1Peter A. J. Waddington and his colleagues state that:

2

“It is not ‘the state’ that knocks on the door in the middle of night, but a police officer – a corporeal being – acting on its behalf. Whatever influence the state exerts is communicated through and mediated by the police who are its agents.” [1]

3Police officers have the power to stop, check, search, interrogate, and even shoot at their fellow citizens, all on behalf of the state. To study police practice is therefore to analyze “the state as a regulatory body acting on deviance and more specifically the mechanisms which make certain persons or groups of people come to see themselves, and/or be seen, as ‘deviant’”. [2] The ethnographic approach to the state in its control procedures is therefore in accordance with a long tradition of research which understands the state “from the bottom up”, beginning with front-line agents who, on a day-to-day basis, not only apply its policies but above all make their own decisions. [3]

4Based on a comparative analysis of standard stop and check policies carried out by police services in Germany and France, we aim to contribute to this topic in two different ways. First of all, we will argue that the decision to ID check individuals allows us to identify styles of police action; in other words how police forces manage routine situations, communicate with the population (in a more or less formal, or more or less confrontational way) and, in broader terms, how they understand their action. [4] Furthermore, in seeking to move beyond “top-down” and “bottom-up” approaches to the state, we combine micro-level observation of police control practices with a more explanatory analysis, mindful of the meso-level (police organizations) and macro-level (police policies) conditions under which checks are carried out. [5]

5By control practices, here we refer to identity checks, interrogations, searches, or arrests carried out in public space. As a demonstration of the coercive capacity of the state, [6] they are also the source of recurring controversy. There have been debates concerning ‘stop and search’ procedures in Great Britain, which the Home Secretary addressed in 2013, [7] and ‘Terry stops’ [8] or roadside checks in the United States of America (we are reminded of the famous expression ‘driving while black’ [9]), as well as debate in France concerning identity checks. [10] Checks are a constant of police activity, to the extent that they are used by all forces, [11] but they can nonetheless vary in their frequency (from rare to systematic), their method (from a simple identity check to a check followed by an arrest), their conduct (from decent and respectful to brutal and violent) and their distribution among the general population (from a concentration on the “usual suspects” to more diverse targets), which requires us to understand why, when, and how these checks are performed and how they are regulated. Regarding the latter perspective, in the two countries studied here, the legal systems (article 78 of the Code of Penal Procedure in France, the Federal Code of Criminal Procedure and the police laws of different Länder in Germany) define in a general way the conditions under which checks are authorized. Whether used for investigative or preventative purposes for an individual suspected of having committed or preparing to commit an offence (judicial police check) or to prevent a breach of public order (administrative police check), checks need to be based on the idea of “reasonable suspicion” (or “concrete suspicion” in Germany). [12]

6In view of the flexibility of this legal framework, officers enjoy some leeway when deciding to undertake checks. Consequently, it is particularly interesting not only to look at what place ID checks have in police-citizen interactions, but also, and most importantly, the proportion of checks instigated by police officers. Police sociology has shown that officers make decisions based on so-called proactive cues, that is to say a “‘shared-recipe’ knowledge about whom to stop for what purpose in particular circumstances”, [13] or on a range of indicators that can vary according to the context, which combine the appearance and behavior of an individual. In analyzing these checks which we describe as “discretionary” – namely those which result from a broad definition of suspicion [14] – we are looking to establish the implicit understanding of police action which officers channel. To approach this empirically, we focus on three questions: 1. To what extent are ID checks used in interactions between the police and the general population? 2. Why do police officers use checks: that is to say, what are their implicit and explicit objectives, and which indicators do they consult? and 3. Who is checked: i.e., which segments of the public are at the greatest risk of being checked? The first two questions focus on the balance between the right to carry out checks (for legal reasons or to maintain public order) and protecting citizens’ private lives and freedom of movement. [15] The third point examines the possible disproportionate checking of certain segments of the population, notably visible minorities. [16] However, for lack of sufficient space, we will only approach the issue of “how” indirectly here (by “how”, we mean the public reaction to checks and the ways they are handled by the police), even if we will inevitably refer to it.

