The emergence of the national medical assistance scheme for the poorest in Mali

ABSTRACT Universal health coverage is high up the international agenda. The majority of the West Africa's countries are seeking to define the content of their compulsory, contribution-based medical insurance system. However, very few countries apart from Mali have decided to develop a national policy for poorest population that is not based on contributions. This qualitative research examines the historical process that has permitted the emergence of this public policy. The research shows that the process has been very long, chaotic and suspended for long periods. One of the biggest challenges has been that of intersectoriality and the social construction of the poorest to be targeted by this public policy, as institutional tensions have evolved in accordance with the political issues linked to social protection. Eventually, the medical assistance scheme for the poorest saw the light of day in 2011, funded entirely by the government. Its emergence would appear to be attributable not so much to any new concern for the poorest in society but rather to a desire to give the social protection policy engaged in a guarantee of universality. This policy nonetheless remains an innovation within French-speaking West Africa.

the so-called vulnerable people who contribute modestly, and poor people who do not contribute at all. The differentiation between these two categories is based on a targeting score specific to urban and rural areas. Policy funding comes primarily from State and local authorities and also from vulnerable people. RAMED is managed by the National Health Insurance Agency (ANAM). The emergence of this policy results from a long process of maturation, trials, and draft laws which were at times forgotten before being resumed. However, the RAMED there fundamentally responds to the need to 'ensure the stability of the political systemi.e. the protection of the monarchy's leadership within a determined process of liberalisation of authoritarianism' (Ferrié et al., 2018). The aim appears to be truly focused on the need to 're-legitimise the monarchy' (Ferrié & Omary, 2019).
In Sub-Saharan Africa, a handful of countries have embarked upon these public policies of social protection for the poorest. In Burkina Faso, the national and international stakeholders have galvanised themselves and coalesced in order to develop such a social protection policy (Kadio et al., 2018b). However, the formulation of its content can be classed more as a compilation of the pre-existing actions, against a background of aid dependency, rather than a genuine and coherent reflection of the interventions needing to be organised in order to meet the needs of the poorest (Kadio et al., 2018a). In Benin, the extensive process of targeting the poorest seen since 2013 has primarily resulted from the impetus of the World Bank (Deville, Escot, et al., 2018). In Senegal, the state has, since the same year, been running a programme aimed at the poorest which not only grants them direct cash transfers but also allows them to become members of the community-based health insurance (CBHI) schemes free of charge. The emergence of this programme in Senegal was largely political and 'without consultation' (Ferrié & Omary, 2019). In Ghana, people classified as the poorest (also beneficiaries of a cash transfer programme, Livelihood Empowerment Against Poverty) are enrolled free of charge on the national health insurance scheme, on paper since 2004 but in reality since 2013 (Umeh, 2018). Political issues, are the main reasons behind the emergence of this policy, even if the decision appears to be more symbolic than based on any true commitment (Kotoh & Van der Geest, 2016). In Rwanda, the poorest also benefit from exemption from the payment of the health mutuals' subscriptions and there too, it was the ideology of the party in power after 1994 and the power balance with the Global Fund to Fight AIDS that resulted in the emergence of this policy and the funding of this subsidy for the poor from 2007 (Chemouni, 2018).
To our knowledge, Mali is the only country in French-speaking West Africa since 2011 to have put in place an ambitious self-funded national policy specifically designed to provide the poorest with free medical cover. The aim of this article is to understand the process that has led Mali to proceed with the formulation of a social protection policy of this type.

The ramed context in Mali
Since 1968, a decree has applied to the regulation of help for the poorest but it has never been applied. In the 1980s, Mali was one of the first countries to roll out a cost recovery scheme whereby health service users were asked to pay at the point of use (Maïga et al., 1999). The nation has since undergone multiple health reform processes, whether to improve the health system's governance and decentralisation, to train personnel or to enhance the quality of care. Faced with the challenge of financial accessibility, Mali put in place several user fee exemption policies from 2000 to 2010. However, these policies were not as effective as had been hoped (Ravit et al., 2018) and the difficulties linked to implementation and funding remain considerable (Touré, 2015). Recourse to healthcare for the sick is still just as low and the most disadvantaged still experience major difficulties obtaining treatment in the formal sector.
