It has become a truism among the police that 10% of their officers cause 90% of their disciplinary problems. Investigative journalists have found police departments where as few as 2% of all officers are responsible for at least half of all citizen complaints. As early as the 1970s, Herman Goldstein observed that problem officers were well known to all, but that all too little was often done to alter their conduct. In 1981, the U.S. Commission of Civil Rights recommended that all police departments implement an early warning system to identify those officers generating frequent citizen complaints or demonstrating identifiable patterns of inappropriate behavior. Still, what are early warning systems and do they work?

An early warning system is a data-based police management tool designed to identify officers whose behavior is problematic. Once problem behaviors are identified, some form of intervention is offered to correct the performance. An early response means the police department is able to intervene before the officer’s behavior damages the department’s efforts or requires formal disciplinary action. The idea is that by alerting the department and assisting the officer, the employee, his or her agency, and the citizens they both serve can all benefit.

While little is known about the use of early warning systems, a 1999 survey by the Police Executive Research Forum (PERF) found that about one-fourth of police and sheriff’s departments serving populations of 50,000 or more had an early warning system in place; half of those had been created since 1994 while more than one-third of departments had introduced theirs after 1996. Another 12% of agencies had planned to introduce a system within a few years of the survey date. Municipal agencies were more likely than county police or sheriff’s departments to have such systems.

Basically, most early warning systems have three distinct phases: selection, intervention, and postintervention monitoring. First, of course, officers have to be selected for the program. Although no standards exist as to how this should be done, there is general agreement about the criteria that should be used. Among these criteria are citizen complaints, discharges of firearms, uses of force by police, involvement in civil litigation, a disproportionate number of resisting-arrest incidents, involvement in high-speed pursuits, and the occurrence of incidents causing vehicular damage.

Although a few police departments rely on citizen complaints alone to identify problem officers, most use a combination of factors such as those just noted instead. The PERF survey showed that among those agencies relying on citizen complaints as at least a factor in their early warning system, most (67%) require three complaints within a specified time frame (overwhelmingly twelve months) for an officer’s inclusion.

Once identified, the primary goal of an early warning system is to intervene with the officer to correct his or her problem behavior(s). Usually this means some combination of deterrence and education. Simply put, it is assumed that the process of being identified and included will induce officers to change their behavior so as to avoid some anticipated punishment. General deterrence assumes that even those officers not selected will alter their behaviors to remain outside of the system’s concern. In addition, any training that follows has the potential to correct officers’ mistakes and help them improve their performance.

For nearly two-thirds of the police departments in the PERF survey, the initial response consisted of a review by the selected officer’s immediate supervisor. Command officers participated at least in counseling selected officers in nearly half (45%) of the surveyed departments, while an equal percentage (45%) made training classes available for groups of officers identified.

Once they have intervened, nearly all (90%) of the agencies in the PERF survey reported that they monitored the selected officer’s conduct so as to determine if the effort had been successful. Although the officer’s immediate supervisor often informally did the monitoring, some departments have instituted formal methods of continued observation, evaluation, and reporting of the officer’s actions. Nearly half (47%) continued to monitor the officer’s performance for at least three years after the initial intervention. The remaining half either did not specify the length of follow-up or noted that monitoring decisions were made on a case-by-case basis.

While there are no experimental evaluations of the impact of early warning systems, case studies of the systems in place in Minneapolis, New Orleans, and Miami-Dade County are instructive. Overall, the systems in place in these cities appear to have been successful at reducing citizen complaints against officers as well as other indicators of problematic police performance. In Minneapolis, for example, complaints against selected officers dropped by 67% within one year of the initial intervention. In New Orleans, the reduction was 64%, while in Miami-Dade the numbers of selected officers having no uses of force went from only 4% to at least 50% following the early warning intervention.

Further, from the New Orleans data it appears that the officers themselves generally had a positive mind-set toward the early warning intervention. In that city, selected officers participated in a Professional Performance Enhancement Program (PPEP) class designed to address the concerns that had led to their selection by the system. In anonymous evaluations of those classes, participating officers gave them an average rating of 7 on a 1 to 10 scale. All officers had at least one positive thing to say about the class experience, and some added specific comments about how the class had helped them. Observations of the PPEP classes found officers most engaged in those components most directly related to the practical problems of policing—especially those incidents that often generate citizen complaints— and least engaged in the more abstract or moralistic components.

Although the use of early warning systems appears to offer some potential benefits, each of the system’s three phases involves a number of complex policy issues. For example, while the criteria involved in officer selection can vary, most systems appear to rely significantly on citizen complaints. The problems related to official data on citizen complaints—including underreporting—have long been documented. As such, a broad range of indicators of problem behaviors is more likely to effectively and accurately identify officers requiring department intervention.

The intervention also can be problematic. In many systems the initial intervention relies on informal counseling between the selected officer and his or her immediate supervisor. Some of these systems required no documentation of the content of that counseling, raising concerns about whether the supervisors know, and delivered, the desired content. Some supervisors may choose to minimize the importance of the intervention by telling the officer not to worry. If so, the selected officer’s behavior may be reinforced rather than corrected. Clearly, further research on the most effective forms of intervention is necessary.

Finally, the types and methods of post-intervention monitoring vary widely. Since the department’s follow-up may be critical to the success of the system and how it is perceived both inside and outside of the department, more research here is important as well.

An effective early warning system is likely to involve a complex, high-maintenance effort that requires a significant investment of administrative and supervisory resources. Unfortunately, some systems appear to be little more than symbolic efforts with little substantive content. It is unlikely that these systems can be effective, especially if the agency otherwise lacks a serious commitment to accountability. When taken seriously, however, the limited information available suggests that early warning systems can become effective management tools. We should remember, however, that an early warning system is but one of many tools needed to raise standards of performance and improve the quality of policing.

Next post:

Previous post: