Power analysis in health policy and systems research: a guide to research conceptualisation

Introduction

Power is defined as the ability or capacity to ‘do something or act in a particular way’ and to ‘direct or influence the behaviour of others or the course of events’.1 Relationships of power shape societies, and in turn, health policies, services and outcomes.2 Power dynamics—or the relational power that manifests in the interaction among individuals and organisations—also influence health systems, or ‘the organizations, people and actions whose primary intent is to promote, restore or maintain health’.3 The universe of power dynamics that are pertinent to the study of health policies and systems includes diverse types and locations of policy, social, implementation and political processes. Power dynamics have also influenced health systems planning and research, by defining what is seen as a health system, and the translation or adaptation of health systems models across distinct geographic contexts over time.4 5

Studying power is thus a core concern of researchers and practitioners working in the field of health policy and systems research (HPSR), an interdisciplinary, problem-driven field focused on understanding and strengthening of multilevel systems and policies.6 Accelerating theoretical development and empirical research on power in this domain is crucial for several reasons. First, it provides a deeper, more nuanced understanding of the mechanisms and structures that lead to social inequities and health disparities.7 Second, it reveals historical patterns entrenched in health and social systems, allowing contemporary policy concerns to be seen in a wider context and lessons to be drawn from these trends.8 Third, analysing power can contribute to the (re)design or reform of health systems to redress imbalances and progress towards improved health outcomes.9

Studies incorporating examinations of power in public health and HPSR have gradually increased in number, including, for example, analyses of accountability, political prioritisation, commercial determinants of health, determinants of universal health coverage and state sovereignty in health agenda setting.10–15 Nonetheless, explicit analyses of power in HPSR remain relatively infrequent.7 16 Lack of a power-specific lens may reflect the continued dominance of biomedical and behaviouralist approaches in health research and funding, limitations stemming from the political economy of research funding and agendas, and reluctance among institutions and individuals to examine their own role in perpetuating existing power dynamics.17 18 Power is also complex to examine conceptually, theoretically and methodologically. Seminal publications providing guidance on different aspects of power research include Erasmus and Gilson’s19 paper on investigating organisational power; the health policy analysis reader edited by Gilson et al,20 and Loewenson et al’s21 methods reader on participatory action research (PAR). Recent resources also provide conceptual overviews of power.7 9 22 However, there remains no comprehensive resource that can serve as a theoretical and methodological starting point for aspiring power researchers, irrespective of disciplinary orientation or area of HPSR interest.16

This paper aims to fill this gap, building on the above-mentioned resources but providing a more consolidated guide to researchers wishing to consider, design and conduct power analyses of health policies or systems. Recognising the expansive and interlinked nature of power relations, we focus this article on the different ways to research power as it manifests in health policies and systems. We also engage with literature on the social determinants of health insofar as these determinants impact health policies and systems.

This project emerged from the Social Science Approaches for Research and Engagement in Health Policy and Systems (SHAPES) thematic working group of Health Systems Global. SHAPES members (SMT, VS, MS and KS) with interest and expertise in power analyses reached out to the wider network and requested other interested researchers and practitioners to join the project. Recognising that expertise can take many forms, no criteria were placed on participation other than an interest in the topic and willingness to contribute to the paper’s development. The group was ultimately comprised of researchers from academic institutions, research organisations and multilateral agencies, in both the Global North (eight) and Global South (six) all of whom have experiential knowledge of assessing and negotiating power in health systems at various levels, and a number of whom have published in this area.

The process to develop this resource began in 2019. Members of the original group (SMT, VS, MS and KS) first prepared an outline of the paper via virtual and email discussions among group members. That outline was then divided into sections on theory, methodology and reflexivity, and section leads were appointed by a process of consensus. Group members volunteered to work on a section or sections based on experience and ability to input. Literature was sourced from database searches combined with expert guidance from group members. Working group leads organised the work of these sections and led drafting. Section drafts were reviewed by each group and then the full group, and two external researchers were invited to provide feedback on specific aspects of the paper. Online supplemental appendix 1 illustrates the iterative process by which the ideas were conceptualised, synthesised and agreed on at different stages of the paper drafting. All authors also read and commented on at least one version of the final paper. As a whole, the project was collaborative and worked from the logic of crowd-sourcing among a diverse set of authors engaged in HPSR.

