- 1 Was able to convey that “insecurity is less of a problem than a solution” (Van Campenhoudt, 1999).
1Following the presidential and legislative elections of 2002, Nicolas Sarkozy was named Minister of Interior and, during the five year term of Jacques Chirac, occupied this position for more than three and a half years (from May 2002 to March 2004 and from June 2005 to March 2007). In the political and media context of the early 2000s, years marked by a continuous escalation of the theme of « insecurity », the substance of the speeches given during and after the electoral campaign was well known and, moreover, not very original when seen in historical context. That a right-wing Minister of the Interior began his term with a « crusade » against crime in order to « finally » bring security to the citizens of France, is not in itself surprising1.
2That he announced and passed new laws intending to reinforce the material, human and legal capabilities of law enforcement is not more suprising. On the other hand, that a Minister of the Interior intended to introduce in his administration a « new management » and impose upon some 237,000 gendarmerie officers and policemen a real « culture of results » is more original and particularly interesting to evaluate. Extending and enlarging upon work begun by Dominique Monjardet (2006), in one of his last texts, concerning the years 2002-2004, we want here to return to the five years during which this « new safety policy » was deployed, to try to evaluate it according to its own « performance » criteria.
3We will first review the official report on the safety policy in place since 2002 and also review the statistical tools available to evaluate this policy (I). We will then examine the evolution of recorded crime (II), then the evolution of the repressive activities of the gendarmerie and the police (III). In conclusion, we will consider the assumptions at the basis of this safety policy, the nature of this « new management » as well other aspects that were also part of the government’s conclusions.
4In his press conference of January 11, 20072, the Minister of the Interior congratulated himself for the results obtained since 2002, particularly in four areas : 1) a « drop in crime » (especially relating to « street crimes») obtained in spite of the continuing rise in personal assaults, 2) an « increase in the activity of the services » perceived as criminal acts identified through police initiatives (rather than through victim complaints), 3) a rise in the closure rate (« while we identified only one criminal in four in 2001, we now arrest more than one in three »), 4) a rise in the number of people arrested and then arraigned.
- 3 As we have already indicated, several years ago (Mucchielli, 2002, 155), this counting of “victims (...)
5Comparing the results of the left wing government that came before (1997-2001) and that of his own (2002-2006), the Minister declared : « While crime has increased by 17.8% between 1997 and 2002, it has fallen 9.4% since 2002. That is to say, that since 2002, more than 1,153,000 victims have been avoided »3. And all of this seemed to him to result from the new way of managing law enforcement that he ordered upon his arrival at Place Beauvau:
« Since 2002, my determination has been total : I have made safety the absolute priority of my ministerial activity, granting the security forces the necessary means, by modernizing them and, above all, mobilizing them. I avoided resignation and renunciation ! Here are the results. They are to be credited to the the police and national gendarmerie. I want to thank these servants of the State for the work they have carried out, without rest ».
6And so, the speech given by the man who a few months later went on to become the President of the Republic, consisted of two observations and an explanatory proposition : the observation of a strong drop in crime, the observation of a clear improvement in the mobilization and performance of police activities, and a proposition explaining the first observation by use of the second. These are the three elements that we are going to test in a systematic examination of the statistical data provided by this same organization (the Ministry of the Interior). We will be adding other data which will allow us to go beyond the limits inherent in the administrative nature of these statistics. But before that, it is important to understand both what this « new management » consisted of and the shaping of the numbers it relied on, as much to lead as to evaluate its action.
- 4 The decree of May 15, 2002 (one week after the proclamation of presidential elections) transformed (...)
7In 2002, after an electoral campaign during which the Right did not stop accusing the Left, in power since 1997, of having lacked « political will », the new government made its priority the lowering of crime. Named as Minister (of the State) of the Interior, and being given an « historic » command of the gendarmerie as well as the police4, N. Sarkozy got down to business as soon as he was appointed. The years 2001 and 2002 had been marked by a strong dissatisfaction as much among the gendarmerie as the police (Monjardet, 2002), dissatisfaction that was conspicuous more for its strongly mediatized public character than for the content of claims made by bureaucrats (denouncing the lack of resources, both material and human, the lack of respect, the feeling of impunity on the part of many « youths », the lack of severity, on the part of the justice system, etc).
- 5 The decree creating bonuses for exceptional results will take effect October 21, 2004. With an env (...)
8With the Law of orientation and programming of internal security (LOPSI), a bill announced as soon as the beginning of July 2002 (and which would finally be voted upon on August 29, 2002), the minister intended to answer the issue of material resources (renovation and construction of premises, increasing the number of bullet-proof vests, weapons-with the introduction of Flash-Balls and, following that, Tasers-, the purchase of new vehicles, payment for overtime, increasing merit bonuses5), and human resources (announcing the creation of 13,500 additional positions in the gendarmerie and the police), all while creating new offences to give more legal tools to law enforcement (Monjardet, 2004). In parallel, The Law of orientation and programming of justice of September 9, 2002, reinforced the legal means for the repression of crime, particularly that of minors.
9In exchange (the reciprocal dimension always being explicit in his discourses), the minister asked the police and the gendarmes for a « complete commitment » in the « war against insecurity » on the one hand, and for the adoption of a « culture of results » on the other-this culture allowing the commitment to be directed by « objectives quantified by the improvement of your efficiency, in terms of reduction of crime, an increase in the clearance rate, and the distribution of your resources for action and training ». The speech of June 26, 2002, given at the Carrousel du Louvre before some 2,000 police cadres (followed by a similar one before the gendarmes), is very clear :
- 6 Some of Nicolas Sarkozy’s speeches as Minister of the Interior are preserved on the Internet site (...)
« Let us hear each other well, so that there are no misunderstandings about the objective : it is not a question of stabilizing crime figures or of obtaining some momentary improvements. We must and we shall reverse the trend. We must and we shall push back crime. […] We cannot accept the explosion of these statistics »6.
10And to reach that objective, each would be assigned objectives, at every hierarchical level, and threatened with loss of career. The Bill of the future LOPSI thus announced, at the beginning of July 2002 :
« A policy of management by objectives has been established. The results obtained in the struggle against insecurity will be regularly evaluated and compared to identified objectives. The local administrators in charge of the police and the gendarmerie will report these results, for their areas of responsibility, and those results will affect the progress of their careers ».
