Domestic humanitarianism: the Mission France of Médecins Sans Frontières and Médecins du Monde

Abstract What are the boundaries of humanitarianism? This question is controversially debated among humanitarian practitioners and scholars, given ever-changing spaces and temporalities of human suffering. This paper explores an understudied site of this controversy: the domestic humanitarian engagement of Médecins Sans Frontières and Médecins du Monde, two NGOs widely regarded as epitomes of liberal international humanitarianism. Their Mission France started in the 1980s to support vulnerable populations in France through medical aid, socio-legal support, and political activism. It has provoked fierce internal opposition ever since, in the name of an inherited vision of humanitarianism as impartial emergency aid. Drawing on organisational documents, archival sources and key informant interviews, we analyse how these conflicts gave rise to an unstable settlement around the diagnostic function of Mission France: Leaders aimed to make the assistance function of Mission France secondary to the advocacy function of drawing attention to health inequity, thus avoiding any long-term substitution for state services. However, the political strategizing demanded by this approach clashed with local volunteer preferences for immediate aid and claims to political neutrality. This conflict about the hierarchy of humanitarian values in the NGOs’ home country sheds new light on the contentious politics of humanitarian witnessing and assistance. Supplemental data for this article is available online at https://doi.org/10.1080/01436597.2021.1916393 .


Introduction
In 2017, Médecins Sans Frontières (MSF, English: Doctors Without Borders) opened a day centre for unaccompanied minors in the suburbs of Paris (Médecins Sans Frontières 2018). In Germany, Médecins du Monde (MDM, English: Doctors of the World) provide medical care and social counselling to people who do not have access to health, while publicly denouncing exclusion from healthcare (Ärzte der Welt 2017, 34-37). In the response to the Covid-19 pandemic, MSF and MDM, like many humanitarian NGOs, provide aid to the underserved in countries around the globe, not only the global South (Kenny, Grant, and Tondo 2020). These activities differ from the dominant conceptualisation of humanitarianism as a practice of ARTICLE HISTORY emergency aid in distant poor or conflict-ridden countries (Barnett 2011;Redfield 2013). They also differ from the dominant image and self-conception of the 'French Doctors' , which were founded in 1971 (MSF) and 1980 (MDM) as internationalist organisations seeking to bring relief to distant sites of suffering. What does this reterritorialization tell us about the politics of humanitarianism and about conceptions of humanitarian spaces?
In this paper, we analyse the contentious evolution of the 'domestic outposts' of MSF and MDM: their program in France called 'Mission France' , in order to understand how the cultural repertoire of international humanitarianism is invoked, adapted, and resisted in spaces inside well-resourced welfare states. 1 We focus on MSF and MDM, whose French engagement started in the late 1980s and has sparked intensive debates about the aims and limits of humanitarianism and its relationship with the welfare state.
Drawing on key informant interviews and archival and online data about Mission France, we retrace the contentious genealogy and implications of a distinct humanitarian strategy marked by diagnostic interventions, political advocacy, and the pledge not to become a substitute for welfare state services. We show that MSF and MDM could invoke their symbolic capital as emergency fighters and thus the fact that their sheer presence has become a marker of humanitarian crisis. This diagnostic function of their engagement -the will to document and witness suffering -has taken centre stage in Mission France as its leaders have attempted to keep the assistance function in check. Pledging not to take over state functions -'non-substitution' -has become a core tenet in a humanitarian strategy that has made immediate aid a tool in the service of political diagnosis and advocacy.
This humanitarian strategy has led to considerable investment in non-medical skills such as social work, legal and political expertise, as well as research and documentation skills for the purpose of diagnosis and advocacy. Yet in both organisations, even though to different degrees, these practices, and often the very existence of Mission France, have remained contentious. In particular, the non-substitution principle has proven difficult to maintain in practice. It has created conflicts between organisational strategists favouring time-limited interventions on the one side, and employees and volunteers on the ground favouring continued assistance on the other. The continued presence of Mission France is thus as much the result of political strategy as of the organisational dynamics it unclenched in local clinics throughout the French territory. Furthermore, in spite of the political clout and growing portfolio of activist and long-term engagement by both NGOs, 'doing politics' remains contentious as many humanitarian actors are protective of a core medical identity.
Our reconstruction of Mission France contributes to a growing literature on the tensions and contradictions within humanitarianism, whose protagonists constantly question the purpose of humanitarian action and the boundaries between humanitarianism and politics, between acute help and long-term development, or between medical and non-medical forms of humanitarian work (Barnett 2011;Redfield 2013;Krause 2014;Rambaud 2015). We make two main contributions to these debates. First, we shed new light on the symbolic and diagnostic function of humanitarian missions. Scholars have emphasised that witnessing and speaking out (témoignage) are central to MSF and MDM practice (Fassin 2008), yet also pointed out that witnessing is held to be secondary to the actual relief work (Redfield 2006, 10). 2 By contrast, the leaders of Mission France made its diagnostic function the core raison d'être, being utterly strategic about their symbolic impact. Second, the dynamics in Mission France clinics, whose personnel would not always comply with the strategy of time-bound interventions, highlight the importance of local staff's incentives and agency. This observation contributes to research on the importance of 'citizen aid' and grassroots humanitarian efforts (Fechter and Schwittay 2019), but stresses that these localised forms of humanitarianism do not necessarily operate separately from established humanitarian institutions. They can also change these institutions from within.
