“The ‘refus d’obtempérer’ is seen as a challenge to the authority of the State”: interview with Sebastian Roché
Sebastian Roché, a sociologist and research director at CNRS, and co-author of “La police contre la rue” (Grasset), is among the few experts who were consulted by the French Assemblée Nationale during the drafting of the report on the increase in police shootings in so-called “failure to comply” (‘refus d’obtempérer’) situations. By highlighting both the efforts and the shortcomings of this parliamentary work, he provides an overview of the current state of the regulation of state violence.
INDEX: On 29 May 2024, the Commission des Lois de l’Assemblée Nationale published a report by MPs Thomas Rudigoz (Ensemble, Macron’s party) and Roger Vicot (Socialist party) on the increase in police shootings during ‘failures to comply’ occurrences. While the two rapporteurs did note an increase in shootings, they did not identify the reason for this trend in the reform of self-defence introduced by article L. 435-1 of the Internal Security Code, voted in 2017, which organisations such as the Human Rights League have described as a “licence to kill” for police officers. You were interviewed by the rapporteurs as part of this mission. Following its publication, what do you think of this report?
Sebastian Roché : To date, this is the most comprehensive analysis of the subject published by the French State. A great deal of work has gone into describing the legal framework, the instructions and circulars sent out by the Directors General of the police and gendarmerie, and the case law, including that of the Criminal Division of the Cour de Cassation. But this work of describing the legislation is accompanied by a refusal to note that the legislation has been transformed by the law of February 2017 – the one that introduced the idea that police officers can interpret the future behaviour of people who do not comply with an arrest order and deem it threatening to the point of using their weapon.
The article that is particularly problematic is L. 435-1 of the French Internal Security Code. This article allows police officers to shoot people who are “likely to cause harm to the life or physical integrity of others in the course of their escape”. This means that the use of lethal force becomes authorised on people who, even if they do not pose a direct and immediate threat to others, could perhaps do so, or are likely to do so. With this article, we lose the notion of imminent danger – which is necessary for the application of the classic self-defence framework. As it stands, the article does not specify what objective factors police officers should rely on to determine whether a fleeing person could be considered potentially dangerous in the future.
These essential elements were changed by the 2017 legislation. But the rapporteurs do not recognise this transformation. So, given the recent increase in the number of fatal shootings, they point to other factors and areas where action is needed, and in particular – and this is a recurring theme – the training of police officers, both initial and ongoing.
INDEX : You are one of a small number of researchers who have been working on these issues for a long time. Do you think your arguments were taken into account by the rapporteurs? What did you tell them?
SR: The two rapporteurs listened carefully to what I told them. On the other hand, they considered that what I told them was of no value in analysing the problem that was at the heart of their work. This was apparent in two ways. Firstly, on the figures for refusals to comply, and secondly, on the role of the law approved in 2017.
On the first point, one needs to bear in mind that in the case of resisting arrest, there are no easily identifiable physical traces of the act. In the case of a burglary, you can see traces of a break-in. When you have a homicide, you have a body, and a doctor can establish the cause of death. In the case of refusal or failure to comply, there is only an account, a subjective interpretation of a situation by an officer. This is even truer when it comes to classifying a refusal to comply as “serious”, because a serious refusal to comply is one that presents a risk – but a risk is a probability that something will happen. It is therefore the officer who subjectively interprets the risk, and its possible, but not proven, seriousness.
I pointed out to the rapporteurs that with such a method, their study was bound to be based on unreliable information, mainly because of the subjectivity inherent in the source of the data. Subjectivity depends on the personal characteristics of the officer, the sensitivity of the society which changes over time, and also on the instructions and guidelines given by the hierarchy (for example, to focus on these “facts”, to consider them as “serious”).
What I’m saying is that we don’t have precise measurement tools for this phenomenon. The rapporteurs felt that I was putting the increase in refusals to comply into perspective. To do this, they relied on statements made by people they met during a study trip to Marseille. Imagine the scene: they go to Marseilles, they ask their interlocutors: “So, how are refusals to comply going?”, and the police and magistrates tell them – as usual – that there are more and more of them, that the situation is very serious. They are also told that not all refusals to comply are recorded. So the rapporteurs conclude that, in reality, the risk is underestimated by the official figures. What I am arguing, based on an analysis of the sources and quality of the data, is that the source is unreliable and that it is therefore difficult to identify robust trends.
On the role of the 2017 law, the rapporteurs take up the broad lines of reasoning that I presented to them and acknowledge that indeed, if we look at both sides of March 2017 – since the law dates from 28 February 2017 – we see a significant increase in the number of fatal shootings. But once that’s said, they argue that you can’t really interpret the difference in the frequency of fatal shootings between before and after March 2017. According to them, 2017 was an unusual year, and therefore not significant. But it is, of course, particularly significant, because it is the year in which the law is changed, and immediately after this change the number of fatal police shootings soars! From a methodological point of view, you can’t suppress data that is embarrassing and claim to be rigorous.