Consequently, within the governmental context of the creation of compulsory medical insurance (AMO) to cover the risk of illness among civil servants and employees from the formal sector (17% of the population), the state arranged the set-up of a non-contribution-based system (the medical assistance scheme, or RAMED, formerly the medical assistance fund (FAM)), aimed at providing free medical care for the poorest (5% of the population). In order to cover the remaining 78% of the Malian population coming under the informal sector, the government decided to rely on the community-based health insurance (Deville, Hane, et al., 2018).
The RAMED is managed by the Agence Nationale d'Assistance Médicale (ANAM), a public administrative body with legal personality and financial autonomy. On its foundation in 2009, the ANAM, which is responsible for the registration of beneficiaries, was placed under the supervision of the Minister for Community Development and Humanitarian Action (MSAH). The RAMED is a scheme aimed at the poorest (indigents), namely 'any person deprived of resources and recognised as such by the local authority covering them', (Act 09-030, 2009). The care package offered is equivalent to that received by beneficiaries of the AMO, excluding the cost of speciality medicines. The RAMED is financed exclusively via public funds, with the state providing 65%. It was originally envisaged that the remaining 35% would come from the local authorities but this provision, which was in many cases neither accepted nor applied, was revised in May 2017 and the share payable by the local councils cut to 15%.

Methods
This involved qualitative research conducted in the form of a holistic case study (Yin, 2009), the case concerned being the RAMED. Without falling into the stagism of the analysis of public policies, we were interested in its emergence (i.e. agenda-setting [Kingdon, 1995]). The empirical data results from two complementary methods.
Firstly, 11 qualitative and in-depth individual interviews were conducted from January 2017 to March 2018 in Bamako, the capital of Mali, using a qualitative sampling strategy (Palinkas et al., 2013) based upon the criteria of knowledge of and involvement in the emergence of the RAMED and, by way of a snowball effect, by following the recommendations of our initial interviewees. Thus, we met with the main persons to have played a central role in the emergence of this policy and representing all of the institutions involved, namely the health ministry (n = 1), the ministry for community development and humanitarian action (n = 9), and the Union Technique de la Mutualité (n = 1). The interviews were conducted in French by LT. They were all recorded and transcribed in full in order to facilitate their analysis.
Secondly, the empirical data also stemmed from numerous (n = 29) documents relating to social protection in general and to the RAMED in particular. Serving to facilitate understanding of the historical development of the process and the different stakeholders involved, these consisted mainly of public policy documents, legislative documents and articles and studies on the subject or on social and health policy from the period concerned.
An inductive qualitative analysis of the content (Paillé & Mucchielli, 2003) of the data resulting from these two collection tools was conducted, while triangulation processes between this data and the discussions on the preliminary analyses with the parties concerned helped to improve the research's internal validity. The results were primarily presented during two deliberative workshops held in Bamako in the presence of various stakeholders concerned by the analysis (June 2018 (n = 16) and February 2019 (n = 33)). Policy briefs also helped to facilitate the discussions with the stakeholders regarding our analyses (http://www.miselimali.org/ Productions_scientifiques.D.htm). The

Results
The history of the RAMED's emergence stretches over a period of almost 20 years, between 1991 and 2009, the year in which the legislative texts were finally adopted. This history was punctuated by four main periods.

1991-1997: A degree of political will
Alpha Oumar Konaré, the first President to be democratically elected following the 1991 coup d'état that put an end to 23 years of military dictatorship, was sympathetic regarding social issues. In 1980, while a civil servant in the national education department, he had been the instigator of Mali's first health mutual, the MUTEC. In 1992, the President expressed the desire to separate the department of social affairs from the health ministry and to bring it under the control of local government ministry. However, this initiative was not regarded favourably and the trade union for personnel from the health and social action sectors arranged marches to oppose it, leading to the project's abandonment. Following a study tour to Tunisia in order to observe the social protection system, the President then expressed the desire to develop social protection in Mali. This resulted in the 1993 declaration of a community development policy.
