Police_Contact

Criminal laws written by Local, State, and Federal entities determine and write laws that are specifically used to direct who are approached by Police Officers. These interactions are a critical and highly complex process surrounded by intense debate and scrutiny. So too is the development of an institution designed to ensure that such laws are followed by police to a point of limiting the civil rights of young people. Limitations or parameters must be set defining the extent of police power granted to this institution to enforce laws. Perhaps Bittner (1970) provides the clearest contemporary description of the rise of policing as one of the last institutional structures of democratic governments to contain law and keep the peace.. Bittner argues that relinquishing to the police the right to use coercive actions goes against middle class values of achieving peace by peaceful means. This is the basis of all contact with people of interest, especially when contacting young people . However, there are instances when only a forceful police action can maintain or restore peace, so the police are given the right to use force. The police are faced with a paradox: to stop violence the police may sometimes need to resort to violence (Sherman, 1980a). Although coercive police tactics are viewed as an evil, they are a necessary evil nonetheless. In 1931 the National Commission on Law Observance and Enforcement (Wickersham Commission) brought national attention to the issue of police violence in the form of brutality for the first time. The commission published the Report on Lawlessness in Law Enforcement that characterized “third degree” police tactics as a major institutional problem (National Commission on Law Observance and Enforcement, 1931). Over twenty years later, Westley (1953) was one of the first scholars to study the police and their views toward the application of force. He discovered that officers rely on force because they view it as an effective means to control those in need of control. Westley also found that force was used as a way to gain public and peer respect. Despite Westley’s early insight, it was not until the mid-1960s that the first wave of scholarly attention was given to police use of force. Much of this attention was directed at excessive force, most notably deadly force. In many cities throughout the country, riots ignited after claims of excessive police force. The President’s Commission for Law Enforcement and Administration of Justice concluded that these allegations of police violence were legitimate. Shortly after, the Law Enforcement Assistance Administration (LEAA) was created with the aim of improving the criminal justice system. In an attempt to professionalize policing, the LEAA provided funds for police officers to attend college. The administration also allocated money for research on criminal justice issues, the first large scale federal effort toward such ends. As a result, most of what is known about police use of force has been acquired over the past 30 years, and a majority of this research has focused on deadly force. Less attention has been given to nonlethal force. Data on police force have generally been derived from four sources: observational studies, officer use of force reports, citizen complaints, and surveys. Each of these mechanisms has strengths and weaknesses. For example, observational research while generally providing rich detail, is rarely used to collect force data as a result of the labor intensiveness of the technique. Conversely, almost all departments collect some type of citizen complaint data, but this mechanism only involves reported cases of excessive force. Nonetheless, police use of force inquiries most often attempt to determine one or more of the following: frequency and prevalence of police use/abuse, degree of variation in types of forceful actions and the application of force within individual encounters, and the causes of police force/abuse. The research presented here follows in this tradition. It examines the extent of police force using observational data, determining how force varies and is applied within individual encounters, and explores the causes of such variation. Accordingly, the following literature review is organized around these three themes. Following this review the limitations of prior work is discussed, which sets the stage for the proceeding inquiry. Observational Studies Reiss (1968) was one of the first researchers to systematically measure the extent of police force. In 1966 he conducted an observational study of police in Chicago, Boston, and Washington D.C. When examining the issue of police force, Reiss was primarily interested in measuring excessive force. He created an expert panel to review incidents where the “...policeman struck the citizens with his hands, fist, feet, or body, or where he used a weapon of some kind – such as a nightstick or a pistol” to determine if excessive force was used (1968: 3). Of 1,565 police-citizen encounters with suspected offenders, 44 involved instances where officers struck citizens in the manner Reiss outlined. Of these 44 encounters, 37 were judged by the review panel to have involved excessive force (2.4%) (1968). Friedrich (1977; 1980) re-examined Reiss’ data. Unlike Reiss, Friedrich relied solely on the observers’ classification of whether force was used at all, and if so, whether he or she believed force was excessive. Using this classification method and based on 1,565 police-suspect encounters, he found that reasonable physical force was applied in 51 cases (3.3%) and excessive force in 29 cases (1.8%). In 1977, a larger scale observational study known as the Police Services Study (PSS) was carried out in 24 departments in three metropolitan areas (Rochester, New York; St. Louis, Missouri; and Tampa, Florida). Using data from this study, Worden (1995) analyzed 1,528 police-citizen encounters involving suspected offenders. He concluded that reasonable force was used 2.4 percent of the time and excessive force was applied 1.5 percent of the time. Another police observational study was done in the mid-1980s by Bayley and Garofalo (1989) in New York City. Of 467 police-citizen encounters classified as “potentially violent,” they found some form of physical force was used about eight percent (37 cases) of the time. 1 As part of a training evaluation regarding police use of force, Fyfe (1988) conducted observational research with the Metro-Dade Police department in Miami, Florida. Of approximately 2,000 potentially violent police-citizen situations, he found that force ‘greater than firm voice commands’ occurred about 12 percent of the time. 2 Klinger (1995) also examined the Metro-Dade data, but looked at a subset of cases classified as disputes. He found that in 241 dispute cases, some form of physical force was used 17 percent of the time. However, he also looked at voice commands and found this type of force was used most frequently (39%). In summary, observational studies have examined the extent to which police use of force occurs. Several of these studies have also attempted to measure the extent of excessive force. From these efforts, it appears that both force in general, and excessive force, are rare events. 3 Use of Force Reports Officer use of force reports is another data collection method used to assess police force. Analyzing 123,500 arrest reports from Rochester, New York between 1973-1979, Croft (1985) found some form of physical force used in approximately two percent of these arrests. In a later study, Croft, along with Austin (1987), examined police use of force in Rochester and Syracuse. Examining data from 1984 and 1985, they found force was used in five percent of arrests in Rochester and in four percent of arrests in Syracuse. 4 Using force reports from custody arrests over a 12-month period in St. Paul, Minnesota, Lundstrom and Mullan (1987) found force was used in 14 percent of the cases. No attempt was made to distinguish between reasonable and excessive force in the Croft (1985), Croft and Austin (1987), or Lundstrom and Mullan’s studies. A few years later, McLaughlin (1992) also looked at use of force reports in approximately 11,000 arrests made by Savannah, Georgia officers. He found physical force was used only one percent of the time. One of the most recent studies to examine use of force reports was conducted by Garner and colleagues (1995) in Phoenix, Arizona. They found the highest rate of physical force to date; of 1,585 arrests over a two-week period, officers used some form of physical force 22 percent of the time. What is particularly interesting in force studies that rely on use of force reports is the variation in rates of force found. As highlighted above, some of these studies (Croft, 1985; McLaughlin, 1992) have found reported force to be very low, similar to what has been found in observational research, while others (Lundstrom and Mullan, 1987; Garner et al., 1995) have found force to be relatively high. Since all these studies use arrests as a base to compute findings it cannot be a matter of differential comparison (e.g., comparison of arrests versus suspects). One interpretation with studies that found lower rates of force may be the result of officer bias in filling out report forms. 5 At the high end it could be the result of counting handcuffing as “use of force.” One benefit of using use of force reports is that they have become a more common data collection technique in recent years for both researchers and police administrators. Many departments now require officers to complete a use of force report whenever an officer uses any form of physical force. An additional benefit is that some reports also include questions pertaining to the level of citizen resistance. This element was often lacking previously, but it can provide a better understanding of the conditions in which force is used. Citizen Complaints Yet another method used to examine use of force is citizen complaints. In 1966 and 1967, Chevigny (1969) examined citizen complaints of the New York City Police department. Of 441 complaints filed, he found that 164 involved allegations of abusive police force. Dugan and Breda (1991) surveyed 165 police agencies in Washington state in 1987 and 1988; analyzing 691 complaints, they found that 123 involved physical force. Although many jurisdictions have recently streamlined the citizen complaint process (better and more accurate procedures), it still appears that measuring use of force through this mechanism is more limited than reliance on use of force reports because many incidents (where force is used) do not produce complaints. In the Reiss (1968) data set, of the 37 cases classified as excessive, only one resulted in a citizen complaint. In a more recent survey, only one out of every three persons who claimed to have been subjected to police physical abuse followed up by filing a complaint (Winick, 1987). Surveys Finally, surveys are also used to determine the extent of police force. In 1966 Bayley and Mendelsohn (1969) surveyed 806 Denver citizens about police brutality. Of these, fifteen percent of Hispanics, nine percent of Blacks, and four percent of Caucasians claimed that they had personally experienced police brutality. Campbell and Schuman (1968) conducted a national survey of citizens in 15 cities for the National Advisory Commission on Civil Disorders. When asked whether they had been “roughed up” by police officers, seven percent of the black respondents and two percent of white respondents claimed that they had. A 1991 Gallup poll found that five percent of citizens surveyed reported having ever been physically mistreated or abused by police, while 20 percent reported that they knew someone who had been physically mistreated or abused (Gallup, 1991). 