ASSESSING ATTITUDES TOWARD MUNICIPAL POLICE IN MEXICO DURING DEMOCRATIC TIMES: A CASE STUDYING

ABSTRACT

One of the key roles of States is to act as primary authorities over their territories and societies to establish order and provide security. Nevertheless, the processes of establishing and maintaining order and social control are complex. As such, different groups and institutions are required to interact effectively to encourage collective behavior. Among the different State agents, the police are directly in charge of fighting crime as well as maintaining social control. Their role in the development of democratic political systems is indeed significant. Police officers’ varying degrees of respect for civil, political, and human rights affect the tone of the relationship between the public and the government. Furthermore, the performance of the police in combating crime and violence can serve as an indicator of government competence. Today, the acceptance of democratic governance around the world has achieved an appeal seldom seen in modern history, yet the spread of democracy has not necessarily allowed all States to establish or maintain order and provide social control. This situation has been especially true for emerging and transitional democracies, as manifested through increasing problems of insecurity. Mexico is one such case of a nascent democracy facing significant insecurity challenges. In an attempt to learn more about the phenomenon of insecurity, this study examines citizens’ perceptions of the Mexican police at a municipal level. By using different approaches of criminal justice as well as a cross-sectional survey design, the study finds that police integrity, political legitimacy, and the perception of escalating neighborhood crime significantly affect varying attitudes toward the police. According to the results, Mexican police administrators and policymakers in general ought to place special attention on issues of integrity and political legitimacy to improve security and other areas of public policy in a country that is currently entering the democratic stage.

RESUMEN

Una de las funciones clave de los estados es actuar como autoridades principales sobre sus sociedades y territorios para establecer el orden y proveer seguridad. Sin embargo, los procesos para el establecimiento y mantenimiento del orden y control social son complejos. Así, diferentes grupos e instituciones necesitan interactuar de forma efectiva para promover el comportamiento colectivo. Dentro de las diferentes instituciones del estado, las policías se encargan directamente de combatir el crimen y también de mantener el control social. El papel de la policía en el accionar de los sistemas políticos democráticos es vital. Los diferentes grados de respeto que los policías ejercen sobre los derechos civiles, políticos, y humanos afectan la relación entre la ciudadanía y el gobierno. Además, el desempeño de la policía en el combate al crimen y la violencia puede servir como indicador de la eficiencia del estado. En la actualidad, la aceptación de los sistemas de gobierno democráticos ha generado un interés raramente visto en la historia reciente, pero la propagación de la democracia no ha necesariamente permitido a todos los estados establecer o mantener el orden y proveer el control social. Esta situación se ha manifestado primordialmente en las democracias emergentes o en transición a través de problemas crecientes de inseguridad ciudadana. México es uno de esos casos de una democracia naciente que enfrenta retos importantes de inseguridad. En un intento para conocer más sobre el fenómeno de la inseguridad, este estudio examina percepciones ciudadanas sobre la policía a nivel municipal en México. A través del uso de diversos enfoques de justicia criminal y encuestas, el presente trabajo reporta que la integridad policial, la legitimidad política, y la percepción colectiva de que el crimen en la colonia está en aumento afectan de manera importante las percepciones del público con relación al desempeño de la policía. De acuerdo con estos resultados, los mandos policiales así como los servidores públicos en general deberían poner especial atención en los asuntos de integridad y legitimidad política para mejorar la seguridad y otras áreas del quehacer gubernamental en un país actualmente adentrado en su etapa democrática.

I

INTRODUCTION

Among the different State agents, the police are directly in charge of combating crime and maintaining social control.¹ Police actions “arguably are the most visible embodiment of the government's exercise of control and authority.”² Police officers exercise wide-ranging discretionary powers in their operations and become gatekeepers in the sense of deciding who will undergo further legal intervention by other social control agencies.³

The role of the police in the development of democratic political systems is indeed significant.⁴ Varying degrees of police officers’ respect for civil, political, and human rights affect the mood of the relationship between the public and the government. Furthermore, the performance of the police in combating crime and violence can serve as an indicator of government competence. Limited police achievements in these public policy areas lead citizens to withdraw their support from the current democratic regime, thus affecting its legitimacy and viability.⁵ Moreover, the effective police agency performance in combating crime and maintaining social control hinges on citizen cooperation.⁶ Citizens can help the police fight crime and violence by voluntarily reporting crimes or identifying criminals.⁷ Most importantly, citizens can significantly contribute to maintaining public order by complying with the current societal laws and rules since “the police cannot be everywhere at every moment to prevent crime or intervene.”⁸

Studies also show that the way in which citizens evaluate the police may impact police enforcement efforts to fight crime and maintain order.⁹ Public dissatisfaction with the police tends to dissuade citizens from cooperating with the legal authority on crime-related activities. In more general terms, perceptions about the police may determine whether the public supports or opposes the current political and legal order.¹⁰ Those who view the police in positive terms tend to be more engaged in the political system or to manifest overall approval of the government. In contrast, negative views about the police may lead citizens to express dwindling confidence in the State's legitimacy.¹¹

Politically speaking, the current acceptance of democratic governance around the world has acquired an appeal seldom seen in modern history.¹² For some authors, “democratization has become one of the dominant international issues of this new century.”¹³ Yet, the spread and intensification of democracy have not necessarily allowed all States to establish or maintain order and provide social control. This situation has been especially true for emerging and transitional democracies. For instance, Latin American democracies generally lack effective social control systems as reflected in the growing problems of common crimes and organized crime.¹⁴ Indeed, the fact that citizen insecurity is on the rise makes Latin America the “most violent region” in the world.¹⁵

Among all Latin American countries, Mexico has been one of the most adversely affected by crime and violence in recent years.¹⁶ Indeed, some authors state that Mexico “finds itself in the midst of the most troubling period of criminality in its modern history.”¹⁷ The Mexican criminal justice system not only faces rising problems of violent common crimes —mainly robbery, property crime, and assault— but also high-profile violence linked to organized crime, such as drug trafficking, arms smuggling, money laundering, and kidnapping.¹⁸

These increasing problems of insecurity have generated tremendous challenges for Mexico's criminal justice system.¹⁹ Most importantly, though, they have unveiled the system's significant weaknesses and the troubling dysfunctions of some law enforcement agencies like the police.²⁰ Generally speaking, Mexican police agencies suffer from funding limitations and problems of coordination that hinder effective efforts to combat crime and violence.²¹ Furthermore, the majority of citizens display limited trust in the police.²² To illustrate this, surveys in Mexico City indicate that 70 percent of the respondents do not trust the police.²³ They also show that approximately 88 percent of all of the crimes committed in Mexico City are not reported to legal authorities. In addition, 57 percent of those citizens who decided to report crimes were afraid to do so.²⁴ This disconnection between the public and the police has evidently hampered police actions to adequately face the increasing insecurity.²⁵ It has also undermined citizens’ confidence in the country's justice sector.²⁶

Why does this gap exist between the public and the police in addressing problems of insecurity in Mexico? This article addresses the question by focusing on how individuals form opinions about police in Mexico, and specifically about the performance of municipal police agencies. A study of public views of local or municipal law enforcement agencies in Mexico is warranted for a number of reasons. Municipal institutions play a significant role in fighting crime and violence.²⁷ Although serious crimes like drug trafficking fall under federal jurisdiction, local authorities can effectively contribute to federal efforts to maintain social control by focusing on other types of crimes.²⁸ Scholars note that approximately 95 percent of all the crimes committed in Mexico fall under the realm of state criminal law.²⁹ Furthermore, local or municipal police departments tend to rely on better information about a neighborhood's crime level and have the potential to improve their relation with the community through their performance.³⁰ Thus, it appears that most of the efforts to reduce Mexico's problems of crime and violence should concentrate on the subnational level instead of the national one.³¹

Using quantitative approaches of criminal justice, a cross-sectional survey design, and multivariate analyses, this article finds that variables like police integrity, political legitimacy, and the perception that neighborhood crime is escalating significantly affect the various attitudes toward the police. According to the results, Mexican police administrators and policymakers in general ought to place special attention on issues of integrity and political legitimacy to improve security and other public policy areas in a country currently transitioning to a democratic stage.

This article seeks to positively contribute to the emerging literature on public perceptions of police at a municipal level in Mexico. It is especially important to emphasize that correct generalizations are difficult to achieve due to marked regional differences among Mexican municipalities. Thus, what follows should be considered an exploratory and provisional interpretation of the topic at hand.

