Public Safety Citizens Advisory Committee Final Report
Recommendations Concerning the Level of Staffing for Patrol in the Takoma Park Police Department
Public Safety Citizens Advisory Committee March 2003
Contents Executive Summary……………………………………………………… 1 Introduction………………………………………………………………. 3 Policing in Takoma Park………………………………………………… 6 Methodology……………………………………………………………... 7 Discussion……………………………………………………………….. 8 Correlation between Population and Police Strength……….. 8 Language Spoken at Home……………………………………. 9 Families with Children under 18………………………………. 10 Population and Crime………………………………………….. 11 Workload………………………………………………………… 12 Calls for Service…………………………………………………. 15 Time Worked and Available for Calls for Service……………. 17 Officers Available……………………………………………….. 22 Response Time…………………………………………………… 23 Findings and Recommendations……………………………………….. 23 Sources……………………………………………………………………. 25 Appendices……………………………………………………………….. 26
Executive Summary As part of the mandate to revisit the implementation of community policing in the City of Takoma Park, one of the crucial elements is the determination of whether current patrol strength of the Takoma Park Police Department is adequate to perform the basic functions of answering calls for service, maintaining visibility, and conducting preliminary1 investigations of crimes and offenses. In order to conduct this assessment, the Records Management System (RMS) was queried, questions asked of management officials, informal conversations held with patrol officers, supervisors and staff. As a result of this analysis, it is recommended that a minimum of 5.28 officers be added to the patrol function. This number is based on backing patrol sergeants out of the equation, with sergeants ultimately performing purely as sector commanders/supervisors (the city is divided into three sectors). There is a second recommendation which is based on also backing corporals out of the patrol equation and having them function as patrol supervisors. Should this occur, the number of officers to be added to the patrol function should be 14.88. The methodology explaining these conclusions can be found in the body of the report.2 It is also recommended that increases in personnel be further analyzed based on an assessment of: • which calls should be considered for other than a mobile response; • supervisory and management styles; • proposed tasks for all personnel under community policing (see Appendix C); • the detective function; • the training function; • changes to the department mandated by accreditation; and • personnel assigned to support functions. It is hoped that the information contained herein will assist the department in meeting its mandate of providing the highest quality police service to the residents, businesses, and organizations of Takoma Park. There is a caveat, however, any analysis and report must be considered fluid due to the increased functions placed on police departments across this nation in this era of increased terrorism and the need for homeland security. As a final note, this study would not have been accomplished with the complete cooperation of the members and staff of the Takoma Park Police Department who provided total access to records, data, and personnel. Thanks are also extended to the staff of the City Manager’s office that provided census data. 1
Investigations of crimes or offenses are initially conducted by patrol officers until it is determined that the services of detectives are needed or required.
1
Andrew Kelemen Chair, Public Safety Citizens Advisory Committee
Gayle Fisher-Stewart Vice Chair, Public Safety Citizens Advisory Committee
2
INTRODUCTION Determining the number of sworn personnel, particularly those There is no one allocated to the patrol function is not a simple task. Police best or simple strength is an imprecise term. Various researchers have tended to operationalize it in a number of ways, the three most way to common being the number of sworn officers, the number of determine the police employees, and the amount of police expenditures level of police (Maguire, 2001). Those using police expenditures justify it on the basis that policing is a personnel intensive industry, and personnel a there is an almost perfect correlation between the number of department police personnel that a jurisdiction employs and the amount it should have. spends on police protection. A handful of researchers have also discussed the difference between absolute and relative police strength (Chamblin and Langworthy, 1996; Slovak, 1986). Absolute strength is the raw number of officers or employees or amount of expenditures in a jurisdiction, while relative strength expresses these variables as a ratio (usually per capita or per unit area). The available resources for local police departments (municipal police, county sheriff) with 100 or more officers varies considerably from 1 to 112 officers per 10,000 residents, with the national average being approximately 26.7 per 10,000 residents. Still, there have been other attempts to understand the determination of police strength that are theory based (Eck and Maguire, 2000). Studies of police strength using theory are typically based (at least implicitly) on one or more of the following three theoretical foundations: (1) rational public choice (or consensus) theory, (2) conflict theory, and/or (3) some form of organizational theory (Nalla, Lynch, and Leiber, 1977). Rational public choice (also known as consensus or economic) theory implies that police organizations grow in response to citizens’ consensual requests for increased public service and protection, often in the face of rising crime rates. This theory implies that local governments and police administrators dole out resources systematically. As stated above, for decades, communities have relied on various methods for determining the appropriate amount of police resources at given times and places. These methods range from informal rules of thumb, such as the need to have at least two officers per thousand residents, to more complex mathematical models implemented in computer software packages (Bayley, 1994; Chaiken, 1975; Larson, 1978; O’Boyle, 1990, Stenzel, 1993). The kinds of variables that have typically been included in these formal and informal systems are the usual suspects: crime rates, population, calls-for-service, and other correlates of police workload. These methods reflect an implicit theory of administrative rationality, suggesting that police strength is a function of a few simple workload variables. As Loftin and McDowall (1982:400) concluded, rational choice or economic models are “too simple to account for the relationship between crime and police strength.” They suggest that models of police strength need to account for other factors in the social and political environment of police organizations. 3
If police strength is not to be based on a rational Regardless of the adjustment to variations in crime rates, then what other method used to factors are important? Conflict theory posits that racial determine police and economic inequality lead the powerful members (elites) in a community to exert political influence over strength, it remains a social control institutions. Such conflict processes would management lead to increases in social control over the powerless. decision. Statistical Tests of conflict theory have concentrated on racial and/or analysis provides economic sources of conflict. These tests are usually done information on which by examining whether variations or changes in the size of to base that decision. “threatening” populations produce differences in police strength. Research on social threat or conflict theory is also overwhelmingly mixed, although there is evidence to support both racial conflict (e.g. Jackson and Carroll, 1981; Liska et al, 1981; Snipes, 1993) and economic or class conflict explanations (e.g., Jacobs, 1978, 1979). Finally, organizational theories examine the effect of processes within police organizations in producing increases or decreases in police strength. This class of theories is the least developed of those discussed. Typically, the only explanation in this line of research is organizational inertia: that changes in police strength are incremental and that the best predictor of police strength in a given year is its value in the prior year (known as a “lag”). Therefore, unlike other perspectives, organizational explanations assume that changes in police strength are based on conditions internal to the organization. According to Nalla and his colleagues (1997:120): “this theoretical perspective assumes that organizational strength, as measured by annual budgets is explained by incremental specification models whereby the present year’s budget is influenced by appropriations in the previous year.” Numerous studies have found evidence to support the organizational inertia explanation (Nalla, 1992; Nalla et al, 1997), however, as with the other approaches, there are also shortcomings to this approach. The first is that it has little serious use in that the lag value of any variable in a time series is nearly always the best predictor of the current value and, second, it is useful for understanding stability in police strength, but it is not useful for understanding growth and decline over time. Overall, organizational explanations in this literature tend not to be well defined, treating police organizations as a “black box.” Even with its problems, the methods used in examining the determinants of police strength continue to grow more sophisticated, with recent refinements suggesting some excellent reasons for mixed findings obtained in the past (Brandl, Chamlin, and Frank, 1995; Snipes, 1993). The relationship between police strength and crime rates is known as a “simultaneous” or “reciprocal” causal relationship because each one is known to cause the other. A variety of specialized methods have been devised by researchers to 4
disentangle simultaneous causal effects, however these methods have been (and remain) the domain of economists. In fact, the majority of research in this arena has been conducted by economists and the result has been that research in this area is not widely known. This lack of acceptance is due to the fact that many of the articles appear in economic journals and the methods used are probably too difficult for criminologists and criminal justice practitioners to understand and use without advanced training in econometrics. In the final analysis, unfortunately, there is no reliable method to determine the resources that a police department needs, because there is no precise and complete listing of the tasks and activities in which police should engage or how long it takes to effectively perform those tasks and activities. However, police In the final managers need to be able to justify increases and/or analysis, police decreases in personnel strength.
strength is a function of informed “guesstimates.”
