A framework for medical power in two case studies of health policymaking in India and Niger

ABSTRACT

Medical professionals influence health policymaking but the power they exercise is not well understood in low- and middle-income countries. We explore medical power in national health policymaking for child survival in Niger (late 1990s–2012) and emergency medicine specialisation in India (early 1990s–2015). Both case studies used document review, in-depth interviews and non-participant observation; combined analysis traced policy processes and established theoretical categories around power to build a conceptual framework of medical power in health policymaking. Medical doctors, mainly specialists, utilised their power to shape policy differently in each case. In Niger, a small, connected group of paediatricians pursued a policy of task-shifting after a powerful non-medical actor, the country’s president, shifted the debate by enacting broad health systems improvements. In India, a more fragmented group of specialists prioritised tertiary-level healthcare policies likely to benefit only a small subset of the population. Compared to high-income settings, medical power in these cases was channelled and expressed with greater variability in the profession’s ability to organise and influence policymaking. Taken together, both cases provide evidence that a concentration of medical power in health policymaking can result in the medicalisation of public health issues.

KEYWORDS:

• Power
• medicalisation
• health policy
• equity
• LMICs

Introduction

Health policymaking in high-income countries (HICs) and low- and middle-income countries (LMICs) has often been shaped by medical doctors, due to their specialised knowledge of health problems; high representation in ministries of health, membership in professional associations and other influential groups; and respected and trusted position in society (De Kadt, 1982; Peterson, 2001). Medical doctors in different specialities exercise power in different ways individually and collectively (Saks, 2015). Specialised medical academics are frequently influential in policy debates (Lewis, 2006), however other medical professionals and diverse actors including politicians, bureaucrats, technocrats, and international and civil society stakeholders can also exercise influence by controlling bureaucratic levers, applying political pressure, or providing funding, among other means (Quadagno, 2004; Sheikh & Porter, 2010).

In the policymaking sphere, power can be used to define areas of participation and action, including closed, invited and claimed spaces, or to shape policy dialogue via dispositional and structural choices around consultation, decisionmaking, and resource allocation (Arts & Jan Van, 2004; University of Sussex Institute of Development Studies, 2011). The power of medical professionals has been discussed in terms of the profession’s ability to secure exclusionary control and safeguard its interests, often with protections from the state (Freidson, 1970; Johnson, 1972), as well as shifts and variations in medical power within and across national health policymaking processes (Peterson, 2001; Quadagno, 2004). While these dynamics have been explored at length in HICs, theoretical and empirical explorations of power in health policymaking in LMICs remain rare, especially around issues of health equity (Embrett & Randall, 2014). Yet attention to power is essential to understanding policymaking processes and outcomes (Erasmus & Gilson, 2008; Walt et al., 2008).

Evidence from HICs suggests that more cohesive networks or organised structures can enhance the power of medical actors (Quadagno, 2004; Rabinowitz & Laugesen, 2010). Rabinowitz and Laugesen (2010) highlight the different policy objectives of groups comprising ‘organised medicine’ in the United States, finding that wider collectives (e.g. the American Medical Association) often pursue economic self-interest, while specialist groups additionally focus on issues of population health. However, few theoretical or empirically-driven analyses exist on these topics in LMICs (Nigenda & Solorzano, 1997; Sheikh & Porter, 2010), despite differing political systems, pressure points and positions of power faced by medical professionals (Jeffery, 1977). Furthermore, medical power in LMICs is often shaped by the historical dominance of biomedicine under colonialism and subsequent post-colonial aspirations towards ‘modernisation’ (Maru, 1985; Zachariah, 2014). As such, practitioners of biomedicine (also called allopathic or ‘Western’ medicine) often emerge as dominant in state health systems compared to traditional, complementary or alternative systems such as Ayurvedic medicine, traditional Chinese medicine, or homeopathy (Lakshmi et al., 2015).

