Recommended Best Practices
Recommendation 7.8 Use early intervention systems to track officer behavior and address needs and deficiencies at the earliest opportunity.
To hold officers accountable, departments should thoroughly and impartially investigate misconduct allegations; identify problem behaviors and poor performance; and mete out consequences. As at any workplace, departments also need nondisciplinary systems to track officer performance objectively over time and to identify potentially problematic behaviors as early as possible.
Officers may fail to meet performance expectations for a variety of reasons, such as insufficient knowledge of the issue at hand (e.g., the nuances of a newly revised policy), deterioration of skills (e.g., insufficient tactical or de-escalation training), or personal stressors (e.g., substance dependency, family conflicts, or insufficient sleep). Supervisors should therefore identify and respond to possible problems at the earliest opportunity to help officers meet professional expectations, develop professionally, and avoid more serious misconduct. Such approaches (which are sometimes multi-tiered) may include referral to an employee assistance program, training, mentoring, and/or other professional growth programs. Specifically, departments should:
Implement early intervention systems. Implementing and maintaining early intervention systems improves supervision, especially at mid-sized (those with 50-999 officers) and larger (those with 1,000+ officers) departments.[i]Indeed, the Commission on Accreditation for Law Enforcement Agencies, the primary law enforcement credentialing authority in the country, has incorporated comprehensive EISs into their standards.[ii]New Jersey’s attorney general, meanwhile, has mandated that all law enforcement agencies implement an “early warning system.”[iii]
EISs identify and respond proactively to behaviors and performance trends that — while not rising to the level of legal or policy violations — nonetheless indicate that officers are not performing at optimal levels. These systems analyze a variety of indicators to identify misconduct and performance problems, such as officer-community relations (e.g., the number of complaints officers receive in a given period) and racial profiling (e.g., demographic data for traffic stops).[iv](For more detail, see Chapter 2.) Departments that carefully and consistently implement EISs have reduced the incidence of misconduct.[v]
That said, EISs are nota substitute for disciplinary systems. Officers should still be held accountable for complying with legal, policy, and performance standards. EISs strive to correct behavior before it leads to misconduct, but they don’t immunize officers from consequences for misconduct that has already occurred. Departments should ensure all officers understand and accept the goals of EISs, even in the face of uncertainty and suspicion.
Because EISs allow for some flexibility in implementation, community and officer input is vital. To inform the development of an EIS, some departments, such as the Austin Police Department, have formed a committee of community members and other stakeholders to identify factors that indicate problematic behavior and to discuss productive interventions.[vi]This is because EISs developed with community and officer input will likely be met with less resistance.
Ensure supervision. EISs are a supplement to, not a replacement for, close day-to-day supervision. Even effective, well-intentioned supervisors inadvertently overlook warning signs about employee performance or miss patterns that only become apparent over time or with the help of data collection.
EISs address this problem. First, they are a repository for information about conduct of interest to officers, departments, and communities (e.g., uses of force, disciplinary actions, complaints by community members, stops, arrests, lawsuits, vehicle and foot pursuits, workplace injuries, etc.) and other data departments are willing and able to track. Second, most EISs have a mechanism that identifies officers who reach pre-defined thresholds for potentially problematic behavior, such as a certain number of uses of force or public complaints over a defined period. Often, departments develop thresholds based on models that identify officers who are statistical outliers in given areas.[vii]More sophisticated systems compare officer conduct to that of colleagues with similar assignments and hours.
Develop effective review processes. EISs that identify officers as statistical outliers (as compared to other officers) or who have had a set number of infractions in a given period, trigger formal review processes. In some departments, the first-level review is conducted by a unit of specialists who administer the EIS. In others, this review is conducted by a supervisor in the officer’s chain of command.
Under both models, the first-level review takes a fresh, retrospective view of officer performance. This includes examining the incident reports that prompted the EIS review as well as recent performance evaluations, supervisory feedback, and, often, relevant body-worn camera or other video footage. The goal is to review materials to identify patterns of potentially problematic behavior, indicators of stress, training needs, and the like. This review should include at least one meeting with the officers in question to discuss the review, address frustration with and/or misconceptions about the EIS, and listen to what officers have to say about underlying incidents or other issues they wish to discuss.