For that reason, there are three generally accepted methods used by police departments to determine the resources needed. The intuitive approach is essentially an educated guess based on the experience and judgment of police managers. The workload approach requires an elaborate information system, standards of expected performance, well-defined community expectations, and prioritization of police activities. To determine workload, the department must be able to identify precisely what the community expects the police to do, which calls must be processed, which calls can be ignored or given a minimum response, and an expected level of service. Although the workload approach is rarely used for an entire department, it is sometimes used internally to determine the needed personnel for functions (e.g. patrol) or programs (e.g. crime prevention). Perhaps the most frequently used method, alluded to This study uses a earlier, is the comparative approach based on the ratio of police officers per 1,000 or 10,000 residents (see appendix modified workload A). However, this method is not very reliable because analysis. departments tend to be diverse with respect to community problems and public expectations, managerial effectiveness, use of technology, policing styles, and the productivity of officers (see appendix B for a sample analysis). There is an added caveat to this method, there are no data using this method for police departments with fewer than 100 officers. Major studies have not been conducted to analyze departments with fewer than 100 officers. Roberg and Kuykendall (1997) provide that they know of no evidence to suggest that even a ratio within extreme parameters (11 to 112 or 13 to 70 officers per 10,000 residents) makes a measurable difference in the crime rate, citizen fear of crime, or the attitudes of citizens. More than likely, at some point, increasing police resources would result in the police becoming more effective; however, this point will vary among 5
communities based on conditions that contribute to, or detract from, the frequency and type of police problems and how effectively the police are managed. POLICING IN TAKOMA PARK In order to determine the number of police officers for the City of Takoma Park, this report focuses on the number of police officers needed in patrol since patrol is regarded as the backbone of policing and provides direct services to the community. It is the most important and visible operation of any police department and the percentage of officers assigned to patrol can range from a high of 64.88% to a low of 32.44%. There appears to be only anecdotal information to support the authorized sworn strength of the department which is currently 41. Of this number, approximately 56% (23)3 of personnel is dedicated to patrol. However, according to Koper et al (2001), in their analysis of small police departments, 81% to 85% of sworn officers serve in field units which also include support units (e.g. detectives, youth and juvenile officers, K-9). Historically, police administrators have envisioned a number of purposes for patrol. O.W. Wilson’s view dominated the conventional wisdom that guided patrol operations for much of the twentieth century. For Wilson (1963), the purpose was to have an officer patrol his beat [and] be alert for conditions that may jeopardize the comfort, safety, and welfare of the people, and take actions to correct improper conditions. Today, with community policing, this early view of patrol has been substantially expanded and reenergized. Alpert and Dunham (1992) summarize the activities that are accomplished by a more traditional patrol unit as: $ $ $ $ $ $ $ $ $ $ $
deterring crime through routine patrol, enforcing the laws, investigating criminal behavior, apprehending offenders, writing reports coordinating efforts with prosecutor, assisting individuals in danger or in need of assistance, conflict resolution, keeping the peace, maintaining order, and keeping pedestrian and automobile traffic moving.
As community policing has been included in departments’ operational strategies, this list has expanded to include community-based or oriented activities. For example, Cordner (1995) advised that community policing requires police to work more closely with citizens 3
This number includes the patrol commander and youth officer. If these positions are backed out, 51% of personnel are dedicated to patrol.
6
and other governmental agencies when attempting to solve problems. Consequently, this list has grown to include activities such as: holding neighborhood meetings, working with other governmental and private agencies, and coordinating youth recreational activities, among others.
$ $ $
Community policing expands basic patrol activities well beyond traditional limits and involves officers in activities that previously were reserved for specialists from other units of the police department (see Appendix C). How well patrol officers can perform traditional tasks and those required under the rubrics of “community policing” continues to be debated and studied. In simple terms, patrol officers are to provide a timely response to citizens’ requests for service whether they relate to community policing or traditional needs. Their presence in a given area allows them to respond more quickly than if they were to be dispatched from a central location. When not responding to calls for service, officers’ patrol activities are designed to create a high visibility for the department and allow them to become involved in officer-initiated activities (e.g. problem solving). Historically, visibility was believed to repress crime. This “omnipresence” (a belief that the police are everywhere) would have a deterrent effect on potential criminals and at the same time increase citizens’ feeling of security. METHODOLOGY In order to assess whether the current staffing level of officers/corporals in patrol is sufficient to provide efficient calls-for-service response, a modified workload analysis was utilized. The department currently does not have a statistical package that can analyze the data found on the Records Management System (RMS), particularly over time. Therefore, information in the RMS had to be physically manipulated in order to obtain needed data. Without physically plotting crimes and other offenses over time on a map, there is no way to view crime spatially. Therefore, there is no recommendation on beat configuration nor is there an analysis of crime by beat. The following sources of data/information were used in this process: • • • • •
Population increases from 1970 to 2000 (from census data); Population characteristics (from census data); A review of Part I (1990 – 2002), Part II, and other offenses for years 2000, 2001, and 2002; Calls for Service (2001-2002); Calls Answered (by Officers/Corporals); 7
• • •
Daily Officer Summaries; Officer Activity Reports; and Response Time Analysis Reports.
DISCUSSION Correlation between population and police strength Various methods of determining police organization strength rely on increases in population. However, when reviewing population for the period 1960 – 2000 (current census data), there has not been a significant increase, particularly between 1980 and 2000. While there was an increase between 1960 and 1970 (+27%) and a decrease between 1970 and 1980 (-13%), between 1980 and 2000, population has remained level (* 3 neighborhoods were annexed in 1996).
Year
Population
2000* 1990 1980 1970 1960
17,299 16,700 16,231 18,455 13,341
Using the measurement of officers per 1,000 population, the city has 2.4 officers per 1,000 residents. This number is below both the national (2.6 per 1,000) and Maryland (2.8 per 1,000) averages (source: National Directory of Law Enforcement Administrators, National Public Safety Information Bureau, 2002). However, a more interesting assessment of population and police strength involves analyzing density characteristics and then determining whether the police department accurately reflects the racial and cultural make-up of the community. Many police agencies take into consideration the racial characteristics, particularly languages spoken, in determining whether strength and strength composition meet community needs. The question to be asked of police department officials and the city manager is whether the current work force reflects the racial/cultural characteristics of the city and whether it is necessary to do so?
8
Population Racial Characteristics Race White Black or African American American Indian or Alaska Native Asian Asian Indian Chinese Filipino Japanese Korean Vietnamese Other Asian Native Hawaiian or Other Pacific Islander Native Hawaiian Other Pacific Islander Some Other Race Two or more races Hispanic
Number
Percent
8,440 5,876 76
48.8 34.0 0.4
754 333 88 107 26 29 103 68 6
4.4 1.9 0.5 0.6 0.2 0.2 0.6 0.4 0.03
3 3 1,287 860 2,494
0.01 0.01 7.4 5.0 14.4
Language Spoken at Home Although this study was unable to determine the percentage of calls coming to the department that are in languages other than English, or the number of calls not received because the caller did not feel comfortable calling because of a lack of fluency in English, or those calls not made because of cultural aspects that inhibit interaction with governmental authority, the following could also be used to assessing strength. Language Spoken at Home Language other than English Speak English less than very well Spanish Speak English less than very well Other Indo-European languages Speak English less than very well Asian and Pacific Island languages Speak English less than very well
Population
%
5,034 2,070 2,265 1,079 1,218 402 447 251
31.6% 13.0% 14.2% 6.8% 7.6% 2.8% 2.8% 1.6% 9
Determining whether to add additional officers fluent in languages other than English would be dependent on the number of officers currently fluent in languages other than English and the languages spoken by the officers. Families with Children under 18 In police parlance, “juveniles” are different. Since, they are not adults, their interactions with the police are covered by different laws, rules, and regulations than adults. While every police officer is familiar with these rules and regulations, most police departments have recognized the need to have police officers Juvenile Arrests for TPPD available (or at least accessible) on every tour of 80 duty who are versed in juvenile law and 70 procedures. This is not to say that “juvenile” or 60 “youth” officers have to be discrete categories of 50 police personnel. If patrol Number officers are not sufficiently Juvenile of Arrests 40 Arrests trained to handle juveniles, 30 adding these officers may be advised. Currently the 20 department has one (1) youth officer. The chart 10 (left) summarizes only arrests for juveniles. 0 1999 2000 2001 2002 However, it can be surmised that the number Year of juveniles with whom police officers come in contact, with no arrest or report, could at least double this number of contacts. Although there are no specific data to indicate that children who live in poverty and/or female headed households are more prone to have negative interactions with the criminal justice system, some criminologists/sociologists and juvenile practitioners believe that there is some merit in police having these data to determine whether programs or personnel are needed to provide preventive/proactive interventions. The following charts show those categories.
10
Category Families with children under 18 Households with individuals under 18 Families below poverty level with related children under 18 Families below poverty level with related children under 5 Families with female householder, no husband present with related children under 18 Families with female householder, no husband present with related children under 5
Number
Percentage (of total Population)
2,121 2,342 200
30.8% 34% 8.7%
102
10.2%
90
16%
30
6%
Population and Crime As stated earlier in this report, when conducting strength analysis, many researchers attempt to make a correlation between increases in population and increases in crime.
20000 18000 16000 14000 12000 10000 8000
Part I Part II Population
6000 4000 2000 0 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02
Population
11
For the data available, crime and population have remained constant with no appreciable spikes or peaks that would allow a conclusion to be drawn that a rise in population is directly related to a rise in crime.
Workload An analysis of work placed on the police officer can provide a more realistic view of whether or not current staffing levels are appropriate for the work being performed. 2002 80 70
Homicide Rape Robbery Assault Burglary Larceny Auto Theft
60 Number of Offenses
50 40 30 20 10
ov N
p Se
l Ju
M ay
M
Ja
n
ar
0
The chart above and the two below show Part I offenses for the years 2002, 2001, and 2000. The total number of offenses for each year are: •
1,222 (2002)
•
1,147 (2001), and
•
1,222 (2000).
