is displayed in academic conferences, the writings of journalists, and books, speeches, and articles. This is generally the arena in which public health scientists can most clearly see their role, that is, contributing to debates among those with a level of expertise in a public health issue. These communities have played a crucial role in preparing the ground for policy development in a number of fields. Most recently the advent of HIV/AIDS as a new and unheralded public health challenge precipitated a widening debate among those with a basic understanding of the area. Thus, gay activists, public health bureaucrats, and academics helped to define the nature of the problem area and the range of potential policy options. This was preparatory to the entry of the issue on the formal institutional agenda and its progress in the policy development process through to implementation (for the New Zealand case, see Lindberg and McMorland (1996)).
The decision-making process is at the heart of public policy. The intensity of interest in a policy area is in direct proportion to the desire for or against change in the existing policy settings. In the case of policy networks of the iron-triangle type, the routinized and predictable nature of these relationships, and the carefully balanced nature of power that they reflect, suggest there is little possibility of change in these cases. To the extent that change does occur in these circumstances, it will be of a relatively major and systemic kind. In the case of issue networks, however, a change in policy settings is more likely, but will be of a more minor variety.
These alternative scenarios indicate two broad types of decision-making process (Walt 1994). In the first instance decisions revolve around issues for which the stakes are high and that require a relatively major mobilization of power and resources. Decisions of this kind are strategic and may have long-term consequences since they change long-standing and stable sets of arrangements. A recent instance has been the rash of health system restructuring that has afflicted most of the advanced economies, prompted by financial stress and ideological challenge. Long-established sets of relationships and assumptions were disrupted by the introduction of a range of new funding and structural arrangements (Ham 1997). In the public health area the introduction of tobacco control initiatives across a number of countries broke an apparent long-standing policy deadlock.
Major and strategic decisions of this kind have also been linked to an idealized rational model of the policy-making process. In principle, policy change, if it is to be mounted, needs to be synoptic and all embracing, if it is properly to co-ordinate all aspects of the issue under consideration. In this sense rational decision-making takes into account longer-term strategic objectives (Alexander 1986). Nevertheless, the concept of rational decision-making is probably better considered a normative one, that is, a desirable objective, rather than providing an accurate description of actual decision-making processes.
At the other end of the spectrum are decisions taken at the margins of the system and involving less central policy goals. These are usually promoted by smaller interests operating within looser, issue-type networks. They do not require the major mobilization of power and resources, and their focus is shorter term and less strategic. In the health system, and indeed any policy sector, most decisions are of this type. They involve bargaining and compromise, and the achievement of relatively small-scale amendments to existing policy settings (Lindblom 1965).
It has been argued that this style of decision-making—the incremental model—is both a better description of the policy process as it is, and a reasonable guide as to how policy should be conducted. There are two reasons why incremental decisions are likely to be the norm. Firstly, there is the inertia of existing arrangements and power relations, such that it is more likely both to achieve agreement and to secure a shift in resources at the margin, and in smaller scale, than in a major and thorough-going fashion. Secondly, the incremental model follows much more closely the traditional way in which bureaucracies work, and it is officials who usually have a determining role in setting the policy agenda.
Most decisions in the health sector are of the incremental type. There are few decisions that are of the major strategic kind, for example population funding equalization or comprehensive tobacco control. The typical decision is one that lies at the margins, for example the extension of an existing service, a new service opened in a small way, a shift in enforcement of current tobacco control legislation, or extending infectious disease control initiatives to a new arena. The problem with such a style of decision-making is that, while it may accurately describe the way in which the great majority of decisions are made, it does little to maintain or ensure the overall coherence of policy. Thus a series of incremental decisions, each clearly justified in its own light as achieving a limited objective, may in the aggregate actually undermine larger policy goals. For example, incremental public health initiatives, while improving health overall, may at the same time deepen social inequalities in health.
It is this concern—with the larger impact of disjointed and uncoordinated incremental decisions—that has fuelled the rational model of decision-making. While decisions may not usually or normally be made in full consideration of their strategic significance, this concept of decision-making is one that is held up by many policy analysts as a desirable modus operandi (Hogwood and Gunn 1984). Realistically this rational strategic approach is likely to occur only in circumstances where decision-makers are provided with a relatively unconstrained and simple set of policy circumstances, for example a major opportunity for political change. As an ideal, this model remains attractive to those involved in public health policy-making, at least as an organizational tool for policy appraisal. Nevertheless, the most effective policy approach is likely to be ‘mixed scanning’, involving a mix of rational planning and incremental adjustment (Etzioni 1967).
