Keywords

Introduction

Croatia, the twenty-eighth member of the European Union , gained independence from the former Yugoslavia in 1991. Since then, the country has embarked on the road toward establishing a more democratic society, in general, and police, in particular. However, the process has been both complex and challenging (Kutnjak Ivković and Klockars 2004, p. 56):

The journey has been far from smooth; the country has experienced a defensive war against the aggression of the Serb-dominated Yugoslav Army and various paramilitary troops, the influx of refugees from Bosnia and Herzegovina , a decade-long governance of a strong, right-wing oriented political party, the strengthening of nationalism, a continued legacy of mismanagement of the economy, the transition into market economy, and a high unemployment rate. All these factors shape the environment in which the Croatian police operate and have a strong impact on the state of police integrity.

Established in the early 1990s, the Croatian police are one of the youngest police agencies in Europe. The organizational history of the Croatian police can be divided into three periods. In the first period, Croatia was still a republic within the former Yugoslavia ; it did not have the legal right to establish an independent police agency in addition to the existing centralized Yugoslav police. In the early 1990s, as it became clear that the war is imminent, the Croatian Parliament passed statutory changes to establish the National Guard Corps (NGC) as a police service within the Ministry of the Interior . Police officers constituted a large portion of the man power for the NGC and, for a period of almost 2 years, the police performed both the defense role and the regular police role (Kutnjak Ivković 2000). The war-related experience exposed police officers to considerable violence, while resulting in the relaxation of official rules and strengthening of the code of silence (see, e.g., Kutnjak Ivković 2004a).

The second period in the history of Croatian policing began after the war ended in 1995. The emphasis was put on democratization of the police. During the late 1990s, the “war-is-over” attitude began to dominate the police administrators’ views. The primary focus of various reforms was on demilitarization, depoliticization, professionalism, demystification, and downsizing (see Kutnjak Ivković 2004a). Eventually, there was a substantial downsizing in the police from 30,000 to 34,000 sworn in the late 1990s (Kutnjak Ivković 2000, p. 79) to 20,000 sworn in 2005 (Kutnjak Ivković 2005). The legal reform followed as well, with the enactment of the new Criminal Procedure Code in 1997 and the new Police Law in 2000.

The third period, marked by further reforms , started in the early 2000s. In 2003, the Croatian version of community policing has been introduced. The society at large put greater emphasis on accountability of police officers and police administrators, resulting in numerous firings and scandals involving top police administrators (e.g., the prime minister fired both the minister of the interior and the chief of police in 2008; the heads of the Vukovar-Srijemska Police Administration were arrested for corruption in 2013; the head of the Splitska Police Administration was removed from the position for the abuse of official position in 2013). At the same time, the country experienced considerable challenges in its attempt to deal with organized crime and widespread corruption (Nacional 2008), resulting in the decline in the level of public support for the police (Kutnjak Ivković 2008).

The Croatian police are a centralized police agency housed within the Ministry of the Interior. On top of the hierarchical structure are the minister of the interior and the chief of police (Police Law 2000), followed by 20 police administrations, and about 200 police stations at the bottom of the structure. The majority of regular police work is performed at the level of police stations, which can be either basic or specialized (e.g., airport, traffic, border). The police currently employ about 20,000 sworn officers or 4.28 police officers per 1000 inhabitants.

This chapter analyzes the societal and police-related factors in the organizational scheme of police integrity theory. The second part of this chapter provides an empirical analysis of survey data measuring the level of police integrity among the Croatian police.

Theory of Police Integrity and the Croatian Police

This chapter relies on the definition of integrity, the theory of police integrity, and the accompanied methodological approach developed by Klockars and colleagues (see, e.g., Klockars and Kutnjak Ivković 2004; Klockars et al. 1997; Klockars et al. 2004a). Police integrity is defined as “the normative inclination among police to resist temptations to abuse the rights and privileges of their occupation” (Klockars et al. 1997). It assumes that police officers are able to resist various forms of temptations, including corruption, use of excessive force, and other forms of abuse, the rights and privileges, to which policing as an occupation exposes them (Klockars et al. 2006). Police integrity could vary across different forms of misconduct (e.g., police corruption, use of excessive force), as well as different levels of seriousness within the same form of misconduct (e.g., within police corruption, acceptance of gratuities vs. theft from a crime scene). Consequently, the related methodological approach incorporates hypothetical scenarios describing not only a variety of forms of police misconduct but also examples of different levels of seriousness within the same form.

The organizational theory of police integrity (see, e.g., Klockars and Kutnjak Ivković 2004; Klockars et al. 1997, 2001), the theoretical organizational theme in this chapter, rests on four dimensions: quality of official rules, quality of the agency’s own internal control of misconduct, curtailing the code of silence, and the influence of the larger environment. This chapter explores each of these dimensions as they apply to Croatia .

Organizational Rules

The first dimension of the theory argues that the quality of organizational rules and the way in which these rules are made, communicated, and understood by the police are critical for the high levels of police integrity (Klockars and Kutnjak Ivković 2004, p. 1.4). The theory predicts that police agencies of high integrity will not only have organizational rules explicitly prohibiting police misconduct but also teach these rules effectively and enforce them when rule-violating behavior occurs. At the same time, police officers in such agencies should know the organizational rules and also support them. The content of the rules, in particular what behaviors are explicitly prohibited by the rules, and the degree to which the rules are enforced could vary drastically across agencies. This should particularly be the case for less serious forms of misconduct, such as the acceptance of free gifts and verbal abuse .

Because the Croatian police are a centralized agency within the Ministry of the Interior , the same set of organizational rules applies to every sworn officer in the country. The evolution of the laws regulating police work in Croatia was affected by the events in the society at large. In 1990, the first Croatian Constitution was enacted, containing the basic provisions regulating citizens’ civil rights. However, the war temporarily halted the changes in the criminal law, criminal procedure, and other police-related laws. The widespread reform of the criminal justice legislation occurred in the late 1990s.