7To address these questions effectively, this article takes an approach which is doubly distinctive. First, it is a comparative study (a method rarely used in social science research on the police [17]), systematically examining the practices and logic behind checks carried out in Germany and France. While there exist numerous fruitful comparisons between Germany and France concerning employment policies, professional relations, or training, there is little available comparative knowledge on police issues, [18] although the combination of similarities and differences between the two countries suggests interesting perspectives. On one hand, Germany and France have similar judicial systems (the Napoleonic penal code), crimes (both from the point of view of acts and of concerns), and even similar institutional responses (increase in crime prevention, policies ranging from oppressive mandates to increasing proximity with the public). [19] Significantly, the few international police classifications that exist include Germany and France in the same continental police model, characterized by a centralized organization, broad responsibilities (including administrative police activity), and a legitimacy which is primarily state-facing. [20] On the other hand, the French centralizing tradition is in contradiction with German federal organization, [21] while urban tensions have resulted in very different levels of collective violence: whereas France has been hit by high profile riots, these are a very rare occurrence in Germany. [22]

8Conducted as part of the Franco-German project Police and adolescents in multi-ethnic societies (POLIS), [23] this research deals with four major cities (Lyon and Grenoble; Cologne and Mannheim). These cities all have large proportions of minorities in their populations: although it is difficult to measure the proportion in Grenoble and Lyon for legal reasons, [24] in Mannheim and Cologne, half of the population under the age of 18 is of immigrant origin, with Turkish as the largest group. [25] In Grenoble and Lyon, minority groups originate from North Africa and to a lesser degree from sub-Saharan Africa.

9This article also draws on a combined use of quantitative and qualitative methods, and aims to combine an understanding of the rationale for ID checks (and non-checks), based on an understanding of how police approach such checks, with measures of how frequently they occur. [26] Alongside a survey carried out with students in secondary schools, the research is based on around 800 hours of direct observation of police at work on the street, around 200 hours in each city. [27] In addition, 65 semi-structured interviews were conducted with police officers in Grenoble and Lyon and around 50 in Cologne (in the Land of North-Rhine Westphalia) and Mannheim (in the Land of Baden-Württemberg). In France, 293 police-population interactions were observed, compared to 247 in Germany. [28] This information allows us to combine quantitative data (by measuring the number of discretionary checks in the overall number of interactions and the different citizens targeted therein, and also by employing the survey of teenagers to measure the prevalence of checks) and qualitative data (by analyzing the values, norms, and objectives which guide the actions of police officers). From this perspective, direct observation is particularly useful for analyzing interactions in context: combined with informal discussion with police officers after checks, this method allows us to analyze (in context) what led to them. [29] In both countries, the observations were carried out during shifts and in quiet moments by observers who could have informal exchanges about the interactions occurring. In Mannheim and Cologne, observations were carried out with patrol teams (responsible for responding to emergency calls, called Streifendienstbeamte in Mannheim and Einsatzbewältigungsbeamte in Cologne), units in charge of police-youth relations (Jugendsachbearbeiter in Mannheim) and community police officers (Bezirksdienstbeamte in Cologne). In Grenoble and Lyon, a higher number of units were observed, to reflect the greater diversity of units operating in public spaces: the two best-known of these are the ordinary patrol units (“Police secours”, in uniform) and anti-crime squads (or BAC – Brigades Anti-Criminalité – in plain-clothes, with a mandate for proactive action against ongoing crime), but we also included specialized uniformed transport units, uniformed units in charge of certain high-crime areas with reinforced equipment (specialized ground units), as well as a myriad of other units, such as “neighbourhood security groups”, or “day units” (generally in uniform, with a more repressive mandate than the emergency squad). Finally, the two regional police forces had so-called “community” (or local) units or youth-relations units, which no longer existed at national level after the beginning of the 2000s and the end of community-based policing. [30]

10The rest of the article will be divided into three parts. First, we will analyse the rationale behind checks, by shedding light not only on police officers’ reasons for action, but also the reasons which lead them not to act. Next, we will examine the targets of checks more specifically, by questioning in particular the overrepresentation of minorities. These first two parts having shown notable differences (that German police officers carry out fewer checks and focus much less on minorities), in the third part we will consider the reasons for the contrast.