II

REVIEW OF THE LITERATURE

1

Attitudes toward the Police

Arguably, the police are the most visible and obtrusive governmental control agency that interact with the general public.³² Citizens come into more frequent and direct contact with the police than they do with other state institutions dealing with control and authority. Walker and Richardson capture this relationship well in explaining that:

Whatever the citizen thinks of the police, they can hardly be ignored. Whereas other public bureaucrats are often lost from the public's view, locked in rooms filled with typewriters and anonymity, policemen are out in the world. They are on the sidewalks, and in the streets, in the schoolyard, and in the shopping mall, cruising, strolling, watching, as both state protectors and state repressors.³³

Some observers may claim that the above-mentioned routine police activities are relatively ordinary or trivial.³⁴ Nevertheless, some researchers argue that these police functions may indeed represent valid opportunities for continually assessing the way in which the police transfer the core values of the current political regime to the public.³⁵ In this sense, most citizens tend to grant legitimacy to the State and display willingness to respect the Rule of Law contingent on their relative perceptions of police personnel behavior. Owing to the close relationship between the public and the police, it is therefore important for administrators to identify citizens’ feelings toward the police in the hope of achieving collective goals such as that of social control. By identifying these public attitudes, administrators may also be in a position to more accurately gauge levels of citizen approval of the existing legal and political order.³⁶

Policy analysts rely on different instruments to assess police performance.³⁷ Examples of these instruments are police station records, crime statistics, and surveys. Yet, the use of surveys is more common owing to some reliability problems regarding the production of police and crime statistics.³⁸ In some cases, police departments may disseminate inaccurate statistics in order to protect or advance their own particular bureaucratic agendas. Policy analysts also prefer conducting citizen surveys because the information is obtained directly from those who receive police services. The results of citizen surveys can provide useful information to attempt to improve relations between the public and the police. They can also serve to increase the level of police accountability.³⁹

However, scholars argue that studying public attitudes toward the police may represent a complex enterprise.⁴⁰ To begin with, a multitude of variables affect how citizens view the police. The complexity increases because many societal groups differ in the way they conceptualize certain aspects of policing. Also, analysts often use different measures to assess attitudes toward the police. Furthermore, citizens tend to discriminate between diffuse and specific support when evaluating the police.⁴¹ They offer diffuse support when judging police performance in general terms and specific support when focusing on specific police practices, policies, or personnel. In some situations, the public may grant low levels of specific support to the police owing to aggressive police action, such as an arrest, but still evince favorable opinions toward the agency as a whole. Moreover, even though the impact of some variables relating to public perception of the police is consistent across studies, other variables still offer inconclusive results.⁴² Researchers are still unable to clearly understand the interactive effects among these variables.⁴³ This relative inconsistency in results precludes scholars from sharing a common methodological framework and calls for caution when attempting to draw theoretical generalizations on the study of attitudes toward the police. Nevertheless, the relatively scant empirical evidence available on perceptions of the police may still prove useful for administrators in terms of policy development.⁴⁴

Researchers generally study attitudes toward the police by considering variables at the individual, police-related, and contextual levels of analyses.⁴⁵ For instance, studies at an individual level reveal that variables such as age, gender, ethnicity, education, income, and social class all have an impact on perceptions of the police.⁴⁶ Younger people and those with higher levels of education, along with minorities, are more likely to display negative attitudes toward the police.⁴⁷ The gender variable shows that women tend to have more favorable attitudes toward the police than men.⁴⁸ In regards to variables like income and social class, some analyses yield inconclusive results, yet it is generally accepted that more affluent citizens or those belonging to a higher social class hold more favorable views toward the police compared to those from lower classes.⁴⁹ Some of the police-related variables include individuals’ prior contact with the police.⁵⁰ Additionally, citizens expect the police to be honest and provide equal treatment to all regardless of gender, ethnicity, or social class.⁵¹ Those who have experienced positive interaction with the police generally express approval towards this institution compared to those who have had negative encounters.⁵² Moreover, citizens evince more negative views toward the police when police officers initiate contact with the public (involuntary contact), but not when the public requests police services.⁵³ Citizens also disapprove of the police when these are perceived as dishonest and/or inconsistent in their treatment of citizens.⁵⁴

At a contextual level of analysis, social scientists include variables dealing with how safe residents feel in their neighborhoods, as well as those relating to personal experiences of victimization when living in these locations.⁵⁵ Other variables focus on the existence of neighborhood conditions, such as unrest or incivilities (bad manners), and varying levels of informal mechanisms of collective security.⁵⁶ On a more general level, scholars argue that citizens’ evaluations of how the broader political and legal systems operate also influence their views of the police.⁵⁷ Residents who perceive themselves as living in dangerous neighborhoods, or who have been victims of crime are more likely to evince negative opinions toward the police. Studies also indicate that citizens are more prone to disapprove of police practices if they live in neighborhoods where visible social and physical unrest prevails.⁵⁸ In contrast, research shows that significant levels of informal collective security correlate with residents’ approval of the police.⁵⁹ Finally, analysts posit direct or positive relationships between the attitudes toward the political and legal system and those toward the police.⁶⁰

Most analyses of these variables show that at an individual level only age and race yield consistent and significant impact on attitudes toward the police.⁶¹ Similarly, contact with the police, police integrity, neighborhood conditions, local crime perceptions, and trust in the legal and political system stand as the most important predictors when studying how citizens perceive the police.⁶² Regarding the rest of the variables, researchers still have not reached consensus on the effects such variables exert on the public's view of the police.⁶³

When conducting surveys via telephone or face to face, researchers generally try to identify attitudes toward the police by asking participants the extent to which they agree or disagree with statements like “the police do a good job,” and then linking the answers with some of the explanatory variables reviewed above.⁶⁴ However, this method precludes more in-depth information relating to the basics of how and why people form their attitudes toward the police. These types of responses have relatively little value in terms of informing police administrators about the specific areas on which to focus in order to improve the public image of the police.⁶⁵ Even the empirical findings on associations found in public perceptions of the police are only partially helpful in providing administrators with clear and specific improvement strategies. For instance, the police may be unable to alter the neighborhood's demographic or economic conditions.⁶⁶

Aware of these potential research limitations, some scholars propose additional strategies to try to obtain more specific information on attitudes toward the police. For example, they suggest using open-ended questions instead of closed ones when asking respondents about their views on the police.⁶⁷ Open-ended questions generate both advantages and drawbacks for researchers.⁶⁸ On the one hand, open-ended questions allow participants to use their own words to express what they believe is important when judging the police instead of choosing from a predetermined pool of answers selected by the researcher. On the other hand, these relatively unstructured questions allow subjects to give a wide variety of responses. This situation may pose challenges to researchers in terms of how to standardize, code, and make use of quantitative techniques arising from a broader range of responses provided by the subjects.⁶⁹

The disadvantages of using open-ended questions in surveys may limit the possibility of drawing generalizations that can be applied to other settings due to problems of the reliability of both the different responses given by the participants and the different ways in which the interviewers interpret and code the information.⁷⁰ In spite of these potential methodological limitations, some scholars argue that the overall advantages of open-ended questions overcome their drawbacks.⁷¹ As Bridenball and Jesilow explain:

For example, if in fact the public does have negative attitudes toward the police, the police need to know why there are negative attitudes, how they are being formed, and what factors are influencing their formation in order for the police to change the circumstances. The responses to the open-ended questions help to more accurately ascertain the participants’ thought with respect to these matters.⁷²

2

The Police in Mexico

Few observers or researchers would disagree on the dark history of police forces in Mexico as being characterized by pervasive corruption, abuse, and an inability to effectively maintain social control.⁷³ Scholars note that the origins of the corrupt nature of the police in Mexico can be partially explained as the result of the country's conflicting history of the creation of the State that began in the early 19^th century.⁷⁴ To achieve or consolidate State power, the rival groups participating in civil wars, independence movements, or revolutions needed to monopolize the means of violence. Irrespective of advancing class, regional, or cultural interests under the banners of either democracy or authoritarianism, the winners sought to control State power by significantly arming the forces capable of securing and maintaining the political elite's supremacy in State leadership.⁷⁵

The ensuing relationship between the State and the police became one of political convenience for the ruling class.⁷⁶ The State would grant the police significant levels of impunity and relative autonomy in exchange for political loyalty.⁷⁷ Impunity and autonomy meant the State condoned or forgave police forces for any wrongdoings against the public. It also meant the State allowed police administrators to manage the police according to their own particular agendas and ignore any internal problems of rampant corruption. In terms of the historical development of the police, López Portillo Vargas notes that police forces “were excluded from the modernization policies of the State and were consolidated as a privileged resource of repression and corruption.”⁷⁸