The Code of Criminal Procedure was enacted in 1997 and subsequently changed several times, finally resulting in the new Code of Criminal Procedure in 2008. Other than simply reinforcing the basic civil rights established by the Constitution (1990), the two Codes of Criminal Procedure introduced extensive changes into the criminal procedure. The Codes revised the predominantly inquisitorial procedure by adding many elements of adversarial procedure. The 1997 Code gave the defendants the rights to be informed about the charges and to receive Miranda-like warnings (Kutnjak Ivković 2004a, p. 202). The 2008 Code of Criminal Procedure authorized the prosecutor to conduct the investigation, established the Office of the Criminal Investigator to assist the prosecutor, introduced plea bargaining, required of the police to audio- and videotape all suspect interrogations, and gave parties the right to select and present the evidence in the case. In addition, the 1997 Code also proscribed the level of force the police were allowed to use during the preinvestigative process (e.g., while executing a search warrant) and determined that, unless one of the conditions for the constitutional exceptions were met, the search of a residence should be performed only on the basis of a judicial warrant.

Another part of the legislation directly regulating police work is contained in the Police Laws (2000, 2011). For the first time, the Police Law of 2000 defined the police as the “public service within the Ministry of the Interior entrusted to perform the tasks enumerated by the law” (Article 2, Police Law 2000). This contemporary view of the police as a public service concerned with the protection of life and property resembles more closely the democratic view of policing. The police tasks are explicitly listed in Article 3. They include traditional police tasks such as the protection of life, rights, safety, and health of individuals; protection of property; prevention and detection of felonies and misdemeanors; and search for and apprehension of the persons who committed felonies and misdemeanors. The police right to use discretion and the police right to use force have been regulated as well (Kutnjak Ivković 2004a). As the Criminal Procedure Law has changed, the need has arisen to have the police activities during pretrial investigation regulated by a separate law, the Law on Police Activities and Rights (2009).

The new 2011 Police Law further streamlined the police toward a democratic agency. One of the most relevant changes includes the newly specified requirements for promotion and advancement, as well as the new procedure for appointment of key administrative personnel. The head of the police is the general police director, a professional (and not a political) function; the director can only be a person who has at least 15 years of experience in policing, out of which at least 10 should be at the supervisory capacity, and holds the rank of police advisor (Article 58, Police Law 2011), the highest rank . The Director is appointed for a period of 5 years and may not be fired at will (Article 59, Police Law 2011). The Police Law establishes the criteria and the process for the appointment of all other supervisory personnel. For example, the heads of the police administrations are appointed for a five-year period after a search conducted by a five-person committee. Once appointed, they are accountable to the director.

As time passed, the laws have become more detailed and sophisticated, but one of the key issues—the frequency with which these laws, should be and actually, are enforced—still remains without a clear answer. Also, familiarity with and the level of knowledge that police officers have about these new rules is an open empirical question. In 1995, a few years after the police had been established and at the time the war was ending, a survey explored officers’ familiarity with official rules (Kutnjak Ivković and Klockars 2004, p. 68). The respondents in the study were asked to identify whether various forms of police corruption described in hypothetical scenarios violated official rules. Although the overwhelming majority of the respondents had no problems recognizing that a theft is a violation of official rules, at most about two thirds of police officers were able to identify the behaviors in the scenarios as rule violating. Kutnjak Ivković (2009) argues that such relatively low percentages are tied to the war-related circumstances, namely relaxed training and rapid hiring.

Results of a study conducted a few years later (Ivanović 2001) suggest that a substantial proportion of police officers, at the time, still had problems with understanding at least some of the rules. A more recent survey of police officers conducted in 2012 reveals that the majority of police officers support the idea that criminal and administrative regulation of police conduct is important for the performance of the police service and indicated that they are well versed with such regulations (Borovec 2013). At the same time, the overwhelming majority of the respondents (95 %; Borovec 2013, p. 6) assess that their knowledge of the rules regulating the use of force is “excellent” or “good,” although, according to their own account, only about two thirds of respondents went through police training (Borovec 2013, p. 14).

Police Detection and Investigation of Police Misconduct

The second dimension of the theory emphasizes the police agency’s own methods of detection , investigation, and discipline of rule violations. These activities could be very heterogeneous, from more reactive activities, such as investigations of corrupt behavior and discipline of corrupt police officers, to more proactive activities, such as education in ethics, integrity testing, and proactive investigations. The theory stipulates that there should be a positive correlation between the existence and use of a sophisticated system of corruption prevention and control, and the level of integrity prevailing in the agency.

The Police Law (2011, Article 5) provides citizens and organizations with the right to submit complaints against police officers. A recent study by the Ministry of the Interior revealed that about 45 % of the surveyed police officers had a complaint filed against them (Borovec 2013). Overall, the respondents evaluated the process that followed as objective. In case a citizen or an organization files a complaint, the heads of the police station or other organizational unit have the responsibility to investigate the complaint and, within 30 days since the date the complaint was filed, inform the complainant about the actions taken. If the complainant is not satisfied with the outcome and/or the procedure taken, the case will be examined by the police complaint board . The board should be composed of one police administrator and two citizens appointed by the Parliamentary Committee on Human Rights and Rights of Minorities. This provision should be the other side of the coin that established police officers’ personal accountability for violation of the laws and official rules (Police Law 2011; Constitution 1990). In fact, to offer an even stronger guarantee, the Criminal Code (1997) establishes that the police officers can be charged criminally if they abuse their office with the purpose of preventing citizens from submitting complaints. Most of police officers in a 2013 study by the ministry positively evaluated the normative regulation of the process and thought that the rules provide equal protection to both the police officer and the complainant (Borovec 2013).

The Police Law (2011) explicitly distinguishes between the criminal procedure and the internal disciplinary procedure (Article 93) and states that the acquittal in criminal court does not automatically imply that the police officer will be relinquished of any responsibility in the internal disciplinary process. The Law enumerates various violations of official rules , establishes the criteria to be used in the internal procedure, and delineates the disciplinary procedure and potential outcomes.

The Police Law (Article 93, 2011) stipulates that a police officer could be held accountable for the violations of official duties if the police officer does not perform them professionally and within the stipulated deadlines; if the police officer does not adhere to the Constitution , laws, and other rules and regulations; or if the police officer, otherwise, dishonors the police occupation . The Law enumerates some less serious violations, such as being rude toward citizens and fellow coworkers, being late for work, or maintaining unprofessional appearance (Article 95). The Law also contains descriptions of serious violations of official rules, such as failing to perform police work, engaging in illegal or criminal activities, revealing confidential information to unauthorized persons, or failing to take medical tests or attend required training (Article 96, Police Law 2011).