Implementation and delivery
Decision-making lies at the heart of the policy process since it alone can precipitate a formal change in policy settings. However, it should be noted that a decision not to alter an existing policy may be just as important as a decision to institute changes formally (Bachrach and Baratz 1963). Thus, the decision not to prosecute a tobacco company that is transgressing existing legislation, for example, has nearly as much effective force as a decision to change the legislation in a more liberal direction. In other circumstances, the ability to prevent an issue coming to decision can be a victory of great strategic importance. For example, the ability of the tobacco industry for years to prevent full consideration of voluntary codes of conduct in a formal decision-making process, which in many circumstances were likely to go against them, was vital in maintaining sponsorship and advertising under existing codes.
In most cases, however, effective decision-making does entail change to existing policy settings. This is usually the desired outcome in much public health policy since typically the promotion of public health objectives involves some modification to a status quo that is dominated by economic and individualistic values (Burris 1997). This being so, the key to policy success is the implementation process. Traditionally, this part of the policy cycle was viewed as relatively unproblematic, being seen as principally an administrative function. However, a series of studies revealed that the objectives of many programmes—agreed and sanctioned in formal policy—were not being achieved in practice. The assumption of ready and straightforward implementation was not fulfilled. From this insight into the incompleteness of traditional models of the policy process sprung the study of implementation and a closer assessment of policy design and policy instruments (Pressman and Wildavsky 1984).
It is perhaps remarkable that the implementation process was ever considered to be an administrative formality. Indeed, the surprise is rather that policies are implemented as intended or even at all, such are the formidable obstacles to moving from formal policy prescription to actual social and behavioural change. The nature of the problems that policy is addressing are usually complex, multifaceted, and sometimes quite intractable. The greatest likelihood of implementation success is in circumstances where the policy has relatively simple technical features, represents a relatively marginal change to the status quo, is implemented by a single agency, has a clear single objective, and is of short duration. It also helps if the policy does not upset powerful interest groups and has relatively low visibility with the public (Walt 1994).
In the political science literature two broad schools of thought about the implementation process have been identified (Mazmanian and Sabatier 1989). For those coming from a more conventional model of the administrative apparatus, successful implementation was seen as issuing in a relatively linear and hierarchical fashion from the source of policy and power, namely from ‘above’. An alternative approach to this conventional administrative model was one that emphasized the active participation of other non-bureaucratic actors and saw implementation as less linear and hierarchical; this reflected a more interactive process from ‘below’ characterized by bargaining among the affected parties.
There are a number of fairly clear preconditions for effective implementation, the principal being the political climate. Before anything can happen, the required political resources must be available; that is, legitimacy, effective administrative apparatus, and relative policy consensus. However, financial, managerial, and technical resources must be available over and above this. Without the required financial support, a policy cannot progress, or can only be implemented incompletely. Finally, management competency and technical capacity are required to secure an effective outcome (Grindle and Thomas 1991).
An important consideration in determining the particular style of implementation is policy content; that is, the type of policy being addressed and kinds of instruments being brought to bear. Thus, one question that has to be asked is whether the policy is addressing an issue of distribution (providing a new service or extending an existing one), redistribution (moving resources from one group to another), or regulation (Palmer and Short 1989). A second set of issues concerns the kinds of policy instrument likely to be considered; that is, market mechanisms, or voluntary, family, and community endeavours, or administrative initiatives such as regulation or directive provision of services (Linder and Peters 1989).
Most public health initiatives are fundamentally regulatory in thrust since they frequently require changes in behaviour and activity relevant to health outcomes. Conversely, most health-care questions are distributive in their implications since they usually involve decisions to introduce or extend the delivery of services to key population groups. The public funding of health care frequently raises redistributive questions, since such resources are generally allocated disproportionately to more needy groups with poorer material circumstances, such as the young, the elderly, and the poor.