2.2 VARIATION IN TYPE AND APPLICATION OF FORCE In addition to examining the extent of police force, some researchers have also attempted to measure variation in types of force police use. In 1991, the Police Foundation (Pate and Fridell, 1993) conducted a national survey of police departments concerning the extent of force by officers, types of force used, number of citizen complaints, and force reporting procedures. Because of the diversity of agencies surveyed, they standardized incidents of force per 1,000 sworn officers. For city departments, they found handcuffing to be, by far, the most frequently used type of force (490 incidents per 1,000 sworn officers). The next most frequent type of force used was ‘bodily force’, with 272 incidents per 1,000 sworn officers. The McLaughlin (1992) study revealed only 133 reports of force out of 11,000 arrests, but further analyses were conducted to examine the type of force used. Forty-five of the 133 incidents involved instances where the officer punched or kicked citizens, 11 in which the officer struck the citizen with a baton, and two where the citizen was maced. The McLaughlin (1992) and Police Foundation (1993) studies highlighted the fact that “[n]onlethal force encompasses a wide variety of police actions” (Klinger, 1995: 171). According to Klinger, “[e]xplicit recognition of the diversity of force that is classified as nonlethal is important not only because it identifies differences in kinds of force but also because it points to differences in severity as well” (1995: 171). Klinger acknowledges that because there are different types of force police use, force can be ranked in terms of severity. Therefore, force can be measured along a continuum ranging from the least to the greatest amount of force. Garner and colleagues (1995) in their Phoenix study, and McEwen (1996) who reviewed police use of force data collection efforts at the national level, also refer to the importance of data collection and analysis in regard to a continuum of force. Despite McLaughlin’s contribution, Klinger (1995) went on to note two deficiencies in McLaughlin’s work and those that came before it. First, prior attempts have failed to examine how various types of force are used within individual encounters. In other words, different types of forceful police actions can take place within single encounters (e.g., grabbing to restrain and striking to subdue). Second, such studies have been limited to classifying only force that is physical in nature. However, force can also be in the form of verbal coercion such as commands and threats. Klinger incorporated these two features into his analysis of the Metro-Dade data. He examined 241 police-citizen encounters involving disputes and found that officers used some form of physical or verbal coercion 164 times in 97 encounters. Clearly, multiple forms of force were used within single encounters. Klinger also analyzed combinations of force within single encounters. Of the 97 cases in which some form of force was used, distinct combinations emerged. In 58 of these cases, officers only used one form of force (56 commands, 2 firm grip). In 22 cases a command and firm grip was used. Pain compliance, a firm grip, and a command were used in another 11 cases. The remaining six cases occurred in numerous combinations. Examining force in this manner discloses several important findings. First, most of the force applied was verbal (58 of 97 cases). This type of force has generally been overlooked in the past, but is obviously a behavior that occurs with some frequency. Second, in 35 of the 41 cases in which physical force was used, it was in the form of either a firm grip or pain compliance hold, those physical behaviors at the lower end of the force continuum. Unless one distinguishes between types of force, all cases (in this study 41 cases) would be grouped together as if the level of force were similar or alike, which is misleading. As evidenced by Klinger, most force was of the lowest level. Third, when higher forms of force were used (e.g., firm hold) within a given encounter, lower forms were also applied (e.g., voice command). Of 97 cases that involved voice commands or some form of physical force, 98 percent of them involved voice commands. The implication is that officers did not increase or immediately apply a higher level of force without using lower forms as well, although Klinger did not employ temporal sequencing. Garner and colleagues (1995) also examined varying types of physical force. Like Klinger, they found that most of the physical force used was at the lower end of the continuum. Of 1,235 cases involving physical force, 918 were in the form of restraints. It is somewhat difficult to discern the significance of the similarities and differences between these two studies. For one, Klinger’s work was based on data collected through observational research and looked only at disputes. The Garner data was based on use of force reports as part of all arrests. Second, Klinger did not include handcuffing as part of physical restraint, while Garner and colleagues did. Nonetheless, several significant issues emerge from these studies. First, inclusion of verbal force is an important part of the nonlethal force picture. Second, force comes in many forms and to understand how officers use their coercive powers, one has to account for these varying types of force. In addition, Garner and colleagues incorporated varying levels of citizen resistance into their analysis, which is a focal point of the research presented here. Besides counting how often the police use force and distinguishing between various types of force, some scholars have explored a slightly different angle regarding the application of force. This involves an attempt to model the micro-processes of how force is used within police-citizen encounters in order to control or manage the encounter. Sykes and Brent (1980, 1983) conducted an observational study in a mid-size city and analyzed approximately 3,000 police-citizen encounters, focusing more on the micro-process of the police-citizen encounter than previous observational studies. They sought to explore the interactional process between officers and citizens, and how officers regulate or control encounters. They proposed that officers are trained to “take charge,” and conceptualized three means of control used to do so: definitional (questioning), imperative (commands), and coercive (the threat or actual use of physical coercion). They found that, for the most part, officers handled situations through definitional and imperative control. The use of coercive tactics were rarely used and usually only after lesser forms of nonphysical tactics failed. Bayley (1986) also looked at how officers managed encounters with citizens and tactics they used to do so. Like Sykes and Brent, he examined the interactional process of police-citizen encounters. He proposed that police-citizen encounters move through three stages: contact, processing, and exit. Examining traffic stop and disturbance encounters, he found that officers use a wide variety of tactics at each of these three stages. Further, he found that decisions made early in an encounter can affect subsequent decisions to use force. For instance, beginning encounters with tactics such as listening, questioning, or seeking information usually led to a less coercive outcome such as a verbal warning or offering advice. Conversely, taking a more coercive approach at the start (e.g., verbal or physical restraint) had a greater likelihood of leading to a more coercive outcome. Another interesting finding was that mere police presence was often enough to quell citizen resistance, thus preventing subsequent use of physical force. Finally, the aforementioned Metro-Dade study conducted by Fyfe (1988) examined actions taken by patrol officers to deal with potentially violent situations (PVs). As with the other research summarized in this section, this study highlights the actual behavior of officers within individual police-citizen encounters. In other words, what is going on in these encounters that help us to better understand how force is applied' However, Fyfe also looked at police actions prior to the point of mobilization. The purpose of Fyfe’s study was to examine the effects of training on how officers handled potentially violent situations, so as to minimize the likelihood of police-citizen violence. Along the lines of Bayley (1986), Fyfe broke down a potentially violent situation into four stages: unassigned time (e.g., time before an officer is dispatched or aware that a PVs is about to unfold), approach (e.g., time between dispatch to PVs and face to face contact), contact (e.g., entry stage), and resolution (e.g., disposition). Among other findings, he found that actions taken prior to involvement in potentially violent situations (e.g., knowledge of the patrol beat, places, people) may reduce the potential for police force (or subsequent use of police force during the encounter). Interestingly, he also found that some officers in certain situations may not have been aggressive enough in handling potentially violent encounters. That is, they failed to take charge when “it was clearly appropriate to take charge” (1989: 22). 2.3CAUSES OF POLICE FORCE The study of police coercion has not been very theoretical. Rather, it has been examined in the context of a broader understanding of police behavior and discretionary decision making, which itself has not been grounded in a solid substantive theory of police behavior. This fact was first highlighted by Sherman in 1980 (1980b). More recently, Riksheim and Chermak, after reviewing previous studies on police behavior regarding the decision to use force, arrest, detect criminal activities, and engage in service behaviors, concluded, “[c]learly, theoretical development [of police behavior] is lagging far behind the quantitative attempts to estimate the relationships between variables” (1993: 377). Nonetheless, the literature has introduced three primary explanatory perspectives that serve as theoretical guides to understanding police use of force: sociological, psychological, and organizational. Sociological Understanding police behavior within a sociological framework often involves examining the social dynamics of police-citizen encounters. This approach has most often been aligned with Donald Black’s theory of law (1976). 