Given that the scholarship on MSF and MDM almost exclusively focuses on these organisations' international work, our study helps to understand a distinct configuration of their activities when they harness their international reputation in the domestic policy arena. Furthermore, many international humanitarian NGOs -not only MSF, MDM, but also Save the Children through their work in the uK, Partners in Health through their work with uS Native Nations, 3 or an entire international 'humanitarian aid machinery' (Haaland andWallevik 2019, 1869) present in European refugee camps -are increasingly present in their home countries in a context of multiple social and health crises. Our study contributes to understanding the importance of such reterritorialization for humanitarian identity especially for liberal international NGOs, which seek to establish coherence and purpose on the basis of a universal notion of humanity as a value beyond politics (Barnett 2020). MSF and MDM, while embodying a specific type of 'French universalism' (Taithe 2004), are also global reference points for the universalist, medical model of international humanitarianism (Krause 2014). Their struggle to navigate blurred boundaries between foreign and domestic sites of intervention, and between medical aid and political engagement, also points to the tensions and limits of liberal humanitarianism more generally.
The remainder of the paper is divided into three parts. The first section situates our argument in the humanitarian literature and introduces the case study and data. Next, we reconstruct the origins and justification of Mission France, which has provoked considerable resistance in the name of humanitarian identity, in spite of compassion with domestic populations. The third section shows how the French doctors strived to solve this dilemma by emphasising the symbolic weight of humanitarian missions, advocating for political change instead of becoming a substitute for universal health coverage. This strategy met its limits in the face of grassroots actors whose commitment to continued aid often ensured that local missions outlived their initial rationale. We conclude by reflecting on the implications of this study for humanitarianism in a post-covid world of debordered health and social policy crises.

Humanitarian politics in the welfare state
Humanitarianism, 'the desire to relieve the suffering of distant strangers' (Barnett 2009, 622), is constantly evolving. Its varied permutations include, but do not exhaust charitable and abolitionist movements in industrialising Europe (Haskell 1985), the co-emergence of domestic charity and international humanitarianism in late 19th century Britain (Roddy et al. 2015); European reconstruction efforts between the two World Wars (Baughan 2012;Hilton 2012), contemporary refugee aid around the globe (e.g. Lewis 2019; Lupieri 2020), faith-based international aid (e.g. Sadouni 2007), and the codification of Western international humanitarianism since the mid-20th century (Krause 2014). The diversity of humanitarian practices is reflected in a rich interdisciplinary literature, which has analysed the contentious boundaries and politics of humanitarianism from different angles. The distinction between humanitarianism, human rights advocacy and development (Barnett 2011), between emergency missions and longer-term engagements (Redfield 2013;Krause 2014), the forms of violence to be addressed by humanitarian NGOs (Bradley 2020), and the risk of erecting a medical government that takes over state functions (Debrix 1998;Falisse 2009) have all sparked intensive debates -both in scholarly writing and within humanitarian NGOs (Rambaud 2015).

Mission France: a contested reterritorialization of international humanitarianism
Our study of Mission France builds on this work on humanitarianism as a contested field of practice, and as a project whose specific local shape is always a configuration of concrete political, cultural, and organisational dynamics (Baughan 2012;Redfield 2013;Taithe 2019). By looking in-depth at the genealogy of Mission France, we move beyond the usual territory of studies of the 'French doctors' , MSF and MDM, given that most scholarly accounts focus on their overseas work (Taithe 2004;Vallaeys 2004;Redfield 2013). Only few studies mention, and even fewer explore in depth, their humanitarian work inside France. Among them is the historian Izambert, who in her study of MSF and MDM's engagement for migrant health in France, considers the NGOs' work as a form of domestic social activism (Izambert 2018b). Therefore, the international origin and humanitarian identity of these NGOs does not figure prominently in the study as Izambert's ambition is to trace the role played by MSF and MDM in extending healthcare access to undocumented immigrants. By contrast, the ethnographies of Fassin (2007; and Ticktin (2011) scrutinise how foreigners receive 'compassionate humanitarian' support by the French state. They highlight that state humanitarianism involves a problematic medicalisation and de-politicisation of suffering -for example when 'the right to asylum is replaced with humanitarian reason' (Fassin 2012, 141). In the words of Ticktin, this implies that 'caring' for vulnerable populations takes precedence over trying to 'cure' them by addressing the causes of their sufferings, thus contributing to maintaining inequalities (Ticktin 2011, 84).