- 7 These measures, thought up at the Paris Prefecture of Police since a study mission in New York in (...)
11Asking to be judged by his results, the Minister announced to his troops that they would be too. And to better oversee the realization of these objectives, the Minister immediately established a monthly analysis of the resulting police statistics (and not just annually). And this, both externally, by transmitting these figures to the press, and internally, by convocation to the minister of prefects (accompanied with the departmental directors of public security, for the police, and commanders of the departmental groupings, for the gendarmerie) from five departments registering the worst results (and from five registering the best), in order to « review and understand the reasons for the developments and to implement an adjustment of resources and of method in order to ensure progress ». In Paris, the prefect of police would even organize weekly meetings with the commissioners responsible for the city districts7.
12This policy and this pressure exerted throughout the police and gendarmerie hierarchy would not be proven incorrect through the first term of Sarkozy as minister (May 2002-March 2004). The speech would be hammered home on every occasion and always in an explicit manner. Thus, for example, on June 24, 2003, before the 53rd graduating class of student police commissioners :
« Ladies and gentlemen, you are aware that to be a superintendant entails many responsibilities. The first of these responsibilities is to achieve the objectives defined by the government. For that, I expect you to be leaders in the culture of results and of assessment. You must establish this culture at the very heart of your teams. And I can tell you that as early as next year, the financial incentives will be in place to reward the most efficient teams ».
13And people should not forget that it was necessary to communicate and publicize this « performance », at all levels of the territory, the minister has to be represented in each member of the security apparatus and each of them having to represent the minister. Thus, for example, the speech of September 26, 2003 in Paris, reminded the prefects gathered together:
« This month I have met more than 20 000 police and gendarmes in the seven zones of defense to tell them what I expect of them, but also to hear their observations. […] I would like to mention a few things. Crime continues to fall […]. It is a success for each one of the police and gendarmes that have committed themselves anew and also of their bosses, therefore it is for you too. These results, I ask you to transmit them each month. I know that it doesn’t always work out that way. Some people believe that a press or Intranet release is enough. They are wrong. You must appear before the press. You must also encourage your departmental Directors of Public Security and your group Commanders to appear before the press as often as possible. Media silence is a mistake ».
14Such is the doctrine and the tools of this « new management of security » put into effect by the new Minister of the Interior, supported by a net budget increase (from the summer of 2002, the Minister announced the power to commit 5.6 billion additional euros between 2003 and 2007) and a management team that was mastering perfectly the cogs of the national police force, understanding well its main bureaucrats and having promoted as much as possible « reliable men » to the main positions of command (Monjardet, 2006, 197).
- 8 The minister was again completely clear in his speech for the inauguration of the OND, October 4, (...)
15Despite the fact that, reaping the fruits of the preceding government’s efforts (it was the Caresche-Pandraud report [2002], in 2003, a National Observatory of Crime (OND) was put into place, theoretically charged with « finally » producing reliable data and avoiding political polemic, the speeches of N. Sarkozy in reality did not stop referring to the single police statistic (called « state 4001 »), in fact they presented it as the ultimate instrument for measuring crimes and offences8. What is it about ?
16What we usually call « crime statistics » in the public debate are in reality the recorded crime statistics of the police and gendarmerie services,which exclude, however, highway infractions, some infractions registered by other administrations (like financial fraud) and above all, violations of the 5th class for which, nevertheless, « the penal treatment is comparable to the least serious offences » (Aubusson de Cavarlay, 1996, 41). This constitutes a very serious problem – nevertheless, one that is never considered by commentators, whoever they are – for the comparison and serialization of police data, and in reality this obliges one to look at legal statistics in parallel. And this detour is very instructive. As indicated in Table 1, the number of violations of the 5th class punished by the justice system outside of highway disputes more than doubled between 2002 and 2006. This concerns assaults against persons (especially voluntary assaults and batteries followed by ITTs -(less than 8 days unable to work), assaults against property (essentially damages), assaults against the environment, and assaults against the economic system (such as the transport or stockpiling of perishable data). Of course, considering the limited number, the integration of violations of the 5th class within police statistics would not change the general direction of the trends (Table 2), but this would still amplify the rise in assaults against persons and contrarily reduce the scale of the decline of damages.
Table 1 : Violations of the 5th class sanctioned by the justice system from 2002 to 2006 (aside from traffic litigation)
Source : Minister of Justice (Statistics from Criminal Records – SDSED)
- 9 The behavior of the victims varies a lot according to the type of infractions involved (for exampl (...)
- 10 Another potential source (but which seems minimal) of divergence between legal and police statisti (...)
17This first problem being identified, what are the recording conditions related to police statistics ? This statistical production is linked at first to the way in which the police and gendarmerie services handle the complaints of victims who come forward9. Firstly, law enforcement can not register the complaint and more or less explicitly rebuff the plaintiff (who can directly bring his complaint to the Public Prosecutor10) ; secondly, simply record the complaint in the local police register ; thirdly, to officially write a statement. Only this last form of handling – if the infraction then qualifies as a crime or offence, but not as a violation – leads to the recording taken into account by the statistics.
18That is why, as rightly said by B. Aubusson de Cavarlay (2005, 10), police statistics do not record the « observed facts » but rather the « facts sent for prosecution » which, for all the reasons we have seen, is not exactly the same thing. This statistical production depends then on what the police « seek and find by themselves » if we can say that, whether through an investigation of the criminal investigations unit, upon the requisitions of the Prefect or the Prosecutor, or often – most frequently – as the product of different daily controls and interventions.
19By all logic (by definition, even), this data cannot but react to changes that intervene in the crimes that are known by law enforcement, but also to their number of officers– and the distribution –, in their proactive modes (their priorities of action) and reactive modes (their ways of dealing with complaints or observed facts). A fortiori when each agent is under the niggling control of his hierarchy in order to obtain good numbers on which his career depends. A researcher, after many experiences of field work with police and the gendarmerie, concludes that : « Within the units of police or gendarmerie, nobody believes in the pertinence of that quantitative data. Each interview brings its share of anecdotes about the best methods for producing ‘good’ numbers, those that are convenient for the bosses, or that allow for the making of categorical claims.