Doing power analyses in health policies and systems research

This paper outlines key considerations and principles for power analyses in health policies and systems research throughout the research cycle. The paper is divided into three sections. The first section starts by discussing the identification of a research topic and presents three overarching empirical ‘sites’–or discrete areas of inquiry–for power-focused HPSR. The empirical sites offer a starting point for study design by providing researchers with ways to reflect on and refine their research question. This section also highlights researchers’ positionality and its influence on the whole research process. The second section provides an introduction to (and tabular summary of) theories useful for analysing power, demonstrating each theory’s relationship to one or more of the empirical sites. Finally, the third section of the paper introduces a selection of methodologies, considers their usefulness in the context of different types of power analyses and discusses how they, too, must be selected with consideration for the research question, the researcher’s positionality and alignment with theory. The ideas presented in this paper apply to all geographic contexts; however, we draw largely on HPSR literature from low-income and middle-income countries. This paper does not engage extensively with the use of specific data collection tools or methods (eg, interviews, observations and document review) associated with a given methodology, as other resources address these topics in detail.19 21 23 24

Identifying a topic

Power is imposed, negotiated and contested in diverse ways in the context of health policy formulation and implementation and health systems functioning. Research into power in the field of HPSR generally focuses on how the ‘expression’ of power enables or blocks health system change or policy implementation and what types of power are implicated in the process.16 20 From these two broad areas of focus, we discern three main sites of empirical work on power in the health policy and systems field, recognising that these three sites overlap significantly. These are: (1) actor relationships and networks; (2) sources of power and (3) societal flows and expressions of power.

In figure 1, we locate each of these empirical sites of power research around an adapted version of Walt and Gilson’s25 seminal Policy Triangle. This figure highlights that applied research on power cannot be conducted in isolation from the actors, context, content, structures and processes of the policy or system in focus. By demonstrating the link between actors, context, and structures and broad areas of power research, the three empirical sites are intended to provide a point of departure for the researcher to consider what is the issue or topic of interest. We expand on each of these empirical sites further below.

Request permission

Figure 1

Three empirical sites of power research in health policy and systems.

Empirical site 1: actor relationships and networks

The role and manifestations of power in actor relationships and networks comprise an important site of empirical research on power in HPSR. We list this site first because we understand health systems as social systems,6 fundamentally shaped by the values, intentions and relationships of the human and organisational actors within them. As illustrated in the central green triangle in figure 1, questions about power relating to actor relationships and networks include foundational enquiries about which individuals and organisations make and influence (health) policy and system decisions, how they relate to one another and why.

Empirical site 2: sources of power

As outlined in Sriram et al 16 and Moon22, a substantial body of theory is directed towards understanding how actors draw on power from particular sources.16 22 Sources of power thus represent a second important grouping of research on power in HPSR. Some methodologies, particularly those based in political science and economic theory, can describe and problematise key sources of power, such as material capital; technical expertise; political and bureaucratic position and influence; and forms of cultural capital and power gained from title, education and knowledge. Resultant research can provide analyses regarding which actors are impacting processes, from where they derive their power and how their actions impact policy and systems. This empirical site focuses our attention on ‘drivers of the drivers’, surfacing the institutions, organisations and attributes that provide a fountainhead of power in HPSR.

Empirical site 3: societal flows and expressions of power

A third empirical site of power research in health policy and systems relates to the societal flows and expressions of power. Research on the exercise of power shows how power is expressed, leveraged and experienced to impact health policy and systems, and ultimately, health inequities. Reflecting the intersection among context, actors and structures, research related to flows and expressions of power can generate insights regarding how formal or informal institutions shape health policy-making and service delivery, or on the impact of prevailing ideologies regarding health policy on service delivery.26 27 Researchers may focus on the ways that health policies and systems shape inequities28 or the ways that different groups have accepted, adapted and subverted health systems, such as the dictates of colonial medicine29 30 or neocolonial or internalised colonial forms of public health practice.31 32

Addressing power within the research process: positionality and reflexivity

In the process of issue identification and throughout the research process, it is critical to recognise the contested relationships of power that shape research itself. The nature of evidence in the fields of global health and health policy and systems research is contested,33–35 and the funding of evidence generation is politicised.18 36 Researchers—whether investigating power or other aspects of health and society—must be willing to consider their own role as actors in a contested process. Health research broadly tends to reward—in professional status, resourcing and publishing–positivist and utilitarian approaches over humanistic and relativistic and/or interpretive ones,36 Northern voices over Southern ones37 and biomedical knowledge over other forms of knowledge.38 Indeed the positionality of researchers is present in the many forms of power and privilege that can distance them from the issues they are analysing. Researchers’ professional positionality in the political economy of global health, as well as their individual lived experiences and attributes relating to race, caste, gender, class, ability and more, can significantly influence the choice of questions and (as discussed further) theories and methodologies used to enact analysis of those issues.

