Ar

Assigned Reading: Thistlethwaite & Wooldredge,  Part 1 Chapter 3-Patrol staffing which is better: One- or Two-officer patrol units? pp. 55-60.
After reading the article answer the following questions in a discussion posting:
You must also comment on two of your classmate’s postings by the end of the week.
1. Do you think the methodology used to complete this study was valid?
2. What limitations do you see, if any, in the way the study was conducted?
3. Based on your knowledge and experience do you think given all of the technological advancements that officers have today this study’s findings are still valid? Why or Why not?
PatRol staffing: Which is BetteR, one-oR tWo-officeR PatRol Units?
Boydstun, J., M. Sherry, and N. Moelter (1977). Patrol Staffing in San Diego: One-or Two-Officer Units. Washington, DC: Police Foundation.
Background Two-officer patrol units became standard in the United States when the automobile replaced foot
patrol. Foot patrol officers walked their beats alone, whereas departments assigned two officers to a car for safety reasons. Officers in cars patrolled larger areas and were spending more time engaged in law enforcement activities perceived as more dangerous (Boydstun et al. 1977). In 1953, the Kansas City (Missouri) Police Department became the first Metropolitan police depart-ment to assign a single police officer to vehicle patrol (Brannon 1956). One-officer patrol was
believed to be more efficient. With each officer having their own vehicle, officers could cover more area, respond quicker, and increase the visible presence of the police in the community. Several other departments across the country adopted one-officer patrol based on these same beliefs. Departments resisting the change held firm to the traditional view that two-officer units were safer. Advocates of the two-officer units also argued that they were more effective in fighting crime. They believed that two-officer units were more proactive in their investigations of suspi-cious activities compared to one-officer units (Boydstun et al. 1977). None of these arguments had ever been empirically tested and the debate over the use of one-versus two-officer units continued for the next 20 years. The San Diego Police Department assigned two officers to each patrol car up until 1954.
Following the example set by the Kansas City Police Department, San Diego gradually imple-mented a one-officer patrol system over the next few years. The civil unrest of the early 1960s caused administrators in the San Diego Police Department to rethink their position on one-officer units (Boydstun 1977). Concern over officer safety resulted in two-officer units assigned to areas of the city perceived as high risk. Beginning in 1964, the department started to use a combination of one-and two-officer units. In response to the energy crisis of 1974, the San Diego Police Department started working on a patrol allocation plan that would help save on fuel costs without increasing personnel costs (Boydstun 1977). The obvious solution seemed to be to revert back to an entirely two-officer patrol force. This would have considerably reduced fuel costs (assuming the number of vehicles on the street was reduced) while keeping the same number of officers on the street. Many officers and administrators perceived two-officer units to be inef-ficient and resisted the proposal (Boydstun 1977). The department had no empirical evidence to help guide this decision until the Police Foundation contacted them. The Foundation was inter-ested in conducting a comprehensive analysis of one-and two-officer patrol units and wanted the study to take place in San Diego as a joint effort with the police department. The objective of the study was to determine the most effective, efficient, and safe allocation of patrol resources by comparing the use of one-and two-officer patrol units. The planning phase of the project lasted over a year. Because this was the first major study of
patrol staffing, the San Diego Police Department and the Police Foundation conducted a nation-wide survey of police departments to determine their use and justifications of one-and two-officer units. Surveys were mailed to 65 police departments serving populations from 36,000 to 8 million. Fifty-two surveys were returned (80 percent response rate). The information obtained from the surveys was used to develop the research design for the study of patrol staffing in San Diego.