For some observers, the gradual transition of Mexico's political regime from an authoritarian State to a democracy would alleviate the problem of the corruption entrenched in the country's law enforcement institutions.⁷⁹ Nevertheless, scholars argue that democratization processes may have created more problems than solutions.⁸⁰ After the end of Mexico's one-party system and the beginning of multi-party competition at all levels of government, some police departments lost the benefits they used to receive from the regime in exchange for political allegiance. The disruption of the previous dysfunctional relationship between the State and the police has compelled some police agents to attempt to recoup lost privileges by increasing police discretion and resorting to even more corrupt practices.⁸¹ As Shirk notes, “police reform has not kept pace with Mexico's democratic regime change.”⁸²

In this sense, problems of corruption and ineffectiveness in achieving adequate levels of social control have stigmatized the police in Mexico.⁸³ Citizens typically view the police as part of the problem instead of the solution.⁸⁴ Newspapers from almost every part of the country continually report incidents of collusion between some police agents and criminal groups, human rights violations, and the police's overall inability to fight crime.⁸⁵

For these reasons, the public evinces very low levels of confidence in the police. When citizens distrust the police, they are significantly less likely to cooperate with the police to solve crimes, thus perpetuating problems of insecurity, corruption, and abuse.⁸⁶ For example, surveys show that 90 percent of Mexicans distrust the police.⁸⁷ As an institution, the police are one of the least respected in Mexico –just slightly above unions, legislators, and political parties.⁸⁸ Yet, as Zepeda Lecuona notes, “the heroism and sacrifice demonstrated by many honest uniformed officers requires that we make an effort to retain all that is beneficial among the police and reform what is not.”⁸⁹

3

Attitudes toward Mexican Police at a Municipal Level

Empirical studies on criminal justice affairs and the police in Mexico are generally limited in number, yet they are increasing. In Mexico, there is still a pattern of insufficient and unreliable data on crime and the police. Scholars indicate that in Mexico “accurate data gathering lags perhaps farthest behind in the law enforcement field.”⁹⁰ The unavailability of data can be explained by a number of factors. Before the 1990s, Mexico did not have a national official crime reporting system.⁹¹ Moreover, police administrators tended to discourage or simply prevent social scientists from studying the intricacies of police agency activities. In terms of academic research, educational programs and academic journals on criminal justice affairs only emerged in the 1990s both in Mexico and in other Latin American countries. Indeed, some Mexican scholars studying the problem of insecurity have had to write their articles in English to disseminate them.⁹² Finally, the reliability of official crime data is still questionable. For instance, law enforcement agencies generally fail to gather information about the victims and the circumstances surrounding the crime.⁹³

In general, most of the available studies focus on the overall performance of the police system in Mexico and/or on suggesting reforms needed to improve it. These studies also pay almost exclusive attention to the problems of insecurity, police corruption, and citizen perceptions of the police in Mexico City. Works on these issues at state or municipal levels, especially those concerning attitudes toward the police, are still scarce.⁹⁴

Nevertheless, there are some exceptions to the scant literature on public perceptions of the police at state and local levels in Mexico. For example, Olivero and Murataya conducted a survey to a non-random sample of 185 citizens from Guadalajara in the summer of 1997 to determine their satisfaction with police services.⁹⁵ Of the 185 participants, 49 (26 percent) had come in contact with the police within the previous 12 months.⁹⁶ Of those 49 citizens, only 1 person called the police to request a service or report a crime. For Oilvero and Murataya, this low number of participants willing to contact the police shows how effectively all citizens in the sample tend to avoid interactions with local law enforcement institutions. In general, the participants see community involvement with the police as a futile enterprise because the police side with the party offering the highest bribe regardless of the crime.⁹⁷

An additional research goal of the Olivero and Murataya project consisted of assessing whether respondents would consider it appropriate to increase the number of police officers on the street to fight crime more adequately.⁹⁸ Instead of the majority of participants responding in the affirmative, less than half of the subjects supported the idea of having more agents patrolling the streets. Olivero and Murataya argue that this negative attitude toward the police can be explained by the inefficiency of the services the police offer the public, which perhaps derives from the rampant problems of corruption in the country's law enforcement institutions.⁹⁹

Brown, Benedict, and Wilkinson led another study that attempts to empirically capture public perceptions of the police in Mexico.¹⁰⁰ In the summer of 2003, these scholars conducted a survey to 303 students from a law school in the city of Tampico, Tamaulipas.¹⁰¹ Using a stratified cluster sampling technique, Brown and his associates gathered data on perceptions toward the police. The researchers asked participants to give their opinions on whether they believed the municipal, state, and federal police were doing a good job in addressing different criminal offenses. They divided the types of criminal offenses according to the functions and jurisdictions of each police agency in question. Even though Brown and his colleagues attempted to use multilevel techniques to analyze the data, they did not find significant results, perhaps owing to the homogeneity of the sample and a significant amount of missing data.¹⁰²

In spite of these research limitations, descriptive statistics show that almost 80 percent of participants evinced negative attitudes toward the overall structure of police agencies in Mexico.¹⁰³ Municipal police forces received the worst evaluations in comparison to state and federal agencies. Brown and colleagues claim that their research is “the most detailed study of public perceptions of the Mexican police conducted to date in that this study gauged diffuse and specific evaluations of municipal, state, and federal police agencies in Mexico.”¹⁰⁴ Aware of the limitations of the quality of the sample, they nonetheless argue that the results of the study can help to empirically confirm the low levels of confidence Mexican citizens have in the police. Results also indicate that national efforts to reform police institutions have been insufficient in improving public views of the police.¹⁰⁵

Kwak, San Miguel, and Carreon also studied attitudes toward the police in Mexico.¹⁰⁶ Their research used data from the 1996 and 2005 surveys conducted by World Values Surveys, an organization dedicated to gathering information from different countries on health issues, politics, economics, social issues, and people's beliefs and values.¹⁰⁷ The total sample used by Kwak and his associates consisted of 3,924 adult respondents from different Mexican states (2,364 from the 1996 sample and 1,560 from 2005). By using multilevel analyses, they tested how factors such as age, education, marital status, levels of political legitimacy to the current political order, degrees of happiness and life satisfaction, religious activity, and size of town affected public confidence in the Mexican police. Statistically significant results show that positive attitudes toward the police are linked to political legitimacy, happiness, life satisfaction, frequency of religious activities, and being married. Negative views are associated with younger respondents, highly-educated people, and those living in bigger cities.¹⁰⁸

According to Kwak and colleagues, the study suffered from certain limitations.¹⁰⁹ Chief among these were the absence of relevant police-related variables associated with public views of the police, such as corruption, police brutality, or citizen-police encounters. Moreover, Kwak and the researchers claim it is difficult to generalize these findings to other settings as the research only included one country in its analysis. Nevertheless, the analysts point out that one can still generate certain policy implications from the study's findings.¹¹⁰ The significant statistical associations among higher degrees of happiness, life satisfaction, and political legitimacy relative to public confidence in the police indicate that governments need to focus on enhancing these variables. The relative success of crime prevention programs like Community-Oriented Policing or Neighborhood Watch depends on public trust toward the police and political order in general.¹¹¹

III

RESEARCH RATIONALE AND CONTEXT

Following the argument that the local level is key to understanding problems of citizen insecurity, this article focuses on municipalities in Mexico. Specifically, the focus centers on a relatively recent case of police misconduct that took place in the municipality of Ensenada, Baja California. Official investigators publicly announced that since 1999 a group of approximately twelve Ensenada policemen began protecting members of a regional powerful drug organization with headquarters in Tijuana.¹¹² These policemen offered the criminal group protection from both authorities and rival gangs. They also orchestrated kidnappings and assisted criminals in drug trafficking operations. Arrests linked to the incident began to take place in 2005. The leader of the corrupt police group would later be captured in 2008. As expected, the reputation and public perception of the police in Ensenada was severely damaged.¹¹³

In order to improve matters regarding crime and violence as well as public perception toward the police, municipal authorities made security issues the cornerstone of its administration policies.¹¹⁴ The main goals were to have an efficient police and to change citizens’ attitudes toward them. Policymakers’ chief expectation was to eventually recoup citizens’ trust by providing an honest, transparent, modern, and efficient police force.¹¹⁵ The municipal authorities made an unprecedented investment in the area of security from 2007 to 2009. This investment helped to fully equip the police force.¹¹⁶ Policymakers claimed that their security program had already started to produce positive results because by the end of 2008 violent and nonviolent crimes in Ensenada had decreased by approximately 16 percent. This decreasing trend in the incidence of crime continued in 2009. Local officials then highlighted that these achievements were indeed significant and unique for the municipality of Ensenada. They also stated that the positive outcomes originating from the local security plan were exemplary for the other four municipalities of Baja California and even for all municipalities in Mexico.¹¹⁷

With this scenario in mind, the goal of this study is to carry out an in-depth analysis on whether government efforts have indeed affected public perceptions toward the local police, and specifically to identify the main driving forces that affect people's attitudes toward police performance after police misconduct scandals.

