Compstat implementation.

The final stage of any new managerial approach involves implementing the process. Compstat, as described in the first two parts of this article, is no different; implementation presents the final challenge. (1) Just as with the design element, Compstat implementation must have the complete support of the leaders of a law enforcement agency. (2) These individuals must ensure that all of their employees understand the importance of the process and witness the high level of commitment that they bring to the entire effort.

[ILLUSTRATION OMITTED]

IMPLEMENTATION

After designing its Compstat program, an agency now must implement the process. To do this, the agency must consider a few key elements: training, the Compstat meeting protocols, the interaction (line of questioning), and the roundtable discussion.

Training

Training for Compstat (or the lack thereof) constitutes a frequent complaint of the participants. But, an agency can accomplish training by first preparing a sample Compstat book that exactly matches the one it will produce every week. The agency must announce ahead of time any subsequent changes to the book so no surprises occur for anyone held accountable for the altered portion. (3)

Next, the agency should conduct a plenary session with all required attendees, chaired by the chief and the facilitators, at the facility where it will hold future Compstat meetings, using the actual equipment. This will give participants the look and feel of the impending meetings and ease their transition once the real meetings begin. During the plenary session, the agency must describe fully the purpose, rationale, techniques, anticipated organizational change, and expectations and ensure that the participants understand all of these issues. For further development, the agency may send participants to an actual Compstat session in a practicing community to observe firsthand what they can expect. (4)

[ILLUSTRATION OMITTED]

Compstat Meeting Protocols

First, the department must identify the day and time for Compstat meetings, which must be the same each week. This is not only imperative to data collection but to consistency, as consistency breeds conformity. Because crime is dynamic and trends emerge and dissipate quickly, particularly due to the commanders' efforts, the department should hold the Compstat session at least on a weekly basis. For example, if a city has four police precincts, one precinct each week becomes the featured command; therefore, the Compstat period for each precinct occurs every 4 weeks (once per month). (5) Participants should expect to spend 2 or 3 hours at each session to cover all of the material.

Next, the department must address the seating arrangement. It should configure the Compstat room in a square (see Figure 1), with assigned seating. The chief and other executives sit at the head of the table; the commanders or designated participants sit on either side of the table; the featured command sits directly opposite the chief, facing the executive audience; the division support staff members sit behind the commanders; and any guests sit in the gallery behind the chief.

The dress code becomes the next protocol to approach. All personnel attending Compstat should appear in uniform-of-the-day attire, preferably the dress uniform. Again, this promotes consistency and sends a message to both participants and observers that Compstat is a formal process worthy of everyone's undivided attention and professional appearance.

The chief opens the Compstat session by welcoming everyone and acknowledging the guests by name (if only a few attend, otherwise by their organizations). The first order of business is accolades. Insofar as possible, commanders should bring their personnel to Compstat for the recognition of outstanding performance. The commander calls the officers before the Compstat group, provides a brief overview of the action that led to the accolade, and commends the officers. This public display of praise strengthens morale and sends the message that individual efforts produce a synergy that contributes to the whole.

The recap, prepared from the scribe's notes of the previous meeting, occurs next. The facilitator reviews the items, and the commanders in question explain what they did to abate the problem. The commanders must narrowly tailor their answers to the recap question without superfluous detail.

Following the recap, the featured command members take their places at the table. The session begins with the featured commander providing an overview of his crime posture and crime-control strategies since he last appeared at Compstat.

   The [overview] is intended to be a comprehensive and informative
recapitulation of criminal activity and police activity within the
command, showcasing what the commander is doing to identify and solve
problems .... The executives may interrupt and direct the [commander] to
focus more closely on a particular issue, or they may interrupt to focus
on a particular case. The overall process of interaction is fluid and
flexible, with few fixed rules.

   The [overview] is the commander's chance to impress executives and
other personnel present at the meeting with his ... knowledge,
leadership talents, crime-fighting abilities, and overall career
potential. This is the commander's [opportunity] to bring problems and
issues (especially those concerning the adequacy of resources and crime
patterns that cross precinct boundaries) to the attention of the
executive staff--in essence, to publicly communicate [his] needs and, in
doing so, to place some of the responsibility and some of the
accountability on the executives. (6)

The key to success for any commander rests with being prepared. This means exhibiting a willingness and commitment to knowing and understanding the data and the underlying conditions within the command, devising effective strategies and tactics, relentlessly following up on initiatives, and possessing the ability to articulate plans and conditions.