In this paper, we examine medical power in health policymaking at the national level in LMICs, focusing specifically on biomedicine. We define medical power in national health policymaking as authority, influence or leverage derived from being a medical doctor.¹ There are hierarchies, factions and differing interests among medical professionals, and individual doctors or medical coalitions can apply medical power to any end. We examine the sources and applications of medical power in health policymaking, drawing on case studies in Niger and India. We present our methods and conceptual framework, and then describe how medical power was used in health policymaking, comparing and contrasting the two cases. We conclude by drawing lessons on the framework’s suitability and discuss how medical power may be used to promote equity-oriented policies.

Case descriptions

Niger

Despite early experiments with village health teams in the 1960s, basic health services remained mainly inaccessible in rural areas in Niger until President Mamadou Tandja launched a campaign to build 2,000 ‘health huts’ after his election in 1999 (Fournier & Djermakoye, 1975; Körling, 2011). Nonetheless the health huts remained under-staffed and under-utilised. In 2005, Nigerien policymakers attended a Dakar meeting on integrated community case management of childhood illness (iCCM), a policy for task shifting to treat childhood illnesses using community health workers (CHWs). Promoted by global child health stakeholders, iCCM became the subject of a follow-up visit to Niger by partners, spurring officials at the Ministry of Public Health (MOPH) to organise a field trial of iCCM in partnership with the World Health Organization (WHO), UNICEF, and the U.S. Agency for International Development (USAID). Following a positive mid-term evaluation in 2007, iCCM was scaled up with funding from UNICEF and the Canadian international development agency starting in 2008. By 2012, iCCM was implemented in all health districts, with subsequent assessments finding generally good quality of care (Bensaid & Gali, 2009; Seidou, 2008). ICCM and related policies were found to have contributed to significant reductions in child mortality rates in Niger in recent years (Amouzou, Habi, & Bensaïd, 2012).

India

The development of emergency medicine (EM) as a medical specialty in India began in the early 1990s, driven by Indian doctors returning from training or work abroad promoting it as a solution to weak systems of emergency care (Sriram, 2017). In the late 1990s, momentum towards EM specialisation increased due to the formation of EM professional associations, active promotion by the private sector, and transnational partnerships. However, stakeholders spent nine years advocating for recognition with the Medical Council of India (MCI), and the interim period saw intense fragmentation amongst the EM network, resulting in uncoordinated advocacy and policy strategies. MCI ultimately recognised the specialty in 2009, opening the door for residency programmes in medical colleges; however MCI’s limited consultation with the EM network resulted in a lack of clarity on requirements and standards. As a result, recognition of 11 EM courses was halted for a period in 2015–2016. While the development of EM increased in number of physicians trained in EM, it also highlights serious challenges in the systematic and planned development of new medical specialties in India.

Methods

Case study methodology is useful for providing holistic, nuanced descriptions of complex social phenomena operating in real-life contexts, including policymaking processes (Walt et al., 2008; Yin, 2014). The two case studies described here were designed and performed separately using case study methodology and used because they differ in relevant, critical ways (governance structures for medical professionals, public-private mix, size and cohesion of the health work force), allowing for an initial exploration of the topic of medical power (Table 1). We draw upon contrast-oriented design in seeking to distinguish between contextual particularities not impinging on questions of theory and unique aspects of cases bringing out features of the theory in development (Skocpol & Somers, 1980).

Table 1. Characteristics of the medical profession in Niger and India.