7 Black posits that the application of law can be explained in terms of social space or distance. In the context of police force, Black’s theory predicts that the police will be more coercive toward those perceived to have lower status, such as minorities and the poor. More directly, this area of explanation examines the structural or situational characteristics in which police and citizens interact. In general, researchers who have applied a sociological approach to police behavior posit that officers rely on, or look to, situational cues as a basis to judge how a particular incident should be handled. Perhaps Bittner’s (1970) statement regarding officers’ “intuitive grasp of situational exigencies” best illustrates this perspective. Thus, applying a sociological approach, the following determinants have been hypothesized to affect the use of police force: citizen race, gender, class, age, demeanor, sobriety, mental state, and emotional state; number of officers present; number of citizens present; location of encounter; and type/seriousness of offense. Reiss (1968) examined his data from a sociological viewpoint. He found that most victims of police force were young, lower-class, and classified as suspects. He also found that 78 percent of the time force occurred in police-controlled settings such as the patrol car, precinct, or public street, and that there were rarely any civilian witnesses. Those who defied the officers’ authority were the most likely to be on the receiving end of undue force. Almost half (48%) of all excessive force victims had either challenged the officers’ authority or physically resisted arrest, and 32 percent were considered deviant in some manner by officers (e.g., intoxicated, homosexual, drugs addicts) (1968). 8 In 1966 and 1967 Chevigny (1969) analyzed citizen use of force complaints made to the New York Civil Liberties Union. Of those complaints that were sustained, in 71 percent of the cases police force was the result of citizen defiance to police authority. Further, defiance was primarily verbal rather than physical. Toch (1969) also examined policecitizen violence. He conducted interviews with both officers assaulted by citizens and citizens assaulted by officers in Oakland. He concluded that most of the violent interactions were the result of citizen disrespect. Toch was also able to analyze the interactional process between officer and citizen. He found that half of the cases that led to excessive force complaints began with a verbal request by an officer followed by a citizen’s failure to abide by the request. At that point, the officer escalated to a command, which the citizen disobeyed and the cycle continued on to physical force. Toch’s study was the first to empirically demonstrate the escalation of police-citizen violence within individual encounters. Friedrich (1977) also used a sociological approach to examine why officers resort to force. Like Reiss, he found support for a sociological explanation. For instance, he found that use of force was more likely if a citizen was antagonistic, agitated, intoxicated, lower-class, suspected of a felony, and if there were bystanders present (e.g., another citizen or officer). Friedrich also looked at psychological and organizational explanations, which will be discussed below. In 1995, Worden re-examined the Police Services Study (PSS) data gathered in 1977. Like Friedrich, he analyzed the determinants of force from several different theoretical frameworks. He too found similar results to Reiss and Friedrich; situational determinants bore the strongest relationship to the use of force. He found suspects that were antagonistic, lower-class, male, intoxicated, and black were more likely to be victims of force (1995). Garner (1996) uncovered several significant predictors of police force toward citizens. Arrestees who were alleged to have been involved in a violent offense, were antagonistic, male, involved in a gang, intoxicated, or known to have a weapon or reputation of being resistant toward the police, were all more likely to be the recipients of forceful police actions. Other factors found to impact officer use of force included the increased presence of both police and bystanders, and officer contact and cover tactics. Another area of explanation often overlooked but nicely illustrated by Sykes and Brent (1983), and to some extent Bayley and Garofalo (1989), involves officer success or failure to control through noncoercive means. As previously mentioned, Sykes and Brent (1983) found that officers rarely resort to force and when they do it is usually after other means of control have failed. In other words, officers resort to force because they are unable to exert control in other ways, thereby leaving force as the only option. This should not be confused with the way most researchers have applied situational variables. Reiss (1968), Friedrich (1977, 1980), Worden (1995), and others looked at situational factors but only in the form of those that are either present or not present in a police-citizen encounter. For example, was the citizen antagonistic, intoxicated, emotionally or mentally disturbed' The fact that these factors increase the likelihood of officers resorting to force does not tell us anything about the officers’ efforts prior to making a decision to use force. It could be that officers “jump” right into applying force more readily with these citizens or it may be that officers have a more difficult time dealing with these people because of the behavior they are displaying, which then results in police force. The problem with looking at the encounter as a whole is that you cannot determine which of these two competing explanations are more accurate. Friedrich (1977) found that police officers treated intoxicated citizens more forcefully than non-intoxicated citizens while controlling for antagonistic behavior on the part of the citizen. This implies that the police treat intoxicated citizens more harshly simply because they are “drunk”, not because of their actions, something that both Westley (1953) and Reiss (1968) alluded to in their research. However, this fails to answer the question of whether the police treated them harshly only because they were drunk or because officers were unable to deal with behavioral manifestations of their drunkenness. But the approach taken by Sykes and Brent distinguishes between the two. Perhaps this issue is a matter of semantics, the primary issue being that officers end up having to use force on intoxicated citizens more readily on average. Therefore, it may not matter if it was a result of officers being unable to control them in ways that are not coercive. What may matter is the fact that the citizen was drunk, which caused the officer to respond with force. Perhaps the answer to this puzzle depends on how one frames the question. For instance, does it matter if the police respond more readily with force when certain factors are present in an encounter' Illustrating this view, one may say “a crazy person is a crazy person and reasoning with him or her usually does not work so there is no use trying it.” Or, does it matter whether the police attempt to exhaust all non-coercive means before resorting to coercive means, regardless of whether certain factors are present' This view may be characterized as “a crazy is a crazy, but sometimes reasoning does work so it is worth trying before becoming forceful.” According to one department’s policy (St. Petersburg) on handling mentally impaired citizens, the latter seems to be the appropriate method. Psychological According to a psychological theoretical perspective, officer characteristics, experiences, views and outlooks are posited to have an effect on police behavior. Simply stated, is there something about particular officers that help explain why they use force and how much they use' This explanation rests on the assumption that officers with certain traits, experiences, or attitudes will respond differently in similar situations. Factors that have been hypothesized to matter include: officer race, gender, education, training, age, length of service, views toward citizenry, attitude toward role orientation (e.g., crime fighter versus service-oriented), workgroup, and patrol assignment. Muir’s (1977) theoretical orientation may best illustrate the role of the individual. Muir was primarily interested in characterizing “good” versus “bad” police officers. He constructed a four-fold typology into which officers could be grouped depending on two dimensions: their view of human nature and their attitude toward the use of coercive authority. Others have also constructed typologies of officers based on various dimensions to explain the use of force (White, 1972; Brown, 1981). However, minimal systematic evidence exists to document the impact of officer belief systems in relation to use of force. Brown’s (1981) study of officers in three departments in southern California was based on observations and interviews with officers concerning how they went about various discretionary decisions. Based partly on Brown’s study, Worden concludes “behavior is not a simple extension of attitudes, as organizational and other social forces can attenuate the impact of attitudes on behaviors” (1995: 35). In short, most explorations regarding the links between attitudes and police behavior have been impressionistic at best. However, as Worden (1995) notes, none of the previous psychological studies solely examined the use of force. As a result, Worden tested several officer attitudes and their relation to the use of force. He posited that officers, “who are most likely to use force could be expected to (a) conceive the police role in narrow terms, limited to crime-fighting and law enforcement, (b) believe that this role is more effectively carried out when officers may use force at their discretion, (c) regard the citizenry as unappreciative at best and hostile and abusive at worst” (1995: 34). He found that officers with negative views toward the citizenry (e.g., believed citizens were unappreciative) were more likely to use both reasonable and improper force. Officers with more favorable views toward force (e.g., believed discretion should be at the officer level) were significantly more likely to use improper force. Finally, he found that how officers conceive the police role did not have an effect. Unlike research attempting to link attitudes with use of force behavior, there has been more of an accumulation of research pertaining to the impact of officer traits and experiences on use of force, but to a lesser extent than might be surmised. Cohen and Chaiken (1972) looked at the relationship between 33 officer background characteristics and various work performance indicators among 1,915 New York City police officers from 1957-1968. One of these measures examined use of force complaints against officers. They found that age, race, and education were all significantly related to citizen complaints. Officers who were white, older, and more educated received fewer complaints. In terms of education, Cascio (1977) also found that more educated officers received fewer complaints. Conversely, Worden (1995) found that officers with a four-year degree were significantly more likely to use physical force classified as “reasonable”, but found no effect on force judged to be “excessive.” Unlike Cohen and Chaiken (1972), most research has not found officer race to be a predictor of force. Friedrich (1977), Croft (1985), Garner et al. (1995), and Worden (1995) all found no difference in terms of officer race. However, Friedrich (1977) did conclude that black officers patrol more aggressively and make more arrests, irrespective of citizen race. As per gender effects, Worden (1995) found none while Garner and colleagues (1995) found that male officers were more likely to use force on male arrestees. Not surprisingly, information regarding gender and race is sparse. Until the past decade or so, police departments were almost exclusively made up of white males. As a result, making valid comparisons has not been possible in many studies. Research has shown that less experienced officers are more active; they patrol more aggressively (Friedrich, 1977), and make more officer-initiated stops (Worden, 1989) than their more experienced counterparts. However, officer experience appears to have little to no influence on officer use of force behavior. Worden (1995) found that officer length of service had no effect on either the use of reasonable or excessive force behavior, while Garner and colleagues (1996) found that officer experience was an “inconsistent” predictor of force. 