CONTEMPORARY GUN PO LICY AND POLICING: AN ANALYSIS OF POLICE CADET ATTITUDES TOWARD CONCEALED WEAPON PERMITTING LAWS By MATTHEW ROBIN NOBLES A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ARTS UNIVERSITY OF FLORIDA 2005
Copyright 2005 by Matthew Robin Nobles
This document is dedicated to Angela, w ho was with me every step of the way.
ACKNOWLEDGMENTS I would like to thank the members of my supervisory committee, Dr. Alex Piquero, Dr. Lonn Lanza-Kaduce, and Dr. Brian Stults, for their vast wisdom and patient assistance throughout the process. I could not have asked for more thoughtful, proficient, or effective guidance for this project. I would also like to recognize Captain Jim Murphey, Associate Director of Criminal Justice Programs, and Chief Daryl Johnston, Director of the Institute of Public Safety at Santa Fe Community College, for their kind support. iv
TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES............................................................................................................vii ABSTRACT.....................................................................................................................viii CHAPTER 1 LITERATURE REVIEW.............................................................................................1 Introduction: A History of Controversy.......................................................................1 Theoretical Origins of Gun Possession.........................................................................3 Concealed Guns: Deterrent or Menace?.......................................................................7 Gun Oriented Policing................................................................................................10 The Role of Police Discretion.....................................................................................14 Policy and Enforcement Challenges...........................................................................18 2 DATA AND METHODOLOGY...............................................................................20 Sample........................................................................................................................20 Method........................................................................................................................20 Dependent Variables...................................................................................................21 Direct-Response Dependent Variables................................................................21 Vignette-Response Dependent Variables............................................................23 Independent Variables................................................................................................24 Other Control Variables..............................................................................................24 3 RESULTS...................................................................................................................26 Discussion...................................................................................................................33 Limitations...........................................................................................................35 Future Research...................................................................................................37 APPENDIX A PARAMETER ESTIMATES FOR MULTINOMIAL LOGISTIC REGRESSION MODELS....................................................................................................................40 v
B POLICE CADET SURVEY QUESTIONNAIRE......................................................45 LIST OF REFERENCES...................................................................................................49 BIOGRAPHICAL SKETCH.............................................................................................53 vi
LIST OF TABLES Table page 1: Descriptive Statistics......................................................................................................25 2: Overall Regression Model Fitness Tests.......................................................................29 3: Independent Effects in Reduced Stepwise Regression Models.....................................30 4: Independent Effects of Cadet Attitudes on Decision to Search.....................................33 5: Independent Samples t-Test Grouped by Survey Date..................................................38 A-1: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q15..41 A-2: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q16..41 A-3: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q25..42 A-4: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q26..42 A-5: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q29..43 A-6: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q24..43 A-7: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q24 with Additional IVs..................................................................................................44 vii
Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Arts CONTEMPORARY GUN POLICY AND POLICING: AN ANALYSIS OF POLICE CADET ATTITUDES TOWARD CONCEALED WEAPON PERMITTING LAWS By Matthew Robin Nobles August 2005 Chair: Alex R. Piquero Major Department: Criminology, Law and Society The academic literature regarding the overall impact of concealed weapons on crime is decidedly mixed. Nonetheless, as a matter of public policy, state legislation authorizing so-called shall issue concealed weapons permits has become politically popular in the past decade. While scholars and lawmakers struggle to answer the larger questions of the effects that guns have on violence, police officers are required to confront the practical problems of enforcement on the street every day. A number of high-profile law enforcement initiatives have attempted to combat illegal guns with specialized patrols. The present study surveys attitudes of law enforcement cadets at a small academy in order to assess their relative support for concealed weapon permitting laws: do police cadets support the legislation, and what impact might these attitudes have on issues of public safety? Influence of police subcultures and the attitude-behavior link are discussed. Limitations and future research questions are addressed. viii
CHAPTER 1 LITERATURE REVIEW Introduction: A History of Controversy Few topics elicit a more polarized reaction in America than that of gun control measures. American history and culture was established and nurtured on conflict, from the earliest European settlers conquest of Native Americans to our countrys emergence as a dominant superpower through two World Wars. In some ways, the American public regards guns very positively: they are viewed as necessary instruments through which such principles as freedom, honor, and civic responsibility are achieved and maintained. Yet undoubtedly, guns figure prominently in the commission of many types of violent crime and are denounced by some as the means by which a new public health epidemic has emerged: homicide is the leading cause of death for young black males in our country. As a matter of public policy, gun control is fervently defended by some and just as passionately decried by others. The implications of gun control policy are no less serious than life and death an irresponsible approach to the problem of gun proliferation has the potential to physically endanger or protect millions of Americans, depending upon which ideological camp is presenting the empirical evidence. Unfortunately, the politicians who directly control gun policy require quick answers that are popular with constituents, which lead to questionable policies based on questionable causal theories on how guns and crime are related. 1
2 The academic community has never reached a solid consensus on the empirical link between guns and crime, in part because the problem is so complex, and in part because the implications are so controversial. Much of the research that seeks to address issues of guns and gun control, however nobly undertaken, seems to become immediately vulnerable to political attack in the media from whichever special interest group is opposed to the findings. Both sides are involved in a propaganda war supporting their particular beliefs, which affects voters and lawmakers, ultimately shaping policy. There are only a precious few points on which both sides agree; one of them is that something must be done to address violent crime in America. Since private gun ownership seems inexorably linked with public policy on gun violence, any initiatives must consider both legal and illegal guns in the equation. Therefore, if the objective of legislatures is to reduce violent crime by dealing with guns, then the most balanced policy approach is one in which any potential value of legal, defensive gun possession is preserved while illegal firearms and those who broker or utilize them are efficiently targeted by law enforcement. Questions remain about how the political and legislative ideals mesh with the practical realities of law enforcement. Do police officers approve of concealed weapon permit laws because of their perceived deterrent value, or do they disapprove because the law requires more work and may present an additional element of danger? Are these attitudes stable over time, or do they vary with street experience? Do these attitudes influence the decision to stop a suspect, perform a search, or make an arrest? The objective of this study is to examine the relationship between concealed weapons permitting laws and police attitudes. Using pilot survey data from police cadets in a mid
3 sized southeastern city, we can lay the groundwork for an investigation of the interplay between police and contemporary gun policy. Theoretical Origins of Gun Possession Considerable scholarship has been devoted to issues of gun ownership. Understanding the psychological and sociological underpinnings of the gun phenomenon is a complex endeavor with many disparate causes. Any meaningful critique of current gun policy in America must consider the various uses and abuses of guns on the part of the public at large. Where does the impetus to own or carry a gun really originate? How are the processes different for legal vs. illegal ownership? Are these processes stable over time, and how strong are contextual influences? Before understanding why gun ownership (and to a lesser degree, gun carrying) behaviors are so prevalent in the United States, it may be useful to examine the demographic profile of a typical gun owner. Gun ownership can serve as a dependent or independent variable according to the research question, but for the purposes of assessing attitudes, gun ownership may be reasonably assumed to be correlated with pro-gun opinions. Survey data from the early 1970s described the average gun owner as Southern, rural, white, Protestant, and upper-class (Wright & Marston, 1975). Few subsequent studies seem to have seriously challenged or fundamentally changed these basic classifications. More recent inquiries have drawn conclusions about gun ownership and attitudes between males and females, with evidence indicating that males are more likely to hold pro-gun beliefs in terms of the right to possess guns, the role that guns play in stimulating crime, and the ability of a gun to protect its owner from victimization (Branscombe et al., 1991). Comparisons of gun control attitudes between countries that have very strict gun laws (United Kingdom) and very permissive gun laws (United
4 States) shows that similar attitudinal gender effects are present despite the context and political landscape (Cooke & Puddifoot, 2000). In a study that deals specifically with attitudes related to concealed weapons, Kleck (1996) finds that crime rates and prior victimization does not predict support for gun permits. Interestingly, he shows that greater support for gun permits occurs among liberals, Jews, people with more education, people with higher income, and people who do not own guns or hunt. These contradictions make synthesis and interpretation of the respective findings difficult. Models of gun ownership vary with respect to the theoretical and practical factors being examined. Most investigators seem to agree that there is no single theory that can explain all gun-related behaviors. Cao, Cullen, and Link (1997), for instance, offer an empirical test of six main explanations for gun ownership. They focus on the distinction between firearms owned for sport and those owned for protection, a potentially important distinction with regard to gun-carrying behavior. In their study, they specify six models of ownership that are not mutually exclusive: economic resources, socialization, racial loathing, fear and victimization, collective security, and conservative ideology. In the economic resources approach, firearms ownership is a result of the availability of the financial means to do so. The socialization model suggests that being indoctrinated into a gun culture, being from the South, or being highly educated may influence your attitude toward guns. Racial loathing frames the issue in terms of white vs. non-white power control. Fear and victimization concerns the perceptions about crime and feelings of vulnerability. Collective security describes succinctly an individuals faith in police and the criminal justice system to maintain civil order.
5 Finally, conservative ideology and related political affiliations have proven to be a significant predictor of gun ownership in past studies. Using survey data, Cao and colleagues find a number of results that they suggest are surprising, including an inverse relationship between gun ownership and prior victimization. With respect to the possession of guns for protective vs. sport reasons, they demonstrate that individuals who own guns for protective purposes are influenced by the socialization model, introduction into a gun subculture, and crime-related attitudes and feelings about safety, while sport gun owners are influenced likewise by the socialization model and gun subcultures. The influence of peers is an important factor that has been examined in prior gun research. Most approaches to peer influence regarding guns take the form of a subcultural explanation for how individuals become acclimated to favorable definitions of guns. In terms of criminological theory, Akers (1973) stated that social learning consisted of differential association, definitions, differential reinforcement, and imitation. These concepts can be utilized to explain any behavior, not just criminal or deviant behaviors. In this case, differential association occurs within the context of peers or family members who own and use guns. Definitions favorable to gun ownership and possibly imitation are followed by differential reinforcement. The indoctrination into a gun subculture is not difficult to understand with respect to family members or friends who share similar hobbies (e.g., hunting). Social learning variables have seen significant empirical attention in the literature. Lizotte and Bordua (1980) write extensively on the topic of subcultural theories of gun ownership. Their regression analyses include a number of social learning variables,
6 including whether members of the family hunt or shoot together. They conclude that gun ownership for sport is indeed subcultural in nature, but protective ownership is not. Oddly, they further discover that sport and protective gun ownership is apparently an independent phenomenon, with no joint occurrence. Findings from Cao et al. (1997) diverge from this conclusion, stating that both sporting and protective gun owners were influenced by social learning variables. Social learning and the role of peers can also be applied in the context of illegal gun behaviors. Dealing specifically with carrying guns rather than ownership, Lizotte et al. (2000) demonstrate that gang membership predicts gun-carrying among young adolescents, while illegal peer gun ownership and involvement in drugs predicts gun-carrying among older adolescents. Participation in a criminal subculture may also lead to higher incidence of crimes where guns are the subject rather than simply a facilitator, including crimes such as gun theft. Wright (1986) concludes from his survey of incarcerated felons that the criminal handgun market is overwhelmingly dominated by informal transactions and theft as mechanisms of supply. Wrights findings specify that only about one-sixth of handguns used in the commission of a felony were legally acquired, demonstrating that the problem of controlling criminal access to concealable guns is rather substantial. Investigation into the gun phenomenon has also produced some intriguing findings with respect to potential risk factors for criminal offending. Schwaner et al. (1999) provide an examination of factors predicting the decision to become licensed to carry a concealed weapon in the state of Kentucky. In their analysis, they use demographic variables such as age, sex, and education to predict which factors influence
7 the decision for individuals to acquire a license. The results show a negative and significant relationship between the dependent variable (decision to become licensed) and the independent variables (age, education, sex [0=male], and household size). The only variable to show a moderated effect in their model was the respondents level of income. Interestingly, however, Schwaner and colleagues demonstrate that the same four factors also predict an intervening variable, heavy drinking, that has a positive and significant (p < .001) effect on the decision to become licensed. In other words, people who reported drinking more were more likely to be licensed to carry a concealed weapon in this sample, presumably increasing weapon availability for these individuals across a range of situations. While this may be only an artifact of other demographic variables (e.g., young males are likely to be higher in both measures), it raises interesting questions about selection effects and certain correlates of gun possession Concealed Guns: Deterrent or Menace? The inspiration to possess firearms for personal and property defense is not new, and scientific investigation on the potential value of legal firearms is fairly positive and conclusive in showing that guns are widely used in real-world defensive scenarios. Results from the National Self-Defense Survey put the estimated number of defensive gun use incidents at around 2.5 million per year, many of which are reported to be located in close proximity to the victims home (Kleck & Gertz, 1995). As for their effectiveness, Kleck and Gertz (1998) observe, the empirical literature is unanimous in portraying defensive gun use as effective, in the sense that gun-wielding victim are less likely to be injured, lose property, or otherwise have crimes completed against them than victims who either do nothing to resist or who resist without weapons.
8 The issue of guns in public places is much more contentious. A great deal of scrutiny lately has monitored the issue of right-to-carry (RTC) laws, statutory provisions that remove discretionary powers on the part of local law enforcement in denying individuals a permit to carry a concealed handgun, absent a compelling legal reason to the contrary. Research by Lott and Mustard (1997) contributed a very controversial finding to the gun policy debate. Using cross-sectional time series data, the authors were able to demonstrate quantitatively that passage of RTC laws worked to reduce the rates of homicide, rape, robbery, and aggravated assaults (more guns = less crime). In theory, the drop in crime was attributable to a deterrence model: RTC laws passed More carry permits issued More guns legally carried Greater incidence of defensive gun use Increase in general deterrence Not surprisingly, scholars in various fields have carefully and thoroughly debunked Lott and Mustards findings with regard to the impact of RTC laws on violent crime. Many subsequent publications found problems with Lott and Mustards original data, which has since been extended, replaced, or otherwise improved; others faulted the statistical methods behind Lott and Mustards results, claiming that coding errors substantially influenced their regressions (Donohue, 2003)1. Some investigators manage to temporarily sidestep the politics of the issue and focus on the empirical questions at hand. A most compelling analysis refuting Lott and Mustards position comes from Kovandzic and Marvell (2003), who used new county-level data on actual concealed-handgun permit holders (compared to Lott and Mustards data on permits issued) to conclude that using the most direct measure of lawful gun 1 Donohues reaction essay (2003) also calls into question Lotts professional ethics, in effect accusing him of promoting a pro-gun agenda with complicity from the NRA and commenting that Lott is suspected of falsifying survey data dealing with rates of defensive gun use.
9 carrying readily available, the number of citizens with valid concealed-handgun permits, we find no credible statistical evidence that permit rate growth (and presumably more lawful gun carrying) leads to substantial reductions in violent crime, especially homicide. Based on the work done since Lott and Mustards initial study, RTC laws have been shown, at best, to have no effect on crime. At worst, they may slightly escalate rates of violence and increase likelihood of lethality. For their part, lawmakers in many jurisdictions seem to be siding with Lott and Mustards interpretation of the role that RTC laws play. At present, 44 states have some type of concealed weapon permitting process, while 34 of those have RTC laws in effect. Additionally, some small municipalities have apparently taken the concept to its logical extremes by mandating gun ownership through local ordinances. Tiny Geuda Springs, Kansas, population 212, made national news when it began requiring residents to own firearms in November 2003. Regardless of the potential deterrent value of guns, pro-gun advocates argue that gun possession by lawful citizens should not be restricted as a matter of principle, and this attitude has become very pervasive thanks in part to a powerful lobby backed by the National Rifle Association. The NRA has succeeded in promoting private gun ownership as a Constitutional entitlement, although scholarly opinion on the applicability of the Second Amendment to private citizens does not always support that interpretation (see Cress, 1984, for a thorough historical examination). In theory, lawful gun owners with legal guns are not the source of the epidemic of gun violence (Lugwig, 2003); in fact, existing research suggests that defensive gun possession can be an asset to citizens and police alike (Kleck & Gertz, 1995). Therefore, in order to reduce violent gun crimes, gun
10 policy must be balanced to combat illegal guns on the street while continuing to tolerate possession of legal guns. Gun Oriented Policing Despite the intellectual tug-of-war in the academic and political arenas, both the pro-gun and the anti-gun camps agree on the basic tenet that guns should not be generally available to certain classes of people, including adults who have a history of violent offending and unsupervised minors.2 To this end, policing strategies have evolved to target guns and gun law violators as a proxy for violent offending, mostly in urban settings with predominately minority populations where higher rates of violent crime exist. If the supply-side logic of gun crime holds true, more seizures of illegal guns and more arrests of illegal gun possessors will lead to a decrease in the quantity of illegal guns involved in street-level acts of violence, similar to a public health model in which disease vectors are controlled to prevent contagion. Consequently, some police forces have developed a strategy of gun oriented policing, which integrates problem-oriented and community-oriented policing with special attention to illegal weapons. The first widely recognized application of gun oriented policing (GOP) was the Kansas City Gun Experiment in the early 1990s. The design was modeled after the body of work examining directed police patrols in geographic hot spots identified as having high crime rates. Police presence was increased in areas reported to have the highest crime rates, and officers were authorized to concentrate fully on patrols with special emphasis on finding illegal guns, eliminating the officers responsibility to respond to 2 Federal and state laws impose restrictions on possession of guns generally and concealed weapons specifically for several general classes of people, including individuals wanted on outstanding warrants, juveniles, illegal aliens, the mentally ill, individuals with prior convictions for domestic violence, and those with a documented history of drug or alcohol abuse.
