Different styles of policing: discretionary power in street controls by the public police in France and Germany

Abstract

By analysing French and German police stop and search on the streets based on embedded observations in police patrols and findings of a large school survey, this article comparatively questions their determinants. Control practices diverge in their frequency: the German police officers control less proactively than their French counterparts. The targets of controls also differ: a concentration on visible minorities is much more pervasive among the French police officers. These divergences may be explained by contrasted professional orientations, especially the importance given to the crime control agenda, and state/society relations.

Full text

Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=gpas20
Download by: [Jacques de Maillard] Date: 26 June 2016, At: 13:01
Policing and Society
An International Journal of Research and Policy
ISSN: 1043-9463 (Print) 1477-2728 (Online) Journal homepage: http://www.tandfonline.com/loi/gpas20
Different styles of policing: discretionary power in
street controls by the public police in France and
Germany
Jacques de Maillard, Daniela Hunold, Sebastian Roché & Dietrich Oberwittler
To cite this article: Jacques de Maillard, Daniela Hunold, Sebastian Roché & Dietrich
Oberwittler (2016): Different styles of policing: discretionary power in street
controls by the public police in France and Germany, Policing and Society, DOI:
10.1080/10439463.2016.1194837
To link to this article: http://dx.doi.org/10.1080/10439463.2016.1194837
Published online: 21 Jun 2016.
Submit your article to this journal
Article views: 9
View related articles
View Crossmark data

Different styles of policing: discretionary power in street controls
by the public police in France and Germany
Jacques de Maillard
a
, Daniela Hunold
b
, Sebastian Roché
c
and Dietrich Oberwittler
d
a
Cesdip-UVSQ-CNRS-Ministère de la justice-UCP and Institut Universitaire de France, Paris, France;
b
Department of
Criminology, German Police University, Münster, Germany;
c
CNRS, Pacte-Sciences Po, University of Grenoble,
Grenoble, France;
d
Max Planck Institute for Foreign and International Criminal Law, Freiburg, Germany
ABSTRACT
By analysing French and German police stop and search on the streets
based on embedded observations in police patrols and findings of a large
school survey, this article comparatively questions their determinants.
Control practices diverge in their frequency: the German police officers
control less proactively than their French counterparts. The targets of
controls also differ: a concentration on visible minorities is much more
pervasive among the French police officers. These divergences may be
explained by contrasted professional orientations, especially the
importance given to the crime control agenda, and state/society relations.
ARTICLE HISTORY
Received 1 November 2015
Accepted 24 May 2016
KEYWORDS
Discrimination; police; stop
and search; proactive
policing; street policing;
comparison
France and Germany are often presented as proximal cases in international comparisons on topics
related to policing. In the typology proposed by Mawby (2008), they are part of the ‘continental
model’defined by a legitimacy directed primarily towards the state, a broad definition of the
police mandate and a rather centralised command structure. Comparative studies of crime policies,
drawing on categories of welfare regimes (Hough et al.2013, pp. 251–254), classify France and
Germany in the same conservatist–corporatist category, characterised by a moderately generous
status-related welfare state, moderately hierarchical society and a penal ideology dominated by reha-
bilitation and socialisation (Lappi-Seppälä 2011, Cavadino and Dignan 2013).
A comparative sociology of policing may wonder whether police behaviours match these macro-
level divisions. Do policing strategies in neighbouring European countries reflect a shared regime of
social and penal policies? More specifically, by analysing how French and German police proceed to
control on the streets, this article questions the sets of values, norms and practical reasoning of police
officers in these two countries. We hence start with a microanalysis of police officers’behaviours on
the streets and later seek to relate their features, set of similarities and differences, to wider organ-
isational, political and cultural factors. By doing so, we echo seminal works exploring the international
contrasts in professional identities, police authority and uses of legal instruments (Banton 1964,
Bayley 1978), and more recent empirically grounded comparisons (Cassan 2011, Body-Gendrot
and Wihtol de Wenden 2014, Devroe and Terpstra 2015, Gauthier 2015) which have underlined
how the use of force, the representation of the public and the training practices differ between Euro-
pean countries.
This paper will focus on the practices of street control undertaken by police officers. The use of the
term ‘police officers’obviously needs further specification considering the varieties of the organis-
ation and policies of security forces within and between the two countries (see below). By street
© 2016 Informa UK Limited, trading as Taylor & Francis Group
CONTACT Jacques de Maillard demaillard@cesdip.fr
POLICING AND SOCIETY, 2016
http://dx.doi.org/10.1080/10439463.2016.1194837
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

control, we mean the practices of discretionary identity checks or stop and search (or stop and frisk),
and arrests that may accompany them. These control practices constitute an important starting point
for comparative analyses because they present a mix of similarities and potential differences (see
Bowling and Weber 2011). Identity checks, searches, questioning and arrests translate concretely
the ability of state representatives to limit the liberties of people. They thus represent a visible mani-
festation of the police legal monopoly of violence and intrusive nature of the state (Weber and
Bowling 2011). They also constitute the most frequent type of interaction between state agents
and citizens (and non-citizens) and, as such, can be a source of controversial relations between
the police and communities. Stop and search in Great Britain (HMIC 2015), Terry stops (especially
stop, search and frisk in New York) or traffic controls (with the famous ‘Driving while black’) in the
US or identity checks in France (Jobard et al.2012, Fassin 2013) have been a focal point of enormous
media, public and political attention. Excessive, unfair and discriminatory controls are seen by many
as damaging police legitimacy and as a major cause of tensions between police and segments of the
population (Hough 2013).
But these control practices may diverge in their frequency (from rare to systematic), in their inten-
sity (from a simple ID check to an arrest), in their handling (from decent and respectful to harsh and
violent) and in their distribution (from a concentration on the ‘usual suspects’to more diversified
targets). In order to fully understand the implications of control policies, researchers need then to
know study how, why, when and where controls are exercised.
ID checks and more generally decisions to control are generally considered as involving dis-
cretion by police officers as legal frameworks (see below) often define very broadly the necessary
requisites to initiate a control. For this paper, we have identified those interactions in which
police officers had discretion (what we call, see below, discretionary controls) in order to
analyse how police officers make use of it in controlling, searching and arresting people.
These decisions to control or not to do so, and then intervene in the lives of citizens are tied
to different styles of policing (Wilson 1968, Muir 1977, Hough 2013), that is, how the police
handle routine situations that bring them into contact with the public. In this research, we
pay attention to the type of police work (police-initiated work or responses to calls), to more
or less formal behaviours (enforcing the law or maintaining order), and to more or less adversarial
types of policing.
Three questions are empirically addressed in this paper: (a) To what extent is control exercised
during encounters, that is, which proportion of interactions between the public and the police are
made of control initiated by the police? (b) Why do police officers control, that is, what are the
implicit or explicit objectives followed by police officers, their proactive cues (Ericson 1982),
when they undertake controls? (c) Who are the controlled, that is, who are the most likely
targets? Each question resonates with important debates in the literature. The first two ones
raise the issue of the balance between the right to control and the right to privacy and
freedom of movement (see, in a large literature, the collection of papers on different national
cases collated by Bowling and Weber 2011). The third one refers to the large debate on the over-
controlling and ‘profiling’of some segments of the population (mainly minorities) by the police (in
a very broad literature, see Bowling and Phillips 2007, Rice and White 2010). The ‘how’issue (how
the control is handled) is not directly addressed here for space reasons, even if we will inevitably
allude to it.
The analysis is developed in four steps. We first precise the legal context (rules constraining control
in both countries) and the methodologies mobilised. Secondly, we analyse what we call the ration-
alities of control in the two countries and the part of discretionary controls, the reasons for enacting
them but also the potential reasons for avoiding controls. We then focus on the issue of discrimina-
tory controls by identifying the proportion of controls devoted to minorities. In Section 4, we interpret
the differences observed in the two countries.
2J. DE MAILLARD ET AL.
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

