The Force Continuum: Prevalence and Characteristics of Police Use of Coercive Force

1.1 Previous research

The purpose of this review is to form an overview of international research on police use of force, perceived as relevant for the focus of this study. The overview forms a base for positioning this study among similar research conducted internationally, and in Norway; and provides a comparative basis for the findings in this study, potentially strengthening reliability of findings if consistent with previous research.

An extensive body of international research has addressed various perspectives on police use of force, including police–public encounters and use of force (Alpert & Dunham, 1997; Alpert et al., 2004; Bayley & Garofalo, 1989; Garner et al., 1996; Garner et al., 2002; Manzoni & Eisner, 2006; McKenzie, 1996; Nickel, 2015); characteristics and prevalence of force (Garner et. al, 2002; Klahm & Tillyer, 2010; Terrill, 2001); moral beliefs and use of force (Noppe, 2016); suspect mental disorder and use of force (Donner, 2012; Johnson, 2011; Rossler & Terrill, 2017); excessive force (Alpert & Smith, 1994; Alpert et al., 2004; Atherley & Hickman, 2014); the force continuum (Garner et al., 1995; Terrill, 2001; Terrill & Paoline, 2013); police use of force and subject resistance (Boivin & Lagacé, 2016); police officer gender and use of force (Alpert & Dunham, 2004; Paoline & Terrill, 2005; Schuck & Rabe-Hemp, 2007); police officer experience and use of force (Croft 1985; Paoline & Terrill, 2007; Terrill & Mastrofski, 2002); prevalence of use of force at the agency level (Terrill et al., 2008; Worden, 2015); deadly force (Geller & Scott, 1992; McElvain & Kposowa, 2008); police use of firearms (Belur, 2010, 2014; Burrows, 2007; Boulton & Cole, 2016; Petersson et al., 2017; Punch, 2010), and the impact of realistic use of force training (Andersen et al., 2016; Oudejans, 2008). This diverse literature has also established several factors affecting police use of force (Klahm & Tillyer, 2010; Riksheim & Chermak, 1993; Sherman, 1980) and the fact that police use of force is related to officer, suspect, and situational characteristics.

Meanwhile, exercise of coercive force by the Norwegian police, and how experiences from such use of force specifically affect operational training, have received limited attention in scholarly literature. Studies to date include assessment of police training and use of apprehension techniques (Lagestad, 2008; Lie, 2010), training in decision-making for police emergency response personnel (Johnsen et al., 2016), use of pepper spray (Holmberg, 2013), police use of firearms in Norway and the Scandinavian countries (Knutsson, 2005; Kuhns & Knutsson, 2010; Strype & Knutsson, 2002), and police officers’ attitudes toward the use of force (Burke & Mikkelsen, 2005). Given the widespread international research on police use of force, it is fair to say that similar research is limited in Norway, and more research is needed. However, a limitation of this study must be acknowledged. The study only focused on ordinary police emergency response officers (IP4), full- or part-time engaged in operational service, including officers in the police special response unit (IP3, similar to UK armed response vehicle). Thus, the study did not include police officers whose specific duties may require the use of force (e.g., those working in police arrests), special units (e.g., the counterterrorism unit), or the five national specialized agencies¹.

1.2 Purpose of the study

The purpose of this study is to provide a general overview of the prevalence and particularly the characteristics of Norwegian police emergency response officers’ use of force.

This includes framing all police use of force (including firearms) in line with the Norwegian police use of force model (the force continuum). A main hypothesis forms the basis of the study: Norwegian police emergency response officers frequently use coercive force in their service, the officers relate to the framework in the use of force model and the exercise of force is concentrated in the lower end of the model. Against the background of a non-existent reporting system for police use of force, the hypothesis’s point of departure is information provided by informal conversations with police emergency response personnel.

The article is structured in six chapters: Following the introductions, in Chapter 2, we describe the police use-of-force framework, to indorse an understanding of the legal framework underlying the Norwegian police's exercise of force, including the use of force model. The research methods are reported in Chapter 3. Chapter 4 presents findings in two sub-sections: descriptive statistics and regression analysis. In Chapter 5, we discuss the main findings of the study in relation to previous research. In addition, a suggestion for a new definition of the use of force is offered. Finally, we present our main conclusions and provide some suggestions for future research.