11 routine service calls. If police could control illegal guns in the high-crime areas, they theorized that they could affect the rates of gun-related violence (Sherman & Rogan, 1995). The theory behind the Kansas City initiative functioned on two levels. First, increased visibility of police would provide general deterrence while aggressive enforcement and targeting high risk individuals would raise the number of police encounters, thereby causing an increase in specific deterrence for those deemed more likely to engage in violent crime. Second, the GOP technique worked on supply-side incapacitation basis: fewer illegal guns meant less gun crime because the enabling mechanism was no longer as readily available. Results from the study were highly encouraging. The increased police patrols resulted in a 65% increase in seizures of illegal guns within the high crime districts and the same areas saw a 49% decrease in violent gun crime, while the control group reported fewer gun seizures and a slight increase in crime (Sherman & Rogan, 1995). A replication of the Kansas City experiment was conducted in two separate beats in Indianapolis in 1997. Many design aspects were similar: the Indianapolis police increased traffic patrols in high-crime areas and began targeting individuals who were believed to be high risk gun offenders using stop-and-frisk techniques. As for the targeted districts, one predominately minority neighborhood and one predominately white neighborhood, both high in violent crime, were chosen for the experimental GOP patrols. The results of the Indianapolis replication were mixed. One of the beats reported similar results to the Kansas City experiment in terms of the effects of new gun-specific patrols on homicide, armed robbery, and aggravated assault, while the other beat reported
12 no effect from the new patrols. Investigators hypothesized that differences in the policing techniques (e.g., more traffic stops vs. more stop-and-frisk encounters) or the racial composition of the targeted districts could account for the discrepancy (McGarrell et al., 2001). Nonetheless, the results from Indianapolis demonstrate at least partial support for the potential of GOP strategies to impact violent crime rates. Perhaps the best-known implementation of a GOP strategy occurred in New York City in the late 1990s. Beginning in 1994, NYPD Commissioner William Bratton instituted a widely publicized, multi-pronged zero tolerance campaign designed to address problems of violent crime. The NYPD approach included a specialized Street Crime Unit tasked with aggressive enforcement of existing gun laws and active pursuit of illegal weapons. NYPD made over 46,000 arrests on gun-related offenses and confiscated more than 56,000 guns from 1994 to 1997. The results of New York Citys efforts to enforce a zero tolerance GOP policy have been shown to be at least partially accountable for that jurisdictions rapid and dramatic decrease in homicide rates during the 1990s. Fagan, Zimring, and Kim (1998) studied the phenomenon and concluded that the precipitous drop in crime rates were not fully explained by extraordinarily high rates returning to baseline, and further documented that the decrease in violent crime could not be explained by population fluctuations or rates of drug use over the period of interest. Zimring and Fagan (2000) concede, however, that the evidence for the decline in crime rates was only circumstantially associated with the change in policing policies in New York City, and comment on the difficulty of accurately measuring such trends in times of general rate declines.
13 Results from other jurisdictions suggest that alternative approaches may also hold promise. Greenes (1999) examination of NYPDs crackdowns states that well-understood methods of community oriented policing and problem oriented policing produced similar crime rate declines in San Diego, but utilized a more benign police presence focused on cooperative efforts rather than an aggressive targeting strategy. The primary negative implication of a zero tolerance policy is that it may damage credibility of the police in community building initiatives, which is apparently an ongoing concern for many urban police forces. However, nationwide policy analyses raise questions over the extent to which community and problem oriented policing programs have been fully and correctly implemented, and further indicate that community and problem oriented policing have not proven to be effective tools for combating urban violence (MacDonald, 2002). New and innovative programs not involving direct-patrol strategies have also shown some promise. One such initiative carried forth by St. Louis police, the Firearm Suppression Program, is premised on obtaining parental consent for searches and seizures at the homes of suspected juvenile offenders. Initial reports demonstrated that the program succeeded in confiscating several hundred guns and managed to do so with minimal community resistance (Rosenfeld & Decker, 1996). Technological advances also make policing guns more effective, as new and more sophisticated methods of detecting concealed weapons become technically feasible, less expensive, easier to deploy, and supportable under current case law (Kamin, 1996). Nevertheless, the bulk of anti-gun policing seems destined, at least in the near future, to occur in the traditional fashion.
14 Application of gun legislation may result in indirect effects for police personnel. The Massachusetts Bartley-Fox gun control law, imposed in the 1970s to deter casual gun carriers who regard concealed weapons as a criminal lifestyle, is one such case. In their analysis of Bartley-Foxs short-term effects in Boston, Pierce and Bowers (1981) suggest that the laws much-publicized unconditional one-year prison term increased public awareness of the need to report gun-related crimes that could be considered less serious (e.g., assault), and reporting of gun assaults went up concomitant with Bartley-Fox. The practical implications for police include greater workload in responding to calls for service, writing reports, performing investigations, and making arrests, all of which may influence the cycle of officer attitudes and behaviors. Additionally, the legislative emphasis on appearing tough on guns logically necessitates in new initiatives for police to pursue. In some cases this may result in new responsibilities for patrol officers (e.g., frisking) which heretofore were discretionary. Gun control laws are intuitively, and in some ways inexorably, associated with the role of police. Evidence suggests that gun-oriented policing strategies can be effective means to controlling illegal weapons and reducing crime, but the evaluation of these strategies (and perhaps formation of future strategies) may be aided by better understanding the factors that police officers consider in the regular performance of their duties. The Role of Police Discretion All policing strategies eventually fall to the street-level officer to be enforced. Conventional wisdom suggests that police officers have considerable discretionary powers in conducting their duties, and that the discretion varies depending on situational factors (e.g., the seriousness of a crime). However, the literature on policing has often
15 challenged the conventional thinking about how officers work. Criminological research has addressed issues of police discretion from a number of different approaches. One route to better understanding behavior is to better understand attitudes. Worden (1989) provides an empirical assessment of the police attitude-behavior link, in which he concludes that evidence demonstrates consistent support for situational factors in decision-making, but very little support for the commonly assumed influence of officer attitudes (i.e., what the officer brings to the situation). Worden posits in his discussion, however, that situational and attitudinal explanations are not wrong but simply incomplete, (pp. 703) suggesting complexity in the theoretical constructs and operational environments alike. He further notes the relative lack of empirical studies on the influence of officer attitudes on behavior, leaving open the plausibility of the police attitude-behavior link. The further examination of police attitudes may be illuminating, both at the individual and at the organizational levels of analysis. One theoretical possibility is that officer discretion is a reflection of officers' attributes, either directly or indirectly. The direct route specifies that personally held values or opinions, such as being conservative with respect to crime and punishment, lead to differential enforcement. A survey of British constables finds that police socialization and training may influence these attitudes, making clear implications for better and more thorough training, as well as seeking a better understanding of police subcultures (Fielding & Fielding, 1991). The indirect route suggests that perceived community reactions to police are predicated on officer attributes. In this way, the proportional use of discretion might be related to perceptions of resistance, which may in turn be influenced by officer characteristics like
16 gender or race. In the case of gender, at least, there is evidence that attitudes do not play such a critical role. A study by Worden (1993) examined conflicting assumptions about women in policing based on sociological theory, and found that male and female officers differ very little in their attitudes and perceptions about occupational factors such as perceived community cooperation. Another possibility is that police culture influences occupational norms, which dictate behavior on patrol. Studies of larger police culture spanning multiple departments provide evidence against the assumptions of common values, attitudes, and norms, and reveal that officers possess a wide variety of perspectives on their occupation that do not fit discreet patterns (Paoline, Myers, & Worden, 2000). This finding is especially important when it relates to the correct implementation of new policing policies, such as community oriented policing, which depend in large part on the effort put forth by individual officers. If officer attitudes and values conflict with the policy, one might expect them to uphold the initiative out of a sense of duty or obligation, but the policy may fail to meet its goals and objectives in the long run because it was not implemented or supported to its fullest potential. If police officers possess discretion and use it in varying degrees depending on situational factors, one might assume that incidents that are more serious would involve less discretion, especially when police have an explicit mandate under the law to enforce certain violations. Non-serious violations, such as traffic violations, are acknowledged as being acceptable situations for use of officer discretion, and may even serve a positive function with respect to preserving organizational norms and promoting officer autonomy (Lundman, 1979). However, even in cases of clear obligation, empirical evidence
17 suggests that reporting serious violations such as child abuse still involves a fair amount of police discretion, which varies with legal factors, such as knowledge of statutory law, and extralegal factors, such as the race of the family (Willis & Wells, 1988). If police discretion plays a role in whatever degree for non-serious as well as serious crimes, it makes sense that the examination of any particular police behavior should require a basic understanding of the attitudes, norms, and values that may influence selective enforcement. The nature of the officer and the crime are not the only deciding factors in discretionary behaviors. Terrill and Reisig (2003) study the value of neighborhood context in use-of-force decisions, which have been traditionally critiqued through dissection of psychological or organizational processes. Their finding is that neighborhood context does matter, even when intuitive factors (suspect resistance, officer training, etc.) are controlled. To the extent that officers decisions and enforcement discretion in use-of-force scenarios varies with the neighborhood setting, logically there may be similar effects for different types of discretion. The value of context is especially relevant in light of past gun oriented policing experiments, in which gun crime hot spots were targeted for special enforcement. Are officers more likely to conduct vehicle searches and seize weapons in high-crime areas? Conversely, are rural officers less likely to enforce concealed weapons laws when violations occur in a small-town (and low-crime) context? Fortunately, there is at least one study that addresses police discretion in rural contexts. Eliason (2003) surveys game wardens in Kentucky and finds wide use of discretion that is influenced by factors such as the seriousness of the offense, intent of the violator, and prior criminal history of the
18 violator. Interestingly, Eliason also queries the wardens on their own history of related deviance, and finds that the wardens past history of wildlife law violation is associated with greater use of discretion. In summary, police discretion is critical to implementation of policies. The amount and nature of discretion, as well its predictors or mediators, is a complex question with no straightforward answer. However, past research suggests that police discretion is real, may relate to attitudes and norms, exists for serious as well as non-serious crimes, and varies with many individual and situational factors. The applicability of police discretion to issues pertaining to guns is a question that deserves more thorough empirical attention. Some research addresses this phenomenon by attempting, with some difficulty, to operationalize vague concepts such as police interest in enforcing gun laws (Moore, 1980). To date, no study has focused on police cadets and the inherent training and socialization issues spurred by contemporary gun policy. Policy and Enforcement Challenges The basic problem with guns is not that the available research is conflicted on whether guns are good or bad for society; indeed, that is a value judgment likely to be influenced by an individuals geographical context, socialization, media influence, economic means, and a host of other factors independent from opinions formed by consuming literature on the subject (Cao, Cullen, & Link, 1997; Dowler, 2002). Moreover, guns are firmly entrenched in American culture and are unlikely to be phased out on a national scale any time soon, so the moralistic view of good versus evil is not fruitful in terms of adjusting policy. Rather, the problem is that the empirical research that does exist on the topics of gun use, gun ownership, and gun control does not seem to
19 form a consensus on exactly what role guns play in commission or defense of crime. There are many contradictory findings to be considered. The disarray in the academic community on issues of gun policy goes beyond political agendas and moral judgments. Zimring (1995) observes that matters of gun policy have only recently become interesting to social scientists, but remarks that matters of interest to gun scholars span many academic disciplines, necessitating multiple theoretical approaches and solutions, each of which must be carefully considered and tested. Not only do different researchers and different disciplines disagree about guns, sometimes even results from the same study can be puzzling. Fagan (2003) points out that even reports demonstrating that guns are not effective in a crime prevention/deterrence role nevertheless go on to identify positive effects of legislation supporting gun possession, effectively confusing the message to policy makers. Without such a consensus, deliberation about the most suitable direction for future gun policy should be regarded cautiously at best. Policing guns is particularly problematic, and police officers are often being asked to do the impossible: provide a consistent enforcement presence that simultaneously preserves lawful gun ownership and deters illegal gun use, in light of political and academic controversies over what, if anything, must be changed to control crime. Their use of discretion in matters of gun crime is an integral part of understanding the whole of gun-related behaviors.
