MSF office

Management at MSF - Part 2

Portrait of Marion Péchayre

MSF-Crash : Team coordinator

Anthropologist specialised in international relief issues. Marion Péchayre has a previous practical experience having endorsed management responsibilities both in the field and at the head office of Solidarités International. She holds a PhD in Development Studies (SOAS) and an MA in International Management (ESCP Europe). She teaches at the Paris School of International Affairs and at the Humanitarian and Conflict Response Institute. Her main publications are: "Politics, Rhetoric, and Practice of Humanitarian Action in Pakistan" (in "The Golden Fleece: Manipulation and Independence in Humanitarian Action" Boulder, Colorado: Kumarian Press, 2012), and "Impartialité et Pratiques de Triage En Milieu Humanitaire. Le Cas de Médecins Sans Frontières Au Pakistan" (in "La Médecine Du Tri. Histoire, Éthique, Anthropologie" PUF, 2014).

Elba Rahmouni

Since April 2018, Elba has been in charge of dissemination at CRASH. Elba holds a Master's degree in History of Classical Philosophy and a Master's degree in editorial consulting and digital knowledge management. During her studies, she worked on moral philosophy issues and was particularly interested in the practical necessity and the moral, legal and political prohibition of lying in Kant's philosophy.

In our first interview, Marion Péchayre discussed a variety of management-related problems at Médecins Sans Frontières: working in silos, the ever-growing number of management tools for monitoring, the endless validation requests, the vanishing role of the individual in favour of pseudoscientific presentations of events and projects, etc. This interview, conducted by Elba Rahmouni, focuses on solutions to these problems and hypotheses on how to improve our ways of working. Rather than offering set recipes, Marion Péchayre advocates an approach based on “practical wisdom” (a concept from the sociology of professions) and deliberative management that every individual and team can apply in their own way.  

0. Introductory remark

In our first interview you said that you wanted to change our ways of working, rather than our structure; can you explain why?

Over the past few years, MSF top management has already been interested in changing our structure: making a reform aimed at empowering the operational managers of the Operations Department, switching from desks to cells, creating new departments (the purchasing and supply department in 2014), clarifying the levels of responsibility and encouraging greater autonomy. Despite these efforts, the siloing problem has only gotten worse since the 1990s. So I said to myself that before tackling organisational chart reforms, we might have to start by thinking in a broader sense about how we work. 

When talking with the participants of a FOOTField Oriented Operational Training is the training course for operational managers.session, both international and national staff, from headquarters and the field, I realised that most of them didn’t want another structural reform giving them more autonomy. They were afraid that “decentralisation” would leave them even more on their own in doing their jobs. What most of them wanted was to participate more in discussing the strategic directions of their projects and support in dealing with their difficulties. So it was less the organisational chart, as such, that was a problem for them than the day-to-day working relationships between departments and with headquarters. 

1. Practical wisdom

You talk about “adapting our management to the unique aspects of our humanitarian work”. What is it about our work that’s unique? 

The heart of our work is providing (sometimes emergency) medical care in resource-limited countries, in dynamic – often volatile – situations. The work has a high degree of uncertainty, comparable in many regards to medical work. 
Sociological research on the medical professions shows that doctors do not rely solely on scientific reasoning when they have to make treatment decisions, to choose the appropriate treatment for each individual patient and care situation. While scientific knowledge is a prerequisite to care, every caregiver also uses their own practical experience and, if necessary, discussions with peers. Sociologist Florent Champy calls this type of reasoning – which mobilises theoretical knowledge (science), practical knowledge (guidelines and procedures), and professional experience (both one’s own and that of one’s peers) – “practical wisdom”. Accordingly, medical decision-making is like gambling on treatment objectives (to alleviate, cure, prevent, reassure, etc.) and treatment tools. A gamble whose outcome cannot, by definition, be 100% guaranteed. 

Aid actors are in a similar situation. Every situation we work in involves some degree of uncertainty regarding the goals of the action (who to help first, for example), the means to employ, and the results of the action, because we cannot anticipate how the patients, the authorities, and our colleagues are going to react. Our decisions and choices are gambles that we must constantly re-assess and justify. While mastery of technical/scientific knowledge and know-how (medical, in particular) is essential, it isn’t everything. The ability to call upon past experience and use inductive reasoning – i.e., what is similar to or different than previous situations – and intuition is equally important. 

Yet we live in a world that places too much value on scientific reasoning. Generally, when we make decisions, we advance seemingly scientific – usually quantitative – arguments. The decisions often appear to be the only possible option, masking the amount of uncertainty and the fact that we always have a choice. 

Rendering the role of experience and professional judgement invisible in our work, however, deprives us of group discussions and the opportunity to compare this experience with other experiences and the experiences of our colleagues. The danger in not seeing that our decisions are more an informed gamble than a logical scientific conclusion is that we believe that our decisions are final and “right”, and must be adhered to at all costs. 

2. Discussion-based management

What are deliberative spaces? 