2.1 The force continuum

Police organizations regularly use models to visualize legitimate and proportionate use of force, and police policies and training in the use of force commonly refer to a continuum of force (e.g. Lohne Lie & Lagestad, 2011; National Institute of Justice, 2018; New Zealand Police, 2019). The dominant use-of-force model has been the force continuum (Terrill & Paoline, 2013). This model indicates that police should attempt to use as little force as possible in any police–public encounter and that they should adapt the severity of the force that they use to escalate and de-escalate the situation, especially depending on the level of the subject’s resistance (Garner et al., 1995; Terrill, 2001). Thus, the force continuum provides descriptions of proportionate use of force and a ranking of coercive measures according to law and policy. Determining the characteristics of reasonable use of force in any given situation is challenging, because the behaviors of all parties in a police–public encounter, including their comments and actions, create situations ranging from courteous to explosive (Alpert et al., 2004). Various terms are used to describe force that is deemed to be unreasonable. Generally, police use of force that is necessary and not excessive is considered reasonable, while force used prior to resistance, or after resistance has ended, is unnecessary (Boivin & Lagacé, 2016). Although situations may arise in which it is considered legitimate to use force before physical resistance, excessive use of force is considered to be the use of more force than is reasonably necessary to gain compliance (Worden, 2015).

2.2 The Norwegian police definition of use of force

The Norwegian police formally define use of force as, “… a forced physical action against a person and against a permanent or movable property in the event of damage to this property” (author translation) (Lovdata.no, 2018a). This narrow definition has two implications. First, the phrase “forced physical action” excludes verbal commands and threats. Second, the phrase is not operationalized, leaving interpretation of “forced physical action” to the police officers’ professional judgement and discretion. Interestingly, the term is not discussed in greater detail in either the Norwegian Police Act or the Police Instructions (Lovdata.no, 2018b). The only guidance for interpreting use of force is in the preparations for general service instructions for police officers, “In relation to persons, however, any forced action must be classified as force, including shoving, pushing and other intrusions and direct physical impact of the body of more ‘trivial’ types” (Auglend, 1988, p. 128). However, this text is not included in police instructions, creating potential uncertainty regarding use of lower-force-continuum physical interventions. Such uncertainties regarding what should be formally regarded as use of force will likely affect police officers’ reporting of their use of force, and thus the data on which future studies are based.

2.3 The Norwegian police use-of-force model

Norwegian police use the Force Pyramid model (Figure 1) to describe the force continuum and to operationalize the guidelines in both law and police instructions (Lohne Lie & Lagestad, 2011, p. 10). This model, which is included in mandatory training for all police students and officers, shows the categories of forcible means at the disposal of police emergency response personnel, with ranked levels of severity of physical and mental injury that may occur as a result of their uses. In addition, the model refers to escalating and de-escalating of use of force in connection with each coercive measure. However, the model does not include descriptions of subject resistance, which is deemed to be proportionate to the level of force used by the police. Thus, the model is neither a complete nor a generally valid description of use of force, but rather a ranking of different coercive measures (Lohne Lie & Lagestad, 2011). Unlike the force pyramid model, an alternative model, the Police Use of Force Model, also includes electronic control devices (ECD, for example TASER) and the use of dogs, in addition to situational assessment (Henriksen, 2016).

Figure 1

Herein, the concept of force is understood to be limited to physical force, in accord with the Norwegian police definition (Lovdata.no, 2018a), although not all respondents in the current study necessarily used the same definition. Nevertheless, only the levels above the dotted horizontal line on the model are considered use of force. These include: (1) apprehension techniques and any physical intervention that is deemed to be force; (2) pepper spray; (3) baton; (4) punches and kicks; and, as a last resort, (5) use of firearms. From January 1, 2019, the Norwegian police introduced a two-year trial use of electronic control devices (ECD) as a new means of force (Nrk.no, 2018). There is currently no information available regarding where on the Force Pyramid ECD will be placed. Use-of-force models have additional limitations, including their inability to depict appropriate officer responses to every police–public encounter, because these are dynamic and complicated situations (Arsenault & Hinton, 2014, p. 138).

