Why is health improvement policy so difficult to secure?

Many governments seek to improve population health, and some seek to reduce health inequalities. Yet, there remains a large gap between their policy statements and actual outcomes. Perennial questions in public health research include: why is this gap so large, why does it endure, and what can be done to close it? This essay uses political science and policy studies insights to address these questions, focusing on the distinctive issues that relate to (1) broad aims like ‘prevention’, (2) specific strategies for health improvement, or (3) new events. On the one hand, the idea of ‘prevention’ has widespread appeal, when governments think they can save money or reduce inequalities by preventing problems happening or worsening. While health protection seeks to inoculate populations against communicable diseases, health improvement strategies, including ‘Health in All Policies’ (HiAP), primarily address non-communicable diseases (NCDs). Further, the coronavirus disease 2019 (COVID-19) pandemic highlights the unequal spread of ill health, showing that preventive health ideas should be at the core of policy. On the other hand, there is a large gap between rhetorical and substantive commitment to prevention, a continuous HiAP implementation gap, and a tendency for COVID-19 health protection to overshadow health improvement. Explaining each problem clearly helps to identify the factors that generally undermine prevention policies and those specific to more detailed strategies like HiAP or events like COVID-19. We do not prioritise leadership or ‘political will’ as the policymaking problem. Instead, we identify the systemic factors that apply to even the most sincere, competent, and energetic policymakers. Health improvement policy is typically undermined by a lack of: clarity about what prevention means in practice; congruity between the prevention agenda (emphasising the need for major change to policy and policymaking) and routine government business; and, capacity to overcome obstacles to policy change.


Introduction
There is a profound and continuous gap between the rhetoric and practice of preventive policymaking.We use political science and policy theories to examine and inform the attempts by public health researchers and advocates to close it.We then use these insights to explore the future of preventive health in the context of COVID-19, focusing on health improvement (also known as health promotion) strategies to prevent noncommunicable diseases (NCDs) such as heart and respiratory diseases, strokes, cancers, and diabetes.
The idiom 'prevention is better than cure' has rhetorical weight across many governments.It sums up the idea that governments can save money and reduce inequalities by engaging preventively, to stop problems before they happen or prevent them getting worse (Kennedy, 2020).Policy initiatives backed by this general idea can be found in sectors including social policy, education, and criminal justice, and cross-sectoral initiatives such as 'preventive spending', and can be traced back to initiatives over the past century (Cairney & St.Denny, 2020).
Further, prevention has particular resonance in public health.A focus on health protection offers the chance to prevent illness by inoculating populations against pandemics of communicable disease.A focus on health improvement offers the chance to improve the social, economic, and physical environment to prevent an epidemic of NCDs and improve health equity.For example, the approach known as 'Health in All Policies' (HiAP) sums up the latter: highlighting the 'social determinants' of health (and unfair health inequalities), describing health as a human right and the need to pursue health equity, and proposing high levels of cooperation across government to produce policies to address NCDs (WHO, 2014).This focus on social or structural determinants challenges a tendency to relate health inequalities primarily to biology ('there is no biological reason for their existence ', Whitehead & Dahlgren, 2006: 4) or individual 'lifestyles' in relation to healthy eating, exercise, and the avoidance of smoking and alcohol.It identifies the profound impacts on population health from (a) environments outside of an individual's control, in relation to threats from others, such as pollution or violence, (b) education and employment, and (c) economic inequality, influencing access to warm and safe housing, high quality water and nutrition, choices on transport, and access to safe and healthy environments (Bliss et al., 2016: S88).It also warns against relating health improvement to health care, since most policy solutions are in non-health sectors (De Leeuw & Peters, 2015).
Most recently, the COVID-19 pandemic has provided stark warnings of the importance of social determinants and health inequalities (Bambra et al., 2021).First, the prevention of NCDs is central to reducing the impact of COVID-19.Vulnerability to major illness and death relates strongly to heart disease, diabetes, and lung conditions (Kluge et al., 2020).Second, it highlights social determinants in relation to self-isolation and social distancing: some people have access to food, private spaces to self-isolate, and open places to exercise away from others; many have insufficient access to food, no private space, and few places to go outside; and, this ability varies widely within and across countries (Shadmi et al., 2020).

Amendments from Version 1
We have made three main changes to address the issues of clarity identified by both reviews: 1. We expanded the initial discussion of three key concepts -clarity, congruity, and capacity -to explain their meaning further.
2. We reorganised key sections -the introductory explanation and discussion of research questions, and methods -to improve essay structure.
3. We have made a series of small changes (largely to aid issues of clarity raised by reviewers) throughout the text.

Any further responses from the reviewers can be found at the end of the article
There is mutually reinforcing harm from infectious disease, chronic disease, and 'social conditions' (often described as a 'syndemic ', Bambra et al., 2020: 965;Todd & Bambra, 2021).
If we combine these reasons to support prevention policies, wouldn't we expect them to rise to the top of national government and international agendas and stay there until policymakers saw major improvements in population health outcomes?The answer from policy research is an emphatic no (Cairney & St.Denny, 2020).Prevention generally receives only lip service, particularly among 'neoliberal' governments that emphasise personal responsibility and resist calls for major state intervention (De Leeuw & Peters, 2015: 987-8;Oni et al., 2019).HiAP studies highlight limited progress.COVID-19 and health protection knocked health improvement off the policy agenda in most countries (WHO, 2020).
In that context, this Essay serves as a way to step back, reflect, and comment on what we know about these enduring policy problems, based on insights from political science and policy process literatures.We ask three separate questions to highlight the relationship between the factors that generally undermine prevention policies (not only in health) but may arise in specific ways in relation to health improvement strategies such as HiAP or government responses to COVID-19.
Section one identifies how to explain the general lack of prevention policy.Generally speaking, why is there such a gap between rhetorical commitment to the broad idea of prevention versus actual policies, practices, and outcomes?
Section two connects broad insights on prevention to the specific study of health improvement strategies such as HiAP.If they are designed to help overcome such problems, why is there still a large and continuous implementation gap?
Section three explores what additional problems are associated with COVID-19 policies, in which a focus on health protection overshadowed and undermined health improvement (WHO, 2020).It uses insights from sections one and two to inform research and practice, focusing on the future of public health policies: what will happen when governments disinvest in their COVID-19 emergency responses and return to a greater focus on health improvement?In public health research, it is common to cite one political science perspective -Kingdon (1984) -to signal that a new 'window of opportunity' for health improvement policy may open.However, sections one and two provide cautionary tales to underpin future strategies.They show that vague political agreement -to mainstream health across government -is no guarantee of substantial action, and the production of yet another new strategy is futile without knowing how consistent it will be with routine government business.The window of opportunity may open to produce a half-baked solution to an ill-defined problem.
This three-stage focus helps to identify how policymakers and practitioners can draw lessons on prevention policy and health improvement during the absence and presence of crises.
What can they learn from each experience to help address this prevention puzzle?