The disciplinary process can be initiated on the basis of the written report from the head of the administrative unit or immediate supervisor (Article 104, Police Law 2011). Before initiating the official procedure, the supervisor has to provide an opportunity for the police officer to respond to the initial charge. In case a police officer is accused of committing what seems to be a less serious violation, the minister of the interior or another supervisor is in charge of conducting the internal investigation and determining the outcome of the case (Article 97, Police Law 2011). However, in case a police officer is accused of committing what seems to be a more serious violation, then the disciplinary board is in charge of conducting the investigation and making the decision (Articles 97–98, Police Law 2011).

The disciplinary process should be conducted without delay and should be public (Article 101, Police Law 2011). The police officer has an active role in the process (e.g., respond to the charges, comment on the evidence, propose evidence to be examined) and may be represented by a lawyer and/or a union member (Articles 102–108, Police Law 2011). If the police officer was accused of committing a less serious violation, the supervisor decides the outcome, while the disciplinary board makes the decision in cases in which a police officer is accused of committing a more serious violation (Article 108, Police Law 2011). Disciplinary options for less serious violations are a written reprimand and up to 10 % salary cut. For more serious violations, police officers’ misbehavior can result in a pay cut of up to 20 % of the salary, prohibition of promotion for 2–4 years, reassignment to another job for 2–4 years, suspended termination for a period of 3 months to 1 year, and termination. Regardless of the severity of the charges, the police officer may appeal the decision (to the disciplinary board for the less serious violations and to the appellate disciplinary board for the more serious violations). Finally, in the cases of more serious violations, the officer could bring a lawsuit at the administrative court and appeal the appellate board decision (Article 108, Police Law 2011).

The disciplinary data are available for the period 1992–1999. In this period, administrative disciplinary procedure has been initiated against a small proportion of the police officers (between 2000 and 3500 annually; Kutnjak Ivković 2004a, p. 212). At the beginning of the period, the war was ravaging the country, and the bulk of the cases included charges of more serious disciplinary violations. As the “war-is-over attitude” started to dominate the scene in the last 1990s, the proportion of less serious charges started to increase and eventually started to constitute the majority of the disciplinary charges (Kutnjak Ivković 2004a). The most frequently applied discipline was fine .

The results of disciplinary data analyses are in complete agreement with the police officers’ perceptions. The 1995 police integrity survey suggested that police officers, if they expect any discipline to be meted out at all, expected less serious forms of discipline (e.g., fine, suspension), even for the most serious forms of corruption (Kutnjak Ivković 2004a). The respondents thought that only the theft from a crime scene would probably result in a dismissal (Kutnjak Ivković 2004a), while similarly serious violations of official rules, such as stealing money from a found wallet and accepting a bribe from a motorist caught speeding, would probably result in less serious discipline.

Curtailing the Code of Silence

The third dimension of the police integrity theory focuses on the code of silence, and the efforts that the police agency is making in curtailing it. Klockars and Kutnjak Ivković (2004) argue that curtailing the code of silence is critical for agencies of integrity. According to the theory, compared to the agencies of high integrity, agencies of low integrity would have strong codes of silence in which police officers would be more likely to tolerate police misconduct without reporting it and supervisors would be more reluctant to investigate police misconduct and discipline police officers who engaged in it .

The historic events in Croatian society affected the parameters of police culture, in general, and the code of silence, in particular. Before Croatia gained independence, many police officers were of Serb ethnicity (about 70 %; Kutnjak Ivković and Klockars 2004). Once the war broke out, a large proportion of them abruptly left the police and joined the paramilitary forces fighting against the newly established Croatian state. At the same time, the Croatian Government passed the law authorizing the Ministry of the Interior to establish the National Guard Corps (NGC) .

Literally overnight, the NGC started recruiting many individuals under the relaxed hiring criteria, while providing minimal police training, if any. Thus, in the early 1990s, there were very few experienced police officers who could socialize recruits into the existing police culture. Consequently, the newcomers are the ones who helped shape the police culture . The exodus of a large number of “grizzled veterans,” the hiring of many people with very limited police training, and the war with the exposure to violence, conversion of police and military roles, and the relaxed disciplinary standards all resulted in the creation of a very tight-knit police culture. The war camaraderie quite likely extended beyond the war itself; after fighting shoulder to shoulder in the war, it would be difficult for a person to report a fellow combatant for any misconduct, particularly a less serious one.

The police integrity survey conducted in 1995, between the two military operations “Flesh” and “Storm,” indicated the presence of the police code of silence (Kutnjak Ivković and Klockars 2004); at least 25 % of the respondents stated that they would not report for any of the behaviors described in the questionnaire, ranging from the acceptance of gratuities to the acceptance of a bribe from a speeding motorist and stealing from the crime scene. However, the code seemed to be even stronger for the forms of misconduct perceived to be the less serious; over 65 % of the respondents stated that they would not report a fellow police officer who accepted a free cup of coffee or any other gifts on his beat (Kutnjak Ivković and Klockars 2004).

Influence of Social and Political Environment

The fourth dimension of the police integrity theory (Klockars and Kutnjak Ivković 2004) focuses on the fact that police agencies are a part of the society at large and, as such, are directly influenced by the events and views held by that society at large (Klockars and Kutnjak Ivković 2004). Police agencies in societies highly tolerant of unethical behavior of public servants should have lower levels of integrity than police agencies operating in societies highly intolerant of unethical behavior of any kind. In a larger environment that supports integrity across the board, police agencies are more likely to set high expectations regarding police integrity as well.

Nepotism and corruption have been well intertwined in the core of the Croatian society. In a 2000 survey, more than 30 % of the respondents in the study reported paying a bribe to a police officer (Derenčinović 2000). A year later, in the 2001 International Crime Victimization Survey , about 15 % of the respondents reported that they were asked to pay a bribe to a governmental official last year (Kutnjak Ivković 2008). A more recent survey, the 2010 United Nations Office on Drugs and Crime (UNODC) survey, documents that 12 % of the Croatian respondents reported giving public officials money, gifts, or favors on at least one occasion during last year (UNODC 2011, p. 16).