	Niger	India
Medical work force	0.3 doctors and 3 total health workers per 10,000 population (Africa Health Workforce Observatory, 2010) Doctors are highly concentrated in cities, especially the capital Niamey Rural regions are served mainly by paid CHWs (Système national d'information sanitaire (Niger), 2014)	6.49 doctors (Global Health Observatory, 2009) and 20 total health workers per 10,000 population (Rao, Bhatnagar, & Berman, 2012) 80% practice in urban areas (Rao et al., 2012)
Governance of medical profession	Regulated by the national MOPH Doctors represented by professional associations including by specialty Union of Community Health Workers created to protect their interests in the mid-2000s	Regulated by a statutory professional council, the Medical Council of India (Rao, Rao, Kumar, Chatterjee, & Sundararaman, 2011) Supplemented by an array of national, state and local associations and unions that do not necessarily work in unison (Banerjee, 2014)
Public-private mix	Upwards of 95% of doctors practice in the public sector (Système national d'information sanitaire (Niger), 2014)	80% of allopathic doctors work in the private sector (Rao et al., 2011)
Medical training	Main public medical school in Niamey, the capital Members of specialist professional associations often train abroad	∼460 medical colleges in both public and private sectors Postgraduate training also offered in private, non-academic medical institutions
Systems of medicine	In addition to biomedicine, traditional African medicine(s) are widely practiced Traditional practitioners are more rarely involved in policymaking	Multiple systems of medicine, with biomedicine dominant in the post-colonial era Traditional forms of medicine emerging more recently as politically ascendant (Lakshmi et al., 2015)

Qualitative methods are useful to trace processes occurring over time, including intended and unintended consequences and potentially divergent views on the issues in question (Maxwell, 2005). The qualitative methodology used in each case study is summarised in Table 2. Both studies used three data collection methods: 1) document review, 2) semi-structured in-depth interviews with participants in the policy process, and 3) non-participant direct observation. The latter method, direct observation, is less often used in health policy studies despite its usefulness in enhancing data quality and credibility, especially as it allows for observing and documenting direct expressions of power and how the ‘rules of the game’ operate or are challenged (Erasmus & Gilson, 2008). Preliminary data analysis varied somewhat between the two case studies. For Niger, process tracing was used to create a descriptive timeline of the policy process, with subsequent exploration of theoretical categories via thematic analysis of coded data. For India, a modified framework approach was used to analyse data deductively and inductively to elaborate themes around power and contribute to theory development (Gale, Heath, Cameron, Rashid, & Redwood, 2013).

Table 2. Overview of methodology for the two case studies.

	Niger	India
Focus of study	Integrated community case management of childhood illness	Hospital-based emergency care & recognition of EM specialty
Data collection	Niamey, Niger (March–August 2012) Document review of policies, scientific articles, ‘grey’ literature, regulations, etc. (N = 103 documents) Semi-structured in-depth interviews (N = 32 policy actors) Direct observation of policymaking forums (N = 3 forums)	India (March 2015–January 2016) Document review of policies, legislation, meeting minutes, scientific articles, news reports (N = 248 documents) Semi-structured in-depth interviews (N = 87 policy actors) Direct observation of meetings and conferences (N = 6 forums)
Analysis methods	Systematic extraction of data from documents Thematic coding of transcripts using pre-defined questions and categories with further memoing and discussion to develop analysis Process tracing and triangulation to synthesise findings Iterative re-analysis using theoretical concepts on power	Selecting and developing analytic frameworks First stage – coding of 46 transcripts; developing themes from inductive and deductive analysis, triangulating between data sources Second stage – additional coding; utilised matrix to confirm and disconfirm themes; analysis of transcripts

Secondary analysis occurred when the conceptual framework emerging from the India case study (Sriram, 2017) was modified and revised in light of the results on power from Niger and applied to previous findings and raw data in both case studies. Building on Sriram (2017), we use a two-step process to examine (1) sources of power and (2) applications of power, allowing for extrapolation of theory (Erasmus & Gilson, 2008; Sriram, 2017). In our conceptual framework, medical power is one source of power in the health policy sphere, within a larger ecosystem of power, and may in fact be derived from other sources of power, such as knowledge claims, bureaucratic power and network power. We identify seven sources of power in health policymaking (Table 3), which may be applied to favour or obstruct certain ends – a dual agency Bachrach and Baratz (1970) call the ‘two faces’ of power (Bachrach & Baratz, 1970). In these instances power may be expressed in more or less visible ways (Foucault, 1994; University of Sussex Institute of Development Studies, 2011). We aim to identify the avatars of medical power in policymaking processes, understand its uses and applications, and compare and contrast findings between the two cases (Walt, Pavignani, Gilson, & Buse, 1999).