20Furthermore, each level of the hierarchy adds its own corrections. There are also effects created by important variations resulting from apparently simple decisions regarding the legal classification and qualification of facts. […] In brief, the more we see the machinery of policing function, (and) the more we are surprised by the credit given to these numbers as they reflect any non-police reality » (Mouhanna, 2002, 577). Concretely, another researcher, who is also an officer of the gendarmerie, reminds us that the practice of « corrupting » statistics, an old practice, can easily be increased by the locals in charge who are at the foundation of the data collection system (here, the commanders of the brigade) : « There are many ways available to them : reclassifying offences as violations – especially in the matter of damages -, not recording complaints, collecting many offences under a single statistical fact, reducing investigations made upon the initiatives of the police, concentrating efforts on a series of cases occurring outside the district (which allow for the counting of definite facts, while the observed facts had been counted by other units and to improve the rate of resolution of crimes and offences of the brigade), recording errors… » (Matelly, 2002, 204).
- 11 Internal audits, like those produced in 1999 in the gendarmerie, or else the Caresche-Pandraud rep (...)
- 12 Aside from the extracts published in the press (for example, Le Canard enchaine of April 23, 2003 (...)
21A remarkable and henceforth unavoidably detailed inventory of these practices has just been accumulated by these two authors (Matelly, Mouhanna, 2007, 60-86). To these are added official documents recognizing these practices11, as well as many police and gendarmerie agents’ testimonies concerning not only the recording of observed facts but also the usual indicators of « police performance » which are the clearance rate and the number of police custodies12.
- 13 See especially three recent series of field work relating to the gendarmerie: Dieu, Mignon (1999, (...)
22In reality, all of the fieldwork has long been converging and highlighting the lack of reliability in the short term evolutions of police statistics when the results were determined in advance and personnel believed their careers to be in jeopardy. Aside from a few zealots and resistance fighters, the great majority of police and gendarmes, « to keep things quiet »,would give the hierarchy what it wants to see13. Consequently, starting in 2002, how would the statistics react to the unseen level of their political solicitation?
- 14 The gendarmes will resist until January of the following year, after which their military discipli (...)
- 15 This theorem of Demonque (alias D. Monjardet using a pseudonym during the time he worked in and wi (...)
23In the short term, the reaction has been clear and the demonstration has already been conveyed (Monjardet, 2006, 193-194). The statistical rupture was both strong and instantaneous, at least for the police. The minister had barely settled into his office and even before he had had the chance to take the slightest measure, draft the slightest circular or sign the slightest decree, the statistics, after years of steadily rising, showed for the month of May 2002 a miraculous drop of 0.69% in the police zone ( and even of 1.6% for the Paris Prefecture of Police). The following month, while the minister’s policy was still at the discussion stage, the drop became spectacular, reaching 7.38% for the police (8.6% in Paris). At the end of the year, the statistics still indicated a total rise of 1.28%, but that was the « fault » of gendarmes who were not quick enough in carrying out the wishes of the minister14. And then, the main thing was achieved : as hoped for, the trend was reversed, granting to the famous « Theorem of Demonque » its most beautiful proof15. But would this drop continue ?
24Table 2 shows the annual public data published by the Minister of the Interior. The three last columns allow us to immediately visualize the changes that arose between 2002 and 2006. However, we will also use retrospective data (here 1994 and 1998, to abide by the same temporal interval), which will allow us to put the evolution into perspective.
25Between 2002 and 2006, the total number of crimes and offences recorded by the police services and gendarmerie dropped 9.4%, from 4.11 to 3.72 million cases. This statistic returned to the level it had reached in 2000 (as in 1984 and 1991). The figure for 2006 is thus far from being revolutionary, rather, it amounts to an earlier state, in a 25 year period during which it always fluctuated between 3 and 4 million recorded cases.
Table 2 : Recorded cases from police and gendarmerie in 1996 and in 2006
Source : Ministry of the Interior
26However, to comment on this figure and this general evolution makes no sense because the so called group « Crime » comprising more than 100 categories of offences for which the nature, the offenders, the victims and the circumstances are very different and whose causes of increase or decrease are many (we will see that it is the same for clearance figures). Like a global rise, a drop can thus mask multiple evolutions and transformations arising simultaneously in criminal behavior, in their legal incrimination, in their denunciation by the victims and in the actions of the police and gendarmerie. Thus, we need to look at the details. To carry on a systematic examination of 107 types of offences defined by police statistics would be at the very least, tedious. We will focus on three main subsets-damages to property (1), damages to people (2) and crimes of « public order » (3).
27On examining Table 2 we notice that it was essentially thefts that had dropped between 2002 and 2006 (by 17%), whereas many of the other categories of crimes and offences increased, sometimes very significantly. In observing the columns that indicate the frequencies (and not the percentages), we observe that this drop allows us to subtract around 500,000 cases between the two dates. In other words, the thefts are so important numerically that their evolution is by itself enough to determine the general look of the whole. And if we refine the analysis by focusing on the details of different categories of thefts, we will notice that this global decline of thefts results on the one hand, from thefts linked to automobiles and motorcycles (going from 1.08 million to 744,000, meaning a decline of around 31%), and on the other hand, burglaries ( going from a bit more than 432,000 to 335,000, or a decline of around 18%).
- 16 This decline in auto thefts as well as burglaries since the middle of the 1990s is confirmed in th (...)
28To them only, with 1.5 million observed cases, these two types of thefts represent close to 37% of the crime registered in 2002, as opposed to 29% in 2006. It is at first these two subsets that draw down the statistical total of registered crime, authorizing the ministerial claim of global success. However, they forbid accepting the minister’s speech which explained this drop as being due to the efficiency of police and gendarmes stimulated and guided by a new management. Indeed, the table indicates that the drop in auto-related thefts has, in reality, been continuous since 1994. And, if 2001 and 2002 were considered, there would have been the same number of burglaries16. Besides, it is in both cases about crimes that are dependant on the complaints of victims and not on police initiatives.
- 17 The proof is that the tendencies observed in France are rarely exceptional in the European context (...)
- 18 Released in the commercial market in 1996, cell phones were owned ten years later by three-quarter (...)