Following the rather lukewarm efforts to deal with corruption in the early periods, the government’s efforts in dealing with corruption have received a major push as the country has faced serious obstacles on the road toward the European Union membership. In the fall of 2008, the prime minister of Croatia fired the minister of the interior and the police chief in a move widely believed to have been a direct consequence of their inability to provide high-quality policing. In 2010, Ivo Sanader, the same former prime minister, has been indicted and arrested for high-level corruption. In 2012, he was convicted of corruption and sentenced to serve a 10-year prison sentence. His second trial is still ongoing. In 2013, two scandals shook the police; the heads of the Vukovar-Srijemska Police Administration were arrested for corruption and the head of the Splitska Police Administration was removed from the post for abuse of his official position.

The general public has not been impressed by the governmental efforts in dealing with corruption. The majority of the respondents in the 2010 Transparency International survey (56 %, Transparency International 2010, p. 47) evaluated the government’s efforts to deal with corruption as ineffective and believed that the level of corruption in the country has increased in the past 3 years. Also, the UNODC study revealed that, while about one half of the respondents perceived that the level of corruption had remained the same in the past 2 years, about one third of the respondents thought that the level of corruption had increased (UNODC 2011, p. 44) .

The country’s ranking on the Transparency International’s Corruption Perception Index (CPI) reflects the changes taking place in the country. Croatia’s CPI score improved from 2.7 (out of 10) in 1999 to 4.4 in 2008 (Transparency International 2008), suggesting a gradual improvement in the government’s dealing with corruption. However, the scores did not continue to improve; they have decreased to 4.1 in 2010 (Transparency International 2010) and 4.0 in 2011 (Transparency International 2011). The picture of corruption portrayed by the Transparency International fits well with the public perceptions hinting at the increase in corruption in the past 3 years.

As the war and war-related violence ravaged the country in the early 1990s, rules protecting human rights were not strictly enforced. The early Human Rights Watch reports (e.g., 1993, 1995, p. 3) document that the Croatian police have engaged in human rights violations , although the prevalence of such behavior seems to have been less extensive among the police officers than among military officers. The nature of the abuse—violent attacks by the police, beatings of the suspects, and failure to provide protection from either the military or the civilian violence (almost exclusively ethnically motivated)—suggests a widespread and systematic nature of the violations at the time. Although police conduct seems to have improved to some extent between 1992 and 1994 (Human Rights Watch 1995, p. 3), the two military operations in 1995 appear to have expanded the extent of the police ethnically motivated mistreatment of citizens.

Although the dominant perception during the second period in the history of Croatian police was that the human rights violations in Croatia stem from the war (Amnesty International 1998, p. 2), the reports began to include descriptions of more “traditional” human rights violations (e.g., use of force against suspects, violations of other citizens’ procedural rights). Toward the end of that period, the European Commission against Racism and Intolerance (2001, Appendix 1) pointed out the positive trend, “the significant democratic changes, which have occurred after the parliamentary and presidential elections at the beginning of 2000, have contributed to the radical changes in a number of policy areas that were previously subject to criticism by various international monitoring instruments.”

In the most recent time period of the Croatian police history, democratization efforts, propelled by the prospect of European Union membership, were in full swing. The nongovernmental organization (NGO) reports no longer described direct police involvement in violence and failure to protect the victims from attacks (Commission of the European Communities 2006, 2007, 2010, 2011). Although the police are still struggling with some of the issues, such as the protection of the lesbian, gay, bisexual, and transgender (LGBT) community when they exercise their right to a free assembly, over time, the police treatment of suspects improved as well. In 2003, the European Committee for the Prevention of Torture (CPT) and Inhumane or Degrading Treatment or Punishment (2007, p. 59) reported that, “the majority of persons interviewed by the CPT’s delegation during the 2003 visit indicated that they had not been ill-treated by the police.”

Measuring Police Integrity

Questionnaire

Our 1995 study of police integrity in Croatia, performed between two major military operations at the end of the war in Croatia , indicated that police officers seemed to be tolerant of corruption. The second version of the questionnaire, fielded in 2008/2009, constituted the empirical foundation of this chapter. It contains descriptions of 11 scenarios that cover a variety of forms of police misconduct—from police corruption and use of excessive force to planting of evidence and verbal abuse.

Each scenario is followed by the same set of questions, used in both the first and the second version of the questionnaire. The follow-up questions ask about police officers’ knowledge of official rules and their opinions about the seriousness of particular rule-violating behaviors, the discipline these behaviors would deserve and would actually receive, and their estimates of how willing they would be to report such behavior. The disciplinary questions have been adjusted to fit the Croatian legal environment.Footnote 1

Finally, the respondents were asked a few demographic questions. To increase the respondents’ willingness to participate in the study, and to exclude the possibility that respondents could be identified, demographic questions have been kept to a bare minimum, inquiring about the length of the respondents’ police experience, rank, assignment, and whether they were employed in a supervisory position.

The Sample

The Croatian police are a centralized police agency within the Ministry of the Interior . On top of the hierarchical structure are the minister of the interior and the chief of police (Police Law 2000), followed by 20 police administrations (which correspond to the 20 counties) and about 200 police stations at the bottom of the structure. The Police Law (2000, Article 13) prescribes the classification of police administrations into categories. Using several criteria (geographic size, population size, number of crimes known to the police, traffic patterns, and geographic location; see Ministry of the Interior 2009), the ministry classifies each police administration into one of the four categories. One of the key roles of police administrations is to oversee the operation of police stations (Police Law 2000, Article 15) in which the majority of regular police work is performed. Based on the nature of the work they perform, police stations can be divided into basic police stations and specialized police stations (e.g., airport, traffic, border). Based on the same criteria used to classify police administrations (Article 15, Police Law 2000), police stations are also classified into three categories.

In 2008/2009, the survey was administered to police officers employed in a stratified representative sample of police stations. The sample of police stations was drawn to represent both police administrations (four categories) and police stations (three categories). Because our interest is focused on regular patrol work, specialized police stations (e.g., border police, maritime police) in which police officers do not regularly patrol the beat were excluded from the sample. We created a 4 × 3 table by cross-tabulating four categories of police administrations with three categories of police stations. East police station was classified into one of the 12 cells (there are no class I police stations which are supervised by class IV police administrations). For the 11 categories populated by police stations, two stations were picked at random from all stations belonging to a specific category, thus yielding 22 stations in our sample. Each of the 22 stations we selected for the study participated in the study and returned questionnaires to us, resulting in a station response rate of 100 %. Most, but not all police officers working in these stations, returned completed questionnaires, yielding the officer response rate of 88 %. Table 4.1 shows the overall number of officers and the distribution of officers by station and administration.

Table 4.1 Sample distribution
Full size table

The overwhelming majority of our respondents are line officers (81 %; Table 4.2) and the rest are supervisors (19 %). Most officers are also very experienced; only one eighth of them have less than 5 years of experience, three quarters have more than 10 years of experience, and almost one half of the respondents had been police officers for more than 15 years (Table 4.2). The respondents performed a range of assignments, with patrol and detective units being the most frequent ones (Table 4.2). About 10 % are women (Table 4.2).

Table 4.2 Respondents’ demographic characteristics
Full size table

The last questions in the questionnaire asked the respondents whether they have told the truth while filling out the questionnaire. About 5 % did not provide any answer to this question. Out of the officers who responded, the overwhelming majority (92 %) said that they did not lie. The analyses that follow include only the answers provided by those respondents who stated explicitly that they did not lie.

The Results

Seriousness

The respondents were asked to evaluate how serious they perceive the behaviors described in the scenarios, as well as to estimate how serious most police officers in their agencies would evaluate them. They were offered answers on a five-point Likert scale, ranging from “not at all serious” to “very serious.”

The respondents’ evaluations of scenario seriousness suggest that the scenarios were generally evaluated to be on the serious side (means are clustered between the midpoint of the scale and the serious end of the scale). However, these evaluations ranged in terms of seriousness, from the scenario evaluated to be the least serious, scenario 8 (describing the cover-up of police driving under the influence (DUI) with the mean at the midpoint), to the scenario evaluated as the most serious, scenario 3 (describing the theft of a knife from the crime scene with the mean very close to the serious end of the scale; Table 4.3). The average evaluations of seriousness enable the classification of these scenarios into three groups.

Table 4.3 Police officers’ perceptions of seriousness and violation of rules
Full size table

The four least serious scenarios (scenario 8: cover-up of police DUI; scenario 1: free meals, gifts from merchants; scenario 6: officer strikes prisoner who hurts partner; and scenario 7: verbal abuse – “Arrest an Asshole Day”) all have means just above the midpoint of our scale (3.00). These scenarios describe the most benign versions of several forms of police misconduct. Scenario 1 (free meals, gifts from merchants) focuses on the acceptance of gratuities, traditionally viewed as the stepping stone toward more serious forms of police corruption. Scenario 8 (cover-up of police DUI) is also a form of police corruption, one involving internal corruption from Barker and Roebuck’s classification (1973). Prior research on police integrity (Klockars et al. 2004a) shows that internal corruption has been classified as one of the least serious forms of police corruption. In this group, of least serious scenarios, there are also two scenarios involving the use of excessive force. Scenario 7 (verbal abuse – “Arrest an Asshole Day”) describes the (ab)use of verbal commands, traditionally viewed as the lowest point on the use of force continuum (National Institute of Justice 2009). Although striking an arrested and handcuffed person should not be acceptable and should be evaluated, as serious (scenario 6: officer strikes prisoner who hurts partner) justification given for the activity—a dose of “street justice”—probably contributing toward making it less serious in the eyes of police officers than it should be.

Out of the four scenarios that can be classified as scenarios of intermediate seriousness (the values of means are mostly between 4.0 and 4.5), two describe misconduct by a supervisor (scenario 5: supervisor offers holiday for errands; scenario 11: Sgt. fails to halt beating). In one scenario, scenario 5 (supervisor offers holiday for errands), the supervisor is proposing a deal that could be classified as internal corruption. In the other scenario, scenario 11 (Sgt. fails to halt beating), a supervisor is not actively using excessive force himself, but is passively watching as his subordinates repeatedly strike and kick a man arrested for child abuse. In other words, his command responsibility, rather than his direct conduct, is at stake. The means for the remaining two scenarios (scenario 9: auto body shop 5 % kickback; scenario 4: unjustifiable use of deadly force) are very close to 4.5 on a five-point scale, suggesting that these scenarios have been perceived as serious. One of the scenarios, scenario 9 (auto body shop 5 % kickback) describes a traditional form of corruption in which a police officer abuses his official position to obtain an illegal kickback (Roebuck and Barker 1973). The other scenario, scenario 4 (unjustifiable use of deadly force) should be viewed as one of the most serious scenarios in the whole questionnaire, if not the most serious. What potentially mitigates the offender’s responsibility in the eyes of the respondents might be the police officer’s prior experience—included in the description of the scenario—in which he was seriously hurt because he did not react on time. Police officers could potentially identify with this situation.

The three most serious scenarios all have means substantially closer to 5 (the “very serious” end of the scale). These three scenarios, plain violations of criminal law, include one scenario with opportunistic theft (scenario 3: theft of knife from crime scene), a scenario illustrating a failure to execute an arrest warrant on a friend (scenario 2: failure to arrest friend with warrant), and a scenario describing falsification of the official report (scenario 10: false report of drugs on drug dealer). All three scenarios are plain violations of official rules in which a police officer does something that he was not supposed to do (e.g., steal the knife from a crime scene) or does not do something that he was supposed to do (e.g., did not arrest a friend, falsified the official report).

A comparison of the respondents’ own estimates of seriousness and how serious they estimated that other police officers in the agency would evaluate the same scenarios revealed several findings. First, for each and every scenario, the respondents evaluated these scenarios as more serious than they thought the other police officers would; the means for evaluations of own seriousness were always higher than the means for others’ estimates of seriousness (Table 4.3). Second, the differences between the means measuring own estimates of seriousness and the means measuring others’ estimates of seriousness were all statistically significant, but were large and meaningful (above the 0.50)Footnote 2 only in five scenarios (scenario 1: free meals, gifts from merchants; scenario 2: failure to arrest friend with warrant; scenario 3: theft of knife from crime scene; scenario 5: supervisor offers holiday for errands; and scenario 9: auto body shop 5 % kickback) the differences were large and meaningful. Third, the relative order of the scenarios, measured through the ranking of the scenarios, suggests that police officers followed the same internal order of seriousness, regardless of whether they expressed their own evaluations or whether they estimated how other police officers would evaluate the scenarios. In fact, the Spearman’s correlation between the rankings of own evaluations of seriousness and others’ estimates of seriousness is very high (Spearman’s correlation coefficient = 0.936, p < 0.001).

