Question 1. What are the ‘common sense’ assumptions underpinning agendas in SRH interventions using PPPs?

A reading of the existing literature on gender analysis of healthcare services suggests that we need to pay attention to several dimensions of programme and service design: the first of these is to asking what is the ‘common sense’ that propels their agenda and direction. Sylvia Chant, for example, has highlighted the ‘feminisation of obligation and responsibility’ within the context of neoliberal economic restructuring (Reference ChantChant, 2008).

In recent years the agenda and direction of programme design for SRH interventions has been driven forward in part by a ‘common sense’ which expresses a widespread concern, particularly on the part of global institutions and international donors, around the high levels of maternal mortality and fertility and a desire to promote gender equality and women's ‘empowerment’ (Reference Filippi, Ronsmans, Campbell, Graham, Mills, Borghi and OrsinFilippi et al., 2006). Yet, the allocation and design of aid within the health sector is a highly political process and is frequently shaped by the ideological stance of the global institutions and the aid donors rather than the health priorities of a particular country or population group (Reference Esser and BenchEsser & Bench, 2011; Reference Storeng and BéhagueStoreng & Béhague, 2014). Critical research has shown how particular health issues, such as maternal mortality, come to dominate the global health agenda at the expense of others, such as cervical cancer (Reference Parkhurst and VulimiriParkhurst & Vulimiri, 2013).

The incorporation of concerns around maternal mortality into the MDGs served to raise the profile of that neglected issue and generated considerable donor funding into seeking rapidly attainable ‘solutions’ to address the high rates that continue to occur across the world (Reference Storeng and BéhagueStoreng & Béhague, 2014). But in an environment in which economic development is prioritised over other social gains, advocates for maternal health have tried to develop an argument that associates it with wider development objectives, and in the process maternal health has come to be portrayed as a means to development rather than an end in itself (Reference GideonGideon, 2014; Reference Mohindra and NikiémaMohindra & Nikiéma, 2010). This offers a clear example of the ways in which the neoliberalization of feminism has played out in the health sector. The push to extend women's access to healthcare services is primarily driven by a rationale of ‘smart economics’. Although women may gain better access to services, at the same time this trend potentially reinforces women's responsibilities and obligations, rather than their empowerment as independent social citizens with rights, and therefore diverts attention away from the wider goals of social justice (Reference Gideon and PorterGideon & Porter, 2016; Reference Storeng and BéhagueStoreng & Béhague, 2014; Reference Yamin and BoulangerYamin & Boulanger, 2013). As Prugl argues,

In the case of SRH services, global and national pressure to meet the 5th Millennium Development Goal (MDG 5) targets of reducing maternal mortality and ensuring universal access to reproductive health, and concurrent increases in funding from donors such as the UK Department for International Development (DFID), the World Bank and the Gates Foundation, provided opportunities for the expansion of PPP models in that sector that seemed to offer rapid and measurable results (Reference Gideon and PorterGideon & Porter, 2016). In the post-MDG context PPPs have gone on to be promoted as an important development financing mechanism in support of the Sustainable Development Goals (see SDG 17.3 and the Addis Ababa Declaration of the Third United Nations Financing for Development Summit, July 2015).

Question 2: On what criteria do the SRH-PPP programmes come to be framed and judged?

The emphasis on achieving maximum gains in SRH in a short period to accelerate processes towards global development goals also created a narrowing of intervention policies to focus on ‘what works’ (Reference Campbell and GrahamCampbell & Graham, 2006). This can be seen as part of a wider process of ‘scientization’ in global health (Reference Drori, Meyer, Ramirez and SchoferDrori, Meyer, Ramirez, & Schofer, 2003) in which the techniques of evidence based medicine (EBM) were used to determine what is or is not effective using randomised control trials as the gold standard (Reference Adams, Biehl and PetrynaAdams, 2013). In the field of SRH PPPs, this emphasis on the easily quantifiable and comparable coupled with the rising influence of health economics also led to a focus on specific packages of care that were deemed cost effective such as use of modern contraception or birth-care in a healthcare facility (Reference Loevinsohn and HardingLoevinsohn & Harding, 2005; Reference Murray, Hunter, Bisht, Ensor and BickMurray, Hunter, Bisht, Ensor, & Bick, 2014). Such defined packages in turn facilitated the emergence of a series of PPP arrangements in which private sector provision of these packages was purchased by the public sector with donor assistance. Maternal healthcare services, for example, were considered well suited for such partnerships because they operated in a well-defined time period and there was a systematic review-supported evidence-base for the package elements with predictable typical costs that could be reimbursed to private providers once a woman has given birth.