Our interpretation of Mission France seeks to make sense of both dimensions -the rights advocacy highlighted by Izambert and the humanitarian 'aid' imaginary emphasised in the works of Fassin and Ticktin. We zoom in on the normative tensions of Mission France and show that it has been accompanied by controversies about the boundaries of humanitarian space and valuable humanitarian work, and that the desire not to become a substitute for welfare state services has given rise to a distinct strategy of demonstrative intervention, political advocacy, and socio-legal action. This humanitarian strategy not only differs from many activities in poorer countries, where international humanitarian NGOs are sometimes factually erecting medical government (Falisse 2009;Redfield 2013). It furthermore sheds new light on two key features of international humanitarianism: the role of diagnosis and witnessing, and the relationship between NGO leaders and actors on the ground.

Diagnosis versus aid, national strategy versus local engagement
First, the study of Mission France reveals that the diagnostic role of humanitarianism has taken centre stage in the domestic context. From the beginning, as the creation of French clinics by MSF and MDM was criticised to be disproportionate to the need on the ground, local actors started collecting data on the dire health conditions and experiences of exclusion of the populations they served. Diagnosis thus legitimises assistance, and making suffering visible by documenting it has become a steady companion to Mission France activities. This observation resonates with Fassin's emphasis on witnessing as a major strategy in humanitarian psychiatry (2008) and with Redfield's account of witnessing (témoignage) in international humanitarian projects (Redfield 2006). However, the diagnostic strategies of Mission France also differ from these accounts in two main respects. First, in Mission France, witnessing and documentation was not only a medicalizing and de-politicizing strategy like the representations analysed by Fassin ( 2007Fassin ( , 2008. It was also a tool of political advocacy that documented social policy failures in the name of equal rights -and thus informed explicit contestation of any sort of two-tier health care with a tier of 'humanitarian citizens' . Second, we show that over time, humanitarian diagnosis moved from secondary to primary importance. Instead of being a companion of assistance, the diagnostic component of domestic projects became a strategic priority taking precedence over direct assistance. The credo of not substituting state services led MDM, and even more so MSF, to make deliberate use of their symbolic power and the political effect of their sheer presence inside France as a demonstration of social policy failure. In a strategy of prioritising advocacy instead of permanent aid, local missions were instructed to remain temporary and only operate to achieve the effect of demonstrating need. The second feature that our study sheds light on is the relationship between professionalised and institutionalised humanitarianism on the one hand, and grassroots or 'everyday' humanitarianism on the other. Many studies emphasise that international humanitarian NGOs have become bureaucratised and professionalised and thus follow managerial or political strategies as much as a field-driven will to help (Krause 2014). By contrast, a growing literature on bottom-up, privately funded and citizen-driven aid emphasises a parallel world of humanitarian action that seeks to keep a distance from established actors such large NGOs, multilateral organisations, or government agencies (Taithe 2019; Fechter and Schwittay 2019). Our study shows that attributes of grassroots humanitarianism, in particular the emphasis on spontaneous and localised support (Lewis 2019), can also affect the practices of big NGOs. We show how employees and volunteers on the ground pursued priorities that diverged from top-down political strategies. Much of the longevity of local MSF and MDM clinics in France can be attributed to the tenacity of volunteers and employees who did not subscribe to the strategic and diagnostic function of their missions.

Method and data
The following sections reconstruct these dilemmas in more depth. We ground our analysis on primary and secondary data. They include archival materials and publications of MSF and MDM, key informant interviews with (former) heads of Mission France, and personal notes and memoirs by several protagonists.
Due to our interest in the historical emergence and evolution of Mission France and the ideological and organisational debates over domestic programs, we selected key informants mostly at the management and executive levels as our interview partners. 4 Our interview partners can be classified into three groups. The first two group are experts who were directly involved in the management of Mission France at the national or regional level, and who had not worked for the NGOs abroad. Their narrative emphasises the role of the French state and healthcare system, while they rarely refer to humanitarianism, which they tend to be critical of. The second group also were directly involved in the management of Mission France at the national or regional level, but who were former international humanitarians. . Despite their international experience, they rarely draw a parallel between humanitarianism at home and abroad. The last group of interviewees also consists in former international humanitarians who afterwards worked in France, but at the headquarters and not on the field. Even if they highlight the differences between humanitarianism at home and abroad, they are much more prone than the second group to reflect on the meaning of humanitarianism at home for humanitarianism in general.
In addition to interviews, we consulted the websites and archives of both organisations. The documentation service of MSF sent us the NGOs' annual reports and the minutes of the board of directors in digital format. As for MDM, Author 2 visited their library service in Paris, where the archives of Mission France (including annual reports for the years 1994-2000, project reports, internal communication and strategy papers, external studies, and advocacy documents) are stored. From MDM's website, we retrieved the most recent annual reports of the organisation (2008-2018) and annual reports on healthcare access and exclusion in France for the years 2009-2018.
Last, humanitarians produce considerable reflexive work on their practice (see Abu-Sada and 2012;Brauman 2006). Hence, we consulted the outputs of MSF's affiliated research centre CRASH (Research Centre on Humanitarian Actions and Knowledge), the journal Humanitaire, which was funded and edited by MDM until it ceased publication in 2015, and books and articles written by former members of the organisations.