Violation of Official Rules

All of the examples of misconduct described in the questionnaire incorporate rule-violating behavior. The respondents were asked whether the behaviors described in the questionnaire could be classified as examples of rule-violating behavior. They could select an answer from three possible choices: “yes,” “no,” and “not sure.”

Indeed, the majority of police officers recognized and labeled the (mis)behavior in each and every scenario as rule violating (Table 4.3). However, the percentages of police officers who did so varied across scenarios, from the scenarios in which a slim majority evaluated the behavior as rule violating (55 % in scenario 1: free meals, gifts from merchants) to the scenarios in which the overwhelming majority did the same (93 % in scenario 3: theft of knife from crime scene; Table 4.3). Scenarios in which two thirds or fewer officers recognized the behavior as rule violating include scenarios with minor forms of police corruption, such as the acceptance of gratuities (scenario 1: free meals, gifts from merchants), minor forms of use of excessive force (scenario 7: verbal abuse – “Arrest an Asshole Day”), scenarios involving “street justice” (scenario 6: officer strikes prisoner who hurts partner), and scenarios involving internal corruption (scenario 5: supervisor offers holiday for errands; scenario 8: cover-up of police DUI). On the other hand, scenarios in which four-fifths or more labeled the behavior as rule violating were scenarios in which the police officer in the scenario engages in severe violations of official rules and criminal laws (scenario 2: failure to arrest friend with warrant; scenario 3: theft of knife from crime scene; scenario 10: false report on drug dealer).

Why would some of the respondents have problems recognizing some of these scenarios as violations of official rules ? It is possible that a substantial minority of the police officers do not know the official rules. Although extensive training was not provided in the war years to each newly hired police officer, in-service training and on-the-job experience, since those early 1990s, should have provided police officers with sufficient opportunities to learn “the basic rules.” The results of Veić’s study conducted in 2001 indicate that a substantial portion of the officers still may have problems recognizing at least some of the rules (Ivanović 2001, p. 12). Indeed, there are at least 10 % of police officers in this study, who seem to be unsure of whether the behaviors constitute violations of official rules (Table 4.3). In four scenarios (scenario 1: free meals, gifts from merchants; scenario 5: supervisor offers holiday for errands; scenario 6: officer strikes prisoner who hurts partner; scenario 8: cover-up of police DUI) the percentage even reaches 20 %. In addition, in three out of these four scenarios with higher percentages of police officers not being sure whether the behavior violates official rules, about 20 % of police officers flatly denied that such behavior indeed violates official rules.

Finally, the respondents’ evaluations of whether the behavior constitutes a violation of official rules is strongly related to how serious they perceive the behavior; the more serious they evaluate the behavior, the more likely they are to evaluate it as rule violating. The ranking of scenarios based on their evaluations of seriousness and the ranking of scenarios based on their evaluations of rule-violating nature of the behavior are very similar (Spearman’s correlation coefficient = 0.882; p < 0.001).

Appropriate and Expected Discipline

The respondents were also asked to express their views about the appropriate discipline for the behaviors described in the questionnaire, as well as estimate what discipline would be meted out in their agencies for such behavior. Based on the norms of the Law on Civil Servants and Public Employees (2001), the answers offered in the questionnaire included: “no discipline,” “public warning,” “fine in the amount of 10 % of the employee’s salary,” “fine in the amount of 20 % of the employee’s salary,” “reassignment,” and “dismissal.”

We use three ways of assessing the respondents’ views. First, we rely on the modal values (Table 4.4). The respondents did not think that the same discipline would be appropriate for all scenarios. Rather, for four scenarios (scenario 1: free meals, gifts from merchants; scenario 5: supervisor offers holiday for errands; scenario 7: verbal abuse – “Arrest an Asshole Day,” scenario 8: cover-up of police DUI), the modal appropriate discipline is (only) “public reprimand,” the most lenient of all forms of discipline. These are the scenarios involving the least serious forms of corruption (scenario 1: free meals, gifts from merchants), the use of the excessive force on the lower end of the use of force continuum (scenario 7: verbal abuse – “Arrest an Asshole Day”), as well as internal corruption (scenario 5: supervisor offers holiday for errands; scenario 7: verbal abuse – “Arrest an Asshole Day,” scenario 8: cover-up of police DUI). With the exception of scenario 5, these are the scenarios perceived to be the least serious and having the smallest number of police officers recognizing them as rule violating as well (Table 4.3). According to our respondents, two scenarios (scenario 6: officer strikes prisoner who hurts partner; scenario 11: Sgt. fails to halt beating) deserve a more severe discipline, a fine in the amount of up to 20 % of the salary. Finally, in five scenarios (scenario 2: failure to arrest friend with warrant; scenario 3: theft of knife from crime scene; scenario 4: unjustifiable use of deadly force; scenario 9: auto body shop 5 % kickback; scenario 10: false report on drug on dealer), the modal response discipline was “dismissal.” Most of these scenarios were evaluated as the most serious and were most likely to have the highest percentage of police officers recognizing them as rule violating (Table 4.3).

Table 4.4 Views about appropriate discipline (should) and expected discipline (would)
Full size table

Second, we relied not only on the modes but also on the percentages of police officers who selected no discipline, some discipline other than dismissal, and dismissal (Table 4.4). The findings of this analysis are similar to the analysis of modal responses. In addition, this analysis reveals that, out of the five scenarios classified as requiring the most severe discipline based on the modal responses, the majority of police officers thought that dismissal is the appropriate discipline in only two scenarios (scenario 3: theft of knife from crime scene; scenario 4: unjustifiable use of deadly force). This finding suggests that dismissal is reserved for rare situations and is only appropriate for the most serious form of corruption and the use of deadly force. For everything else, some discipline is appropriate, but it should not be dismissal.

Third, we also ranked scenarios based on the modal appropriate discipline for each scenario (Table 4.4). Because the answers are recorded on the ordinal scale with six possible categories, several scenarios share the same ranking. Nevertheless, a comparison of the ranking of appropriate discipline with the ranking for seriousness shows that they are strongly correlated (Spearman’s correlation coefficient = 0.839, p < 0.01); the more serious the respondents evaluated the scenarios to be, the more likely they were to think that the appropriate discipline should be harsher. Similarly, a comparison of the ranking of appropriate discipline with the ranking for rule-violating behavior shows that the two are strongly related as well (Spearman’s correlation coefficient = 0.897, p < 0.001). The more likely the respondents were to evaluate the behavior as rule violating, the more likely they were to think that the appropriate discipline should be harsher.