Interaction

After the overview, the facilitator asks a series of direct, probing questions concerning current investigations, quality-of-life conditions, and crime-control strategies. The questioning, at times, may become adversarial, especially if the commander failed to implement a plan of action. Commanders should maintain their focus and never lie as a means to extricate themselves from difficult questioning. The following example provides a typical line of questioning that might arise in a robbery pattern involving two suspects using a blue vehicle:

Facilitator: I see that six robberies involving the same vehicle occurred between August 3 and August 20 in your precinct. These six robberies account for a 6-percent increase over last week, a 2-percent increase from last month at this time, and a 10-percent increase over this time last year. Explain the increase to me and what you are doing about it.

Commander: Chief, my crime control officer identified the pattern 2 weeks ago and assumed that the drug trade in the area was fueling the problem. As the robberies occurred between 2200 and 0300, I increased patrols during those times. I issued a directed patrol order for the train station because four of the six robberies occurred in that vicinity. I notified the transit police officers who patrol the interior and a portion of the perimeter of the train station, and I briefed them on the details. I personally spoke to the chief of the transit police and advised him to post crime bulletins where commuters could see them. The transit chief said that he also would increase patrols during those hours and review surveillance tapes from fixed cameras outside the station. I assigned two officers to a task force assembled by the robbery division. The MO and the vehicle description are the same on each of the robberies, but I am not certain if one person committed all six robberies or if there are six separate suspects.

Facilitator: Robbery division commander, what are you doing about it?

Commander: Chief, I deployed a task force of detectives and precinct personnel who are working exclusively on this investigation. I have some officers in uniform, others acting as decoys, and a team conducting street surveillance. So far, the results have been negative. I will continue these tactics for another week; however, if the results remain the same, I will reassess the tactics. I issued a crime bulletin to all commands, here is a copy. On Tuesday, we will have the latest victim meet with the police artist to develop a sketch. The earlier victims could not identify the gunman because he wore a mask, but the last victim struggled with the suspect and managed to pull off the mask. The earlier victims said that they could identify the voice, a deep male voice. When an apprehension occurs, I will obtain a voice exemplar and have the victims listen to it. As for the mask, I had it sent to forensics for analysis.

Facilitator: Crime scene commander, what is the disposition of the mask?

Commander: Chief, hair fibers were recovered from inside the mask. It is being tested for DNA right now. Once I have the DNA analysis, I will run it through our database to search for a comparison. I should know something by the end of the day. No other evidence was recovered from any of the other crime scenes.

Facilitator: Place that on the recap: crime scene commander to provide results of DNA testing on the suspect's mask by 5 p.m. today. Obviously, robbery is the motive. The precinct commander asserted that the local drug trade is fueling the problem, and it is a high-narcotics area. Let me see the map of narcotics complaints. Narcotics division commander, tell me what you are doing about this? Is there a nexus between the drug trade in the area and the robberies?

Commander: Chief, on Monday, Tuesday, and Thursday of last week, I conducted 12 different buy-bust operations, two during the a.m. and two the during the p.m. hours, which yielded 12 arrests for sale and possession of cocaine or heroin; 15 field interviews, 5 of which resulted in arrests for outstanding warrants; 6 traffic summonses; and 4 vehicles impounded. Unfortunately, the prisoner debriefings were negative. We continue working to establish a connection to the drug trade and to ascertain the suspects' identities.

Facilitator: Gang division commander, is the vehicle description listed in the gang database? Do any vehicles of known gang members match this vehicle?

Commander: Chief, I am not sure. I will check on that and advise the robbery division by the end of the day.

Facilitator: Place that on the recap: gang division to review the database of identified gang members to see if any similar vehicles fit the description and to notify the robbery commander by 5 p.m. today. Auto crimes commander, have any blue vehicles been impounded since August 20, the date of the last incident?

Commander: Chief, three blue vehicles were impounded. I assigned a single detective to investigate all three. I also notified crime scene to print each vehicle and advise me of the results. Two of the three vehicles were stolen; the other was towed for street cleaning.

