Toronto Police Accountability Coalition


Paired Up: Fact, Fiction and the Politics of Two-Officer Police Patrols


September 19 2017 - Chris Armstrong

Paired Up: Fact, Fiction and the Politics of Two-Officer Police Patrols

Christopher J. Williams
(Toronto Police Accountability Coalition)

I. Introduction

This unit is operating under protest. If the City of New York needs a martyr, let it be this unit  so stated an NYPD officer in a message to the central dispatcher as he patrolled Manhattan in the late 1970s. His invocation of potential martyrdom, linked to his insistence that this unit will handle any and all jobs, was a form of personal protest in response to a pilot project devoted to determining the feasibility of one-officer patrols. It was an anomalous situation, to be sure, yet it was nonetheless indicative of the more-heat-than-light character of certain discussions, debates and disquisitions about one-officer vs. two-officer patrols, particularly as these options apply to evening/night shifts. Modes of argumentation involving logic and emotion, discourses of danger and emphases on efficiency, quantitative data and qualitative accounts, and so forth, are such that no single perspective prevails above all others. As a practical consequence, therefore, some police organizations have requirements for two-officer patrols during evenings and nights whereas others do not.

Among police organizations that do have such requirements, collective bargaining agreements are the typical means by which evening/night two-officer patrols are articulated and sustained. In Toronto, for example, the collective agreement between the Toronto Police Association and the Toronto Police Services Board states, in section 22.01, that all uniform patrol cars, except those assigned to traffic duties, shall be manned by two fully trained and armed police officers while on patrol between the hours of 4:00 p.m. one day and 4:00 a.m. the following day or during such other continuous period of twelve hours per day as shall be designated by the Board to coincide with the period of peak patrol activity.

In other jurisdictions the universality of this approach  as indicated by the reference to all uniform patrol cars  is absent and, instead, a hybrid approach prevails whereby a specified complement of two-officer units conduct patrols during evening/night shifts featuring a majority of one-officer units. Section 08-02 of the collective agreement in Windsor furnishes an example of this stipulation: During the hours of 5:00 p.m. and 7:00 a.m. no one-officer units shall be deployed until&eleven (11) two-officer units have been deployed, providing there are available personnel on duty. Embedded within this requirement is a recognition that some (or most) situations can be ably handled by officers riding solo, a topic to which we will return.

Crucially and interestingly, sharp disagreements between police associations and police boards can arise regarding evening/night patrol standards even when the standards are part and parcel of a collective agreement. In Durham Region in 2007, for instance, arbitration hearings were conducted pursuant to a dispute between the Durham Regional Police Association and the Regional Municipality of Durham Police Services Board. The core issue pertained to whether a particular section of the collective agreement conflicted with powers conferred onto police chiefs by the Police Services Act. More specifically, section 19.01 of the agreement states that units deployed for two-Member uniform patrol function will be staffed, at minimum, by one fully qualified Member and one fully trained Member who has completed a minimum of eight (8) shifts with a Qualified Coach Officer. A minimum of nineteen (19) such units will be deployed from the Night Shift complement between the hours of 2000 hours and 0400 hours. Like Windsor, this is a hybrid approach with a majority of night units (54 out of 73) being one-officer units. Turning to the Police Services Act, section 41(1) states, in part, that the duties of a chief of police include...administering the police force and overseeing its operation. So what issue lay at the core of the dispute? In essence, the Board and the Chief argued that the 8:00pm to 4:00am two-officer unit stipulation ran contrary to the operational autonomy of the Chief.

In her summary of the arguments advanced by the Chief, arbitrator Paula Knopf noted that the Chief viewed two-officer units as an unnecessary carry-over from previous eras; he also cited the negative resource implications of such units while contending that viable alternatives were readily available. To wit:

The Chief objects to the contractual provisions relating to the requirement of two officer patrol units during the night shift. It was conceded that two officer units have often been designated to attend to particularly difficult policing situations, for both the safety of the officer as well as the safety of the community. However, it was stressed that this arrangement is not the only alternative available to address the situation. The Chief argues that unnecessary two officer patrol units can have a significant impact on the operations when they tie up valuable resources.

And, with additional specificity, Knopf mentions deployment options, as put forth by the Chief, that could serve as functional substitutes for two-officer units:

It was recognized that the requirements regarding two officer units may have been implemented in the past with some safety concerns in mind. However, it was suggested that any safety concerns could better be addressed by deploying two constables in separate vehicles, or by invoking the tactical support teams to situations that require the presence of more than one officer.

Perhaps notably, the word safety appears four times in these two passages, an indication of the extent to which notions of occupational danger can heavily influence decisions to introduce or maintain two-officer evening/night units. But putting aside the popular intuitive assumption that two officers must be better than one, is it indeed the case that one-officer patrols create more dangerous situations for officers (and possibly civilians) than two-officer patrols? Formulating an answer to this question is the task to which we now turn.