Table 3. Types of power in health policymaking and their intersection with medical power.

	Working definition	Intersection with medical power
Medical power	Authority, influence or leverage derived from recognised membership in general medicine or the specialties	–
Political power	Ability to affect policy decisionmaking based on position in government or a political party	Medical doctors usually occupy high appointed positions in ministries of health, and sometimes other important positions in government; Centralisation can bestow greater power to medical doctors at national level; Historical link between biomedicine and (colonial) authorities in many LMICs.
Knowledge claims	Formalised knowledge of technical content as it relates to the relevant policy	Medicine is one of the most highly specialised and esoteric bodies of knowledge; Specialists have even rarer bodies of knowledge; Non-health professionals have limited knowledge claims on clinically oriented policies; Medical professionals from high-income settings often have greater access to codified medical knowledge, for example, the published literature.
Bureaucratic power	Control or influence over the administrative machinery through which policies are designed and implemented	Ministry of Health personnel with broad responsibility for national health policy include many medical doctors; Self-regulatory role accorded to medical profession results in regulatory capture.
Financial power	Control or influence over resource flows from international, national or sub-national budgets, loans or grants, or private sources	Medical sphere linked to economically powerful actors such as pharmaceutical and insurance companies, technical device manufacturers, and for-profit hospitals; Medical officials at international agencies often have the power to allocate funding for in-country activities to support decisionmaking, policy implementation, research, etc.
Network/corporate power	Influence, authority or access to resources derived from formal and/or informal organisation of policy actors	Medical networks are often highly cohesive and comprised of élites; Medical doctors are often connected via shared educational experiences; Elite public sector medical colleges are closely involved in developing health policy; Medical professions associations tend to be highly organised and influential in health policymaking.
Social-structural power	Control or influence over policy decisions and outcomes based on demographic/social characteristics	Medical professionals benefit from the high social status traditionally accorded to healers by society; In many contexts, medical doctors are more likely to be male, and come from advantaged family and ethnic backgrounds.

Both case studies received ethics approval from the relevant bodies.

Results

Networks of medical professionals dominate discussions about policy content

In both India and Niger, policy discussions and decisionmaking tended to be dominated by medical doctors, often specialists in the relevant clinical domain, organised into networks that magnified their influence. In Niger, a small group of paediatricians and to a lesser extent neonatologists at MOPH or at major partner agencies such USAID, the WHO and UNICEF were the most vocal and influential parties in discussions to adopt iCCM. Nigerien paediatricians, including those working for external agencies, were small in number (one respondent said approvingly ‘now they are more than 40’) and often shared educational experiences, creating informal links reinforced by membership in professional associations:

We are all from the same generation, we all went to school together; we have known each other since childhood. (Child health officer, international agency)

These ties created familiarity, confidence and a basis for consultation between medical policy actors sitting at government and international agencies: as this respondent said, ‘relationships matter a lot’.

Some medical doctors operating outside the central network of paediatricians were involved in policy discussions in a more peripheral way. And while paediatricians consulted with other stakeholders, including frontline health workers and representatives of the national malaria programme and the National Communication Service for Health, these actors were frequently side-lined during policy workshops where consequential decisions were made. Yet while medical doctors (specifically paediatricians) spoke most frequently in these settings, many appeared less familiar with daily implementation realities than frontline actors in rural communities. As one respondent recounted:

We were at a [policy workshop], and representatives of community-based organisations were invited, but did you hear them speak? [No.] It’s a problem! Even though everything that we were discussing, it was for them. (Project officer, international NGO)

Similarly in India, the development of the EM speciality was promoted by EM specialists from metropolitan areas and HICs, including members of the Indian diaspora, who worked largely in hospitals rather than in policymaking bodies. Organised through professional associations, these specialists actively advocated for MCI, the main regulatory body for medical education, to formally recognise EM. Indeed MCI exercised near exclusive decisionmaking power despite the Ministry of Health and Family Welfare (MoHFW) nominally bearing shared responsibility for medical education.