Methods and approach
This Essay employs narrative synthesis to identify general and specific lessons for research and practice.Section 1 synthesises insights from our previous work that applied political science and policy theories to multiple case studies of prevention and public health policies (including Boswell, 2016a;Boswell 2016b;Boswell et al., 2019;Cairney et al., 2019;Cairney, 2020;Cairney & St.Denny, 2020).We generate a simple framework that categorises common obstacles to policy and policymaking reforms, using alliteration -describing clarity, congruence, and capacity -to aid the narrative.Section two connects these broad insights to the qualitative systematic review of HiAP studies that Cairney and St.Denny co-authored (with Heather Mitchell) in Open Research Europe.The detailed account of our systematic review methods, reproducibility, ethics, and further data can be found in Cairney et al., 2021b.Here, we focus on its take-home messages, which we use in section three to narrate key issues of policymaking in the COVID-19 era.
The final part of our approach is to seek feedback from practitioners on the usefulness of this essay when considered in different contexts, and to compare our insights with theirs.
To that end, our first step was to share insights in an academicpractitioner workshop in Scotland, in May 2022 (summarised in Cairney et al., 2022a).Our aim is to use the flexibility of the Open Research Europe platform to prompt further reflection and update the Essay in line with future workshop activity.

Section 1. The general problem with prevention policies: explaining the gap between rhetorical commitment and actual practices and outcomes
Prevention is a recurrent theme in efforts to improve policymaking and respond to complex policy problems.It relates to issues that cut across policy silos and sectors, to tackle problems like inequalities (Cairney & St.Denny, 2020), climate change (Bradford et al., 2018), social deprivation (White, 2017), and criminal justice (Sherman & Eck, 2002).It travels across levels of policymaking, in the documents of the WHO (Mendis, 2010), EU (Mackenbach, 2006), national governments (DHSC, 2018;National Preventative Health TaskForce, 2009), and subnational governments (Craig & Robinson, 2019;Haynes et al., 2020).Its logic is boosted by idioms and metaphors: 'prevention is better than cure'; 'a fence at the top of the cliff is better than an ambulance at the bottom' (DHSC, 2018;Shallowe et al., 2020).It links closely to 'futures thinking', 'horizon-scanning' or 'anticipatory governance' (Guston, 2014;Kuosa, 2016;van Rij, 2010).Almost everyone sees prevention as an idea worth pursuing.Yet, actual practices and outcomes fail to match this rhetoric, contributing to a cycle of enthusiasm in theory and despair in practice (Cairney & St.Denny, 2020).
A synthesis of political science and policy studies insights helps to explain such dispiriting outcomes.These explanations begin with general reference to two foundational concepts (Cairney, 2020;Cairney et al., 2022b).First, bounded rationality describes how policymakers deal with the limits to their ability to process information and make choices.They draw on cognitive and organisational shortcuts: 'rational' ways to prioritise trusted sources of information, and 'irrational' ways to draw on their gut instincts and emotions to draw quick conclusions (Cairney & Kwiatkowski, 2017).Health studies respond to this dynamic by promoting 'evidence based policymaking' and generating more information to reduce policymaker uncertainty (Cairney, 2016).Political science accounts highlight the importance of ambiguity: there are many ways to define or 'frame' a policy problem, prompting policy actors to exercise power to draw policymaker attention to their preferred framing at the expense of others, and generating demand for evidence in that context (Cairney et al., 2016).
Second, policymaking complexity describes the environments over which policymakers have limited knowledge and even less control, summed up with reference to five concepts: 1. Actors.There are many policymakers and influencers spread across many levels and types of government (otherwise known as venues of authoritative choice).

2.
Institutions.Each venue has its own rules and norms.Formal rules are written and well-understood.
Informal understandings are unwritten and often communicated non-verbally.

3.
Networks.Elected policymakers delegate most policy to bureaucrats, who seek information and advice from interest groups and experts.Policymakers and influencers trade access for advice, forming networks that often exclude most other actors.