Deliberation is one way to ensure that our operational gambles are as informed, and free of risk, as possible. It is also a way to manage the irreducible gap between what the organisation “prescribes” (procedures, standards. etc.) and the “reality” of the work (procedures sometimes have to be adapted, standards cannot always be met, the unexpected happens, etc.) Deliberative spaces allow us to discuss such gaps and how to manage them. Anything that can happen routinely, as expected, and without interfering with others does not need, perhaps, to be discussed; the rest, however – the unusual, the problematic – does. And meetings should be devoted primarily to that, rather than to reporting in detail what we have done – something that can be put down in writing somewhere for better coordination. 

Professional disagreements among colleagues are a key element in deliberative spaces, because explanation and discussion of such differences is precisely where temporary new rules come from. It is not a question of seeking consensus at any cost, but of hearing multiple viewpoints and arguments that the team leader, or the person who brought the issue to their colleagues, can use to make a decision. Using this approach, the group can decide, together, to deviate from certain rules without giving everyone carte blanche to make new rules on their own, in their corner.

What discussion-based management is NOT

  • A revolution. Discussing professional experiences and opinions and making decisions in uncertainty are practices that already exist. But it would be good to value them more and devote more time to them.
  • More meetings. It does not necessarily mean having more meetings, but altering the content of existing ones. Now, team meetings are often used for reporting and coordinating around planning. In deliberative spaces, teams buckle down and discuss the difficulties, and more specifically the gap between what had been planned or what the protocol says (prescribed work) and what is really happening (real work).
  • The end of objectives. While there is a difference between results-based management and discussion-based management, discussion-based management does not mean that objectives which are essential for setting a common path and getting organised disappear. But in discussion-based management, objectives and planning are not the only topic on the agenda discussions are guided by what is frustrating our plans, even if that means revisiting the objectives and, in some cases, revising them.
  • Anarchy. While rules and procedures need to be adapted sometimes, that does not mean letting everyone decide to do whatever they want. Deviations from the rules are regulated by group discussions with co-workers or peers. Individuals wishing to deviate from the rules must submit their idea for discussion so that their colleagues can assess the possibility and the attendant risks from their own standpoint, thus sharing the responsibility for not following procedure.

3. The subsidiarity principle

Who should decide? Must we move toward less centralisation?

Right now it seems to me that the issue is less about centralised or decentralised decision-making than about coproduction, or having several people involved in discussions prior to a decision. There should be a variety of deliberative spaces: some between people with different jobs at a given hierarchical level, and others between the different hierarchical levels (field, capital, and headquarters).

The subsidiarity principle holds that people should refuse to make a decision that someone at a lower level could make. That rule of thumb does not always tell people what is, or is not, their responsibility.  I think it’s a mistake to think that guidelines, however exhaustive, can address that problem. I think we need a few major rules, in writing, on each person’s responsibilities (annual budget validation by project, responsibilities related to purchasing procedures, etc.) to provide a clear, simple framework. But for the rest, people have to be able to regulate as they go along, consulting their superiors or colleagues, asking first for their opinion – not their approval. It seems to me that this would help people better understand the constraints faced by their colleagues, de-compartmentalise the work, and ultimately lead to jointly-constructed decisions. 

4. Listening to everyone, or only the strongest?

Is discussion-based management the most inclusive possible way to make decisions, or a system that encourages and legitimises individuals with the best public speaking and argumentation skills? 

I prefer an organisation that values an exchange of ideas and viewpoints to one that encourages individuals to follow and enforce rules. 

That being said, stating an opinion is an action by which a person exposes himself – taking a not always obvious risk (that of disagreeing with others). Everyone needs time to find their feet on a team, to get to know the others well enough and learn how to express their point of view. If such spaces were routine, people could practice and get accustomed to talking things over. 

In my opinion, it’s important for managers to share their doubts, rather than act like the authority-who-knows, and set an example by presenting their choices not as truths but as hypotheses. MSF should also offer support (via training courses or mentoring) to help all members develop their argumentation skills.  

5. Personal satisfaction and social mission

Is the goal of working this way to make work more satisfying to individuals, or to better accomplish our social mission? 

We became interested in ways of working because we noticed that bureaucratic rigidity was slowing our work and weighing it down. There were also complaints and frustrations at various levels; operational managers felt that the “support functions” were taking on more of a monitoring than a support role, and at the same time, the support departments were complaining that they were not being considered enough or included in discussions about operational choices. 

Several studiesMathieu Detchessahar, ‘Faire Face Aux Risques Psycho-Sociaux: Quelques Éléments d’un Management par la Discussion’, Négociations, no. 1 (2013): 57–80; Christophe Dejours, Le Choix: Souffrir Au Travail n’est Pas Une Fatalité (Bayard Paris, 2015); and Yves Clot, Travail et pouvoir d’agir, Travail humain (Paris: PUF, 2008).have shown that deliberative spaces not only help avoid the kinds of friction that occur when people are not consulted beforehand (and the workload involved in managing that friction), but they also, most importantly, give each person control again over the meaning of their work, inasmuch as everyone’s opinions counts. 