This period was characterised by a high level of planning activity. In the area of health, this took the form of institutional reforms and the creation of a health development plan (PDS) 1998-2007, which for the first time made reference to eventual guaranteed UHC through the mutuals (for the liberal professions and the self-employed), compulsory medical assistance (for employees and civil servants) and a medical assistance fund for persons not covered by any other insurance scheme and recognised as indigent.
The technical and financial partners (PTF) provided support throughout the PDS's planning phase, but they began to focus more on the reinforcement of the offer via the implementation of the ambitious sectoral decentralisation plan in the field of health, displaying a marked disinterest in the social issues. The Malian government was thus forced to impose inclusion of the social sector in the shape of the Programme de Développement Sanitaire et Social (PRODESS 1998(PRODESS -2002.
"The partners regarded social affairs as an extension to health. Even the World Bank did not at the time want social action to be included in the PRODESS. Its representatives would say that it was not possible to take taxes from people to come and assist the poorest in Mali when they had their own poor. That was the idea they put in the partners' heads and it took all the energy of Modibo Sidibé at the Ministry of Health to get social action included": a former executive at the community development ministry.
An ad hoc group, set up within the framework of the development of a ten-year plan and placed directly under the responsibility of the Prime Minister, was tasked with development of a social protection policy.
Against a background of structural adjustment, the foreign experts, galvanised for the development of the tenyear plan, brought forward the idea of planning a third strand to the PRODESS entitled "alternative health funding methods with AMO, FAM and mutuals": a DNDS manager.
Technical capacity for the management of sickness insurance in Mali was then practically nonexistent (Letourmy & Ouattara, 2006). In 1997, the ad hoc group commissioned a feasibility study which presented an institutional set-up of the different schemes envisaged (AMO and FAM) and an estimation of the technical costs and the contributions according to different healthcare baskets. On this basis, the ad hoc group put forward a draft law that established a social protection code, but this was never adopted by the government or submitted to a vote due to fears of the financial consequences for the state. The only real achievement was the production in 1996 of a legislative and regulatory system governing the mutuals. The Union Technique de la Mutualité (UTM) was then created in 1998, with French and Belgian assistance at a technical level.
With regard to the poorest, the results generated during this period were decidedly mediocre. The political will of the government to establish a genuine community development policy resulted in the first ever entry in the state budget of a funding line devoted to actions and programmes for community development, at a level of 43 million CFA francs in 1992 and up to 110 million in 1997 (Sidibe, 2017).
From 1998 to 1999: The first phase of interruption It was during this period that a first interruption occurred in the social protection policy's process of emergence. The ten-year plan of which it formed a part was the subject of lengthy consultation, which delayed the progression of the work (Sidibe, 2017). Moreover, this plan needed to be rolled out in two five-year programmes (PRODESS), the finalisation of which was delayed. The implementation of the PRODESS's institutional architecture proved to be a major struggle.
"There was a lack of proactivity regarding the set-up of the PRODESS's different bodies. All of these bodies needed to be set up but that didn't happen until 1999 and 2000. And then the funds were not entirely available because there were several partners involved and multilateral aid to be coordinated. All of this meant that it wasn't until 2000 that the matter was able to be resolved, after quite a bit of delay": a former executive from the MDSSSPA Ministry.
Lastly, the reformist trend generated some concerns. The presidential vision of social protection struck some as avant-gardist. Doubts regarding both the state's will and technical capacity to put in place such systems for the benefit of the population were clearly expressed. Against a newly democratic background, the imposition by the state of a system that was compulsory for some (civil servants from the AMO) and non-contribution-based for others (the poorest from the FAM) seemed anachronistic and provoked some resistance.
"In the end, only the community-based health insurance (mutual) aspect was developed. The AMO and the medical assistance fund, meanwhile, had to be halted and time out taken, because people didn't understand that the aim was to introduce compulsory sickness insurance. The group's ad hoc work had begun in 1996, just five years on from the 1991 revolution, which led to a coup d'état and democracy being introduced. In fact the people had begun to feel free and this promise of freedom did not fit in well with any concept of obligation. So people were saying to each other than nothing should be compulsory any more": a former MSSPA ministry executive.
The anticipated institutional upheaval in the health field also caused concerns, particularly the administrative and sectoral decentralisation that would grant certain powers to community stakeholders and to elected representatives.