In the following empirical sections, we strive to do justice to this rich corpus of data by providing long quotes from the interviews and the archives. Due to space restrictions, we could not do it for each of our empirical arguments and in those cases, we provide the corresponding quotes and their analysis in Supplementary Appendix 1c. Given that the aim of the empirical analysis is to allow a first theorisation of the meaning of humanitarianism at home for humanitarianism at large, the empirical analysis mostly focuses on the commonalities between MSF and MDM Mission France. When justified though, the differences between their respective Mission France are highlighted.

Debating the boundaries of humanitarian work
When in the second half of the 1980s, MSF and MDM decided to open domestic missions on French territory, this may have seemed like an obvious step: It was a reaction to domestic exclusion and dramatic health inequity. Yet this project met considerable resistance within the organisations as it challenged the dominant conception of these NGOs' missions as international aid providers. This section analyses the specific drivers of humanitarian engagement in their domestic setting and shows how the rationale for domestic interventions sparked heated debates about humanitarian identity.

A clear humanitarian cause?
The head of MDM's Mission France throughout the 1990s told us the 'mythical' origin of the program: Initially, at the very beginning of Mission France […], actually, the many fathers of Mission France say the same thing: they passed by, they were rue de la clef [name of a street], they saw this man who left the hospital and had an undried plaster cast, while it was raining cats and dogs, his plaster cast was disintegrating, and they were saying to themselves: 'but what is that? In our country, which is so rich, we let this lad outside under the rain with his plaster cast that is disintegrating, what is happening?' And they said to themselves: 'we cannot tolerate it, so we are going to do something' . (Interview 10) Hence the decision to start Mission France was driven by the outrage, if not the shame felt by MDM's doctors, who could not bear the fact that in such a rich country as France, homeless people were not properly cared for (but rather stigmatised, see Brücker, Pierquin, and Henry 1994). The quote above echoes a story told by Alain Deloche, one of the founders of MDM's France, in his memoirs. He notably recalls that his mother tirelessly asked him why he went so far away to help people, whereas there were also poor people in France (Deloche 1993). Similarly, the creation of MSF's Mission France was based on the observations that some people could not access the healthcare system and the willingness to address this situation that humanitarian doctors judged abnormal and unacceptable (Interview 3; Interview 7). Thus put, the classic humanitarian 'will to act' (Taithe 2004) provided an unambiguous impetus to become active inside France.
Beyond the humanitarian drive to act, there were also organisational and pragmatic reasons to opt for a domestic mission. First, professional leadership played an important role in both NGOs. At MDM, Alain Deloche was one of the founding members not only of Mission France but also of the NGO, and its president. In this endeavour, Deloche benefitted from the support from another founding members of MDM, Bernard Kouchner. At that time, Kouchner was starting a political career and managed to convince François Mitterrand to visit MDM's first clinic in Paris, thus putting to the fore the issue of healthcare exclusion and MDM's Mission France (Izambert 2018b, 96). As for MSF, the program was supported by Rony Brauman, then president of MSF France. Professional leadership continued to play a role in the development of Mission France since both MSF's and MDM's national programs mostly expanded throughout the 1990s when they were respectively led by the same persons for almost a decade, who were strongly attached to the programs.
Second, the problem identified by the humanitarians -unmet healthcare needs -was one of the symptoms of a bigger issue gaining prominence in the public debate in the late 1980s: exclusion (Brücker, Pierquin, and Henry 1994;Benamouzig 1998). As the director of MSF explained to us, the issue of new forms of poverty was given widespread media coverage following publication of the Wresinski report on poverty and precariousness in 1987 (Interview 3), 5 thus turning it into a public and a political issue, while this had not been the case beforehand (Interview 8). Given such public sentiment, it was straightforward to justify Missions France's establishment with reference to nearby distress.
The impulse to act was further undergirded by pressure from donors and heightened competition for private donations. For instance, Coluche, a popular comedian and entertainer, created the Restos du Coeur in 1985. This association provides free meals and food aid for poor people in France. It became increasingly difficult for the French doctors to justify their restrictive international focus, all the more that they started receiving letters from their donors in the 1980s urging them to pay attention to poverty in France and not only abroad (Izambert 2018a). It is, however, unclear to what extent fundraising played a significant role in the further development of Mission France as this aspect was not mentioned in our sources after this initial phase. In addition, given that the relationship between MSF and MDM had a competitive dimension, leaders of both NGOs felt propelled to prove their relevance by engaging in domestic sites. As Rony Brauman puts it: 'The fact that MDM started (a program in France) was not irrelevant in explaining our willingness to do it (to open a program in France)' (Interview 3).

Policing the boundaries of humanitarianism
However, from the outset the domestic engagement also met resistance within both NGOs. Criticisms challenged the value of Mission France from various angles.