3.1 Questionnaire

Due to the lack of an adequate reporting system for police use of force in Norway at the time of the survey, no national reporting form existed on which to base this survey. Thus, a questionnaire was developed, based on established police use-of-force reporting systems in six other countries.² The questionnaire focused on common core data and questions used by other police organizations, adapted to Norwegian policies, including the definition of use of force and the use-of-force model (Figure 1). Four categories divided the questions thematically: The police officers, their use of force, the subject (against whom force was used) and experiential learning. Only single-choice questions were used, and respondents were not required to answer every question to complete the survey. Four experts, representing each of the disciplines within police training in the use of coercive force and firearms at PHS, validated the professional content and questions. In January 2018, respondents received the electronic questionnaire, in which they were asked: “About how many times in service did you use force against a person (s) throughout the year 2017?”, and specifically “How many times in service did you use force against a person (s) in the month of December 2017?” In the main section of the questionnaire, the participants were encouraged to respond to “The last time you used force in the service, can you describe the specific incident…”, followed by questions on characteristics of this specific incident. Finally, a few questions related to experiential learning from such incidents were also included, but not reported herein.

3.2 Sampling

The Norwegian police employs approximately 10 000 sworn officers (Politiet.no, 2018). This survey focused only on the police emergency response personnel in the operational service. These officers are the police emergency response personnel who handle the vast majority of police–public encounters and are most likely to use force in service. The survey was distributed via their workplace e-mail to 3772 potential respondents, of which 1637 police officers completed the survey anonymously. The survey only addressed police emergency response officers, although a very limited number of other employees, such as civilian employees, also received the questionnaire (but did not respond). Although a few other employees received the survey, it is not likely that this influenced the results to a considerable degree, since they did not respond. However, it may have influenced the response rate of 43% negatively, by making it appear to be somewhat lower than it actually is.

3.3 Data

To the best of our knowledge, this survey was the first time that Norwegian police emergency response officers were asked to comprehensively report on their use of force, using a specific reporting form. The data analyzed herein are uses of force self-reported by police emergency response personnel in all 12 police districts in Norway. A possible limitation is therefore that only the police versions of these incidents were included; however, police-reported data are the most readily available regarding police–public encounters, despite their potential biases (Garner et al., 2002). In addition, these data were obtained retrospectively, and several officers expressed difficulty remembering how many times they had used force in the past year. Thus, reporting may have been affected by police officers’ reporting the more severe, if not the most recent, incidents, because these were recalled in greater detail. Although the survey findings were supported by focus group interviews, the use of force reported herein may be more severe than that which was actually used.

Police emergency response personnel are expected to be familiar with the definition of use of force based on their cumulative education and instructions. Despite this, feedback on the survey indicates that a more precise demarcation is needed for the definition of use of force, particularly when reporting use of physical interventions at the lower levels of the continuum. Thus, uncertainty regarding what respondents considered to be use of force may also have affected these data, at least to a minor degree.

One challenge to researching police use of force is capturing the various aspects of the assignments and development of the situations in question. Alpert and Dunham (2004) have pointed out that police interventions are dynamic interactions between police officers and subjects, and thus “researchers and analysts know that police interventions cannot be understood as phenomena that unfold consistently and inevitably to a particular end” (Boivin & Lagacé, 2016, p. 192). Reports of their use of force may only capture the moment of confrontation, without the contextual nuances of perceived threat or the turning point in the situation. Nor does this study take into account the police officers’ intentions when using force. For example, one respondent described a situation in which a girl was physically forced away from the edge of a building, from which she intended to jump in order to take her own life. This incident was recorded as use of physical force, consistent with a typical arrest, illustrating the diversity of situations in which police officers may resort to use of force.

3.4 Analysis

Two methods were used for data analysis. First, frequency analysis was used to provide descriptive statistics, to present an overview of response options and the corresponding percentage of respondents (Ringdal, 2018). Second, linear regression analysis describes the relationship between the dependent variable (prevalence of use of force in December 2017) and the independent variables (e.g. gender and age) (Johannessen et al., 2016). This analysis contributes to causal explanations for the prevalence of the use of force, by measuring the effect that variables such as age and gender (independent variables) could potentially have on the prevalence of respondents' use of force in December (dependent variable).