This first-level review typically leads to a proposed remedial, nondisciplinary intervention, which may include a referral to an employee assistance program, increased supervision (e.g., supervisor ride-alongs), counseling, training, and coaching. Typically, an EIS panel, committee, or other officials experienced in EISs conduct the first-level review and propose interventions to ensure they are consistent with prior interventions for the same type of misconduct and are relevant, fair, and adequate.
In some instances, only minor interventions are proposed, such as increasing coaching and counseling. Review of video footage may, for example, identify officers who take unnecessary risks when stopping motorists. Even when they don’t lead to interventions, EISs benefit officers, departments, and communities because they foster communication between officers and supervisors and provide valuable information about officer conduct. These interventions should be viewed as learning opportunities.
Create sufficient data storage. Because EISs are driven by data, special efforts are needed to protect data integrity and ensure that data warehouse(s) are capable of responding to sophisticated queries whenever necessary. Several EIS software programs are commercially available, but software is no substitute for procedures and business rules that ensure that data are entered correctly and on a timely basis. Officers seeking to implement EISs should look to — and learn from — the many other departments that use and benefit from them.
Train supervisors to use early intervention systems. Supervisors should be trained to use EIS software, to examine past performance impartially, and to provide corrective supports in a manner that encourages officers to correct problematic behavior. EISs often prove to be valuable training tools for instructors as well. Well-built systems have officials who are trained to mine the data to find out, for example, how many foot pursuits result in the use of force, whether certain tools, tactics, or techniques are ineffective, and so on. Data-rich EISs also make it easier to identify how successfully — or unsuccessfully — training instructors prepare officers for duty.
Supervisors should track interventions, along with remedial steps or recommendations regarding officer conduct, in an electronic database system. Many systems also enable supervisors to track officer progress and hold supervisors accountable if they fail to follow through.
Implement simplified EISs in small departments. Smaller departments may not have the resources to implement an electronic EIS, but they can still institute processes to track officer performance and spot red flags. Because smaller departments have fewer officers and are often in less densely populated areas, leaders likely have fewer interactions to track and thus may be able to develop data systems with Excel or other widely available software.[viii]
[i]Int’l Ass’n of Chiefs of Police, Midsize Agencies Division, https://www.theiacp.org/working-group/division/midsize-agencies-division (The IACP defines mid-size agencies as those comprised of 50 to 999 sworn officers. Large size agencies would be those with 1000 or more sworn officers).
[ii]Samuel Walker Ph.D., Early Intervention Systems for Law Enforcement Agencies: A Planning and Management Guide, U.S. Dep’t of Justice, Community Oriented Policing Services, 8 (2003), (citing Comm’n on Accreditation for Law Enforcement Agencies, CALEA Standards for Law Enforcement Agencies Standard 35.1.5 (2001) (“A comprehensive Personnel Early Warning System is an essential component of good discipline in a well-managed law enforcement agency. The early identification of potential problem employees and a menu of remedial actions can increase agency accountability and offer employees a better opportunity to meet the agency’s values and mission statement.”)), https://ric-zai-inc.com/Publications/cops-w0085-pub.pdf.
[iii]Gurbir S. Grewal, State of N.J. Office of the Att’y Gen., Attorney General Law Enforcement Directive 2018-3 Statewide Mandatory Early Warning Systems (Mar. 20, 2018) https://www.nj.gov/lps/dcj/agguide/directives/ag-directive-2018-3.pdf.
[iv]Samuel Walker, Ph.D et al., U.S. Dep’t of Justice, Nat’l Inst. Of Justice, Early Warning Systems for Police: Responding to the Problem Police Officer, Research in Brief 7 (July 2001), http://www.ncjrs.gov/pdffiles1/nij/188565.pdf.