First, many critics did not consider France a valid humanitarian site as the severity of the health needs of MF's target population was debated very controversially. Not all members of the two NGOs agreed that healthcare exclusion was really alarming in France. For instance, in the 1987-1988 annual report of MSF, one could read that: Stormy discussions and serious disputes had been required to reach an agreement within Doctors Without Borders about the indisputable fact: that some people are excluded from healthcare, in the country of social security. No need to talk about this debate, except to emphasize that the majority of us, focusing on far-away distresses, had been blind for a brief or longer period, to distresses that grew at the corner of our streets. (Brauman 1988, 8) Humanitarians' blindness to nearby sufferings might be explained by two factors. The first one is an idealisation of the French healthcare system when abroad, and, as a result, an impossibility to come to terms with its shortcomings. This idealisation is captured by the following quote from Bernard Kouchner, who is a founder of both MSF and MDM: From afar, faced with the bare essentials, misfortune, destitution, in poor countries, in the face of extreme difficulty, we dream of the smooth running of French hospitals, the efficiency of French medicine and the social protection system of our country, which an exotic experience of more than 20 years allows us to consider as one of the best in the world. The shortcomings of this support system are all the more unacceptable, especially for us, upon return from our faraway missions. (Kouchner 1987, 144) The second reason is related to this faith in the French healthcare system: When not abroad, humanitarian doctors worked as part of this public healthcare system, where they felt that they cared for all the patients coming to the hospitals (Interview 2). As explained to us by former MSF president Jean-Hervé Bradol, many MSF members realised that in the hospitals they worked in, the administrative staff barred access to people without proper documentation (Interview 2).
Yet even as the proponents of MF managed to render visible discriminatory practices and factual exclusion (Interview 7; Deschamps and Luciolli 1990; Médecins du Monde 1995, 6), staff at both MSF and MDM kept challenging the domestic engagement with arguments related to the identity of the NGOs. Both MSF and MDM were predominantly founded by physicians (Vallaeys 2004) and dominated by physicians as well (Redfield 2013;Krause 2014, 93). This medical identity shaped which kind of humanitarian work was considered valuable. In France, the NGOs set up activities both to address the vulnerabilities of their patient population and, as explained in more depth in Section 3, to make the state fulfil its responsibility. Yet beyond medical care, the NGO provided social and legal counselling to their patients. Indeed, healthcare was not the priority for most of the patients, some being homeless or undocumented. In those situations, curative medicine was of limited impact (Interview 2; Lasne 2010). In the words of Bradol: 'How can you treat the cancer or coronary syndrome of someone who lives on the street, who does not have money in the morning when they wake up to eat during the day' (Interview 2). Hence, even though until the late 1990s, the main aim of both NGOs was precisely to obtain an extension of the public healthcare coverage so that all French residents could access healthcare, they also had to address the specific needs of excluded populations. They provided their patients with care and social counselling to access social benefits (Interview 2) and developed programs for specific groups such as drug addicts and immigrants. This trend continued after the passing of the universal healthcare law in 1999, when MSF and MDM re-oriented their programs towards groups particularly suffering from socio-economic exclusion: migrants and Roma people, as well homeless people (Médecins du Monde 2002, 6; Neuman 2012).
Many members of MSF contested the value of these interventions as they overstretched the humanitarian-medical identity of the organisations: Did the organisations even have the necessary expertise to conduct such a program? Were there not already better qualified and experienced associations to address these issues, for example organisations working for migrant rights (Médecins Sans Frontières 1996; see also Escafré-Dublet 2014)? Consequently, the value of humanitarian work at home had never been fully acknowledged at MSF. The following quote from the former head of Mission France in the 1990s illustrates this point: So, there was really, at least that how I lived it, a mix of cultures, that is it (domestic programs) is not the dominant culture, we need to know it and accept it. And the day when things stop, when the people who support them leave, when there will be another executive at the head of MSF when other directions are taken, it is not the culture that will stay. It is certain. It is not the dominant culture. At some point, if there are priorities, if there is a war where MSF is able to significantly intervene, it would be sidelined, and it is normal. (Interview 5) Noëlle Lasne's retrospective impression was that Mission France did not fit with a dominant culture of international humanitarianism, and was of lesser importance in the NGO in comparison to international programs. The subsequent evolution of Mission France after her departure support this interpretation, as domestic humanitarianism remained precarious and contested within MSF (Interview 2;Neuman 2012).
But also at MDM, headquarters contemplated, during the initial development phase of Mission France, the possibility of transforming the program into a separate and new association given that the activities conducted by Mission France did not resemble international activities at the time (Gouriou 1993, 223). 6 This proposal to outsource Mission France did not materialise. Mission France expanded in both NGOs as they discovered the extent of poverty and exclusion in France. Yet, the persistence of their domestic engagement demanded a genuine rationale to counter the sceptics and their own doubts.

Diagnosis, aid, and the non-substitution principle
To reconcile the will to help with their humanitarian identity, the NGOs sought to adapt and re-articulate humanitarian reason for a domestic context. To redefine the boundaries of humanitarianism for the context of the French welfare state, the humanitarians increasingly invoked the principle of non-substitution as a legitimation and operational guidance. This idea of not releasing the state from its responsibility, and of using humanitarian action for activist purposes became the overriding ambition of Mission France. In the following we retrace the emergence of this principle and the frictions concerning its application.