1. Context of the study: legal framework and methodological background
1.1. Reasonable suspicion and discretionary controls
In both France and Germany, the legal framework defining conditions of controls are broadly similar.
1
Controls may be used for investigatory or preventive purposes in relation to an individual suspected
of a specific offence. They must rely on ‘reasonable suspicion’by police officers. The public cannot
refuse to be submitted to an ID check (even if this ID check is illegal). In addition to suspicion by
an agent, stop and search may be triggered as part of a penal policy as decided by judicial authorities
(article 78.2 of the Code of Criminal Procedure). For example, in certain areas of Lyon and Grenoble –
known for being drug-dealing places –police are given authorisation by the local prosecutor to stop
and search anybody regardless of his behaviour. This also applies to Germany, where police is able to
define so-called danger zones which permit ID-checks irrespective of behaviour (known as ‘gefähr-
liche Orte’, dangerous zones).
Legal rules give the police officers a large leeway to decide what course of actions to choose. As a
consequence, we define discretionary controls as ID checks and/or stop and search based on an
extensive definition of ‘reasonable suspicion’. Identifying these discretionary controls is a way of iso-
lating them from other controls and ID checks that fall out of the decision of an agent, such as clear
infringement to the law (over speeding) or an indisputably suspicious behaviour (being aggressive,
exchanging a small package in a drug-dealing zone for example). Discretionary controls are spoken of
when the officers decide to control on criteria external to the possibly delinquent behaviour of the
person: an individual out of place, the general condition of a car, clothing and the attitude of a person
(for instance ironic looking). In all these cases, police officers decide to control based on a set of
factors that we could call ‘proactive cues’, that is, ‘shared recipe knowledge about whom to stop
for what purpose in particular circumstances’(Ericson 1982, p. 86).
2
1.2. Methodology
This article is based on a joint French–German project (called POLIS, Police and Adolescents in Multi-
Ethnic Societies) funded by the two Funding research agencies in the countries (the Agence nationale
de la recherche and the Deutsche Forschungsgemeinschaft).
3
The research was multi-methods and
multi-sites. Four cities have been chosen: medium-sized cities between 300 and 400,000 inhabitants
and large cities of 1 million inhabitants (Grenoble and Lyon on the French side and Mannheim and
Cologne on the German side). These cities all have high proportions of minorities in their populations.
If the measure is difficult in Lyon and Grenoble due to the census counting, half of the population
under 18 is from a migrant background in Cologne and Mannheim, with Turkish being the largest
group (Oberwittler and Roché 2013). In Lyon and Grenoble, minorities are from North Africa and,
at a lesser degree, from Sub-Saharan Africa. The research was primarily based on around 800
hours of direct observation of police patrol, with approximately 200 hours in each of the cities. More-
over, 65 semi-structured interviews were conducted with police officers in Lyon and Grenoble and
about 50 in Cologne in the Land of North Rhine-Westphalia
4
) and Mannheim (in the Land of
Baden-Württemberg). In France, 293 police–citizen encounters have been observed, against 247 in
Germany.
5
As the main focus of the study was on police–adolescents relations, police officers
were partly selected on the basis that they specialised on this age group, the observations are to
a certain extent biased towards interactions involving young people. The observational data are sup-
plemented by standardised data from a large school survey which was conducted in all four cities
among young adolescents (ca. 13–16 years) using an identical design and questionnaire. The ques-
tionnaire follows the tradition of previous youth studies including the International Self-Reported
Delinquency Studies (Oberwittler 2004, Junger-Tas et al.2012) and includes questions on self-
reported delinquency, routine activities, social bonds and family situations of adolescents, as well
as a section with detailed questions on their contacts with police officers (i.e. experience at last
POLICING AND SOCIETY 3
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