4.
Ideas.Each venue or network operates with reference to well-established beliefs about the nature of the policy problem and the feasibility of each solution.
5. Context and events.Policymakers must respond to socioeconomic factors and crises that are not in their control (Cairney et al., 2019: 6-7).
These concepts inform studies of: (1) multi-centric policymaking, in which policymaking power is spread across many venues rather than concentrated in a single core executive, (2) governance rather than government, in which policymaking is often characterised by blurry boundaries between those who make, influence, and deliver policy, and (3) complex policymaking systems, in which policy outcomes seem to 'emerge' in the absence of central government control and 'implementation gaps' are routine features, not bugs (2019: 7; see Harris et al., 2014;Harris et al., 2015 for applications in public health policy).
We use these insights to identify three causes of the prevention puzzle.This distinction often relates to an unclear ecological metaphor: upstream measures seek to change the social environment, while midstream and downstream measures target companies or mitigate the effects of the environment on at-risk groups or individuals (McMahon, 2021a;McMahon, 2021b).
For example, Boswell's (2016a) study of political debate on obesity in Australia and the UK uncovers competing narratives that lie beneath superficial consensus on the value of prevention.For some advocates, prevention would mean engaging in radical policy and policymaking changes, seeking (1) fundamental socio-economic transformation to alleviate the social inequalities that drive obesity, and/ or (2) the strict regulation of the production and marketing of food to mitigate 'obesogenic environments'.Outside this world of public health, moderate ideas enjoy greater currency.Support for preventing obesity typically means investing in medical interventions, or social marketing campaigns to encourage populations to make healthier choices.The shared commitment to prevention in principle eventually gives way to an intense but often hidden political contest over the meaning in practice.
This temporary agreement is a recurrent feature across the many settings, sectors and issues for which prevention holds so much appeal.Policy research shows that ambiguity provides a valuable resource to policymakers as it enables widespread buy-in for general ideas (see Yanow, 1996;Zahariadis, 2003).
In practice, low-profile contests over the meaning of prevention generally get resolved in favour of minimal policy change (Cairney & St.Denny, 2020).
Congruity: prevention is out of step with routine government business Ambitions for major changes to policy and policymaking flounder when they are incongruous with 'business-as-usual' policymaking.A wide variety of terms exist to describe causes of policy inertia in policymaking systems, including 'inheritance before choice' (the first act that new governments make is to accept the policy commitments of their predecessors), 'policy succession' (most 'new' policies are actually modifications of old programmes), and 'path dependence' (choices made in the past contribute to the rules and practices that still govern the present) (Hogwood & Peters, 1983;Mettler & SoRelle, 2018;Pierson, 2000;Rose, 1990).They all suggest that, while governments find the language of radical policy change useful, it does not really sum up what they do in practice.This is particularly true when the adoption of radical prevention initiatives relies on a 'leap of faith ' (Botterill, 2006) that policymakers are unwilling to take, or when those initiatives would present a challenge to their higher salience aims (such as to prioritise particular economic models, or high funding for healthcare).
Major policymaking reforms are difficult enough when everyone generally agrees on what to do, and present often-insurmountable obstacles when aims and outcomes are contested.
The first obstacle is political short-termism.The promise of prevention is in long-term and uncertain outcomes.However, most decisions are made on a short timeframe between elections, which focuses attention on more immediate and tangible costs and benefits (MacKenzie, 2016).Or, the urgent crisis today crowds out the potential crisis tomorrow (Hammond & Smith, 2017;Mazey & Richardson, 2021;Rhodes, 2011;van Dorp & Hart, 2019).The most recent example is COVID-19, in which the urgent needs of health protection have enabled massive intervention often at the expense of hard-fought efforts to enable health improvement and tackle the long-term drivers of chronic disease (WHO, 2020).
The second obstacle is the bureaucratic politics largely hidden from public view.In some narratives, there are clear intentions to reproduce the status quo.Here, we see how institutions -'the way things work around here' -are reinforced by the networks of actors that influence, make, and deliver policy (March & Olsen, 1984;Rhodes et al., 2006;Schmidt, 2010;Streeck & Thelen, 2005).These rules, norms, and routine practices help mollify intervention.Behind the scenes, powerful actors use these rules to thwart radical action (Hawkins & Holden, 2014;Miller & Harkins, 2010).In other narratives, these dynamics relate to complex policymaking systems, in which the interaction between many actors -reproducing a multiplicity of (often contradictory) rules -can undermine new initiatives regardless of intent (Cairney, 2020: 104-7).
Consequently, there are regular efforts to 'institutionalize' prevention in policymaking: creating organizational structures or dedicated agencies within government to offer leadership and challenge short-termism and delay (Boswell et al., 2019;Smith, 2020).In practice, these agencies account for a tiny proportion of a sector's budget and frequently fail to disrupt policies and outcomes, with powerful rival organizations in and out of government rendering their status precarious (Cairney et al., 2021b: 15).The radical nature of prevention makes it difficult for agencies to establish credibility and value, especially among the actors whose practices they challenge, and when the evidence for the impacts of interventions are hard to demonstrate during normal policymaking timelines.They become marginalized by the 'business-as-usual' practices of evaluation and accountability in policymaking.

Capacity: low support for major investments with uncertain rewards
Capacity can relate to agency, to describe the limited 'will' to pursue policy change, or context, to describe the limited resources that governments have to coordinate policy change in systems beyond their control.The latter can result from choice, to restrict the role of the state or share responsibilities, but also necessity, to reflect the limits to central control that we describe at the beginning of Section one (Cairney et al., 2019: 6-7).These factors help to explain why policymakers frequently conclude that many government reforms are not feasible in relation to the immediate constraints in which they take decisions.Indeed, the predominant story about policy capacity, especially in liberal democracies, is of state retrenchment and increasing reliance on third and private sector actors to make and deliver policy (Hardiman & McCarthaigh, 2017;Wu et al., 2015).
Preventive policymaking has some additional problems with capacity.First, the common perception is that many preventive measures are prohibitively costly with no immediate payoffs.
Preventive measures might be economical on long-term timescales; indeed, that is their founding logic, regarded as 'self-evident' in public health (Capewell & Capewell, 2018).However, the immediate reality is of tight budgets and limited bureaucratic capacity.New initiatives struggle to win support, given the tendency of governments to see their current balance of taxing/spending as not for negotiation.Even interventions with a history of investment and widespread legitimacy can face budget cuts.
Second, there is often a lack of capacity or commitment to overcome major opposition to policy change.Government coordinative capacity is limited when relying heavily on governmental and non-governmental organisations to make and deliver policy.In theory, they could combine delegation with strong regulation, but in practice most governments tend towards a 'hands off' approach (partly to help avoid blame for policy outcomes, Cairney et al., 2019: 9;24-5).Further, when preventive measures entail 'hard choices' that impact on powerful industries, governments prefer to persuade than compel (Godziewski, 2022).
These general and specific limits combine in different ways.
For example, the Sure Start early intervention programme in the UK was initially treated as a high-priority and successful programme to detect and address social, health and educational issues in families.However, the cross-cutting nature of 'families policy' presented general problems of coordination across multiple departments (most of which did not treat it as a priority), and it became difficult for the Labour government to measure and describe its success in ways that were conducive to long-term political support (Cairney & St.Denny, 2020: 93-8).Indeed, it was then scaled back as part of the austerity measures of the new Coalition government in 2010 (Torjesen, 2016).
This kind of experience differs from the more contested arenas of public health -such as tobacco, alcohol, and food policies -in which there is clearer opposition to state intervention.
In such cases, industry actors are relative experts at exerting influence throughout the policy process, and enjoy a particular advantage in impacting the low-profile institutions and practices that turn vague agreement (on prevention as a goal) into tangible action (Boswell, 2016b).The classic examples are voluntary or self-regulated compliance schemes for industry, which symbolize low government commitment and capacity to enforce policy change (Baggott, 1986; although examples such as tobacco control show the potential to shift towards regulation -Cairney & St.Denny, 2020: 149-53).Global trends highlight a shift to asking food companies to restrict and monitor marketing of unhealthy products to children, and alcohol and sports betting companies to fund social marketing on responsible drinking and gambling.However, public health actors are naturally sceptical of the claims made for these schemes (often based on their cynical use in the name of tobacco control, Cairney et al., 2012).Most evidence suggests that these initiatives fail to deliver on their purported 'efficiency' benefits of working with, rather than regulating, industry (Hastings, 2012;Lacy-Nichols et al., 2020).Nevertheless, their prevalence speaks to the broader point: that policymakers feel reliant on private sector actors and prefer to engage in collaboration than outright hostility (Boswell et al., 2019;Godziewski, 2022).Cairney et al. (2021b: 8-13) show that this starter's kit informs a wider collection of commonly-dispensed advice on how to foster HiAP success (which they describe as the HiAP 'playbook'): 1. 'Use well-established ways -such as by using the WHO starter's kit -to get from talk to action, and to sustain long-term commitment' (2021: 8).
2. 'Raise awareness and connect HiAP to a government's values and policy agendas' (2021: 8).For example, frame HiAP aims to be consistent with a government's overall vision and useful to its core business.
3. 'Focus on win-win solutions to foster trust-based intersectoral action' (2021: 9).Generate the sense that HiAP provides mutual gains between health and other sectors, such as building trust and establishing the payoffs to collaboration.
4. "Avoid projecting a sense of 'health imperialism'" (2021: 10).Avoid the sense that HiAP represents interference in non-health sectors -undermining their core business and provoking reactions against HiAP -by contributing to shared agenda across sectors.
5. 'Identify policy champions and entrepreneurs' (2021: 10).Key actors use their knowledge, networks, and skills to kickstart HiAP and ensure regular progress.
6. 'Use HiAP to promote the routine use of HIAs' (2021: 10).HIAs are formal processes to help identify the value of non-health initiatives to reducing health inequalities.
7. 'Do not rely on a traditional cost-benefit-analysis case for HiAP ' (2021: 11).Find ways to establish the value of HiAP that do not rely on a narrowly defined economic case.