IV

DATA, METHOD AND ANALYTIC STRATEGY

1

Data

At this point, it is important to highlight that in order to test hypotheses about public attitudes toward the police, this study borrowed variables from models used by some of the authors mentioned in the Literature Review section. Most of these studies drew on U.S. communities as research sites. Variables on race, ethnicity, and immigration were all significant in affecting public views on the police. Nonetheless, based on the research purposes of the particular work, such variables were excluded for two reasons. In terms of race and ethnicity, Mexico's population is not as heterogeneous as that of the United States.¹¹⁸ As to immigration, Mexico is generally considered a country that sends migrants abroad instead of receiving them. This situation marks a clear contrast with the U.S. case.

To gauge citizens’ attitudes toward a variety of public policies, especially those linked to criminal justice affairs, social scientists tend to rely more on survey research instead of official statistics to avoid biased reporting. As a result, this project followed the lead of other scholars working on similar topics and used survey research as the primary data collection method. The data reported in this article were collected from a survey administered to locals living in different neighborhoods in Ensenada during the fourth quarter of 2009.

When performing the exploratory data analysis, one issue was identified to have compromised the data. Specifically, the gender variable exhibited discrepancies in the population of Ensenada. While census parameters indicated 49.63% of females in Ensenada, the sample showed 57%. The discrepancies between sample and population parameters were ascribed to a higher likelihood of finding more women at home than men during the schedules selected to conduct the survey. As a result, the data were adjusted by means of a statistical software package in order to obtain similar proportions to those reported by official statistics for Ensenada.

2

Sample

For purposes of convenience, the local police department of Ensenada divided the neighborhoods into different police districts. To select the sample, members of the city police were contacted to obtain authorization to gather information about which neighborhoods correspond to each police district. Police department staff provided a list of the neighborhoods. The neighborhoods to be surveyed were randomly selected from this list. Once the neighborhoods of interest were selected, the police department also provided a list of the grids covering neighborhood streets that officers tended to patrol regularly. The streets on the grids were also randomly selected. Finally, households were systematically chosen to participate in the surveys. Following Jesilow, Meyer, and Namazzi, one to three dwellings from each street were selected.¹¹⁹

The author of this research and five other paid collaborators conducted the surveys. Only one person living at the dwelling in question was interviewed and all of the participants were at least 18 years old. Most of the interviews took place on weekday afternoons and some took place on weekends. The approval rate was about 70 percent and the final sample consisted of 221 participants.

3

Dependent Variable

Drawing from Benson, attitudes toward the police were measured using public perceptions of police performance as the dependent variable.¹²⁰ Participants were asked: “How would you rate the overall police service in the two or three blocks around your home?” The response options were: 1 (very poor), 2 (inadequate), 3 (adequate), 4 (good), or 5 (excellent). For purposes of analysis, options 1 and 2 were considered negative evaluations and options 3, 4, and 5, positive evaluations.¹²¹ To code this variable, a dichotomous or dummy variable was created in which 1 represented a “negative evaluation” and 0, a “positive evaluation.” In this sense, the outcome variable became “negative attitudes toward the police.” The independent variables then measured the likelihood of participants showing negative attitudes toward the police.¹²²

4

Independent Variables

The independent variables impacting negative attitudes toward the police were divided into three groups: individual, police-related, and contextual. The first group included demographic variables such as age, gender, education, and household income. Age was entered as a continuous variable. Gender was a categorical variable, coded 1 (male) and 0 (female). Education and income were ordinal variables. Education ranged from 0 (no high school diploma) to 3 (college graduate or higher). Household income ranged from 1 (less than $3,500 a year) to 7 (more than $20,500 a year).¹²³ Following Benson, education and income were combined to create a composite measure called social class. The categories for social class were “lower class” and “middle and upper class.”¹²⁴ Lower class consisted of only those participants with no more than “high school diploma” and a “household income” of less than $3,500. The rest of the participants who did not fulfill the above conditions were considered “middle and upper class.” The coding for social class was 1 for “lower class” and 0 for “middle and upper class.”

The next block of items dealt with police-related variables. The “contact with the police” variable was measured with the question: “Have you had any contact with the police in the last 12 months?”¹²⁵ It became a categorical variable coded as 1 (yes) and 0 (no contact). Police integrity was measured through two survey items.¹²⁶ The first item asked participants to express their level of agreement or disagreement with the following statement: “Policemen in your neighborhood are basically honest.” The second item asked: “The police in your neighborhood treat all citizens equally according to the law.” Response categories ranged from 1 (strongly disagree) to 5 (strongly agree). Police integrity was measured by using the combination of these two items to show acceptable reliability or internal consistency (Cronbach's alpha = .758). Based on Benson, the composite measure of police integrity was calculated as the sum of these two items.¹²⁷ Higher scores indicate higher police integrity.

The third group of variables included measurements at a contextual level of analysis. The variables were “broken windows,” “neighborhood crime,” “personal violent victimization,” “collective efficacy,” and “political legitimacy.” To test the broken windows variable, open-ended questions were used. The questions were: “What do you best like about where you live?” and “What do you least like about where you live?”¹²⁸ Following Bridenball and Jesilow, the focus was just on the aspects participants disliked about their neighborhoods.¹²⁹ Participants’ responses were then divided into categories based on whether or not complaints were police-related.¹³⁰ Police-related complaints about neighborhoods were coded as 1 and non-police related complaints as 0.

Drawing again from Benson, the following item was used to elaborate on levels of neighborhood crimes: “Do you think crime in your neighborhood has increased, decreased, or stayed the same in the last year?”¹³¹ Responses indicating that crime “had decreased” or “stayed the same” corresponded to the “not increasing” category and the “had increased” option to the “increasing” category.¹³² This crime-related variable was coded as 1 (increasing) and 0 (not increasing). In keeping with Sampson, Raudenbush, and Earls, personal violent victimization was gauged by asking: “While you have lived in this neighborhood, has anyone ever used violence, for instance in a mugging, fight, or sexual assault, against you or any member of your household anywhere in your neighborhood?”¹³³ Respondents had to respond yes or no, which was coded as 1(yes) and 0 (no).

The existence of informal security measures in neighborhoods was also an important factor affecting attitudes toward the police. Developed by Sampson and associates, collective efficacy was used to denote informal collective security.¹³⁴ Collective efficacy results from the combination of two composite measures: informal social control, and social cohesion and trust. To measure informal social control, the following questions were asked: (a) “How likely is it that your neighbors could be counted on to do something if children were skipping school and ‘hanging out’ on a street corner?” (b) “How likely is it that your neighbors could be counted on to do something if children were spray-painting graffiti on a local building?” (c) “How likely is it that your neighbors would do something if children were showing disrespect to an adult?” (d) “How likely is it that your neighbors could be counted on to do something if a fight broke out in front of their house?” and (e) “How likely is it that your neighbors could be counted on to do something if the fire station closest to your home was threatened with budget cuts?”¹³⁵ Social cohesion and trust were measured with five statements: (a) “People around here are willing to help their neighbors,” (b) “This is a ‘close knit’ community,” (c) “People in this neighborhood can be trusted,” (d) “People in this neighborhood generally do not get along with each other,” and (e) “People in this neighborhood do not share the same values.”¹³⁶

For social cohesion and trust, the possible answers ranged from 1 (strongly disagree) to 5 (strongly agree). In regards to informal social control, options fluctuated between 1 (very unlikely) to 5 (very likely) with the exception of items (d) and (e) which were reverse coded.¹³⁷ Reliability scores for social cohesion and trust, and informal social control were acceptable (Cronbach's alpha = .684, and Cronbach's alpha = .802). The combination of the informal social control measure and levels of social cohesion and trust generated the total collective efficacy measure. Internal consistency for collective efficacy was likewise acceptable (Cronbach's alpha = .807). Collective efficacy was calculated as the sum of the two items in question. Higher scores indicate higher collective efficacy.

Based on Kwak, San Miguel, and Carreon, the probable relationship between attitudes toward the police and political legitimacy was also tested.¹³⁸ Political legitimacy was measured by using two items. Both items sought to capture respondents’ trust toward government institutions other than the police. The questions were: “How much do you trust the municipal government?” and “How much do you trust the courts and the judicial system?” Responses ranged from 1 (no trust at all) to 5 (a lot of trust). For statistical purposes, the two items were combined to obtain a composite measure of political legitimacy. Internal consistency between the two items was acceptable (Cronbach's alpha = .743). Political legitimacy was computed as the sum of these items. Higher values mean greater political legitimacy.