contact) and their attitudes towards the police (cf. Hunold et al.in press). The survey was conducted
in the paper and pencil mode in classes during school hours, in Germany in 2011 and in France in
2012.
6
These data enable us to mix quantitative (measuring the proportion of discretionary controls and
the various categories targeted by these controls) and qualitative (analysing the objectives, values,
norms that guide the behaviour of police officers based on their explanations of their actions)
data. Direct observation is particularly relevant as it enables to describe the concrete exercise of con-
trols. By informally discussing with police officers after the controls, it also allows us to infer the con-
textualised reasons that have generated the decision to control. In both countries, observations were
undertaken by trained observers during ride-along, and during the downtimes observers could
debrief informally about their interventions. In Mannheim and Cologne, observations have been
made with response teams (patrol officers), youth officers
7
and with community policing units.
8
Thus, this study covered the work of ‘ordinary’German policemen (and women) who constitute
the largest subgroup of the profession in German cities, as well as of specialised officers whose
tasks are more tuned to variations of community-oriented policing. In Lyon and Grenoble, reflecting
the diversity of police organisations, observations have been realised in a broader variety of units
(response teams, plain-clothes units with a crime-fighting mandate (Brigade anti-criminalité), uni-
formed units specialised in transport, uniformed unit dedicated to specific areas with reinforced pro-
tective equipment (Brigade spécialisée de terrain)). However, we did not observe any community
policing units, as community policing reforms introduced at the end of the 1990s have been since
politically challenged by the political right (especially by Nicolas Sarkozy when he was minister of
interior between 2002 and 2005) and suppressed from the French National police (Roché 2005).
As a consequence, the observed police units in the two countries differ to some extent in terms of
tasks and code of conduct. Nevertheless, these differences display the reality of everyday police
work in the streets. The observational data are supplemented by standardised data from a large
school survey which was conducted in all four cities using an identical design and questionnaire
which contained detailed questions on contacts with police officers, including the experience of
the last contact. Thanks to the very large sample sizes (13,500 in France and 7300 in Germany),
the survey offers unique opportunities to analyse the experiences and attitudes of adolescents
from ethnic minorities (as well as of native adolescents) without quickly running into the problem
of small numbers. Nearly, 2800 Maghrebian/African adolescents in France and 1400 Turkish adoles-
cents in Germany participated in the survey (Hunold et al. in press).
2. Rationales of controls: proactive ID checks and searches in police–citizens
encounters
In France, discretionary controls are clearly an important part of the toolbox used by police officers on
the streets, whereas this is true to a lesser extent among German police officers. The advantages and
the drawbacks of controlling are weighted differently by police officers in the two countries.
2.1. Rationales of controls
To measure the extent of the discretionary controls among all the interactions (Table 1), we have iso-
lated those involving a discretionary control (see the definition in the methodological note).
Differences are striking. In Lyon and Grenoble, more than one in four interactions are made of a
discretionary control undertaken by the police (27.3%), whereas the proportion is one in eight in
Cologne and Mannheim (12.6%). This discrepancy calls for a closer scrutiny of the concrete inter-
actions in order to understand the different approach of French and German police. Discretionary
controls (stop and search but also simply ID checks) may, in principle, serve several purposes that
we can divide into three categories: crime fighting, asserting authority and collecting information.
4J. DE MAILLARD ET AL.
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

Firstly, control may be seen as a way of detecting and arresting people engaged in, or planning,
crime (Bowling and Weber 2011). Reasons related to crime investigation or detection are pervasive
among the French police officers, some of them having a rather low threshold to define suspicious
behaviours. Clothing, daytime and skin colour (as we will see later) are crucial factors, and they are
often combined. One example from the observed interactions may serve as emblematic illustration
here:
I22 (Grenoble): The police officers see two young boys on a scooter, without a helmet and with a large bag. They
suspect it could be drug-related. They decide to stop and check the boys.
The location is important: we are in a deprived suburb of Grenoble nearby a housing project. The
issue of drugs is also crucial: in Grenoble and Lyon, police officers are often looking for drugs, not if
limited to personal use, however. In the interactions we witnessed, they most of the time underen-
forced the law (by simply destroying the hashish) when the quantity was an indication of personal
use only. In some (rare) cases, police officers have a maximalist conception of control in relation to
drugs traffic, as the example below illustrates:
I91 (Lyon): We are nearby Lyon-Part-Dieu railway station. Two males from Maghreb ask the two police officers for
directions. The police officers asks them to show their ID and if they have any drugs. They take a look at their ID,
tell them ‘it is your lucky day, no frisk’and let them go after giving them the direction.
German police officers, too, are searching for drug-related crimes. The decision for a discretionary
control mainly relies on appearances or location:
I58 (Mannheim): The officers decide to control juveniles who are loitering at the street corner because of their
clothes which are suggestive of drug use.
However, as crime control is of little importance in general, corresponding actions were rarely
observed. This is due to the structure and tasks of the units –community police officers and
patrol police –that are potentially in charge of proactive identity checks. Although community
police officers predominantly operate proactively and are instructed to conduct identity checks, in
fact they rarely practice controls. While patrolling the streets they rather focus on relationship man-
agement with the public –for what they are also responsible. In most instances, community police
officers already know the persons loitering at suspicious places and thus avoid checks, unless they do
not know the persons or have a reasonable suspicion for a crime. Therefore, most of the contacts
between young people and community police officers are informal (Hunold et al. in press) In contrast,
patrol police are predominantly dealing with citizen calls, and thus rarely perform proactive controls
(see below), simply because they do not have enough time for this kind of work.
The second rationale is asserting authority. It is expressed when police officers yearn to express
that they are in control of the streets. Discretionary controls that may result here from an attempt
to attest a visible presence in a neighbourhood show that ‘we are not afraid of them, that we are
in control’as one French police officer told us. Controls may also result here from tense interactions
Table 1. Discretionary controls.
Total interactions (abs.) Discretionary controls (abs.) Share of discretionary controls (%)
France
Total 293 80 27.3
Assumed natives 71 10 14.1
Visible minorities 205 64 31.2
Mixed groups 17 6 35.3
Germany 247 31 12.6
Total
Assumed natives 115 14 12.2
Visible minorities 120 15 12.5
Mixed groups 10 2 20
POLICING AND SOCIETY 5
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

between the public and police officers. ID checks are the product of a latent conflict. Exchange of
sights, ironic expression or insults may trigger controls:
I287 (Lyon): three young guys from Maghreb are in front of a snack. The police pass in car and the sergeant hears
‘go away’. They immediately make a half-turn and go to control the three guys’.
In this kind of situation, control is connected to a disciplinary rationality (Gauthier 2015). Control
may resemble a form of ‘street justice’(Van Maanen 1978, Fassin 2013) for those who disrespect the
police or do not accept the police definition of the situation: the person is immobilised for a few
minutes, may be searched in front of other persons, is questioned potentially intrusively. This use
of discretionary controls reveals the poor relationship between the police and some segments of
the youth. Our results suggest that they are more frequent in France: we did not observe any such
incidences in Germany, although some German police officers mentioned in interviews that they
may handle some situations this way.
Collecting information is the third rationale. Controlling ID and questioning is a way of getting
information on (groups of) individuals. In France, the ID controls undertaken in hallways of social
estates are particularly emblematic: their purpose is to collect information on individuals shared after-
wards with other local actors (municipal agents and social housing companies). This logic is fuelled by
interorganisational processes: police, municipalities, the judiciary, social workers and housing estates
may be part of working groups in which information is shared between individuals (according to
certain rules of confidentiality). In the two countries, ID checks may also be realised at the
demand of the transport companies when they fine people without their IDs.
In Germany, the collection of information follows a different logic. Community police officers and
youth officers, in particular, decide for a control when they notice youngsters in the streets they do
not yet know personally. Their aim is not to share information with other local actors but to get into
contact, generally in an informal manner, with these juveniles:
When I don’t know a group of juveniles and they don’t suit to the neighbourhood, then I control them, talk to
them in leisurely manner and then I just ask what they are doing here and then I see how they react. (Community
police officer, Cologne)
French police officers use control mainly in the three first logics, German police officers more in
the two last ones. This implies that not only German police officers control less but they also
control differently, with a lesser law enforcement tone.
2.2. Avoiding controls
The issue of controls can be analysed from the opposite angle: when police officers could have done
a control but have avoided it. Two different situations must be distinguished here: either officers do
not control citizens because they know them or they deliberately avoid control for other reasons. The
first situation reveals the interpersonal linkages between the police and public and the second a
certain reflexivity of police officers about the consequences of control.
2.2.1. Non-control and interpersonal linkages
In the first configuration, control is unnecessary by the mere fact that police officers know the indi-
viduals, and more specifically the youngsters. In the four cities, such situations may occur. For
instance, in I90 (Lyon), police officers search for someone suspected of a robbery. When getting
closer to individuals matching the suspects’description, they realise that they are people they
know and who are no criminals and hence decide to avoid the stop and search. Non-control
results from an interpersonal knowledge. According to our data, the difference is a matter of
degree: interpersonal connections are simply more frequent in the German case. In France, we
have only two cases in which interpersonal connections may lead to an absence of control,
whereas the frequency is much higher in Germany. Let us give the following examples: ‘A juvenile
6J. DE MAILLARD ET AL.
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