The statistical study that we published in September 2022 in the journal Esprit on trends in fatal shootings after 2017 [with Paul Le Derff of Université de Lille and Simon Varaine of Université Grenoble Alpes, ed.] had a precise methodology, which the rapporteurs did not understand. In this study, we compared the two periods before and after 2017 for the police with the two periods before and after 2017 for the gendarmerie. This difference-in-difference approach is used, for example, in medicine to assess the effectiveness of drug treatments. The rapporteurs did not really see that this was a causality test. They simply said that it was at best a correlation, with no causal effects. Added to this is the fact that they insist on discarding the year that interferes with their reasoning. In the end, all this means that the effect of the 2017 legislation is not recognised.
INDEX : If on the one hand MPs admit that there is a problem with the increase in fatal shootings during refusals to obey, but on the other they refuse to acknowledge the effect of the 2017 law, then what exactly are they advocating?
SR: There is disagreement between them on this point. The Socialist party rapporteur Roger Vicot writes in the report that, even if we can’t really know with absolute certainty whether this law has had an effect on the number of fatal shootings, there is still some doubt. So it would be a good idea to remove one word from the law, the one that introduces the idea that police officers and gendarmes should guess at the risks likely to occur in the future. The idea would therefore be to remove from the law this notion of the “susceptibility” of harm, because the police have no way of knowing whether or not an individual might commit harm to others while fleeing.
The other rapporteur, Thomas Rudigoz, from the Ensemble group, argues that the law should not be rewritten, as it has already been approved.
In fact, most of the report’s recommendations concern officer training. For the rapporteurs, if the problem is not the legislation, then it must be the training – and so they focus upon initial and in-service training in the use of weapons.
On this issue, they observe that within the police there is a huge training deficit. Police officers are supposed to fire three rounds of ammunition a certain number of times in the course of the year, but only 60% or 70% of them do so, and the commission was unable to obtain precise information from the National Police Headquarters about the profiles of officers who actually train. They point to a major problem: it is impossible to establish how many of the officers deployed on the street are properly trained. This overall average of 60-70% also includes officers who work behind desks, and it is currently impossible to distinguish between them and those who patrol public spaces. This could mean that 70% are well trained, but behind desks, and 45% in the street. The rapporteurs asked the DGPN [the director-general of the French police, ed.] to provide them with these figures, but the DGPN stated that it was not in a position to supply them, as it did not itself have any data on who among its officers underwent ongoing weapons training. This problem had also been pointed out by the Cour des Comptes, which, in addition to considering that firearms training was insufficient and poorly done, had also pointed out that the police hierarchy did not keep a list of its officers who had failed their firearms training. The rapporteurs themselves conclude that all this gives the impression of “an inability on the part of the system to exclude those who do not have the required aptitude or behaviour”, echoing a conclusion already made by the Cour des Comptes in a previous report.
The final point that the rapporteurs make in the report is that the gendarmerie has a much better ‘professional culture’ than the police, and that the police should improve their training in social issues, improve police-population relations, develop better control of weapons by their officers and change the relationship they have with them. All this is extremely worrying! The rapporteurs admit in black and white that in France we have a police force that does not know how to manage its relationship with the public or with its own weapons.
That said, by focusing on training, they exonerate themselves as MPs – after all, they are the MPs who voted for this law. And yet, the law did trigger an application circular, a set of guidelines the style of which was a clear break with the previous ones, which in turn legitimised the firing of weapons because of the likelihood of future danger. And the deaths that go with it.
INDEX : If, as you say, the members of parliament are absolving themselves of any need to reform article L. 435-1, and the government doesn’t seem to want to put a stop to this police drift either, all that’s left at institutional level is the judiciary. On 3 October, for example, the two officers of the Stains police were put on trial for having fired eight shots at the vehicle of Nordine A. and Merryl B. in 2021. Article L. 435-1 and the question of the legitimate response of police officers to a situation perceived as a refusal to comply are at the heart of these proceedings. In your opinion, is the judicial route a realistic approach to breaking the deadlock over the State’s handling of refusals to comply?
SR: It’s important to use this route, because it’s the only one left among the institutional routes. In the Assemblée Nationale, MPs do not want to question their own institution, nor the previous decisions of their own political parties. In the case of this report, the two MPs who led the work are for one a member of the Socialist Party, which passed this law in 2017, for the other a Macronist MP, whose party has always found it to be a very good law. They are not well placed to assess it objectively. In some countries, there have been major reports following protests which themselves followed deaths. These reports were produced by independent rapporteurs. This is the case of the Kerner Commission in the United States, or the Macpherson Commission in Great Britain. It was independent figures who conducted the investigations and came to conclusions that shook up the way the State and the police operated. But in France, things don’t work like that.