CHAPTER 2 DATA AND METHODOLOGY As previously stated, the goal of this research is to explore the nature of police attitudes toward concealed weapon permitting legislation. Such an exploration necessitates direct queries about police attitudes, as well as an investigation into the demographic and situational determinants that are empirically associated with officer attitudes. This investigation takes place at the individual level of analysis, examining the nature and predictors of officer attitudes. Sample The subjects surveyed were police cadets at a law enforcement training academy in a small southeastern city (approximate population 110,000). The group as a whole was representative of a typical police sample with respect to their backgrounds. In particular, all of the surveyed individuals passed a rigorous screening process that included a criminal history check and drug test prior to their academy admission. These factors become important when considering the degree to which attitudes regarding concealed weapons are influenced by the past behavior (criminal vs. non-criminal) of the possessor. The survey response rate was 85%, and the sample size upon which the analysis is based was 68. Method Since no publicly available data source offered measures of the attitudes and opinions of police with the desired degree of detail, an original questionnaire was developed specifically for this study. The instrument was intended to gather 20
21 demographic and situational data, as well as to quickly assess individual respondents attitudes pertaining to concealed weapons legislation. Prior to its administration, the instrument was reviewed by the training director for the academy in order to critique the surveys terminology and ease of response. The validated questionnaire was then distributed to two separate groups of police cadets at the beginning of their class period on February 10, 2005, and April 4, 2005, and completed surveys were collected anonymously at the end of class. The two different dates reflect a scheduling convenience only, not an attempt at longitudinal comparison. The data discussed here are cross-sectional in nature, and there was no overlapping of participants across the two administration dates. Dependent Variables The dependent variable for this analysis is cadets attitudes toward concealed weapons legislation. This concept was operationalized several different ways on the survey instrument in an effort to provide different approaches to the question at hand. Seven dependent measures of interest can be divided into two logical sub-types. The first type was a direct question asking for the individuals attitude or opinion pertaining to a given type of gun legislation or its effects. The second type was an indirect question asking the individual to read a hypothetical vignette and then specify which behavior would be most likely for him or her in that situation. Each dependent variable featured a Likert-style response continuum from more to less with a neutral choice included for every question. The variables are described in detail below. Direct-Response Dependent Variables Respondents were asked their opinions of concealed weapons laws, ranging from = strongly disapprove to = strongly approve (Q15). This item is intended to be as
22 straightforward as possible, and provides the foundation for the individuals attitude schema on concealed weapons. Subjects were asked their opinion of the effect of concealed weapons laws on violent crime, ranging from = more likely to decrease violence to = more likely to increase violence (Q16). The pairing of a consequence (violence) to the more abstract notion of the legislation forces the respondent to more fully consider the impact of his or her position. Respondents were asked their opinion of concealed weapon licenses holders, ranging from = more likely to decrease violence to = more likely to increase violence (Q29). This item differs from the previous measure by making the distinction between concealed weapons laws vs. concealed weapons holders, potentially revealing a bias. In theory, concealed weapons laws (an abstraction) and people who carry concealed weapons (an application) share a fundamental construct that is intended to promote lawful conduct. If weapons holders are judged differently than weapons laws, perhaps the survey subjects are expressing doubt in the practicality of weapon-carrying behaviors in real-world scenarios. Subjects were asked their opinion on whether concealed weapons laws made their duties more difficult, ranging from 1 = much more difficult to = much less difficult (Q25). Since the surveyed population consists entirely of police cadets rather than experienced officers, this item is capturing the individuals expectations rather than his or her history. This item may yield additional insight with a comparative sample of veteran police officers.
23 Respondents were asked whether they would change existing concealed weapons laws, ranging from = much more restrictive to = much less restrictive (Q26). Modification of existing laws is not the purview of police, except in the grander sense that each of us has an opportunity to participate in the legislative process via elections in a representative democracy. However, asking the cadet what he or she would do with that responsibility invites him or her to examine the issue from a different perspective, as an architect rather than an enforcer of the law. Vignette-Response Dependent Variables Subjects were provided with the following mini-vignette: Consider this hypothetical scenario. While out on patrol, you pull over a car for a minor traffic violation, such as speeding. When you approach the driver, he states that he is in possession of a concealed weapon and that he is licensed to carry the weapon. Two follow-up questions asked the respondent of the likelihood that they would write the offender a traffic ticket, ranging from 1 = more likely to = less likely, and the likelihood that they would request to search the offenders vehicle, ranging from = more likely to = less likely. In both cases, the neutral response read = equally likely, regardless of this information. The vignette items approach the boundary that exists between officer attitudes and officer behaviors, perhaps moving as closely as possible to recording behaviors without actually observing them. Functionally, these items are advantageous because the cadets being surveyed have never been in the scenario described, therefore a similarly oriented direct-response item (e.g., Have you ever) would not be appropriate. Another potential methodological concern is the relatively low base rate for weapons violations as compared to more routine occurrences such as traffic violations, simple assaults, and
24 petty property crime. Utilizing the hypothetical scenario allows greater flexibility than direct observation of officer behaviors, which may be difficult to capture because the convergence of factors described in the scenario may be infrequent and the interpretation may be subject to observer bias. Independent Variables The independent measures of interest deal with a variety of experiential factors relating to experience with guns generally and concealed weapons specifically. Respondents were asked if they had prior military experience, (recoded to 0=No, 1=Yes) (Q11d); if they were a member of the NRA (0=No, 1=Yes) (Q13); if members of their immediate family belonged to the NRA (0=No, 1=Yes) (Q14); if they had ever personally carried a concealed weapon (0=No, 1=Yes) (Q17); if they had ever possessed a license for carrying a concealed weapon (0=No, 1=Yes) (Q21); and if they had ever personally come into contact with legal (Q27d) or illegal (Q28d) concealed weapon holders (recoded to 0=No, 1=Yes). Other Control Variables Several demographic factors were selected as additional control variables for the regression models, including Q1, respondents age group (age 19-29 [coded 0] and ages 31+ [coded 1]); Q2, gender (male [coded 0] vs. female [coded 1]); Q3, race (white [coded 0] vs. non-white [coded 1]); and Q4, ethnicity (non-Hispanic [coded 0] vs. Hispanic [coded 1]). Finally, a dummy variable was used to control for the survey administration date (2/10/05 [coded 0] vs. 4/4/05 [coded 1]). The univariate analysis of the samples demographic and control variables reveals a trend toward younger cadets (62.7% in the 19-29 age group vs. 37.3% in the 30+ group), more males than females (72.5% vs. 27.5%), more whites than non-whites