How should researchers engage with these challenges? There is no straightforward mechanism by which to operationalise critical reflexivity. Instead, building on the work of Sultana,39 Citrin,40 Mafuta et al,41 Abimbola37 Keikelame and Swartz42 and Pratt,43 we offer a set of questions in table 1 to guide reflection on power as it impacts a given research project. Researchers should consider: for whom they are designing and conducting data collection and analysis and writing up findings? And, how does this influence ‘bad habits’ that pervade global health research?44 However, discussions of power dynamics as they manifest in politics, social norms and otherwise is not a straightforward endeavour. Those who are brought in to collaborate in research processes, whether they be community members, health services representatives or funders, might be uncomfortable with an explicit focus on power relations. Shining a light on power asymmetries could create risks for collaborators or participants.

A conscious nurturing of critical reflexivity within all stages of a research process is a necessary component of ethical and rigorous praxis. However, analysing power while simultaneously maintaining awareness of the power relationships that structure the research endeavour itself is no easy feat. These questions and processes demand a more deliberative, bottom-up, time consuming approach to defining and answering research questions than is often enacted in HPSR. Prospective researchers of power should factor this time into their work. Since the political economy of global health and health policy and systems research can create incentives that undermine reflective, inclusive and transparent approaches to defining and answering research questions,18 these considerations should be taken into account from this initial step through the dissemination of findings and beyond.

Refining the research question with theory and methodology

The three empirical sites provide a launching pad for considering avenues for power inquiry for health policy and systems. In moving from a topic of interest to a more specific research question on power, and in conjunction with considerations of their own position and power, the researcher must consider their epistemological foundation (ie, what do we consider knowledge and how do we know it), the theories that provide a relevant analytical scaffolding, and concurrently, the methodologies that will enable appropriate collection, collation and analysis of data to that end.45

Pairing theory with methodology

Different theories are better suited to analysing power asymmetries characterising each of the three empirical sites. With regards to empirical site 1, theories with potential for exploring actor relationships and networks may include Weber’s three sources of authority47: street-level bureaucracy58; feminist standpoint theory,50 critical race theory48 and Bourdieu’s fields.59 Theories particularly relevant to examining the sources of power (empirical site 2) include Barnett and Duvall’s taxonomy of power,60 Bourdieu’s ‘fields’,59 Gramsci’s concept of cultural hegemony61 and feminist approaches.50 62 Theories relevant to expansive questions regarding how power is expressed and manifest in society at large (empirical site 3) may include Foucault's concept of knowledge/power,63 Veneklasen and Miller’s ‘expressions of power’64 and Lukes’ three faces of power.65

While theory helps to shape what we ask about power in HPSR, methodology shapes how we ask it and how we interpret the findings (figure 2). Below we provide an overview of 10 methodologies (broadly defined) that are of use in the context of the three empirical sites. The organisation of the methodologies under the empirical sites is merely illustrative. While some methodologies may be closely associated with a given empirical site (eg, social network analysis is associated with actor relationships and networks), many others are not. In conjunction with ongoing reflexive considerations of positionality, researchers choosing a methodology should consider their theoretical and epistemological position and the context of the research question, since the assumptions underlying the application of methodologies can be different (eg, the difference between an objectivist case study and an ethnography). Selection of methodologies should also consider for whom the research is being conducted, and whether the aim is to generate or further refine a theory or produce more immediately actionable findings. A summary table of these methodologies may be found in online supplemental appendix 3.

Request permission

Figure 2

Linking empirical sites, theory and methodologies for research on power in health policy and systems research.

To further make this point, table 3 provides illustrative examples of possible combinations of research question, theory and methodology. The inclusion in the table of two research questions at each of the different levels of health policy and systems function (micro, meso and macro) is intended to demonstrate (although incompletely) the breadth of potential inquiry as well as to showcase the specificity sometimes required to enable effective theoretical and methodological linkage. A key point made clear by the repeat listings of theories and methodologies across the various questions in table 3 is that there are many valid combinations of theories and methodologies.