5

Analytic Strategy

In order to evaluate how the independent variables affected negative attitudes toward the police, logistic regression was used as the principal analytical technique. This technique was selected because the operational definition given to negative attitudes toward the police resulted in a variable that was binary or dichotomous in nature.¹³⁹ Moreover, logistic regression allows for both continuous and dichotomous independent variables in the model.¹⁴⁰ Instead of directly estimating the expected value of the dependent variable (Y) based on selected independent variables, logistic regression also relies on the combination of predictors, but models the dependent variable on the probability that Y = 1. The assumption is that Y has been coded 1 = yes and 0 = no.¹⁴¹ For the purpose of this research, the focus was to determine the probability of a participant showing a negative attitude toward the police based on a set of independent variables. Adopting a pragmatic approach, a statistical software package was used to perform the relevant descriptive and multivariate analyses to help answer the key research question.

V

FINDINGS

1

Descriptive Statistics

As indicated in Section IV-1, the gender variable had to be adjusted, but the remaining sample demographic variables were consistent with the census population parameters for Ensenada. All descriptive data of the variables included in this study are presented in Table 1.

TABLE 1.

DESCRIPTIVE STATISTICS (N = 221).

Variables	Mean	SD	Min	Max
Attitudes toward the police	.35	.478	0	1
Negative (1)	(34.8%)
Positive (0)	(65.2%)
Age	36.18	13.487	18	76
Gender	.50	.501	0	1
Male (1)	(50.4%)
Female (0)	(49.6%)
Social class	.29	.453	0	1
Lower (1)	(28.6%)
Middle/upper (0)	(71.4%)
Contact with police	.47	.500	0	1
Yes (1)	(46.6%)
No (0)	(53.4%)
Police integrity	5.75	1.935	2	10
Broken windows	.41	.493	0	1
Police-related complaint (1)	(40.9%)
Non-police related complaint (0)	(59.1%)
Neighborhood crime	.27	.445	0	1
Increasing	(27.1%)
Not increasing	(72.9%)
Victimization	.24	.430	0	1
Yes (1)	(24.3%)
No (0)	(75.7%)
Collective efficacy	32.92	6.915	10	50
Political legitimacy	5.87	2.138	2	10

Note: The percentages for categorical variables are listed in parentheses.

Regarding the respondents’ evaluations of the police, a little bit more than one-third of respondents (n = 77) indicated that the services the police provided were “very poor” or “inadequate.” Almost two-fifths (n = 85) perceived the police provided “adequate” services, and just over one-quarter (n = 59) rated their services as “good” or “excellent.” Overall, about two-thirds of participants (n = 144) had positive perceptions about police performance and the remaining third (n = 77) held negative views.

Although not used for statistical analyses and underscoring the relevance of using open-ended questions in conducting surveys, participants had the opportunity not only to evaluate police performance, but also to openly mention the specific things they liked or disliked about the municipal police. On the one hand, 51% (n = 112) had positive comments about the local police and 49% (n = 109) refrained from mentioning positive items. Some participants provided more than one positive comment and 137 responses were counted. Most of the praise for the police referred to the newly acquired equipment and the promptness with which they responded to citizens’ calls. On the other hand, 73% of the interviewees (n = 161) complained about the police and 27% (n = 60) did not share their opinions on the subject. Again, some individuals offered more than one negative comment. A pool of 237 complaints was created. Corrupt practices and arrogance stood as the main complaints from the public. Tables 2 and 3 specifically list the positive and negative comments that participants had about the police, along with their frequency and corresponding percentages.

TABLE 2.

PRAISE ABOUT THE ENSENADA MUNICIPAL POLICE.

	Frequency	Percentage
Praise to the police
Good equipment	36	26.3
Good response time	31	22.6
Officers are efficient/reliable/professional	21	15.3
More police presence/good visibility	13	9.5
Officers are kind/respectful	9	6.6
I feel protected	8	5.9
Officers are cooperative	8	5.9
The police engage in community-oriented policing	4	2.9
Good appearance	4	2.9
New facilities	1	.7
Discounts on fines	1	.7
Honesty	1	.7

TABLE 3.

COMPLAINTS ABOUT THE ENSENADA MUNICIPAL POLICE.

	Frequency	Percentage
Complaints about the police
Corruption	53	22.4
Arrogance	46	19.4
Slow response time	25	10.6
Abuse of authority	23	9.7
Poor police tactics	19	8
Police officers do not respect the law they enforce	12	5
Police collude with criminals	8	3.4
Service is not community-oriented	8	3.4
Lack of security	7	3
Lack of judgment	6	2.5
Police brutality	5	2.1
Poor appearance	5	2.1
Lack of professionalism	4	1.7
“I dislike everything”	4	1.7
Weak enforcement of laws	3	1.3
Police officers are ignorant	2	.8
Police officers are not trustworthy	2	.8
Lack of moral values	1	.4
Expensive equipment	1	.4
Lack of equipment	1	.4
Not enough officers	1	.4
Apathy	1	.4

Continuing with open-ended questions, participants were asked to give possible suggestions on how the municipal government could specifically improve security issues in the city. Nine-tenths of them (n = 197) agreed to do so. They were also free to offer more than one suggestion. After reviewing the answers, a set of 291 responses was obtained. Among the many suggestions, the public emphasized tackling corruption and promoting police honesty and accountability, having more police presence in their communities, and professionally training police officers. A detailed categorization of the suggestions is provided in Table 4.

TABLE 4.

SECURITY-RELATED SUGGESTIONS MADE TO THE MUNICIPAL GOVERNMENT.

	Frequency	Percentage
Tackling corruption and promoting honesty and accountability of police officers	43	14.8
More police presence and patrolling	41	14.1
Professionally training police officers	32	11
Better salaries	18	6.2
Better police recruitment processes	14	4.8
Better police equipment	13	4.5
Police operational efficiency	13	4.5
Better police response time	12	4.1
Police need to get involved with the community	12	4.1
Police intelligence/strategic planning	11	3.8
Better infrastructure services (traffic signs, speed bumps, street lighting, recreational facilities)	10	3.4
“Just do the job!”	9	3.1
Being more respectful and fairer treatment by the police	9	3.1
Police officers need to show serious commitment to the police profession	6	2.1
Police rotation	5	1.7
More cooperation from the public	5	1.7
Crime prevention programs	4	1.4
Efficient enforcement of laws	4	1.4
Better city budget management	4	1.4
Focusing on strengthening society's values	3	1
The police need to trust citizens	3	1
Use of surveys	2	.7
Better municipal government services (less bureaucracy) More military presence in the area	2	.7
No arrogance and no abuse of power	2	.7
Harsher penalties for criminals	2	.7
More financial resources for security	2	.7
Better police facilities	2	.7
Lower traffic fines	2	.7
Independent agencies auditing the police	1	.3
Providing support to crime victims	1	.3
Improved intergovernmental coordination	1	.3
Reduction of police personnel	1	.3
Public servants need to set the example of	1	.3
appropriate behavior	1	.3

Taking into consideration variables at the contextual level of analysis, ninety-one percent of the participants (n = 202) agreed to share the things they liked best about the neighborhood. Some respondents gave more than one answer and a pool of 272 responses was obtained. Of the 272 responses, a quiet neighborhood ranked first (32%) followed by location (18%), aesthetics/amenities (16%), good neighbors (9%), city infrastructure (8%), and good security (5%), to name a few.

For items indicating the things respondents least liked about where they live, the responses were separated into complaints that were related or unrelated to the police. This categorization allowed for the creation of the “broken windows” variable that was subsequently used in the statistical analyses. 95% of respondents (n = 211) complained about the neighborhood and the other 5% (n = 10) did not have any complaints. The complaints of 57% (n = 121) were considered unrelated to the police. Some of them gave more than one complaint and a set of 191 complaints was obtained. Among the complaints, they mentioned lack of neighborhood infrastructure¹⁴² (34%), a lack of neighborhood aesthetics or amenities¹⁴³ (20%), traffic (8%), stray dogs (7%), noise (7%), strained relations between neighbors (6%), inconvenient locations (5%), and abandoned houses/lots (3%), among others. The other 43% of the participants (n = 90) complained about issues that they thought the police could intervene to fix them. Participants also were allowed to express more than one complaint and a total of 119 responses were counted. The most frequent police-related complaint was the perception that they were living in neighborhoods that “lacked security.” Table 5 lists all of the police-related complaints, their frequency, and the corresponding percentages.

TABLE 5.

POLICE-RELATED COMPLAINTS ABOUT NEIGHBORHOODS.

