Introduction
Gauging performance on tasks provides valuable information that individuals and groups can use to excel. It is difficult to improve performance and functioning if accurate and relevant information about performance is not available. Imagine young athletes trying to succeed in a chosen sport without an idea about how they are doing, or students trying to determine their comprehension of subject matter without getting a report card or feedback from teachers. In both cases the individuals require information about their performance on selected tasks in order to improve. The same idea holds true for organizations, including police departments. Unfortunately, the potential of effective performance measurement systems has not been realized in police departments.
Measuring the performance of the police extends back to 1930, when the Federal Bureau of Investigation’s Uniform Crime Reporting (UCR) Program was launched and the crime clearance rate was proposed as an indicator of police effectiveness. Despite relatively early beginnings and well-known limitations of commonly used performance indicators, performance measurement has not evolved a great deal in police departments. The end of the twentieth century saw performance measurement play a significant role in many organizational development and management innovations that were taking place in the public and private sectors. Interest in making police performance measurement more effective also blossomed during this time.
Measuring police performance at the organizational level can serve two important purposes: ensuring accountability and promoting organizational learning. In a democracy, the police are accountable to the public through the political process. In this way a local community has a voice in what police do and how they do it. The process of holding police departments accountable is believed to promote effective and fair operations because evidence of shortcomings is expected to increase demands for improvement and evidence of success is expected to be rewarded. The public and individuals in positions of authority who make decisions about the police, including the allocation of resources and personnel decisions, must have accurate and relevant information that can be used to assess police performance. Without this information informed decisions cannot be made, undermining the potential for accountability to ensure quality performance.
A second purpose of performance measurement is to provide information that can ultimately improve police operations through organizational learning. A learning organization has been described as one that takes advantage of experiences, including those of other organizations, to improve its functioning. The assumption is that organizations can process information, learn, and then change. Without accurate and relevant information about performance or the ability to process it, organizational learning suffers. By measuring what is important about performance and building the means of using that information effectively, a police department is in a better position to take advantage of experience and make appropriate adjustments. Reformers advocated for police departments to become learning organizations in the late 1990s; improved performance measurement is a necessary part of this process because measurement provides the necessary raw material.
In addition to accountability and organizational learning, the performance measurement process itself holds the potential to affect police behavior. A performance measurement system sends subtle messages to employees and external audiences about what the organization values. By measuring certain aspects of police work a department sends messages about what it values most. This can motivate personnel to concentrate their efforts on what is being measured and possibly overlook other aspects of their work. This idea is reflected in the phrase “what gets measured gets done.”
Problems the Police Face Today
In the general sense, measuring abstract phenomena, such as the quality of a service that has been provided and whether services have been provided in a fair manner, is challenging. This is no different for police departments; designing and utilizing an effective performance measurement system is a complicated process. First, police performance is multidimensional. In other words, police departments perform a wide range of tasks, and they may do some things better than others. In addition, police are responsible for employing just and fair procedures, meaning the way police do their jobs is an important dimension of performance.
A second complicating factor is that police have a complex mission. For indicators of performance to be useful they must be grounded in a clear understanding of what a police department does, what it is expected to do, and what it seeks to accomplish. It is critical that performance measures conform to what organizations do and what they intend to accomplish. A mismatch means the department is limited in how it can use feedback. Several problems can result from a disjunction: A police department may not learn about important dimensions of its behavior if those dimensions are not measured; when a department does not know how well it is performing important tasks or whether it is achieving intended outcomes, it is not in a position to evaluate, learn, and make appropriate changes; it is difficult to make sense of performance feedback that is only marginally related to what a department is attempting to accomplish; and a department might be judged unfairly if performance indicators focus on outcomes over which a department has little influence.
For these reasons it is difficult to identify a ”bottom line” of policing. To treat crime as the ”bottom line” of policing overlooks the fact that crime reduction is not an acceptable outcome if police employ unjust methods, and it ignores the fact that police provide many services not directly related to serious crime. Useful performance measurement schemes must include multiple indicators of what police do (that is, outputs), how it is done (that is, processes), and what is accomplished (that is, outcomes).