25 (70.6% vs. 29.4%), and more non-Hispanics than Hispanics (90.2% vs. 9.8%). Fewer respondents had military experience (23.5%), and fewer were themselves members of the NRA (9.8%) or had family members who belonged to the NRA (17.6%). A minority of the sample reported carrying a concealed weapon in the past (23.5%) and fewer reported being licensed to do so (13.7%). Descriptive statistics for all variables are available in Table 1. Table 1: Descriptive Statistics Min. Max. Mean SD Dependent Variables Q15: Opinion of concealed weapons laws 1 5 3.88 .913 Q16: CCW laws effect on violent crime 1 3 2.31 .874 Q23: Hypo: write ticket? 2 3 2.99 .122 Q24: Dummy Hypo: search vehicle? 0 1 .32 .471 Q25: CCW laws make duties more difficult? 1 5 2.70 .723 Q26: Restrictions on CCW laws? 1 5 2.53 .922 Q29: Opinion of concealed weapons holders 1 3 2.25 .910 Independent and Control Variables Q11: Dummy Military experience (0=None) 0 1 .24 .427 Q13: NRA member? (1=No) 1 2 1.16 .371 Q14: Family NRA members? (1=No) 1 2 1.21 .410 Q17: Carried concealed weapon? (1=No) 1 2 1.33 .473 Q21: License for concealed weapon? (1=No) 1 2 1.18 .386 Q27: Dummy Contacts with legal CCW holders (0=No) 0 1 .60 .493 Q28: Dummy Contacts with illegal CCW holders (0=No) 0 1 .45 .502 Q1: Dummy Age Group (0=19-29 years) 0 1 .63 .486 Q2: Dummy Gender (0=Male) 0 1 .22 .418 Q3: Dummy Race (0=White) 0 1 .24 .427 Q4: Dummy Ethnicity (0=Non-Hispanic) 0 1 .10 .306 DATE: Dummy Survey Date (0=2/10/05) 0 1 .24 .427
CHAPTER 3 RESULTS Cadets from this sample seem generally in favor of concealed weapons policies. The majority (70.2%) responded that they approve or strongly approve of concealed weapons laws, while only 4.5% reported that they disapprove or strongly disapprove; 25.4% responded that they were neutral. A majority (55.9%) reported that concealed weapons laws made their jobs no more difficult, but many more cadets reported that the laws would make their jobs somewhat more difficult or much more difficult (33.8%) than somewhat less difficult or much less difficult (7.4%). When asked what their position would be if they could control weapons laws, the largest group of cadets (48.5%) reported that they would make the laws somewhat more restrictive or much more restrictive, while many (42.6%) said they would make the laws no more restrictive, and only 8.8% reported that they would make the laws somewhat less restrictive or much less restrictive. When asked about their opinions regarding the effects of concealed weapons laws on violent crime, the majority (57.4%) reported that they believed the laws had no effect, while 26.5% reported that they believed the laws decreased violence, and 14.7% reported that they believed the laws increased violence. These trends were consistent with reported opinions of concealed weapons holders, for which 55.9% believed there was no effect on violence, 30.9% believed violence was decreased, and 11.8% believed violence increased. 26
27 The hypothetical vignette presented to cadets was designed to assess their likely course of action in the event of encountering a citizen with a legal concealed weapon on a routine traffic stop. Two follow-up questions were asked, pertaining to a request to search the vehicle and the decision to write a traffic citation. The majority of cadets (62.7%) reported no effect in the decision to request a vehicle search, but the remaining cadets were divided on the decision: 19.6% were more likely to request a search, while 17.6% were less likely to request a search. With regard to the decision to write a traffic citation, the overwhelming majority (98.5%) reported no effect in either direction with respect to the presence of a concealed weapon. Consequently, this outcome variable was eliminated from further analyses. In summary, police cadets surveyed in this sample favor concealed weapons laws generally, but the data indicate that many believe that concealed weapons make their job more difficult and that many would make the laws more restrictive if they could. Further, the majority of cadets believed that concealed weapons laws and concealed weapon holders have no effect on violent crime, but more cadets believed that the phenomenon decreased violence than increased it. Finally, it appears that the majority of cadets surveyed report that their decision to request a vehicle search or write a traffic ticket in light of their knowledge about the presence of a concealed weapon was largely unchanged. The next portion of the analysis examines the predictors of concealed weapons perceptions. Further analysis employed multinomial logistic regression due to the multiple categories used for each dependent variable. A multiple regression was estimated for each dependent variable using the same independent and control variables:
28 age, gender, race, ethnicity, military experience, NRA membership, family NRA membership, past history of concealed carry, possession of license for concealed carry, the number of contacts with legal/licensed concealed weapons holders, the number of contacts with illegal/unlicensed concealed weapons holders, and the date of the survey. Based on established demographic and subcultural characteristics of gun owners (Wright & Marston, 1975; Caento, 1979; Cao, Link, & Cullen, 1997), the expected profile of a typical pro-carry respondent is one who is middle-aged, male, white, non-Hispanic, has no military experience, and who may be exposed to a pro-gun subculture (either legal or illegal). Prior history of concealed carry and possession of a license to do so are expected to be significant predictors of present pro-gun attitudes. Regression models were estimated using stepwise backwards-elimination to identify the independent variables that exhibited a significant effect on each dependent variable of interest. Stepwise removal was accomplished by testing the likelihood ratio of each factor until only statistically significant independent variables remained. Overall model fitness was tested by comparing -2 log likelihood values for the null model (in which all parameter coefficients are zero) against the final model. The chi-square statistics and significance values are reported in Table 2. In all cases, the selected models outperformed the null models in predicting the dependent variable. Follow-up analysis of independent effects in each model demonstrates that various factors were statistically significant stepwise predictors for different models. Results are presented in Table 3, and parameter estimates for each model are shown in Appendix A.
29 Table 2: Overall Regression Model Fitness Tests Model -2 Log Likelihood Chi-Square df Q15: Opinion of Concealed Weapons Laws Intercept Only (Null Model) 102.625 Final Model 71.635 30.990** 9 Q16: Concealed Weapons Laws Effect on Violent Crime Intercept Only (Null Model) 86.032 Final Model 63.474 22.558** 4 Q25: Concealed Weapons Laws Make Duties More Difficult? Intercept Only (Null Model) 78.129 Final Model 46.492 31.637* 16 Q26: Restrictions on Concealed Weapons Laws? Intercept Only (Null Model) 81.285 Final Model 59.068 22.217** 8 Q29: Concealed Weapons Holders Effect on Violent Crime Intercept Only (Null Model) 77.262 Final Model 65.316 11.945* 4 Q24: Hypothetical Vignette: Request Search of CCW Holders Vehicle? Intercept Only (Null Model) 90.054 Final Model 67.196 22.858** 6 = p .10, ** = p .05, *** = p .01 Gender appears to be a consistent predictor of outcomes for almost all dependent variables, although it is not always the strongest factor in each model. Crosstabular examination reveals that females in the sample were more likely to disapprove of concealed weapons laws, more likely to believe that the laws increased violence, more likely to report that the laws made their duties more difficult, and were more likely to request to search a vehicle driven by a concealed weapon holder under the provided vignette.