12
2001
70 Homicide Rape Robbery Assault B&E Larceny Auto Theft
60 50 Number of Offenses
40 30 20 10
N ov
Se p
Ju l
M ay
Ja n
M ar
0
2000 70 60
Rape Robbery
40
Assault Burglary Larceny
30 20
Auto Theft
10
ov N
p Se
Ju l
M ay
ar M
n
0
Ja
Number of Offenses
Homicide
50
13
The increase/decrease in offenses can be shown for each crime category: 2000 Crime Homicide Rape Robbery Assault Burglary Larceny Auto Theft TOTALS
2001
Total 1 5 97 160 137 494 148 1042
Crime Homicide Rape Robbery Assault Burglary Larceny Auto Theft TOTALS
Total 1 4 70 162 150 562 198 1147
2002 %Change 0 -20 -28 +3.7 +9.4 +12 +25
Crime Homicide Rape Robbery Assault Burglary Larceny Auto Theft TOTALS
Total 0 6 92 148 107 639 230
%Change -100 +33 +24 -8.0 -29 +12 +14
1222
The greatest number of offenses has occurred consistently in the category of larcenies.4 However, larcenies are difficult to prevent and/or solve if the victims are not more circumspect about their property. It would be difficult to justify increasing the number of police officers, for example, by establishing a special squad, in an attempt to decrease larcenies. Reduction in larcenies may be more in the realm of community education. However, the same is not true for auto theft (and attempts). Although largely opportunistic in nature (vehicles, if not garaged, are in plain view of potential criminals), and while owners have some responsibility in protecting their vehicles, depending on the nature of the thefts, police can have an affect. A specialized squad focusing on types of vehicles most likely to be stolen, the hours most likely to experience thefts, and the locales that experience thefts can have a deterrent affect. The amount of time and effort needed to achieve a reduction in these thefts is not realistic for patrol officers. Based on protracted and intensive crime analysis, these officers must be able to change tours of duty, locations, and methods, not usually possible for patrol officers. When assessing Part I offenses, one of the most widely used (if not effective) tools to measure “success” in policing is clearance (arrest) rates. For obvious reasons, larcenies are not included in this analysis nor are Part II offenses (because of all the offense/incident classifications included in this category).
4
Larcenies are thefts of property not physically on or in possession of a person. For example, a wallet is placed on a counter in a store and is stolen or a bicycle is left on the front lawn of a house and is stolen.
14
Offense
Year
Total/Cleared
Year
Total/Cleared
Year
Total/Cleared
Homicide Rape Robbery Assault Burglary Auto Theft
2000 2000 2000 2000 2000 2000
1/0 5/4 97/25 160/93 137/36 148/10
2001 2001 2001 2001 2001 2001
1/1 4/0 70/10 162/56 150/13 198/7
2002 2002 2002 2002 2002 2002
0/0 6/3 92/18 148/52 107/14 230/14
Clearance rates for the above offenses, across years, range from a low for Auto Theft (5%) to a high for Rape (53%).5 Arrest Rate %
Offense Rape Homicide Assault Robbery Burglary Auto Theft
53% 50% 43% 20% 16% 5%
Calls for Service 14000 12000 Calls for Service
10000 8000
2000
6000
2001 2002
4000 2000 0
Part I
Part II
Other
Crime rates alone are not an adequate measure of workload. There must be a comparison of all offenses/incidents reported to the police department and all calls for service. It is important to recognize the overlap between the two categories. However, they provide
5
It should be noted that offenses are cleared under the UCR (Uniform Crime Reporting Guidelines) when a physical arrest has been made. Most police departments consider offenses cleared when either an arrest has been made or when an arrest warrant has been issued for an identified suspect.
15
a truer measure of work conducted by patrol officers. The chart above compares the number of Part I and Part II offenses (2000 – 2002) with all other offenses/incidents. “Other” comprises more than six times the number of offenses reported in the other two categories.6 Reported offenses still do not totally reflect the workload placed on patrol. Although a small portion of calls for service may be directed toward other units within the department, the majority of calls require some type of response from patrol. 19400 19200 19000 Calls for Service
18800
2000 2001 2002
18600 18400 18200 18000 17800
Year
Calls for Service
2000
18,428
2001
18,633
2002
19,288
As the charts above indicate, between 2000 and 2002, there was a 4.5% rise in calls for service, although they have remained fairly constant over the years examined. To further analyze calls for service, data on calls answered by patrol officers were obtained. This was necessary because not all calls require a response from a patrol officer. Here, again, only calls answered by patrol officers were analyzed. Sergeants 6
“Other” can include calls such as “suspicious person,” “kids loitering,” etc., which may or may not result in a report or other police action.
16
and above were screened out because these numbers distort the number of personnel available to answer calls. These data were obtained from Patrol Activity Sheets kept by the Patrol Commander that list calls answered by officers for each month during the two years reviewed, 2001 and 2002. Calls for Service (2001 and 2002)
3000 2500 Calls for Service
2000 2001 2002
1500 1000 500 0
ov N
p Se
ly Ju
ay M
ar M
n
Ja
For 2001, the number of calls answered by officers was highest for March, May, and September. For 2002, there was a minor increase throughout the year with the exception of July, September, November, and December. However, even with these minor dips, the number of calls answered remained fairly constant over each of the two years. When pulling the RMS data for 2001 and 2002, data indicate that the average time spent on each call answered was twenty-one (21) minutes. This reflects and a two (2) minute decrease from 2000 (23 minutes). Time Worked and Available for Calls for Service Prior to the advent of community policing, police (patrol) managers were primarily concerned with how fast patrol units responded to calls for service (response time), the number of arrests made, and the number of traffic tickets or citations written. However, community policing required that time be spent by patrol officers in preventive (proactive) activities, working with the community, and “managing” their assigned patrol areas. This required that an assessment be made of how much time was being spent on traditional police activities and how much was available for community policing. 17
As a point of reference, the following chart (for 2002) shows the number of hours actually worked, the number of hours worked, but unavailable (court, training, prisoner processing, etc.), and the number of hours unavailable (e.g, personal leave, workers compensation, military leave, etc.). As the data indicate, for the majority of hours, patrol officers were available and on the street to answer calls for service and engage in community policing activities.
4000 3500 Number of Hours
Hours Worked
3000 2500
Hours Unavailable
2000 1500
Hours Worked, Not Available
1000 500
Ju l Se p N ov
Ja n M ar M ay
0
To fully explore the ability to target patrol officers’ time in order to fulfill community policing expectations, additional analysis is required. The charts that follow show the number of hours worked versus the number of hours unavailable, but not available for calls for service. The first chart indicates that for the year 2001, an average of 4000
Hours Worked vs. Hours Worked, but Unavailable for Calls
3500 3000 2500
Hours Worked 2001 Hours Unavailabl e
2000 1500 1000 500
D ec
Se p O ct N ov
Ju n Ju l Au g
0 Ja n Fe b M ar Ap r M ay
Number of Hours
18
twenty-five percent (25%) of the patrol officers’ time was unavailable for patrol activities. Hours worked ranged from a high of 3,554 hours (June) to a low of 2,545.5 hours (December), with the average available hours being 3,139 hours. In terms of hours worked, but unavailable for patrol activities, the range was a high of 1,302 (June) to a low of 228 hours (September), with an average being 782 hours. Hours Worked vs. Hours Worked, but Unavailable for Calls
4000 3500 3000 2500
Number of Hours
Hours Worked 2002
2000
Hours Unavailable
1500 1000 500
Ju l Au g Se p O ct N ov D ec
Ja n Fe b M ar Ap r M ay Ju n
0
For 2002, officers were unavailable to answer calls for service an average of 16% of the time. For hours worked, the range was a high of 3,405.5 for November, to a low of 2,604.5 hours for February. The average was 2,941 hours. In terms of hours unavailable, the range was a high of 803.5 hours (September) and a low of 220 hours for May, with the average being 482.5 hours. “Working, but unavailable” is still only part of the equation in determining how much time is available for community policing activities. Time spent “Out of Service” on calls for service or other dispatched activities must also be analyzed. This analysis was also conducted for the two tours of duty currently utilized, 10 pm – 8 am and 8 am – 10 pm. Data prior to September 2001(when three tours were utilized) had to be collapsed into the two tours. Due to the volume of data available, but not user-friendly, a decision was made to use data from the 15th of each month for the years 2001 and 2002.
19
For 2002, looking at the 15th of each month, a total of 37,440 hours were worked on the 10 pm – 8 am tour of duty. Of this 14,958 hours were spent “out of service” (40%). On the 8 am – 10 pm tour of duty, again for the 15th of each month, a total of 41,040 hours were worked and a total of 11,487 was spent “out of service” (28%). When averaged across the two tours, 34% of worked hours were spent “out of service.”
2002: Hours Available vs. Hours Out of Service
5000 4500 4000 3500 3000
Number of Hours
10 pm - 8 am Available 10 pm - 8 am OOS
2500 2000 1500 1000 500
D ec
Se p O ct N ov
Ju n Ju l Au g
Ja n Fe b M ar Ap r M ay
0
When looking at 2001, the following was revealed. On the 10 pm – 8 am tour of duty, 29% of the time worked was spent “out of service.” For the 8 am – 10 pm tour, 30% of the time worked was spent “out of service.” When combined, the total percentage of time spent unavailable for other activities was 30%.