	Frequency	Percentage
Lack of security	25	21
Gangs	14	11.8
Loitering	13	10.9
Vandalism	12	10.1
Drugs (use, addicts, dealing)	11	9.2
Graffiti	10	8.4
The police (inadequate police protection, police harassment, and lack of patrolling)	9	7.6
Transients or homeless	7	5.9
Traffic violations	6	5
Crime/delinquency/violence	6	5
Groups of teenagers or others hanging out and harassing People	2	1.7
Bad neighbors (fighting, arguing, violence, being hassled by neighbors)	2	1.7
Street fights	1	.8
Disorder/unrest	1	.8

2

Multivariate Analyses

Logistic regressions were used to assess the likelihood of citizens to express negative attitudes toward the police. At an individual level, it was expected that age would lower the likelihood of respondents with negative attitudes toward the police. In contrast, being male and belonging to a lower class would increase it. Contact with the police and police integrity correspond to the police-related factors. Whereas previous contact with the police was thought to increase the probability of holding negative views, police integrity was expected to reduce it. As to contextual variables, the expectation was that broken windows, neighborhood crime, and personal violent victimization would increase the likelihood of negative attitudes toward the police. Collective efficacy and political legitimacy were thought to produce the opposite effect.¹⁴⁴

Before conducting the logistic regressions, tests were conducted to identify whether multicollinearity or outliers were present in the model. Neither multicollinearity (Variance Inflation < 1.29; Tolerance > .77) nor outliers were present.

Three logistic regression models were run. The first model included individual variables: age, gender, and social class. The second model added police-related variables: contact with the police and police integrity. The third model added contextual variables: broken windows, neighborhood crime, personal violent victimization, collective efficacy, and political legitimacy. Table 6 displays the results of the logistic regressions of negative attitudes toward the police using independent variables. In Table 6, b stands for unstandardized logistic regression coefficients (standard errors are presented in parentheses), OR for the odds ratio of coefficients, and βR for semi-standardized logistic regression coefficients. Based on Roncek, semi-standardized coefficients were obtained by multiplying each unstandardized coefficient by its corresponding standard deviation.¹⁴⁵ Variables showing larger coefficients (in absolute values) were considered stronger predictors in the equation.¹⁴⁶

TABLE 6.

PREDICTING NEGATIVE ATTITUDES TOWARD THE POLICE.

	Model I			Model II			Model III
Variable	b (SE)	OR	βR	b (SE)	OR	βR	b (SE)	OR	βR
Age	-.007	.993	-.094	.012	1.012	.162	.018	1.019	.243
	(.011)			(.013)			(.014)
Male	.259	1.295	.123	.278	1.320	.139	.091	1.096	.046
	(.284)			(.315)			(.357)
Lower class	-.273	.761	-.124	-.044	.957	-.019	.104	1.109	.047
	(.333)			(.374)			(.417)
Contact with police				.017	1.018	.008	-.345	.709	-.173
				(.323)			(.384)
Police Integrity				-.574^***	.564	-1.111	-.536^***	.585	-1.037
				(.101)			(.114)
Broken Windows							.472	1.603	.233
							(.361)
Neighborhood crime							1.902^***	6.697	.846
							(.408)
Victimization							.316	1.372	.136
							(.462)
Collective efficacy							-.013	.987	-.089
							(.028)
Political legitimacy							-.177^*	.838	-.378
							(.091)
Constant	-.432	.649		1.938^***	6.946		2.325^*	10.22
	(.436)			(.683)			(1.206)
Nagelkerke R^2	.015			.254			.440

N = 221

^*P < .10

** P < .05

^***P < .01

For Model I, none of the individual-level variables were statistically significant on negative attitudes toward the police. Model II revealed improved explanatory power when police-related variables were included in the equation. The overall model was significant (-2 Log Likelihood = 240.772, model X² = 44.953, with 5 degrees of freedom, and P = .000) and the independent variables were able to account for 25% of the variance in negative attitudes toward the police. In Model II, police integrity was the only statistically significant variable (P = .000).

Model III, the final and most complete model, added the contextual variables. The overall regression model was statistically significant (-2 Log Likelihood = 200.833, model X² = 84.893, with 10 degrees of freedom, and P = .000). Predictors in the model were able to explain 44% of the variation regarding the participants with negative attitudes toward the police. Police integrity remained statistically significant (P = .000). Neighborhood crime and political legitimacy were also statistically significant (P = .000 and P = .051). In this model, controlling for the rest of the independent variables, each unit-increase in police integrity decreased the odds of negative attitudes by 42%. Similarly, political legitimacy reduced the odds of negative attitudes toward the police. On average, each one-unit increase in political legitimacy was associated with a .177 decrease in the log odds of negative attitudes. As to the odds ratio, each one-unit increase in political legitimacy increased the odds of negative attitudes by .838 or lowered them by 16%. The perception of increasing levels of crime in the neighborhood while keeping the rest of the predictors constant increased the odds of negative attitudes by a factor of 6.697 (a 597% increase in the odds of negative attitudes). Overall, police integrity remained the strongest predictor of negative attitudes toward the police indicated by the magnitude of its βR.

VI

DISCUSSION AND CONCLUSION

In the multivariate analysis, the most important explanatory variables were police integrity followed by perceptions of neighborhood crime and political legitimacy. The rest of the independent variables lacked statistical significance. Despite showing the expected signs and even significant correlations with negative attitudes (.231, P < .01; -.211, P < .01), broken windows and collective efficacy did not become significant determinants when incorporated into the complete logistic regression model. The results contrast with those of Cao and associates who find both variables statistically significant predictors of attitudes toward the police.¹⁴⁷ Moreover, the results are not in line with those of Jesilow and colleagues who find broken windows the best predictor of negative attitudes.¹⁴⁸ Nonetheless, the previous studies conducted in U.S. cities did not include variables like police integrity, perceptions of neighborhood crime, or political legitimacy. This situation may help explain the differences in causes of negative attitudes toward the police. Based on the results of previous studies like those of Bridenball and Jesilow, it was also expected that contact with the police would exert a significant impact on negative attitudes.¹⁴⁹ Yet, this was not the case. The fact that most participants in the current study voluntarily contacted the police may indicate the limited influence this predictor may have on unfavorable views toward the police.

As to the variables that marked statistical significance, public perceptions of the severity of neighborhood crime exerted a significant impact on negative attitudes toward the police. The results are consistent with Benson, who finds that citizens are more critical towards the police when they perceive neighborhood crime on the rise.¹⁵⁰ These results also coincide with those of Chermak, McGarrell, and Gruenewald, in whose model the perception of crime in one's neighborhood is found to be a consistent predictor of attitudes toward the police.¹⁵¹ As concerns about neighborhood crime increase, citizens tend to display more unfavorable views of the police.¹⁵²

Political legitimacy was another variable exerting a relatively strong impact on negative attitudes toward the police. The more citizens trust political institutions, the less likely they will have unfavorable views toward the police. These results are similar to those of Kwak and associates who find a statistically significant relationship between political legitimacy and public confidence in Mexico's police.¹⁵³ Using a sample of Latin American countries, Cao and Zhao also find trust in the political system a strong predictor of confidence in the police.¹⁵⁴ Overall, the results of this article support Albrecht's and Green's argument who claim that “public attitudes toward the police do not exist in isolation, but are a part of broader complex of attitudes toward the system of legal justice and its various representatives.”¹⁵⁵

Nonetheless, the results show that in the Mexican context and within the boundaries of this study, police integrity tends to markedly overshadow the influence of the rest of the predictors. For this project, police integrity measured police honesty and whether the police treated everyone according to the law. In keeping with the model, positive perceptions of police integrity significantly dampen the likelihood of individuals holding negative attitudes toward the police. The results are relatively consistent with the comments on the things citizens like least about the police and the suggestions for improving public security in the city. As Tables 3 and 4 show, reducing corruption and encouraging police accountability are the aspects most citizens deem key in improving citizen security in general and police services in particular.