To check or not to check: police thinking and professional reflexivity

11In France, discretionary checks are a much-used procedure, which is only partly true in Germany. The advantages and disadvantages are evaluated differently by police officers from the two countries. The marked contrasts between the decision to carry out a check, or indeed not to, bring us back to the implicit definitions of police activity.

Avoiding checks

31The issue of checks can also be approached from another angle, namely when police officers avoid going as far as checks, whether because they know the people in question, or because they foresee the negative consequences of a check. The first situation reveals a mutual understanding between police and public; the second a reflexivity on the part of police officers about their own practices.

Disproportionate checking of minorities? Comparing national patterns

49The second aspect of our research concerns the targets of discretionary checks, and in particular the potential over-representation of young men [39] from minority groups. Obviously, one of the questions asked is about the discriminatory practices noted in the study, [40] and how to measure them. Qualitative studies have difficulties in robustly demonstrating discrimination by the police because of the specific material they collect, which is composed primarily of interviews and does not allow for potential racial profiling to be objectified. [41] Here, we propose combining three sources: quantified interactions (Table 2), the results of the survey on teenagers (Figure 1), and finally the qualitative elements taken from the direct observations. The first two techniques allow us to identify the processes of disproportionate checking, and the second is more useful as we attempt to identify explanatory elements. On this last point, we will return in particular to the question of links between racism and discriminatory practices.

Focus or balance? National contrasts

50The data collected during the direct observation sessions highlight that discretionary checks make up a much larger proportion of interactions with visible minorities [42] (31.2%) than with white people (14.1%) in France, whereas in Germany, proportions of checks per 100 people are more equal between the majority population (12.2%) and visible minorities (12.25%) (see Table 2).

Table 2

Distribution of discretionary checks (based on appearance)

	N	Proportion of discretionary checks in total number of interactions (%)
France
Total	293	27.3
Apparently white	71	14.1
Visible minorities	205	31.2
Mixed groups	17	35.3
Germany
Total	247	12.6
Apparently white	115	12.2
Visible minorities	120	12.5
Mixed groups	10	20

Distribution of discretionary checks (based on appearance)

51These results coincide with those of the survey carried out in secondary schools, the results of which are shown in Figure 1, where the prevalence rate of discretionary checks over the last twelve months is less than 30% for Germany regardless of the group they belong to, while in France the rate is 21.6% for young people of French origin and 36.7% for young people of African origin. If we look at the profiles of the young people most frequently the subject of checks (five times or more during the last year), young people of African origin have a three times higher representation (17% versus 5%), while no major difference can be seen in German cities.

Figure 1

Prevalence of checks initiated by police (Germany on the left, France on the right)

Prevalence of checks initiated by police (Germany on the left, France on the right)

52Even if we cannot compare these results with the available population (those present in the area at the time of the check), they show regular, substantial differences. These conclusions support those of existing studies which, on the basis of quantitative sets, tend to establish discriminatory practices on the part of French police officers. [43] More so in France than in Germany, the ethnic element constitutes one of the factors associated with discretionary checks on the part of the police; the “police quarry” (gibier du police) [44] consists of young North African or black men dressed in street clothes, and this is the phenomenon which we seek to understand. In particular, the disparities revealed here go hand in hand – as we will see – with particular ways of thinking by the police.