29In fact, the main factors determining the level of police registrations are completely outside of police action17. In certain cases it could involve conditions related to car and residence insurances. It also involves, and no doubt for the most part, a technical factor : the improved protection of vehicles and and domiciles of owners. The case of car thefts is known (Mucchielli, Saurier, 2007, 55-56) and the impact of this factor is very great. It has a particlar influence on the evolution of certain types of thefts against the private individual (with or without violence), in particular the thefts of cellphones, for which the market has truly exploded over the last ten years18.
- 19 In practice, these two categories are not airtight. When the victim of a purse snatching or the th (...)
- 20 Politically silenced since the end of the 1990s, the connection between crime and socio-economic p (...)
30If certain thefts declined strongly these last years, police statistics suggest that such does not apply to the other predations that touch the daily lives of our citizens. Table 2 indicates a continuous rise since 1994 of thefts with or without violence against private citizens19. This leads us to believe that the thefts continue to develop, particularly in the dual context of a consumer society in constant development and of processes of socio-economic exclusion (school failure. mass unemployment of young people with little or no education) which have been growing for a quarter of a century and have fueled an increasingly marked social and territorial segregation20.
31Finally, it is worth noting that this problem of thefts is not completely separable from that of « violence ». In their execution, some thefts required that their perpetrators threaten or carry out acts of physical violence, especially if the victim refused to let him have his way. The studies of victimization show that in the suburbs of Paris, around half (sometimes locally even more) of all types of violent acts reported by victims are actually linked to theft or its attempt (Zauberman, Nevanen, Robert, Didier, 2006 ; Zauberman, Nevanen, Robert, Lotodé, Fouquet, 2006). This is important data for consideration in the contemporary public debate on « violence ».
32Returning to Table 2, we observe this time a very significant rise in the assaults against persons between 2002 and 2006, from close to 24% (meaning an increase of 72,000 observed cases). We understand that this increase constitutes the achilles heel of the minister’s discourse, with his political opponents and some journalists continuing to insist on this fact to relativize the certainty of the self-satisfied minister. Nevertheless, we will see that the minister’s policy and the action of law enforcement are not more responsible for this increase than they are for the drop in auto thefts and burglaries.
33Let us check the details of these assaults on persons and observe that, with the exception of the homicides, attempts and mortal blows (which have had important annual variations in a trend of global decrease since the middle of the 1980s) these are all of the large categories of offences that have increased between 2002 and 2006. The strongest increases first involve verbal violence (« attacks on the dignity and the personality », which increased 53%, and « threats and blackmail », which increased by 24%). Following that are the voluntary blows and injuries (non lethal), which represent the largest figure, with an increase of 31%. After that are offences against the family and the child, a category which includes abuses of children, offences which relate to the custody of children and the non-payment of child support. Finally, we come to the « attack on mores », which include many infractions of very different types (from pimping to exhibitionism, rapes, incest, sexual harassment and other assaults of a sexual nature).
34In what way is the implementation of security policy responsible for these increases in observed cases ? One is tempted here to come up with a radical answer : in no way. There are four supporting arguments.
35Firstly, Table 2 indicates that this increase in the assaults on people is not limited to the period of 2002 to 2006. The continuity with the preceding periods is, on the contrary, striking.
36Secondly, while they confirm, in a general way, the trends of the police statistics in relation to thefts and burglaries, the victimization surveys contradict them on the matter of interpersonal violence, at least for physical violence. The national victimization survey, conducted each year since 1996 by the INSEE, distinguishes among three types of violence : violence characterized as physical (for which the criterion is the same as that defined by administrative statistics : a temporary [greater than 8 day] inability to work (ITT in French), simple physical violence (where the inability lasts 8 days or less), and verbal violence. Reading the serialization of these surveys (Lagrange, Robert, Zauberman, Pottier, 2004, 312-316 ; Le Jeannic, Vidalenc, 2005 ; Le Jeannic, 2006, 639, Rizk, 2005), we observe that, from the end of the 1990s and until 2005, the proportion of people 15 years and older claiming to have been victims of an assault during the two preceding years is stable, at around 7%.
37Looking at the details, the only rising trend is that of verbal violence. And these observations are corroborated by other surveys in the general population, especially the Health Barometer conducted by the National Institute for Prevention and Health Education (INPES) on the basis of a representative sample almost three times as large as that of the INSEE (around 30,000 people 12 years and older), questioning people about violent acts they had suffered or perpetrated. The comparison between 2000 and 2005 shows a slight percentage drop of people claiming to have been attacked during the preceding twelve months as well as a stability in the percentage of those who claimed to have struck or injured someone (Léon, Lamboy, 2006). It is thus necessary to face the evidence : that even though our society, since the 1990s, has not ceased the alarm about a « rise in violence », the global volume of blows exchanged between people has remained stable. It thus becomes necessary to search elsewhere for the reasons behind the continuing and strong rise being recorded by the police and gendarmerie. And we can offer two, which will constitute our third and fourth arguments.
38Thirdly, it is necessary here to introduce into the analysis a new factor affecting the recording of crime : the legislative evolution. The crime that is pursued by law enforcement and punished by the justice system is defined by the penal law. Flowing from that, when the law changes, crime changes too ; more precisely, when the law changes the definition or even the nature of the criteria which constitute the offences, their recording necessarily experiences modifications. As in the case of technical evolution and its consequences for the protection of property, it is a factor that is generally neglected by the commentators, even though it can sometimes prove to be decisive. And that is the case with assaults on persons.
39Without here resuming a demonstration that was presented elsewhere (Mucchielli, 2008), since the beginning of the 1990s, the legislator has not stopped changing the law in order to toughen the repression of assaults on people, whether they involve physical violence (successive broadenings of the definition of « voluntary blows and injuries »), sexual violence (broadening of the definition of rapes, creating the offences of sexual harassment and hazing) or verbal violence (creation of the offence of moral harassment at work).
40Moreover, it focused its efforts to repress violence against certain categories of persons (minors, « agents of public authority » and, more generally, people « entrusted with serving the public »), certain locations (especially schools) or some circumstances (demonstrations, acting « as a group »). Therefore, the police and gendarmes can now criminalize behaviors that were difficult to criminalize in the past, and now they are presented with victims that could not have been conceived as such (and sometimes not by themselves either) in the past.