Facilitator: For the recap, the auto squad commander will present the findings of the investigation of the three blue vehicles and the crime scene commander will determine whether latent prints or other forensic evidence was recovered from any of the vehicles by 1 p.m. Monday, August 23. Robbery division commander, what is the victimology? Do the victims have a criminal history?

Commander: Chief, all six victims have prior drug arrests. Two are on probation, and I notified the county probation department of that fact on August 16.

[FIGURE 1 OMITTED]

Facilitator: Robbery commander, check with parole and DOC to determine who currently is on parole and living in the area, as well as who recently was released from prison and moved to the area.

Commander: Chief, the task force already is working on that. I should have some answers by next Wednesday, August 25.

Facilitator: For the recap: robbery commander to identify the results of parole and DOC inquiry on parolees' residences by Wednesday, August 25. Let's move on. To the commanders, I want intense monitoring of this investigation. Advise my office the minute something breaks.

The interaction during Compstat is dynamic. No standard questions exist, except for a few that the chief always will want answered.

* What was the motive? Was it robbery, jealousy, revenge, thrill, bias, dispute, domestic, debt?

* What is the victimology (i.e., a complete history of the victim, including lifestyle, personality traits, and employment)? Other important factors include the victim's age, occupation, family background, reputation, likes and dislikes, drug/alcohol use, financial troubles/stability, religious beliefs, routines/habits (e.g., checking mail or walking the dog), criminal history, connection to area or suspect, DOC history, and gang affiliation, along with name and vehicle record checks pertaining to the victim, the last known person the victim spoke to or was seen with, and the circumstances, enemies, or any known reason why someone may have wanted to harm the victim.

* What is the profile of the offender? This includes the offender's "signature" (i.e., identifiable characteristic) and other known information, such as financial troubles/stability, religious beliefs/fanaticism, reputation/propensity for violence, drug/alcohol use, known hangouts, NCIC inquiry, outstanding warrants, criminal history, likes/dislikes/obsessions/infatuations/perversion, the last known person the offender spoke to or was seen with and the circumstances, enemies, connection to area or victim, DOC history, gang affiliation, and name and vehicle record checks.

* Was the incident suppressible? (7) Could patrol or a proactive street-crime unit have prevented the incident? Could detectives have been more assertive?

* Why is performance up or down? Patrol or investigative strategies and tactics, motor vehicle checkpoints, supervision, motivated employees, morale, vacation, sick time, and personnel strength can affect performance.

* What connection does the suspect have to other crimes? Multiple victims who identify the same suspect, the suspect's MO or "signature" matches other similar unsolved crimes in the area, and the examination of forensic evidence and surveillance tapes represent three ways to determine a connection.

* What progress has been made to date? Statements taken, polygraph administered, suspects identified, warrants issued or served, composite sketch, and assets seized can measure progress.

* What is the deployment and strength level? The number of sector cars, overlap/umbrella cars, walking posts, overtime detail, special units, and uniformed and plainclothes personnel can provide data.

* Is there a nexus to gangs, drugs, or organized crime? Drug rip-off, drug kingpin, gang leader/member, member of crime family or criminal enterprise, and potential for vertical prosecution (e.g., RICO) or an enhanced prosecution/sentence can show a link.

* Does recovered forensic evidence have a connection to other crimes? Comparing samples, such as DNA, trace evidence, bullets or shell casings, pry or tool marks, latent prints, impressions (tires and shoes), written documents, audio and video tapes, liquids, paint chips, and shards of glass, along with examining computer hard drives and Internet history can reveal a connection.

* What is the plan of action or what are the next steps? Develop a task force; serve warrants; raze buildings; tow derelict autos; padlock notorious businesses; issue summonses/motor vehicle enforcement; seize assets; conduct inspections of buses, taverns, bodegas, ATMs, convenience stores, gas stations, and taxi cabs; present case to a grand jury; seek civil enforcement (nuisance abatement); increase patrols; and initiate overtime constitute some next steps.

Above all, the chief does not want a recitation of the incident report. On occasion, a summary may prove useful, and, if so, the chief will ask for it. Otherwise, reciting the incident report amounts to a superfluous detail and makes commanders appear to be temporizing because they are unprepared. Commanders must expect a variety of questions unique to each investigation.