Analysts generally explain public corruption as “the misuse of public office for private gain.”¹⁵⁶ In Mexico, official promises and efforts to eradicate police corruption and other types of corruption have come and gone for decades. Since the early 20^th century, incoming federal administrations have promised to seriously tackle corruption.¹⁵⁷ Yet reality consistently shows a radically different story. Not even two democratic federal administrations that did not belong to the Partido Revolucionario Institucional [Institutional Revolutionary Party] were able to curb corruption when they held power from 2000 to 2012.¹⁵⁸ Corruption in Mexico may have even increased in these two presidential terms.¹⁵⁹ Moreover, the return of the Partido Revolucionario Institucional (PRI) in December of 2012 has not brought any better news on the corruption front.¹⁶⁰ In the early 1980s, Riding made a reflection about Mexico's corruption that is no less applicable for today's circumstances: “even good intentions are crushed by reality: the pledges of incoming administrations to clean up corruption look naïve or cynical six years later.”¹⁶¹

Admittedly, there have been some positive changes to Mexican police institutions recently.¹⁶² Chief among these are larger budgets for police departments.¹⁶³ More financial resources have allotted for more and relatively better police personnel, equipment, and technology to combat crime and maintain social control more efficiently. Along these lines, Table 2 on praise about the Ensenada police shows that the respondents most frequently extolled the acquisition of “good” equipment. Continuing with changes that have been made to Mexican policing, there is now a renewed official discourse and consensus on the need to professionalize the police and develop institutional mechanisms to facilitate coordination among the different law enforcement agencies belonging to the municipal, state, and federal levels.¹⁶⁴

Yet, official promises, good intentions, and even actions undertaken to change police agencies in Mexico have continuously been overshadowed by undisputed evidence of police corruption, abuse, and operational inefficiency.¹⁶⁵ Under this scenario, it is difficult for the average Mexican citizen to understand police behavior. On the one hand, citizens know about official efforts to improve police services, but on the other, they are constantly bombarded by news of growing insecurity and persistent official corruption.¹⁶⁶

Researchers have identified some underlying causes of police corruption actually provided by the police officers themselves. At a basic level, a low salary “promotes and even justifies corruption.”¹⁶⁷ Aside from this, others argue that most join the force unconcerned about the limited income because they anticipate getting rich by simply wearing the uniform.¹⁶⁸ Institutional limitations also play a role in fostering corruption. In some cases, unscrupulous police academy instructors sell exams or ranks, or condone arrests in lieu of a bribe. In other cases, instructors charge for bullets if students want to participate in shooting training sessions.¹⁶⁹ Furthermore, police officers oftentimes need to obtain some basic benefits and working tools such as life insurance, uniforms, or bulletproof vests on their own.¹⁷⁰ From a cop's perspective, it is better to incur these costs than to lose the opportunity to patrol the streets. Being out to the field generally allows patrolmen to achieve the sought-after “extra income.” The additional money most likely offsets incurred costs and permits the officer to make a profit.¹⁷¹

Yet, police officers do not pocket this additional income in its entirety. Any money from extortion needs to be enough to “buy” the job and to satisfy the quotas demanded by their immediate superiors.¹⁷² Some police chiefs may charge for posting an officer in a specific geographical sector or riding a patrol car or motorcycle.¹⁷³ Viewed under this light, some police officers have to resort to extortion to compensate for inadequate working conditions, low salaries, and significant pressure from superiors to meet the required financial quotas.¹⁷⁴

In their defense, policemen argue that one key factor often omitted when assessing police corruption is the deliberate participation of citizens.¹⁷⁵ On the corrupt police-citizen relationship, some policemen say:

People think that all police officers are corrupt, but corruption starts with citizens because it's easier for them to speed up their business and save time by “greasing the wheels” with cash. The government allows many things. Corruption is rampant, and since we don’t get good social benefits because of the economy, a policeman allows himself to be corrupted.

I would like to ask citizens and the media: why are they so keen on putting the blame on us if there is corruption everywhere in this country? Even several government authorities and leaders have stolen money from the Mexican people.

There are others who steal millions and get immunity. But when a policeman steals four pesos, he is persecuted.¹⁷⁶

Although not all Mexicans are corrupt, it is indeed widely acknowledged that corruption permeates Mexican society.¹⁷⁷ From the policemen's statements, it is possible to identify the explicit assumption that corruption is justified because everyone participates, particularly those holding important governmental posts. If those running the State, supposedly chosen to lead efficiently, honestly and based on the Rule of Law, are blatantly corrupt, then society in general may lack incentives to behave differently.¹⁷⁸ Indeed, analysts agree that the great fortunes associated with “grand-scale” corruption are made at the highest governmental levels.¹⁷⁹ The repertoire of such tricks may include “kickbacks, theft, embezzlement, [and] insider deals,” among others.¹⁸⁰

For example, recent news media accounts detail the inner workings of how federal monies earmarked for public security are handled. A group of Mexican mayors publicly denounced some federal congressmen for assigning resources contingent upon the fulfillment of certain requirements by using federal officials as intermediaries.¹⁸¹ Some mayors reported that from 2006 to 2012 they were approached by federal bureaucrats from the security area who “suggested” forms of handling these resources. The assigned resources were specifically aimed at the improvement of municipal police departments, which would in general terms consist of higher salaries, police training courses, and better operating facilities. The resources also help to buy modern communication and surveillance devices and other implements for the job such as weapons, patrol cars, bulletproof vests, and uniforms.¹⁸² Some federal bureaucrats provide mayors with a specific list of suggested suppliers to fill the needs of municipal police departments. The distribution of resources would then be blocked, delayed, expedited, or even increased in subsequent years, contingent upon whether mayors do business with the recommended providers.¹⁸³

Nevertheless, federal bureaucrats are not the only actors who are involved in corrupt schemes in the distribution of federal monies. High-level local policymakers and police officers themselves also play a role in corruption. Local bureaucrats can participate in corrupt activities by hiring private companies whose owners are relatives or close allies. To illustrate this, municipal authorities can hire one of these suppliers and overpay him or her artificially inflated prices when buying police gear such as bulletproof vests, among others.¹⁸⁴ The mechanics of this form of corruption is not new. Writing in the early 1980s, Riding noted similar corrupt schemes between top bureaucrats and contractors.¹⁸⁵ As to the police, some police chiefs can also dishonestly and arbitrarily manipulate resources to benefit close aides or punish adversaries or the rank and file.¹⁸⁶

In a democratic setting, the pernicious effects of corruption are strongly associated with lower confidence in the legitimacy of the regime. Declining trust in political institutions tends to inhibit collective actions between the public and the government (through law enforcement agencies) to jointly address issues like citizen insecurity, among others.¹⁸⁷ As noted above, citizens interact more frequently and directly with policemen than with public servants from other government offices. As such, citizens generally view police performance as a reflection of the performance of government as a whole. Whereas an incompetent police force tends to alienate the public from political institutions, an effective and honest police enhances the scope of mutual cooperation and strengthens the regime's legitimacy.

Some Mexican scholars ascribe Mexico's persistent problems of disorder and corruption to a fundamental lack of legitimacy in the political regime.¹⁸⁸ For example, history shows that Mexicans have long lived in conditions of a weak Rule of Law, which presupposes accountability both for State and non-State actors, as well as equal access to justice under the law.¹⁸⁹ Rule of Law also allows citizens to interact with each other in coherent and predictable environments that protect property rights and the enforcement of contracts.¹⁹⁰ Adherence to laws and rules largely hinges on the degrees of people's trust toward the regime.¹⁹¹ In Mexico, most people perceive the law as subservient to the wealthy and powerful. Citizens usually obey the law to avoid punishment instead of willingly working with the authorities. Low credibility of the regime, Rule of Law, and political institutions reduces the State's capacity to effectively enforce laws and maintain social control.¹⁹²

In a scenario of widespread disrespect for the law, citizens adopt a rather cynical view of government leaders and their actions.¹⁹³ Even governmental actions targeting corruption or other illicit acts are viewed with suspicion. Owing to significant inconsistencies in law enforcement, citizens wonder whether the arrests of some corrupt politicians, union leaders, or criminal bosses are legitimate or politically motivated.¹⁹⁴ Citizens get even more confused on learning that these individuals are indeed arrested but receive special privileges while in prison.¹⁹⁵

Given the omnipresence of the State in the lives of its citizens, the conduct of policymakers strongly influences how citizens behave.¹⁹⁶ If citizens are aware that State officials are corrupt, it is very likely that most citizens will behave in a similar fashion. Citizens might even adopt an “anti-State/pro-society bias”;¹⁹⁷ that is, they may promptly dismiss any official discourse to improve matters and embrace all the proposals from civil society that demand accountability. Most citizens will comply with the law, but hesitantly and out of fear. At the same time, they will protest against the state of current affairs by misbehaving and attempting to get away with illicit acts like most policymakers and other political actors do. An environment like this produces feelings of disillusionment and pessimism in the public when considering how to reverse the situation.¹⁹⁸ It may also lead citizens to finally tolerate, and even excuse, official corruption in exchange for relatively effective public policy.¹⁹⁹

When ordinary residents become aware of the mismanagement of public policy programs like those targeting insecurity, enlisting citizen assistance and cooperation may be extremely problematic. Without the participation of citizens, the original goals of joint actions between the public and the police regarding crime prevention disappear. Kratcoski, Das, and Verma note:

If the community is to have a major impact on crime prevention, citizens must have a value system grounded in the beliefs that it is wrong to violate laws, that the rights of others must be respected, and that those who do not accept these premises and do violate the laws should be punished.²⁰⁰

If we strictly apply the above criteria to the Mexican case of securing citizen involvement in crime prevention, it can be argued that the prospects of joint efforts with the police, at least in the short-term, are discouraging. For instance, the value system based on the beliefs that violating laws is wrong is not necessarily in place. Some surveys show that most Mexicans have a relatively high tolerance of illegality. Indeed, there is a widespread view that violating the law is not serious.²⁰¹ Littering, not using a seatbelt, or buying bootleg goods is generally not seen as illegal.²⁰² Moreover, some would categorize fellow citizens as “stupid” if they obey the law in cases where adverse consequences of non-compliance are minimal or nil.²⁰³ The signals some citizens send is that it is fine to disobey the law and that one only needs to be crafty to avoid “getting caught.”²⁰⁴ Environments featuring these characteristics are unsuitable for obtaining citizens’ cooperation in crime prevention or in any other area of public interest that requires collective actions.