41Fourthly, behind this legislative evolution, it is in reality the whole society that for thirty years has effected a veritable transformation in its relationship to violence, by transforming the status of violence, by stigmatizing it more and more, by delegitimizing it more and more. From that, a paradox that is only apparent : the general feeling of an increase in violent behaviors can perfectly well accompany an increase in their denunciation but also of stagnation, even a reversal of their actual frequency. In reality, our society no longer accepts violence, no longer allows it any legitimacy and, finally, no longer recognizes meaning in it.
42This transformation of the relationship to violence concerns all areas of social life, from police violence to interfamilial violence. We perceive there a profound sensitivization of our societies in favor of the protection of the weakest persons, especially women and children, and the rejection of all forms of brutality, often attributed to « male chauvinism » even to masculinity. Therefore, contemporary discourses condemning violence, if they have, doubtless, just a small effect (at least in the short term) on the frequency of behaviors, this effect serves to nourish the increasing denunciations of these behaviors. And from this point of view, the political discourses of the period 2002-2006 are not distinguishable from those of the preceding periods.
- 21 At the Department of the Interior, we refer to infractions revealed by the action of services (IRA (...)
43Whereas the evolution of crimes against property and persons are thus essentially independant of the activities of the police and gendarmes, that is not the case for those that we can include under the term « crimes against public order ». The reason for it is simple : in order to become aware of and eventually pursue those committing theft, burglaries and assaults, the police depend almost completely on the complaints of victims, whereas in the case of crimes against public order the contrary is true, they depend mainly on police initiative ; whether through control on the streets or through the operations of the criminal investigation units (which can take weeks, months and sometimes years before producing results)21.
- 22 We will not refer here to the totality of “destruction and damage to property”, a particular categ (...)
44Therefore, having received real orders (sometimes matched by objectives set in advance, especially related to the policing of foreigners) to show greater repressive efficiency, it is not surprising to observe an increase, sometimes spectacular, in the recording of some of these offences between 2002 and 2006. Three groups clearly illustrate this process, which in 2006 represented some 300,000 cases, or 8% of the total of registered crimes and offences22.
45Firstly, the period from 2002-2006 saw a particularly significant rise (52%) in offences related to the policing of foreigners. We can recognize the effect of the orders given by the Minister of the Interior to harden the control of foreigners found in irregular circumstances. And we presume that the many « illegal work » infractions are also discovered as a result of these controls.
46Secondly, the period is characterized by an increase in drug offences that is also very strong (40%). Looking at it in detail, we observe that it essentially involves an increased penalization of simple drug users (who represent 76.5% of the total infractions from this category and who represent the strongest increase : 43%), knowing that in around 90% of the cases it is smokers of cannabis who are involved. This trend existed before 2002 but has since accelerated.
- 23 Maybe because it exists in the only category of “violence”of which the recording is based complete (...)
47Finally, it is necessary to single out infractions towards people entrusted with public authority (IPDAP), for which the progression is strong between 2002 and 2006 (increase of 15%) even if it is less than the preceding period23. This is a group of acts that has been increasing regularly since the beginning of the 1990s, particularly since the end of this decade and the beginning of the following one (Jobard, Zimolag, 2005). This evolution shows both a continuing degradation of relations between the police and the population (especially certain categories of people who are more often than others the objects of identity checks), as well as the police habit of judicializing the « frictions » arising in particular during controls and stop and search, in relation to the demands made regarding damages that are often claimed in the civil courts, in parallel with criminal procedures.
48The statistics service traditionally publishes a summary of data regarding the repressive activity of law enforcement (clearance rates , the number of people placed in police custody, questioned and imprisoned). This is the second aspect of this official report.
49The increase in the clearing rate is, after the decline in crime, the second of the minister’s arguments to explain the success of his policy. This rate has indeed progressed by 8 points between 2002 and 2006, moving from 26.3 to 34.3% relative to all the recorded cases.
50This increase is more than an evolution : it is a small revolution. If the drop in recorded crime was only returning its global volume to the level already reached seven years earlier, this rise in the global rate of solution would be a true reversal of the historical trend. Indeed, if, from the point of view of the police, it has been a long time that the rate of clearance oscillated between 20 and 25% of recorded cases) ; the clearing rate of the gendarmes had gone from close to 75% in the middle of the 1980s to 32% in 2002 (Table 3).
Table 3 : The evolution of clearances by the police and gendarmerie (1986-2006)
Source : Ministry of the Interior
- 24 From 1986 to 2006, the number of police and gendarmes had, on the contrary, gone from 209,000 to 2 (...)
51But this collapse of clearances in the gendarmerie, as well as the persistent low level of clearances among the police, is not due to a relaxation of practices, to bad will, to a defect of management or to a reduction in staff24. In the 1950s, police and gendarmes solved half of the crimes and offences recorded, in 2002 only a quarter. Meanwhile, two important phenomena occurred (Robert, Aubusson de Cavarlay, Pottier, Tournier, 1994, 79sqq). First, the evolution of lifestyles was marked by a very strong concentration of the population within urban zones, then in peri-urban ones.
52Now, unlike village life or that of a small provincial town marked by strong interfamiliarities and forms of community surveillance in a reduced space from which one only occasionally leaves, life in the large cities and now also in the new cities – even of modest size – peri-urban zones, are marked by anonymity, the absence of community life as well as surveillance of homes and cars, in a much-expanded space that we desert daily to go to work that is further and further away. Secondly, from the 1960s, we witness a very strong increase in thefts and burglaries, which was linked to the spread of new consumer goods. From then on, police and soon after, the gendarmes, found themselves facing crime that was a lot more intense and a lot more anonymous.
53A third indispensible element of analysis : independant of the evolution of techniques (which in reality only play a marginal role for most types of offences), the main element in the success of investigations by the police and gendarmes is the assistance of the population (Mucchielli, 2006). We then understand that they are helpless in the face of a multiplication of anonymous petty crimes, especially theft, burglaries and, more recently, property damage. This evolution has greatly concerned the police and their urban territories since the 1960s, and it has affected the gendarmerie in its turn during the 1980s and above all the 1990s.
- 25 There, where the police culture favors the flagrante delicto (Robert, Aubusson de Cavarlay, Pottie (...)
54From there, the observed disarray – following others, in the last big fieldwork project (Mucchielli, dir., 2007) – of gendarmes nevertheless served by a research doctrine stressing proximity with and intelligence obtained from the population25, but which meet in turn with an accumulation of complaints against X which they are rarely able to satisfy. For all these reasons, the strong rise in the rate of crime clearance since 2002 is more than surprising. To try to understand, let us examine the details of types of offences.