the experiment
A team of researchers led by John Boydstun (1977) used a quasi-experimental research design to determine the effects of one-officer and two-officer patrol units on several patrol operations in San Diego, California. In 1977, the city had a population of approximately 697,000 residents with a police force of 1,077. Like most departments, patrol operations were divided into three, 8-hour shifts. Seventy-one percent of officer assignments were one-officer units and the remaining were two-officer units. Two-officer units were primarily used in areas designed as “high risk.” The department utilized a computer-assisted dispatch system for assigning calls for service. To control for preexisting differences across patrol areas, a sample of 16 police beats with a total of 44 patrol units were selected. Beats differed according to police officers’ perceptions of risk. This allowed researchers to assess differences in outcomes in areas the police perceived as safe compared to areas perceived as higher risk. The beats were then divided into four groups based on the existing type of patrol. The first area involved no change in officer assignment. One-officer units were assigned in previously one-officer areas (first control group). In the second area, one-officer units were assigned in areas where there were previously two-officer units (first experimental group). In the third area, two-officer units were assigned in areas that were previously one-officer units (second experimental group). The final area consisted of two-officer units in previously two-officer areas (second control group). Dividing the beats into these groups allowed researchers to compare one-officer units with two-officer units across all study areas. Beats were assigned to one of the four groups according to socio-economic and demographic variables in order to control for population differences between areas. A total of 306 police officers patrolled the study areas. The experiment officially began on October 26, 1975, and lasted an entire year. The start
date itself was randomly determined. Much of the data, however, were derived from 12, 7-day study periods selected throughout the year. Four weeks separated each study period so that researchers were able to collect data at different time periods all throughout the year without collecting an unmanageable amount of data. Researchers used several sources of archival data, including dispatch and departmental records, crime statistics, and officer journals. Four outcome measures were examined: patrol unit performance, patrol unit efficiency, patrol officer safety, and officer attitudes and perceptions.
Results
A substantial amount of data was collected for this experiment despite efforts to reduce the amount of data. Findings for each of the outcome measures are summarized below.
PatRol Unit PeRfoRmance Data from dispatch records, officer journals, and arrest and complaint reports were used to determine the impact of one-versus two-officer units on patrol performance. Several performance indicators were examined, including the number of calls for service, number of arrests, traffic warnings, field interrogations, open business checks, citizen contacts, and total miles patrolled. No significant differences were found between the one-officer and the two-officer patrol units, with the exception that calls for service assigned to one-officer units resulted in more arrests and formal reports, whereas two-officer units produced more traffic citations. Findings were consistent across shifts, beats, and previous patrol staffing areas. Researchers also examined the number of citizen complaint reports filed against officers and found that two-officer units received more complaints. In a more thorough analysis, researchers separated out two-officer ambulance units from regular two-officer patrol units because of the differences in activities between the two types of units. This analysis refuted the differences in number of arrests, formal reports, traffic citations, and citizen complaints.
PatRol Unit efficiency Researchers examined dispatch and department records as well as police officer activity reports to determine if there were any differences in the use of police officer backup, response time, officer initiated activities, and cost-effectiveness between one-officer and two-officer patrol units. One-officer units requested backup more frequently than two-officer units, but only on the second shift (3 p.m. to 11 p.m.). No differences were found overall in response time; however, it was revealed that there was a delay in dispatch time for one-officer units. This finding was attributed to two-officer units receiving more high urgency calls. Two-officer units were more efficient in their responses (spending less time servicing calls for assistance), and had more time (almost one hour more per shift) available for officer initi-ated activities. This finding suggested that two-officer units were more efficient in their response and activities, but it is also important that there were no differences found for the performance indicators. To determine the patrol unit cost, researchers considered the cost of servicing calls for service and the cost of officer initiated activities. One-officer units were more cost-effective except for calls where two-officer units were recommended (40 percent of all calls).
PatRol officeR safety Crimes, vehicle accidents, and injury reports were used to assess patrol officer safety between one-and two-officer units. Two-officer units were involved in more resisting arrest confrontations and critical incidents, but were not more likely to be assaulted or injured. No differences were found in the number of vehicle accidents.