20
2002: Hours Available vs. Hours Out of Service
5000 4500 4000 3500 3000
Number of Hours
2500
8 am - 10 pm Available 8 am - 10 pm OOS
2000 1500 1000 500 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2001: Hours Available vs. Hours Out of Service
4000 3500 3000 2500 10 pm - 8 am Available Hours OOS
2000 1500 1000 500
Ju l Au g Se p O ct N ov D ec
0 Ja n Fe b M ar Ap r M ay Ju n
Number of Hours
21
2001: Hours Available vs. Hours Out of Service
4000 3500 3000 2500
Number of Hours
8 am - 10 pm Available Hours OOS
2000 1500 1000 500
D ec
Se p O ct N ov
Ju l Au g
Ja n Fe b M ar Ap r M ay Ju n
0
Officers Available Another aspect of determining how much time is available for community policing activities is the number of officers available. Again, there is no hard and fast rule for determining minimum and maximum (optimal) staffing levels. According to department officials, the optimum level of personnel on the street is six (6), the minimum is three (3). Again, when assessing time available for 2001 and 2002 using the 15th of each month, average staffing levels were also obtained.
Year/Tour
% Out of Service
Staffing Level
2002 2002 2001 2001
40 28 29 30
4.33 4.83 4.00 3.92
Although using the arbitrary date of the 15th of each month, the data indicate that the optimal staffing level was rarely achieved. Another point is that when taking into account the percentage of time “out-of-service,” “actual” staffing levels provide a more realistic 22
assessment of the number of officers on the street as opposed to the number of officers assigned at the beginning of a tour of duty. Response Time Most researchers/practitioners have agreed that no patrol beat should be so large that calls for service cannot be answered in three minutes. Although an analysis was not conducted by beat, the “Response Time Analysis” by offense type and shift in the RMS system was queried for the years 1999 through 2002. Records indicate that average response time is three (3) minutes. FINDINGS AND RECOMMENDATIONS In order to obtain a more clearer assessment of time available for preventive or community policing activities, times “out-of-service” and ‘time working, but unavailable” must be combined. For the year 2001, an average of 54% of patrol officers’ time was unavailable and “out of service.” For the year 2002, an average of 50% of officers’ time was unavailable and “out of service.” There is no recognized standard of how much time should be spent for various activities or time unavailable for providing response, however, in 1992, the International City/County Management Association recommended that no more than 35% of an officer’s time be spent responding to calls for service. However, ICMA’s recommendation does not take into account the amount of time “working, but unavailable.” Therefore, for purposes of this study, the two times are combined and then compared with the 35% recommendation. If the “standard” of 35% is utilized, then an additional 3.3 officers are required to bring the 50% (combined Out-of-Service and Working, but Unavailable) down to 35%. With a relief factor of 1.6 that allows for leave and days off, maintaining this level of staffing would require a total of 5.28 additional officers. However, if there is a decision that allows sergeants to act solely as sector commanders/supervisors and have the six (6) corporals totally assume street supervisory duties, if the corporals are backed out of the equation, a total of 14.88 officers would need to be added to patrol (6 officers to replace the six corporals, added to the 3.3 officers to reduce time to 35%, times the relief factor of 1.6). To reiterate comments from the beginning of this report, determining personnel strength of any police department finally comes down to managerial decisions based on available data and the needs of the community. In order to provide data on a protracted basis on which to make managerial decisions, the following is also recommended: • on-going analysis of data, over time, with a statistical software package designed to analyze vast amounts of disparate data;
23
• • • • • •
more detailed accounting of patrol activities and time spent that is recorded in both the RMS system and through use of a detailed officer “run sheet;” detailed analysis of the activities of support personnel as they relate to the patrol function; comprehensive accounting of all current and proposed patrol activities; temporal and spatial analysis of crime and other incidents; assessment of alternative response to crime and incidents; and the role of management and supervision within the Takoma Park Police Department.
24
Sources: Alpert, G. and R. Dunham (1988) Chaiken, Jan (1975) Chamlin, Mitchell (1996) Cordner, G. (1995) Eck, John and E. Maguire (2000)
Gaines, L., E. Kappeler, J. Vaughn Koper, Christopher et al (2001)
Loftin, Colin and D. McDowall (1982) Nalla, Mahesh K, M. Lynch (1997)
Roberg, R. and J. Kuykendall (1997) Wilson, O.W. (1963)
Policing Urban America. Prospect Heights, ILL, Waveland Press Patrol Allocation Methodology for Police Departments, Rand Corporation, California “The Police, Crime, and Economic Theory: A Replication and Extention,” American Journal of Criminal Justice, 20:165-182 “Community Policing: Elements and Effects,” Police Forum, 5(3):1-7. “Have Changes in Policing Reduced Violent Crime? An Assessment of the Evidence,” in The Crime Drop in America, edited by Alfred Blumstein and Joel Wallman, 207-265, New York, Cambridge University Press Policing in America, 2nd ed., Cincinatti, OH, Anderson Publishing “Hiring and Retention Issues in Police Agencies: Readings on the Determinants of Police Strength, Hiring, and Retention of Officers, and the Federal COPS Program,” Urban Institute, Washington, D.C. “The Police, Crime, and Economic Theory: An Assessment,” American Sociological Review 47:393-401. “Perspectives on the Growth of Police Bureaucracies, 1948-1984: An Examination of Three Explanations,” Policing and Society, 3:5161. Police Management, Los Angeles, Roxbury Press Police Administration, 2nd ed., New York, McGraw-Hill
25
Appendix A. Comparison of Municipal Police Department Staffing in the States of Maryland and Virginia (10,000 - 20,000 residents) $ $
National average of officer per 10,000 residents: 26.7/10,000 ( 2.6 per 1,000) Maryland average of officer per 10,000 residents: 28.7/10,000 (2.8 per 1,000) (Both national and Maryland numbers based on departments having 100 or more sworn personnel)
Maryland
(No distinction made between full-service and less-than-full-service departments)
City/Town
Population
# Officers
Ratio (per 1000 residents)
Comments
Aberdeen
14,000
39
2.9
more than Takoma Park
Bel Air
10,000
30
3.0
more than Takoma Park
Cambridge
12,500
44
3.5
more than Takoma Park
Crofton
10,000
6
.6
Easton
10,000
40
4.0
Elkton
13,000
29
2.2
Greenbelt
21,096
53
2.5
more than Takoma Park
Havre de Grace
10,000
24
2.4
more than Takoma Park
Hyattsville
15,000
29
1.9
Mt. Rainier
11,800
14
1.1
Ocean City
10,500
13
1.2
Ocean Pines
10,500
13
1.2
Princess Anne
15,000
14
.9
more than Takoma Park
26
7
Sparrow’s Point
15,000
12
.8
Takoma Park7
18,600
42
2.3
These numbers are listed in the cited report and do not reflect the current police strength of 41 or the current population of 17,333.
27
Virginia
(No distinction made between full-service and less-than-full service departments)
Bristol
18,281
50
2.7
City Sheriff (see below)
Bristol
18,281
60
3.2
Police Dept. (see above)
Christianburg
16,000
25
1.5
Colonial Heights
17,500
7
.04
City Sheriff (see below)
Colonial Heights
17,000
50
2.9
Police Dept. (see above)
Culpepper
10,600
30
2.8
Falls Church
10,000
21
2.1
City Sheriff (see below)
Falls Church
9,900
32
3.2
Police Dept. (see above)
Fredericksburg
20,700
9
.4
City Sheriff (see below)
Fredericksburg
24,900
85
3.4
Police Dept. (see above)
Front Royal
12,869
34
2.6
Herndon
17,732
50
2.8
Martinsville
16,000
48
3.0
City Sheriff (see below)
Martinsville
16,162
54
3.3
Police Dept. (see above)
Poquoson
11,300
17
1.5
Radford
15,900
26
1.8
City Sheriff (see below)
Radford
13,914
26
1.8
Police Dept. (see above)
Williamsburg
55,000
13
.2
City Sheriff (see below)
Williamsburg
11,500
29
2.5
Police Dept. (see above)
Source: National Directory of Law Enforcement Administrators...., National Public Safety Information Bureau (2002) Highlighted areas have a higher officer-to-resident ratio than Takoma Park. 28
Appendix B Comparative Analysis Approach: San Jose Police Department In 1993, the San Jose Police Department had approximately 16 officers per 10,000 residents (1,312 officers). San Jose could be compared to either one or a group of police departments to determine the differences between the San Jose mean ratio of officers to residents with that of the comparison department or group. The difference is multiplied by 82,which is the approximate number of units of 10,000 population (820,000) in San Jose in 1993. Six possible comparisons are provided. 1.
Using the national average (1993) of approximately 22 officers per 10,000 residents (for local police departments with more than 100 officers), San Jose should have had 492 more officers.
2.
Other than San Jose, there are four cities in California with 1,000 or more officers. This group has a mean officer to-resident ratio of 18.2 per 10,000. Based on this comparison, San Jose should have had 180 more officers.
3.
Of municipalities with 100 or more officers, in 1993 Atlantic City, New Jersey, had the highest officer-to-resident ratio at 112:10,000. Compared with this city, San Jose should have had 7,872 more officers.
4.
Compared with Washington, D.C., the city with more than 1,000 officers that has the highest officer-to-resident ratio at 70:10,000, San Jose should have had 4,420 more officers.
5.
Compared with Indianapolis, the city with more than 1,000 officers that had the lowest officer-to-resident ratio at 13:10,000, San Jose had 246 too many police officers.
6.
Compared with Anchorage, Alaska, the city with more than 100 officers that had the lowest officer-to-resident ratio at 11:10,000, San Jose had 410 too many officers.