From 2000 to 2005: Intensive discussion and study activities against a background of doubts
In 2000, the government achieved a major institutional reform: the creation of a ministry for social and community development and the elderly (MDSSPA). Its aim was to develop a whole new conception of the social sector.
The new body believed that the old forms of social protection, of which the effectiveness and human quality remained undeniable, could not cover the entire field. It wanted to encourage a social protection policy which resulted from straightforward assistance and involved those concerned and the population as a whole. (Sidibe, 2017).
Two strong initiatives underline this new ministry's desire to bring about emancipation, which was hampered by a lack of human, technical and financial resources: -the research by International Labour Organization (ILO) technical assistance, which funded the conducting of three studies on the analysis of the social protection system. -the set-up of a social protection orientation council (COPS) by the MDSSPA.
2001 saw the holding of the first social development sittings to give content to this new ministry's policy.
"It was at that point that the idea was revived of having guaranteed medical cover via a social security scheme. But it wasn't clear what instrument should be used, which is why we called upon a French expert for a preliminary evaluation of what could be set up in Mali as a mechanism to help alleviate medical insecurity": a former MDSSPA executive. This national event marked an important turning point in the field of social development. Three main orientations were adopted by the government as a result of the work conducted at these first sittings: the consolidation of national solidarity and the reinforcement of the struggle against exclusion, the strengthening of social protection and the fight against poverty.
In order to ensure the reliability of the community system deployed in the field of health, interest in social protection clearly increased, as it had in the 1990s, and centred on effectiveness, in terms of asking how the poorest could be taken care of (Maïga et al., 1999).
"With the new policy, the CSCOM had to rely on user fees generated by these services. All of the studies conducted had yielded positive results, except for those on the issue of poverty. The poorest couldn't pay. It was clear that if free membership was introduced, the CSCOM would not survive. So it was necessary to find a solution for them. In order to avoid excluding the poorest, at the first PRODESS, it was concluded that the alternative system needed to consist of two strands: the CBHI for those able to contribute and, secondly, for the system to remain viable, we need to cover the issue of poverty": a DNDS executive.
In 2002, the declaration of a policy for the extension of social protection was adopted by the Council of Ministers. It would be followed in November 2004 by a national action plan for the extension of social protection in Mali (2005Mali ( -2009. In 2005, two national experts were appointed for the synthesis of the work conducted and the proposal of an institutional organisation plan for the AMO and the FAM. On the basis of their conclusions, complementary studies were requested, 'relating to the parameters and quantitative elements of the systems proposed both for the AMO and for the FAM' (Letourmy & Ouattara, 2006). A second feasibility study put forward different scenarios relating to the system's institutional organisation and the updating of the data followed by a national consultation regarding its conclusions (Letourmy & Diakité, 2003).
Concerning the FAM, which would later become the RAMED, we did not find any mention of analytical studies on the context, existing systems or current practices regarding the coverage of the poorest by the health structures, and their limitations. To our knowledge, a single local experiment with a medical assistance fund project has been conducted by MSF-Luxembourg (MSF-L), successively at two district hospitals (CSREF) in the region of Sikasso, Sélingué in 2001 and at Bougouni in 2003. However, after three years of operation, the FAM in Sélingué had supported just 2% of hospitalised patients, of which only 9% were recognised as poorest and entitled to full user fee exemption (Noirhomme & Thomé, 2006).
The Malian state has struggled to establish its credibility with regard to its declared aims, with numerous doubts arising in relation to its ability to equip the country with such systems: poor nation, problems of governance, health system deficient or corrupt, etc. 'Another issue was that the people didn't believe in social protection': an MSAH executive.
The background was that of a social sector which was struggling to separate itself from the field of health in order to reinvent itself around more ambitious policies. This longstanding control of social protection by health, initially institutional up until 2000 (a single minister for both sectors), then financial within the framework of the PRODESS funding, and finally symbolic served to maintain a somewhat wait-and-see approach on the part of social contributors, who were also faced with a chronic deficit in terms of human resources. The funding from which the new ministry benefited was the subject of bitter debate within the health sector, as it included no support from partners and was always well below that required to fulfil its declared aims.