Again, many public health matters naturally lend themselves to the deployment of regulatory policy instruments and/or the direct provision of services. Typically, a public health initiative requires change to existing patterns of behaviour or activities. Frequently this involves some regulatory or administrative intervention and usually this can only be delivered and enforced by an officially sanctioned agency of the state. A striking departure from this tradition was the almost universal reliance on non-statutory mechanisms in the response to HIV/AIDS across the developed world (Kirp and Bayer 1992). Indeed, non-governmental organizations are playing an increasingly important role as partners in public health initiatives (Walt 1994).
Three scenarios for knowledge in action
The framework developed in this chapter has been organized around three strands of deliberation. In the first instance it has been argued that for any public health initiative to occur it requires a clear definition of the problem to be addressed in a manner that makes it susceptible to policy intervention. This requires an embedding of the issue within an action-oriented social science framework. Secondly, even where there is a clear definition and a solid professional consensus around that definition, the political climate has to be favourable to its progression onto the policy agenda. Finally, an issue, once well defined and favoured by the political environment, has to pass through a relatively well-established set of processes if it is to be implemented effectively.
Stated in this fashion, the picture is of a relatively orderly and linear progression from problem formulation, through the garnering of political support, to policy development and implementation. Clearly, this is an overly rationalistic representation of what is usually a far more complex and disordered process. Indeed, one authority sees these three strands of interest operating as three relatively independent streams (Kindgon 1995). Thus, there are at any time a number of ‘problems’ awaiting policy solution, some of them of long standing, others being relatively new and potentially urgent. Then, there is an established distribution of power and a range of interest groups focused on policy questions which, when taken together, provide a series of political opportunities for administrative action. Finally, there is a range of policy solutions and settings either in position or in process that are available in the immediate political culture and institutional environment.
The conventional view of translating knowledge into action is to see it as a progression from problem definition through advocacy and political support to policy outcome. There are a number of areas where this model applies. Thus, the advent of HIV/AIDS presented a novel problem that could not be ignored and for which there were no clear policy precedents. Problem definition proceeded to policy development at a reasonably low political threshold, only requiring high-level political support once a series of relatively unprecedented policy options were proposed.
Many public health problems, however, are of long standing and do not offer any clear-cut technical solutions. Furthermore, there is frequently a political readiness to deal with these issues; examples include teenage pregnancy, youth suicide, mental health, and poor diet. Still more intractable are issues of health inequalities and social determinants of health (Marmot and Wilkinson 1999). These are problems in search of the right policy prescription. The mobilization of political resources will still be required in these circumstances, but the principal requirement is a feasible policy option.
Finally, there are those public health problems about which a clear problem definition and policy options have emerged, but that await the correct political climate. For example, according to Berridge (1999) passive smoking was a ‘scientific fact waiting to emerge’. The public acceptability of passive smoking was shaped by a complex number of factors, only one of which was the scientific evidence. It was consistent with a growing individualistic and environmental ethos, giving medical and scientific legitimacy to a position that had originated from a moral, and then a personal rights, issue.
Each of these three scenarios requires a slightly different approach to the knowledge–action link. In the case of new problems, or existing problems about which there is new knowledge, the definitional process and its links into feasible policy options is a key step, as exemplified in the case of HIV/AIDS. In the case of long-standing problems for which the balance of power is not a crucial consideration, the development of feasible policy options is the essential prerequisite to moving the issue along to some resolution. Finally, where a problem is clearly defined and feasible, and well-established policy options are available, the issue of political climate is salient. This can either be nurtured through advocacy and lobbying or has to await a significant shift in the balance of power at the level of the formal political process.
Traditionally, disciplines of a more technical kind like epidemiology, biostatistics, and a range of laboratory sciences have been at the core of public health. It is through the application of this expertise that some of the most startling breakthroughs in prevention have been achieved. Increasingly, however, as public health has moved from the more conventional issues of disease prevention to include those of health promotion, and as the awareness of larger policy questions has impinged on public health practice, so the salience of the policy process has become more marked. This chapter has outlined a framework for the analysis of the interface between the public health sciences and the policy process. It has argued that a consideration of policy should permeate the ‘basic sciences’ of public health—in the sense that the conceptualization of fundamental public health issues should be permeated by the social sciences in order to facilitate links to policy. The approach should be one that is interdisciplinary with the social sciences at a conceptual level. Furthermore, once the move is made into the application of this core public health expertise, the political climate and the policy process are both crucial determinants of effective translation of knowledge into effective public health practice.