Table 3

View inline|View popup

Illustrative combinations of theory and methodology paired with research questions on power in HPSR

Useful methodologies for empirical site 1: actor relationships and networks

Stakeholder analysis is an actor-oriented methodology useful for examining the power differentials of key policy and health system actors, ranging from frontline healthcare workers to national level policy makers.20 Stakeholder analysis is most commonly used prospectively, as a tool for researchers and practitioners to understand the feasibility of a given policy and to develop responses to likely challenges in implementing that policy.66 Stakeholder analysis can also be used retrospectively, as a stand-alone study or in combination with political economy and case study approaches. Stakeholder analysis is also commonly used to consider sources of power, described in further detail below.

Actor interface analysis focuses on understanding individual actors (rather than organisations), examines policy through the lens of power struggles between individuals and explores how this behaviour is embedded in actors’ lived experiences and values, called actor lifeworlds.67 68 When used to study health policy, actor interface analysis examines how interactions among different actors shape the implementation and outcomes of the policy. Where actors interact, collaboration, contestation or resistance can be identified and analysed. This methodology brings an actor-centric lens to the study of power in policy implementation as compared with other (more institutionally focused) methodologies and helps to examine how policy-related decisions and action are shaped by the actors themselves.67 69 70

Social network analysis is the quantitative study of relationship patterns among actors, with actors being broadly defined to potentially include people, groups or organisations.71 72 This methodology draws from sociology and mathematical foundations of graph theory to illuminate how the nature of actors and ties (eg, number, strength and type of tie, such as friendship, supervisory relationship and whether information, resources or beliefs were shared) enable expressions and tools of power (eg, money, pressure, influence and knowledge) to be concentrated, spread or blocked.73 In the field of HPSR, social network analysis can be used to analyse the health system structure as it functions, including through informal personal relationships, rather than as it is formally defined.74 This can inform policy makers about how ties among actors can influence the diffusion and implementation of health reforms and programmes; how social networks influence governance and financing structures; as well as informing the public about how policy makers may be using power to include or exclude certain actors.71 75 76

Useful methodologies for empirical site 2: sources of power

Case study design is a form of empirical inquiry characterised by an ‘intense focus on a single phenomenon within its real-life context’77 and is particularly useful in situations where boundaries between the phenomenon of interest and the context are blurred. In relation to power in HPSR, case study research has most commonly been used to produce exploratory and explanatory accounts focusing on different actors’ expressions of power (formal and informal, overt and covert) to answer ‘how?’ and ‘why?’ certain health policy or system features exist and to assess efforts to change power dynamics.20 78 By combining an interpretivist (seeking to understand individual and shared social meanings) and critical (questioning one’s own and others’ assumptions) analytical approach, researchers may use this methodology to consciously account for the ways in which broader social and political environments influence both macropower and micropower dynamics.79 80 Comparative case studies can be used for theory building or theory testing.

Political economy analysis is a methodology used to identify and describe structures such as government and the law; resources (labour, capital, trade and production) and how they are distributed and contested in different country and sector contexts, and the resulting implications for policy and indicators of well-being.81 Of relevance to HPSR, political economy can draw on both quantitative and qualitative methods to explore the nature of the political landscape through mapping the power and position of key actors. Political economy can also explore how the distribution of resources influence relationships and through this the feasibility and trajectory of policy reform over time.81 82 Reflecting their roots in the comparatively more positivist paradigms of political science and economics, these methodologies have been used for purposes of explanation and hypothesis testing in HSPR, including in the context of evaluations and policy design. Consistent with HPSR’s multidisciplinary orientation, political economy methodologies can nonetheless be developed and deployed in a way that accommodates—or even centres—interpretive goals.

Big data analytics examines high volume, biological, clinical, environmental and behavioural information collected from single individuals to large cohorts at one or several time points.83 Big data analytics can uncover patterns in health outcomes and health behaviours84; health policy (eg, resourcing and implementation fidelity)85; and health system function (eg, provider behaviours).86 87 When applied in conjunction with a power lens, big data analytics can reveal important and often masked trends or patterned experiences, prompting further explanatory work or evaluative action.88 For example, Yu et al 89 use big data analytics to explore the influence of private medical providers in promoting unnecessary medical interventions.89 Big data analytics may also help identify systemic issues such as discrimination, information asymmetry and patient-provider dynamics and their influence on care quality. Nonetheless, given its volume as well as its potential interest to profit seeking entities, big data presents unique challenges for ethics, boundaries and reflexivity. Researchers should carefully consider the potential misuses of the data, the extent to which the data accurately represents the factors of interest (construct validity) and which individuals and groups are overlooked in analyses that focus on the mean (or median).90