takes away his hood when the police officers aims a pocket lamp on him, they know each other’(I10,
Cologne), ‘the men are drinking alcohol in front of a house and hear music from a car radio, officers
and men seem to know each other’(I16, Cologne), ‘officers and juveniles have a chat about police
work, they know each other’(I21, Cologne) and ‘the young man is talking to his friends, the officer
and the young man know each other’(I24, Cologne). These examples suggest that control would
be more limited in Germany due to more interpersonal linkages between police officers and citizens.
But this kind of police–public relationship is specific to community police officers.
9
This is, however, not only an issue of interpersonal knowledge but also of professional values and
norms. In France, police officers might control someone because he is known for not carrying his ID
card: ‘A young Maghrebin is well-known for never having his ID on him. He is checked for that reason.
The police officers teach him. His brother swears he will have it on him next time’(I234 Grenoble). In
this interaction, if police officers underenforce the law (as they should have taken him to his home or
to the police station to prove his identity), they deliberately control him to remind the necessity of
carrying his ID card: the interpersonal knowledge does not lead to a non-control but rather to the
contrary.
2.2.2. Non-control and police reflexivity
In the second configuration, police officers eschew control because they consider that its advantages
in terms of police work do not overcome its drawbacks in terms of their relations with the youths. In
France, this kind of attitude, which remains rare, is expressed by experienced police officers seeking
to avoid unnecessary tensions with the youths. One sergeant, for instance, insisted on the need for
controlling only teenagers they did not know and expressed it clearly during controls (‘we control you
because we don’t know you, but we won’t do it again’). It is more widespread among German police
officers. In the two quotations below, officers show a lot of caution to use ID checks. They adapt their
behaviour to the nature of the interaction (the control not being a requisite of the start of the inter-
action) and may use information from other colleagues to identify the persons they do not know:
Well when I see a bunch of young people trying to tamper with something, that’s something I notice. When this
happens in my area for example, I would go to them on a very normal way and talk to them about commonplaces,
ordinary things. And through this dialogue I will see if it’s interesting to control them or not. Well, I wouldnever go
to them and to let the ‘big man’come out, I’m here now and you all give me your ID now. (Community police
officer, Kalk, Cologne)
As long as they simply hang around somewhere in the area where they grew up, and they hang around everyday,
all the time, then I don’t need to control them. I can ask some people, listen, he looks like that and that, well I ask
older colleagues, community police officers, he looks like this and that I think he is from Turkey, he is always
wearing that and that, who is he? (Patrol officer, Kalk, Cologne)
These situations reveal that common practices of restraint may also rely on different logics of
specific units. The community police officers in Cologne have an interactional notion of their work:
they must be accepted by the youths to get information, adapt their approach depending on the
current interaction. Patrol officers have a more distant relationship, but they may avoid unnecessary
control by getting information by other means. They also have, in general, more reactive tasks in
which they have less time for proactive activities, thus reducing any opportunity of discretionary
control.
In France, reflexivity about the antagonistic effects of control is rather limited. The interaction 91
mentioned above is emblematic of a conception in which ID checks and searches are unquestioned
routine interactions. Control as a professional routine may also be illustrated by the interaction I122:
I122 (Lyon): The transit police stop two males (around 25 years old) in a subway station, dressed in a sporty
manner and ask them for their IDs. They say they are social workers and are taking a group of kids out. They
are indeed with a group of five kids. ‘OK, but you still need to have an ID,’says one of the police officers. The
social workers say they don’t have them. The police officer lectures them about the fact that it’s irresponsible
to not have their ID while taking care of children. The two men are a bit irritated but acknowledge the police
officer is right. They let them go.
POLICING AND SOCIETY 7
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