HiAP and capacity
There have also been efforts to build capacity to ensure sustainability for the HiAP agenda.Proponents have tried to establish a credible track record for HiAP initiatives, and to generate strategic support at a high level.One focus has been on establishing a robust and legitimate toolkit for analysis and evaluation, with HiAP as the driver and processes such as HIA (playbook 6) as contributors.This has meant challenging the value of simplistic cost-benefit approaches to understand overall costs, return on investment, or efficiency (playbook 7).
Advocates understand it is important to demonstrate the economic value of HiAP, but hold that it is difficult to make a short-term 'business case' because: ' 'the dominant narrative of HiAP in theory does not correspond to the meaning of HiAP in practice.The former is an ambitious strategy to address the social determinants of health with radical policy change across multiple sectors, facilitated by intersectoral action and high strategic commitment to produce support for better policies.The latter is an ambitious strategy on paper only, representing moderate policy change at best and a negative commitment at worst, particularly when the funding and allocation of staff is minimal in relation to the wider sector.' (2021: 28).
In other words, there only appears to be high commitment to HiAP when we view that commitment in isolation.It does not seem so impressive when we relate it to a much larger commitment to the status quo, in which there is no additional commitment to economically redistributive policies, or to boost welfare state provision, and a high and enduring commitment to treating highly-funded healthcare as the main solution to health inequalities.Further, a vague long-term commitment to fostering preventive population health and wellbeing approaches does not compete well with specific and short-term commitments to maintain reactive services.In that context, the idea that intersectoral action could help overcome the status quo is unrealistic.Indeed, the veneer of collaboration across and outside of government 'helps dilute the ambitions of HiAP enthusiasts, not produce policies they favour' (2021: 28).

Section three. The new COVID-19 problem: explaining its additional negative impact
All of these HiAP studies were written pre-COVID-19.However, building on Cairney et al. (2021b), we argue that this new policymaking dynamic is impossible to ignore, not least because it adds a new twist to the HiAP narrative, with implications for health improvement more generally.
First, COVID-19 should have prompted governments to treat health improvement as fundamental to public policy.Many had made rhetorical commitments to public health strategies to prevent NCDs, and COVID-19 reinforces this rationale.Social determinants relate to health improvement (health inequalities resulting from factors such as income and social and environmental conditions) and health protection (unequal resources to live and work safely).Further, COVID-19 had a visibly disproportionate impact on the mortality and health of people with underlying health conditions associated with NCDs (Bambra et al., 2021).Second , the opposite happened: health departments -in many countries -postponed health improvement and moved resources to health protection (WHO, 2020).Third, these events are particularly dispiriting since attention to one aspect of public health comes at the direct expense of the other.Practitioners in health improvement are used to dealing with slow progress in relation to fostering cooperation with other sectors and being undermined by other agendas (usually economic, relating to 'austerity' or 'neoliberal' policies) (Cairney et al., 2021b).In this case, they or their colleagues have been obliged to contribute to the reduced status of health improvement by shifting their energies to protection.
In that context, experiences from prevention policy provide a profound cautionary tale for the future.Many governments will be 'rebooting' or rethinking their health improvement strategies for a post-COVID world, and health improvement advocates will return to the language of 'windows of opportunity' for progress and the strategies summed up by the HiAP playbook.Yet, these approaches have not served them well, to the extent that it is worth stepping back to reflect on their future impact.Cairney et al.'s (2021b) review of HiAP studies suggests that there is limited internal learning from experience thus far.Most proponents still treat policymaking as a technical exercise and use policy theories instrumentally, to find the right language to define the problem, the solutions that work, and the right model of intersectoral action and implementation.In other words, the alleged solution is to refine the playbook.This approach continues to underestimate the impact of politics on policy, in favour of functionalist arguments: identifying which policies should be selected, and how policymaking should work, rather than what actually happens (Cairney et al., 2022b).
An alternative is to take politics and policymaking complexity more seriously, reflecting on experiences so far, and identifying ways to take them into account.These processes of reflection, analysis, and learning should be guided by a clear map of the challenges which are: (1) relatively general features of policymaking and delivery, and ( 2) more specific to health improvement ambitions and principles.