[v]SeeSamuel Walker, Ph.D et al., U.S. Dep’t of Justice, Nat’l Inst. Of Justice, Early Warning Systems for Police: Responding to the Problem Police Officer, Research in Brief 2 (July 2001), http://www.ncjrs.gov/pdffiles1/nij/188565.pdf(finding that a study of three departments showed significant reductions in problem behaviors and encouraged changes for both officers and supervisors); See, e.g., Thomas D. Bazley, et al., Early Intervention Program Criteria: Evaluating Officer Use of Force, 26 Just. Q. 107 (March 2009) (suggesting improvements to EIS implementation by demonstrating through a small study that merely measuring the number of times an officer uses force as compared to their peers is insufficient to identify all the officers that need to be enrolled in an early intervention program. Departments should also take into account whether the force applied equals the resistance encountered, thus identifying and enrolling those officers who use excessive force, albeit infrequently, into early intervention programs.), https://www.tandfonline.com/doi/pdf/10.1080/07418820801989742; Samuel Carton et al., Identifying Police Officers at Risk of Adverse Events, 22nd ACM SIGKDD Conference on Knowledge Discovery and Data Mining (2016), https://www.kdd.org/kdd2016/papers/files/adf0832-cartonAemb.pdf(suggesting ways to improve EIS systems, such as assigning risk scores to police officers and subsequently dispatching officers to neighborhoods and incidents with those scores as a filter. For example, not dispatching a “hot” officer to a high intensity incident in a high risk neighborhood.); But see,Robert E. Worden, et al., Intervention with Problem Officers: An Outcome Evaluation of an EIS Intervention, 40 Crim. Just. & Behav. 409, 415 (2013), https://www.researchgate.net/profile/Robert_Worden/publication/258129275_Intervention_With_Problem_Officers_An_Outcome_Evaluation_of_an_EIS_Intervention/links/55133d650cf23203199baca6.pdf(reporting that rates of complaints declined following implementation, but expressing skepticism and finding that evidence is neutral and does not support the emphasis being placed on adoption of these systems).
[vi]Samuel Walker Ph.D, et al., Supervision and Intervention within Early Intervention Systems: A Guide for Law Enforcement Chief Executives, U.S. Dep’t of Justice, Community Oriented Policing Services 32 (Dec. 2005), https://www.policeforum.org/assets/docs/Free_Online_Documents/Early_Intervention_Systems/supervision%20and%20intervention%20within%20early%20intervention%20systems%202005.pdf.
[vii]Robert E. Worden, et al., Intervention with Problem Officers: An Outcome Evaluation of an EIS Intervention, 40 Crim. Just. & Behav. 410- 411 (2013), https://www.researchgate.net/profile/Robert_Worden/publication/258129275_Intervention_With_Problem_Officers_An_Outcome_Evaluation_of_an_EIS_Intervention/links/55133d650cf23203199baca6.pdf;seeFrank Hughes, Lisa B. Andre, Int’l Ass’n of Chiefs of Police, Police Chief Magazine, Problem Officer Variables and Early-Warning Systems, http://www.policechiefmagazine.org/problem-officer-variables-and-early-warning-systems/ (“approximately 10 percent of police officers can cause, or have caused, 90 percent of the problems in law enforcement agencies.”).
[viii]See Samuel Walker Ph.D., et, al., Supervision and Intervention within Early Intervention Systems: A Guide for Law Enforcement Chief Executives, Community Oriented Policing Services,31 (Dec. 2005), https://www.policeforum.org/assets/docs/Free_Online_Documents/Early_Intervention_Systems/supervision%20and%20intervention%20within%20early%20intervention%20systems%202005.pdf (“[S]maller systems cannot do as much, but they are less expensive and easier to create. Some departments still use paper files or simple Excel spreadsheets. A police department must determine the size, capacity, and purpose of the system at the outset.”);Samuel Walker Ph.D., Early Intervention Systems for Law Enforcement Agencies: A Planning and Management Guide, U.S. Dep’t of Justice, Community Oriented Policing Services, 21 (2003), https://ric-zai-inc.com/Publications/cops-w0085-pub.pdf (“Medium-sized and small agencies do not need as complex and sophisticated a system as do large departments.”).