Table 4 : The evolution of clearances between 2002 and 2006
Source : Ministry of the Interior
- 26 It relates to receiving, procuring, drug related offences, offences related to bars, alcohol and t (...)
55The first surprise in studying the details concerning rates of clearance by categories of offence (Table 4) derives from the significant number of those which exceed 100%. As specified by the Reminder Letter of September 6, 1999, sent to all the services of police and gendarmerie by the Central Directorate of the Criminal Investigation Unit, « concerning the indexes 44, 45, 55, 56, 57, 58, 59, 60, 61, 69, 70, 71, 74, 75, 76, 77, 79 and 86, the number of cleared cases must be equal to the number of recorded cases. Indeed, it concerns offences that are recorded and solved at the same time » (cited by Matelly, Mouhanna, 2007, 144). Eighteen out of 107 offence categories would therefore have to reflect a clearance rate of 100%26. Actually, we see them in Table 3 (which details only some of the offences), illegal work in addition.
56What are these « anomalies » the result of ? The official explanation of the Minister of the Interior – that they were « catching up » from one year to the next, some offences could be solved long after their recording – does not hold water. If it is in itself exact (especially in a criminal matter where the investigations can be very long), it does not allow for the consideration of errors that, far from being exceptional and counterbalancing from one year to the next, are on the contrary, constants and accumulate from one year to the next. A classic reason is the double counting of the same offence by several services (Aubusson de Cavarlay, 1997, 2). This reason would suffice to close this parenthesis on the « anomalies » if they would be distributed in a random manner over time.
57Such is not the case. Matelly and Mouhanna (2007, 159sqq) have constructed a series of these anomalies from 1995 to 2005, which demonstrate two things. The first is that these overestimates of the clearance rates have strongly increased since 2003, whereas they had fallen in the previous five years. Thus, the general rise in the clearance rate was accompanied by a parallel rise in « anomalies » leading to a still greater inflation of the clearance rates.
- 27 So it is, after all, a new application of the Theorem of Demonque that we can thus formulate: “The (...)
58The second is that the curve of the overestimate of clearance rates bonds strangely with governmental periodicity. Strong in 1995 and 1996, under a right wing government which had already pressed for police efficiency, the overestimate diminished from 1997 to 2002 under the government of the left, rising again after the return to power of the right, in the context of the « new management of security ». It is therefore difficult to not believe in the existence of internal arrangements designed to help better respond to objectives determined in advance27.
- 28 In his press conference of January 2007, the Minister of the Interior announced that he created ab (...)
59This first enigma resolved, let us return to Table 4 which gives the details of the changes that arose between 2002 and 2006 in relation to case clearance, both in terms of percentages and raw frequencies. All in all. with raw frequencies almost constant28, the police and gendarmes cleared some 200,000 extra cases (198,462 to be precise).In which categories of offences did they achieve this ? Table 4 shows only increases, but among these, the percentages and raw numbers vary considerably. Moreover, a strong increase in the rate of clearance of one offence does not necessarily contribute much to the total if the frequency is weak.
60To facilitate the analysis, we have created Table 5 which presents the « ten higher increases » of the number of cases solved between 2002 and 2006, by distributing them between two categories according to which the offences are discovered by the initiative of the services or through the complaints of the victims. These 10 offences (out of the 107 of this statistical index) explain just about 80% of the supplementary 200,000 cases solved in 2006.
Table 5 : The « ten greatest increases » in the number of cleared cases between 2002 and 2006
Source : Ministry of the Interior
Interpretation : with 36 861 additional cases cleared in 2006, use of drugs contributed 18,6 % to the total number of the additional cleared cases.
61As for the police initiatives, we observe without surprise – taking into account what we have seen of the evolution of recorded cases – that the police and gendarmes have above all raised their rates of case clearance by pursuing more marijuana smokers and foreigners with irregular documentation (see Table 5). By themselves, these two categories of offences explain close to 35% of the 2000,000 supplementary cases cleared in 2006. The rest consist of the IPDAPs, possession of weapons and clandestine labor (which we can assume are tied to the undocumented situations of foreigners), the whole of this explaining 42% of the rise in the number of solved cases.
- 29 Regarding riots, because we cannot determine the role of proactivity (flagrante delicto) and that (...)
62In other words, in terms of offences which are essentially discovered by means of identity controls in public places, we can conclude here that the increase in cleared cases proceeds simply from the intensification of identity controls in the public places and the more intensive verbalization of offences discovered (or provoked) through that. We can also add that this daily intensification of police pressure is probably focused on the populations that constitute « targets » and ordinary « clients » of the police in uniform and of BACs (anti-criminal brigades), in particular the young men from immigration (Zauberman, Lévy, 2003). Indeed, the most intensified daily police surveillance occurring over the five years studied has focused on them, both before and after the riots of November 200529.
- 30 The fifth being simple thefts against individuals, of which we recall (Table 2) that they constitu (...)
63As for the complaints of the victims, Table 5 shows that four out of five cited offences are assaults on people30. With more than 34,000 cases solved in 2006, voluntary blows and injuries contributed 17.5% to the global rise in the number of cases solved compared to 2002. Following that is verbal violence (threats, blackmail and attacks on diginity) which contribute 10.5%. And finally, offences against the family and the child with 4% (being 7,708 supplementary solved cases, of which close to two thirds are made up of offences relating to child custody and non-payment of support). The police and gendarmes have thus solved more attacks against persons where the presumed perpetrators are most often known through the complaints of victims.
64Certainly, these assaults on persons rose in the recorded cases between 2002 and 2006, part of the effect here is due to structure. However, the rates of solution rose in all the categories of assaults on persons, from 4 to 8%. Nothing allows us to think that the nature of these offences changed in the space of five years, it thus remains necessary to explain these superior 2006 performances as compared to 2002.
65Logically, two hypotheses are possible : either these bureaucrats really dedicated more time and energy to these files, or they treated some of the complaints differently (in the police station registers, or in the violations of the 5th class, as we have seen), when they did not simply refuse the victims with the mildest cases in order to concentrate on the others (an hypothesis that could also apply to some of the thefts). The second hypothesis adds to a general question asked at the beginning of this text. The first hypothesis can find support by observing the evolution of police arrests.