3.5 Validation of results

To strengthen the internal validity of the findings, two focus group interviews were conducted with police emergency response personnel from four medium-sized police districts. The interviews included both male and female officers, in both competency categories (see full details in Section 4.1.1), employed by both police stations and smaller police offices, and having professional experience ranging from 1 to 25 years of service. When asked how the survey findings corresponded to their own operational service experiences, they generally indorsed the results presented herein. For example, a majority of informants generally agreed that the prevalence of use of force seemed recognizable, although it was commented that the regularity seemed somewhat low. Regarding police officer characteristics, differences between gender and prevalence of use of force were not remarked on, except by one informant who stated that female police officers use less force than male officers do. Several informants agreed with the findings that younger police officers seem to use force more frequently than their older colleagues do. The informants largely agreed with the characteristics of the use of force or did not make any specific comments. Regarding the distribution of different forms of use of force, according to the levels of the Norwegian use-of force-model (Figure 1), the informants made no comments except for one who specifically stated it seemed to be correct. Regarding findings on subject characteristics, the informants agreed that they most commonly use force against men and that subjects are usually somehow affected/intoxicated; one informant elaborated on this, stating that the number of subjects with mental disorders exposed to police use of force was less than expected. However, the officers emphasized major differences within and between police districts regarding use of force, particularly due to the location of specific institutions or individuals having an extensive impact on the prevalence of their use of force. These findings have not been broken down based on police district. However, because the respondents were from all the national police districts – and with an overall response rate of 43% – the results are likely to have external validity and thus are generalizable to all districts.

4.1 Descriptive statistics

4.1.3 Subject characteristics

Subjects exposed to police use of force (n = 1547) were 85% male and 13% female. In 2% of the reported incidents, force was used against both men and women. The officers (n = 1571) reported the subjects’ ages within 5-year ranges, with 49% aged 21–30 years and 94% presumed to be aged 50 years or younger. The subjects (n = 1572) were unarmed in 82% of incidents and in the remainder were armed with weapons/objects including knives (9%), objects used to strike (1%), firearm replicas (1%), firearms (1%), and other dangerous objects (6%). The subjects (n = 1571) were reported to be intoxicated and/or to have a mental illness in 92% of the police–public encounters in which coercive force was used (Figure 2).

Figure 2

4.1.5 Use-of-force characteristics

Police emergency response personnel reported various reasons for their use of force (Figure 3).

Figure 3

The police officers (n = 1571) used less forcible means (e.g., verbal commands, warnings) in 82% of the incidents before any physical force was used. In 14% of incidents, no warnings were given, and in 4%, the police threatened use of weapons (all categories) before coercive force was used. It is common practice in the Norwegian police for two officers to patrol together; these respondents (n = 1549) reported that, in 66% of the incidents, there were two officers who used force against the subject(s); in 10%, only one officer used force; and, in the remaining 24%, three or more officers were involved. For the question regarding who led the assignment, it was reported (n = 1551) that 18% were led by an on-scene commander, 69% were led by another police officer (e.g., one of the officers themselves), and in the remaining 13%, there was no defined commander.

In 87% of the incidents (n = 1562), physical force (e.g., “come-along holds,” apprehension techniques) were reported as the first type of force used. In 6%, pepper spray was used first, 1% involved telescopic baton, 1% involved kicks and punches, 5% involved threats with the use of a firearm, 0.1% involved firing a warning shot, and 0.1% involved using a firearm against a person as their first coercive measure. On the question of how much force was necessary to gain control of the situation, only minor escalations were reported (n = 1574) from the initial use of force: using physical force was sufficient in 83%, pepper spray in 8%; telescopic baton in 2%; kicks and punches in 2%; threats with a firearm in 5%; firing a warning shot in 0.2%; and use of a firearm in 0.1%.

Injuries to the parties involved and to bystanders were also recorded. Among the situations reported (n = 1577), no injury, or injury without any need for treatment (e.g., minor bruises) for the subject(s), occurred in 98% of cases. In 2%, injuries to the subject(s) required treatment. The same frequency was reported (n = 1574) for injuries to the police officers themselves. Bystanders (n = 1574) were reported to be injured and in need of treatment in 1% of the situations in which force was used. These findings show that in the majority of situations, two officers use force against the subject(s) and that the officers predominantly took the lead on these assignments. In the vast majority of situations, the police officers tried less coercive measures (e.g., verbal warnings) before they escalated to the use of physical force. When used, force was concentrated at the lower end of the force continuum, a level that generally did not escalate the situation and seldom caused injury.

4.2 Regression analysis

The questionnaire data followed a similar structure to several existing reporting forms for police use of force. However, some variables were recoded for analysis. Linear regression was used for analyses of a selection of independent variables, and data were analyzed using IBM SPSS 24.