Vouchers and franchising approaches thus gained popularity in health and development policy circles in the early 2000s (Reference Koehlmoos, Gazi, Hossain and ZamanKoehlmoos, Gazi, Hossain, & Zaman, 2009; Reference Murray, Hunter, Bisht, Ensor and BickMurray et al., 2014; Reference RavindranRavindran, 2011). Social franchising schemes in the health sector are networks of private-for-profit health practitioners linked through contracts to provide a socially beneficial service under a common brand, while health voucher schemes, copied from the education sector, are designed to stimulate demand for a specific health service through the distribution of pre-paid vouchers for services. Voucher schemes, which we return to in the case study, are seen as a means of encouraging use of services like family planning, treatment of infectious diseases, immunizations, and maternal and child health services through subsidising costs that may otherwise deter the user (Reference Gupta, Joe and RudraGupta, Joe, & Rudra, 2010, p. 4).

Advocates tend to frame success in terms of increased effectiveness. But despite widespread donor support for these PPPs in SRH there appears to be little evidence to support this. For example, despite claims that social franchising can improve access to and the quality of family planning services (Reference MontaguMontagu, 2002), reviewers find it difficult to draw firm conclusions about their effect (Reference Agha, Mehryar Karim, Balal and SoslerAgha, Mehryar Karim, Balal, & Sosler, 2007; Reference Koehlmoos, Gazi, Hossain and ZamanKoehlmoos et al., 2009). Advocates suggest that social franchises could extend coverage to areas which are under-served, but in practice the franchises tend to operate in areas that are already covered by services and where private providers wish to live and work. Private clinics are typically recruited into a franchise when already in business, and private concerns are known to often congregate in areas with sufficient population and transport routes to provide an assured market for their services (Reference Ravindran and FonnRavindran & Fonn, 2011, p. 93). The Merrygold franchise in Uttar Pradesh, for example, is in direct competition with public providers offering free contraceptive services, suggesting that they are operating in areas which are relatively well served (Reference Ravindran and FonnRavindran & Fonn, 2011) and low-income groups are less likely to use the services compared to better-off women. In the words of a WHO report, there is ‘a trade-off between serving the poor, providing a full range of reproductive health services and financial sustainability of the franchise’ (2007, p. 8).

Critical attention needs to be paid to the choice of measurements selected to assess progress towards meeting goals around reducing maternal mortality and how the choice of targets can lead to unintended consequences for individual's health and well-being (Reference Yamin and BoulangerYamin & Boulanger, 2013), and the balance between reconciling the need for global standards with the need to take account of local realities (Reference SpanglerSpangler, 2012). It is important to determine how far prevailing norms acknowledge the importance of women's health knowledge or create spaces for women to define their own health needs and to challenge programmes where there is a deficit (Reference GideonGideon, 2014). Relevant local sources of information around women's health may not be taken into consideration in decision making because of the attractiveness and ‘scientific’ nature of the Disability Adjusted Life Years (DALY) calculations or EBM), both of which have been critiqued because of their inherent male bias (Reference GoldenbergGoldenberg, 2006, Reference Goldenberg2010; Reference SundbySundby, 1999). A number of empirical studies have shown that ‘local level’ knowledge is often sidelined within dominant technical approaches to healthcare (Reference BerryBerry, 2014; Reference EriksonErikson, 2012). Community level health workers are those most likely to understand the challenges that women face in accessing maternity care and how this relates to broader societal and infrastructural challenges including gender norms, access to cash and transport (Reference Theobald, MacPherson, McCollum and TolhurstTheobald, MacPherson, McCollum, & Tolhurst, 2015, p. 5). Yet the higher respect accorded to statistical evidence has meant that long-term involvement of community workers and their field-based knowledge is no longer considered valid and such actors often no longer have a voice in the development process (Reference Adams, Biehl and PetrynaAdams, 2013; Reference EriksonErikson, 2012; Reference MishraMishra, 2014).