The non-substitution principle
Non-substitution addressed many of the criticisms levelled against Mission France and legitimated the domestic interventions both internally and externally. The rationale underlying the non-substitution doctrine was that in a developed country like France, the state had the capacity to extend healthcare coverage so that the public healthcare system could provide for all: 'France (is) a rich, medicalized country, with public hospitals, doctors every fifty meters' , as a former head of MSF's MF highlighted (Lasne 2010). Therefore, the health needs of the target population of the domestic programs did not result from a lack of medical services, but from their exclusion from the healthcare system (Lasne 2010). This observation hints at the fact that the French doctors did not only demand the state to provide health for all because it could, but also because it should. With the non-substitution doctrine, the associations insisted that their role did not consist in alleviating the shortcomings of the state, but in asserting the health rights of their patients so that they could benefit from the same services as other citizens: 'in rich, xenophobic, discriminatory societies, it has never been so important to assert that the poor are not humanitarian citizens, but citizens like the others' (Médecins Sans Frontières 1995b, 5). While this quote reflects core humanitarian principles of universalism and equality, it also hints at a fundamental value, if not a 'political project' for Mission France (Neuman 2019), namely the defense of a universal public service (Interview 1; Interview 5). In other words, the non-substitution doctrine aimed at making the French healthcare system live up to the expectations of the humanitarian doctors.
Another important aspect of the non-substitution doctrine is that it requires the NGOs to identify and characterise the gaps in the healthcare system in order to, first, diagnose healthcare exclusion, and second, formulate precise demands to the state. For that purpose, the NGOs have collected socio-medical data on their patients and their stories since the inception of Mission France. The data and the stories were then communicated through press releases, reports and books (Deloche 1993, 331;Deschamps and Luciolli 1990) and MDM even institutionalised these data collection efforts through the publication of a yearly report on health exclusion in France (Médecins du Monde 2020). As a result, the members of Mission France managed to make their patients visible in the eyes of fellow humanitarians and the French government and healthcare institutions. In addition, the data collection efforts were not only used as an internal justification tool, but also as a diagnostic instrument. Indeed, the quantitative data coupled with personal stories of the dramatic consequences of negative life changes and experiences of discrimination enabled the members of Mission France to document and campaign around the extent and mechanisms of healthcare exclusion (Deloche 1993;Deschamps and Luciolli 1990;Interview 7;Médecins du Monde 1995), and thus to precisely identify the gaps in the healthcare system -and ask for them to be filled.
Last, non-substitution implies that on the field the missions should be of limited scope. Initially, the NGOs indeed chose to set up small and short-term missions because they expected the state to react quickly, enabling them to close Mission France after a few months (Interview 3; Mamou 1998). Moreover, the activities carried out in the field were designed to contribute to achieving this goal. Beyond providing immediate aid, these projects generated data about exclusion, and local staff tried to help those excluded to get access to services through socio-legal support (Interview 5, see also subsection 3.2). Hence domestic missions were not thought to have a lasting impact on the operations of the NGOs, and their identity as well.
This strategy of non-substitution led to a strong emphasis on diagnosis, not only as a complement to humanitarian aid as in most international missions (Redfield 2006), but as its primary aim. In the case of Mission France, the NGOs are willing to diagnose a problem only, but then expect the state to solve it, whereas abroad they are sometimes ready to assume both functions, giving only residual role for the state (Interview 8; Redfield 2013). In Chad, for instance, MSF Belgium did not provide healthcare, but tried to fully reorganise the healthcare system and infrastructure in a country affected by years of civil war and drought (Falisse 2009, 41-44): Instead of being solely an emergency medical organisation, it almost assumed the role of the 'Health Ministry' (Falisse 2009, 42). In this regard, NGOs have much less expectations from the host states abroad (Interview 8), where a life as a 'mere humanitarian citizen' seems apparently much more imaginable.
Nevertheless, the plan for a short-term and small-scale mission never materialised, not only because the state did not react as soon and as much as expected, but also because non-substitution clashed with contradictory humanitarian principles.

The politics of using aid strategically
The strategy for Mission France has created tensions for both NGOs, which sought to reconcile a targeted and demonstrative deployment of domestic missions with staff's desire to sustain aid on the one hand, and with their medical and apolitical humanitarian image on the other.
