SEXUAL ORIENTATION IN STATE HATE CRIME LAWS: THE INTERPLAY OF SOCIAL CONSTRUCTION AND PUBLIC POLICY by LAURA E. VALCORE B.A. Sociology Eastern Illinois University 2007 MCJ University of Colorad o Denver 2009 A thesis submitted to the Faculty o f the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Doctor of Philosophy Public Affairs 2014
ii This thesis for the Doctor of Philosophy degree by Laura E. Valcore has been approved for the Pub lic Affairs Program by Angela Gover Chair Mary Dodge, Advisor Stacey Bosick Valerie Jenness July 25, 2014
iii Valcore, Laura E. (Ph.D., Public Affairs) Sexual Orientation in State Hate Crime Laws: The Interplay of Social Construction and Public Pol icy Thesis directed by Professor Mary Dodge ABSTRACT Hate crimes are an identified social problem caused by bias and prejudice against certain social groups and societal responses to them are important to understand because of the unique damage that bia s motivated crimes cause to victims and communities. Previous studies identified factors involved with the diffusion and differentiation of state hate crime laws. The timing of legal adoption and the strength of interest groups have been identified as ke y variables that determine the form and content of state hate crime laws, but studies to date have failed to fully explain the wide variation of bias categories. The addition of sexual orientation as a bias category is a specific area of concern because c rimes against gays and lesbians often are more violent than other hate crimes and are sometimes committed by law enforcement personnel. Only 30 of the 45 states that specifically criminalize hate or bias motivated violence include sexual orient ation as a protected category. This dissertation explores the social construction of gays and lesbians in a sample of six states in order to help explain the differentiation in state hate crime laws. Gays and lesbians continue to be viewed as deviant, which often p revents them from gaining hate crime prot ections. Furthermore, e xisting research fails to address the impact of hate crime laws on included social groups. Thus, t wo research questions are asked: 1) Is the social construction of gays and lesbians related to the inclusion or exclusion of sexual orientation as a bias category in state hate crime law? 2) Does the addition of sexual orientation to a state hate crime law impact the social construction of
iv gays and lesbians in that state? Data are drawn from a s ample of 12 daily newspapers. Content analysis determines the social construction of gays and lesbians in each state. Categorical and time series analyses are used to test the hypotheses Results indicate that a positive dependent construction is associa ted with target groups included in state hate crime protections and that inclusion fails to have a positive impact upon th eir construction. One implication is that additional hate crime laws and bias categories may be unnecessary. The form and content of this abstract are approved. I recommend its publication. Approved: Mary Dodge
v DEDICATION I dedicate this work to all of the members of the LGBTQ community around the world who continue to fight for equality and acceptance daily
vi ACKNOWLEDGMENTS I would like to thank my advisor, Mary Dodge, for believing in me and directing me toward an academic career. H er encouragement and support contributed to my decision to pursue a doctorate. I must also express gratitude for the members of my dissertation committee, Angela Gover, Stacey Bosick, and Valerie Jenness. T heir hard work and recommendations were instrumental in the successful completion of this dissertation Additionally, I would like to express my gratitude to Loren Cobb for his statistical cons ulting services. H is assistance on the time series analysis was essential in the data analysis
vii TABLE OF CONTENTS CHAPTER I. INTRODUCTION ................................ ................................ ................................ ........... 1 Defining Hate Crime ................................ ................................ ......................... 4 Defining Social Construction ................................ ................................ ............ 6 Sexual Orientation in the United States ................................ ............................ 7 Impact of Hate Crime Laws on Social Groups ................................ ................. 9 Dissertation Overview ................................ ................................ .................... 11 II. LITERATURE REVIEW ................................ ................................ ............................. 13 Introduction ................................ ................................ ................................ ..... 13 Scope of the Problem ................................ ................................ ...................... 13 Consequences of Hate Cr imes ................................ ................................ ........ 18 Purpose and Impact of Hate Crime Laws ................................ ....................... 19 Development of Hate Crime Law in the United States ................................ ... 23 Social and Political Status of Gays and Lesbians ................................ ........... 31 Summary ................................ ................................ ................................ ......... 35 III. SOCIAL CONSTRUCTION FRAMEW ORK ................................ ........................... 38 Introduction ................................ ................................ ................................ ..... 38 Framework Overview and Assumptions ................................ ......................... 38 Gays a nd Lesbians in the SCF Matrix ................................ ............................ 42 Criticisms of the Framework ................................ ................................ .......... 44 Previous Applications of the SCF ................................ ................................ ... 45 Operationalization of Social Construction ................................ ................ 46 Operationalization of Political Power ................................ ....................... 47 Quantitativ e Applications ................................ ................................ ......... 48