30 Table 3: Independent Effects in Reduced Stepwise Regression Models Effects -2 Log Likelihood of Reduced Model Chi-Square df Q15: Opinion of Concealed Weapons Laws Intercept 71.635 .000 0 Q2: Gender 77.951 6.317** 3 Q17: Carried Concealed Weapon? 84.129 12.495*** 3 DATE: Survey Date 78.002 6.368** 3 Q16: Concealed Weapons Laws Effect on Violent Crime Intercept 63.474 .000 0 Q2: Gender 74.238 10.764*** 2 Q4: Ethnicity 76.602 13.128*** 2 Q25: Concealed Weapons Laws Make Duties More Difficult? Intercept 46.492 .000 0 Q1: Age 57.772 11.279*** 2 Q2: Gender 58.703 12.211*** 2 Q3: Race 55.241 8.749** 2 Q11: Military Experience 53.563 7.071** 2 Q14: Family NRA Members 56.280 9.788** 2 Q17: Carried Concealed Weapon 52.204 5.712* 2 Q21: Licensed for Concealed Weapon 54.977 8.484** 2 Q27: Contacts with Legal CCW Holders 55.363 8.871** 2 Q26: Restrictions on Concealed Weapons Laws? Intercept 59.068 .000 0 Q3: Race 67.325 8.257** 2 Q14: Family NRA Members 70.831 11.763*** 2 Q27: Contacts with Legal CCW Holders 69.920 10.851*** 2 Q28: Contacts with Illegal CCW Holders 68.708 9.640*** 2 Q29: Concealed Weapons Holders Effect on Violent Crime Intercept 65.316 .000 0 Q2: Gender 70.370 5.053* 2 Q27: Contacts with Legal CCW Holders 72.254 6.937** 2 Q24: Hypothetical Vignette: Request Search of CCW Holders Vehicle? Intercept 67.196 .000 0 Q2: Gender 74.737 7.541** 2 Q3: Race 75.849 8.653** 2 DATE: Survey Date 77.404 10.208*** 2 = p .10, ** = p .05, *** = p .01
31 These findings are consistent with other studies examining attitudes of women toward gun issues (Branscombe et. al, 1991; Cooke & Puddifoot, 2000). A Gallup poll conducted in June 2005 indicated that a majority of women (53%) believed that only government officials should carry handguns in public places. More than three quarters of women (76%) reported that they felt less safe in a public place that permitted concealed firearms. Race and ethnicity were significant in several models. Non-whites who participated in the survey were more likely to be neutral on whether the law makes their duties more difficult and whether they would place additional restrictions on the licensing process. Non-whites were also less likely to request to search a concealed weapon holders vehicle under the hypothetical vignette. Hispanics in the sample were more likely to believe concealed weapons laws decreased violent crime. These findings diverge somewhat from prior literature on gun control attitudes. First, race is generally not found to be a statistically significant predictor. Caento (1979), for example, found significant effects of gender for both gun ownership and attitudes toward gun control, but found no significant effects for race (white vs. non-white). Second, there appear to be virtually no peer-reviewed studies that make the distinction between Hispanics and non-Hispanics in sampling gun control attitudes. In most cases where ethnicity is considered, the categories appear to be collapsed into whites vs. non-whites for the purpose of analysis. The effects of prior experience with guns and the prior exposure to pro-gun subcultures were significant in two models. Respondents who had previously carried a concealed weapon were, unsurprisingly, more likely to approve of concealed weapons
32 laws. This group was also more likely to respond that concealed carry laws made their official duties no more difficult (neutral). Military experience was significant in only one model, with those having military backgrounds more likely to report neutrality on whether the law makes their duties more difficult. Those who reported known contacts with concealed weapon holders were more likely to report that the laws made their duties more difficult, but paradoxically were less likely to express a desire to make the licensing process more restrictive. Family members NRA membership was significant in two models, including the one predicting placing restrictions on the acquisition of civilian licenses to carry concealed weapons. The objective of the current study is to measure attitudes and their effects, however, it is also possible to consider attitudes as an independent variable predicting behavior. Although cadet behavior was not directly measured, it might be reasonably approximated by the hypothetical scenario asking whether the subject would request a search of a vehicle whose driver was in possession of a concealed weapon. A follow-up analysis was conducted, in which the subjects projected decision to request a vehicle search was the dependent variable and two measures of attitudes pertaining to concealed weapons laws and concealed weapons holders were introduced as independent variables. All previous independent and control variables were retained in the model. The results indicate that neither of the two independent variables concerning attitudes about concealed weapons laws or concealed weapons holders were significant predictors of decision to request a vehicle search; rather, the significant predictors are the same as in the previous model (see Table 3) Results are presented in Table 4, and parameter estimates are included in Appendix A.
33 Table 4: Independent Effects of Cadet Attitudes on Decision to Search Effects -2 Log Likelihood of Reduced Model Chi-Square df Q24: Hypothetical Vignette: Request Search of CCW Holders Vehicle? Intercept 67.304 .000 0 Q2: Gender 75.590 8.286** 2 Q3: Race 76.476 9.172*** 2 DATE: Survey Date 76.637 9.333*** 2 = p .10, ** = p .05, *** = p .01 In conclusion, the available evidence from this sample of police cadets suggests that individual characteristics such as gender, race, and ethnicity play a role in influencing attitudes toward concealed weapons policies. Experiential factors (past history of carrying concealed weapons, possessing a license to carry) and socialization or subcultural factors (family NRA membership and military experience) found limited but inconsistent support. Discussion The objective of this study was to describe and explore the attitudes of police cadets toward concealed weapon permitting laws by (1) directly surveying active police cadets, (2) utilizing multiple operationalizations of gun-related experience and subcultural socialization effects as independent variables, (3) controlling for an assortment of demographic variables, and (4) employing multiple potential measures of cadet attitudes as dependent variables. The results of the analyses indicated that several models held predictive validity for the outcome variables employed, and that different factors were significant in predicting the dependent variables. Police behavior was not directly measured, but the data suggest that opinions toward concealed weapons laws are
34 not a significant predictor of the decision to request a vehicle search in the presence of a concealed weapon. Although preliminary, these findings are interesting in several ways. First, there appears to be diversity in attitudes and beliefs toward concealed weapon policies among members of the law enforcement community. This suggests that police cadets in this sample operate without a groupthink bias either toward or against gun control. It also hints at the possibility of individual officer discretion in gun enforcement situations. Second, the factors predicting the attitudes of interest demonstrate that different individual characteristics hold value depending on the attitude under examination. In some cases, the regression models held substantial predictive accuracy, but it seems unlikely that one construct (e.g., social learning) can account for the majority of variance in all of the models. If that were the case, the same independent variables would have been significant in a majority of models, when in fact only one factor (gender) appeared in all but one of the models. Third, because a single construct does not emerge to explain cadet attitudes, it seems plausible that other predictive factors may yet await detection. This indicates an opportunity for a more detailed investigation that accounts for new contextual variables not utilized here, such as measures of individual history, family structure, religiosity, political affiliation, education, income level, and so forth. If these measures can be successfully incorporated, it may be possible to draw a more complete picture of the personal or experiential attributes that matter most. Fourth, from a policy evaluation perspective, there may be inferences to be drawn from these data regarding the sustainability and enforceability of current shall-issue concealed weapon permit laws based on reactions from police on patrol. For example, escalating fear and strongly held
35 anti-policy attitudes may suggest a questionable long-term outlook for the currently popular policies. On the other hand, the relatively ambivalent (and even supportive) attitude of police cadets toward concealed weapon legislation may indicate a tacit acceptance. Longitudinal analysis may help to address many of these future aims. Limitations This study was not without limitations. First, the research design employed is not conducive to inferring causation in any of the models. The target population was members of the law enforcement community. However, the lack of a discreet control group is defensible when only attitudes of police are of interest. Comparison to a control group may be beneficial in that they permit inferences about selection effects for police cadets, but a non-police control group does little to address the relationship between the surveyed attitudes and police behavior. Second, because the sample size was limited, and because the survey was administered at only one site, the generalizability of the findings is an issue. Until replication occurs, ideally across several different locations and with more participants, these conclusions should be regarded as preliminary findings for follow-up investigations. Third, the value of context must be recognized in interpreting the findings. The state of Florida was a pioneer in the implementation of a shall-issue permit law, and as such, Florida has one of the least restrictive policies on permits for concealed weapons anywhere in the United States. It is conceivable that surveyed attitudes pertaining to these laws may be influenced by either a more general knowledge of practical realities on the street, or by a resignation to the fact that public policy dealing with concealed weapons in Florida is unlikely to change in the near future. It is conceivable that a similar study, conducted outside the state of Florida, would generate remarkably different outcomes. Finally, the use of stepwise regression techniques may
36 indicate a potential weakness in the analysis. Although stepwise procedures are not generally embraced for traditional theory testing purposes due to problems with artificially high pseudo-R2 values, difficulty with collinearity, and other mathematical concerns, in this case the stepwise approach offers a convenient tool for evaluating the statistically significant factors in a model that is not grounded in a formal theoretical framework. In the current study, multicollinearity was not problematic. The sample and data also presented some specific limitations that must be acknowledged. First, the survey was administered on two different dates as a matter of availability, and the effects of the split may be confounded by salient current events. The date of the second administration coincided with the introduction of a very public amendment to Floridas laws pertaining to the use of force by civilians. In the amendment, the so-called castle doctrine was extended into the public domain for those citizens who use lethal force in self-defense (Florida Department of Agriculture and Consumer Services, 2005). In essence, legally armed citizens were no longer required to retreat from an attacker before using force under the proposal, whereas existing law required at least a reasonable effort to escape the threat before using force. The practical implications of this proposal made headlines worldwide (Elliott, 2005), and the accompanying debate undoubtedly permeated law enforcement communities as it did with the general public. A control variable was added to the models to account for any variation in attitude reporting due to these events, and the results indicated that the effect was significant in two of the six regressions. In a way, the status of gun legislation in Florida at the time of this survey could be viewed as an opportunity to draw comparisons
37 in preand post-law enforcement opinions. Replication of this survey after the law takes effect in October 2005 may yield additional insights. The survey date is related to another limitation of this study. For confidentiality reasons, no personally identifying information was recorded from the participants in this sample. Although the date of the survey is known, not all cadets began their academy experience at the same time, so calculating the amount of time that a particular cadet had been enrolled in the police academy at the time of the survey is impossible. This presents a constraint when considering the potentially significant effects of police socialization and training, in that it is very difficult to quantify the effects of more experience on the respondents attitudes of interest with these data. A cadet who has been at the academy longer is presumed to have been more socialized than a brand new cadet, and it seems likely that socialization plays a major role in gun-related beliefs and behaviors based on past research (Cao, Link, & Cullen, 1997). Subsequent analysis using an independent samples t-test shows that the means of the dependent variables for the two groups (separated by survey date) differs on only one dimension, with the group surveyed in February more likely to approve or strongly approve of concealed weapons laws. The results suggest that any potential effects of experience and socialization are minimal for this sample, which is reasonable for a sample of police cadets. Results are presented in Table 5. Future Research One potentially important issue largely unaddressed by this investigation is the degree to which cadets or inexperienced rookie officers are socialized into a subculture or fraternity of policing. The presence of such a subculture may fundamentally alter belief systems instilled at the academy. It seems plausible that the formal or informal tutelage
38 of more experienced mentor officers may outweigh a cadets training in some instances. A future project may involve a survey of relevant attitudes with a sample of veteran officers, in order to compare the two groups and derive the value of real world experience and/or acculturation for police. Furthermore, the extent to which these factors influence the implementation of new policing strategies, whether related to gun control or not, may be interesting from a policy evaluation perspective. Table 5: Independent Samples t-Test Grouped by Survey Date t df Mean Difference Std. Err. Diff. Q15: Opinion of concealed weapons laws 2.428** 27.654 .582 .240 Q16: CCW laws effect on violent crime -1.475 31.006 -.327 .222 Q24: Hypo: search vehicle? -1.296 34.063 -.245 .189 Q25: CCW laws make duties more difficult? 1.585 26.611 .251 .158 Q26: Restrictions on CCW laws? -.160 26.537 -.029 .180 Q29: Opinion of concealed weapons holders .269 24.262 .069 .257 = p .10, ** = p .05, *** = p .01 Future investigations may find success in further exploring the nature of police cadet training with respect to concealed weapons. One possible implication for the safety first mentality taught in police academies is that any concealed weapon, whether legal or illegal, is a threat until neutralized. Therefore, the psychology of a routine traffic stop where concealed weapons may be present likely extends well beyond a favorable vs. unfavorable view on the presence of concealed weapons into a series of decisions and procedures, any or all of which may reflect an individual officers attitudes or beliefs. Qualitative interviews with cadets and instructors, critical assessment of academy
39 curricula, and systematic observation of police-citizen encounters may yield additional avenues for exploration on this topic. Broader inquiry into the domain of gun policy and policing issues may address rationalization of violent gun crime and legal gun defense from a multidisciplinary approach (e.g., psychological, sociological, biological), and will very likely involve further replications and program initiatives with the cooperation of the law enforcement community. The effects of fear of crime, attitudes toward police, and the impact of self-control theories related to gun use all have promise as further avenues of investigation. Hopefully, policymakers will eventually come to recognize that the answers to questions of gun policy will never be as simplistic as more guns or fewer guns, but will require layered, integrated solutions to ensure that guns are regulated and managed in a responsible way. Clearly, the need for continued investigation and innovation on the gun problem is as pronounced now as it has ever been.