The decision to control is taken rapidly by one police officer, although there is no particularly sus-
picious behaviour. When he becomes aware that they are social workers accompanied by kids, he
decides to teach them of the necessity to carry their IDs (although he could again have taken
them to the police station). Another example illustrates a weak understanding of antagonistic effects:
I49-52 (Grenoble): After a rather calm shift, the patrol officers wanted to show some ‘real police activity’and
perform some stop and search in an area known for being near a drug-dealing place. In about 10 minutes,
they controlled five males, whose one of them will be particularly tense.
These interactions were handled professionally by police officers (calm and respectfully). In an
informal debrief after these interactions, they expressed anger, emphasising that their authority
was constantly challenged, but did not reflect on the fact these controls had been completely
random.
This gives us a rather contrasted image of the French case: police officers have to deal with tense
interactions when they operate controls (and they professionally handle situations most of the time),
but most of them have a limited reflection on the resentment that the repetition of controls may
engender, nor on the very limited hit rate of these controls. This leads us to a paradox: if French
police officers admit that ID checks may degenerate, they still maintain that to control is part of
their powers and as such their use should not be discussed. In Germany, the power to control
seems to be handled in a more flexible way depending on workload, situational factors and behav-
iour of persons concerned. Additionally, most of the police officers are aware of negative outcomes of
control practices. This leads to the reserved control patterns we have observed.
3. Controlling minorities
The second aspect of the research concerns the targets of discretionary controls. If contacts focus
mainly young males in the two countries, the main question is on the potential overrepresentation
of minorities (see for England, Bowling and Phillips 2007; for the US, Rice and White 2010). Here again,
our findings are quite divergent between France and Germany.
3.1. An overrepresentation of minorities? Evidence from different methods
We address this crucial question by drawing on both the observational and the survey data collected
as part of our study. The observed control interactions can with some caution be categorised accord-
ing to the ethnic backgrounds of the controlled persons. However, it seems appropriate to be careful
and to avoid false inferences from the visible appearance of persons. We therefore use the rather
broad categories of ‘assumed natives’(French resp. German) and ‘visible minorities’which in
France are predominantly persons from Maghrebin and Sub-Saharan African backgrounds and in
Germany from Turkish or Italian backgrounds. During participant observations in Cologne and Man-
nheim, we were able to distinguish ethnic origins only very broadly, that is, Eastern vs. Southern
Europe origin. In police jargon, all people from Southern European, Turkish and Arabian backgrounds
alike are simply called, Southern’, with skin and hair colour being the essential criteria. ‘Mixed’
denotes groups of persons from different backgrounds
10
(see Table 1). In France, most discretionary
controls are concentrated on minorities (64 out of 80), whereas in Germany, the proportion is very
much equal between the majoritarian population and visible minorities. In terms of their relative
shares of all observed interactions by ethnic groups, 31.2% of interactions with visible minorities
but only 14.1% of interactions with assumed natives in France are discretionary controls, whereas
the shares are around 12% for both groups in Germany.
These results are in line with those of the standardised school survey which are displayed in Table
2where the prevalence rate of police-initiated controls over the last 12 months is ca. 30% in Germany
for all male adolescents regardless of their ethnic group, whereas in France this rate is ca. 20% for the
native French and 39% for boys from Maghrebin origins (see also for convergent results, FRA 2010).
8J. DE MAILLARD ET AL.
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

We concentrate on male adolescents here because they are across ethnic groups more likely to be
controlled by the police. Looking to high-frequency controls (five or more during last year) only,
the Maghrebin boys even have a threefold higher likelihood compared to native French boys
(17% vs. 5%), whereas no differences can be observed in German cities. Not reported here, these strik-
ing differences between France and Germany and the strong overrepresentation of migrant groups
in France remain after controlling for a range of sociodemographic and behavioural factors. Taking all
quantitative evidence together, we find strong indications of ethnic profiling by the French but not
by the German police forces. Our results give also credit to the hypothesis of a biased-policing of
French police officers (in particular see Jobard et al.2012 for systematic observations of 525 identity
checks conducted in several Parisian locations).
11
Such a discrepancy between France and Germany is also evident if we look at interactions in a
more qualitative way. Three configurations can be identified (as overcontrol is more frequent in
France, most of these configurations will be mainly illustrated by French examples). The first one is
some specific interactions in which discrimination, if impossible to prove, may be presumed. One
interaction (91, Lyon) mentioned above is characteristic: two police officers check the IDs and
comment on the ‘lucky day’of two Arabs who came to ask for directions. The stereotypes associated
with drug dealing by young Arabs constitute an underlying cause. The clothing, the location, the time
of the day or the workload of the patrol may be cause for controls combined with the physical
appearance. The case below (I12) is typical of this uncertainty: although the observer may speculate
on a potential discriminatory reasoning by the police officer, the situation in itself does not offer any
clear evidence of discrimination. It shows, however, a very low threshold for determining a suspicious
behaviour (see above):
I12 (Grenoble): A 16 years old boy from Maghreb is walking with a backpack in his hands. The police check him, to
make sure it is not stolen. He has no ID; the police take him to his place to check his ID.
I59 (Cologne): Two juveniles (Turkish background, streetwear style) are talking to each other in the front of a park;
the officers decide to control them without any obvious reasons.
In the second configuration, overcontrol results from targeting practices of specific police units.
Direct observation is here again useful to identify a series of discretionary controls in which minorities
may have been targeted. Two examples may be given: during their shift, a uniformed patrol unit in
Grenoble had eight encounters with the public, all of them with minorities, with five discretionary
controls (May 2011). In the second example, a transit police unit in uniform in Lyon had eight inter-
actions during their shift, all of them with minorities (November 2012). The four discretionary controls
undertaken in the first part of the shift (in a train) were all targeted at minorities.
The third configuration results from the poor relationship between the police and minorities:
discretionary control stems from prior tense interactions (which refers to the second motive: assert-
ing authority). In France, several controls have been initiated in response to ironic or defiant
Table 2. Police-initiated discretionary contacts during last 12 months by male adolescents (self-reports, by ethnic background).
Countries of origin NTotal prevalence
Frequency of contacts
1–23–55+
Cologne and Mannheim
Native German 1601 29.8 19.4 6.2 4.2
Turkey 706 30.5 19.3 6 5.2
Europe 344 25.4 13.1 7.6 4.7
Other 405 29.9 17.3 6.7 5.9
Mixed native/migrant 404 33.6 21.5 6.9 5.2
Lyon and Grenoble
Native French 3300 19.7 11.7 3.5 4.5
Maghreb 911 39.1 14.4 7.9 16.8
Europe 366 26.5 13.7 4.9 7.9
Other 686 26.6 11.8 5 9.8
Mixed native/migrant 1091 27.1 13.1 5.2 8.8
Note: POLIS youth survey, boys only (N= 3460 in Germany, N= 6354 in France).
POLICING AND SOCIETY 9
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