9 Finally, the amount and type of training officers receive has been hypothesized to have an impact on use of force decisions (Fyfe, 1988). Nonetheless, this has been largely unexplored. Even the Garner (1996) study, one of the most recent inquires on use of force, failed to account for differences in officer training. 10 One reason this may be absent in most studies is the difficulty of obtaining training information that can be linked to individual officers. Another reason may be that researchers consider such variables as length of service and education to be adequate proxies for training. Organizational Organizational theory looks to the role organizations play in shaping officer behavior. Two predominant organizational theories have emerged in the past 30 years. In 1968, Wilson proposed an organizational theory that emphasized the formal structure of an organization and the political environment within which it operates. The theory presumes that through organizational rules, regulations, standard operating procedures, incentives and disincentives, combined with direction from top administrators, a style of policing develops. In other words, a common vision becomes part of each officer’s mind set of how to handle the everyday aspects of police work. Thus, similar incidents that occur at the street-level will be handled similarly by different officers in that department. Wilson (1968) proposed that within a department, one of three styles of policing develops: legalistic, service, or watchman. For use of force purposes, this theory suggests that departments classified as watchman are the most likely to engage in forceful police actions. Others have also emphasized the influence of organizational policies. Ground breaking work done by Fyfe in the 1970s demonstrated the impact of restrictive administrative policies on lethal police force. After the New York City police department adopted a restrictive firearm policy in 1972, police firearm discharges decreased by some 30 percent. Further, Sherman and Cohn (1986) examined police shootings in big city departments between 1970 and 1984 and found that the number of citizens killed by police shootings was cut in half. A second organizational theory looks to the impact of the informal structure (police culture) rather than the formal structure (Brown, 1981). This theory presumes that the police culture serves to protect and isolate officers from internal and external scrutiny more than it does to forge a particular style of policing. Instead, operating through the been largely unexplored. Even the Garner (1996) study, one of the most recent inquires on use of force, failed to account for differences in officer training. 10 One reason this may be absent in most studies is the difficulty of obtaining training information that can be linked to individual officers. Another reason may be that researchers consider such variables as length of service and education to be adequate proxies for training. Organizational Organizational theory looks to the role organizations play in shaping officer behavior. Two predominant organizational theories have emerged in the past 30 years. In 1968, Wilson proposed an organizational theory that emphasized the formal structure of an organization and the political environment within which it operates. The theory presumes that through organizational rules, regulations, standard operating procedures, incentives and disincentives, combined with direction from top administrators, a style of policing develops. In other words, a common vision becomes part of each officer’s mind set of how to handle the everyday aspects of police work. Thus, similar incidents that occur at the street-level will be handled similarly by different officers in that department. Wilson (1968) proposed that within a department, one of three styles of policing develops: legalistic, service, or watchman. For use of force purposes, this theory suggests that departments classified as watchman are the most likely to engage in forceful police actions. Others have also emphasized the influence of organizational policies. Ground breaking work done by Fyfe in the 1970s demonstrated the impact of restrictive administrative policies on lethal police force. After the New York City police department adopted a restrictive firearm policy in 1972, police firearm discharges decreased by some 30 percent. Further, Sherman and Cohn (1986) examined police shootings in big city departments between 1970 and 1984 and found that the number of citizens killed by police shootings was cut in half. A second organizational theory looks to the impact of the informal structure (police culture) rather than the formal structure (Brown, 1981). This theory presumes that the police culture serves to protect and isolate officers from internal and external scrutiny more than it does to forge a particular style of policing. Instead, operating through the protection of one another, officers are able to develop their own unique styles. For instance, in the three departments Brown (1981) studied, he found that individual, not departmental, styles of policing emerged within each of the three departments. Similar to Brown, Manning (1989) examined the role of the informal structure. Although Manning concedes that first-line supervisors may have some influence on officer behavior, he concludes that the formal aspects of a police organization have little impact on officer behavior. Environmental criminologists like to point out that it is easier to alter or manipulate environmental features than to alter the behavior of criminals and their victims. In a similar fashion, researchers see promise in organizational theories because of the view that the organization is more readily manipulable (at least the formal aspect). The problem, however, is that it is difficult to test adequately many of the organizational theories that have been proposed because of the expense and difficulty of collecting information on multiple agencies. The extant research on testing organizational hypotheses regarding use of force are limited to a few studies. Friedrich (1977), using the Wilson (1968) typology, found little support for a net organizational effect, although findings were in the predicted direction. Worden (1995) also examined Police Services Study (PSS) data from an organizational perspective. He examined three organizational characteristics: degree of bureaucratization, emphasis placed on crime-fighting activities by the chief, and a measure for informal culture. Of these, only the degree of bureaucratization was significantly related to the use of “reasonable” force. The more bureaucratized the department, the greater the likelihood force was used. According to Worden, it would appear that in smaller departments with fewer levels of hierarchy, the formal aspects of the organization may matter. In summary, situational explanations seem to offer the most insight on why officers use force. Those factors that most consistently predict the use of force are characteristics of the citizens and their behavior (e.g., demeanor, antagonistic or aggressive behavior, intoxication, and gender). Both organizational and individual officer characteristics have failed to explain much in terms of police force. 2.4LIMITATIONS OF PRIOR WORK ON POLICE There are several elements of these previous studies that have hindered our understanding of police use of force and contact between Police Officers and youth. The three highlighted here form much of the basis for how I originally set out to fill the existing gaps in the extant research. First, prior works have primarily examined the application of police contact in the form of what has been termed as sometimes non-specific reasoning and focusing on extreme police violent attitudes such as indiscriminate contact. Second, they have tended to conceptualize police contact and emotional and physical force dichotomous: excessive force/non-excessive force, physical force/nonphysical force. Third, they often examine or view the application of police force in the context of a static rather than a dynamic process. The following describes these limitations in more depth. Many scholars have struggled with a variety of terms to describe police violence after contact and what reasons are that makes it justifiable to our society. These include: excessive use of force, use of excessive force, brutality, unauthorized force, wrongful force, unjustified force, misuse of force, and unnecessary force. While these phrases are interchangeable to some, others note fine distinctions. For example, use of excessive force can be defined as more force than needed to gain compliance in any given incident, while excessive use of force may be defined as using force in too many incidents (Adams, 1995). Fyfe (1997) makes the distinction between brutality (a willful and knowingly wrongful use of force) and unnecessary force (force used by well-meaning officers ill-equipped to handle various incidents). Worden (1995) also distinguishes between different types of force. He defines excessive force as that which is more than required to subdue a citizen, and unnecessary force as that which precedes a citizen’s resistance or continues after citizen resistance has ceased. One reason for the recent attention placed on defining police violence in more finite terms may be the result of apparent progress made in the control of such behavior. Through a variety of control mechanisms (e.g., criminal laws, administrative policies, greater minority political empowerment, civil protections) it would seem that such extreme behavior is more infrequent today despite some highly publicized incidents to the contrary. Thus, several policing experts are beginning to further examine types of force that are not overtly egregious, but may be unnecessary (in the sense as defined by Fyfe). Klockars (1995) argues most police agencies gauge officer use of force using minimum standards. That is, the criteria for the legitimate use of force is that which is not a criminal violation, prevents any civil liability, and is of a nature that will not bring embarrassment to the department. These standards are necessary, but are they sufficient' Several police experts believe they are not. Klockars states, “[we] would not find the behavior of a physician, lawyer, engineer, teacher, or any other professional acceptable merely because it was not criminal, civilly liable, or scandalous and it is preposterous that we continue to do so for police” (1995: 17). He calls for broadening the focus of excessive force to include, “...the use of any more force than a highly skilled police officer would find necessary to use in [a] particular situation” (1995: 18). Even dating back to Bittner there has been an awareness that “the skill of policing consists in finding ways to avoid its use” (1974: 40). Thus, many scholars argue that further insight on police use of force requires examining all instances of police force, not just those “artificially” labeled excessive force. Even further, one must look not only at instances when some aspect of physical force was employed, but all instances when physical force might have been used but was not. Just as there are instances when force may be excessive, there are times when force may be needed but is not used. 12 So, failure to include all cases when physical force may be reasonably needed prevents one from uncovering important factors associated with the use or nonuse of physical force. 