Police officeR attitUDes Surveys were administered to the entire central patrol division (390 officers). As mentioned previously, surveys are ideal for measuring individuals’ attitudes and perceptions. Researchers included both closed-and open-ended questions. A closed-ended question provides a list of items from which the respondent is instructed to select. Open-ended questions leave a blank space for the respondent to write a response. Half of the police officers responded that unit staffing was important or very important to police work. Officers who had previously worked two-officer units were more likely to indicate that unit staffing was impor-tant. A little more than half (52.4 percent) of all officers working in the department indicated a personal preference for two-officer units, while 74.6 percent of the officers participating in the study expressed a personal preference for two-officer units. Survey results further revealed that younger officers with lower ranks were more likely to prefer the two-officer units. Researchers also wanted to determine if personal preference depended upon such factors as salary implica-tions, partner, shift, and beat assignments. Fewer officers indicated a preference for two-officer units with these considerations. Questions regarding perceptions of performance advantage and unit efficiency revealed that officers only slightly favored the two-officer units. Police officers were neutral in their perceptions that two-officer units were safer, with a few exceptions. Officers assigned to one-officer units were more likely to perceive two-officer units as safer, as did officers of a lower rank with more patrol experience. No significant differences between one-officer and two-officer units were found in the questions measuring officer morale and job satisfaction. Officers indicated that two-officer units were less likely to use excessive force against a suspect. Finally, researchers discovered that officers with experience working two-officer units in beats perceived as high risk expressed the strongest preference for two-officer units. Researchers concluded that the data did not support the continued use of two-officer patrol
units in San Diego. One-officer units performed just as well if not better than two-officer units on most of the outcomes examined. One-officer units also allowed the department to save money. According to the study, 18 one-officer units cost less than 10 two-officer units. The only excep-tion to this recommendation was a suggestion to continue the practice of assigning two-officers to the police ambulance units where the research findings revealed that two officers were needed.
limitations
There was one major obstacle that had to be addressed by researchers in the San Diego study. A key feature of an experimental research design is the ability of the researcher to exercise con-trol over all of the experimental conditions. This would have required the random assignment of police officers to the study units to control for police officer characteristics that might be related to the outcome measures. Random assignment was not feasible because it would have disrupted existing staffing patterns and personnel policies (i.e., the department allowed officers to pick their beats based on seniority). The inability to randomize officer assignment brings into ques-tion whether officer characteristics influenced the results of the study. To help address this issue, the researchers devised a hypothetical staffing assignment (which would have been used if the existing system of unit assignments had been suspended during the study period) and compared it to the actual assignments. No significant differences were found, leading researchers to con-clude that individual officer characteristics would not have biased the results. Because of the lack of random assignment the study was offered as a quasi-experiment. Boydstun et al. measured police officers’ attitudes and perceptions of both one-and two-officer units, but their research did not address the attitudes and perceptions held by police administrators who were responsible for making decisions on patrol staffing. A study by Richard Larson and Thomas Rich (1985) found that police administrators believed one-officer units were safer because officers were more alert and used better judgment in their decision-making. Another drawback to the San Diego study had to do with the data collected on officer injuries. Boydstun et al. reported only on the number of injuries, not the type or severity of the injuries (Wilson and Brewer 1992).