On very highly technical things [the MoHFW] won’t argue with [MCI] too much. (Former central government official)

Previously, MoHFW participated in discussions on specialisation when there was a perceived impact on population health, such as when advocating for expansion of the family medicine speciality. However these efforts were often met with inertia from India’s complex system of postgraduate medical education. MoHFW was not directly involved in decisions around recognising EM.

Members of the EM network were connected through shared medical college backgrounds, experiences at elite medical institutions, and geographic proximity. Networks also facilitated EM stakeholders’ access to doctors in policymaking roles at MCI and policymaking exercises such as curriculum development committees. Stakeholders from the All India Institute of Medical Sciences, a prestigious public sector medical college in New Delhi, were involved in curriculum development, and frequently collaborated with MCI.

Also influential were private sector actors, with access to resources and international partnerships, whom some respondents said were responsible for pushing EM:

[The for-profit private hospital network] had taken this initiative … any new change in healthcare in India was always driven by private sector. (Private sector stakeholder)

Ideological differences on the role of private-sector and international actors fragmented actors into multiple EM sub-networks working independently and sometimes competitively with one another. Still, formal and informal policymaking spaces were occupied by a limited set of medical specialists, which did not include frontline emergency care providers such as rural medical officers or Casualty Medical Officers.

Broader health systems and infrastructural changes require the intervention of non-medical actors

In Niger and India, medical policy actors conceded the need for wider health systems improvements and upstream interventions to improve health outcomes, but tended to remain focused on medically curative treatment. Broader health systems improvements were pursued at the behest of non-medical policy actors. In Niger, medical doctors working on child health policy at government and non-government institutions frequently demonstrated awareness of the limitations of Niger’s health system, for example, mentioning the insufficient number of health workers and limited reach of the formal health system. However the major health systems improvements that served as the foundation for iCCM task-shifting were accomplished as a political initiative, which was initially opposed by medical stakeholders at MOPH and elsewhere. This began in 2001, when then-President Mamadou Tandja launched a campaign to build a large network of rural ‘health huts’ using a financing stream from the Heavily Indebted Poor Countries initiative that he directly controlled:

The head of state at the time wanted to increase the health coverage rate and health huts were the easiest way to do it. (Child heath officer, partner agency)

Tandja encountered strong resistance from the Nigerien medical establishment housed at MOPH, who considered that he ought rather to extend the formal health system, instead of creating a new layer of health centres that would provide ‘inferior medicine’:

The entire health staff of the MOPH thought that the advent of health huts would be a regression of the health system. (Official, CHW trade union)

Because of Tandja’s political supremacy (he would be soundly re-elected in 2004) and direct control over funding, some 2,000 health huts would be built by 2007 (Dalglish, Surkan, Diarra, Harouna, & Bennett, 2015). However the health huts remained under-utilised (many were empty) because of the incongruence between Tandja’s vision and MOPH staffing structure and policies, which arguably favoured higher-level medical professionals who resisted working in remote, poorly equipped structures. This situation was resolved when MOPH was able to ‘re-appropriate’ the health huts via the advent of iCCM, a policy bearing the imprimatur of medically prestigious international agencies (notably WHO and UNICEF) beginning around 2005–2006.

In India, EM at tertiary level was prioritised over other solutions for improved facility-based emergency care, such as strengthening health systems at lower levels. EM stakeholders in India said the broader health system challenges were too steep, and it was better to address issues within their radius, such as generating specialists through postgraduate medical education. Unlike in Niger, there was no political process to push forward a larger infrastructural vision. Instead EM stakeholders hoped for a trickle-down impact:

Even then we didn’t really talk about [health system issues] because we knew we couldn’t do anything. One hospital in Pune, one hospital in Ahmadabad can’t change anything. (International emergency physician)

You cannot individually train all the carers. What you can do is you can train the trainers. So these trainers in turn will train others, and then the training will percolate down to the grassroots level. (Private sector stakeholder)

EM stakeholders said that when postgraduate courses became more common, change would spread throughout the system, a view reinforced by international stakeholders accustomed to residency-based training in HICs. However, a few respondents said this approach would result in an urban, private-sector orientation, limiting the ability to serve poor and rural populations:

There is India and there is Bharat [word for ‘India’ in Indian languages]. The fallacy of emergency medicine is that no one is looking at Bharat. (Private sector stakeholder)

Some respondents said integrating EM into the undergraduate medical curriculum or working with practitioners of traditional medicine could promote wider access to service delivery, but these suggestions did not gain policy traction. Furthermore, other emergency care policies, such as the expansion of pre-hospital care services, suffered from a lack of coordination and prioritisation across policy efforts.