Alexander, E. (1986). Approaches to planning: introducing current planning theories, concepts and issues. Gordon and Breach, New York.
Alford, R. (1975). Health care politics. ideological and interest group barriers to reform. University of Chicago Press.
Atwood, K., Colditz, G.A., and Kawachi, I. (1997). From public health science to prevention policy: placing science in its social and political contexts. American Journal of Public Health, 87, 1603–6.
Bachrach, P. and Baratz, M.S. (1963). Decisions and nondecisions: an analytical framework. American Political Science Review, 57, 1947–52.
Beaglehole, R. (1990). Epidemiology and health policy: how do we stop the band playing? New Zealand Medical Journal, 103, 323–5.
Bergthold, L.A. (1988). Purchasing power: business and health policy change in Massachusetts. Journal of Health Politics, 13, 425–51.
Berridge, V. (1999). Passive smoking and its pre-history in Britain: policy speaks to science? Social Science and Medicine, 49, 1183–96.
Berridge, V. and Stanton, J. (1999). Science and policy: historical insights. Social Science and Medicine, 49, 1133–8.
Brownson, R.C., Newschaffer, C.J., and Ali-Abarghoui, F. (1997). Policy research for disease prevention: challenges and practical recommendations. American Journal of Public Health, 87, 735–9.
Burris, S. (1997). The invisibility of public health: population-level measures in a politics of market individualism. American Journal of Public Health, 87, 1607–10.
Chapman, C. and Lupton, D. (1994). The fight for public health. Principles and practice of media advocacy. BMJ Publishing, London.
Cobb, R.W. and Elder, C. (1983). Participation in American politics: the dynamics of agenda-building. Johns Hopkins University Press, Baltimore, MD.
Conrad, P. (1992). Medicalization and social control. Annual Review of Sociology, 18, 209–32.
Dahl, R. (1961). Who governs? Democracy and power in an American city. Yale University Press, New Haven, CT.
Davis, P. (1994). A sociocultural critique of transition theory. In Social dimensions of health and disease. New Zealand persectives (ed. J. Spicer, A. Trlin, and J.A. Walton), pp. 162–75. Dunmore Press, Palmerston, New Zealand.
de Leeuw, E. (1993). Health policy, epidemiology and power: the interest web. Health Promotion International, 8, 9–52.
de Leeuw, E. and Polman, L. (1995). Health policy making: the Dutch experience. Social Science and Medicine, 40, 331–8.
Doll, R. and Hill, A.B. (1950). Smoking and carcinoma of the lung. British Medical Journal, September 30, 739–48.
Doll, R. and Hill, A.B. (1964). Mortality in relation to smoking: ten years’ observations of British doctors. British Medical Journal, 1, 1399–410.
Duckett, S. (1984). Structural interests and Australian health policy. Social Science and Medicine, 16, 959–66.
Etzioni, A. (1967). Mixed-scanning: a ‘third’ approach to decision-making. Public Administration Review, 27, 385–92.
Fletcher, S.W. (1997). Whither scientific deliberation in health policy recommendations? Alice in the wonderland of breast-cancer screening. New England Journal of Medicine, 336, 1180–3.
Florin, D. (1998). Influencing national policy; public health medicine and coronary heart disease. Journal of Public Health Medicine, 20, 80–5.
Frazier, P.J. (1984). Public and professional adoption of selected methods to prevent dental decay. In Social sciences and dentistry: a critical bibliography (ed. L.K. Cohen and P.S. Bryant), pp. 84–144. Quintessence Publishing, London.
Gillroy, J.M. and Wade, M. (ed.) (1992). The moral dimensions of public policy choice. Beyond the market paradigm. University of Pittsburgh Press.
Goodin, R.E. (ed.) (1996). The theory of institutional design. Cambridge University Press, New York.
Greenberg, M. (1992). Impediments to basing government health policies on science in the United States. Social Science and Medicine, 35, 531–40.
Grindle, M. and Thomas, J. (1991). Public policies and policy change. Johns Hopkins University Press, Baltimore, MD.
Ham, C. (ed.) (1997). Health care reform. Learning from international experience. Open University Press, Buckingham.
Hersey, J.C., Collins, J.W. Gershon, R., and Owen, B. (1996). Methodologic issues in intervention research—health care. American Journal of Industrial Medicine, 29, 412–17.