66After the counting of the recorded and cleared cases, the annual volumes of the Minister of the Interior offer three series of statistics concerning suspects : the first regarding the numbers held in police custody, the second regarding those under suspicion and the last regarding the number of people incarcerated. This last series is the most fragile. According to the official publication, « most of the police and gendarmerie services count as ‘incarcerated’, people who, after having been arraigned before a judge, are the object of an incarceration », except in Paris and the inner suburbs where they register only arraignments« because of a preliminary filter which is the TGI reception center in Paris ». If we can assume that this precision does not affect the issue of the comparison in time, we remain here at the point where the decisions of the police and the judges meet, the series of those incarcerated referring to the appeal of provisional detention before judgement (Aubusson de Cavarlay, 1997, 4).
67The comparison of the years 2002 and 2006 (Table 6) mainly highlights an increase of close to 40% of the people in custody, almost two times more than the rise in those under suspicion, and almost nine times greater than the slight rise in incarcerations. In more detail, we see that the trend in custodies confirms that of the case clearing because the strongest increases relate to the crimes of public order (foreigners, vandalism, drugs and IPDAP) and the assaults against persons, whereas the weakest relate to thefts. And this is shown in the context of arrests, where the strongest rise relates to the foreigners with irregular documentation and smokers of marijuana (10.5 more numerous than the traffickers), ahead of assaults against persons.
68As for decisions regarding incarceration, their trends are multiple. If the judges partially follow the police in the matter of drugs (in reality, essentially for the traffickers), they show themselves to be more cautious in relation to assaults on persons (incarcerating only in criminal cases and the most serious offences), regarding the group of cases involving foreigners, « urban violence » (incarcerating only some arsonists) and regarding that which implicates the police themselves (the judges clearly perceiving a difference between violence and simple outrage). Finally, in financial and economic matters, they agree with the police very little, and not at all regarding theft (except when the thefts were committed with violence, but we have seen that those increased neither in the recorded nor the cleared cases).
Table 6 : Evolution in the number of police custodies, under suspicions and incarcerations between 2002 and 2006
Source : Ministry of the Interior
- 31 See for example F. Tassel, The minister who doesn’t like justice, Libération, September 22, 2006.
69We guess here that, in general , this growing gap between police pressure and the provisional detentions can only reinforce the traditional accusation aimed at judges by the police, according to which the first ones always free those that the second ones would bring before the law. The denunciation of « laxness in justice » logically follows, repeated strongly over the course of the final trimester of 2006 by a Minister of the Interior while in the final phase of his electoral campaign for the presidency of the Republic31.
70We can, however, suggest a completely different interpretation of this evolution ; indeed, if the strong increase in the number of custodies incontestably means the intensification of police pressure and that the strong increase in the number of persons questioned are logically linked to the increase in the rate of case clearance, this does not at all speak to the gravity of the cases for which the people are brought to justice, nor the solidity of the police investigations.
- 32 An illustration producing a “magnifier effect” (or like a magnifying mirror) of this growing gap b (...)
71Hence, if a good part of the additional cases brought by the police are in reality minor and/or the proofs are docuumented in haste and thus contain more and more procedural irregularities or are insufficient proofs (the price of haste and of the production of figures for the sake of figures imposed by this « new management » of security32), it becomes a lot less surprising that the judges don’t automatically follow the police in the policy of « zero tolerance » imposed on them by their minister and their hierarchy.
- 33 It is particularly true in relation to youth justice (Mucchielli, 2007a).
72At least not in the carceral sense that one would expect. For if the justice system does not stop, in reality, to try and satisfy the political injunction of a systematic penal response, it does it by enlarging the range of measures and sanctions at its disposal, reserving the use of provisional detention for cases deemed to be the most serious and for people deemed to be the most dangerous33.
73In summary, it seems to us that the intensification of police pressure desired by the minister caused a largely artificial increase in cleared cases as well as a strong increase in police custodies and those under suspicion, the whole, concerning for the most part, offences lacking in gravity and more tolerated before 2002.
« Police Commissioner [Prefecture of the police of Paris, co-responsible for the attempt to import the new statistical evaluation system from New York] : We can say that it is because we made more arrests that there are fewer burglaries. There is a correlation…more or less fictitious when we think about it, to be honest. »
Researcher : What ?
Commissioner : Because I don’t believe that burglaries only depend on police action. Of that I am firmly convinced. […]
Researcher : Therefore, in spite of all this work, you still only loosely connect the activity of the police and the evolution of criminality ?
- 34 Conversation recorded in 2004 by Didier (2005, 70-71).
Commissioner : Naturally. It’s a step which allows us to make each one responsible for what does not go well in his district. […] But really, we have to be serious, we are not going to say that it’s the police who, by what they do, create more or less criminality »34.
74When the time comes for the assessment, two empirical conclusions are to be noted. The first is that the evolution of crime is dependant on a multitude of factors, most of which defy police action. Thus, the declining trend seen in car thefts and burglaries has no direct connection to the initiatives of law enforcement ; in the same way, the evolution of reported violence stems from the profound transformations in French society upon which the police and gendarmes had practically no control (except for the intake of victims). The second is that, no doubt, aside from the issue of policing foreigners and drug use, there is no direct correlation between the level of police pressure and the effectiveness of their action.
- 35 It is the opposite that would have been surprising. Indeed, international studies had been converg (...)
- 36 As shown by the general increase of the prison population (from 48,594 prisoners on January 1, 200 (...)
75So the strong increase in controls, identity checks and persons taken into custody certainly allowed-by means of all the arrangements in statistics recording that we examined – the display of performance indicators that conformed to the wishes of the minister, but the increase did not result in a transformation of the crimes that were brought into the justice system (except in the two cited cases of public order). In a general way, the idea that the intensification of police action would be the principal factor (directly by deterrence or indirectly through arrests) in reducing the global volume of crime is thus a false idea35. The only indisputable impact of this policy is the intensified repression of petty crime by the police and, in part, by the justice system36.
- 37 However, the story of this administrative and political enterprise remains to be written (some ele (...)
- 38 The left came to power for the first time in 1981, during the Fifth Republic. Since then, alternat (...)