Getting clarity on what health improvement means in the COVID-19 era
Clarity can make the difference between minimal and maximal policy change.Advocates should not take for granted that the concrete measures required to put into practice the programmatic ideas associated with health improvement will flow naturally from initial commitments.Indeed, the opposite may be true.Securing high levels of national and international in-principle commitment to a vague idea can represent an alternative to substantive change (Cairney & St.Denny, 2020).
The challenge lies in clarifying what a health improvement strategy reboot might look like.For example, does it come with a detailed plan of action with a starter's kit and playbook or represent a more general 'preventive philosophy' for government?Is it a vehicle for the relatively top-down implementation of specific interventions, or to encourage collaborative governance in which health actors play an equal role?It is always tempting to answer 'yes' to all aims, but policymakers resolve ambiguity by making specific choices with resource implications.
If so, securing in-practice support for measures to redistribute resources, refocus policy horizons beyond the short-term, and bind sectorally disparate actors to common goals, represents a second level of policy change that needs to be negotiated and secured.Yet, as with most preventive policymaking initiatives, the HiAP literature contains a major internal contradiction that undermines this progress (Cairney et al., 2021b).Frankly, many HiAP advocates want to take control to produce policy instruments with specific ends (to reduce health inequalities) and give up control to work collaboratively with policy actors across and outside of government.They want to encourage collaboration to generate widespread ownership of policy change and reserve the right to reject the outcomes as not conducive to the HiAP agenda.In that context, if it is understandable to criticise policymakers for using ambiguity to avoid specific commitments, it is also reasonable to challenge health improvement advocates to address the ambiguity inherent in their approach.
Enhancing congruity between health improvement initiatives and the 'new normal' Key elements of health improvement seem particularly vulnerable to disconnection with established practices and routines of policymaking.By definition, a radical policy agenda is out of step with the way things are done.The short-term pressures of the pandemic response have also seen governments strip back resources for longer-term health improvement.
Recovering that knowledge, impetus and infrastructure will be a much harder feat than before, even if there is a 'window of opportunity' to do so.Crucially, the congruity challenge follows directly from a preferred definition of health improvement strategy based on two competing options.
First, if advocates see HiAP primarily as a global agenda to be adopted in similar forms in new domestic contexts, congruity is about altering radically the policies and structures of governments.If so, Cairney et al.'s (2021b) review identifies two congruity challenges that may never be overcome: specific HiAP units are too small and uninfluential to generate coordinative capacity, challenge business-as-usual government, or encourage new and sustainable ways of working; or, the formal reorganisation of coordinative mechanisms exacerbates coordination problems at the expense of more effective informal collaboration measures.
Second, if HiAP is a way to encourage intersectoral action and 'collaborative governance' (Ansell & Gash, 2008), congruity is about playing one small part in a policymaking project.Policy scholarship describes essential practices, including: (1) incorporating formulation and implementation challenges into policy design; (2) embracing collaboration in policy design and implementation; and (3) understanding that 'implementation' is an often-misleading term, since policy often 'emerges' in the absence of central control, requiring adaption, learning, and persuasion to respond effectively (Ansell et al., 2017;Crowley et al., 2020: 141-162).

Building capacity to sustain the health improvement agenda
The goal of self-sustaining capacity to deliver health improvement needs to be grounded in policymaking reality, as described by studies of public administration and policy theory-informed empirical studies.The former identify how to foster intersectoral action, such as when Carey & Crammond (2015: 1022-8; see also Greer & Lillvis, 2014) describe a supportive governance 'architecture', skilful and flexible leadership, a manageable number of aims, and a powerful narrative to represent a common purpose.The latter explain the limits to coordinative capacity, including the strong logic for policy specialisation and silo working, and a tendency for the distribution of policymaking responsibilities to relate weakly to the task.In other words, it is unrealistic to expect all actors to come together to produce 'coherent' policies and practices (Cairney et al., 2021a;Cairney et al., 2022).Intersectoral, collaborative, 'mainstreaming' initiatives require different actors spread across numerous sectors and operating at different levels to 'pull together' to deliver a multitude of complementary services.However, it would be a grave mistake to equate functional requirements with actual policymaking.
We have seen major cross-sectoral coordination activities in the crisis response to COVID-19, but these are short-term arrangements.Meanwhile, the formal architecture and informal relationships that existed for joined-up governance on health improvement have been de-funded and de-prioritised.Building back capacity for health improvement represents an acute governance problem, to harness the complex political and organisational systems required to produce ambitious collective impacts.

Conclusion: Approaching health improvement as a governance problem
Our focus on clarity, congruity, and capacity helps streamline the health improvement 'playbook' -focused on maximising coherent and effective inter-sectoral action -and relate it to conditions of complexity over which all policy actors have very limited control: 1. Clarity.Encourage and exploit a 'window of opportunity' to generate high levels of acceptance for health improvement in principle and 'lock in' a clear and detailed strategy (backed by regulations, earmarked finance, organisations, and guidance) to secure a lasting mandate.

2.
Congruity.Align health improvement to existing initiatives to capitalise on synergies and bring service delivery partners 'on board'.