13 Police Force as a Dichotomy Another common deficiency in many studies to date involves conceptualization of the dependent variable: police force. Such research often seeks to distinguish between excessive and reasonable force in accordance with the law or police policy. Similarly, studies that attempt to explain the causes of police force often do so using a dichotomous dependent variable (force versus no force; reasonable force versus excessive force), or in some cases, a trichinosis dependent variable (no force, reasonable force, excessive force). There are two problems with this approach. First, these studies usually only measure physical force. However, force can also be nonphysical in the form of commands and threats. Second, they often fail to take into account the varying degrees of physical (restraint techniques, pain compliance holds, striking) and nonphysical (commands, threats) force. Future researchers are advised to broaden the scope of the dependent variable to include a continuum of force. This would include not only physical forms of police force, but nonphysical actions as well. Although most police departments train officers to apply varying levels of force in accordance with the level of resistance encountered by citizens, researchers almost universally fail to take this into account in their analyses. In any given police-citizen encounter, officers can and do apply numerous forms of both physical and nonphysical force. We need to better understand how all forms of police force are applied, in what circumstances, and to what extent. Static versus Dynamic Process A final limitation regarding prior work on police use of force involves the process of how it is applied. Studies that seek to explain or predict use of force decisions too often look at the police-citizen encounter as if it were a single discrete event, without noting the developmental nature of interaction over time within that event. For instance, Worden (1995) in his analysis, found that antagonistic citizens were significantly more likely to be on the receiving end of some type of physical force. This is an important finding, but we still do not know the nature or extent of such antagonistic behavior, only that it occurred at some point during the encounter. Further, we do not know whether the citizen’s antagonistic behavior preceded or followed an officer’s coercive behavior. Examining police-citizen encounters statically also fails to account for the number of times the citizen and officer displayed such forms of behavior. Unless one more closely probes the interactional process in which the encounter takes place, he or she will miss out on valuable cues as to what prompts use of force incidents. The weakness of the approach taken by most researchers is that it fails to account for “sequential” behaviors displayed by both the citizen and the officer throughout the entire encounter. Most analyses look solely at what factors are present or absent during an encounter and fail to capture the numerous behaviors of both citizens and officers occurring during the encounter. As stated by Fridell and Binder, the police-citizen encounter must be “seen to encompass a pattern of interaction between an officer and an opponent and multiple decisions by both” (1992: 386). That is, like transactions between nations, transactions between people have a history, and what one does at any point will be based on the nature of the history up to that point. RELEVANCE OF RESEARCH Section 210402 of the Violent Crime Control and Law Enforcement Act of 1994 calls for an uniform collection of police use of force data. The intent is to gain a better understanding of how and why officers resort to force. The United States Attorney General delegated responsibility of this directive to the National Institute of Justice (NIJ) and Bureau of Justice Statistics (BJS). These agencies have recently released their latest report concluding (National Institute of Justice, 1999, pgs. vii-x): What we know with substantial confidence: • Police use force infrequently. • Police use of force typically occurs at the lower end of the force spectrum, involving grabbing, pushing, or shoving. • Use of force typically occurs when police are trying to make an arrest and the suspect is resisting. What we know with modest confidence: • Use of force appears to be unrelated to an officer’s personal characteristics. Such as age, gender, and ethnicity. • Use of force is more likely to occur when police are dealing with persons under the influence of alcohol or drugs or with mentally ill individuals. • A small proportion of officers are disproportionately involved in use of force incidents. What we do not know: • The incidence of wrongful use of force by police. • The impact of differences in police organizations, including administrative policies, hiring, training, disciple, and use of technology on excessive and illegal force. • Influences of situational characteristics on police use of force and the transactional nature of these events. The report notes that force appears to be rare; more research is needed on what factors are related to force; we still know little about the transactional nature of the police-citizen encounter, and so on. Unfortunately, this report tends to mirror the previous two rather than breaking much new ground, rehashing the same state of affairs. Admittedly, several studies highlighted in this latest report aim to improve our understanding of force, but we are still very much in the infancy of really pinpointing, in more precise terms, exactly how and why officers resort to force. Repeated citizen surveys, or analysis of citizen compliant records or use of force reports, simply cannot answer many of the questions that linger. For example, according to the report, previous research findings “...do not address the transactional, or step-by-step unfolding, of police-public encounters. Was suspect resistance the result of police use of force, or did police use force after experiencing suspect resistance'” (ix, 1999). This same conclusion was reached in the original report, which only serves to emphasize just how little we truly understand. We are not going to get any closer to the answer by conducting another survey, by allocating money for another researcher to look at use of force reports, or by doing yet another secondary analysis on citizen complaint data. It is clear that the unfolding of citizen and police actions within an encounter is recognized as a crucial link to a better understanding, yet researchers continue to rely on means that in no way can disentangle such behaviors adequately. On one hand the research community says it recognizes the need to do so, but on the other it continues to attempt to answer this question using inadequate data and inappropriate methods that simply cannot break down the dynamic process of a police-citizen communication method. Establishing the Need for the Police Training Curriculum The OJJDP conceptual framework suggests that (a) DMC may be present at multiple points of contact within the juvenile justice system, (b) DMC often results from multiple factors, and (c) the extent and the nature of DMC differ across jurisdictions (OJJDP, 2009). Given these considerations, OJJDP encourages jurisdictions to assess where DMC is present within the juvenile justice system and to identify DMC contributing mechanisms in a given community prior to selecting strategies for reducing DMC. Strategies are defined as appropriate if they target the contact points where DMC is present and if they address the identified contributing factors. In the state of Connecticut, the Juvenile Justice Advisory Committee (JJAC) had three studies conducted to assess minority over-representation and DMC in the juvenile justice system (Richetelli, Hartstone, & Murphy, 2009). The baseline study, conducted in 1991-1992, found that Black and Hispanic juveniles were overrepresented at several points of contact for police, court, and corrections. Some of these disparities were “neu- tralized” by legal and social variables, but many were not. (Hartstone & Richetelli, 1995). The second study, a reassessment—conducted in 1998-1999—found not only a decrease in baseline disparities but also a need to continue the work on reducing DMC (Hartstone & Richetelli, 2001). Of particular interest in this study were data from the earliest police decision points where most juveniles enter the juvenile justice system. The study reported the over-representation of minority juveniles in the juvenile justice system finding that, in 1998, Black and Hispanic juveniles accounted for only 22% of the State’s 10- to 16-year-old population, but 49% of the juveniles referred to court. Using data from a sample of police written incident reports, the study also determined that no differences were observed across race and ethnicity in police decisions to arrest and refer juveniles to court. To address this conflicting information, the JJAC recommended that a task group look into issues that would affect police handling of juveniles prior to, or in circumstances that would not result in, a written incident report. The third study, conducted in 2005-2006, was published in 2009 after the development of the police training curriculum and found continued disparities (Richetelli et al., 2009). This study is available at www.ct.gov/opm/dmc. The Police/Juvenile Task Group was established in response to the JJAC recommendation. It was comprised of volunteers from state and local law enforcement, the Police Officer Standards and Training Council, the juvenile prosecutor’s office, and the JJAC. It began its task in April 2004 and, after lengthy discussion, determined that the most effective way to accomplish its task of addressing law enforcement contact