II. Comparative Dangerousness

In a 1992 article in the Journal of Criminal Justice, the authors concluded, following a review of multiple studies on the relative risks linked with one-officer and two-officer units, that the results indicate no special danger associated with single-officer patrols&[and] call into question the widely espoused position regarding the greater danger associated with single-officer patrol. Nonetheless, 15 years later a study of the Long Beach Police Department, conducted by the City of Long Beach, advanced the claim that officers on solo patrols were getting shot with disturbing regularity due to their use of attention-consuming onboard computers:

Recent upgrades to the Mobile Data Computers (MDC's) in the police cars has dramatically increased the ability of the Officer in the field to access files and databases critical to effectiveness and safety. They also have made it possible for Officers to write their reports in the Patrol car, and, in some cases, file them from the field. There are several problems with the current system, which make report writing in the car problematic. With a preponderance of one-officer cars, officer safety becomes a valid concern. An Officer with his or her attention focused on the computer screen to type reports is vulnerable to a variety of threats, and the LBPD has had several Officers shot over the years, while distracted inside the car and not paying attention to their surroundings.

Several shootings of distracted and unsuspecting officers would, presumably, have resulted in some of them getting killed, yet in the quarter century leading up to the publication of the report (1982-2007) three police officers had been killed, none of whom were shot in the course of functioning as one-officer units. One officer (Karl Simons) was killed in a highway accident, another officer (Edward Davenport) was killed due to a fall from a ladder in a pistol range, and a third officer (Daryle Black) was shot dead  while in a patrol car with his partner. Taking this into consideration, readers might draw their own conclusions about the veracity of the assertion that one-officer units in Long Beach are especially prone to being targets of gunfire.

In fact, contrary to standard propositions pertaining to this topic, some observers and analysts argue that two-officer patrols are at greater risk of having their safety compromised than their solo counterparts. Standing in support of one-officer patrols, a 1990 New York Times editorial articulated this perspective:

The city's strong police union argues that one officer cars are unsafe, noting that during an experiment with solo cars in the late 1970s, an officer in one of those cars was killed. But officers riding together are killed, too. In fact, there is some evidence that two officer cars could be more dangerous because officers working together may develop a false sense of security and lose alertness.

Gavin de Becker, a renowned security expert and Senior Fellow at the UCLA School of Public Affairs, observes that although two-officer units are overwhelming preferred by officers, such a preference is not necessarily in accord with the aim of self-preservation. Holding the view that machismo is a vice and cautiousness is a virtue, he declares that all cops want two-man cars. You have a buddy, someone to talk to. But one-man cars get into less trouble because you reduce bravado. A cop by himself makes an approach that is totally different. He is not as prone to ambush. He doesn't charge in. He says, Im going to wait for the other cops to arrive. He acts more kindly. He allows more time.

Indirect acknowledgements of de Beckers perspective can occasionally be gleaned from official documents such as a 2014 report on Houston Police Department staffing produced by the Police Executive Research Forum. The Policy Preferences section of the report includes this observation: There is a preference inside the department to have more patrol officers available in two-officer cars. This is thought to enhance productivity by increasing the level of self-initiated activity and enhancing officer safety. Forms of self-initiated activity that are aggressive and intrusive qualify not only as manifestations of bravado, but also as modes of policing that can be antithetical to the aim of bolstering safety, a point that would presumably be appreciated by police officers who themselves are inclined to say any encounter could be a deadly encounter. If two-officer units and proactive policing are linked such that more of the former leads to more of the latter  as the report suggests  then the antagonistic (and often avoidable) conflicts that may arise carry physical risks for all involved parties.

Shifting to figures that disaggregate killings of police officers based on whether the deceased were patrolling alone or with partners, a Statistics Canada study reveals that 133 police officers were killed on duty from 1961 to 2009. Of these killings, a slight majority occurred in the context of two-officer patrols: There has been some debate within the policing community surrounding the issue of one-person versus two-person vehicle patrols. Of those officers killed while on vehicle patrol, 54% were assigned to two-officer vehicles and 46% to one-officer vehicles. One question bedevilling information along these lines relates to the matter patrol mode prevalence: during the time period under examination  in this case 48 years  what percentage of patrols involved one officer? In the absence of an answer one cannot definitively declare that one-officer patrols are overrepresented (more vulnerable) or underrepresented (less vulnerable) among patrol modes. The same limitation applies to an FBI report about officers feloniously killed which states that of the 46 police officers killed while on duty in 2014, 16 (34%) were alone and unassisted at the times of the attacks. Nonetheless, both sets of numbers would seem to suggest that one-officer patrols are perhaps no more dangerous than two-officer patrols, and that the two other possibilities  equally dangerous or less dangerous  are more likely.