Police departments have traditionally relied on a limited set of indicators to monitor their performance, thus limiting their ability to learn and improve. These indicators include crime rates, crime clearances, arrests, and response times. While these measures are easily collected and inherently appealing, they are also limited in important ways. In their commonly used forms these indicators do not facilitate police accountability or organizational learning.
Crime rates are often measured with arrests, victimization surveys, and crimes reported to the police. A number of significant criticisms have been levied against the use of crime rates for the purposes of accountability and departmental learning. Crime control and prevention are only one part of what the police are expected to accomplish. A near-exclusive focus on measuring crime is problematic because other dimensions of police performance that are also centrally important, such as the use of fair and legal procedures and police responses to nonserious crime problems, are overlooked. In this case a department cannot understand how well it is meeting other responsibilities and how it can improve in these other domains.
Police have only a limited ability to affect some types of crimes. Recognizing that a host of societal factors combine to change levels of some crimes means it makes little sense to hold police accountable for any observed changes. For this reason crime rates do not offer the ideal and ”bottom line” performance indicator they might seem to represent. In addition, police departments can influence crime data through a host of means that are independent of actual levels of crime in the community. Police agencies can inadvertently discourage crime reporting by community members and more overtly manipulate crime statistics to make it appear a department is keeping crime at an acceptable level. Thus, indicators of crime may not provide an accurate sense of police performance at keeping communities safe.
The crime clearance rate is also an inherently appealing indicator of police success because it measures how often police solve crimes that come to their attention. More specifically, clearance rates are used as indicators of investigative success. The clearance rate is calculated by dividing the number of crimes that are defined as ”cleared” by the number of crimes that are officially recorded by the police. Despite their inherent appeal, crime clearance rates are also limited for monitoring police performance.
Clearance rates are not measured consistently across agencies, and their calculations can be manipulated. Thus, it is problematic to compare the performance of different departments. One department may appear to perform better than another simply because of the way clearance rates are calculated. For example, one police department may define crimes as cleared when a suspect is identified but not actually taken into custody, while another may employ strict criteria and define a crime as cleared only when a suspect is arrested. Similarly, the arrest of a single suspect can lead to several clearances if police personnel report that the suspect is responsible for other crimes.
Defining an event as a crime is easily manipulated, making clearance rates unreliable. Rather than officially define an event as a crime, personnel can categorize the event as “unfounded,” which means the incident is not officially counted as a crime. This reduces the value of the denominator in calculations of clearance rates, thus increasing the clearance rate and making it appear that a department is more effective at solving crimes. That the information used to calculate clearance rates is easily manipulated has led experts to conclude that the clearance rate is a poor indicator of how police are performing.
A third commonly used measure of police performance is the number of arrests and citations issued. These are often considered indicators of police output because they measure work that seeks to accomplish the goal of keeping communities safe. As with the use of clearance rates, comparing police departments in terms of arrest rates is inappropriate because of inconsistent definitions and localized pressures that can affect how often officers make arrests. Departments that record arrests early in the process of detaining individuals would appear to be more active than departments that record arrests later in the process, such as when an individual is booked. In this case both departments might be taking people into custody at the same rate, but one would appear to perform better only because of the definition used.
On a more abstract level, it is hard to judge whether relatively high arrest rates represent desirable or undesirable outputs. In some cases creative, nonarrest solutions may hold greater potential to achieve desired outcomes, such as increased community safety. For instance, in the 1990s many police departments were experimenting with creative partnership-based responses to the mentally ill that sought to reduce the chances of arrest. Furthermore, a near exclusive reliance on counting arrests means that the way police perform their duties takes a backseat.