The second question about discipline asked the respondents to estimate what discipline their agencies would mete out. One of the options was no discipline, that is, that their agency would not discipline a police officer at all. For only one case, scenario 5 (supervisor offers holiday for errands), the modal discipline was “no discipline” at all, but a more detailed analysis across the three categories of discipline (Table 4.4) clearly suggests that, although about one third of the respondents indeed expected no discipline, about two thirds expected some discipline, but mostly minor in terms of its severity. Similarly, there are two scenarios (scenario 1: free meals, gifts from merchants; scenario 7: verbal abuse – “Arrest an Asshole Day”) describing minor forms of police corruption and excessive force in which the mode was “public reprimand,” and the analysis of percentages across the three discipline categories shows that the majority of the respondents expected some discipline. On the other hand, for five scenarios (scenario 2: failure to arrest friend with warrant; scenario 3: theft of knife from crime scene; scenario 4: unjustifiable use of deadly force; scenario 9: auto body shop 5 % kickback; scenario 10: false report on drug on dealer) the modal discipline was dismissal. However, the analysis across the three categories of discipline reveals that in only two of these scenarios (scenario 3: theft of knife from crime scene; scenario 4: unjustifiable use of deadly force), the scenarios perceived to be most serious in the questionnaire, which over 90 % of the respondents evaluated as rule violating, the majority of the respondents expected discipline to be meted out, while in the other three scenarios (scenario 2: failure to arrest friend with warrant; scenario 9: auto body shop 5 % kickback; scenario 10: false report on drug on dealer), despite the fact that the modal discipline was dismissal , the majority of the respondents expected some discipline, but not dismissal.

We also compared the respondents’ estimates of appropriate and expected discipline. First, both the analysis of modes and the analysis of percentages across three categories of discipline suggest that, in general, the views about appropriate and expected discipline are similar. The largest differences in both modal choices and differences in percentages of at least 10 % appear in two scenarios (scenario 4: unjustifiable use of deadly force; scenario 5: supervisor offers holiday for errands). In both of these scenarios, the respondents approved of somewhat less serious discipline than they expect their agencies to mete out. Second, the analysis of ranks confirms the overall findings of close relation (Spearman’s correlation coefficient = 0.930, p < 0.001). Third, the modal discipline was the same in 8 out of 11 cases. In two of the scenarios with differences (scenario 5: supervisor offers holiday for errands; scenario 6: officer strikes prisoner who hurts partner), the appropriate modal discipline was harsher than what they expected that the agencies would mete out, thus suggesting that there is a substantial group of police officers who evaluated the actual discipline to be too lenient. On the other hand, in one of the scenarios involving internal corruption (scenario 8: cover-up of police DUI), the expected modal discipline was more severe than the modal appropriate discipline , indicating that there is a substantial group of police officers who evaluated the expected discipline as too harsh.

Willingness to Report Misconduct

The last two questions after each scenario asked the respondents to express how willing they would be to report misconduct and to estimate how willing other officers in their agencies would be to do so. The answers ranged on a five-point Likert scale from “definitely not” to “definitely yes.” The answers conveying their own (un)willingness to report misconduct, help us assess the extent and nature of the code of silence.

The analysis of the results (e.g., mean values, percentage of the officers not reporting) shows that the code of silence varies across scenarios, from 51 % of the respondents saying that they would not report a cover-up of police DUI (scenario 8) to 12 % who said that they would not report a failure to arrest a friend with an outstanding warrant (scenario 2) or an unjustifiable use of deadly force (scenario 4; Fig. 4.1). The respondents’ adherence to the code of silence is directly related to their perceptions of scenario seriousness (Spearman’s correlation coefficient = 0.936, p < 0.001), likelihood of recognizing it as rule-violating behavior (Spearman’s correlation coefficient = 0.809, p < 0.01), and the severity of appropriate discipline (Spearman’s correlation coefficient = 0.839, p < 0.01). The analysis of the mean values suggests that scenarios could be divided into three categories (Table 4.5).

Fig. 4.1
figure 1

Percentage not reporting by scenario number

Full size image
Table 4.5 Police officers’ perceptions of willingness to report
Full size table

First, there are four scenarios (scenario 1: free meals, gifts from merchants; scenario 6: officer strikes prisoner who hurts partner; scenario 7: verbal abuse – “Arrest an Asshole Day”; scenario 8: cover-up of police DUI) in which the mean values are below the midpoint of the scale, suggesting that the behaviors described in these scenarios would be well protected by the code of silence. In addition, a separate analysis of the code of silence (measured as the percentage of the respondents who said that they would not report) shows that about one half of the respondents for each of these scenarios said that they would not report a police officer who engaged in misconduct described in the scenario. These four scenarios include the least serious forms of police corruption (i.e., the acceptance of gratuities) and the use of excessive force, as well as the two scenarios with internal corruption . These four scenarios have been perceived as the least serious, least likely to be recognized as rule-violating behavior, and deserving the least serious forms of discipline (Tables 4.3 and 4.4).

Second, there are four scenarios (scenario 5: supervisor offers holiday for ­errands; scenario 9: auto body shop 5 % kickback; scenario 10: false report on drug on dealer; scenario 11: Sgt. fails to halt beating) in the middle group with the means between 3 and 4 and with about one quarter of the respondents saying that they would not report (Table 4.5). These scenarios include a false report, serious forms of corruption (kickback), hitting an arrested suspect as a measure of “street justice,” and not reacting to the beating. They have been classified in the middle of the scale in terms of their seriousness, likelihood that the respondents will label them as rule violating, and severity of the discipline (some discipline, but not dismissal).

Third, there are three scenarios (scenario 2: failure to arrest friend with warrant; scenario 3: theft of knife from crime scene; scenario 4: unjustifiable use of deadly force) for which the means are above 4 and are on the reporting side of the scale. At the same time, the percentage of the police officers who stated that they would not report is the smallest for these scenarios (between 11 and 16 %). These scenarios describe the most serious forms of corruption and use of excessive force, which are not only violations of agency rules but also violations of criminal law as well. These scenarios were evaluated as the most serious by the respondents, were most likely to be recognized as violations of the official rules, and merited dismissal from the agency.