Hinman, A.R. (1997). Quantitative policy analysis and public health policy: a macro and micro view. American Journal of Preventive Medicine, 13, 6–11.
Hogwood, B.W. and Gunn, L.A. (1984). Policy analysis for the real world. Oxford University Press.
Holtgrave, D.R., Doll, L.S. and Harrison, J. (1997). Influence of behavioral and social science on public health policymaking. American Psychology, 52, 167–73.
Howlett, M. and Ramesh, M. (1995). Studying public policy. Policy cycles and policy subsystems. Oxford University Press.
Immergut, E.M. (1992). Health politics: interests and institutions in Western Europe. Cambridge University Press.
Jasanoff, S. (1990). The fifth branch. Science advisers as policymakers. Harvard University Press, Cambridge, MA.
Jeffery, R.W. (1989). Risk behaviors and health: contrasting individual and population perspectives. American Psychology, 44, 1194–202.
Jordan, A.G. (1986). Iron triangles, woolly corporatism and elastic nets: Images of the policy process. Journal of Public Policy, 1, 95–123.
Kindgon, J.W. (1995). Agendas, alternatives, and public policies. Little, Brown, Boston.
Kirp, D.L. and Bayer, R. (ed.) (1992). AIDS in the industrialized democracies. Passions, politics, and policies. Rutgers University Press, New Brunswick, NJ.
Klein, R. (1990). Research, policy, and the National Health Service. Journal of Health Politics, Policy and Law, 15, 504–23.
Kuhn, T.S. (1970). The structure of scientific revolutions. University of Chicago Press.
Lane, J.-E. (1993). The public sector. Concepts, models and approaches. Sage, London.
Latour, B. (1987). Science in action. How to follow scientists and engineers through society. Open University Press, Buckingham.
Lee, P. R., Benjamin, A.D. and Weber, M.A. (1997). Policies and strategies for health in the United States. In Oxford textbook of public health (3rd edn) (ed. R. Detels, W.W. Holland, J. McEwen, and G.S. Omenn), pp. 297–321. Oxford University Press.
Leichter, H.M. (1991). Free to be foolish. Politics and health promotion in the United States and Great Britain. Princeton University Press, Princeton.
Lembruch, G. and Schmitter, P. (1982). Patterns of corporatist policy-making. Sage, Beverly Hills.
Levenstein, C. (1996). Policy implications of intervention research: research on the social context for intervention. American Journal of Industrial Medicine, 29, 358–61.
Lewis, J. M. and Considine, M. (1999). Medicine, economics and agenda-setting. Social Science and Medicine, 48, 393–405.
Lindberg, W. and McMorland, J. (1996). ‘From grassroots to business suits’: the gay community response to AIDS. In Intimate details and vital statistics. AIDS, sexuality and the social order in New Zealand (ed. P. Davis), pp. 102–19. Auckland University Press.
Lindblom, C. (1965). The intelligence of democracy: decision making through mutual adjustment. Free Press, New York.
Linder, S.H. and Peters, G.B. (1989). Instruments of government: perceptions and contexts. Journal of Public Policy, 9, 35–58.
Link, B.G. and Phelan, J. (1995). Social conditions as fundamental causes of disease. Journal of Health and Social Behavior, Special Issue, 80–94.
Lomas, J. (1990). Finding audiences, changing beliefs: the structure of research use in Canadian health policy. Journal of Health Politics, Policy and Law, 15, 525–42.
Lomas, J. (1991). Words without action? The production, dissemination, and impact of consensus recommendations. Annual Review of Public Health, 12, 41–65.
McKinlay, J.B. (1993). The promotion of health through planned sociopolitical change: challenges for research and policy. Social Science and Medicine, 36, 109–17.
Marmot, M. and Wilkinson, R.G. (ed.) (1999). Social determinants of health. Oxford University Press.
Mazmanian, D.A. and Sabatier, P.A. (1989). Implementation and public policy. University Press of America, Lanham, MD.
Mechanic, D. (1993). Social research in health and the American sociopolitical context: the changing fortunes of medical sociology. Social Science and Medicine, 36, 95–102.
Mechanic, D. (1995). Emerging trends in the application of the social sciences to health and medicine. Social Science and Medicine, 40, 1491–6.