Useful methodologies for empirical site 3: societal flows and expressions of power

Discourse analysis entails close examination of the use of language in texts (such as laws, policies, strategy documents or news media articles) and oral communication (such as transcribed interviews, debates or speeches) to describe the ways in which communicative acts construct shared understandings of what is normal91 92 and what is possible, legitimate, or true.63 Discourse analysis should include the study of what is present in the text, as well as what is assumed or ignored, shedding light on often unacknowledged material asymmetries and social hierarchies that pervade health policy-making at all levels.93 94 In this way, discourse analysis can expose and problematise dominant paradigms in global and domestic health policy-making, such as the ways that standard epidemiological risk factors obscure structural inequities,95 the assumption that the private sector will act in the public interest96 or that a primary function of government reproductive health programmes is to decrease the fertility rate, rather than enable reproductive autonomy.97

Ethnographers seek to understand how humans in groups interact, behave and perceive, and how meaning and value are established. Ethnography can build rich and holistic understanding of people’s perspectives, practices and cultural context98 and focuses on depth over breadth, immersive observation in natural settings (eg, non-experimental conditions), exploratory (rather than hypothesis testing) research and describing the meaning and function of human action in context.99 100 While ethnography has its origins in colonial conceptions of ‘culture’ and colonial motivations to study them and has thus been frequently used to ‘study down’,101 ethnography has also been employed to research ‘up, down and sideways’.102 This includes work focusing on institutions and politics, political legitimacy, moral universes, tacit knowledge and discourses to provide insight into how power is constructed, solidified and wielded within and beyond health systems,103 104 the development and normalisation of certain forms of knowledge105 and the implicit or explicit privilege or denigration of individuals or marginalised groups accessing healthcare.106

Participatory action research (PAR) seeks to build new understandings of power while also changing power relations. PAR seeks to shift control over the construction of knowledge and truth from the historically privileged to the historically marginalised107 108 and increase participant understandings of injustice (conscientisation)109 in order to build solidarity110 and transform systems and institutions. PAR explores and recognises different sources of power (eg, social position, nationality and cultural knowledge) and applications of power (eg, via citizen-led collective action111). This research methodology typically entails the use of tools, such as community meetings, resource mapping, problem identification, visioning and diaries that draw out the priorities and perspectives of the communities participating, rather than reflecting a priori theory. It is apt for exploratory questions, as well as for bringing stakeholders together to cocreate solutions to health systems challenges.112

Historical research aims to generate or regenerate explanatory narratives relating to past events, places or people. Historical evidence includes visual, audio and text-based materials (archival material, communications, policy documents and project reports) and first-person accounts (oral histories). The study of history can illuminate broad power-related themes that continue to be relevant, such as the interface between individual liberty and domestic governmental health objectives113; medical experimentation, social control and scientific racism114 115; corporate profit making, governmental interference and population health116; and global health as a vehicle for state-craft, diplomacy, population control and Western-centric conceptions of charity.8 97 117–119 Historical studies also offer broader explanatory value as ‘cases’ for the development of theory related to power27 28 and as case studies for contemporary policy debates. Insofar as traditional historical approaches can privilege written work, it may omit the perspectives of historically oppressed groups. To combat this tendency, alternative methods such as participatory oral historical or community-based sourcing of visual, audio and text-based records not located in ‘official’ repositories open up alternative analytical possibilities.

Conclusion

More research on power in health policy and systems is needed. Linking empirical inquiry with theory and methodologies, with attention to positionality strengthens the rigour of such research and can help improve the depth and breadth of knowledge regarding root causes of inequities in health. This paper guides readers through the multiple stages involved, and a range of theories and methodologies that may be used, in developing a study focused on power in health policy and systems. It also seeks to push the HPSR field to challenge the political economy of research and destabilise hierarchies of knowledge through greater honesty about how power dynamics influence the research endeavour itself. Through the analysis of power in health policies and systems, we encourage researchers to expand the boundaries of how we may address inequities of health, to surface new insights, theories and approaches pertaining to power and, ultimately, to contribute to a more just world.