[ILLUSTRATION OMITTED]

Once the crime presentation ends, the performance presentation begins, during which the featured commander must answer for how his command compares with others. As with any other portion of Compstat, the commander must articulate why performance has increased or decreased, what action plans he has created, and where he expects to adjust strategies if necessary.

Roundtable Discussion

During the presentations, other commanders should not interject issues unrelated to the discussion, causing the facilitator to engage in boundering. (8) Instead, they should make notes and save their comments for the last portion of Compstat, the roundtable. At this time, the chief polls all of the other commanders and asks if they have anything to discuss. Commanders can debate training issues, announce other city or department initiatives, review budget issues or procurement problems, and handle other similar items. When the roundtable discussion ends, the chief thanks everyone for attending and dismisses them.

ADAPTABILITY

Adaptability stands as one of the distinctions about Compstat. It is easily adaptable to subdivisions of the organization, such as internal affairs (IA), or to other segments of government. When organized properly, IA Compstat can reduce personnel complaints while lessening corruption and increasing integrity. This promotes a much higher degree of overall organizational discipline, and "a well-disciplined work force is in voluntary compliance with the rules and regulations of the organization and works efficiently to attain the goals of the organization." (9)

Purchasing and procurement can provide another element subject to Compstat. Scrutinizing the purchasing process will ensure that the agency acquires much-needed equipment and materiel as expeditiously as possible. Compstat can identify unscrupulous vendors and practices, as well as uncover contractual problems, system delays, and funding obstacles. Because purchasing often involves other elements of local government, participants may include the city manager/business administrator, the budget director, and the purchasing agent.

How the Philadelphia Police Department uses the process to focus on specialty units can illustrate Compstat's adaptability. Because of its size and decentralized command structure, the department holds Compstat meetings every 4 weeks that focus exclusively on its specialized units, including SWAT, canine, mounted, aviation, bomb disposal, environmental response, marine, and accident investigation. At these meetings, participants identify and discuss performance measures, such as the number of cases involving barricaded persons handled by SWAT personnel, the number of vehicle pursuits engaged in by aviation officers, and the number of code enforcement violations issued by environmental response officers. (10) As another example, Baltimore uses CitiStat, a variant of Compstat, to monitor all of the city's operations. (11)

[ILLUSTRATION OMITTED]

CONCLUSION

Law enforcement agencies can do well embracing the Compstat process. "By adopting a flexible, accountability-driven law enforcement structure, cities that have made little progress to date can achieve reductions on par with the most dramatic declines in urban crime during the last decade, while those cities that already experienced success can continue to force crime down to ever lower levels." (12) Crime rates among the cities practicing Compstat reveal the program's true success. In New York City over the last 10 years, crime came down 64 percent; in Philadelphia, crime fell 23 percent between 1995 and 2002; in Baltimore, crime decreased 31 percent between 1995 and 1999; and in Newark, crime declined 51 percent between 1995 and 2001. (13)

Law enforcement agencies need invest only a negligible amount of money to implement Compstat. The key is for law enforcement agencies to structure for success. "Creating that structure requires extensive central data collection and analysis and constant feedback and review of the effectiveness of police programs. Perhaps, most important of all, a culture of accountability must be instituted within the structure. At every level, from the whole city to a single street, the law enforcement personnel entrusted with preventing crime must take responsibility for [mistakes] and be recognized for success." (14)

Compstat--a transparent accountability system that objectively measures performance and holds those responsible open to scrutiny--offers more than police rhetoric. It offers favorable, achievable results for large or small law enforcement agencies. (15)

The author thanks his friend and colleague Chief Anthony F. Ambrose of the Newark, New Jersey, Police Department for his inspiration and insight concerning this article.

Endnotes

(1) Jon Shane, "Compstat Process," FBI Law Enforcement Bulletin, April 2004, 12-21; and "Compstat Design," FBI Law Enforcement Bulletin, May 2004, 12-19.

(2) For illustrative purposes and to maintain clarity, the author refers to the leaders of law enforcement organizations as chiefs and employs masculine pronouns for these individuals, as well as other command-level personnel, throughout the article as needed.