Source:
Roberg, R. and J. Kuykendall (1997), Police Management, Los Angeles, Roxbury Press. 29
Public Safety Citizens Advisory Committee March 2003
Contents Executive Summary……………………………………………………… 1 Introduction………………………………………………………………. 3 Policing in Takoma Park………………………………………………… 6 Methodology……………………………………………………………... 7 Discussion……………………………………………………………….. 8 Correlation between Population and Police Strength……….. 8 Language Spoken at Home……………………………………. 9 Families with Children under 18………………………………. 10 Population and Crime………………………………………….. 11 Workload………………………………………………………… 12 Calls for Service…………………………………………………. 15 Time Worked and Available for Calls for Service……………. 17 Officers Available……………………………………………….. 22 Response Time…………………………………………………… 23 Findings and Recommendations……………………………………….. 23 Sources……………………………………………………………………. 25 Appendices……………………………………………………………….. 26
Executive Summary As part of the mandate to revisit the implementation of community policing in the City of Takoma Park, one of the crucial elements is the determination of whether current patrol strength of the Takoma Park Police Department is adequate to perform the basic functions of answering calls for service, maintaining visibility, and conducting preliminary1 investigations of crimes and offenses. In order to conduct this assessment, the Records Management System (RMS) was queried, questions asked of management officials, informal conversations held with patrol officers, supervisors and staff. As a result of this analysis, it is recommended that a minimum of 5.28 officers be added to the patrol function. This number is based on backing patrol sergeants out of the equation, with sergeants ultimately performing purely as sector commanders/supervisors (the city is divided into three sectors). There is a second recommendation which is based on also backing corporals out of the patrol equation and having them function as patrol supervisors. Should this occur, the number of officers to be added to the patrol function should be 14.88. The methodology explaining these conclusions can be found in the body of the report.2 It is also recommended that increases in personnel be further analyzed based on an assessment of: • which calls should be considered for other than a mobile response; • supervisory and management styles; • proposed tasks for all personnel under community policing (see Appendix C); • the detective function; • the training function; • changes to the department mandated by accreditation; and • personnel assigned to support functions. It is hoped that the information contained herein will assist the department in meeting its mandate of providing the highest quality police service to the residents, businesses, and organizations of Takoma Park. There is a caveat, however, any analysis and report must be considered fluid due to the increased functions placed on police departments across this nation in this era of increased terrorism and the need for homeland security. As a final note, this study would not have been accomplished with the complete cooperation of the members and staff of the Takoma Park Police Department who provided total access to records, data, and personnel. Thanks are also extended to the staff of the City Manager’s office that provided census data. 1
Investigations of crimes or offenses are initially conducted by patrol officers until it is determined that the services of detectives are needed or required.
1
Andrew Kelemen Chair, Public Safety Citizens Advisory Committee
Gayle Fisher-Stewart Vice Chair, Public Safety Citizens Advisory Committee
2
INTRODUCTION Determining the number of sworn personnel, particularly those There is no one allocated to the patrol function is not a simple task. Police best or simple strength is an imprecise term. Various researchers have tended to operationalize it in a number of ways, the three most way to common being the number of sworn officers, the number of determine the police employees, and the amount of police expenditures level of police (Maguire, 2001). Those using police expenditures justify it on the basis that policing is a personnel intensive industry, and personnel a there is an almost perfect correlation between the number of department police personnel that a jurisdiction employs and the amount it should have. spends on police protection. A handful of researchers have also discussed the difference between absolute and relative police strength (Chamblin and Langworthy, 1996; Slovak, 1986). Absolute strength is the raw number of officers or employees or amount of expenditures in a jurisdiction, while relative strength expresses these variables as a ratio (usually per capita or per unit area). The available resources for local police departments (municipal police, county sheriff) with 100 or more officers varies considerably from 1 to 112 officers per 10,000 residents, with the national average being approximately 26.7 per 10,000 residents. Still, there have been other attempts to understand the determination of police strength that are theory based (Eck and Maguire, 2000). Studies of police strength using theory are typically based (at least implicitly) on one or more of the following three theoretical foundations: (1) rational public choice (or consensus) theory, (2) conflict theory, and/or (3) some form of organizational theory (Nalla, Lynch, and Leiber, 1977). Rational public choice (also known as consensus or economic) theory implies that police organizations grow in response to citizens’ consensual requests for increased public service and protection, often in the face of rising crime rates. This theory implies that local governments and police administrators dole out resources systematically. As stated above, for decades, communities have relied on various methods for determining the appropriate amount of police resources at given times and places. These methods range from informal rules of thumb, such as the need to have at least two officers per thousand residents, to more complex mathematical models implemented in computer software packages (Bayley, 1994; Chaiken, 1975; Larson, 1978; O’Boyle, 1990, Stenzel, 1993). The kinds of variables that have typically been included in these formal and informal systems are the usual suspects: crime rates, population, calls-for-service, and other correlates of police workload. These methods reflect an implicit theory of administrative rationality, suggesting that police strength is a function of a few simple workload variables. As Loftin and McDowall (1982:400) concluded, rational choice or economic models are “too simple to account for the relationship between crime and police strength.” They suggest that models of police strength need to account for other factors in the social and political environment of police organizations. 3
If police strength is not to be based on a rational Regardless of the adjustment to variations in crime rates, then what other method used to factors are important? Conflict theory posits that racial determine police and economic inequality lead the powerful members (elites) in a community to exert political influence over strength, it remains a social control institutions. Such conflict processes would management lead to increases in social control over the powerless. decision. Statistical Tests of conflict theory have concentrated on racial and/or analysis provides economic sources of conflict. These tests are usually done information on which by examining whether variations or changes in the size of to base that decision. “threatening” populations produce differences in police strength. Research on social threat or conflict theory is also overwhelmingly mixed, although there is evidence to support both racial conflict (e.g. Jackson and Carroll, 1981; Liska et al, 1981; Snipes, 1993) and economic or class conflict explanations (e.g., Jacobs, 1978, 1979). Finally, organizational theories examine the effect of processes within police organizations in producing increases or decreases in police strength. This class of theories is the least developed of those discussed. Typically, the only explanation in this line of research is organizational inertia: that changes in police strength are incremental and that the best predictor of police strength in a given year is its value in the prior year (known as a “lag”). Therefore, unlike other perspectives, organizational explanations assume that changes in police strength are based on conditions internal to the organization. According to Nalla and his colleagues (1997:120): “this theoretical perspective assumes that organizational strength, as measured by annual budgets is explained by incremental specification models whereby the present year’s budget is influenced by appropriations in the previous year.” Numerous studies have found evidence to support the organizational inertia explanation (Nalla, 1992; Nalla et al, 1997), however, as with the other approaches, there are also shortcomings to this approach. The first is that it has little serious use in that the lag value of any variable in a time series is nearly always the best predictor of the current value and, second, it is useful for understanding stability in police strength, but it is not useful for understanding growth and decline over time. Overall, organizational explanations in this literature tend not to be well defined, treating police organizations as a “black box.” Even with its problems, the methods used in examining the determinants of police strength continue to grow more sophisticated, with recent refinements suggesting some excellent reasons for mixed findings obtained in the past (Brandl, Chamlin, and Frank, 1995; Snipes, 1993). The relationship between police strength and crime rates is known as a “simultaneous” or “reciprocal” causal relationship because each one is known to cause the other. A variety of specialized methods have been devised by researchers to 4
disentangle simultaneous causal effects, however these methods have been (and remain) the domain of economists. In fact, the majority of research in this arena has been conducted by economists and the result has been that research in this area is not widely known. This lack of acceptance is due to the fact that many of the articles appear in economic journals and the methods used are probably too difficult for criminologists and criminal justice practitioners to understand and use without advanced training in econometrics. In the final analysis, unfortunately, there is no reliable method to determine the resources that a police department needs, because there is no precise and complete listing of the tasks and activities in which police should engage or how long it takes to effectively perform those tasks and activities. However, police In the final managers need to be able to justify increases and/or analysis, police decreases in personnel strength.
strength is a function of informed “guesstimates.”