76And so, the « new management of security » will have had as a paradoxical consequence, the ruining of the credibility of its tool of evaluation. Indeed, that the police statistics are invoked by a Minister of the Interior to evaluate – or at least to provide indications of – his security policy, or that of his predecessor, is not in itself new. Such was doubtlessly not the desired goal when, at the beginning of the 1970s, the Minister of the Interior was able to create this collection of data then published annually until today (thus taking over from legal statistics which had become confidential)37. But it is indeed what these statistics are used for mainly on the political level, above all since the national elections have instituted the alternation of the right and the left and that the theme of security has constituted the heart of the right’s electoral argument38.
- 39 AFP, 30 July 2007.
- 40 This “index of operational evaluation” is an example. It has to be filled out daily and was put in (...)
77Far from introducing some sort of break, and while giving the illusion of understanding the criticisms formulated by the parliamentary mission organized by the previous government (by setting up the OND, with the previously mentioned limits), N. Sarkozy pushed this logic further than anyone in the past, going as far as fixing national objectives in advance, indicating not just a tendency but henceforth a pre-calculated result as well. Having become President of the Republic, he has written to his successor Minister of the Interior, Michèle Alliot-Marie, a mission letter indicating that : « crime must continue to drop. from 5% over two years for general crime, from 10% for street crime » (as well as « the rate of solution for crimes and offences must reach 40% »)39. And to be quite sure of reaching it, aside from the possible statistical arrangements, the local bureaucrats will now see their assessments (and their bonuses) directly connected to the nature of their daily interventions40.
78Therefore, we do not understand very well how the commentators (especially the journalists) can and will be capable again of acting as if they seem to believe in the « beautiful story » that is told to them (Salmon, 2007) and to grant even a minimum of credit to the figures which clearly no longer have any connection to the reality of crime or policing. The police statistics have become purely a tool of internal administrative control and external political communication. On this latter level, the strategies have proven to be very effective.
- 41 In a poll produced on January 27-28, 2006 (BVA, Observer of Political Personalities), to the quest (...)
- 42 No doubt it is not an accident if, in the course of the first semester of 2007, in a new election (...)
79Unable to convince public opinion that his report was positive and his action truly effective41, the minister mostly succeeded in imposing the idea that he had been more active than his predecessors and would act better than his rivals in the future. The fact remains that the price of this strategy (which we can describe as the strategy of scorched earth) is the loss of credibility of the police statistics as an instrument for measuring crime and police performance and, beyond that, a dangerous example of the instrumentalization of public statistics given to the French political class42.
- 43 See above, note 7. The New York experience of reforming police management is original in the sense (...)
80Finally, the policy led by N. Sarkozy from 2002 to the beginning of 2007 requires that we comment for a last time on the original attempt at introducing business management into public administration. To conceive it, N. Sarkozy borrowed from the spirit of the times. The idea of introducing a new mode of management into the higher levels of police forces, by measuring and sanctioning their « performances », is typical of the Anglo-Saxon style of the New Public Management, which developed considerably in the 1990s (Boltanski, Chiapello, 1999 ; Power, 2005 ; Robert, 2007), and which was singularly and famously illustrated in the management of the New York police by its boss W. Bratton at the request of the new mayor, R. Giuliani43. It has already been said that N. Sarkozy was inspired by it. But the French version only amounted to quite a superficial import in the sense that it was based on a triple simplification.
- 44 Aside from the classic work of Skogan and Hartnett on Chicago (1997), one can also look at the man (...)
81The first simplification was in applying a uniform and authoritarian model of management to all of the national territory, instead of attempting to define objectives adapted to local realities in concert with local officials. In other countries, on the contrary, the community police programs experienced at least partial success when they were elaborated in partnership with the locals44.
- 45 The French failure is not an isolated one. In Spain, the reform « Policia 2000 » was based on a le (...)
82The second simplification consisted of evaluating police performance solely on the basis of police statistics, that is to say, on an administrative metric of police activity that the police on the beat fill out themselves and which they would then send up the chain of command. In such a system, « the priority given to the numeric evaluation led to the favoring of quantity over quality », and « the production of ‘good numbers’ progressively overrode the objectives of the police forces » (Matelly, Mouhanna, 2007, 10, 255). Indeed, by definition, such a system could only create submission to the hierarchy and varying re-arrangments of reality, with the foundation of security policy (the reduction of risks for citizens) progressively being lost sight of or, to put it another way, becoming more often dependant solely on the professional good will and conscience of the police officers45.
- 46 It is another well established international conclusion which leads us to assume that the totality (...)
83The third simplification was in reducing not only the evaluation process of police performance but even more broadly, reducing the profession of the police and gendarmerie to its repressive component. Certainly, the actual tools only encouraged and allowed for the « evaluation » of its repressive aspects : solving cases, custodies, arrests, people incarcerated. But the police and gendarmes know that – except of course in the services specializing only and integrally in judicial investigations (criminal policing brigades, research brigades and sections in the gendarmerie) – the largest part of their time is devoted to other missions46.
84The gap between, on the one hand, a global image and an evaluation relevant only to the repressive component and, on the other hand, a daily activity that goes completely beyond this sector, creates a type of general tension of identity that is classic to police and gendarmes. And we know that their professional socialization tends over time to reinforce the focus on the repressive component and this image of « police as hunters of criminals », the only one valued by the institution (Monjardet, 2005, 298). However, other representations of identity, corresponding to other aspects of everyday work, have always coexisted and many law enforcement personnel value their relationships to the population, the values of help and protection (this is especially so for the gendarmes).
- 47 Dominique Monjardet (1996) distinguished among three: the Criminal Investigation Department, the p (...)
85There are in reality several facets of policing47 that therefore require several professional identities and systems for the evaluation of performances. The recognition of these different needs, of these different values, of the necessity to create different indicators of performance and of partnerships not strictly limited to security stricto sensu has thus been at the heart of all the reflections on « community » police (in France, « local » police, or still, « the public peace » police) in Western countries for thirty years. Hence, as has already been understood by D. Monjardet (2006, 199-201), security policy led by N. Sarkozy has doubtlessly and in reality been a regression in the sense that, aside from discrediting police statistics, it threatens to make even more narrow the professional culture of the police, to discredit the idea of prevention, to distance the police institution a bit more from the notion of public service and thus to degrade even more the general relationship between the police and the population.