Question 3: To what extent do the SRH-PPP interventions take into account, reinforce or confront the existing realities of gendered social and economic life?

The narrowly focused approach of many initiatives has meant that structural influences on well-being are frequently overlooked (Reference Gideon and PorterGideon & Porter, 2016; Reference RavindranRavindran, 2011). Within PPPs women tend to be constructed as ‘deficient subjects who deserve investment’ (Reference PruglPrugl, 2017, p. 44) and as ‘responsible mothers, consumers and citizens who help others’ (Reference SatoSato, 2016, p. 163). There is a failure to acknowledge that pathways into poverty and poor health are frequently gendered and a tendency to marginalise unpaid care work and non-quantifiable, non-marketable values (Reference Prugl, Grosser, McCarthy and KilgourPrugl, 2016; Reference RazaviRazavi, 2012). While many interventions have focused on improving the MMR through increasing women's access to formal birth facilities these types of interventions fail to address the underlying inequalities that contribute to the exclusion of particular groups of women from services in the first place. This is evident in the Brazilian case where maternal deaths occur most frequently among poor black women and some of the most significant factors contributing to this maternal mortality are the high frequency of caesarean sections, illegal abortions, and regional and socio-economic inequalities in health (Reference Gideon, Leite and Alvarez MinteGideon, Leite, & Alvarez Minte, 2015, p. 259). In Nicaragua the criminalisation of abortion in all circumstances, even if a woman's life is at risk, has undoubtedly contributed to the high levels of maternal mortality among the more marginalised sectors of society (Reference KvernflatenKvernflaten, 2013). These critical factors underpin high rates of MMR in certain parts of the world but are often not factored into the framing of interventions which seek to promote better access to packages of biomedical services.

Health policies also often contain implicit assumptions about the unitary nature of households which are highly problematic given the complexities of intra-household relations and family structures (Reference GideonGideon, 2014). Complex decision-making processes which may be highly gendered can play a significant role in shaping where women (and of course other family members) seek healthcare and what happens to them afterwards. A detailed longitudinal follow up of women in Burkina Faso whose lives were saved by biomedical intervention during a pregnancy complication – an apparent success story – has demonstrated the subsequent loss of status and personal impact of having become a ‘costly’ wife and a drain on a household's meagre resources (Reference Murray, Akoum and StorengMurray, Akoum, & Storeng, 2012; Reference Storeng, Murray, Akoum, Ouattara and FilippiStoreng, Murray, Akoum, Ouattara, & Filippi, 2010). In the context of the growing marketisation of healthcare and increasing out-of-pocket payments, it is essential that policy makers consider the gendered allocation of resources within households. While there is not extensive empirical research in this area, research in Latin America has shown that women tend to be more negatively impacted by the marketisation of healthcare services compared to men (Reference Ewig and Hernández BelloEwig & Hernández Bello, 2009).

We now employ our three questions to consider a voucher and contracting approach in more detail, drawing on empirical data from a case study on experiences with implementation of the Sambhav scheme in Uttar Pradesh, India.Footnote ³

What ‘common sense’ assumptions underpinned the voucher scheme?

Like with other types of PPP intervention in the health field, policy support for healthcare vouchers has been driven by an assumption that high maternal mortality and fertility rates can be reduced through increased targeted healthcare utilisation by the poor – including childbirth in medical facilities. The second underlying assumption is that low usage is primarily attributable to poor quality and inefficiency of public sector healthcare and the costs to the users of access, and that private providers will be more effective. In this scenario voucher and contracting programmes are purported to increase poor women's demand for services, ‘empower’ them as service users by giving them means to exercise choice, and create incentives for providers to be ‘innovative, cost effective, and responsive to their clients’ (World Bank, 2005, p. x).