3.
Capacity.Nurture critical mass, in terms of the actors working to deliver health improvement objectives, and encourage champions to show leadership and bridge the gaps between levels and sectors.
Each element of this strategy will inevitably face difficulties determined by the political, social, and organisational dynamics of each context.These challenges include: to generate or sustain attention to, and enthusiasm for, health improvement; political or social pressure to prioritise other state activities (including the management of the economy) over health improvement or prevention; and, resistance from key service delivery actors unwilling or unable to change operational procedures or organisational culture.Further, solving all of these issues while avoiding trade-offs -for example between top-down control and bottom-up ownership of, and compliance with, delivery activities -is impossible.Nevertheless, by approaching health improvement as a policymaking problem, it may be possible to identify clearly, and mitigate against, some of the most glaring pitfalls while engaging with inevitable governance dilemmas.
There is no straightforward way to take action on health improvement.In fact, our contemporary context -of unprecedented attention for public health -makes achieving health improvement aspirations more difficult.Attention, resources, and relationships poured into these efforts have been disrupted and displaced by crisis management.In these circumstances, the old HiAP playbook looks unpromising.While a focus on clarity, congruity, and capacity does not solve these problems, it encourages health improvement advocates to maximise their chances by being clear on what they seek to achieve and consistent and realistic in how they seek to achieve it.
The weaknesses of the article are drawn from this structure.I think all of them could be addressed by light textual refinements and perhaps a clear limitations section that would specify the applicability of the ideas outside the health improvement literature and context in which it receives its empirical content here.The two themes of my comments, which I hope are constructive, is that the discussion downplays the importance of raw politics and shows a strong influence from one particular literature and topic, HiAP, on what seems to be a more general theory.Reading Julia Lynch's work might be very helpful.
First, there is the possibility that all of this is cheap talk by governments, knowingly replacing policies that might work with ineffective ones; or knowingly adopting policies that won't work because no plausible alternative policies exist that might work.HiAP is almost tailor-made for governments to look busy if that is what they need to do, and also to win some support from health advocates who think half a loaf is better than no bread at all.As they note, "Securing high levels of national and international in-principle commitment to a vague idea can represent an alternative to substantive change."Is lack of clarity, in particular, a structural feature of HiAP or a deliberate feature introduced by policymakers for some reason such as coalition maintenance or reluctance to actually adopt an effective policy?
Second, the whole of interest group politics and, I think, party politics is to be found in one-half of one C (capacity).The authors are wholly aware of the existence of these powerful forces, but downplay them.This is not a big issue in the HiAP literature they review (since it is mostly about events within government, which means interests are mostly shadowy forces behind what look like noncooperative departments) but it is rather a big issue in the COVID-19 debate, where a variety of interests (e.g.workplaces that did not want to protect customers and workers) and partisan actors (often of the populist radical right) had a big influence on the implementation and adoption of policies.Much prevention policy directly harms somebody's business and leads to opposition.Is it right to collapse that into one component of capacity?Does the essay need a fourth C of, I don't know, clout?Otherwise the conclusion that "there only appears to be a high commitment to HiAP when we view that commitment in isolation" doesn't really have an obvious antecedent.
Third, HiAP is a very weird policy idea, with a superficial appeal (something for nothing!) and huge problems starting with a willful defiance of everything we know about bureaucracies and a name seemingly chosen to irritate people who work in "all policies."It disguises all sorts of interests behind what appear to be problems of bureaucratic coordination -a refusal to produce healthy transport systems might have something to do with lobbies for road transport, toll operation, and road building rather than fatheaded bureaucracy in the transport ministry (or, better, that seeming fatheadedness is simply one manifestation of a whole interest constellation).Further, it is primarily evaluated by health promotion and public health analysts with a particular way of thinking and a reluctance to give subtle political explanations for policy outcomes.Insofar as the case is the data validating the text, it needs contextualization of how it demonstrates the main theory; if it's illustrative, that should be made clear and some discussion of the limits of this example would be helpful.The plain-language summary contains a nice description of the cases that HiAP and COVID-19 responses are; it is not really followed through in the text but could be the basis for a simple categorization of which issues fit well with the general theory and which issues would require using it with different emphases.
Fourth, the essay underplays the distinction between adoption and implementation.A lot of the HiAP literature is about the relative failure of a plan, reorganization, or budgeting technique.In other words, implementation of something lighter than specific legislation.Does the theory work equally well to explain adoption of anything at all, adoption of the specific instruments (already indicators of government seriousness), and implementation of any specific instrument?
Finally, the bibliography needs work.I think a software program introduced gremlins; it is particularly hard to identify the authors' own work correctly.There are several one sentence paragraphs in the plain language summary and introduction.Perhaps these could be combined with preceding/subsequent paragraphs?
Is the bold text in the final paragraph above the section 2 heading ('The triple threat...') meant to be a heading?This looked a bit incongruous so I suggest maybe removing and integrating the point into the paragraph.

2.
These are just presentational suggestions so feel free to ignore and no need for any further formal revisions or responses.There is significant overlap between the content covered in answering these questions such that there seemed to be some repetition in what was covered under the clarity, congruence and capacity headings.Q1 is probably the question that is less clear out of the two.Is this question asking why such policies are not adopted, or is this about implementation?If the latter, how is it different to Q2? If, however, this question is more about non-decisionmaking, perhaps that could be made clearer.

1.
What is the relationship between the key terms 'prevention', 'health improvement' and 'health-in-all policies'?Are they all synonymous with each other?Or is one of them an overarching term of which the others are subsets?Or is the problem that all these terms are used, but each have somewhat different ambits?I ask this because some of the points made in the article may apply more directly to some but not all of these terms.Health improvement, for example, generally covers what happens in health services as well as broader social determinants and inter-sectoral aspects of health.

2.
The discussion about clarity on p5 is very insightful.I couldn't help wondering whether the issue about clarity is about the definition of problems or of solutions or both.Much of the discussion in these paragraphs seems to be about the lack of clarity around problem definition.However, it could be argued, a la Lindblom that this doesn't really matter as long as there is some broad agreement about particular solutions.So if we take Sure Start as an example, many government agencies and interest groups may support Sure Start for many different reasons -many of which have little to do with health improvement.As long as the 'programme' is reasonably clear, does fuzziness about the problem(s) meant to be solved really matter?Regarding obestity, we have a rather different dynamic in which there is agreement that 'something needs to be done', but little agreement about what.So where 3.
precisely is the challenge around clarity -in problem definition or solution specification?
I think the article works best when discussing the challenges of inter-sectoral and health-inall approaches to prevention and health improvement (of which Sure Start seems to be the most tangible example).For this type of initiative, the arguments about clarity, congruity and capacity work very well.However, the authors seem to be referring to more 'traditional' public health policy concerns in the section on p6 under 'capacity'.Here the power of industry is discussed, and this seems most relevant to debates about accessibility, advertising and pricing of alcohol, junk-food, tobacco and gambling (what some refer to as the commercial determinants of health).In these types of issues, is it really capacity that is lacking, or is it power, plain and simple?States do have plenty of regulation capacity, but they are often not willing to use it in the face of real and anticipated industry opposition.At the very least, these are very different types of capacity constraints to those faced by intersectoral policy initiatives such as Sure Start.

4.
For Question 3, what exactly is the irony referred to?Some may argue that a focus on protection is inevitably in conflict with a focus on prevention as public health policy resources and capacities are inherently limited.A broader question about COVID-19 and the relationship between protection and prevention is whether the things that the authors are referring to are generalizable or specific to country contexts.In my own country, both agendas are playing out simultaneously.