"15% of the PRODESS budget was allocated to social protection. And even getting to that 15% level required negotiations with ministers because at the start, we only got 5%. That was around 2006, I think. The World Bank at the time placed us at a disadvantage by saying that it only wanted a single contact. And when there were two ministers, the World Bank said that its sole contact would still be the Health Minister": a former MDSSPA executive.
"Another element of blockage was the fact that there were no available partners, due to a lack of belief": an MSAH executive.
Eventually, this new Ministry for Social and Community Development and the Elderly, which was responsible for guiding and uniting the stakeholders around a process for the development of a social protection system, had to build itself up and equip itself with new skills in terms of social insurance.

2006-2009: The system's slow maturation and institutional tensions
At the time, UHC was not an international priority and the partners had few solutions to offer to ensure support for the AMO (Noirhomme & Thomé, 2006) and the RAMED. This period was characterised by the project's slow advancement, small steps, lack of skills in the social sector, and fear of the reactions that these social reforms might generate. Moreover, despite the new President (elected in 2002) Amadou Toumani Touré (ATT) displaying a stronger desire than ever to obtain a national consensus, progress was decidedly sluggish.
The formulation of these policies and that of the AMO in particular accentuated tensions and opposition to the project and highlighted the weak contribution from other sectors.
Locally, the health care providers regarded this initiative, which affected them in more ways than one, as a dual threat: -as civil servants, they were likely to be members of the AMO and feared losing the benefits they had acquired compared to the other bodies, -as health professionals, since the deployment of these social protection systems would need to be accompanied by service control mechanisms, would favour generic medicines over specialist ones, and would not take into account the private sector in which they had special interests (Letourmy & Ouattara, 2006).
The social partners (trade unions in particular, see case of higher education staff [Goita, 2018]) have adopted a fairly ambivalent position regarding the set-up of the AMO. Mali's two major unions officially support the principle but fear the levelling of situations between civil servants that the AMO implies.
"The head of the Workers' Trade Union Confederation of Mali (CSTM) was against it and so all the unions linked to his confederation rejected the AMO. The UNTM, the other main trade union, was for the AMO and all the unions affiliated to it followed his line. Up until today, when the UNTM takes one direction, the CSTM takes the opposite one. These are conflicts that predate the issue of the AMO and are more or less attributable to political or union-related matters or opposition between individuals. One thing's for sure though: they never agree": a former MDSSPA executive.
Concerning the RAMED in particular, there have been fewer issues and fewer conflicts between those involved. The comparison with the history of the AMO is therefore heuristic. The FAM became the RAMED at the meeting of the general secretaries of the different ministries for the adoption of the text. Basically, they took the view that it amounted to a scheme, even though it is not contribution-based, and that it was therefore necessary to change its name. Apart from this issue, the RAMED appears very simple to implement compared to the AMO, which required construction ex nihilo, and its set-up generated very little debate.
"The discussions on the RAMED were concluded more rapidly. The targets were known and accepted, and the contributions were calculated. And it was settled. But on the AMO side, there was much more debate, as it provoked more opinions than the RAMED": a former MDSSPA executive.
In the end it was a version of the system that was both vague and unrenewed that evolved up until the formulation of the legislative texts submitted to the National Assembly. Laws and decrees concerning the creation of the ANAM and the CANAM (Mali's national health insurance fund), and of the AMO and the RAMED, were adopted in July 2009. The adoption of these texts did not of course indicate the finalisation of the formulation of these policies. Those responsible were aware that the texts needed to be clarified and considered that the discussion work needed to improve the formulation of the systems was now the responsibility of the teams recruited, both at the CANAM and the ANAM.

Discussion
This research is one of the first studies in French-speaking West Africa to uncover a long process of emergence of a public policy in favour of access to healthcare for the poorest populations. However, the study shows that the process has been long, chaotic, political, and largely ignored by international donors.