Medical power and trade-offs between occupational control and social equity concerns

In Niger, iCCM necessitated transferring certain clinical functions to lower level health workers to increase access for under-served populations, particularly in rural areas. Initial resistance of policy actors with medical power, including paediatricians at MOPH and professors at prominent teaching hospitals, was often expressed as concern around quality of care. Devolving responsibility to lower-skilled providers would translate into ‘second-rate medicine’ for the poor, some said, especially as this regarded prescribing antibiotics for pneumonia and providing care for newborns:

Imagine, presently, when you speak about newborns, what will be the answer of paediatricians? Even general practitioners are not skilled enough to take care of newborns – how then would you expect CHWs to administer care? It would be criminal! (Senior manager, international organisation)

However other holders of medical power, including paediatricians and specialists working at international agencies such as WHO, UNICEF and USAID (of Nigerien or other nationality), were able to present ethical and practical arguments in favour of task shifting:

The main argument for all these great professors was to say ‘you are working in a facility that is well-equipped with equipment, staff and materials, but presently where do children go to seek care? Not necessarily in town. (International consultant)

Such arguments mitigated resistance to task-shifting among Nigerien doctors, as the imperative to provide care for the country’s large, under-served rural population outweighed the fears over quality of care. Medical doctors may also have felt less threatened by the reform, as they were personally unlikely to occupy posts in rural health huts. On the other hand, resistance from lower-level health cadres arose against the increasing prominence of CHWs:

Nurses and midwifes thought that the training of CHWs would further ruin the health system. (Official, CHW trade union)

In India, occupational interests around EM coincided with a massive expansion of the private sector. The private sector actively took on EM, helping found a professional association, initiating training courses, and sponsoring conferences. Beyond allowing for more organised, systematic emergency care, respondents said EM was seen by corporate hospitals as a way to create an ‘image of [a] hospital which will be a good destination in emergencies’ and obtain a competitive advantage (Healthcare Executive, 2014). Said one respondent:

There is a corporatisation of medicine, and a lot of corporate hospitals that are coming up want specialists to satisfy the needs, real or artificial. (Public sector stakeholder)

Respondents from both public and private sectors suggested the latter’s motives in advocating for EM were largely financial and reputational, to the detriment of an adequate investment in building health systems, though other respondents countered this perspective by noting the limits of private actors’ ability to influence sector-wide change.

EM stakeholders have more recently worked to extend access to lower levels of the health system via training and empowering medical professionals in more peripheral positions. Stakeholders have started organising pre-hospital care programmes, frontline emergency care strengthening, short course training programmes, and disaster relief programmes; some EM stakeholders recognised that the ‘trickle down’ impact of specialists on the health system would take time, and noted the need for ‘Indianised’ systems of emergency care that could reach lower levels of the health system. For example, MoHFW, in consultation with EM stakeholders, piloted a programme to train medical officers and other frontline health workers in basic emergency care procedures. However these approaches have remained sporadic, and are not seen as key pillars of professional associations’ agendas.

Discussion

In the two contrasting contexts analysed here, medical doctors, mainly specialists working in government offices, regulatory structures, professional societies, medical colleges, and international agencies, dominated formal and informal policymaking discussions. Medical power was concentrated in professional networks whose members set the policy agenda and maintained control over policy content and points of debate, often to the exclusion of other voices in the policy dialogue. In Niger, the involvement of a powerful non-medical actor, the country’s president, shifted the terms of the debate by driving broad health systems improvements and creating an infrastructural incentive to engage in task-shifting to benefit under-served populations. However in India, the policy development process remained cloistered within a small set of specialist EM stakeholders, and issues relating to population-level access to care were largely delinked from this agenda.