Rationales given for development of the Sambhav scheme in Uttar Pradesh typified such arguments in favour of health vouchers. In the lead-up, a USAID-commissioned report declared that ‘the government of India have recognized that unmet reproductive and child health needs outstrip their capacity and financial resources’ and there was ‘great potential to tap the private sector to expand the provision of quality RCH [reproductive and child health] services’ (Population Technical Assistance Project, 2004, Appendix A). Later programme documents made similar claims (IFPS Technical Assistance Project, 2012a).

It could be argued that the specific SRH aims of the scheme were really secondary to its fundamental ideological purpose, that of encouraging and extending commercialisation processes in public sector healthcare in India. Two rhetorical tropes were used throughout the Sambhav scheme's design and implementation to support a framing of this process: partnership and innovation. Partnership as Standing has observed, ‘has a nice cuddly sound to it’ (Reference Standing2007, p. 519), but in reality the Sambhav scheme's ‘partnership’ was a contracted service delivery model that had little precedent in Uttar Pradesh's healthcare systems and which minimised the direct involvement of the state in provision while creating a range of opportunities for private sector actors. Funding was channelled through a non-governmental organisation established by USAID. Government healthcare service providers were not involved in the programme but the state government was nonetheless required to monitor and regulate services and to provide free tetanus vaccines and contraceptives to participating private providers for their use with voucher users. Programme designers hoped that ultimately the government would take on the role of subsidising the private providers after the withdrawal of USAID funding.

‘Innovation’ was also a euphemism for commercialisation. A USAID-commissioned background document for the Sambhav scheme made clear the link between innovation and commercialisation in provision of healthcare:

Programme documents described an experimental framework for development of the voucher and contracting scheme (IFPS Technical Assistance Project, 2012b). This consisted of five stages – develop, design, demonstrate, document and disseminate. Aside of a fondness for alliteration, this framework reveals that the emphasis was not on testing this ‘innovation’ as might be anticipated, but rather on demonstrating it as a service delivery model. That ideological mission was reflected in the focus of the programme evaluations.

On what criteria was the voucher scheme framed and judged?

This driving concern to demonstrate that the Sambhav scheme was an appropriate service delivery model for SRH in Uttar Pradesh was reflected in the very limited focus of the three programme evaluations, which were designed to show voucher uptake and to glean ‘lessons’ for refining future implementation. The first two evaluations, by Futures Group in 2012, and Ipsos Research in 2013, described how many vouchers had been used, while making claims about increased uptake of maternity care that did not account for the concurrent effects of other programmes (IFPS Technical Assistance Project, 2012a; Ipsos Research, 2013). In a revealing foreword to the Futures Group evaluation report, Director of USAID India's Health Office Kerry Pelzman explained the tactical nature of the Sambhav scheme as a ‘prototype’ model for future service delivery:

A third evaluation, also published by USAID in 2013, again highlighted how many vouchers had been used, but this time did not make any claims to success in terms of effect on uptake of services. The report provided a more critical analysis of the Sambhav scheme's service delivery model, on the grounds that it had little prospect of financial sustainability once aid was withdrawn. Despite these reservations the report's conclusions, which also drew on experiences with three other PPPs in northern India (all of which had poor prospects for sustainability), still stated that the Uttar Pradesh state government should ‘focus on health systems strengthening by exploring options for PPP for health in order to expand quality care and services’ (Reference Andina, Roy, Bhawsar, Ghosh and RuduchaAndina, Roy, Bhawsar, Ghosh, & Ruducha, 2013, p. 78). Thus concerns with the PPP models were secondary to the need to expand private sector involvement in public healthcare provision, and the gendered dimensions of the programme and implications for women's health were incidental.