The question that follows is why it is that most Mexicans display strong attitudes toward disobeying the law. Citizens tend to disobey the law because they perceive Mexico's legal system and enforcement institutions as inefficient and influenced by the rich and powerful. Citizens distrust policymakers in charge of administering justice and confer low legitimacy to the government and its institutions. The principal cause of mistrust and illegitimacy is the actual and perceived corruption of government officials.

Issues of police integrity and political legitimacy are two of the most important findings in this study of negative attitudes toward the police. Both variables are likely to lower citizens’ propensity to display negative attitudes. To improve public views of the police and citizen security, police administrators and politicians in general need to design, implement, and evaluate strategies that reduce corruption and gain citizens’ trust.

To mitigate police corruption in Mexico, some analysts recommend reviewing and altering police department compensation and personnel management policies.²⁰⁵ They consider the U.S. military a successful guiding model of effective workforce management that relies on economic incentives that may be applied to the Mexican case.²⁰⁶ Mexican policymakers and police administrators ought to focus on restructuring promotion schemes, pay officers more, and use a seniority-based pay system. Productivity bonuses are also needed and recruitment standards must be stricter.²⁰⁷ Merit ought to be the main guideline for promotions, salary increases, bonuses, and recruitment procedures rather than favoritism or any other form of corrupting influence. The proper implementation of these mechanisms is thought to increase police productivity and foster honesty in Mexico's police departments.²⁰⁸

Naturally, political and law enforcement leaders must be honest themselves and work in the public interest to increase the likelihood of the successful implementation of the above mentioned police-related changes in human resources polices.²⁰⁹ It can be said that honest leadership in any type of official agency is key in reducing corruption. As Tanzi puts it:

When the top political leaders do not provide the right example, either because they engage in acts of corruption or, as is more often the case, because they condone such acts on the part of relatives, friends, or political associates, it cannot be expected that the employees in the public administration will behave differently.²¹⁰

Even though the causes of and solutions to corruption are indeed multifaceted, researchers tend to agree that the critical ingredient to successfully curb official corruption is the political will of an honest and visibly committed leadership that shows no tolerance for corruption.²¹¹ More police in the streets, up-to-date security programs, more enforcement tools, or more security-related laws will be insufficient to improve Mexico's security crisis unless political leaders seriously commit themselves to strengthening the legitimacy of the regime through effective public policies and a reduction of acts of corruption. Indeed, it will be difficult, if not impossible, to attain a functioning and efficient Mexican police (and thus lower levels of insecurity) if they perform within a corrupt and even authoritarian political regime.

Having said this, has democracy made a difference in ending problems of police officer and politician corruption, crime and violence, as well as the culture of unaccountability in Mexico? Not really. Therefore, it appears that corruption, insecurity, and a lack of accountability are only symptoms of a more profound structural problem in Mexico's political system that is now in the democratic stage.

For Rose and Shin, the problem is that most third-wave democracies, including Mexico, tend to democratize backwards.²¹² Democratization backwards means that, unlike first-wave democracies, policymakers from the latest wave of democratization opted for introducing competitive elections before securing the basic institutions of a modern State: Rule of Law, a free and participative civil society, government accountability, and the effective separation of powers.²¹³ Whereas policymakers in advanced democracies made sure to first consolidate a strong and modern State before allowing universal suffrage and other political benefits, those from third-wave democracies mostly chose an inverse democratization sequence.²¹⁴

Adopting this logic, policymakers and politicians running democracies that democratized backwards now face a number of significant challenges. They need not only to organize, compete for, and monitor relatively clean elections, but also to simultaneously deliver effective public policies, build up Rule of Law and accountability mechanisms, while ensuring citizens’ meaningful participation in political processes.²¹⁵ The challenges are daunting. Even for policymakers genuinely committed to efficient and honest governments, the magnitude of the challenges may resemble instances like “trying to build an airplane while you’re flying it”²¹⁶ or “attempting to repair or replace a faulty parachute while free falling.”²¹⁷

The case of increasing insecurity in Mexico attests to the lack of a relatively modern State capable of ameliorating this particular and pressing problem. First, citizen complaints about police extortion, arrogance, repression, impunity, and corruption all point to weak Rule of Law. Second, citizens tend to grant the police and other governmental institutions low levels of legitimacy. As a result, citizens gradually develop feelings of alienation from the broader scope of political and legal systems. Strong feelings of alienation make citizens’ meaningful participation in Mexico's political life less likely and may also adversely affect the creation of a strong civil society. Third, the collusion among the members of different branches of government to tolerate corruption denotes deliberate actions to avoid accountability. One case in point is the corrupt schemes between some members of the legislature and the executive branch when distributing security-related monies to states and municipalities. For these reasons, it can be argued that the construction of a modern State in Mexico is still a work in progress.

According to Rose and Shin, policymakers operating democracies lacking the basic institutions of the modern State are likely to face one of three scenarios.²¹⁸ One is the completion of the democratization process by building up a modern State. The next scenario is the total abandonment of democratic practices for alternative political models. The last one is a democratic regime where relatively free elections may persistently and indefinitely coexist with weak institutions of a modern State. Rose and Shin call the last political model an incomplete or “broken back” democracy.²¹⁹

Broken back democracies entail certain economic and political costs though. Economically speaking, domestic and foreign entrepreneurs tend to limit investments in market-oriented economies (like Mexico) featuring a weak Rule of Law.²²⁰ Without a relatively strong Rule of Law, the full protection of property rights and the enforcement of contracts are unpredictable. On the political front, it is important to highlight that even though free elections are necessary for democratization, they are not enough to complete or consolidate a democracy.²²¹ Broken back democracies generally consist of two interacting actors: on the one hand, a set of corrupt, untrustworthy, and unresponsive policymakers; and on the other, a resigned or complacent public accepting their rulers’ limited performance. Citizens may claim that an incomplete democracy “is a lesser evil” in comparison with the totalitarian or authoritarian regimes of the past.²²² When a broken back democracy is institutionalized, citizens enter a stage of “idiotization.”²²³ According to Rose and Shin, idiotization refers to “the conscious rejection of the obligations of a citizen…at the individual level, this is a rational reaction to a government that is unaccountable and unresponsive to demands.”²²⁴

The price citizens pay for a broken back democracy is rather high. A broken back democracy tends to preclude citizens from truly reaping the benefits of a consolidated democracy and even those of a market economy. Focusing on the pernicious effects of a weak Rule of Law, particularly manifested in high levels of corruption, Tanzi states:

The widespread disillusion among the population of some economies in transition and some developing countries with both market economies and democratic processes is very much provided by the widespread corruption that prevails in these countries and that is wrongly attributed to the market economy and the democratic processes.²²⁵

Based on the information provided in this study, and using Rose and Shin's categorizations of democracy, it can be argued that Mexico is on the path of completing democratization, yet it hangs by a very thin thread. Mexico seriously risks falling into the broken back democracy category, or even into that of the total repudiation of democracy, especially if problems of low levels of political legitimacy and ubiquitous corruption remain largely unaddressed in the immediate future.

The causes of and solutions to problems of crime and violence are complex. Many variables enter into the equation such as extreme poverty, misdistribution of wealth, a lack of education and employment opportunities, poor health, and even inadequate housing.²²⁶ Nonetheless, and based on this study of the Mexican case, it seems that the root causes and possible solutions to the problem are of a more political nature. Yes, Mexico has made progress on many fronts, yet more is needed to unleash the country's significant potential. The construction of modern State institutions, an honest and committed political leadership working in the public interest, and citizens’ genuine engagement in the country's political life are key in improving not only the security situation and thus the reputation of the police, but also other public policies and institutions of Mexico's nascent democracy.