We also compared the respondents’ own willingness to report with their estimates of others’ willingness to report (Table 4.5). The mean values suggest that, in all 11 scenarios, the respondents seem to be somewhat more willing to say that they would report than they estimated that others would. Although the differences between the means for their own willingness to report and the means for others’ willingness to report are statistically significant in 10 out of 11 scenarios, these differences are meaningful in only two scenarios (scenario 2: failure to arrest friend with warrant; scenario 3: theft of knife from crime scene).Footnote 3 Finally, a comparison of the rankings shows that their own willingness to report and their estimates of others’ willingness to report are very strongly related (Spearman’s correlation coefficient = 0.980, p < 0.001).

Conclusion

Although Croatian police are one of the youngest police agencies in Europe , their history has been burdened with serious challenges. Not only has the transition toward a democratic agency been hampered with the war but the war experience—be it from the perspective of a victim, soldier, or both—affected the police officers as well. Hiring standards and police training have been relaxed while the enactment of the relevant legal norms has been put on hold. Eventually, relevant legal statutes (e.g., Criminal Law, Criminal Procedure Law, Police Law) have been enacted and the legal environment has been created for the successful transition . As time passed, the society at large has become more engaged and successful in dealing with corruption and nepotism , as well as in addressing ethnic-based violence and hostility.

The Ministry’s recent survey revealed that the respondents’ familiarity with the use of force rules was “excellent” or “good” (Borovec 2013, p. 6). Consistently, the results of our study indicate that most, but not all police officers, know the official rules. Over two thirds of police officers recognized most of the examples of misconduct as rule-violating behavior, but there was still a substantial minority of police officers who did not. Police officers were less likely to recognize as rule-violation behavior which they evaluated as less serious (e.g., acceptance of gratuities, dispensing a dose of “street justice” on a handcuffed arrestee, covering up a DUI accident, getting favors from a supervisor in exchange for running errands). In such cases, a larger percentage of our respondents tended to say that they were not sure whether the described behavior violated the official rules. Such cases indicate a potential conflict between official agency rules and the actual practice in the agency. Accordingly, an increased proportion of our respondents expected, or approved of, no discipline at all or very mild discipline for such behaviors.

The respondents’ evaluations of misconduct seriousness vary substantially across the scenarios. However, the second version of the questionnaire that we used in this study allowed us, for the first time, to look at potential patterns of whether police officers differentiate across different types within the same form of misconduct. The results clearly show that the respondents had no problems distinguishing among different forms of corruption; they evaluated the acceptance of gratuities and internal corruption as the least serious forms of corruption and a kickback and theft from a crime scene as the most serious forms of corruption. These findings fit well within the parameters of corruption seriousness established in our earlier work. In particular, our first questionnaire, focusing exclusively on police corruption, has been used in more than a dozen countries and overwhelmingly shows the same pattern (Alain 2004; Chatta and Kutnjak Ivković 2004; Edelbacher and Kutnjak Ivković 2004; Johnson 2004; Klockars et al. 2004a, 2004b; Kremer 2004; Kutnjak Ivković and Khechumyan 2014, 2013; Kutnjak Ivković 2004b; Kutnjak Ivković and Sauerman 2012, 2013a, 2013b; Kutnjak Ivković and Shelley 2005, 2007, 2008, 2010; Newham 2004; Pagon and Lobnikar 2004; Pounti et al. 2004; Punch 2004; Torstensson Levander and Ekenvall 2004).

The use of excessive force, on the other hand, has only been addressed in this second questionnaire. The police officers’ perceptions of seriousness of different forms of (mis)use of force seem to be strongly related to the use of force continuum. The forms that violate the rules, at the lower end of the use of force continuum (e.g., abuse of a verbal command), at the same time, are evaluated to be less serious than forms of misconduct at the higher end of the use of force continuum (e.g., use of deadly force).

The behaviors described in the questionnaire are violations not only of the administrative rules but also of criminal law. Based on the results of our study, a potential outcome for a police officer who engaged in the described behaviors, and is eventually officially processed, would not necessarily involve a severe discipline. Only for the two most serious cases (i.e., the use of deadly force, theft from a crime scene) would the majority of the respondents support and expect dismissal. For all of the other forms of misconduct, the majority favored and expected some discipline, but definitely less serious than dismissal. The more lenient discipline that police officers expect the police agency to mete out—if they are reading the disciplinary threat correctly—signals that the agency is not as serious about curtailing police misconduct. On the other hand, if the police agency is actually firing police officers for a range of other forms of misconduct, the police officers are neither getting the message loudly nor clearly.

Lastly, our results constitute clear evidence that the code of silence exists among the Croatian police officers. However, the code of silence is far from a flat prohibition of reporting because it varies dramatically. On one hand, about one half of the respondents would protect in silence a fellow officer’s DUI accident, administering a dose of “street justice,” or verbally harassing citizens. On the other hand, about one tenth of the respondents would protect a police officer who unjustifiably used deadly force or who failed to execute an arrest warrant. Compared to the findings of the 1995 survey (Kutnjak Ivković and Klockars 2004), in which at least one quarter of the respondents said that they would not report a fellow police officer for any of the behaviors described in the questionnaire, the results of our 2008/2009 survey suggest that the extent of the code overall has decreased because only about one tenth of the respondents said the same. The strong code of silence that shielded the police officers who fought together during the war seems to have weakened. This is probably a combination of two factors: Improving attitudes of seasoned police officers and the arrival of the more recently trained, post-war police officers. As time passes, the war-related influences weaken, institutions of control strengthen, and the country moves along the path toward democracy.

However, our results show that, even in our more recent survey, about 10 % of the police officers would not report the most serious examples of police misconduct. The good news is that the majority of the officers would not protect such behavior in silence and that the code has weakened overall. The bad news is that a nontrivial proportion of the police officers still would tolerate even the most outrageous forms of misconduct in silence. While it is not reasonable to expect that the code of silence will disappear completely as the Croatian police become a truly democratic police agency, the expectation placed before any police agency of integrity is that police culture and individual police officers will not be tolerant of serious forms of misconduct. Police officers who abuse their right to use deadly force should be certain that they would be reported by their peers and that the official disciplinary and criminal procedure would be initiated.