gestures of minority persons, whereas no equivalent cases could be found in Germany. In these
cases, the interaction often starts with a more symbolic exchange (police and youths watching
each other for instance) and then deteriorates (from irony to insults), leading to a control. These
types of control, caused by the sometimes rather defiant attitudes of young minorities, arise
more generally from the long history of strained relationship between the police and some seg-
ments of young minorities. For all these reasons, apparent ethnicity, especially in France, is one
of the factors to initiate a control.
3.2. Explaining overcontrol: stereotyping, suspicion and control
The main interpretation relies on the link existing between specific stereotypes, the formation of sus-
picion and course of action (Bowling and Phillips 2007, p. 957). In both countries, police officers have
stereotypes towards the public. In Germany, most of the interviewed and observed police officers also
have ethnic stereotypes, as this quote from a patrol police officer illustrates: ‘When I think about it
quickly …the Germans steal scooters, the Russians booze and beat others and the Turks deal with
drugs.’But the linkage between stereotypes and the course of action differs in the two countries.
An idea commonly shared by French police officers is that the minorities would be more often
delinquents, and it would therefore be more rational to control them. In various informal discussions,
police officers explained why overcontrol is rational: it is simply the best way to improve their hit rate.
Speaking of his colleagues as kids receiving candies, a sergeant from a juvenile unit in Lyon explained
us: ‘If you give a kid two bags containing candies and tell him that one bag has eight sweet candies in
it, the other only three, which one do you think he’s gonna choose?’.
Three illustrations taken from our observations give an idea of the rationalities of overcontrol of
ethnic minorities. The first one is the classical linkage between stereotyping, the formation of suspi-
cion and control, without any underlying hostility.
The van is passing the police car in the town centre of Lyon. The driver seems to have a coloured skin. ‘How does
the driver look?’asks one of the officers. ‘I think it’s a Pakistani,’retorts the sergeant. ‘It would interest me if it was a
Gypsy’, replies the driver. We follow the van, and stop it a few blocks away. (Field notes)
In this case, there is no hostility expressed towards minorities, either before, during or after the
control (and this will be confirmed during the remaining of the shift), but traditional crime-fighting
reasoning. We may have equivalent situations in Germany:
The officers stop a car driven by a Roma’s woman. They justify the stop with the fact that they thought the kids on
the back seat were not belted (but this is not the case). The woman is already known as Roma and by name by the
police officers. (I28 Cologne)
The second rationale is also a consequence of stereotyping that conducts to repeated controls. In
a train (see above), two police officers start controlling three young Arabs (one is in on parole or con-
ditional release), then check the ID and the train ticket of an Afghan asylum seeker (whose papers are
in order), then control two Arabs (both with criminal records, and a bit of hashish), the fourth one
targets two Arabs one of which has a criminal record and no papers. The police officers do not
arrest any of them for the small amount of hashish (although they could legally have). Answering
a question of the observer on how they choose to control one person rather than another, the ser-
geant commented: ‘It’s a matter of look, attitude and anyway 80% of offenders are Arab, there is
nothing racist in that, it’s a reality.’Interestingly, all interactions were handled softly; the sergeant
did not hesitate to cool the atmosphere when the youths controlled expressed exasperation.
The third situation illustrates a control in which the rationale of control is affected not only by
stereotypes but also by an implicit hostility. A police officer stops a damaged car and controls the
driver (who happens to be an Arab) near Grenoble. The control is courteous and the police officer
chooses not to fine the driver for a minor infringement to the traffic regulation code. When debriefing
informally about the reason of the control, the police officer comments:
10 J. DE MAILLARD ET AL.
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

I chose this car because I saw that it was badly damaged …and I also saw the hand of Fatma, thus that means it’s
an Arab or an African, they often put it to avoid that their car could be f. up …These people often don’t have their
papers in order …so for us, it is a good hit rate.
The police officer admits having targeted the car because of a religious sign and on the basis of
stereotypes (‘they often do not have their papers in order’) but he adds an element of negative preju-
dice towards this group by implying the hand of Fatma would be a way of avoiding to get their car
damaged by (implicitly) other Maghrebins or Africans. In this case, statistical and categorical discrimi-
nation (Reiner 2010) are blurred. From this point of view, racism (as a set of belief and sentiments) and
discrimination (as attitudes and practices) may be tied for a minority of police officers.
In Germany, there seems to be a more differentiated idea of suspicion not predominantly focused
on ethnicity. Rather, many police officers associate certain crimes with visible attributes as clothing.
For example, ‘street’and ‘hip-hop’fashion is seen as a signal for drug-related crimes. Lifestyle and
other behavioural criteria appear to be more relevant than ethnicity for the decision to conduct
an identity check. And those criteria are strongly connected to notions of place: lifestyle signs that
are associated with drug-related crimes become salient at places known for those crimes. Place is,
in particular, relevant for the proactive work of community police officers, as this quotation illustrates:
‘When we patrol the streets for identity checks, we selectively approach venues that are already
known for drug dealing juveniles.’The same approach also applies to other crimes such as graffiti
and vandalism. In addition, dress styles may influence officers’decision-making by transporting
messages of social status which are flatly equated with decency: ‘I would decide whether they are
well-dressed –come from a good family –or whether they are dressed like juveniles who fuck up.’
In contrast to Jobard and Levy (2009, p. 32) who identified clothing as a ‘racialized variable’because
they mainly observed minority youths wearing typical styles associated by the police with crime, in
Germany, style patterns are not primarily ascribed to ethnicity but rather to social disadvantage. This
may reflect the fact that urban areas of concentrated disadvantage in Germany are characterised by
ethnic diversity.
4. Discussion and conclusion
Stopping, searching, frisking and arresting people are activities common to police all over the world.
It is therefore crucial to understand comparatively how, when, where and with whom this process is
accomplished. In this perspective, our research stresses four main findings.
The first one is a note of caution: the coherence and homogeneity of national practices, values and
professional norms must not be overestimated: there are intra-national contrasts dependent upon the
type of units and the profile of the police officers considered (for similar results in England and Wales,
see HMIC 2015). Community police officers in Cologne avoid control most of the time and have an
individualised knowledge of youngsters, although patrol officers in Cologne and Mannheim may
check IDs but in more reactive ways. Variations have also been identified in the French case, some
experimented police officers showing caution in resorting to control. However, our study is based
on more than one monograph in each country which strengthen the possibility of generalisation.
The second main finding is that two dominant professional orientations can be distinguished. In
France, a more proactive street control style dominates, whereas in Germany, a more informal and
reactive style of policing prevails. The categories of Wilson (1968) are useful to set these professional
practices into perspective. German police officers could be said close to the order maintenance style:
their concern is to ensure order in the community, operational autonomy and informalism are valued.
Police officers will, for instance, look for other ways to collect information than resorting to IDs checks.
French police officers would be closer to the legalistic style: they give priority to law enforcement,
relations and interpersonal contacts are more limited. However, as we have seen, the same French
officers often underenforce the law (either for limited possession of hashish, for the absence of IDs
or for infringement to the Highway Code). There is a paradoxical formalism, with a crime control
POLICING AND SOCIETY 11
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