viii Application of the SCF to State Hate Crime Laws ................................ ......... 49 Research Questions ................................ ................................ ......................... 50 Summary ................................ ................................ ................................ ......... 51 IV. RESEARCH DESIGN AND METHODOLOGY ................................ ...................... 53 Introduction ................................ ................................ ................................ ..... 53 Multiple Time Series Design ................................ ................................ .......... 53 Hypotheses ................................ ................................ ................................ ...... 54 Operationalization ................................ ................................ ........................... 56 Social Construction ................................ ................................ ................... 56 Political Power ................................ ................................ .......................... 57 Sample ................................ ................................ ................................ ............. 59 Methods ................................ ................................ ................................ ........... 65 Content Analysis ................................ ................................ ....................... 66 Coding ................................ ................................ ................................ ....... 66 Variables ................................ ................................ ................................ ... 67 Analysis ................................ ................................ ................................ ........... 67 Summary ................................ ................................ ................................ ......... 68 V. RESULTS I ................................ ................................ ................................ .................. 69 Sample Descriptives ................................ ................................ ........................ 69 Missing Data ................................ ................................ ............................. 69 Research Question 1 Target Groups ................................ ............................... 70 The SCF Matrix ................................ ................................ ........................ 74 Colorado/Oklahoma ................................ ................................ ............ 76 Texas/Alabama ................................ ................................ ................... 78 Tennessee/North Carolina ................................ ................................ ... 79
ix Summary of H 1 and H 2 Results ................................ ................................ 80 Research Question 1 Summary ................................ ................................ 82 VI. RESULTS II ................................ ................................ ................................ ............... 84 Colorado/Oklahoma ................................ ................................ .................. 87 Texas/Alabama ................................ ................................ ......................... 91 Tennessee/North Carolina ................................ ................................ ......... 95 Summary of H 3 Results ................................ ................................ ............. 98 Research Question 2 Summary ................................ ............................... 100 Results Conclusion ................................ ................................ ........................ 101 VII. DISCUSSION ................................ ................................ ................................ ......... 103 Research Question 1 ................................ ................................ ..................... 103 Research Question 2 ................................ ................................ ..................... 109 Policy Implications ................................ ................................ ....................... 113 Limitations ................................ ................................ ................................ .... 115 Future Research ................................ ................................ ............................ 120 Conclusion ................................ ................................ ................................ .... 121 REFERENCES ................................ ................................ ................................ ............... 123 APPENDIX ................................ ................................ ................................ ..................... 136 A. CODING FRAMEWORK ................................ ................................ ......................... 136
x LIST OF TABLES T able IV .1 State Demographic Frequencies ................................ ................................ ............. 62 IV.2 Newspaper Sample based on Passage of Statute and Political Leaning ................. 6 5 V.1 Sample Frequencies ................................ ................................ ................................ 6 9 V.2 Missing D ata ................................ ................................ ................................ ............ 70 2 7 3 5 V .6 Power .. 7 6 V I.1 .. 8 5