One might think that discussion-based management is too time-consuming, given that our operations often require quick action. I would say first that it’s obviously not a question of discussing everything. Next, in my opinion, in the end it takes less time to discuss problems than to solve the problems caused by a lack of discussion. According to the psychiatrist and work psychologist Christophe Dejours, we live in a society made ill from overwork. He claims that some of that overload can be directly blamed on the fragmentation of work – in particular, when each person has to spend a substantial amount of time reconstructing information already possessed by others.   

Hence our hypothesis (which is perhaps naïve, and requires testing) is that discussion-based management improves both personal satisfaction and the overall effectiveness of the organisation.  

6. The managers 

How does a “good” manager go about implementing deliberative management?

Managers or team leaders play an essential role. First, they have to be able to identify important subjects to discuss as a team, like “conflicting standards” between departments and other difficulties encountered when conducting activities. To do this, they must take an interest in the real, day-to-day work of their team members, and in the problems that each of them faces: for example, filling a position for which the compensation is determined by pay scales out of sync with the labour market; following a medical department recommendation that deviates from national policy or requires additional resources; sending accounting data or medical statistics in a timely manner despite computer problems; or following conflicting orders such as to reduce the exposure to insecurity while maintaining medical quality; etc. 

Unlike in the business world, where management professionals often have little understanding of employees’ real jobs, MSF managers generally know the job of the people they supervise. Many have done it themselves, as logisticians, administrators, caregivers, activity managers, field coordinators, etc. However, they are not always familiar with the real work of their colleagues in other departments. I think we have to encourage them to hone their knowledge of “humanitarian work”, in all its dimensions, even if that means doing “micro-internships” with colleagues in other departments (for example, having the medical coordinator spend several days with the Human Resources coordinator, who explains the issues and constraints of the job, and vice-versa). 

In discussion-based management, the manager has to pay particular attention to formalisation. Formalisation makes it possible to:

  • Check whether what was said verbally was understood correctly. It is useful to write to the participants after a meeting to make sure that everyone agrees on what was said; this is not always obvious.
  • Continue to reflect. Having to put things in writing helps clarify ideas. Sometimes people don’t realise that their arguments aren’t as obvious as they seem, or aren’t sufficiently supported, until they are put in writing.
  • Make information available beyond just the people who participated. Once transcribed, discussions can be archived and distributed to anyone who might be interested.
  • Keep a record for the future. Given that our choices are gambles and not self-evident “correct” decisions, we have to see what they produce and in some cases adjust them.  This requires additional discussion, for which it is good (even necessary) to share what was already said. Having a record of the past – which permits re-evaluation – is therefore very important. 

7. The institution 

What has been done to develop discussion-based management in-house?

We developed the practical aspects of this approach when revising our team management training course. The Field Management Training (FMT) – that we renamed WoW (Ways of Working) – is now anchored in this self-critical examination of our practices. There is a new Learning & Development policy for MSF France employees, and we're beginning to see its concrete manifestations: for example, the reformatted individual end-of-mission assessment, which looks at what individuals have done and learned, and checks whether they took an interest in their colleagues’ and team members’ work constraints, rather than judging, in a normative way, their skill level in a series of areas. 
When we introduce this approach, particularly in trainings, I sense that people are drawn to it, overall. Fortunately, some did not need our advices and were already working this way, and I hope that gives them some arguments for convincing their colleagues and bosses, if necessary. 

As an example, several people involved in Île-de-France Covid-19 response told me how exciting the team dynamic at their project had been, despite all the obstacles they encountered – precisely because they had set aside daily discussion time for sharing their questions and discomfort, sometimes for airing their differences, and ultimately for fine-tuning what they were doing. Listening to them, I thought how these exchanges seemed to have brought them together by creating, from day to day, a shared narrative around the goal of helping those most marginalised by the system in France. In that case the team was able to compensate for the perverse effect of working in silos. Several of them did, however, comment on the lack of discussion spaces with higher-ups, with the result that vertical fragmentation compromised the quality of their work. 

Any team leader can implement this kind of approach at his level with his team, independent of what’s happening higher up; it gives individuals an opportunity to influence decisions, and even leave their mark on certain projects. Yet the only way to institutionalise this practice is to adopt it at every level, so that team leaders set an example – including at the very top of the organisation.

8. In closing… 

On a personal note, this reflection process gave me the urge to return to the field to work directly with the teams. So I am leaving on mission again, for two years, hoping to put this deliberative management approach to the test, on the theory that it is the best way to work, and that the results should be observed and – especially – discussed.

To cite this content :
Marion Péchayre, Elba Rahmouni, “Management at MSF - Part 2”, 20 juillet 2020, URL :

If you want to criticize or develop this content, you can find us on twitter or directly on our site.




Ou peut on trouver la première partie?


Add new comment

Cette question consiste à tester si vous êtes ou non un visiteur humain et à éviter les demandes automatisées de spam.