III. Differentiated Duties and the Logic of One-Officer Patrols

In the opening paragraphs of this piece it was noted, in connection with the Windsor Police Service, that hybrid approaches to evening/night patrols are founded on recognitions that a variety of duties can be suitably assigned to one-officer patrols. Among the range of duty-related categorization schemes that have been devised, one of the most useful is based on two factors: (1) the importance/seriousness of incidents in need of a police response, and (2) the appropriate number of responding officers. Specifically, the authors of a one-officer feasibility study in New York offer the following classificatory details:

We were able to group incidents into four aggregate categories  H2, H1, L2, L1. H jobs have the highest dispatch priorities and L jobs have the lowest while the number following H or L indicates the number of one-officer cars to be initially dispatched. H2 jobs are, as expected, the most severe incidents and include many types of crimes in progress such as robberies, burglaries, and assaults as well as reports of shots fired and calls to assist a police officer. H1 jobs are typically ambulance cases and residential incidents for which there is less potential danger to responding officers. L2 is the smallest category, including only pickups of emotionally disturbed persons. The L1 category includes incidents such as past burglaries and larcenies, reports of a disorderly person or noise, and street accidents.

If, as sensationalist media coverage would suggest, the vast majority of jobs are H2 jobs then an ironclad case could be made for requirements that all evening/night patrols be conducted by two-officer vehicles. But such is not the case because most police work deals with incidents that are closer to petty than profound, as the article confirms in the case of New York City: For L1s, which are typically incidents such as a report of a past burglary or a noisy party, one officer seemed to be quite adequate. The existence of large numbers of these LI jobs was, of course, the main motivation for considering a one-officer program in the first place.

Candid and nuanced statements concerning sensible and defensible approaches to officer deployment are more likely to be issued by relatively detached academic researchers than police officials, but the latter have been known to be straightforward in supporting forms of patrol that are in line with actual  rather than quasi-mythical  features of police work. Several years ago, for example, a Washington D.C. police information officer offered this take on why a blend of units, both solo and paired, was appropriate:

Most of the patrol cars in Washington by far carry one man now. We agree that in most cases the one-man unit is more efficient because the wide majority of police work is in nonemergency type calls. But it is those emergency calls that we have to be concerned about, and on those we want to be able to send two men. We think it's best to have a mixture of (one-man and two-man) units.

In Canada, as well, the bulk of police duties qualify as jobs that fall under the L1 rubric. An examination of Statistics Canada figures in the publication Police-Reported Crime Statistics in Canada, 2015 shows that for every firearm incident (use of, discharge, pointing) there were 45 incidents of disturbing the peace; for every incident of aggravated assault there were 84 incidents of mischief, and so forth. Patterns of this sort are the basis for an old joke about a field researcher who accompanies officers on patrol, only to be disappointed by the mundane nature of the experience and then told you should have come along last Thursday  there was a lot of action. The humour is derived, of course, from the implausibility of the claim that the serious action always seems to take place when researchers are absent. Levity aside, L1 jobs predominate in both Canada and the US, a state of affairs which has rather obvious implications for decisions about suitable forms of patrol during evenings and nights.

IV. The Optional Character of Two Officer Patrols

In Toronto, as noted in the opening paragraphs of this study, evening/night two-officer patrols are required based on collective agreement stipulations established following a 1974 arbitration award; having been in place for over four decades, this requirement is seen in some circles as necessary, unquestionable and wholly preferable to the alternative, namely, one-officer patrols. But, of course, this limiting view obscures the fact of alternatives  plural  the likes of which once prevailed in Toronto prior to 1974:

Before the 1974 award, there was a substantial number of two-officer patrols operating in the Metropolitan area. The evidence is that these patrols were assigned to areas where violent crime or confrontation with police was more common, or where a back-up car would be slow arriving because of distances. For the most part, these cars were assigned during the period of peak activity, and they were assigned for the primary purpose of safety, to ensure a quick response by enough officers to deal with the problem. In 1969 Chief Adamson (then Deputy Chief) reported&that 59 out of 160 cars on the 7 p.m. to 3 a.m. shift were manned by two officers.

This hybrid approach, with just over one-third of the patrols as two-officer patrols, has life in present-day police organizations who, again, recognize that differentiated duties allow for the utilization of different patrol modes. The result is that in numerous jurisdictions throughout North America two-officer patrols are considered optional, not mandatory, during evenings and nights.

Pursuant to the aim of getting a sense of how various police services orient themselves vis-à-vis the question of two-officer patrols, we sent e-mail messages to 21 police organizations in Canada and the United States. Here are the questions we posed, with Toronto as the key reference point:

In connection with a 1970s Toronto arbitration decision, all evening/night police patrols are required to have two officers in each vehicle. Questions: (a) Does your department have a similar evening/night two officer patrol requirement? (b) If yes, is this rooted in management practice or is it legislated? (c) With respect to the practice of your department (two officer patrol requirement or no requirement) have any assessments been conducted regarding monetary costs or savings?