For that reason, there are three generally accepted methods used by police departments to determine the resources needed. The intuitive approach is essentially an educated guess based on the experience and judgment of police managers. The workload approach requires an elaborate information system, standards of expected performance, well-defined community expectations, and prioritization of police activities. To determine workload, the department must be able to identify precisely what the community expects the police to do, which calls must be processed, which calls can be ignored or given a minimum response, and an expected level of service. Although the workload approach is rarely used for an entire department, it is sometimes used internally to determine the needed personnel for functions (e.g. patrol) or programs (e.g. crime prevention). Perhaps the most frequently used method, alluded to This study uses a earlier, is the comparative approach based on the ratio of police officers per 1,000 or 10,000 residents (see appendix modified workload A). However, this method is not very reliable because analysis. departments tend to be diverse with respect to community problems and public expectations, managerial effectiveness, use of technology, policing styles, and the productivity of officers (see appendix B for a sample analysis). There is an added caveat to this method, there are no data using this method for police departments with fewer than 100 officers. Major studies have not been conducted to analyze departments with fewer than 100 officers. Roberg and Kuykendall (1997) provide that they know of no evidence to suggest that even a ratio within extreme parameters (11 to 112 or 13 to 70 officers per 10,000 residents) makes a measurable difference in the crime rate, citizen fear of crime, or the attitudes of citizens. More than likely, at some point, increasing police resources would result in the police becoming more effective; however, this point will vary among 5
communities based on conditions that contribute to, or detract from, the frequency and type of police problems and how effectively the police are managed. POLICING IN TAKOMA PARK In order to determine the number of police officers for the City of Takoma Park, this report focuses on the number of police officers needed in patrol since patrol is regarded as the backbone of policing and provides direct services to the community. It is the most important and visible operation of any police department and the percentage of officers assigned to patrol can range from a high of 64.88% to a low of 32.44%. There appears to be only anecdotal information to support the authorized sworn strength of the department which is currently 41. Of this number, approximately 56% (23)3 of personnel is dedicated to patrol. However, according to Koper et al (2001), in their analysis of small police departments, 81% to 85% of sworn officers serve in field units which also include support units (e.g. detectives, youth and juvenile officers, K-9). Historically, police administrators have envisioned a number of purposes for patrol. O.W. Wilson’s view dominated the conventional wisdom that guided patrol operations for much of the twentieth century. For Wilson (1963), the purpose was to have an officer patrol his beat [and] be alert for conditions that may jeopardize the comfort, safety, and welfare of the people, and take actions to correct improper conditions. Today, with community policing, this early view of patrol has been substantially expanded and reenergized. Alpert and Dunham (1992) summarize the activities that are accomplished by a more traditional patrol unit as: $ $ $ $ $ $ $ $ $ $ $
deterring crime through routine patrol, enforcing the laws, investigating criminal behavior, apprehending offenders, writing reports coordinating efforts with prosecutor, assisting individuals in danger or in need of assistance, conflict resolution, keeping the peace, maintaining order, and keeping pedestrian and automobile traffic moving.
As community policing has been included in departments’ operational strategies, this list has expanded to include community-based or oriented activities. For example, Cordner (1995) advised that community policing requires police to work more closely with citizens 3
This number includes the patrol commander and youth officer. If these positions are backed out, 51% of personnel are dedicated to patrol.
6
and other governmental agencies when attempting to solve problems. Consequently, this list has grown to include activities such as: holding neighborhood meetings, working with other governmental and private agencies, and coordinating youth recreational activities, among others.
$ $ $
Community policing expands basic patrol activities well beyond traditional limits and involves officers in activities that previously were reserved for specialists from other units of the police department (see Appendix C). How well patrol officers can perform traditional tasks and those required under the rubrics of “community policing” continues to be debated and studied. In simple terms, patrol officers are to provide a timely response to citizens’ requests for service whether they relate to community policing or traditional needs. Their presence in a given area allows them to respond more quickly than if they were to be dispatched from a central location. When not responding to calls for service, officers’ patrol activities are designed to create a high visibility for the department and allow them to become involved in officer-initiated activities (e.g. problem solving). Historically, visibility was believed to repress crime. This “omnipresence” (a belief that the police are everywhere) would have a deterrent effect on potential criminals and at the same time increase citizens’ feeling of security. METHODOLOGY In order to assess whether the current staffing level of officers/corporals in patrol is sufficient to provide efficient calls-for-service response, a modified workload analysis was utilized. The department currently does not have a statistical package that can analyze the data found on the Records Management System (RMS), particularly over time. Therefore, information in the RMS had to be physically manipulated in order to obtain needed data. Without physically plotting crimes and other offenses over time on a map, there is no way to view crime spatially. Therefore, there is no recommendation on beat configuration nor is there an analysis of crime by beat. The following sources of data/information were used in this process: • • • • •
Population increases from 1970 to 2000 (from census data); Population characteristics (from census data); A review of Part I (1990 – 2002), Part II, and other offenses for years 2000, 2001, and 2002; Calls for Service (2001-2002); Calls Answered (by Officers/Corporals); 7
• • •
Daily Officer Summaries; Officer Activity Reports; and Response Time Analysis Reports.
DISCUSSION Correlation between population and police strength Various methods of determining police organization strength rely on increases in population. However, when reviewing population for the period 1960 – 2000 (current census data), there has not been a significant increase, particularly between 1980 and 2000. While there was an increase between 1960 and 1970 (+27%) and a decrease between 1970 and 1980 (-13%), between 1980 and 2000, population has remained level (* 3 neighborhoods were annexed in 1996).
Year
Population
2000* 1990 1980 1970 1960
17,299 16,700 16,231 18,455 13,341
Using the measurement of officers per 1,000 population, the city has 2.4 officers per 1,000 residents. This number is below both the national (2.6 per 1,000) and Maryland (2.8 per 1,000) averages (source: National Directory of Law Enforcement Administrators, National Public Safety Information Bureau, 2002). However, a more interesting assessment of population and police strength involves analyzing density characteristics and then determining whether the police department accurately reflects the racial and cultural make-up of the community. Many police agencies take into consideration the racial characteristics, particularly languages spoken, in determining whether strength and strength composition meet community needs. The question to be asked of police department officials and the city manager is whether the current work force reflects the racial/cultural characteristics of the city and whether it is necessary to do so?
8
Population Racial Characteristics Race White Black or African American American Indian or Alaska Native Asian Asian Indian Chinese Filipino Japanese Korean Vietnamese Other Asian Native Hawaiian or Other Pacific Islander Native Hawaiian Other Pacific Islander Some Other Race Two or more races Hispanic
Number
Percent
8,440 5,876 76
48.8 34.0 0.4
754 333 88 107 26 29 103 68 6
4.4 1.9 0.5 0.6 0.2 0.2 0.6 0.4 0.03
3 3 1,287 860 2,494
0.01 0.01 7.4 5.0 14.4
Language Spoken at Home Although this study was unable to determine the percentage of calls coming to the department that are in languages other than English, or the number of calls not received because the caller did not feel comfortable calling because of a lack of fluency in English, or those calls not made because of cultural aspects that inhibit interaction with governmental authority, the following could also be used to assessing strength. Language Spoken at Home Language other than English Speak English less than very well Spanish Speak English less than very well Other Indo-European languages Speak English less than very well Asian and Pacific Island languages Speak English less than very well
Population
%
5,034 2,070 2,265 1,079 1,218 402 447 251
31.6% 13.0% 14.2% 6.8% 7.6% 2.8% 2.8% 1.6% 9
Determining whether to add additional officers fluent in languages other than English would be dependent on the number of officers currently fluent in languages other than English and the languages spoken by the officers. Families with Children under 18 In police parlance, “juveniles” are different. Since, they are not adults, their interactions with the police are covered by different laws, rules, and regulations than adults. While every police officer is familiar with these rules and regulations, most police departments have recognized the need to have police officers Juvenile Arrests for TPPD available (or at least accessible) on every tour of 80 duty who are versed in juvenile law and 70 procedures. This is not to say that “juvenile” or 60 “youth” officers have to be discrete categories of 50 police personnel. If patrol Number officers are not sufficiently Juvenile of Arrests 40 Arrests trained to handle juveniles, 30 adding these officers may be advised. Currently the 20 department has one (1) youth officer. The chart 10 (left) summarizes only arrests for juveniles. 0 1999 2000 2001 2002 However, it can be surmised that the number Year of juveniles with whom police officers come in contact, with no arrest or report, could at least double this number of contacts. Although there are no specific data to indicate that children who live in poverty and/or female headed households are more prone to have negative interactions with the criminal justice system, some criminologists/sociologists and juvenile practitioners believe that there is some merit in police having these data to determine whether programs or personnel are needed to provide preventive/proactive interventions. The following charts show those categories.
10
Category Families with children under 18 Households with individuals under 18 Families below poverty level with related children under 18 Families below poverty level with related children under 5 Families with female householder, no husband present with related children under 18 Families with female householder, no husband present with related children under 5
Number
Percentage (of total Population)
2,121 2,342 200
30.8% 34% 8.7%
102
10.2%
90
16%
30
6%
Population and Crime As stated earlier in this report, when conducting strength analysis, many researchers attempt to make a correlation between increases in population and increases in crime.
20000 18000 16000 14000 12000 10000 8000
Part I Part II Population
6000 4000 2000 0 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02
Population
11
For the data available, crime and population have remained constant with no appreciable spikes or peaks that would allow a conclusion to be drawn that a rise in population is directly related to a rise in crime.
Workload An analysis of work placed on the police officer can provide a more realistic view of whether or not current staffing levels are appropriate for the work being performed. 2002 80 70
Homicide Rape Robbery Assault Burglary Larceny Auto Theft
60 Number of Offenses
50 40 30 20 10
ov N
p Se
l Ju
M ay
M
Ja
n
ar
0
The chart above and the two below show Part I offenses for the years 2002, 2001, and 2000. The total number of offenses for each year are: •
1,222 (2002)
•
1,147 (2001), and
•
1,222 (2000).