with youth prior to written incident reports was to focus on patrol officers. Referrals to court showed no evidence of DMC, and these decisions are often the responsibility of youth officers and supervisors experienced in working with youth. The discussion developed into a training focus on the interplay between line officers and juveniles when they meet in the community in nondangerous situations. Rather than diversity or cultural competency training, which was already available in the state, the group wanted to ensure that patrol officers had the knowledge they needed to differentiate problematic adolescent behavior at SEIR on December 12, 2012 from typical adolescent behavior, as well as skills to deescalate situations involving agitated or defiant youth and prevent situations from escalating in the first place. With a direction in mind, the Police/Juvenile Task Group requested technical assistance from the OJJDP to assist it in exploring police training strategies that may have been successful in other states. No strategies specific to patrol officers, DMC, and working with youth were identified. The task group then began the long task of developing its own curriculum— Effective Police Interactions With Youth. Effective Police Interactions With Youth Training Curriculum The Effective Police Interactions With Youth training curriculum is designed to reduce DMC by decreasing the likelihood that interactions between police officers and young people will lead to arrest, particularly for minority youth. The training targets the earliest point in the juvenile justice system—the initial contact between patrol officers and young people. The goals of the curriculum are to (a) increase patrol officer awareness of DMC, (b) increase patrol officer knowledge of youth behavior and strategies for interacting effec- tively with youth, (c) improve police attitudes toward young people, (d) increase the likeli- hood that police–youth interactions will have positive outcomes for youth, and (5) increase the likelihood that youth will respond positively toward police officers (Ouellette, 2006). The training program of 5.5 hours of instruction was designed for delivery by two certified police trainers in a classroom setting. Trainers use a variety of instructional techniques, including slide presentations, video clips, class discussions, small group activities, and individual activities. This creates an interactive environment that builds on participants’ existing knowledge and provides opportunities for them to share and learn from each other. Additional details of the training content are available in a videotape online (State of Connecticut Office of Policy and Management [OPM], 2008). Research Hypotheses The goal of this evaluation was to determine the effects of the Effective Police Interactions With Youth training curriculum on police officers’ knowledge and attitudes. To accomplish this goal, the following hypotheses were tested: Hypothesis 1: Training group participants will achieve higher knowledge test scores on posttest than on pretest. Hypothesis 2: Training group participants will achieve greater increases in knowl- edge test scores between pretest and follow-up than control group participants. Hypothesis 3: Training group participants will achieve higher scores on positively worded attitude items (and lower scores on a negatively worded attitude item) on posttest than on pretest. Hypothesis 4: Training group participants will achieve greater increases in attitude test scores between pretest and follow-up than control group participants over the course of testing. at SEIR on December 12, 2012. Method Evaluation Design An experimental design was used to evaluate changes in patrol officers’ knowledge of, and attitudes toward, youth following the 1-day training program. The State of Connecticut Office of Policy and Management (OPM) and the Juvenile Justice Advisory Committee (JJAC) invited the participation of local police departments. Participating departments were offered first choice of dates for the free patrol officer training with Police Officer Standards and Training Council Review Training Credit and eligibility to apply for the JJAC Police and Youth Grants for 2007/2008. Each participating local police department was required to provide a list of names of patrol officers with limited experience working with youth. The number of officers on the list was to be 20% of the department’s total number of certified officers. The department was to send half of the officers it listed to the training. The specific individuals to be trained were randomly selected by the evaluator. The department also had to have non-trained control group officers complete a pretest questionnaire and both the trainees and the control group officers had to complete a post test questionnaire several months after the training. The training and evaluation project had the support of the Connecticut Police Chiefs Association (CPCA). A total of 33 local police departments in Connecticut volunteered to participate. Training officers from each department submitted lists of patrol officers who were randomly assigned to be in either the training or the control group. The two groups were then compared with respect to knowledge and attitudes regarding typical adolescent behavior, police–youth interactions, and DMC. Knowledge and attitudes were measured using a questionnaire that was designed specifically for the purpose of this evaluation. The training group received the questionnaire on three occasions—once immediately before the training, once immediately after the training, and then 5 to 7 months after the training. The control group received the questionnaire twice—once during the pretraining period and again during the 5- to 7-month follow-up period. A total of 468 officers from various police departments throughout the State of Connecticut participated in this evaluation project. Measurement A survey instrument (police questionnaire) was developed specifically for the purpose of evaluating the Effective Police Interactions With Youth training program. The pretest version of this instrument contained 7 demographic questions, 19 knowledge questions, and 7 attitude questions. The demographic questions included name, department, gender, age, race and ethnicity, years of police experience, and years of experience as a youth officer. The 19 knowledge questions were multiple-choice and true-and-false questions designed to measure patrol officers’ knowledge about equal treatment of diverse youth in the juvenile justice system (5 questions), adolescent development (7 questions), and effective behavioral strategies for police interactions with youth (7 questions). The 7 attitude questions were designed to measure police officers’ attitudes regarding youth at SEIR on December 12, 2012pqx.sagepub.comDownloaded from LaMotte et al. 167 and police–youth interactions. Attitude questions were measured using a 10-point Likert scale, with 1 indicating strong disagreement and 10 indicating strong agreement. The post test version of this instrument contained 2 demographic questions, the 19 knowledge and 7 attitude questions from the pretest version of the questionnaire, and 5 open-ended items aimed at evaluating participants’ satisfaction with the workshop and their sense of its overall quality. The follow-up version of the questionnaire contained the same demographic, knowledge, and attitude items as the post test but did not include the satisfaction questions. A copy of the instrument is available online (Sanderson, Kosutic, Griggs, & Anderson, 2008). Development of the Survey Instrument The police questionnaire underwent careful development to ensure its suitability as a measure of change. Four members of the evaluation team attended one training session and read and reread the instructor’s manual for the Effective Police Interactions With Youth training curriculum. On the basis of their knowledge of the curriculum, evaluators constructed an initial pool of 89 multiple-choice and true-and-false questions designed to measure patrol officers’ knowledge of the three topics covered in the training curriculum. Of these, there were 36 questions about equal treatment of minority youth (Topic 1), 25 questions about adolescent development (Topic 2), and 28 questions about effective police–youth interactions (Topic 3). From the initial pool, team members selected 25 items that best covered each of the three major topics. In addition, team members developed 8 questions about patrol officers’ attitudes regarding youth and police–youth interactions. The questionnaire was then sent to the program developers, who made wording changes in items and response choices. Pilot Test 1. The first version of the questionnaire was pilot tested in a sample of 17 training participants. It was administered immediately before and immediately after the training. Following data collection, each knowledge item was examined in terms of response variability. Items with little (n = 5) and no (n = 2) variability were changed or dropped from the questionnaire. To test the utility of the pilot questionnaire as a measure of change, knowledge items were summed and converted to a total score, on the basis of the percentage of items answered correctly. Participants’ pre- and post-test scores were compared using a paired-sample t test. Results indicated that the average post test score was significantly higher than the average pretest score (M = 77% vs. M = 63%, t = -5.9, p < .001). In addition, average pre- and post-test scores on each of the attitude items were compared. Results indicated that participants’ average scores increased on all attitude items. However, differences between pre- and post-tests were statistically significant on 4 of the 8 items. Based on these findings, it was concluded that the questionnaire had good potential as a measure of change but that additional development was needed. Seven knowledge questions and/or response choices with little or no variability were modified, 11 items were eliminated, and 7 new items were added on the basis of feedback provided by training participants to the program developers. One attitude item was dropped from pqx.sagepub.com at SEIR on December 12, 2012. Downloaded from 168 Police Quarterly 13(2) the questionnaire. Finally, 3 closed-ended and three open-ended questions were added to the post test questionnaire to assess participants’ reactions to the training. Pilot Test 2. The second version of the questionnaire was tested in a sample of 28 training participants. As with the initial pilot test, the questionnaire was administered immediately before and immediately after the training. Data analysis indicated that all knowledge items had adequate variability. Pre- and post-test knowledge questions were summed and converted to total scores on the basis of the percentage of items answered correctly. Participants’ pre- and post-test scores were compared using a paired-sample t test. Results indicated that the average post test score (M = 76%) was significantly higher than the average pretest score (M = 46%, t = -11.6, p < .001). Average pre- and post-test scores on each of the attitude items also were compared. Post test scores were significantly