Historically, police departments have dispatched officers to respond quickly when the public calls for service. Since it has become expected in many cases that police will be rapidly dispatched, it seems reasonable, on its face, to measure response time and use this information to gauge police performance. Rapid response is predicated on the belief that if police can get to the scene of a crime quickly, then the chances of arresting a suspect will increase. Research has shown that rapid police response is not responsible for appreciably increasing the chances of arresting a suspect. As with arrests, it is not clear that rapid response represents high levels of police performance. Not all calls for service, even about serious crime, demand a rapid response, and in some cases citizens are satisfied even if an officer will not be dispatched to the scene. Thus, agencies that work hard to reduce the time between call receipt and officer arrival on the scene are not necessarily making effective use of resources. Nevertheless, rapid police response is associated with citizens’ positive assessments of police services.
Perspectives at the Turn of the Century
At the turn of the century, an innovative spirit was afoot in policing, and performance measurement systems were considered important mechanisms for improving policing and advancing reforms. Police practitioners and researchers were concerned with the state of police performance measurement, with generating new methods that would reflect accurately and comprehensively on police performance, and with ways of making better use of available performance indicators to improve the quality of policing. This heightened awareness about performance measurement was partially the result of community policing and problem-oriented policing reforms. These reforms forced a reexami-nation of the police mission, police organization, and how police do their jobs. In 2004, the National Academy of Sciences’ Committee to Review Research on Police Policy and Practices concluded that measures of police performance commonly utilized were not useful for assessing how well police were filling contemporary conceptions of their mission.
Effective performance measurement has the potential to advance the practice of both community policing and problem-oriented policing because a police agency has the chance to learn about what it is doing well and how it may be failing in terms of reform efforts. For example, problem solving is promoted under both major reforms—community policing and problem-oriented policing—and represents a systematic method for officers to address significant community problems.
A performance measurement scheme must tap into the major components of problem solving for a police department to gauge how well specific elements of the practice are being implemented and whether it is achieving intended results. When a performance measurement system excludes indicators of problem solving, a police agency misses the chance to obtain feedback about how it might improve. These same ideas apply to many aspects of community policing, including police departments’ efforts to engage and collaborate with communities, to reduce citizens’ fear of crime, and to improve citizen satisfaction with police services.
A reconsideration of performance measurement systems means police departments must carefully and openly consider new sources of data that reflect on their performance and facilitate organizational learning rather than limit themselves to readily available administrative sources of data. New sources of data include employee surveys, direct observations of police-community encounters and community conditions, and information from other public and private organizations.
It is not reasonable to assume that police will abandon crime rates as an indicator of performance, but an improved measure of crime is available. Research has demonstrated the value of using crime rates that are adjusted for factors over which police have little control, such as shifting demographics and economic conditions. An adjusted measure of crime can indicate how much influence the police can have on crime. When combined with additional measures of police performance, adjusted crime rates have the chance to become an important part of a comprehensive performance measurement scheme.
Citizen surveys also represent an important source of information on performance that many police departments have been collecting. Citizen perceptions have long been recognized as an important and useful indicator of police performance. Public perceptions of the police are important because police effectiveness depends heavily on support from citizens, police are accountable to the public, and police services are frequently delivered at the point of contacts with citizens. Police can obtain valuable information from citizen surveys that is not easily obtained from other sources of data, including information about crime and community problems, citizen willingness to cooperate with police, and assessments of officer performance during specific encounters.
General community surveys seek to include individuals that represent a broad cross section of a particular area. One problem with general surveys is that a large portion of the community has not had recent contact with the police. This means that citizen perceptions of the police can be formed through mechanisms not directly related to police performance. Contact surveys help to overcome this problem by including individuals who have had some recent contact with the police, including arrestees, crime victims, and drivers. Police must pay careful attention to the way citizen surveys are constructed and implemented because important problems can result from surveys that are poorly designed and implemented.
It is important to include a representative sample in general and contact surveys. A sample of citizens that is disposed toward holding favorable (or unfavorable) perceptions of the police reduces the chances that police will learn about how the entire community and important subsections truly perceive performance. The inclusion of only certain citizens in a sample can distort results. While it may be tempting to exclude individuals who are presumed to hold generally unfavorable views of the police, these individuals can provide valuable insights a department can use.