(3) To better manage spontaneous events, the Newark Police Department established a "24-hour rule," which states that a commander is not expected to answer for the 24-hour period immediately preceding a Compstat session. A shooting, for example, that occurred overnight will not be part of the regular Compstat meeting; however, the commander must be able to answer basic questions about the incident, but not the usual intimate details. Inquiries about an event within the last 24 hours constitute "breaking the 24-hour rule." If the chief breaks the 24-hour rule, it usually occurs after the recap items. The rule exists so commanders are not surprised by questions that arise from a spontaneous event and because they have not had sufficient time before Compstat to explore the details.

(4) Some agencies practicing Compstat include Newark, NJ; Philadelphia, PA; Boston, MA; Indianapolis, IN; Baltimore, MD; New Orleans, LA; Broward County, FL; Washington, DC; Austin, TX; Seattle, WA; Mount Vernon, NY; Durham, NC; Lowell, MA; Longmont, CO; Maryland State Police; Los Angeles, CA; Minneapolis, MN; Chicago, IL; San Diego, CA; and, of course, New York, NY; see P.P. McDonald, Managing Police Operations: Implementing the New York Crime Control Model--Compstat (Belmont, CA: Wadsworth, 2002), 26.

(5) Generally, an overlap of data analysis occurs with specialty commanders (e.g., robbery, narcotics, homicide, and auto crimes) who participate each week. The 2-week interval gives these commanders a better picture of trends and patterns than the normal 1 week of data.

(6) V.E. Henry, The Compstat Paradigm: Management Accountability in Policing, Business and the Public Sector (New York, NY: Looseleaf Law Publications, 2002), 227. See also http://www.nyc.gov/html/nypd/html/chfdept/strategies.html.

(7) An incident is deemed suppressible if 1) it occurred on view, that is, the location was visible from the street or 2) it occurred as part of a pattern of recidivist behavior by an identified suspect. For example, there is an active domestic violence investigation pending by the police department, and the suspect has been identified. If the detectives fail to make an arrest or take some other enforcement action (e.g., serve a restraining order) and the suspect commits another domestic violence incident, the second incident is deemed suppressible.

(8) Boundering is a questioning technique used by an interviewer to prevent an interviewee from straying from the subject matter being discussed. See Donald Schroeder, Frank Lombardo, and Jerry Strollo, Management and Supervision of Police Personnel (Binghampton, NY: Gould Publications, 1995), 133.

(9) Ibid.

(10) Philadelphia Police Department, The Philadelphia Compstat Meetings (Philadelphia, PA, 2003); retrieved on May 6, 2003, from http://www.ppdonline.org/ppd_compstat.htm.

(11) R.M. Linden, Working Across Boundaries: Making Collaboration Work in Government and Nonprofit Organizations (San Francisco, CA: Jossey-Bass/John Wiley and Sons, 2002); retrieved on June 1, 2003, from http://www.josseybass.com/WileyCDA/Section/id-10980.html. For more information, see http://www.baltimorecity.gov/news/citistat/.

(12) G.L. Kelling and Ronald Corbett, The Center for Civic Innovation at the Manhattan Institute, Civic Bulletin 32, "This Works: Preventing and Reducing Crime" (New York, NY, March 2003).

(13) New York City crime figures from the New York City Police Department's CompStat Unit for the week of April 21 through April 27, 2003; retrieved on May 6, 2003, from http://www.nyc.gov/html/nypd/pdf/chfdept/cscity.pdf; and Philadelphia, Baltimore, and Newark crime figures retrieved from U.S. Department of Justice, Federal Bureau of Investigation, Crime in the United States, for the years 1995 through 2001.

(14) Supra note 12.

(15) "Case studies conducted in six New York City police precincts in 2000 show that precinct commanders use Compstat technology to identify when specific types of crime, such as robbery and burglary, become unusually serious problems. Incidences of such crimes fell after the commanders employed specifically devised tactics to combat the identified problem"; see G.L. Kelling and William H. Sousa, Jr., The Center for Civic Innovation at the Manhattan Institute, Civic Report 22, "Do Police Matter? An Analysis of the Impact of New York City's Police Reforms" (New York, NY, 2001), 1-2; retrieved on May 4, 2003, from http://www.manhattan-institute.org/html/cr_22.htm.

Captain Shane is the commanding officer of the Policy and Planning Division of the Newark, New Jersey, Police Department.