orientation on the one hand and recurring practices of underenforcement on the other, that would
deserve further investigation in future research.
Thirdly, interactions with minorities display significant differences. In France, interactions are
characterised by a pattern of overcontrol of minorities, whereas proportions are evenly distributed
in Germany. Explaining these variations is a difficult task, as often discrimination is based on physical
appearance mix with other social attributes (age, sex or clothing). Direct observation and informal
talks with police officers following controls provide support for the existence of a linkage between
stereotypes, formation of suspicion and decision to control, although in Germany, police officers
show a great deal of care in dealing with these issues. The existence of heterogeneous practices
according to units, especially on the French side, is also a dimension to be taken into account.
If this research has mainly focused on the descriptive dimensions, an important question is to
relate these practices to organisational policies, policing policies undertaken and, more globally, to
the relations between the police and the public. Our reasoning is here more speculative, but three
different sets of reasons (politico-institutional, professional and social) may be put forward.
The first reasons are the policing policies and mandates existing in each country. The priority given
to the ‘fight against crime’and the ‘culture of performance’in policing policies in France between
2002 and 2012 have encouraged practices of control, as a way of getting immediate results in the
fight against crime (policies whose effects have been discussed in the literature, see Roché 2012,
Jobard and Maillard 2015). More generally, the interpretation of the police mandate is broader in
Germany than in France: police is often described as ‘dein Freund und Helfer’(‘your friend and
helper’). This orientation towards the public also translates itself into training policing: ‘interaction
with the public’is a component of training curricula in police schools in Baden-Württemberg and
North Rhine-Westphalia, whereas interactions with the public are rather seen through the angle of
the protection of the police officer in France. Cultural and religious diversity are also addressed in
German police schools but not in French ones as a dedicated and well-identified set of courses.
Therefore, policing concepts/styles vary in the two countries with regard to the police mandate,
the content of training or the value given to the relation with minorities.
This leads to our second series of reasons, which are the implicit values, norms and assumptions
that guide police officers. Even if too homogeneous conceptions of national police cultures should be
avoided, several striking differences appear: controls are seen as a legitimate way to exercise power
and as a multi-purpose activity (control crime effectively, assert authority and collect information) in
France, although the antagonistic effects of controls are poorly internalised. In Germany, there is a
heightened awareness of these negative effects. These reflections are shared by police officers
(more or less extensively) and internalised through a process of socialisation. It is also to be noted
the ‘we’/‘they’divide was more commonly expressed by French police officers during interviews.
One may even say that policies directed at ID checks are rather made of non-decisions. In France,
stop and search and ID checks are not registered by police officers in any systematic manner, there
are no guidelines on how to conduct them, no immediate supervisors, and, even more, administra-
tors have a poor knowledge of the controls undertaken. During our observations, it has never been a
topic of discussion and debrief between street police officers and their supervisors in France as in
Germany. As Jobard et al. comment (2012, p. 357), ‘ID checks are a widespread practice, and, para-
doxically, benefit from a legal cloak of invisibility.’In both countries, no clear policy exists relating
to the practice of control. However, discretionary checks are less relevant in daily work in Germany
compared to France. This is to be explained by several factors that reflect a different cop culture.
German police officers are not always ‘hunting for the good case’(Lukas and Gauthier 2011,
p. 191). Patrol police officers are mainly concerned with conflict resolution by handling citizen com-
plaints and emergency calls, thus time for proactive controls is limited. Community police officers
indeed are in charge of proactive controls but their work is predominantly focused on the mainten-
ance of [good] relationships with the local public. Their habitus is consistent with watchman style
(Wilson 1968, p. 141). In consequence, crime fighting is not considered as a primary task in their
everyday work (Hunold 2015, p. 443).
12 J. DE MAILLARD ET AL.
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

This leads us to our final comments. The level of trust and legitimacy of the French police is lower
than in Germany (Hough et al.2013, see also Lévy 2016). Perceptions of unfairness are also in disfa-
vour of the French police (FRA 2010, Hough et al.2013), as are the experiences of violence during
interactions by the police (Oberwittler and Roché 2013). These various elements raise serious suspi-
cions on the causal link between the crime control style of policing prevailing within the French
police and its lack of social legitimacy. Our data imply that by multiplying controls, French police offi-
cers put themselves in situations where they have to deal more frequently with defiant attitudes and
conflicts.
Notes
1. In France, ID checks are regulated by the article 78 of the Code of Criminal Procedure. In Germany, they are
defined by the Federal Code of Criminal Procedure and by the police laws of the different Länder.
2. These suspicious behaviours were identified through direct observation and/or informal debrief with police offi-
cers. Among the interactions with an ID check and/or search, we have excluded: (a) controls resulting from a clear
law breaking (for instance the absence of valid tickets and IDs when controlled in public transport, breach of the
Highway Code or –more rarely –criminal act) and (b) controls resulting from a suspicious behaviour (such as
someone throwing away something as the police arrive, an apparently too young person driving a car,
persons involved in a violent dispute, etc.)
3. Grant reference: ANR-08-FASHS-19, Pacte research unit (Sciences Po, CNRS, University of Grenoble Alpes) and DFG
AL 376-11/1, Max Planck Institute for Foreign and International Criminal Law, Freiburg.
4. Police are in the hands of Länder (apart from the Border Control police and the Criminal Federal police, but they
are not studied here).
5. By police–citizen encounters, we mean more or less durable contacts (either verbal or physical) between police
and citizens. It can be a mere verbal exchange or a sequence of interactions entailing verbal and physical
contacts.
6. In France, the survey sample is representative of the young adolescent population in the metropolitan area of
Grenoble and of Lyon which are the major municipalities of the second largest and wealthiest region of
France, Rhône-Alpes. The sample population resembles urban France in terms of age, sex and school level. It
departs from the rural parts of the countries in terms of SES since such areas display more farmers or workers
as parents’professions than the urbanised sectors. In Germany, the survey population is representative of the
young adolescent population in the cities of Cologne and Mannheim, two large cities in the western part of
Germany with very high shares of migrant populations. The sample of schools includes public and private sec-
ondary schools of all academic levels excluding those for special needs. Thanks to the very large sample sizes
(overall ca. 13,500 in France and 7000 in Germany) and the high share of adolescents with migrant backgrounds,
the survey offers unique opportunities to analyse the experiences and attitudes of adolescents from ethnic
minorities (as well as of native adolescents) without quickly running into the problem of small numbers.
7. The work of youth officers is usually characterised by the supervision of serious juvenile offenders or specific
controls of young people during special events.
8. In Cologne and Mannheim, we did not observe plain-clothes units whose main mandate is to search for drugs.
These units may have more crime control-oriented practices than the ones we have observed.
9. In contrast to France, the concept of community police is well established in Germany. The 16 states (Länder) and
to some extend local police forces are responsible for the implementation of community policing. In Cologne,
each police precinct employs several community police officers who are in charge of local networking (e.g.
school visits) and crime prevention by proactive police work. The community police officers are frequently in
contact with local juveniles. Usually, they have already worked on their beat for several years or decades. In Man-
nheim, community policing is less elaborated, but ‘reach out’work with adolescents is an important task of special
youth officers whereas patrol police officers are in charge of proactive police work as identity checks.
10. Contrary to other analyses, we are not here capable of comparing these numbers to the available population at
the time of control. If this limits the reliability of results, the use of the available population as benchmark has
been discussed. And our method authorised more qualitative input in the analysis (see below).
11. After controlling for the benchmark population, the authors demonstrate that ‘depending on the location, a Black
was 3.3 times more likely to be stopped than a White, with similar figures for Arabs (1.8/14.8 times more likely)’
(Jobard et al., 2012, p. 364).
Disclosure statement
No potential conflict of interest was reported by the authors.
POLICING AND SOCIETY 13
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016