xi LIST OF FIGURES Figure III .1 Social Construction Matrix ................................ ................................ ..................... 4 1 IV.1 State Hate Crime Laws ................................ ................................ ........................... 61 V.1 Colorado Matr ix ................................ ................................ ................................ ....... 77 V.2 Oklahoma Matrix ................................ ................................ ................................ ..... 77 V.3 Texas Matrix V.4 Alabama Ma 79 V.5 Tennessee Ma V.6 North Carolina M V I.1 Colorado Social Construct 89 V I.2 Oklahoma Social Construct 90 VI.3 Colorado Power Ti 90 VI.4 91 VI.5 Texas Social Constructio VI.6 Alabama Social Construction VI.7 Texas Power Time VI.8 Alabama Power Ti VI.9 Tennesse e Social Constructi VI.10 North Carolina Social Constru 96 VI.11 Tennessee Power Time Se 97 VI.12 North Carolina Power 97
1 CHAPTER I INTRODUCTION This dissertation has two distinct but related goals concerning hate crime laws and public policy. First, the work aid s in the development of a public policy theory related to gay and lesbian status in the Social Cons truction Framework. Second, the work is designed to increase overall understanding of state level variations in hate crime laws. Specifically, the Social Construction Framework is used to quantitatively determine the social and political status of gays a nd lesbians in a sample of U.S. states. The relationship of social construction to the inclusion of sexual orientation as a bias category in state hate crime law is examined, and in turn, the impact of that inclusion upon gays and lesbians is explored T he results of the study, thus, have both theoretical and practical implications. Hate crime is a recent development in criminal law (Chakraborti & Garland, Smith, 2001, p. 4 79), a new policy domain (Grattet & Jenness, 2001) and a part of the growing field of hate studies (Blitzer, 2005/06) Hate crime has existed as a distinct category of criminal law for only 36 years and as a field of academic study for about 20 years, so there is still much to learn about the impacts of the laws upon potential victims and offenders, as well as possible residual effects upon society. The study of hate crimes is important because thousands of people are victims in the United States each yea r (Sandholtz, Langton, & Planty, 2013 ) The effects of being targeted for a criminal event community (Cogan, 2002; Herek, Gillis, & Cogan, 1999). This study focus es o n a
2 specific aspect of social identity, sexual orientation, because antigay hate crimes are unique, disproportionately common, and severely under reported (Gerstenfeld, 2013). Hate crime law was developed with the intention of providing special attention to and redress for minority groups who were historically harassed, discriminated against, and abused by both the criminal justice system and society at large (Chakraborti & Garland, 2009). In the 1960s and 1970s civil rights and victim rights groups comb ined their efforts and convinced states to implement hate crime laws that provide stronger punishments for crimes motivated by bias or prejudice (Jenness & Grattet, 2001). The first state hate crime statute was passed in California in 1978 ( Grattet, Jenne ss & Curry, 1998). Today, f orty five states have some version of a hate or bias motivated crime statute but the form and content of each varies and little is known regarding the direct impacts of the laws (Gillis, 2013; Meyer, 2014). The laws were intend ed to have a positive effect on minority groups and to decrease negative treatment by the criminal legal system, although their efficacy remains unknown (Meyer, 2014). In the U.S. two waves, or tiers, of hate crime laws were passed. The social groups incl uded as bias categories in the first wave of hate crime statutes were those that were historically or traditionally agreed upon by society to be worthy or deserving of specific protection under the law groups based on race, religion, and ethnic or nationa l origin (Grattet, Jenness, & Curry, 1998). Gender, disability and sexual orientation were rarely included as bias categories in the original hate crime statutes. Political interest and advocacy groups, such as Human Rights Campaign and the Anti Defamat ion League, fought to add those categories during the second wave and were largely successful at both federal and state levels (Jenness, 1995; Grattet & Jenness, 2001). But, there is still a
3 great deal of variation in bias categories between the states. Only thirty out of forty five states currently include sexual orientation as a bias category (Gillis, 2013) Fifteen of those also include gender identity, or protection for transgendered or transsexual persons. Some states include other categories such as gender, disability, age, and/or political affiliation (Schweppe, 2012) Not all states explicitly list the groups tha t are covered by their statute. This disserta tion helps begin to explain state variations regarding sexual orienta tion as a bias catego ry in hate crime laws. Research to date has revealed some of the processes through which hate crime laws were proposed and developed (Jenness & Grattet, 2001), developed a typology of hate crime offenders (McDevitt, Levin, & Bennett, 2002) analyzed facto rs in hate crime victimization (Herek, Cogan & Gillis, 2002; Lyons, 2006), discussed issues with the reporting and collection of hate crime data ( Nolan & Akiyama, 1999), and examined aspects of hate crime law enforcement (Jenness & Grattet, 2005), implemen tation (Bell, 2002), and prosecution ( McPhail & Jenness, 2005/06). But, there are still many unanswered questions. W hat factors for example, explain the wide variation in the content of state hate crime laws? Are states passing hate crime laws that ref lect public opinions? What impact are the laws having upon included or excluded social groups? This dissertation begins to address some gaps in the academic literature by considering the interplay of social construction and public policy specifically as it relates to sexual orientation as a bias category in state hate crime law. This chapter introduces the key concepts and goals and defines the key terms of the study.