First, a problem encountered by both NGOs was whether and how to limit the duration of local assistance in order to exert political pressure. Recurrent questions dwelled with by Mission France leaders were: If NGOs step in, do they let the state off the hook? But can you decide not to act in the face of apparent need and inequality? The NGOs responded differently to this dilemma. MDM adopted a relaxed line when it came to interpreting nonsubstitution. Thanks to the mobilisation of MDM members at the local level -who were mostly volunteers -, a dozen of Mission France programs opened throughout the countries just in few years after the opening of the first clinic in Paris (Gouriou 1993). The engagement of these volunteers was valued within the organisation and the domestic presence of MDM gradually became part of its identity (Interview 10; Izambert 2018a). While acknowledging the risk of becoming a long-term substitute for welfare state services, Jean-François Corty explained to us, the NGO strived to complement its aid with witnessing and political advocacy: The risk, the key question we face in our associations, is the risk of being a substitute for the state. But at some point, if the state is doing nothing, and that you, you say: 'ah, no, I am not acting because it is substitution' , then: what happens to the people in the field? And then, to solve that dilemma, it seems to me that the answer is: we act, we treat people because there is a need, but we collect information about what we see, what we do. And we do advocacy to make the public authorities do what they were supposed to do. (Interview 4) By contrast, MSF sought to enforce a tighter interpretation of the non-substitution principle by drastically limiting the number of programs. The headquarters indeed sought to restrict the expansion of the programs opened by local sections in Marseille, Lyon, and Lille (Interview 7). An episode in the mid-1990s illustrates the tensions this could produce. The head of MSF's Mission France decided to stop the program in Lyon because the local authorities and another NGO agreed to take over MSF's program. She also argued that closing the mission would force the authorities to act more quickly (Médecins Sans Frontières 1995a). Yet, the Lyon volunteers strongly opposed the closure, holding that the alternative structures were not functional yet. They warned that, if the program would close, patients would be abandoned, thus violating the relationship of trust they had managed to build with the patients. The Lyon team also highlighted that the patients received more humane care in their program than in public hospitals. Even though many members of MSF belonging to both the board of directors and the Lyon team disapproved of the top down decision-making process and expressed concerns regarding care for their patients, the program eventually closed. In other instances, this rationale was criticised for being too dogmatic and thus preventing MSF from intervening despite urgent needs (Emmanuelli 2005, 88-89;Médecins Sans Frontières 1993). Yet the strategy to limit aid in the name of political advocacy was maintained.
What is true of both NGOs, however, is that a more strategic and time-limited use of aid faced opposition especially from local actors, the staff and volunteers at local clinics or support centres. These had their own commitments and resources in their communities resembled the grassroots resources typical of smaller and privately-run aid organisations known from the literature on citizen aid (Paugam et al. 1997;Fechter and Schwittay 2019). While both NGOs had mechanisms such as steering groups in place to contain the local autonomy of their domestic projects (see Médecins du Monde 1993), grassroots agency did play a considerable role in initiating local projects as well as sustaining them way beyond their original rationale (Izambert 2018a).
Second, to keep domestic missions temporary and exert pressure on the state involves that the NGOs need to engage in political activism. This creates tensions with another humanitarianism tenet: the principle of political neutrality. Distance from the state is often asserted to be a core principle of international humanitarianism, considered to allow NGOs to focus on health and survival within a zone of neutrality. Evidently, this principle has created problems and tensions throughout humanitarian history. The gap between the image of apolitical lifesavers of humanitarian NGOs, on the one hand, and their speaking out activities, on the other hand, is often debated within the humanitarian community (Barnett 2011;Redfield 2013). A case of high symbolic importance for this stance towards politics and neutrality has been MSF's public denunciation of the manipulation of aid during the Ethiopian famine in the mid-1980s, which led to its expulsion from Ethiopia. While MSF's leadership justified its témoignage on the ground that 'remaining silent in the face of such a tragedy is to become complicit in it' (Davey 2015, 226), most other NGOs on the field condemned MSF's stance. For them, the priority and only concern was to provide relief (Davey 2015, 226).
Mission France has been no less political. Notably, though, the NGOs' engagement with the authorities did not only take the form of public denunciation, but also more complex forms of lobbying and cooperation including the acceptance of public funding in the case of MDM. A former head of MDM's Mission France, Nathalie Simonnot, explained to us that she had to fight to impose that Mission France would accept public funding. Contrary to her colleagues, she did not see public money as a threat to the NGO's independence. She rather perceived it as a way to both develop the program and force the public authorities into taking an interest in the issue at stake. As she puts it: 'we have a stronger strike force when we are financed (by public money) because we also have discussion forums' (Interview 9).
While MSF renounced public funding in the name of political independence, it closely engaged with policymakers. For instance, during the drafting of the universal healthcare coverage law, MSF met regularly with the rapporteur in charge of drafting the law and edited a newspaper sent to associations, healthcare authorities, the social security, the government and the legislative, in which the NGO highlighted the gaps in the healthcare system and made recommendations on how to address them (Interview 5). Even though head of Mission France Noëlle Lasne asserted that, at that time, they voluntarily 'did politics' to reach a specific goal, namely to advance the right to healthcare, her team also felt it had way too much political influence, while only being an NGO (Interview 5). This concern echoes MSF's long-standing self-understanding as a 'lone ranger' in the international domain, which the NGO seeks to maintain even when it closely collaborates with policy makers and ministries of health (Healy, Aneja, and DuBois 2020).