higher for 5 out of the 7 attitude items (p = .005). These results led to a conclusion that the questionnaire was an effective measure of change. Following the second pilot test, the program developers suggested that 3 knowledge items be dropped and 1 new knowledge item be added. Therefore, the final version of the questionnaire included 19 knowledge items and 7 attitude items. The posttest version also included 5 satisfaction items. Identifying Youth At Risk of Delinquency Seasoned law enforcement officers in departments around the country have come to recognize early warning signs for later delinquency. Responding to calls, officers enter homes where youth have been ex- posed to domestic violence, drug and alcohol abuse, gang activity, neglect, and other criminal behavior. Officers see youth who have been exposed to crime and violence on the streets, in their schools, and among their peers. Many experienced officers know delinquent youth whose first encounters with law enforcement were as victims of crime or as family members of someone who was arrested. Officers frequently recognize that such victimization experiences and exposure to criminal and delinquent fam- ily members are related to later offending. Current research on the risk factors that distinguish youth who are more likely to become involved in delinquency from those who are less likely to do so con- firms and expands on what some law enforcement officers already know. Risk factors can be defined as conditions in the environment or in the individual that predict an increased likelihood of developing delinquent behavior (Brewer et al., 1995). Risk factors for delinquency and violence are generally described in five categories: community, individual, peer group, school related, and family (Brewer et al., 1995; Hawkins et al., 1998). Community risk fac- tors include poverty, physical deterioration, availability of drugs, and high crime rates. Individual risk factors include child- hood hyperactivity, aggressiveness, and risk taking. Peer group risk factors in- clude association with a peer group that has favorable attitudes toward delin- quency and gang membership. School- related risk factors include early and persistent antisocial behavior and academic failure. Finally, family risk factors include family conflict, family management problems (e.g., failure of caretakers to set clear expectations, lack of supervision, and excessively severe punishment), and favorable attitudes toward and involve- ment in crime and violence (for further discussion of risk factors for delinquency see Gottfredson and Polakowski, 1995; Howell, 1997; Hawkins et al., 2000). OJJDP’s longitudinal, prospective re- search on the causes and correlates of delinquency has found that delinquency and violent behavior stem from the accumulation and interaction of risk factors in the five categories described above (Thornberry, Huizinga, and Loeber, 1995; Hawkins et al., 1998). The probability of violence and delinquency increases (sometimes dramatically) with increases in the number of risk factors (Office of Juvenile Justice and Delinquency Prevention, 1998). For example, a study of 411 South London boys found that the percentage of boys convicted for violence more than doubled in the presence of 1 risk factor but increased tenfold in the presence of 4 or 5 risk factors (Farrington, 1997). Researchers have also identified a number of protective factors that provide a buffer against risk factors (Hawkins, Catalano, and Miller, 1992; Thornberry, Huizinga, and Loeber, 1995). These include individual factors (e.g., high intelligence and positive social orientation), factors related to social bonding (e.g., supportive relationships with family members or other adults), and healthy beliefs and clear standards of behavior (e.g., norms that oppose crime and vio- lence). Because protective factors also tend to have cumulative effects, youth who have or are exposed to a large number of protective factors show greater resilience in coping with the risk factors in their lives than do those with fewer protective factors. Although the understanding of risk and protective factors is increasing, questions remain about how police and sheriff’s departments can best use this information. Law enforcement administrators who want to prevent delinquency may be discouraged by the perceived practical difficulties of coordinating a prevention program, especially because most departments are already very busy just responding to calls for service. Administrators at the Westminster, CA, Po- lice Department considered these issues when they created the SHIELD program. Instead of designing a program in which services are delivered directly by the police department, they developed a co- ordinated mechanism that uses a multidisciplinary team to identify at-risk youth and connect them to existing services in the community. Westminster has not eliminated every deficit in services for at-risk youth. For example, services that target non-English speaking youth in a culturally appropriate way are still needed, and treatment options for children under age 5 remain limited. Nevertheless, by even identifying needs that it cannot immediately fulfill, SHIELD has allowed Westminster to begin working on solutions for affected youth. Police Roles and Responsibilities Police officers are the most visible officials in the criminal justice system. They introduce citizens to the justice process. That introduction ranges from taking a report from a victim or witness to a crime, to issuing a traffic citation, to questioning or taking into custody a suspect in a misdemeanor or felony offense. Police are charged with preventing crime and enforcing the law. They are given the authority to make arrests, to use reasonable physical force when necessary, and to take persons charged with crimes into custody. Society entrusts a great deal of authority to police, but also expects a lot from them. Police are expected to provide public order and safety; to prevent crimes from occurring, and find and apprehend offenders when crimes occur; and to perform a variety of law enforcement functions without violating con- institutional rights. In reality,traditional police patrol does little to prevent crime. Police in most cases react to crime after it has already happened, responding to citizen calls, reporting to crime scenes, conducting investigations, and tracking and apprehending offenders. The fact that police are called upon for many services besides law enforcement makes their job even more difficult. Police officers actually perform three roles in fulfilling their law enforcement responsibilities: law enforcement, order maintenance, and service (Wilson, 1968). The public and the police themselves have viewed the law enforcement function as the primary and most important task, and little attention was given to the others, which were considered less important and not “real law enforcement.” Law Enforcement The traditional law enforcement role of police is to detect and investigate crimes and to apprehend those responsible for committing crimes. Police attempt to detect crimes through regular police patrols and by responding to complaints of victims and statements of witnesses. The traditional law enforcement role gives police visibility to the public as they “protect and serve.” There are some additional challenges in policing crimes such as drug dealing, gambling, and prostitution, where there are no witnesses or clearly identified victims. To enforce laws against the so-called victimless crimes (or more appropriately termed “consensual” crimes, because persons involved are willing participants), police work as undercover officers to detect the crimes and make arrests. The law enforcement role includes enforcement of traf- fic laws and parking violations, and it is here that officers have the most interaction with the general public as law enforcers.To finalize their law enforcement role and ensure that suspects are brought to trial, police engage in interrogation of suspects, collection of physical evidence at a crime scene, and presentation of the evidence in court. Order Maintenance The order maintenance function of police involves crowd control during events such as parades, large public gatherings, music concerts, and sports competitions in indoor and out- door stadiums; and patrolling on foot, on bicycle, on horseback, or in vehicles and on streets, sidewalks, and in public parks. The order maintenance function parallels the law enforcement role when officers intervene to control disorderly behavior. The order maintenance role is less clear (both to the public and to many police) than the law enforcement role, mainly because the behaviors being controlled are less clearly defined. “Disorderly behavior,” for example, generally refers to behavior that disturbs the public peace,but the exact definition and an officer’s determination whether the behavior warrants official intervention depend on the neighborhood location and the time during which the disturbance occurs. The officer’s role may be that of telling participants of a loud party to quiet down or dispersing a group of juveniles who are loitering on a street corner or in front of a business establishment. Service Function The third role of police is that of providing services to the public. This may include providing aid or assistance to persons in need,such as calling a tow truck for a stranded motorist; transporting abandoned or neglected children to a hospital or shelter facility; delivering a baby whose mother did not make it to the hospital on time. The service function often results in a combination of functions, such as when one officer transports abandoned children to a shelter,and another officer locates the parent(s) and initiates a child abuse investigation (a law enforcement function). The service function more recently has come to include an educational component, such as when police are assigned to schools to assist in the education of children and youth on the dangers of drugs and how to avoid drug abuse. The three primary roles of police are very different on a number of dimensions: criminal versus noncriminal, urgent versus routine, and dangerous versus relatively safe (Dorne & Gewerth, 1995). Police officers generally view the law enforcement function as the primary role, while order maintenance and service tasks have been typically regarded with mixed feelings, ranging from ambivalence to disdain (Moore, 1992). Police officers hold varying opinions of the importance of each of the roles, and they do not undertake these three functions with equal degrees of enthusiasm. They are given considerable autonomy and independence in carrying out their law enforcement roles and are allowed to place greater or lesser importance on a given role depending on their assigned patrol area and individual circumstances. In a reading in this section, Susan Guarino-Ghezzi recommends that police reexamine the order maintenance and law enforcement roles as they pertain to juvenile offenders and victims. Juvenile Offending and Police Discretion Police have considerable discretionary power in handling juvenile matters, ranging from reprimand and release, to transporting a juvenile