Medical power, as understood from sociological theory in high-income contexts, has largely focused on utilising professional status to protect the interests of medical doctors, ensure occupational control (Saks, 2016), and garner ‘social privilege, economic power and political influence’ (Quadagno, 2004), despite some efforts within the profession to emphasise public interest (Starr, 1982; Stevens, 2001). Our analysis shows that despite similar motivations (seen in the exclusionary control of specialists in India, or initial resistance to the devolution of curative practices to lower-level providers in Niger), there appear to be key differences in how medical power is organised, channelled and expressed in LMICs compared to HICs, particularly as this concerns the profession’s ability to organise and influence policymaking. In high-income settings, organised medicine evolved to be ‘delicately’ held together by professional associations and other self-regulating structures (Bucher & Strauss, 1961), a relative cohesion that allows organised medicine to more effectively shape and influence policy. Our case studies show issues of organisation and cohesion taking on different dimensions. In India, organised medicine is highly pluralistic, and the interests of the profession are variously represented by the MCI, national and state professional associations, and unions. While EM stakeholders were successful in gaining recognition, their fragmentation might have blunted their influence over a longer-term policy agenda. Conversely in Niger, where the overall number of medical professionals remains quite small and geographically concentrated, cohesion within the stakeholder network increased the power of the profession in policymaking. Nigerien medical specialists wielded significant power in deciding the contents of health policies, albeit within broader health systems parameters influenced by political actors.

By focusing on medical power in the health policymaking process, we attempt to engage with other frameworks for studying power in public policy. These range from theoretical analyses in the Foucauldian or Bordieusian veins to more practical applications such as the Swedish development agency’s ‘Power Analysis’ tools (Pettit, 2013), the ‘Drivers of Change’ framework based on political economy work at the United Kingdom development agency (Nash, Hudson, & Luttrell, 2006), and the ‘Powercube’ developed by scholars at the University of Sussex (University of Sussex Institute of Development Studies, 2011). Our framework attempts to synthesise work on the sources of power likely to drive health policymaking at national level in LMICs. As such our findings underscore the intersection of medical power with other sources of power (Table 3) and how it is ‘parlayed’ into other forms (Quadagno, 2004), including bureaucratic and financial power; network/corporate power; and technical expertise. We found that medical power was not uniformly held or applied, as perceptions of technical expertise can vary by institutional affiliation or country of origin, leading to power asymmetries within stakeholder groups (Mosse, 2005; Walt et al., 1999). For example, in both Niger and India, we observed a higher regard for technical expertise originating from Western sources such as international agencies, returned members of the diaspora, or experts trained in Western institutions. Hierarchy, seniority and regionalism can also affect policymaking dynamics, creating communications and perceptions gaps among medical doctors, including on how evidence is valued (Behague, Tawiah, Rosato, Some, & Morrison, 2009).

Our case studies contribute further evidence that one consequence of the concentration of medical power in health policymaking is a medicalisation of public health issues. In neither case were underlying causes of public health problems seriously addressed, such as improved sanitation to reduce childhood diarrhoea in Niger, or improved road safety to prevent catastrophic injury in India. However medical actors cannot always effectively tackle non-medical or even health systems-wide issues, as Indian respondents clearly articulated. In Niger, broad reform of the health system only occurred when a powerful non-medical actor (the country’s president) exercised non-medical (political and financial) levers of power outside the formal health policy development process (Dalglish et al., 2015). A comparison of the outcomes in these two cases suggests that medical power may not be oriented towards mitigating the lack of ‘staff, stuff, space, and systems’ that plague health services in many LMICs (Farmer, Kleinman, Kim, & Basilico, 2013). Yet even beyond medical doctors, the broader public health community has been accused of following a ‘disease du jour mentality’ that ‘creates silos, disproportionately funds biomedical solutions, and … seldom gain[s] the kind of ongoing political attention and economic resources [needed] to improve the public’s health’ (Gostin & Powers, 2006).