4 Defining Hate Crime Despite legal consensus about the constitutionality and applicabil ity of hate crime laws, there is no universally accepted definition of hate crime (Chakraborti & Garland, 2009; Gerstenfeld, 2013; Jacobs & Potter, 1998). E motion al bias and institutional and personal prejudice s related to definitial nonconcensus hinder eq uitable enforcement (Gillis, 2013; Mahoney, 1998; Meyer, 2014). Herek (1989), a prominent scholar defined her or his membership in a minority group; they include viol ent assaults, murder, rape, and property crimes motivated by prejudice, as well as threats of violence or other acts of 948). Legal definitions of hate crime vary among the states and at the federal level (Jacobs & Potter, 1998; Lawrence 1994). The federal Hate Crime Statistics based on race, gender and gender identity, religion, disability, sexual orientation, or ethnicity, including where appropri ate the crimes of murder, non negligent manslaughter; forcible rape; aggravated assault; simple assault; intimidation; arson; and destruction, federal definition after t he passage of the Matthew Shepard and James Byrd, Jr. Hate Crime Prevention Act in 2009 ( Hong, 2009). Hate crime is a social construct which lends itself to imprecise definitions T wo key components of hate crime that have consensus include the bias or prejudice of the offender and the social identity of the victim (Jacobs & Potter, 1998). In other words, a hate crime occurs because the offender is motivated by prejudice against the social group(s) to which the victim belongs (Gerstenfeld, 2013).
5 Ther e are four primary types of state hate crime statutes including crime/penalty enhancement, substantive offenses, institutional vandalism, and data collection (Gillis, 2013). S everal states also have statutes mandating law enforcement training for hate cri mes (Smith & Foley, 2010). This dissertation maintain s a focus on the first two types, state statutes that specify criminal sanction or penalty enhancement for hate and bias motivated crimes. Institutional vandalism laws deal only with religious based cr imes and are therefore distinct. Data collection policies fail to represent an active response to crime, and lack the symbolic power of a law that punishes biased behavior (Gillis, 2013). For those reasons, and following previous research neither instit utional vandalism nor data collection statutes are considered for this study. Two different forms or models of hate crime laws appear in state and federal statutes. The first, the racial animus model, is based on the animus the perpetrator holds toward t he victim (Lawrence, 1994) The definition of hate crime found at the federal level in the Hate Crime Statistics Act and Hate Crime Prevention Act follow this model. The second, the discriminatory selection model, punishes the behavior of discriminatory selection and is unconcerned with the motivation (Lawrence, 1994). The discriminatory selection model is exemplified by the Anti Defamation League (ADL) model statute that was constitutionally upheld by the U.S. Supreme Court in Wisconsin v Mitchell (199 3) and has been widely adopted by many states. T he ADL model statute distinguishes hate crimes from non hate crimes during the sentencing phase when a sentence enhancement is applied because of the discriminatory selection of the victim. Lawrence (1994) argued that the discriminatory selection model is insufficient to determine guilt of a bias crime and should only be used as an indicator of animus, but nevertheless, 43 states and the
6 District of Columbia have adopted laws based upon or similar to the ADL model ( Anti Defamation League, 2012 ). Th e definition of hate crime and model of law are important concepts, although they are extraneous to the current work which examines only the states that have enacted law s whi ch specifically criminalize hate or bia s motivated violence. Defining Social Construction been created by politics, culture, socialization, history, the media, literature, religion, and ram, 1993, p. 335). Public policy design is an important force in the social construction of targeted groups. Policy designs shape institutions and culture, impact public opinion and social constructions, and determine the distribution of political and e conomic resources (Carney, 2010; Ingram & Schneider, 1991). Social constructions are influenced by numerous forces. The media, moral entrepreneurs, policymakers, historical events, and the courts are identified as sources of influence in the constructio n of numerous social groups throughout the course of U.S. history (Schneider & Ingram, 2005). Definitions of crime, and hate crime specifically, are socially constructed (Jacobs sult of inherently political processes that reflect consensus only among those who control or determination by state legislatures as to what is criminal law, the actions determined by the courts to be violations of criminal law, and who the criminal legal system treats as criminal offenders are developed and propagated through a process of social construction