Milio, N. (1988). Making healthy public policy; developing the science by learning the art: an ecological framework for policy studies. Health Promotion, 2, 263–74.
Muir Gray, J.A.M. (1997). Evidence-based healthcare: how to make health policy and management decisions. Churchill Livingstone, New York.
Murray, C.J.L. and Lopez, A.D. (1996). Evidence-based health policy—lessons from the Global Burden of Disease. Science, 274, 740–3.
Nelson, B.J. (1984). Making an issue of child abuse. Chicago University Press.
Northridge, M. (1995). Annotation: public health methods: attributable risk as a link between causality and public health action. American Journal of Public Health, 85, 1202–4.
O’Neill, M. and Pederson, A.P. (1992). Building a methods bridge between public policy analysis and health public policy. Canadian Journal of Public Health, 83 (Supplement 1), S25–30.
Palmer, G. and Short, S. (1989). Health care and public policy. Macmillan, Melbourne.
Payer, L. (1992). Disease mongers. How doctors, drug companies and insurers are making you sick. Wiley, New York.
Pearce, N. (1996).Traditional epidemiology, modern epidemiology, and public health. American Journal of Public Health, 86, 678–83.
Pressman, J.L. and Wildavsky, A.B. (1984). Implementation: how great expectations in Washington are dashed in Oakland. University of California Press, Berkeley, CA.
Read, M. (1992). Policy networks and issue networks: the politics of smoking. In Policy networks in British government (ed. D. and R. Rhodes Marsh). Clarendon Press, Oxford.
Remington, R.D. (1990). From preventive policy to preventive practice. Preventive Medicine, 19, 105–13.
Richmond, J.B. and Kotelchuck, M. (1991). Co-ordination and development of strategies and policy for public health promotion in the United States. In Oxford textbook of public health (ed. W.E. Holland, R. Detels, and G. Knox). Oxford University Press.
Ripa, L.W. (1993). A half-century of community water fluoridation in the United States: review and commentary. Journal of Public Health Dentistry, 53, 17–44.
Robins, L. and Backstrom, C. (1994). The role of state health departments in formulating policy: a survey on the case of AIDS. American Journal of Public Health, 84, 905–9.
Rochefort, D.A. and Cobb, R.W. (1993). Problem definition, agenda access, and policy choice. Policy Studies Journal, 21, 56–71.
Rose, G. (1985). Sick individuals and sick populations. International Journal of Epidemiology, 14, 32–8.
Rothman, K.J. and Poole, C. (1985). Science and policy making. American Journal of Public Health, 75, 340–1.
Rutten, A. (1995). The implementation of health promotion: a new structural perspective. Social Science and Medicine, 41, 1627–37.
Sato, H. (1999). The advocacy coalition framework and the policy process analysis: The case of smoking control in Japan. Policy Studies Journal, 27, 28–44.
Schmid, T.J., Pratt, M., and Howze, E. (1995). Policy as intervention: environmental and policy approaches to the prevention of cardiovascular disease. American Journal of Public Health, 85, 1207–11.
Scotch, R. (1989). Politics and policy in the history of the disability rights movement. Milbank Quarterly, 67, 380–400.
Snider, D.E. and Satcher, D. (1997). Behavioural and social sciences at the Centers for Disease Control and Prevention. American Psychology, 52, 140–2.
Spasoff, R.A. (1999). Epidemiological methods for health policy. Oxford University Press, New York.
Spector, M. and Kitsuse, J.I. (1987). Constructing social problems. Aldine de Gruyter, New York.
Stacey, M. (1992). Regulating British medicine: the General Medical Council. Wiley, Chichester.
Stallones, R.A. (1982). Epidemiology and public policy: pro- and anti-biotic. American Journal of Epidemiology, 115, 485–91.
Steinmo, S. and Watts, J. (1995). It’s the institutions, stupid! Why comprehensive national health insurance always fails in America. Journal of Health Politics, Policy and Law, 20, 329–71.
Tesh, S. (1988). Hidden arguments: political ideology and disease prevention policy. Rutgers University Press, New Brunswick, NJ.
Walt, G. (1994). Health policy. An introduction to process and power. Zed Books, London.
Williams-Crowe, S.M. and Aultman, T.V. (1994). State health agencies and the legislative policy process. Public Health Reports, 109, 361–7.