Further thinking is needed on how to encourage public health thinking to promote sustainable, systems-wide and multi-sectoral policy solutions to population health problems. Medical professionals remain central to health policy networks even where there have been long-standing efforts to reduce their dominance (Lewis, 2006), though their power can be restrained by political actors or competing claims from other interest groups (Jeffery, 1977; Nigenda & Solorzano, 1997). Findings from our case studies support the need to acknowledge and encompass other sources of knowledge and experience beyond those of medical doctors, including from lower-level frontline providers, public health experts and community representatives (Sood & Ananthakrishnan, 2012); value diverse types of knowledge and skills, such as negotiation and consensus-building, practical experience, and logical or ethical reasoning (Dalglish, Rodríguez, Harouna, & Surkan, 2017); and include community members in research to design health policy (George, Mehra, Scott, & Sriram, 2015). Formally including these voices in policymaking processes is necessary, however given the entrenched nature of current structures and interests, greater understanding of power dynamics in health policymaking will be required to design effective strategies for achieving this goal.

Study limitations

One major limitation of this analysis is that the two case studies were not specifically commissioned to contribute to theory-building on the role of power in health policymaking. However we believe our data were rich enough to shed light on this topic, and have been careful not to draw conclusions beyond what such a comparison can support. The study of power in policymaking presents unique challenges because while its influence is pervasive, power is rarely explicitly discussed; furthermore secrecy and intentional opaqueness are key facets of power in bureaucracies (Olivier de Sardan, 2004; Weber, 1946). As such the ability to draw meaningful conclusions relies on carefully collected data and the interpretative skill of the analyst. We attempted to improve validity using triangulation, peer debriefing and member checking during preliminary data analysis, and triangulation and peer discussion during secondary analysis. Another potential limitation is that authors of this paper hail from similar disciplinary backgrounds, possibly occasioning analytical blind spots, particularly as none of us have medical or clinical backgrounds. However we were careful to self-reflect on positionality, especially as this relates to accessing elite (medical) respondents, and analyse how this could have influenced our results (Mikecz, 2012; Sriram, 2016; Walt et al., 2008).

Conclusion

Frameworks for studying power in health policy are evolving. We present an in-depth look at one type of power, medical power, which we found to be pervasive in two case studies of policymaking in Niger and India, where medical actors strongly influenced policy choices in different ways. In India, medical actors controlled the policy agenda around specialisation, resulting in the adoption of a relatively narrow, specialised policy; whereas the involvement of powerful political actors in Niger brought about broader-systems wide improvements later expanded upon by medical actors. Further research is needed to strengthen theoretical categories around power in health policymaking and test them against empirical data, especially to understand how medical power can be channelled towards achieving broader health gains and improving health outcomes of less powerful population groups.

Acknowledgements

The authors sincerely thank Drs Kabir Sheikh and Sara Bennett for reviewing an early version of our manuscript and providing helpful comments and guidance. Additional thanks are due to Reviewer #1, whose engagement with our manuscript improved its final form, and to editors at GPH for their thoroughness and attention to detail. The contents are solely the responsibility of the authors and do not necessarily represent the official views of any funders.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Sarah L. Dalglish http://orcid.org/0000-0002-7218-5193

Additional information

Funding

The Niger case study was funded by UNICEF (#43114640) and the USAID TRAction project (FY11-G06-6990). Both UNICEF and USAID staff advised the study team, but did not substantively affect the study design, instruments or interpretation of data. Fieldwork for the India case study was supported by the American Institute of Indian Studies and the Department of International Health at the Johns Hopkins Bloomberg School of Public Health. V. S. is currently supported by the Agency for Healthcare Research and Quality [grant number T32 HS000087] (PI: Jane Holl, MD, MPH).

Notes

1. The role of other medical providers in nursing, para-medical fields, or lower level cadres like community health workers, is a related point that should be more fully considered elsewhere.