From police practices to political and organizational contexts

73After reporting the main findings from our analysis, we will attempt to propose an explanatory framework. First, the consistency and homogeneity of national values, norms, and practices should not be exaggerated. While community police agents in Cologne avoid checks most of the time and have an individual knowledge of young people, most patrol agents don’t have this knowledge at their disposal and many of them show an inclination towards undertaking checks in a reactive way. In France, certain experienced agents demonstrate a great deal of care in the use of checks, unlike many of their colleagues who take little account of their collateral effects. Our research tends to support the idea that two different professional styles exist: in France a more proactive checking style prevails, driven by the fight against crime, whereas in Germany a more reactive and informal logic dominates. In Lyon and Grenoble, identity checks are seen as a way of exercising power, and as a way of meeting several objectives, while their negative effects on the public are little taken into account, contrary to the approach in Mannheim and Cologne. Interactions in France are also characterized by a heavy over-representation of minorities compared to the majority population, while proportions are much more balanced in Germany. Direct observation, informal interviews with police officers and statistical data suggest that a link exists between ethnic stereotypes, creation of suspicion, and the carrying out of checks, whereas in Germany police consider indicators other than visible minority status. German police are close to what James Q. Wilson [59] calls the “order maintenance” style: their first concern is to maintain social order (public calm) in the community, and value is attributed to operational autonomy and informality. For example, officers seek to gather information by other methods rather than using checks to obtain information on individuals. The style of French policing is more legalistic: they give priority to applying the rule of law – particularly in relation to their mandate of crime-fighting – and interpersonal relationships are more limited. [60]

74By focusing on the practice of checks, we are situating ourselves on the micro level of agent behavior, and must now turn to the extrapolation of more general explanations, by combining these results with the more meso- (organizational) and macro-level elements (public policies). The explanation of observed differences leads us to abandon the idea of a strict approach in terms of relations of monocausality and to consider instead the combined effect of a set of factors (political, organizational, and professional) embedded in layers of social relations. Although recent police policy offers a partial explanation, it exaggerates the structural tendencies at work in both countries.

75First, these differences are found in recent policies carried out in the two countries. In France, the priority given to the “fight against criminality”, the “culture of performance”, and the “policy of results” between 2002 and 2012 has encouraged proactive practice, with a higher number of checks being associated with greater efficiency in the fight against crime, [61] while in the two Länder, we observe no equivalent policy which values interrogations as the centerpiece of the work of police on the street. In Cologne and Mannheim, officers have little incentive to carry out checks in order to conduct interrogations and thus meet objectives on bringing individuals to justice. In Grenoble and Lyon, although not omnipresent, this logic constitutes a background context for action, with many agents complaining on their breaks of the perverse effects of the “race for numbers”. [62] Similarly, the disappearance of community policing in France constitutes a major difference with the two German Länder.

76However, the influence of these recent policy variables should not be exaggerated; the reasons are also structural, combining three factors: training, organizational direction, and professional routines and ethos. “Relations with the public” are an important component of training policies in the two Länder and the tenets of different religions and cultures are taught, while in France the relationship with the public is not a central issue, and is neither included in initial training, nor in in-service training. Significantly, when we interviewed officers about their training in public relations, they talked about professional intervention techniques (les gestes et techniques professionnels d’intervention (GTPI)), which are self-defense techniques.

77These differences must also be understood in terms of the organizational context: the checks carried out during our period of observation [63] were not the subject of internal questions. In France, their numbers are not recorded, there is no guide on how to carry them out (outside of the penal code and a few lines in the code of ethics), nor is it an issue the agents discuss with middle management. During our observations, checks were never the subject of debriefing between the agents and their supervisors. The organization of police units also differs: the rarity of plain clothes officers with an anti-crime mandate in Germany [64] marks a notable difference with the French anti-crime squads (BAC), for whom these checks are strongly connected to the fight against crime. [65] Police missions in both countries – and the organizational structure which translates these priorities – thus indirectly affect checking policies.