As regards the possibility of a legal challenge to the current situation, the difficulty is that the judges tend to have the same interpretation as the politicians, i.e. one that interprets any form of refusal to comply as an attack on the State, a “challenge to the authority of the State and a direct challenge to the authority of the forces of law and order”. I am quoting a phrase from the report by the two MPs, which is highly emblematic and which, in the light of the judgements handed down, seems to be shared by some judges. For many of them, condemning a police officer is tantamount to condemning the action of the State. The State wants to be able to command, and the instrument of its authority is the police. So by condemning a police officer for illegitimate use of force, a judge – who is part of the State, just like the government and the Assemblée Nationale – would be condemning the fact that the State is trying to impose its authority on the population.
Magistrates find themselves in this uncomfortable position: they are supposed to read the law, and they do, but they are also the instruments of political domination, to use Max Weber’s expression – in other words, the army, the police and the judiciary are the instruments of a State’s domination of a territory. So how can they challenge these neighbouring instruments, the other agents of state domination, the police? I think that magistrates find it very difficult to do so, because they are in some way trained to protect the State, they evolve in a professional culture of protecting the State.
INDEX : There is a perception that the discourse adopted by police actors has a strong influence on both the French political and judicial milieux. While the social sciences teach us that violent crime is declining, representatives of the police institution tend to propagate images and feelings of unbridled criminality and permanent threat, which tend to justify measures such as the reform of self-defence. Why, in your opinion, does this discourse enjoy such credence, as in the case of this parliamentary report, when the various indicators available to us should temper, if not disqualify, such alarmist discourse?
SR: What is striking about the report is the credence given, without any particular verification, to a number of assertions made by police officers or magistrates. It’s as if some people should be taken at their word and others not. To take the example of the rapporteurs’ visit to Marseille, they report that they met police officers and magistrates who told them that not all refusals to obey orders were recorded, but who did not provide any proof of this phenomenon, nor were they able to measure the proportion of refusals to obey orders that were not recorded. Despite this, these assertions are sufficient for the rapporteurs. There is an “authority effect”, meaning that what is said by people in positions of authority in the administrative apparatus is deemed to be true, as if senior civil servants could not make mistakes or tell untruths, let alone lie. This possibility has been ruled out.
There are really two standards. There is a bonus given to officers. These are organizations which, if they protested against the government, would be a considerable thorn in the side of a government if they decided to refuse to carry out their duties, to throw their uniforms on the ground, to go on a strike – which they have done in the past, often in response to sanctions taken against their colleagues.
There is a kind of leniency with regard to the requirement for proof when it comes to statements made by police officers, and little willingness to demand precise information about the use of force and the risks of the job. What we see in the report is that when it comes to training, the MPs have asked a lot of questions, demanded proof, etc. But when it comes to the reality of the risks, the MPs have not been willing to take any action. But when it comes to the reality of the risks, the reality of the danger, this is much less the case. The rapporteurs tended to take at face value the notion of risk as presented to them by the police hierarchy. But it is the risk that justifies the use of violence. If you question the reality of the risk, you call into question the justification for using violence in a given situation – and, ultimately, the authority of the State. This was not the case.
INDEX : The State defends itself. What do you see as the prospects for change?
SR: There are MPs and members of the government who clearly stand in solidarity with state officials rather than with the victims of state violence. There are also judges, who are not necessarily in solidarity with the government, but who consider that their role is to participate in the defence of the State; and that to criticise the way in which the police use force would be to deprive the State of an essential means of governing. Today, because of their institutional position, these people constitute the line of defence of the police power when it comes to the use of force and weapons.
On the other side, there is an undermining of violent practices. We are in the middle of the battle for recognition of the violence perpetrated by public institutions. A battle that, as far as the police are concerned, is highlighting their excesses in a number of concrete cases: for example, fatal shootings against unarmed people who pose no direct threat, as is often the case when people refuse to comply, and particularly when it regards the deaths of passengers in vehicles.
This fight is part of a wider transformation of sensibilities. The work of numerous activist groups is helping to remind people that the law must apply to everyone, both in terms of the rights it confers and the acts it prohibits. There is also the work of organisations such as yours, which collect evidence, particularly photo and video evidence, and make it public. There is also the work of intellectuals, academics and certain political leaders, who produce knowledge about, for example, the racist bias of the police.
All these dynamics are now unfolding in the public arena, in battles, testimonies, studies, activist initiatives… For the moment, we don’t know whether they will influence court decisions, but in any case, we can see that they are beginning to have an impact on public debate.