APPENDIX A PARAMETER ESTIMATES FOR MULTINOMIAL LOGISTIC REGRESSION MODELS
Table A-1: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q15 Effects Strongly Disapprove vs. Neutral Approve vs. Neutral Strongly Approve vs. Neutral B SE Exp(B) B SE Exp(B) B SE Exp(B) Q15: Opinion of Concealed Weapons Laws Intercept ------------------Q2: Gender (0=Male) ------1.502** .753 4.492 ------Q17: Carried Concealed Weapon? (0=No) -------------3.267** 1.298 .038 DATE: Survey Date (0=2/10/05) ------------------Reference category is Neutral. = p .10, ** = p .05, *** = p .01 Table A-2: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q16 41 Effects Likely to Decrease Violence vs. Neutral Likely to Increase Violence vs. Neutral B SE Exp(B) B SE Exp(B) Q16: Concealed Weapons Laws Effect on Violent Crime Intercept ------------Q2: Gender (0=Male) ------------Q4: Ethnicity (0=Non-Hispanic) ------------Reference category is Not likely to influence violence (neutral). = p .10, ** = p .05, *** = p .01
Table A-3: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q25 Effects More Difficult vs. Neutral Less Difficult vs. Neutral 42 B SE Exp(B) B SE Exp(B) Q25: Concealed Weapons Laws Make Duties More Difficult? Intercept ------------Q1: Age (1=19-29 years) ------------Q2: Gender (0=Male) -1.704* .887 .182 ------Q3: Race (0=White) 2.024** .970 7.566 ------Q11: Military Experience (0=None) ------------Q14: Family NRA Members (1=No) ------------Q17: Carried Concealed Weapon (1=No) ------------Q21: Licensed for Concealed Weapon (1=No) ------------Q27: Contacts with Legal CCW Holders (0=None) ------------Reference category is No more difficult (neutral). = p .10, ** = p .05, *** = p .01 Table A-4: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q26 Effects More Restrictive vs. Neutral Less Restrictive vs. Neutral B SE Exp(B) B SE Exp(B) Q26: Restrictions on Concealed Weapons Laws? Intercept ------------Q3: Race (0=White) 1.593** .720 4.920 ------Q14: Family NRA Members (1=No) ------------Q27: Contacts with Legal CCW Holders (0=None) ------------Q28: Contacts with Illegal CCW Holders (0=None) ------------Reference category is No more restrictive (neutral). = p .10, ** = p .05, *** = p .01
Table A-5: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q29 43 Effects More Likely to Decrease Violence vs. Neutral More Likely to Increase Violence vs. Neutral B SE Exp(B) B SE Exp(B) Q29: Concealed Weapons Holders Effect on Violent Crime Intercept ------------Q2: Gender (0=Male) ------------Q27: Contacts with Legal CCW Holders (0=None) -1.586* .868 .205 -1.991* 1.163 .136 Reference category is Not likely to influence violence (neutral). = p .10, ** = p .05, *** = p .01 Table A-6: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q24 Effects More Likely to Search vs. Neutral Less Likely to Search vs. Neutral B SE Exp(B) B SE Exp(B) Q24: Hypothetical Vignette: Request Search of CCW Holders Vehicle? Intercept ------------Q2: Gender (0=Male) -1.904* .993 .149 ------Q3: Race (0=White) -------2.556*** .995 .077 DATE: Survey Date (0=2/10/05) ------------Reference category is No effect (neutral). = p .10, ** = p .05, *** = p .01
44 Table A-7: Significant Parameter Estimates in Reduced Stepwise Regression Model for Q24 with Additional IVs Effects More Likely to Search vs. Neutral Less Likely to Search vs. Neutral B SE Exp(B) B SE Exp(B) Q24: Hypothetical Vignette: Request Search of CCW Holders Vehicle? Intercept ------------Q2: Gender (0=Male) -1.858* .995 .156 ------Q3: Race (0=White) -------2.692*** .974 .068 DATE: Survey Date (0=2/10/05) ------------Reference category is No effect (neutral). = p .10, ** = p .05, *** = p .01
APPENDIX B POLICE CADET SURVEY QUESTIONNAIRE Q1: Please indicate your age: Q2: Please indicate your gender: 1 Male 2 Female Q3: Please indicate your race: 1 White 2 Non-white Q4: Please indicate your ethnicity: 1 Hispanic 2 Non-Hispanic Q5: Please indicate your zip code of residence: Q6: Please indicate the total number of years you have been a police officer: Q7: Please indicate your marital status: 1 Single (never married) 2 Married 3 Divorced or separated Q8: Please indicate your political affiliation: 1 Republican 2 Democrat 3 Other/third party 4 No political affiliation Q9: Did you vote in the November 2004 Presidential election? 1 No 2 Yes Q10: Please indicate the highest level of education you have completed: 0 Lower than high school 1 High school or GED 2 Some college or AA/AS 3 Undergraduate or BA/BS 4 Graduate school or MA/MS/MBA/etc. 5 Postgraduate/professional school or PhD/MD/etc. 45
46 Q11: Please indicate your military experience: 1 No military experience 2 Active duty 3 Reserves/National Guard 4 Other (please specify) Q12: If you have served in the military, please indicate which branch: Q13: Are you now, or have you ever been, a member of the National Rifle Association? 1 No 2 Yes Q14: Do any members of your immediate family (parents, siblings, spouse) belong to the National Rifle Association? 1 No 2 Yes Q15: Please indicate your opinion of concealed weapon laws generally: 1 Strongly disapprove 2 Disapprove 3 Neutral 4 Approve 5 Strongly approve Q16: In your opinion, are concealed weapon laws: 1 More likely to cause a decrease in violent crime 2 More likely to cause an increase in violent crime 3 Not likely to influence violent crime either way Q17: Before you became a police officer, did you ever carry a concealed weapon? 1 No, never carried a weapon 2 Yes, carried a weapon Q18: If YES to above question, please indicate the type of concealed weapon that you carried (mark all that apply): 1 Handgun 2 Fixed-blade knife (e.g., hunting knife or dagger) 3 Folding knife (e.g., pocket knife) 4 Stun gun 5 Billy club 6 Other (please specify) Q19: If YES to above question, please indicate the number of times per week that you typically carried your weapon(s):
47 Q20: If you answered yes to previous, please indicate most likely reason for concealed carry (otherwise skip to next question): 1 Enjoy hunting, recreational shooting, or competitive shooting sports 2 Personal protection 3 Other (please specify) Q21: Before you became a police officer, did you possess a license or permit to carry a concealed weapon? 1 No, never had a license 2 Yes, had a license Q22: If YES to above question, please indicate the state(s) from which you acquired a concealed weapons license or permit: Consider this hypothetical scenario. While out on patrol, you pull over a car for a minor traffic violation, such as speeding. When you approach the driver, he states that he is in possession of a concealed weapon and that he is licensed to carry the weapon. Q23: Regarding the hypothetical scenario above, are you: 1 More likely to write him a traffic ticket 2 Less likely to write him a traffic ticket 3 Equally likely to write him a traffic ticket, regardless of this information Q24: Regarding the hypothetical scenario above, are you: 1 More likely to request to search his vehicle 2 Less likely to request to search his vehicle 3 Equally likely to request to search his vehicle, regardless of this information Q25: In your opinion, do concealed weapons laws make your duties as a police officer: 1 Much more difficult 2 Somewhat more difficult 3 No more difficult (same as before) 4 Somewhat less difficult 5 Much less difficult Q26: If you could personally control concealed weapons laws, would you make them: 1 Much more restrictive 2 Somewhat more restrictive 3 No more restrictive (same as current laws) 4 Somewhat less restrictive 5 Much less restrictive Q27: How many times have you personally had contact with an individual carrying a concealed weapon legally (e.g., possessing a valid permit)?