78Finally, these styles are closely intertwined with the relatively autonomous professional ethos [66] of the policies established. The profile of “the hunter” (or the “new centurion” in the crusade against crime) [67] is less prevalent in Germany than it is in France, while the idea that the police officer is “your friend and aide” (dein Freund und Helfer) permeates the representations of both German police and public. [68] In France the them/us separation is much more a structuring factor in the way in which police officers talk about their relationship with the public in general, the narrative of events involving defiance of the police is more commonly shared, [69] in a context where relations between the police and the public seem much more tense.

79These four factors (police policy, training, organizational direction, professional ethos) support a complementary relationship: for example, training contributes to forging a professional ethos oriented to varying degrees towards an interactional view of the job, an ethos which itself influences the acceptable organizational policies at the heart of the police service. From this point of view, the effects of the so-called “independent” variables are not isolated: they direct the practice of checks on the basis of the interrelations between them.

80* * *

81At the crossroads between judicial logic and public order, checking practices bring priority targets and the orientation of police work into focus. It is therefore crucial to understand precisely how, when, where, and on whom checking processes – ordinary police practices – are accomplished. In urban societies which are similar in many ways (socio-urban inequalities, proportion of immigrant population, urban security issues, etc.), it is essential to know if checks by law enforcement are also used in similar ways. Based on our Franco-German comparison, we have made three contributions to this question.

82First, we have identified what we will call, after James Q. Wilson, [70] styles of policing; that is to say, relatively consistent ways of doing things, choices of priorities, and targets. The importance of checks in interactions in France is embodied in a police force which situates itself in a position of control (a “legalistic” style as Wilson would say) and thus resorts to a coercive tool, while police in Cologne and Mannheim more often find themselves in situations where they are responding to the population’s demands and where they have to settle conflicts between different parties (the “order maintenance” style as Wilson would put it).

83Such results, based on a large-scale ethnographic research project that allows for quantification of contacts, address a gap in knowledge when it comes to the comparative analysis of the police. The few comparative works available too readily confine themselves to constructing models which have little connection to administrative and professional realities, [71] or are too centered on the few formal institutional variables, [72] tendencies which can be accounted for by the inherent difficulties of any comparative research undertaken on police issues. Yet, to use only this example, ID check law diverges only minimally between Germany and France. Comparing tacit knowledge, reasons for action or understandings that in the profession “go without saying” thus constitutes a promising avenue, building on the tradition of ethnographic works that seek to analyze the state from the bottom up. [73] Linking micro dimensions (practices) to the more meso- and macro-level variables (organizational direction, professional sub-cultures, police policies, training) represents a promising project for the comparative analysis of not only state administrations but also, more generally, public organizations. On the one hand, work such as this which seeks to reproduce professional understandings allows us to reintegrate multiple professional and administrative layers into comparative analyses, which otherwise run the risk of being centered only on elites, discourses, and programs of action. On the other hand, international comparison allows us to relate observations of practice to the political and institutional conditions in which these activities are carried out.

84Finally, a comparative analysis such as this has the merit of denaturalizing practices considered universal. The Franco-German comparison allows us to show that ID checking practices are undeniably an invariable element of police practice, but their use can vary as a function of professional and political contexts. It also highlights that confrontations between police and minorities are not self-evident in contemporary society. These results should be seen in the context of other international studies, which all tend to show that the level of confidence in the police is higher in Germany than in France, and that the perception of the existence of unfair treatment or even experience of violent treatment is also not in France’s favor. [74] Converging bodies of work have elsewhere revealed that checks seen as unfair, discriminatory, and excessive can damage confidence in institutions and become sources of conflict between the police and certain segments of the population. [75] In combination, these results tend to challenge the prevailing direction in France: by increasing the number of checks, French police officers put themselves in situations where distrustful and divisive attitudes are nurtured, and by doing this, they contribute to weakening the social legitimacy of the police, without gaining a measurable reduction in criminality. [76]