7 during which history, tradition, the news media, scholarly research, publi c opinion, and legislative priorities are all weighed and considered (Mogul, Ritchie & Whitlock, 2011; Schneider & Ingram, 2005). Hate crimes and the hate crime laws often are leading news stories and topics of debate in nationally recognized social and p olitical venues (Chakraborti & Garland, 2009). But state level politics are more often focused on local opinions and assumptions that all states respond to national or external pressures similarly are questionable (Lax & Phillips, 2009). This study, ther efore, examines social construction in one jurisdiction -the state level. A theory of the public policy process, the Social Construction Framework, is utilized and further developed in this dissertation (Ingram & Schneider, 1991; Schneider & Ingram, 1993 ) This study is the first to utilize the SCF to examine hate crime laws and aiding in the development of the framework by quantitatively testing and plotting its key variables is an important goal of the dissertation. Sexual Orientation in the United Stat es Sexual orientation as a unique and identifiable characteristic is a fairly recent development in human thought and understanding. Gays and lesbians were not recognized as a distinct social group until the late 19 th and early 20 th centuries when doctor s and psychiatrists began to treat homosexuality not as aberrant behavior but as a personality characteristic or biological defect that distinguished gay and lesbian persons from the general population (Frank, 2009 ; Halperin, 1990 ). Prior to thi s change in professional perspective about same sex behavior, the re was no homosexual or heterosexual identity. During this time period, identif ication as gay or straight was unnecessary because same sex attraction was not associated with a specific personality or
8 identifiable type of person. Sexual orientation as a distinctive social category, thus, is a social construct of our culture (Thorp, 1992). For hundreds of years, g ays and lesbians have constituted a deviant group in American society. Homosexuality a nd gender nonconformity have been persistently associated with danger, depravity, disorder, disease, and violence to the point that assumptions of criminality are an enduring problem for queer people and a systemic issue within the criminal legal system (M ogul, Ritchie, & Whitlock, 2011). Pervading queer criminal archetypes, such as the gleeful gay killer or the sexually degraded predator, are spread through the media and influence the actions and decisions of people in the criminal legal system from polic e officers to judges and prison wardens (Mogul, Ritchie, & Whitlock, 2011) The queer community is heavily policed and punished for defying heteronormative expectations and gender rules (Mogul, Ritchie, & Whitlock, 2011). Sodomy laws were being enforced as recently as 2013 (Bennet Smith, 2013). Police raids of gay social establishments throughout the late 20 th and early 21 st century were a common occurrence that often included physical brutality and verbal abuse (Chakraborti & Garland, 2009; Mogul, Ritc hie & Whitlock, 2011). Even today, harassment, threats, and bias related violence against the LGBTQ community at the hands of police and private security officers is a daily occurrence (Mogul, Ritchie & Whitlock, 2011). Many scholars argue that g ay men a nd women have had their sexualities and private lives policed in ways that heterosexuals never experience (Chakraborti & Garland, 2009; Levy, 2003). Gay communities and queer groups struggled for decades to organize politically, gain legitimacy and asser t equal standing with the dominant groups in American culture
9 and capitalist society (Epstein, 1987; Frank, 2009; Knopp, 1986). Negatively constructed groups however, find it difficult to organize for political action (Schneider & Ingram, 2005). The Sto newall Riots in New York City in 1969 are widely regarded as the beginning of the gay civil rights movement in the U.S (Mogul, Ritchie, & Whitlock, 2011). By asserting a more positive group identity and joining the mainstream culture during the late 20 th and early 21 st century, gays and lesbians have managed to gain political power and challenge their status as a group that is harmful or damaging to society (Schneider & Ingram, 2005). Social movements, interest groups, and anti violence projects brought the issue of anti gay and lesbian violence to the attention of society, made it credible to