significance and subsequent Research
For several decades, the San Diego Police Department had been staffing patrol units according to traditional assumptions of the advantages and disadvantages of one-and two-officer units. Department administrators were faced with the pressure of allocating their patrol resources in the most effective and efficient manner possible without compromising the safety of their offi-cers. With assistance from the Police Foundation, the department participated in a study that would better inform their decision-making. The experiment involved a sophisticated research design that allowed researchers to control for type of area, shift differences, and former patrol staffing and assignments. Results challenged the belief that two-officer units were better and safer for the officers. The findings provided the department (as well as others around the country) the justifications needed to employ one-officer units, even though this would counter the personal preferences of many of their own police officers. Police officer unions dictated patrol staffing in some departments. Changing police practice on the basis of a single study can be problematic if the results cannot be replicated. Replication is important in establishing the external valid-ity of experimental findings. The extent to which results from one experiment can be repeated with other experimental studies strengthens the generalizability of the findings. Results from replication studies of patrol staffing have produced findings consistent with the San Diego study. Edward Kaplan (1979) offered a model-based evaluation of one-versus two-officer patrol units. He used data from the San Diego study to demonstrate some of his mathematical models. Several performance measures were examined, and his results favored the use of one-officer units. One-officer units covered more area, had faster response times, and increased patrol frequency and visibility, all while reducing costs and without increasing the risk of injury. Using data from the Kansas City Response Time study (discussed above), David Kessler
(1985) replicated the San Diego study. In his analysis, he included additional control variables in an effort to explain one of the findings from San Diego: that response time for two one-officer cars was faster than one two-officer car. This finding could have been attributed to the type of call patrol units were dispatched to, whether or not a patrol unit was patrolling outside of their assigned beat, whether or not the officer(s) were inside their cars when dispatched, or whether or not the patrol unit waited for a “cover” car. Even with the additional control variables the response time for two one-officer cars was significantly faster. The replication studies discussed so far involved an examination of the impact of patrol
staffing on police-related outcome measures. In 1982, Scott Decker and Allen Wagner expanded this research by exploring the influence of patrol staffing on citizen outcomes. The data for their study were from a medium-sized city in the United States. They found that in situations where citizens filed a complaint against a police officer, more citizens were injured as a result of the police–citizen encounters compared to police officers. The chances of injury were higher when citizens encountered a two-officer unit than for a one-officer unit. Citizens were also more likely to be arrested and charged with assaulting a police officer as a result of their encounters with two-officer units. Decker and Wagner’s findings were significant regardless of how dangerous the area was (measured by the violent crime rate). Patrol staffing appeared to have significant implica-tions for citizens as well as for the police. It was also suggested that two-officer units approached their encounters with citizens differently. A study of over 1,000 Australian police officers found that two-officer units were more likely to encounter citizens who failed to comply with the offi-cers’ requests/commands (Wilson and Brewer 2001). Results from the San Diego Patrol Staffing Experiment revealed no significant differences in patrol unit performance measures. Research conducted in the United Kingdom suggests that the type of crime police respond to might influ-ence the effectiveness of two-officer patrols. Two officer units were found to be more successful in making an arrest for burglary (Blake and Coupe 2001). Recent research has shown that perceptions have not changed much since results from
the San Diego experiment were published. Alejandro del Carmen and Lori Guevara (2003) pub-lished a study of police officer attitudes related to one-and two-officer patrol units. Surveys were administered to 50 police officers in a North Texas police department. The officers’ responses indicated that their perceptions of performance did not depend on whether they were working as part of a one-or two-officer patrol unit. Officers expressed a preference for two-officer units in certain circumstances. Officers believed that two-officer units should be used in the evening and at night as well as in communities where a significant number of residents were suspicious of the police. The perception that two-officer units were more effective was prevalent also. A major-ity of the officers indicated that two-officer units could observe more, respond faster to calls for service, increase police visibility, and deter more crime than one-officer units. With regard to perceptions of safety, however, most officers believed that one-officer units were just as safe as two-officer units (provided the officers were properly trained). Today, the use of one-officer patrol dominates small-to medium-size police departments
while larger departments continue to use a combination of one-and two-officer patrol units. The San Diego study of patrol staffing, along with the replication studies, provided police depart-ments with the knowledge to make more informed decisions. Police departments no longer allocate resources based on traditional assumptions of how to achieve the most effective, cost-efficient use of their patrol resources.

References
Blake, L., and R. Coupe (2001). “The Impact of Single and Two-Officer Patrols on Catching Burglars in the Act.” British Journal of Criminology 41:381–96.
Brannon, B. (1956). “A Report on One-Man Police Patrol Cars in Kansas City, Missouri.” Journal of Criminal Law, Criminology & Police Science (July–August):238–52.
Decker, S., and A. Wagner (1982). “The Impact of Patrol Staffing on Police-Citizen Injuries and Dispositions.” Journal of Criminal Justice 10:375–82.
del Carmen, A., and L. Guevara (2003). “Police Officers on Two-Officer Units: A Study of Attitudinal Responses