The way survey items are worded can also affect results. Questions that are worded in vague and general terms will likely yield mostly positive results. This creates the opportunity for police departments to present themselves in a positive light rather than for purposes of gauging performance and organizational learning. These potential problems have led experts to recommend that police departments become more sophisticated in the way they collect survey data from citizens and in the ways they use this information. Very little is known about the way police departments use information gathered through citizen surveys.
Prior to selecting sources of data and collecting information, it is necessary for police departments to identify the performance dimensions they value. Police departments have traditionally lacked theoretical guidance about the dimensions they should assess. As interest in police performance measurement grew in the late 1990s, experts recommended sets of performance dimensions for departments to consider measuring. These included, for instance, the extent to which police hold offenders accountable, ensure the safety of public places, and limit fear of crime. Departments were also encouraged to measure the process dimension of police performance. The process aspect is unique because it focuses attention on measuring the manner in which police perform their duties rather than emphasizing measures of what police do and traditional outcomes. The process dimension of police work is illustrated by Stephen Mastrofski’s (1999) six aspects of police service that he terms “policing for people”: attentiveness, reliability, responsiveness, competence, manners, and fairness.
Including the manner in which police perform their duties in performance measurement systems is critical because scholars and practitioners recognize that the process of policing has important implications for the outcomes of policing. Research and theoretical work in the 1990s demonstrated that the way police treat citizens during encounters can affect citizen cooperation with the police and compliance with the law, two important outcomes for the police. In addition, fair and just treatment by the police is an important police product, independent of how such treatment affects crime and community order.
A Comparative Approach to Performance Measurement
One broad framework for rating the performance of organizations is with a comparative performance measurement system. This approach has been used across a range of service providers, including hospitals, local governments, and schools. Experts have recommended that this approach be used to evaluate police performance. Comparative performance measurement in policing would consist of measuring the performances of police departments and then comparing them over time and place. This allows the performance of a police department to be judged against the performance of others or against itself in a previous time period so that the department can learn how it is doing. Without some reference of comparison, it is difficult for individuals and organizations to understand and interpret their current level of performance.
Comparative performance measurement has typically come in the form of an organizational report card. The school report card, well known to teachers, children, and guardians, illustrates the value. A school report card communicates information about student performance to several people and can motivate students to excel through several mechanisms. This same logic can be applied to organizations; a performance grade is shared with a broad audience, and poor performance can motivate a department to search for ways to improve.
Organizational report cards differ from traditional forms of performance measurement because those traditional forms are largely self-assessments; the report card represents externally focused performance measurement. The grade an organization receives is communicated to a broad external audience, facilitating accountability. When performance indicators are not shared with external audiences, the chances for accountability are reduced, along with opportunities to improve performance.
The police would seem to be well-suited for organizational report cards because reports cards offer a unique mechanism for enhancing public accountability. The challenge for comparative performance systems is to yield appropriate comparison groups because police departments operate in distinct environments and face unique problems.
Policing scholars have advocated for national monitoring systems that would be akin to report cards. For instance, Tom Tyler (2002) recommends that police agencies conduct ongoing surveys of the citizens they serve to monitor public perceptions of police fairness and legitimacy. In 2004, the National Academy of Sciences’ Committee to Review Research on Police Policy and Practices recommended a national effort to collect data on several aspects of police work. Finally, more than half of the states in the first decade of the twenty-first century mandated that police agencies collect and report data on traffic stops. The purpose was to monitor potential racial disparities in patterns of traffic enforcement and report individual agency performance to the public.
Unresolved Issues
Despite the innovative spirit that dominated policing from the 1980s into the twenty-first century and the belief that performance measurement systems have the power to facilitate improved service delivery, police organizations have not advanced their performance measurement systems. Advancing performance measurement schemes and promoting organizational learning represent processes that will persist into the twenty-first century as long as police departments are serious about implementing community policing and realizing the reform’s full potential. The challenge lies not only in accurately measuring several important aspects of police performance with appropriate sources of data but also in determining how the information will be used to promote more effective, efficient, and fair police work.