Funding
This work was supported by Agence Nationale de la Recherche and Deutsche Forschungsgemeinschaft.
References
Banton, M., 1964.The policeman in the community. London: Tavistock.
Bayley, D., 1978.Forces of order. Policing modern Japan. Berkeley: University of California Press.
Body-Gendrot, S. and Wihtol de Wenden, C., 2014.Policing the inner city in France, Britain and the US. Basingstoke:
Palgrave.
Bowling, B. and Phillips, C., 2007. Disproportionate and discriminatory: reviewing the evidence on police stop and search.
Modern law review, 70 (6), 936–961.
Bowling, B. and Weber, L., 2011. Stop and search in global context: an overview. Policing & society, 21 (4), 480–488.
Cassan, D., 2011. Une ethnographie de l’intégration professionnelle du gardien de la paix et du police constable. Déviance
et société, 35 (3), 361–383.
Cavadino, M. and Dignan, J., 2013. Political economy and penal systems. In: S. Body-Gendrot, M. Hough, K. Kerzsi, R. Lévy
and Snacken, S., eds. The routledge handbook of European criminology. Abington, MA: Routledge, 280–294.
Devroe, E. and Terpstra, J., eds., 2015. Plural policing. European journal of policing studies, 2 (3), 235–343.
Ericson, R., 1982.Reproducing order: a study of police patrol work. Toronto: University of Toronto Press.
Fassin, D., 2013.Enforcing order: an ethnography of urban policing. Cambridge: Polity Press.
FRA (European Union Agency for Fundamental Rights), 2010.Data in focus report: police stops and minorities, report.
Gauthier, J., 2015. Originescontrôlées. Polices et minorités en France et en Allemagne. Sociétés Contemporaines, 97, 101–127.
HMIC, 2015.Stops and search powers 2: are the police using them effectively and fairly? National report.
Hough, M., 2013. Procedural justice and professional policing in times of austerity. Criminology and criminal justice, 13 (2),
181–197.
Hough, M., Jackson, J., and Bradford, B., 2013. Trust in justice and the legitimacy of legal authorities: topline findings from
a European comparative study. In: S. Body-Gendrot, M. Hough, K. Kerezsi, R. Lévy and S. Snacken, eds. The routledge
handbook of European criminology.London: Routledge, 243–265.
Hunold, D., 2015. Ethnic minority recruitment and policing ethnically diverse neighbourhoods in Germany. European
journal of policing studies, 2 (4), 422–441.
Hunold, D., Oberwittler, D., and Lukas, T., in press.‘Can I see your identity card please?’Negotiating authority in police-
adolescent encounters. Findings from a mixed-method study of proactive police practices towards adolescents in two
German cities. European journal of criminology.
Jobard, F. and Lévy, R., 2009.Profiling minorities a study of stop-and-search practices in Paris. Available from: http://www.
opensocietyfoundations.org/reports/profiling-minorities-study-stop-and-search-practices-paris.
Jobard, F., and Maillard, J. de, 2015.Sociologie de la police. Paris: Armand Colin.
Jobard, F., et al., 2012. Measuring appearance-based discrimination: an analysis of identity checks in Paris. Population-E,
67 (3), 349–376.
Junger-Tas, J., et al., 2012.The many faces of youth crime. Contrasting theoretical perspectives on juvenile delinquency across
countries and cultures. New York: Springer.
Lappi-Seppälä, Tapio, 2011. Explaining imprisonment in Europe. European journal of criminology, 8 (4), 303–328.
Lévy, R., 2016. La police française à la lumière de la théorie de la justice procédurale. Déviance et société, 40 (2),
139–164.
Lukas, T. and Gauthier, J., 2011. Warum kontrolliert die Polizei (nicht)? –Unterschiede im Handlungsrepertoire deutscher
und französischer Polizisten. In: D. Oberwittler and R. Behr, eds. Polizei und Polizieren in multi-ethnischen Gesellschaften.
Pfaffenweiler: Centarus, 113–117.
Mawby, R. I., 2008. Models of policing. In: T. Newburn, ed., Handbook of policing. London: Willan, 15–40.
Muir, W. K., 1977.The police: streetcorner politicians. Chicago, IL: Chicago University Press.
Oberwittler, D., 2004.Multilevel analysis of neighbourhood contextual effects on serious juvenile offending. The role of
subcultural values and social disorganization. European journal of criminology, 1 (2), 201–235.
Oberwittler, D. and Roché, S., 2013. Experiences, perceptions and attitudes. ESC newsletter,3,4–10.
Rice, S. and White, M., eds., 2010.Race, ethnicity and policing: new and essential readings. New York: NYU Press.
Reiner, R., 2010.The politics of the police. 4th ed. Oxford: Oxford University Press.
Roché, S., 2005.Police de proximité. Paris: Seuil.
Roché, S., 2012. Les politiques de sécurité intérieure. In: J. de Maillard and Y. Surel, eds. Les politiques publiques sous
Sarkozy. Paris: Presses de Sciences Po, 237–258.
Van Maanen, J., 1978. The asshole. In: P. Manning and J. Van Maanen, eds. Policing. A view from the street. Santa Monica,
CA: Goodyear Publishing Company, 221–238.
Weber, L. and Bowling, B., 2011. Stop and search in global context. Introduction. Policing & society, 21 (4), 353–356.
Wilson, J.Q., 1968.Varieties of police behaviour. Cambridge, MA: Harvard University Press.
14 J. DE MAILLARD ET AL.
Downloaded by [Jacques de Maillard] at 13:01 26 June 2016