Finally, 'doing politics' also implies, for both NGOs, to invest considerably in non-medical activities such as socio-legal counselling. Given the important self-understanding of humanitarians as engaging in emergency aid and thus taking a 'minimal approach' to care, social work is often considered an unaffordable 'luxury' (Interview 2). In France, however, providing care became routinely accompanied, if not superseded, by social counselling to help vulnerable people access state service. For some, it was an asset of Mission France that it demonstrated the importance of social work and pushed the NGOs to widen the scope of the services they provided (Interview 2). Nevertheless, this was a contested process and it is still incomplete. For instance, in the early 2000s, MSF Mission France opened a sheltering program to host and provide care to severely ill homeless people (Médecins Sans Frontières 2004). The program, however, suffered from staunch opposition from the executive board, which argued it was too much oriented towards social support and did not fit the core identity of MSF, namely humanitarian medicine in war zones. It was eventually shut down (Interview 2).
These tensions show how the protagonists of Mission France have struggled to maintain a neutral, medical image, while at the same time becoming deeply involved in social policy. De facto, the boundaries between humanitarian-medical aid, social support, and political engagement have become blurry. Even though contested, new layers of humanitarian activity such as socio-legal counselling and lawmaking have been added to what is considered humanitarian. Nevertheless, continued reference to humanitarian neutrality and the NGOs' medical image had their role in making sure that their domestic engagement had the intended symbolic weight. Having MSF and MDM work on domestic health problems signals urgency and social policy failure, partly because these NGOs symbolise a commitment to human life and dignity that is beyond politics.
Medicine and neutrality continue to play a key role in Mission France today: 'during the discussion over the re-opening of the mission at MSF in 2015 in the midst of the refugee crisis, the fear of "doing politics" was a core concern and many wondered whether MSF could be of any help given that most of the refugees were not sick' (Neuman 2019). 7 At MDM too, medicine remains the encompassing frame of reference for its actions: The NGO, for instance, justifies its advocacy work in favour of the housing rights of Roma population by explaining that evictions have negative health effects (Interview 6). In the midst of political struggles about social policy and equity, medical universalism might not provide operational guidance for humanitarian aid, but serves as a frame of reference for social critique.

Conclusion
The political and geographic boundaries of humanitarianism are constantly evolving. Visions of neutrality and a prerogative of medicine must be reconciled with the reality of conflicts and protracted health challenges in many parts of the world; and in each context, humanitarian organisations find genuine settlements between varied moral economies of aid. One site of such settlement is the domestic engagement of humanitarian NGOs that are best known for their international work. As we have shown through a historical analysis of the case of MDM's and MSF's venturing into French health policy, such reterritorialization has demanded considerable rethinking of their purpose and identity as liberal and universalist humanitarians.
They have used their symbolic capital as international life savers to put pressure on the French government, in a strategy that made conscious use of the symbolic impact of their engagement on French territory. This prominence of the diagnostic and symbolic function of their work went beyond the reigning self-conception that immediate aid is the priority, and witnessing a complement to it. It created challenges because local volunteers not always endorsed the strategic restraint of aid in the name of non-substitution and political pressure, and because it required considerable investment in social work, politics and lawmaking, activities which many members of MDM, and even more of MSF, saw as a threat to the non-political identity of their NGOs.
The move towards advocacy and social engagement in Mission France prefigured, and in some cases influenced broader developments in the NGOs programs. Beginning in the 1980s, and ever more systematically in the 1990s, the NGOs went beyond emergency interventions: MDM launched risk reduction programs -first experimented in France -in Burma and Vietnam in the late 1990s (Martin 2020). A decade later, Bradol pushed for the provision of social support to patients in MSF reconstructive surgery program in Amman because he had discovered the value of social work through Mission France (Interview 2). Yet, wherever humanitarians do rally for social justice and health policy, they must navigate the tensions between providing emergency support and asking states to ensure health equity. Our analysis suggests that the response to such tensions hinges not only on the strategies of NGO leaders, but also on the resources and commitment of local employees and volunteers. Where NGO members have local control of their missions, they may raise different claims on the state than they would as foreign interveners. Across the globe, the organisational set-up, domestic ownership, and reporting lines of local humanitarian chapters vary considerably. While in some countries, international NGOs only open field offices to carry out a specific program, in others, they support or allow the creation of proper chapter by former local staff for fundraising or/and project management purpose. Yet in some cases, field offices do transform into local chapters, for example in the case of MSF South Africa (Fox 2014, 185-198).
Writing this conclusion in a historical moment -the ongoing Covid-19 pandemic -one may be tempted to assume that questions of bare life, live-or-death triage decisions are what ultimately matters in health policy around the globe. Yet, the structures that make certain people more vulnerable than others -for example people in refugee camps, overcrowded prisons, or broken healthcare systems more generally -are building up over time. Political neutrality -a principle in high esteem in situations of war and physical violence -hits its limits where more subtle forms of violence are at play. In a time of debordered health and economic crisis, as citizens and NGOs look for new ways to counter humanitarian emergencies in all parts of the world, including industrialised countries, we will see further difficult compromises between the impulse to help and the wish to hold governments to account.