12
2001
70 Homicide Rape Robbery Assault B&E Larceny Auto Theft
60 50 Number of Offenses
40 30 20 10
N ov
Se p
Ju l
M ay
Ja n
M ar
0
2000 70 60
Rape Robbery
40
Assault Burglary Larceny
30 20
Auto Theft
10
ov N
p Se
Ju l
M ay
ar M
n
0
Ja
Number of Offenses
Homicide
50
13
The increase/decrease in offenses can be shown for each crime category: 2000 Crime Homicide Rape Robbery Assault Burglary Larceny Auto Theft TOTALS
2001
Total 1 5 97 160 137 494 148 1042
Crime Homicide Rape Robbery Assault Burglary Larceny Auto Theft TOTALS
Total 1 4 70 162 150 562 198 1147
2002 %Change 0 -20 -28 +3.7 +9.4 +12 +25
Crime Homicide Rape Robbery Assault Burglary Larceny Auto Theft TOTALS
Total 0 6 92 148 107 639 230
%Change -100 +33 +24 -8.0 -29 +12 +14
1222
The greatest number of offenses has occurred consistently in the category of larcenies.4 However, larcenies are difficult to prevent and/or solve if the victims are not more circumspect about their property. It would be difficult to justify increasing the number of police officers, for example, by establishing a special squad, in an attempt to decrease larcenies. Reduction in larcenies may be more in the realm of community education. However, the same is not true for auto theft (and attempts). Although largely opportunistic in nature (vehicles, if not garaged, are in plain view of potential criminals), and while owners have some responsibility in protecting their vehicles, depending on the nature of the thefts, police can have an affect. A specialized squad focusing on types of vehicles most likely to be stolen, the hours most likely to experience thefts, and the locales that experience thefts can have a deterrent affect. The amount of time and effort needed to achieve a reduction in these thefts is not realistic for patrol officers. Based on protracted and intensive crime analysis, these officers must be able to change tours of duty, locations, and methods, not usually possible for patrol officers. When assessing Part I offenses, one of the most widely used (if not effective) tools to measure “success” in policing is clearance (arrest) rates. For obvious reasons, larcenies are not included in this analysis nor are Part II offenses (because of all the offense/incident classifications included in this category).
4
Larcenies are thefts of property not physically on or in possession of a person. For example, a wallet is placed on a counter in a store and is stolen or a bicycle is left on the front lawn of a house and is stolen.
14
Offense
Year
Total/Cleared
Year
Total/Cleared
Year
Total/Cleared
Homicide Rape Robbery Assault Burglary Auto Theft
2000 2000 2000 2000 2000 2000
1/0 5/4 97/25 160/93 137/36 148/10
2001 2001 2001 2001 2001 2001
1/1 4/0 70/10 162/56 150/13 198/7
2002 2002 2002 2002 2002 2002
0/0 6/3 92/18 148/52 107/14 230/14
Clearance rates for the above offenses, across years, range from a low for Auto Theft (5%) to a high for Rape (53%).5 Arrest Rate %
Offense Rape Homicide Assault Robbery Burglary Auto Theft
53% 50% 43% 20% 16% 5%
Calls for Service 14000 12000 Calls for Service
10000 8000
2000
6000
2001 2002
4000 2000 0
Part I
Part II
Other
Crime rates alone are not an adequate measure of workload. There must be a comparison of all offenses/incidents reported to the police department and all calls for service. It is important to recognize the overlap between the two categories. However, they provide
5
It should be noted that offenses are cleared under the UCR (Uniform Crime Reporting Guidelines) when a physical arrest has been made. Most police departments consider offenses cleared when either an arrest has been made or when an arrest warrant has been issued for an identified suspect.
15
a truer measure of work conducted by patrol officers. The chart above compares the number of Part I and Part II offenses (2000 – 2002) with all other offenses/incidents. “Other” comprises more than six times the number of offenses reported in the other two categories.6 Reported offenses still do not totally reflect the workload placed on patrol. Although a small portion of calls for service may be directed toward other units within the department, the majority of calls require some type of response from patrol. 19400 19200 19000 Calls for Service
18800
2000 2001 2002
18600 18400 18200 18000 17800
Year
Calls for Service
2000
18,428
2001
18,633
2002
19,288
As the charts above indicate, between 2000 and 2002, there was a 4.5% rise in calls for service, although they have remained fairly constant over the years examined. To further analyze calls for service, data on calls answered by patrol officers were obtained. This was necessary because not all calls require a response from a patrol officer. Here, again, only calls answered by patrol officers were analyzed. Sergeants 6
“Other” can include calls such as “suspicious person,” “kids loitering,” etc., which may or may not result in a report or other police action.
16
and above were screened out because these numbers distort the number of personnel available to answer calls. These data were obtained from Patrol Activity Sheets kept by the Patrol Commander that list calls answered by officers for each month during the two years reviewed, 2001 and 2002. Calls for Service (2001 and 2002)
3000 2500 Calls for Service
2000 2001 2002
1500 1000 500 0
ov N
p Se
ly Ju
ay M
ar M
n
Ja
For 2001, the number of calls answered by officers was highest for March, May, and September. For 2002, there was a minor increase throughout the year with the exception of July, September, November, and December. However, even with these minor dips, the number of calls answered remained fairly constant over each of the two years. When pulling the RMS data for 2001 and 2002, data indicate that the average time spent on each call answered was twenty-one (21) minutes. This reflects and a two (2) minute decrease from 2000 (23 minutes). Time Worked and Available for Calls for Service Prior to the advent of community policing, police (patrol) managers were primarily concerned with how fast patrol units responded to calls for service (response time), the number of arrests made, and the number of traffic tickets or citations written. However, community policing required that time be spent by patrol officers in preventive (proactive) activities, working with the community, and “managing” their assigned patrol areas. This required that an assessment be made of how much time was being spent on traditional police activities and how much was available for community policing. 17
As a point of reference, the following chart (for 2002) shows the number of hours actually worked, the number of hours worked, but unavailable (court, training, prisoner processing, etc.), and the number of hours unavailable (e.g, personal leave, workers compensation, military leave, etc.). As the data indicate, for the majority of hours, patrol officers were available and on the street to answer calls for service and engage in community policing activities.
4000 3500 Number of Hours
Hours Worked
3000 2500
Hours Unavailable
2000 1500
Hours Worked, Not Available
1000 500
Ju l Se p N ov
Ja n M ar M ay
0
To fully explore the ability to target patrol officers’ time in order to fulfill community policing expectations, additional analysis is required. The charts that follow show the number of hours worked versus the number of hours unavailable, but not available for calls for service. The first chart indicates that for the year 2001, an average of 4000
Hours Worked vs. Hours Worked, but Unavailable for Calls
3500 3000 2500
Hours Worked 2001 Hours Unavailabl e
2000 1500 1000 500
D ec
Se p O ct N ov
Ju n Ju l Au g
0 Ja n Fe b M ar Ap r M ay
Number of Hours
18
twenty-five percent (25%) of the patrol officers’ time was unavailable for patrol activities. Hours worked ranged from a high of 3,554 hours (June) to a low of 2,545.5 hours (December), with the average available hours being 3,139 hours. In terms of hours worked, but unavailable for patrol activities, the range was a high of 1,302 (June) to a low of 228 hours (September), with an average being 782 hours. Hours Worked vs. Hours Worked, but Unavailable for Calls
4000 3500 3000 2500
Number of Hours
Hours Worked 2002
2000
Hours Unavailable
1500 1000 500
Ju l Au g Se p O ct N ov D ec
Ja n Fe b M ar Ap r M ay Ju n
0
For 2002, officers were unavailable to answer calls for service an average of 16% of the time. For hours worked, the range was a high of 3,405.5 for November, to a low of 2,604.5 hours for February. The average was 2,941 hours. In terms of hours unavailable, the range was a high of 803.5 hours (September) and a low of 220 hours for May, with the average being 482.5 hours. “Working, but unavailable” is still only part of the equation in determining how much time is available for community policing activities. Time spent “Out of Service” on calls for service or other dispatched activities must also be analyzed. This analysis was also conducted for the two tours of duty currently utilized, 10 pm – 8 am and 8 am – 10 pm. Data prior to September 2001(when three tours were utilized) had to be collapsed into the two tours. Due to the volume of data available, but not user-friendly, a decision was made to use data from the 15th of each month for the years 2001 and 2002.
19
For 2002, looking at the 15th of each month, a total of 37,440 hours were worked on the 10 pm – 8 am tour of duty. Of this 14,958 hours were spent “out of service” (40%). On the 8 am – 10 pm tour of duty, again for the 15th of each month, a total of 41,040 hours were worked and a total of 11,487 was spent “out of service” (28%). When averaged across the two tours, 34% of worked hours were spent “out of service.”
2002: Hours Available vs. Hours Out of Service
5000 4500 4000 3500 3000
Number of Hours
10 pm - 8 am Available 10 pm - 8 am OOS
2500 2000 1500 1000 500
D ec
Se p O ct N ov
Ju n Ju l Au g
Ja n Fe b M ar Ap r M ay
0
When looking at 2001, the following was revealed. On the 10 pm – 8 am tour of duty, 29% of the time worked was spent “out of service.” For the 8 am – 10 pm tour, 30% of the time worked was spent “out of service.” When combined, the total percentage of time spent unavailable for other activities was 30%.