5.
Another point made under the answer to Q3 is the tension between top-down, hierarchical stipulation and allowing room for bottom-up collaborative experimentation.This seems to be a much broader point about a fundamental tension of governance in any policy area.Perhaps it is not surprising that we would find advocates within public health for both approaches.Are they the same advocates?Is this substantively different or unique to what is happening in other policy areas?Or is there something particularly peculiar about the way this governance tension plays out in health prevention and improvement? 6.
Is the topic of the essay discussed accurately in the context of the current literature?Yes

Is the work clearly and cogently presented? Yes
Is the argument persuasive and supported by appropriate evidence?Yes Does the essay contribute to the cultural, historical, social understanding of the field?Yes Competing Interests: No competing interests were disclosed.
Reviewer Expertise: Health policy, public policy, policy implementation I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
prevention and its successful translation (and implementation) into policy outputs using relevant concepts from political science and policy studies (i.e.bounded rationality, policy complexity).As such, it address an important real world problem and one which has presented both a conceptual and empirical challenge to health policy actors and researchers.The attempt to deepen understanding through the application of cross-disciplinary theoretical insights is to be welcomed.The authors do so by identifying 3 key factors -clarity, congruity and capacity -as explaining the prevention 'gap' in health policies, which are applied to analyse health in all policies and covid 19 responses as policy case studies.
While the first and the third of these are intuitively plausible factors, congruity, and the mechanisms through which they impact on policy are set out in relation to previous studies, the concept of congruity, and its explanatory power in this framework, are less clearly explained.It could be argued that a lack of congruity between rhetorical commitments to prevention and the capacity of the policy-making architecture to deliver this restates the problem that the article seeks to explain rather than identifying characteristics of the policy process that account for this.I think this section (and the relevant passages of the introductory and background sections introducing the concept), therefore, need to be revised to clarify precisely what is meant by congruity and what the specific insights to be gained from this concept are.What it seems to be about is the inability of sedimented policy systems to facilitate 'radical' policy change?But this seems to be less of an issue of congruity with health prevention than a more general issue of path dependence and institutional memory and the default to the status quo?Another issue is how this interacts with issues of political will to force change despite this inertia.Can the issue of committed policy entrepreneurs be completely set aside as the introduction seems to suggest?Related to this, I think that the concept of radical policy change needs to be more clearly introduced and explained.Does this simply mean doing prevention effectively or is it a specific model of prevention (for example, on page 5, column 2, 1st paragraph, it is unclear if 'radical advocates' are the same as 'proponents in public health' or they hold two different positions)?
The article could also benefit from some restructuring and material being moved elsewhere.The first paragraph of section 1 reads more like introductory material to me.This would be useful as the introduction at present doesn't set out clearly the main focus and argument of the article.I suggest revising so that it frames the article more clearly for the reader.Related to this, the methods section seems to contain material that belongs more logically elsewhere while not really explaining completely clearly what was done.I suggest you consider moving the material in paragraph 1 to the introduction and adding an additional paragraph on how the article was conceived and executed.Paragraph 2 reads like a summary of the argument or introduction to the main argument and could go in the intro or as an opening to the main sections (currently labelled 1, 2, 3).
On page 5 the point about temporary agreement: is it temporary agreement or people just talking about 2 different things using the same label?I'm not sure that it is primarily about temporality so I suggest rephrasing.This paragraph also seems to contain important issues -contests over meaning and to define policy issues -which could be further unpacked and explored as reasons for lack of clarity and the prevention gap.Engagement with the literature on policy framing could assist.Are there citations or evidence for the final sentence?
The case studies are well presented and explained and the inference drawn about the specific findings on these and about health policy are generally are well made.

Minor points:
Abstract: Line 3, change to '…policy statements on the one hand and practices and outcomes on the other.'?

○
Health prevention and improvement are both used.Are these the same in your view or do they differ?Perhaps use just one here and/or define the terminology in the main article.Plain language summary: This leads with covid but the article is not principally about this.The meaning of the first sentence is also unclear.

○
Is 'social determinants of health' understood without explanation to non-specialists?Likewise the idea of these being 'distributed evenly' and the protection prevention distinction?
the topic of the essay discussed accurately in the context of the current literature?Yes Is the work clearly and cogently presented?Yes Is the argument persuasive and supported by appropriate evidence?Is the topic of the essay discussed accurately in the context of the current literature?Yes Is the work clearly and cogently presented?Yes Is the argument persuasive and supported by appropriate evidence?Yes Does the essay contribute to the cultural, historical, social understanding of the field?Yes Competing Interests: No competing interests were disclosed.I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.Reviewer Report 02 September 2022 https://doi.org/10.21956/openreseurope.16281.r29986© 2022 Hawkins B. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Ben Hawkins MRC Epidemiology Unit, University of Cambridge, Cambridge, UK The authors have engaged with, and responded thoughtfully to, my previous comments.I think the argument is now much clearer and I am happy to recommend indexing it in its current form.My only further (very minor) comments are:

1 Reviewer
the topic of the essay discussed accurately in the context of the current literature?Yes Is the work clearly and cogently presented?Yes Is the argument persuasive and supported by appropriate evidence?Yes Does the essay contribute to the cultural, historical, social understanding of the field?Yes Competing Interests: No competing interests were disclosed.Reviewer Expertise: Health policy, policy studies I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.Version Report 18 July 2022 https://doi.org/10.21956/openreseurope.16031.r29568© 2022 Tenbensel T. This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Tim Tenbensel 1 Associate Professor, School of Population Health, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand 2 Associate Professor, School of Population Health, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New ZealandThank you for the opportunity to review this article.I am very interested in the work of these authors and I have found their contributions on the topic of health improvement and prevention policy to be incredibly useful in both teaching and research.It is great to see this synthesis.I am very sympathetic to the overall argument that there is something about health improvement and prevention policies that is extremely challenging in the context of 21 st century government, policy and governance.The work of these authors effectively distils a broad body of policy studies literature relevant to the questions they pose.Overall, this is already a very strong and well-argued contribution to the literature.The comments and suggestions I have may be useful in sharpening and strengthening it further, hopefully.The article is structured according to three research questions outlined on p3-4.Having looked at the whole article, it wasn't quite clear to me how Q1 and Q2 relate to each other.