Research confirms the importance of historical authenticity of public policies and the temporality of analyses (Laborier & Trom, 2003). The process was discontinuous, consisting of successive periods of intense activities and periods of interruption. This process resonates with the theoretical proposals of True et al. (2007) and their proposal for a punctuated equilibrium. Sabatier & Weible, (2014) suggests studying public policies over a long period of time, in order to understand and see possible changes. Indeed, our study confirms the national challenges of being part of international processes and goals in favour of the poorest population and universal health coverage (UHC). The Millennium Development Goals (MDG) were followed by the Sustainable Development Goals (SDGs) and international declarations in favour of UHC and leaving no one behind (69th World Health Assembly (WHA) in 2016). However, these normative approaches remain declarative, with little hold at the national level. The perpetual highlighting of a (rhetorical) concern for the poorest population in international declarations seems to have difficulty translating into the emergence of national policies, particularly in West Africa, not to mention the ultimate challenges of their implementation, which we have not addressed here. The equity-focused and human rightsbased approach still has a long way to go in this region (Ooms et al., 2019;Robert, Lemoine, et al., 2017) as in Benin and Senegal (Paul et al., 2019). We are still far, as in Senegal, from a definition of UHC which recognises that its objective is the 'creation of citizen subjects and autonomous subjects' (Mladovsky, 2020).
This case study also shows, despite international declarations, that Mali has mostly found itself alone in tackling the problem of access to care for the poor. While there have been some strategies to learn from other countries in a timid policy transfer process, these have had little effect, unlike Rwanda, where political will has been stronger (Chemouni, 2018). As for Ghana's social protection programme where those in charge in Mali travelled to Brazil (Foli, 2016), it amounted to a new demonstration of the role (and ineffectiveness) of study tours in the public policies transfer in Africa, especially in Mali (Gautier et al., 2019). This type of strategy is so important that certain international organisations have produced guides for organising such study tours (Kumar & Watkins, 2017). In a West African context where official development assistance plays an important role in State budgets and the diffusion of ideas (see below), the research found very little involvement of international partners in the process of emergence and, above all, formulation of this policy, apart from a brief pilot experiment carried out by an emergency NGO which did not seek to continue it. This analysis shows, once again, the crucial role played by international aid organisations in the emergence of policies in Africa and the challenges involved in obtaining their support (Gautier & Ridde, 2017;Lee & Goodman, 2002).
In addition, the absence of investment in pilot projects demonstrates the low degree of importance granted to this issue by international donors. In fact, this region of Africa, including Mali, has seen a large number of experiments funded by donors concerning the extension of user fees for healthcare (Ridde, 2015), but very few, if any, regarding healthcare access for the poor. It was only in 2019 that Senegal organised a pilot project for the integration of its free healthcare policies within its national health insurance system (Daff et al., 2020). In Burkina Faso, user fees exemption pilot projects have supported decision-making on national policy for pregnant women and children (Ridde & Yaméogo, 2018), but this has not been the case for the poorest population despite more than 10 years of pilot projects (Ouédraogo et al., 2017). Conversely, when the situation in South-East Asia is examined, it is clear how many of these pilot projects have been instrumental in the emergence of social protection policies in favour of the most deprived (Asante et al., 2019). However, several user fee exemption policies (HIV in 2004(HIV in , caesarean in 2005(HIV in , anti-malaria for children and pregnant women in 2006 have been put in place in Mali, decisions which were highly political and technically contested in their formulation (Touré, 2015). Championed by the Health Ministry and widely supported internationally (Robert, Samb, et al., 2017), these initiatives occupied the forefront of the political scene and attracted widespread attention, eclipsing the development of the social protection system and health cover for the poor.