20
2002: Hours Available vs. Hours Out of Service
5000 4500 4000 3500 3000
Number of Hours
2500
8 am - 10 pm Available 8 am - 10 pm OOS
2000 1500 1000 500 0 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
2001: Hours Available vs. Hours Out of Service
4000 3500 3000 2500 10 pm - 8 am Available Hours OOS
2000 1500 1000 500
Ju l Au g Se p O ct N ov D ec
0 Ja n Fe b M ar Ap r M ay Ju n
Number of Hours
21
2001: Hours Available vs. Hours Out of Service
4000 3500 3000 2500
Number of Hours
8 am - 10 pm Available Hours OOS
2000 1500 1000 500
D ec
Se p O ct N ov
Ju l Au g
Ja n Fe b M ar Ap r M ay Ju n
0
Officers Available Another aspect of determining how much time is available for community policing activities is the number of officers available. Again, there is no hard and fast rule for determining minimum and maximum (optimal) staffing levels. According to department officials, the optimum level of personnel on the street is six (6), the minimum is three (3). Again, when assessing time available for 2001 and 2002 using the 15th of each month, average staffing levels were also obtained.
Year/Tour
% Out of Service
Staffing Level
2002 2002 2001 2001
40 28 29 30
4.33 4.83 4.00 3.92
Although using the arbitrary date of the 15th of each month, the data indicate that the optimal staffing level was rarely achieved. Another point is that when taking into account the percentage of time “out-of-service,” “actual” staffing levels provide a more realistic 22
assessment of the number of officers on the street as opposed to the number of officers assigned at the beginning of a tour of duty. Response Time Most researchers/practitioners have agreed that no patrol beat should be so large that calls for service cannot be answered in three minutes. Although an analysis was not conducted by beat, the “Response Time Analysis” by offense type and shift in the RMS system was queried for the years 1999 through 2002. Records indicate that average response time is three (3) minutes. FINDINGS AND RECOMMENDATIONS In order to obtain a more clearer assessment of time available for preventive or community policing activities, times “out-of-service” and ‘time working, but unavailable” must be combined. For the year 2001, an average of 54% of patrol officers’ time was unavailable and “out of service.” For the year 2002, an average of 50% of officers’ time was unavailable and “out of service.” There is no recognized standard of how much time should be spent for various activities or time unavailable for providing response, however, in 1992, the International City/County Management Association recommended that no more than 35% of an officer’s time be spent responding to calls for service. However, ICMA’s recommendation does not take into account the amount of time “working, but unavailable.” Therefore, for purposes of this study, the two times are combined and then compared with the 35% recommendation. If the “standard” of 35% is utilized, then an additional 3.3 officers are required to bring the 50% (combined Out-of-Service and Working, but Unavailable) down to 35%. With a relief factor of 1.6 that allows for leave and days off, maintaining this level of staffing would require a total of 5.28 additional officers. However, if there is a decision that allows sergeants to act solely as sector commanders/supervisors and have the six (6) corporals totally assume street supervisory duties, if the corporals are backed out of the equation, a total of 14.88 officers would need to be added to patrol (6 officers to replace the six corporals, added to the 3.3 officers to reduce time to 35%, times the relief factor of 1.6). To reiterate comments from the beginning of this report, determining personnel strength of any police department finally comes down to managerial decisions based on available data and the needs of the community. In order to provide data on a protracted basis on which to make managerial decisions, the following is also recommended: • on-going analysis of data, over time, with a statistical software package designed to analyze vast amounts of disparate data;
23
• • • • • •
more detailed accounting of patrol activities and time spent that is recorded in both the RMS system and through use of a detailed officer “run sheet;” detailed analysis of the activities of support personnel as they relate to the patrol function; comprehensive accounting of all current and proposed patrol activities; temporal and spatial analysis of crime and other incidents; assessment of alternative response to crime and incidents; and the role of management and supervision within the Takoma Park Police Department.
24
Sources: Alpert, G. and R. Dunham (1988) Chaiken, Jan (1975) Chamlin, Mitchell (1996) Cordner, G. (1995) Eck, John and E. Maguire (2000)
Gaines, L., E. Kappeler, J. Vaughn Koper, Christopher et al (2001)
Loftin, Colin and D. McDowall (1982) Nalla, Mahesh K, M. Lynch (1997)
Roberg, R. and J. Kuykendall (1997) Wilson, O.W. (1963)
Policing Urban America. Prospect Heights, ILL, Waveland Press Patrol Allocation Methodology for Police Departments, Rand Corporation, California “The Police, Crime, and Economic Theory: A Replication and Extention,” American Journal of Criminal Justice, 20:165-182 “Community Policing: Elements and Effects,” Police Forum, 5(3):1-7. “Have Changes in Policing Reduced Violent Crime? An Assessment of the Evidence,” in The Crime Drop in America, edited by Alfred Blumstein and Joel Wallman, 207-265, New York, Cambridge University Press Policing in America, 2nd ed., Cincinatti, OH, Anderson Publishing “Hiring and Retention Issues in Police Agencies: Readings on the Determinants of Police Strength, Hiring, and Retention of Officers, and the Federal COPS Program,” Urban Institute, Washington, D.C. “The Police, Crime, and Economic Theory: An Assessment,” American Sociological Review 47:393-401. “Perspectives on the Growth of Police Bureaucracies, 1948-1984: An Examination of Three Explanations,” Policing and Society, 3:5161. Police Management, Los Angeles, Roxbury Press Police Administration, 2nd ed., New York, McGraw-Hill
25
Appendix A. Comparison of Municipal Police Department Staffing in the States of Maryland and Virginia (10,000 - 20,000 residents) $ $
National average of officer per 10,000 residents: 26.7/10,000 ( 2.6 per 1,000) Maryland average of officer per 10,000 residents: 28.7/10,000 (2.8 per 1,000) (Both national and Maryland numbers based on departments having 100 or more sworn personnel)
Maryland
(No distinction made between full-service and less-than-full-service departments)
City/Town
Population
# Officers
Ratio (per 1000 residents)
Comments
Aberdeen
14,000
39
2.9
more than Takoma Park
Bel Air
10,000
30
3.0
more than Takoma Park
Cambridge
12,500
44
3.5
more than Takoma Park
Crofton
10,000
6
.6
Easton
10,000
40
4.0
Elkton
13,000
29
2.2
Greenbelt
21,096
53
2.5
more than Takoma Park
Havre de Grace
10,000
24
2.4
more than Takoma Park
Hyattsville
15,000
29
1.9
Mt. Rainier
11,800
14
1.1
Ocean City
10,500
13
1.2
Ocean Pines
10,500
13
1.2
Princess Anne
15,000
14
.9
more than Takoma Park
26
7
Sparrow’s Point
15,000
12
.8
Takoma Park7
18,600
42
2.3
These numbers are listed in the cited report and do not reflect the current police strength of 41 or the current population of 17,333.
27
Virginia
(No distinction made between full-service and less-than-full service departments)
Bristol
18,281
50
2.7
City Sheriff (see below)
Bristol
18,281
60
3.2
Police Dept. (see above)
Christianburg
16,000
25
1.5
Colonial Heights
17,500
7
.04
City Sheriff (see below)
Colonial Heights
17,000
50
2.9
Police Dept. (see above)
Culpepper
10,600
30
2.8
Falls Church
10,000
21
2.1
City Sheriff (see below)
Falls Church
9,900
32
3.2
Police Dept. (see above)
Fredericksburg
20,700
9
.4
City Sheriff (see below)
Fredericksburg
24,900
85
3.4
Police Dept. (see above)
Front Royal
12,869
34
2.6
Herndon
17,732
50
2.8
Martinsville
16,000
48
3.0
City Sheriff (see below)
Martinsville
16,162
54
3.3
Police Dept. (see above)
Poquoson
11,300
17
1.5
Radford
15,900
26
1.8
City Sheriff (see below)
Radford
13,914
26
1.8
Police Dept. (see above)
Williamsburg
55,000
13
.2
City Sheriff (see below)
Williamsburg
11,500
29
2.5
Police Dept. (see above)
Source: National Directory of Law Enforcement Administrators...., National Public Safety Information Bureau (2002) Highlighted areas have a higher officer-to-resident ratio than Takoma Park. 28
Appendix B Comparative Analysis Approach: San Jose Police Department In 1993, the San Jose Police Department had approximately 16 officers per 10,000 residents (1,312 officers). San Jose could be compared to either one or a group of police departments to determine the differences between the San Jose mean ratio of officers to residents with that of the comparison department or group. The difference is multiplied by 82,which is the approximate number of units of 10,000 population (820,000) in San Jose in 1993. Six possible comparisons are provided. 1.
Using the national average (1993) of approximately 22 officers per 10,000 residents (for local police departments with more than 100 officers), San Jose should have had 492 more officers.
2.
Other than San Jose, there are four cities in California with 1,000 or more officers. This group has a mean officer to-resident ratio of 18.2 per 10,000. Based on this comparison, San Jose should have had 180 more officers.
3.
Of municipalities with 100 or more officers, in 1993 Atlantic City, New Jersey, had the highest officer-to-resident ratio at 112:10,000. Compared with this city, San Jose should have had 7,872 more officers.
4.
Compared with Washington, D.C., the city with more than 1,000 officers that has the highest officer-to-resident ratio at 70:10,000, San Jose should have had 4,420 more officers.
5.
Compared with Indianapolis, the city with more than 1,000 officers that had the lowest officer-to-resident ratio at 13:10,000, San Jose had 246 too many police officers.
6.
Compared with Anchorage, Alaska, the city with more than 100 officers that had the lowest officer-to-resident ratio at 11:10,000, San Jose had 410 too many officers.
Source:
Roberg, R. and J. Kuykendall (1997), Police Management, Los Angeles, Roxbury Press. 29