○
Second paragraph: The opposite of what?○ Third paragraph, sentence 2: the logic is unclear.Next sentence: If these gaps are inevitable why bother trying to reduce/end them?Rephrase.○ Final paragraph is unclear to me.○ Main text: I found the numbering of the main sections of the article confusing and suggest deleting or editing to reflect their position in the text.○ P4 column 1 last paragraph before 'clarity' heading.Has a long complex sentence that was unclear.Rephrase?○ Heading 'interpreting covid in that old context' doesn't sound quite right.Rephrase?○ Next heading: should it be 'the' (not 'a') covid 19 era? ○ Is the topic of the essay discussed accurately in the context of the current literature?Yes Is the work clearly and cogently presented?Partly Is the argument persuasive and supported by appropriate evidence?Yes

The triple threat of low clarity, congruity, and capacity.
The general dynamic of prevention policy suggests that vague rhetorical commitment gives way to power and politics in policymaking systems that defy control.A lack of clarity about the meaning of prevention enables initial buy-in and momentum from a broad coalition of actors.However, these actors have incommensurate ideas about what prevention ought to entail.The more ambitious and radical hopes of public health researchers and practitioners tend to get dashed as this ambiguity is thrashed out in policy work.These hopes tend to lack congruity with 'business-as-usual' in comparison with more moderate solutions.Governments also feel they lack capacity to invest in ambitious and radical policies, or -in many cases -to tackle powerful industry actors head-on.The result is a perpetuating cycle of enthusiasm and frustration, where bright ideas and good intentions founder on the mundane realities and constraints of policymaking(Cairney &  St.Denny, 2020).

The problem with health improvement strategies: explaining the 'implementation gap' in Health in All Policies
Cairney et al.'s (2021b)des much-needed context during the study of Health in All Policies (HiAP).If we know that vague commitment and high aspirations tend to translate into unfulfilled expectations, we can explain limited progress.Yet, health improvement advocates seek also to overcome such obstacles, to see HiAP as an ambitious and highly supported strategy, containing a coherent narrative, and supported by (whatCairney et al., 2021b call)a 'playbook' to turn strategy into action.If so, unlike the vague concept of prevention, HiAP strategies appear to have more substance.A government commitment to HiAP comes with a model and set of instructions to carry it out.In that context, we might expect more progress since there appears to be more clarity and commitment.However,Cairney et al.'s (2021b)review shows that HiAP encounters very similar problems.First, it is not as clearly defined as it first appears and its playbook (or list of actions to foster high commitment and cooperation) does not help to deliver improved health equity.Second, researchers frequently bemoan a temporary 'implementation gap' which may be better interpreted as the routine consequence of governance in complex policymaking systems.3."Identifyevidence-based 'upstream' solutions" to address and mitigate inequality as early in people's lives as possible (2021: 7).4.'Promote intersectoral action and collaborative governance' to ensure a coherent response across (and Cairney et al.'s (2021b: 12-20)reflections on empirical studies suggest that HiAP is too abstract to translate into action in a straightforward way(Huang et al., 2019: 2).Rather, its meaning and definition can vary markedly(Storm  et al., 2014: 184).As such, 'every HiAP initiative is uniquely designed and governed, and so it is challenging to understand how to translate studies of one case to others'(Shankardass  et al., 2018: 2).Indeed,Cairney et al. (2021b: 12-20)find that almost all of the terms essential to the meaning of HiAP remain ambiguous, including 'social determinants of health', 'upstream' measures, collaboration, 'political will'(Baum et al.,  2020; Carey & Friel, 2015; de Leeuw, 2016; O'Flynn, 2016;  Post et al., 2010).Therefore, any sense of coherence would be misplaced, contributing to confusing advice on how to operationalise and deliver HiAP aims.HiAP and congruityMany HiAP advocates have tried to ensure congruity with policymaking norms and routines.A key focus has been on framing HiAP as consistent with a government's overall vision and useful to its core business (playbook point 2), in the hope that HiAP becomes mainstreamed throughout government policy and an accepted way to judge performance(Freiler  Delany et al., 2016: 889; see also Breton, 2016: 383-4).A perception of external interference can lead to defensive interactions, poor collaboration, jurisdictional conflicts, and opposition to the extra health-focused work that comes at the expense of current commitments(Gottlieb et al., 2012: 158; Guglielmin  et al., 2018: 287-90; Lawless et al., 2012: S15; Newman et al.,  2014: 54; Oneka et al., 2017: 836; Synnevåg et al., 2019: 7).
Cairney et al. (2021b) relevant HiAP study describes (from the perspective of HiAP researchers) a dispiriting 'implementation gap' that is not overcome by the playbook.In practice, HiAP encourages considerable wriggle room as specialists interact and collaborate with other policyactors.etal., 2013: 1070; Greaves & Bialystock, 2011: 407; Molnar  et al., 2016: 2-3).Examples include framing HiAP as: a way to reduce the unsustainable burden on health services(Kickbusch  et al., 2014: 187-8); 'stimulating economic productivity'(Delany et al., 2016: 888); and, essential to 'EU core values such as solidarity, equity and universality'(Bert et al., 2015: 45).Cairney et al. (2021b)identify a large number of studies which highlight limited progress in relation to overcoming siloed, sectoral conflicts.Key issues include the sense of 'health imperialism' that arises when 'problems and the necessary actions are defined from the viewpoint of the health sector only' (Yet, in practice, these cross-sectoral gains are not easily won, nor sustained over time.Without careful deliberation and engagement to sustain buy-in, HiAP strategies and initiatives still trigger professional-identity-driven opposition to 'health imperialism' (despite attempts to address this problem in playbook 4).Further, such problems with informal ways of working are not solved by formal reorganisations.Holt et al.'s  (2018)account of HiAP in Danish municipalities concludes that sectoral re-organisations 'tend to reproduce the organizational problems they are intended to overcome', suggesting that 'It is time to dismiss the idea that intersectoral action for health can be achieved by means of a structural fix' (2018: 48).
Are they just (as opposed to primarily) about NCDs or as covid showed is it also about susceptibility to other morbidities including infectious diseases?Not sure what 'tethered to political reality' means.Rephrase/clarify.Next line: systemic policymaking dynamics is a bit unclear.Could this be reworded?Maybe 'characteristics of the policy-making system'?