In addition, this research highlights the importance of the appropriation of public policies by African governments and the weight of the ideas and funding from international partners in the field of health (Gautier & Ridde, 2017;Kuhlmann et al., 2019) and social protection (Foli, 2016). In the case in point, and contrary to the norm, the national party was in a position to be able to stand strong so that its ideas on the role of the social sector could prevail. The role of these ideas in the emergence and transfer of public policies (Kuhlmann et al., 2019) was also confirmed in this West African country's social policies (Béland & Ridde, 2016). The issue of the compulsory nature of membership of the CBHI and the state's role of stewardship had been central to the emergence of the UHC in Rwanda (Chemouni, 2018), in contrast to in Mali and Senegal . Indeed, a 2016 study conducted in Mali demonstrated that it was 'the obligation that had been the decisive factor in the rejection of social insurance' (Goita, 2018). One another striking aspect revealed by this analysis in Mali is the repeated use of the same experts. It was surprising to observe the lack of diversity in the expertise deployed to offer advice to the government. Most of the reports were written by the same experts, who are usually from France or trained in France. This lack of diversity shows how much the role of the technical entrepreneurs of spreading ideas is central in this context (Gautier et al., 2019(Gautier et al., , 2020, as well as the role of consultants in public policy in West Africa. Finally, this case study highlights the challenges of health intersectorality, despite international calls for SDGs and science (Hussain et al., 2020). In Mali, the new Ministry for Social and Community Development and the Elderly was forced to impose its leadership on the other ministries involved, which were not always interested or cooperative, such as the Health Ministry, which had the supervision of this health insurance project taken away from it. Since Alma-Ata in 1978 up until the present day and UHC (Hussain et al., 2020), this issue of intersectoriality keeps emerging, and is central to the current construction of public policies in Africa (Blas et al., 2016). The intersectoriality remained very difficult to implement, and the famous advocacy coalitions (Sabatier & Weible, 2014) were not working towards the consideration of equality in this context. The extremely lengthy duration of the process and discussions, along with solutions of which the stability worked against the necessary changes for greater fairness, recalled the theoretical proposals of Hall (1993) and the paradigms of public policy. Hall (1993) clearly showed, however, just how important it was to take into account tensions between 'puzzling' and 'powering' and, in the present case, the asymmetry of powers between the network of stakeholders, particularly between international and national ones. The consideration of the role of the ideas of the stakeholders in Mali active in maintaining the relative stability of social protection policies became essential (Daigneault, 2014).
However, a slow but positive process of change can be observed in the health sector in favour of social protection systems. This paradigm shift is subtle (Daigneault, 2014), and analysis of these policies clearly needs to be conducted over a longer term (Sabatier & Weible, 2014). From total disinterest towards this 'utopia' at the start, social protection has become established as an alternative to user fees exemption policies. Imposed without any consultation by the country's highest authorities, the latter are in fact particularly poorly received by health professionals and even considered harmful to the viability of the health structures and the personal interests of professionals. Furthermore, the health professionals believe that the structuring of these policies in relation to social protection mechanisms has not been sufficiently considered or planned in advance by those responsible (Touré, 2013(Touré, , 2015. The AMO, therefore, appears to be the best option for health professionals, provided that it is possible for its implementation conditions to be negotiated. Furthermore, in perpetual competition and on a quest for visibility and power, the consequences of which for public policies are well-known (Béland, 2010;Hall, 1993), they generally adopt opposing positions regardless of the subject under discussion, and irrespective of the consequences.

Conclusion
The analysis of the emergence of this ambitious policy in favour of Mali's poorest citizens demonstrates the importance of the role of the policy entrepreneurs when it comes to breaking away from processes which are secular in terms of sector. This is made all the more remarkable by the fact that access to healthcare for the poorest has not been a priority, either for the national authorities or for the international partners. But this attempted break with tradition has occurred over a lengthy time span and against the background of a battle between sectors, most notably between health and social protection. It is almost as if the issue of poverty and its inclusion in social protection policies is a foil to the principle of universality, which has struggled to find a place within a system that especially benefits the poorest. The complexity and low effectiveness of the CBHI have restricted the decisionmakers in Mali in their development of UHC for the two extremes of the social continuum, i.e. the very poor and the less poor. However, this emergence of a policy in favour of the poorest is not for the moment synonymous with genuine consideration for the most deprived. The analysis of the formulation and then the study of its actual practical application would certainly show us all the importance (or not) that those in charge give to their fellow citizens living in poverty. The fact that the Malian government has decided (in June 2019) that social protection will once again be integrated within the Health Ministry to the detriment of the Social Action Ministry does not augur well in this regard.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work has been conducted with the financial support of the initiative 'Innovation for mothers and children's health in Africa' which was granted by the Ministry of External Affairs, Commerce and Development of Canada, the Canadian Institutes of Health Research, the International Development Research Centre. These organisations did not have any role in the design of the study, the collection, the analysis or other parts of the research.