48 Q28: How many times have you personally had contact with an individual carrying a concealed weapon illegally (e.g., not possessing a valid permit)? Q29: In your opinion, are legal concealed weapon permit holders: 1 More likely to cause a decrease in violent crime 2 More likely to cause an increase in violent crime 3 Not likely to influence violent crime either way
LIST OF REFERENCES Akers, R. (1973). Deviant behavior: A social learning approach. Belmont, CA: Wadsworth. Bankston, W., Thompson, C., Jenkins, Q., & Forsyth, C. (1990). The influence of fear of crime, gender, and Southern culture on carrying firearms for protection. Sociological Quarterly, 31(2), 287-305. Branscombe, N., Weir, J., & Crosby, P. (1991). A three-factor scale of attitudes toward guns. Aggressive Behavior, 17, 261-273 Caento, D. (1979). The domestic arms race. Journal of Communication, 29(2), 39-46. Cao, L., Cullen, F., & Link, B. (1997). The social determinants of gun ownership: Self-protection in an urban environment. Criminology, 35(4), 629-650. Cohen, J. & Ludwig, J. (2003). Policing crime guns. In J. Ludwig & P. Cook (Eds.), Evaluating gun policy (pp. 217-250). Washington D.C.: Brookings. Cook, P. & Ludwig, J. (2000). Gun violence: The real costs. Oxford University Press. Cooke, C. & Puddifoot, J. (2000). Gun culture and symbolism among U.K. and U.S. women. Journal of Social Psychology, 140(4), 423-433. Cress, L. (1984). An armed community: The origins and meaning of the right to bear arms. The Journal of American History, 71(1), 22-42. Donohue, J. (2003). The final bullet in the body of the more guns, less crime hypothesis. Criminology & Public Policy, 2(3), 397-409. Dowler, K. (2002). Media influence on attitudes toward guns and gun control. American Journal of Criminal Justice, 26(2), 235-247. Eliason, S. (2003). Throwing the book versus cutting some slack: Factors influencing the use of discretion by game wardens in Kentucky. Deviant Behavior, 24, 129-152. Elliott, G. (2005, April 9). Fight or flight? Florida gun folk to be given option. The Weekend Australian, World section, p. 17. Fagan, J. (2003). Guns, science, and social policy. Criminology & Public Policy, 2(3), 359-361. 49
50 Fagan, J., Zimring, F., & Kim, J. (1998). Declining homicide in New York City: A tale of two trends. Journal of Criminal Law & Criminology, 88(4), 1277-1307. Fielding, N. & Fielding, J. (1991).. Police attitudes to crime and punishment. British Journal of Criminology, 31, 39-53. Florida Department of Agriculture and Consumer Services. (2005, May 5). Link to new Use of deadly force law effective October 10, 2005. Retrieved June 10, 2005, from http://licgweb.doacs.state.fl.us/weapons/law_deadly_force.html Gallup Organization. (2005, June 14). Public wary about broad concealed firearms privileges. Retrieved June 24, 2005, from http://www.gallup.com/ Greene, J. (1999). Zero tolerance: A case study of police policies and practices in New York City. Crime & Delinquency, 45(2), 171-187. Kamin, S. (1999). Law and technology: The case for a smart gun detector. Law and Contemporary Problems, 59(1), 221-262. Kleck, G. (1996). Crime, culture conflict and the sources of support for gun control: A multilevel application of the general social surveys. American Behavioral Scientist, 39(4), 387-405. Kleck, G. & Gertz, M. (1995). Armed resistance to crime: The prevalence and nature of self-defense with a gun. Journal of Criminal Law and Criminology, 86, 150-187. Kleck, G. & Gertz, M. (1998). Carrying guns for protection: results from the National Self-Defense Survey. Journal of Research in Crime and Delinquency, 35(2), 193-225. Kovandzic, T. & Marvell, T. (2003). Right-to-carry concealed handguns and violent crime: Crime control through gun decontrol? Criminology & Public Policy, 2(3), 363-396. Lafree, G. (2000). Explaining the crime bust of the 1990s. Journal of Criminal Law & Criminology, 91(1), 269-305. Lizotte, A., Boruda, D. (1980). Firearm ownership for sport and protection: two divergent models. American Sociological Review, 45(2), 229-244. Lizotte, A., Krohn, M., Howell, J., Tobin, K., & Howard, G. (2000). Factors influencing gun carrying among young urban males over the adolescent-young adult life-course. Criminology, 38(3), 811-834. Lott, J. & Mustard, D. (1997). Crime, deterrence, and right-to-carry concealed handguns. The Journal of Legal Studies, 26(1), 1-68.
51 Ludwig, J. (2003). Evaluating gun-policy evaluations. Criminology & Public Policy, 2(3), 411-418. Lundman, R. (1979). Organizational norms and police discretion. Criminology, 17(2), 159-171. MacDonald, J. (2002). The effectiveness of community policing in reducing urban violence. Crime & Delinquency, 48(4), 592-618. McGarrell, E., Chermak, S., Weiss, A., & Wilson, J. (2001). Reducing firearms violence through directed police patrol. Criminology & Public Policy, 1(1), 119-148. Moore, M. (1980). The police and weapons offenses. Annals of the American Academy of Political and Social Science, 452(1), 22-32. Paoline, E., Myers, S., & Worden, R. (2000). Police culture, individualism, and community policing: Evidence from two police departments. Justice Quarterly, 17(3), 575-606. Pierce, G. & Bowers, W. (1981). The Bartley-Fox gun laws short-term impact on crime in Boston. Annals of the American Academy of Political and Social Science, 455, 120-137. Rosenfeld, R. & Decker, S. (1996). Consent to search and seizure: evaluating an innovative youth firearm suppression program. Law and Contemporary Problems, 59(1), 197-220. Schwaner, S., Furr, L., Negrey, C., & Seger, R. (1999). Who wants a gun license? Journal of Criminal Justice, 27(1), 1-10. Sherman, L. & Rogan, D. (1995). Effects of gun seizures on gun violence: Hot spot patrol in Kansas City. Justice Quarterly, 12(4), 673-694. Stell, L. (2001). Gun control and the regulation of fundamental rights. Criminal Justice Ethics, 20(1), 28-33. Terrill, W. & Reisig, M. (2003). Neighborhood context and police use of force. Journal of Research in Crime and Delinquency, 40(3), 291-321. Wilcox, P. (2002). Self-help? Examining the anti-crime effectiveness of citizen weapon possession. Sociological Focus, 35(2), 145-167. Willis, C. & Wells, R. (1988). The police and child abuse: An analysis of police decisions to report illegal behavior. Criminology, 26(4), 695-716. Worden, A. (1993). The attitudes of women and men in policing: Testing conventional and contemporary wisdom. Criminology, 31(2), 203-241.
52 Worden, R. (1989). Situational and attitudinal explanations of police behavior: A theoretical reappraisal and empirical assessment. Law & Society Review, 23(4), 667-711. Wright, J. (1986). The armed criminal in America. Research in Brief, National Institute of Justice, 1-5. Wright, J. & Marston, L. (1975). The ownership of the means of destruction: Weapons in the U.S. Social Problems, 23(1), 93-107. Zimring, F. (1995). Reflections on firearms and the criminal law. Journal of Criminal Law & Criminology, 86(1), 1-9. Zimring, F. & Fagan, J. (2000). The search for causes in an era of crime declines: some lessons from the study of New York City homicide. Crime & Delinquency, 46(4), 446-456.
BIOGRAPHICAL SKETCH Matt Nobles is a Gainesville native and life-long Gator fan, graduating from Gainesville High School in 1997. He received a B.S. in psychology with dual minors in business administration and criminology from the University of Florida in 2001, followed by a B.A. in criminology in 2003, and was elected to the honor society of Phi Kappa Phi. Matt anticipates pursuing a Ph.D. in criminology at the University of Florida, where he will continue to contribute leadership and service. 53