Given that only seven organizations responded and supplied answers to our questions we are not able to tout the virtues of a representative sample, needless to say. Still, the information featured in the following table is at least suggestive of the extent to which Torontos two-officer stipulation does not hold in other cities.

Table 1: Evening/Night Two Officer Patrols
Police Department
Two Officer Requirement?
Basis for Requirement
Assessment of Costs or Savings?
Collective Agreement
No, but a workload assessment  including a consideration of two officer patrols  is in the early stages.
Collective Agreement
Yes, for some night patrols.
Collective Agreement
No Response.

As Table 1 demonstrates, two of the responding police services have a universal two-officer patrol requirement, and one  Windsor  has a partial two-officer requirement which, as previously discussed, is based on a sequential rollout system whereby the evening/night shift (5:00pm to 7:00am) begins with eleven two-officer units going on the road followed by an unspecified complement of one-officer units. In all instances these stipulations are embedded within collective agreements, an unsurprising fact in light of the wariness that police unions (the primary proponents of mandatory two-officer units) often have towards police management; insofar as the latter sometimes hold the view that deployment decisions should be the exclusive prerogative of police chiefs (as argued in Durham Region), the former predictably insist on the assurances provided by collective agreements.

As for the question of whether these organizations have undertaken studies of costs or savings associated with their deployment standards, a spirit of incuriosity seems to prevail: none of the seven have done so, though one respondent made the point that, with respect to one-officer units, the costs and savings might be mutually offsetting: Obviously there are incremental costs associated with having additional cruisers. However this can be offset by the fact that many calls only require one officer to attend as opposed to two (thereby gaining a labour savings).

This reiteration of the proposition that during evenings and nights an abundance of duties can be handled by one-officer units runs against the safety-oriented claims of interested parties (e.g. the Toronto Police Association) who contend that the opposite is true. The inclusion of the Houston Police Department (HPD) in our study is notable in relation to these claims and counterclaims because the HPD, unlike the Toronto Police Service, has no two-officer unit stipulation. Why is this notable? First, because Houston is a much more dangerous city than Toronto to the extent that homicides are a suitable indicator of serious violence in general. Specifically, Houston had a homicide rate of 13.1 per 100,000 residents in 2016 (302 homicides with a population of 2.3 million) while Toronto, in striking contrast, had a rate of 2.6 for the same year (74 homicides with a population of 2.8 million). Second, HPD officers do not seem to face intensified risks as a result of engaging in solo patrols during evenings/nights. Using one measure, officer deaths due to gunfire from 1997 to the present, we find that four officers were killed while patrolling alone; three were killed during morning hours (8:30am, 9:00am, and 9:45am) and one was killed at 5:30pm. And if we shift our attention to a Canadian city with no two-officer requirement, Calgary, details on police deaths show that 1992 was the last time an officer was killed while on solo patrol. On the basis of these facts, therefore, rhetoric centred on images of one-officer units as easy prey or sitting ducks seems to be more hyperbolic than empirically supportable.

V. Concluding Remarks

The variability of police organizations regarding the question of whether all evening/night patrols should feature two-officer units leads, obviously, to one overarching conclusion: universal two-officer units are a choice, not a necessity. The choice was made in Toronto due, in large measure, to police union alarm-sounding about officer safety but the solidity of this position is undermined by some of the information we have highlighted, as well as the lack of police union enthusiasm for measures that would likely bolster officer safety. Mandatory annual fitness testing, random drug testing of officers, stiffer departmental penalties for officers caught driving drunk  all of these, if adopted together, would make police officers in Toronto (and elsewhere) less susceptible to injury or death. But many police unions  including the Toronto Police Association  view these measures as either undesirable or downright odious, irrespective of their safety-enhancing merits.

Outside of Toronto, however, police thinking tends to be less rigid and more in tune with the capabilities of one-officer units. Allowances for good judgement are also operative because, as some organizations contend, the absence of a two-officer unit requirement does not mean officers are all alone at night, isolated and friendless. As one respondent stated, We have no requirement for two officer patrols over night. With that said, it is often encouraged and single officers will typically team up to form a partnership on night shifts whenever possible. Another respondent noted that in his department two-officer unit deployments are linked to situational factors: Yes, there are occasions that we have two officer patrols however our decision to utilize this practice is predicated on operational or service oriented needs. These statements evince a form of calm rationality that stands in opposition to the longstanding alarmism of those who insist, against good evidence, that two-officer units are the only suitable patrol mode during evenings and nights.

Resource Issue(s):