The ideological project – to promote a model for commercialisation of public healthcare services via a SRH scheme – has been very apparent. Despite the clear weaknesses and very limited scope of research and evaluations, ‘lessons’ promoting the Sambhav scheme model have been widely circulated and have been influential in India and in the global health arena. The scheme recently featured as a positive example of ‘innovation’ in two publications by the Public Health Foundation of India (itself a public-private partnership): a ‘white paper’ entitled Innovative Ways to Meet Health Challenges of Urban India (Reference Arora, Neogi and MisraArora, Neogi, & Misra, 2011) and a book Innovations in Maternal Health: Case Studies from India (Reference Satia, Misra, Arora and NeogiSatia, Misra, Arora, & Neogi, 2014). At an international level, ‘success’ experiences with the Sambhav scheme were presented at the 2009 and 2011 International Conference on Family Planning. The former, in Uganda, was followed by the launch of a voucher and contracting programme in Uganda that was funded and managed by the conference's leading organisations: Bill and Melinda Gates Foundation, Johns Hopkins Bloomberg School of Public Health and Makerere University School of Public Health. Documentation on the Sambhav scheme is typical of a wider policy landscape that prioritises narrow interpretations of EBM at the expense of critical discussion on how policies are formulated, framed and implemented (Reference Adams, Biehl and PetrynaAdams, 2013; Reference Ghaffar, Gilson, Tomson, Viergever and RottingenGhaffar, Gilson, Tomson, Viergever, & Rottingen, 2016; Reference Murray, Hunter, Bisht, Ensor and BickMurray et al., 2014). Given the inherent problems in contracting private providers to deliver public SRH services and likely implications for women's health and position in society, it is important that evaluations of PPPs go beyond short-term assessments of service use and predictable analyses of programme ‘challenges’ if policy-makers and practitioners are to learn any meaningful lessons where interventions are replicated on a large scale.

What gendered aspects of social and economic lives are reinforced or challenged by the voucher scheme?

So what part do such voucher and contracting schemes play in empowering women? Most importantly from a gender perspective, the programme evaluations and the subsequent promotional literature on the Sambhav scheme (and on voucher and contracting approaches more widely) failed to examine the programme's interplay with gender norms. Analysis of our empirical data reveals that the Sambhav scheme reinforced gendered household and community roles during voucher distribution. Easy notions of ‘empowering’ individual women failed to take into account that other more powerful family members – specifically a woman's husband and/or mother-in-law – played key roles in decision-making on where she would seek care. Community workers did little or nothing to challenge this, as one explained during an interview:

Moreover, any potential for ‘empowerment’ and ‘choice’ for women was moderated by (typically) male managers and the female community workers who worked under them, who had vested interests in promoting a narrow interpretation of the Sambhav scheme, publicising and validating their own activity while (in the case of the community workers) minimising their own personal travel costs. Those community workers were women who lived in or near slum areas and who met minimum criteria for literacy and marital status. They were typically better educated than other women in the slum areas and had prior experience with community programmes, and their monopoly of access to information on the programme's details further distorted their local relationships.

Community workers withheld vouchers from women service users in order to prevent vouchers being damaged or lost, and instead preferred to be personally visible accompanying women to nearby participating hospitals. In some cases therefore women were never themselves aware that the vouchers had a physical form nor that they had a choice of which hospital to go to in the scheme, they simply knew that their local community worker could help them to obtain cheap care at a specific hospital – typically a nearby hospital as community workers did not feel they were adequately reimbursed for transport costs. This can be seen as an instance of women taking power away from other less knowledgeable women, and signifies that while community health workers play an important role in building relationships between communities and healthcare systems (Reference GilsonGilson, 2005; Reference MishraMishra, 2014), they can also subvert those relationships. It is therefore important not to over-romanticise the role played by community workers, and to understand the context in which their activities take place.

The fetishisation of the paper voucher reflected its function within the project as a tool for monitoring and accounting; programme managers would only process payments to community workers and hospitals if a voucher had been physically submitted. The programme also created other pressures on the community workers by setting targets that they must meet and by competing with the government maternity services. The government's Janani Suraksha Yojana [Safe Motherhood Scheme] paid a woman 1000 rupees (£10) if she gave birth in an urban government facility. To meet their targets the community workers therefore tried to convince families to use their voucher scheme and not government services by presenting this good quality of care versus cash. The pragmatic response of the community workers was encouraged by the precarious nature of their employment at the front line of service delivery and reveals ways in which the gendered nature of health systems reflects and reinforces power relations in communities.