to detention and referral to juvenile court. Discretion is important in police work, for the officer’s decision to intervene in any suspected law violation is the first stage in the juvenile justice process. Officers use their discretion in deciding whether or not to take official actions with offending juveniles or simply order them to “move on,”“break it up,”or “get on home.”Most police contact with juveniles is nonofficial, and police make an arrest and take juveniles into custody in only a small percentage of cases. In a study of police responses with juveniles in two cities,Myers (2002) found that police took juveniles into custody in only 13% of their encounters with juveniles. Most of the police– juvenile encounters involved noncriminal matters, such as public disorder (22%), traffic offenses (14%), nonviolent conflicts (9%), and suspicious situations (7%); and about one fourth (27%) involved violent or nonviolent crimes (Myers, 2002, p. 123). In 2003, 20% of juvenile arrests were handled within law enforcement agencies, 71% were referred to juvenile court, and 7% were referred directly to criminal court. The remaining 2% were referred to a welfare agency or to another police agency. The proportion of arrests referred to juvenile court increased from 1980 to 2003, from 58% to 71% (Snyder & Sickmund, 2006, p. 152). Police discretion has been criticized because some believe that police abuse their broad discretionary powers, and that they base their decisions on extralegal factors other than the offense. Extralegal factors such as sex, race, socioeconomic status, and individual characteristics of the offender have been shown to make a difference in police officers’ decisions of whether or not to take official actions. Girls are less likely than boys to be arrested and referred to juvenile court, but they are often referred more than boys for status offenses such as running away or disobeying parents (Armstrong, 1977; Chesney-Lind, 1977). Researchers have reported differing results on the importance of race in police discretion. Some studies report few differences when controlling for offense seriousness and prior record. African American and other minority youths seem to be involved in more frequent and serious offenses than whites, so it is difficult to determine whether they are singled out more by police for official action. There is some evidence of racial bias, however, as minority youths have often been targeted more by police for official intervention (Wolfgang,Figlio,& Sellin,1972,p.252).Some critics of police discretion also contend that lower-class youths are processed into the justice system for the same offenses for which middle- or upper-class juveniles are simply reprimanded and released to their parents. Police and juvenile officers justify this use of discretion on the basis that middle- and upper-class youth are more likely to be corrected without refer- ral to the justice system because their parents have the resources to provide their children with the necessary supervision and corrective services. Merry Morash (1984) found that an older juvenile with a prior record and who fits the image of a serious delinquent is more likely to be referred by police to the juvenile court. A juvenile’s demeanor and attitude make a difference in a police officer’s use of discretion.A youth who is polite and respectful is more likely to get off with a reprimand, while a negative and hostile attitude is likely to result in a court referral (Lundman, Sykes, & Clark, 1990; Piliavin & Briar, 1964). y Race as a Factor in Juvenile Arrests The issue of race is a major concern in the criminal and juvenile justice systems. It is an undisputed fact that racial and ethnic minorities (especially African Americans) are disproportionately represented at each stage of the system: in police arrests, in jails and detention centers, in courts, and in correctional facilities. Research studies are mixed, however, as to whether that disproportionate representation is a result of racial bias in police arrest,prosecutors’ decisions, and judicial sentencing (Conley, 1994; Wordes, Bynum, & Corley, 1994). African American youth are overrepresented in juvenile arrests when compared to their proportion of the population.Black youth,who accounted for 17% of the juvenile population in 2005,were involved in a disproportionate number of juvenile arrests for robbery (68%), murder (54%), motor vehicle theft (43%), and aggravated assault (42%) (Snyder, 2007, p. 9). The question is whether the over-representation of black juveniles in police arrest rates is due to racial bias, or to the greater involvement of black youth in violent crimes. Violent crimes are more likely to be reported, detected, and result in a police arrest. To answer this question, Pope and Snyder (2003) analyzed National Incident-Based Reporting System (NIBRS) data from law enforcement agencies in 17 states,with a large sample of 102,905 juve- nile offenders. They found no significant effects of race in police arrest decisions, and they were able to identify some characteristics that differentiated the crimes of white and nonwhite juvenile offenders. Compared to nonwhites, white juvenile offenders were: • Less likely to have multiple victims • More likely to act alone • More likely to commit crimes indoors • Less likely to possess a non-personal weapon (firearm, knife, or club) • Less likely to offend against adults • Less likely to offend against members of another race • More likely to commit crimes against family members; equally likely to commit crimes against acquaintances; but less likely to commit crimes against strangers (Pope & Snyder, 2003, p. 4). The findings revealed that the crime incident characteristics that increased the odds of arrest for violent crimes were largely the same for white and nonwhite offenders,with one important exception: Victim’s race was correlated with arrest probability for nonwhite juvenile offenders, but not for white offenders. A nonwhite juvenile offender therefore was more likely to be arrested if the victim was white than if the victim was nonwhite. More research must be conducted on police arrest patterns, using larger samples that are more representative of the nation. Arrest patterns may differ among states and within regions of states and of the nation. Race and ethnic background may be a factor in police decisions to arrest juvenile offend- ers, but based on research evidence it is clear that several other factors influence officers’ deci- sions. In summary, the factors that may affect police officers’ decisions to arrest a juvenile or to take less formal actions without court referral include factors relating to the: • Offense (seriousness, type, time of day, gang related, use of weapon) • Youth’s record or status (prior police contact or arrest, school record, probation status) • Offender (age, gender, race, social class, demeanor) • Complainant (present at the scene, desire to prosecute, age, gender, and race) • Location of the offense (type of neighborhood, low- or high- crime area) • Parents (attitude, present at the scene or at home, concern, and ability to supervise) • Officer (training and experience, view of justice system and diversion, workload) • Police department (enforcement policies, community policing, or problem-solving emphasis) (Sanborn & Salerno, 2005, pp. 137–139). Police discretion is necessary, and the juvenile justice system could not function without some use of discretion. Juvenile courts in urban areas have a backlog of cases, probation offi- cers’ caseloads are too high for them to provide adequate supervision, and correctional facilities are becoming overcrowded. The system must concentrate on those juvenile offenders who pose the greatest risk and need official intervention to prevent further offending. Alternatives to Police Arrest and Custody A police officer may refer a minor offender to a youth services bureau, a community agency such as a Big Brother Big Sister program, or a similar delinquency prevention program. In the majority of cases where police have reason to believe that a juvenile has committed an offense,the youth will be taken to the police department juvenile bureau for questioning, may be finger- printed and photographed, and then will be taken to the intake unit of the juvenile probation department where a decision will be made to detain the youth or release to the parents. • Questioning, warning, and release in the community. The least severe sanction is when an officer questions a youth for a possible minor offense, and gives a warning and reprimand on the street without taking formal actions. • Station adjustment. Police may take a youth into custody and to the station, record the alleged minor offense and actions taken, give the youth an official reprimand, and release the youth to the parents. The parents are generally contacted first and may be present when the youth is reprimanded. In smaller cities the youth may be placed under police supervision for a short period of time. • Referral to a diversion agency. Police may release and refer a juvenile to a youth service bureau (YSB), Big Brother Big Sister program, runaway center, or a mental health agency. Diverting minor offenders from the juvenile justice system to a YSB that provides counseling and social services is considered preferable for many first-time offenders and troubled youth. • Issuing a citation and referring to juvenile court. The police officer can issue a citation and refer the youth to juvenile court. The intake probation officer accepts the referral,contacts the parents if the police have not already done so, and releases the youth to the parents on the condition that they will report to the court when ordered to do so. The intake officer then determines whether a formal delinquency petition should be filed. In some states the decision is made by the prosecuting attorney assigned to the juvenile court. • Taking to a detention center or shelter home. The police officer can issue a citation, refer the youth to the juvenile court, and take him or her to a detention center. The intake officer at the detention center then decides whether to hold the juvenile or release him or her to the parents. Juveniles are detained when they are considered dangerous, when there is a lack of parental supervision, or when there is a high probability that they will not report to the court when ordered to do so. If a detention center is felt to be too restrictive, and an appropriate parent or foster home is not available, the youth may be placed in a shelter care facility that might be either a private home or a group home. Most states now provide for a detention hearing within a day after the youth’s referral in which a judge or referee must determine whether there is sufficient reason to continue to detain the juvenile. In cities without a separate juvenile detention center, juveniles who cannot be released to their parents are confined in a separate section of the county jail, or may be transported to a juvenile facility in another county. There has been a national effort to remove juveniles from adult jails. Removing juveniles from their homes and detaining them in juvenile centers is considered a last resort.