legislators, and ensured that it would be recognized as a social problem (Bernstein, 2002; Grattet, Jenness & Curry, 1998; Jenness, 1995, 1999; Schneider & Ingram, 2005). As a result, 30 out of 45 states have opted to include sexual not including sexual orientation within the laws sends a message that homophobia and its consequences are accep 304 5). This study proposes that the social construction of gays and lesbians in the states may be a key factor in determining why a state would or would not choose to include sexual orientation in its hate crime law. Impact of Hate Crime Laws on Social Groups To date, extant research has failed to verify the effects of hate crime laws upon bias motivated violence, intergroup relations, or social groups. Gerstenfeld (2013) acknowledged that measuring potential positive impacts, such as the sym bolic importance of the laws for included social groups, or the increased public attention to the issue, would be extremely difficult. In fact, scholars have only speculated about the
10 potential impacts of hate laws on social groups. One concern is that g roups that are not easily identifiable or readily accepted as bias categories for example, the elderly or the homeless, will be upset about their exclusion. Jacobs and Potter (1998) assert ed: e selective depreciation of their Another common argument against hate crime laws is that they draw more attention to intergroup conflicts and may actually increase, rather than decrease, the occurrence of bias crime. Jacobs and Pot ter (1998) place hate crime laws under the realm of identity politics and argue that the laws encourage individuals to focus on their membership in specific identity groups and encourage identity groups to view themselves 31). This delineation between groups and group members may increase resentment rather than build unification. But, no empirical research has been conducted that confirms or denies those arguments. The symbolic value of hate crime laws should not be under estimated, however, as condemnation of violence against specific identities is particularly important in a diverse society. Government can show solidarity with stigmatized groups by declaring that singling them out is unacceptable (Chakraborti & Garland, 2009). In addition to examining the relationship between social construction and hate crime laws, this dissertation is the first attempt to examin e the impact of the inclusion of sexual orientation in a state hate crime law upon the social construction of gays and lesbians in particular state s A primary question of interest is whether or not hate crime laws increase negative opinions of gays as a social group by drawing attention to their difference, as predicted by Jacobs and Potter
11 Considering the v ariation in state hate crime laws, the role of social construction, the socio political status of gays and lesbians, and the lack of research regarding impacts, this study addresses two research questions: 1. Is the social construction of gays and lesbians re lated to the inclusion or exclusion of sexual orientation as a bias category in state hate crime statutes? 2. D oes the addition of sexual orientation to a state hate crime law impact the social construction of gays and lesbians in that state? This study makes three primary contributions to the exisiting literature. First, the work contributes to the growing body of research that utilizes the Social Construction Framework to examine public policy process and design by developing a replicable, quantitative appr oach to determining social construction that can be utilized and verified in future studies. Second, it is the first study to utilize the SCF to examine hate crime laws and adds to our understanding of the existing variations in state hate crime laws by h elping explain how the social construction of target groups relates to bias categories in state hate crime laws And third, the research takes the first step toward examining the social and political impact of hate crime laws on included social groups, a topic of concern that remains un addressed. Dissertation Overview This dissertation examines the interplay of social construction and public policy, specifically examining sexual orientation as a bias category in state hate crime laws. Chapter 2 review s t he scope of the hate crime problem in the U.S. and why it is an important topic that deserves further attention The purpose, impacts, and development of hate crime laws are also discussed, specifically as they relate to sexual orientation and
12 the constru ction of gays and lesbians in the U.S. Forty five states have a law criminalizing bias motivated violence but only 30 include sexual orientation as a bias category. Chapter 3 provides a discussion of the Social Construction Framework and its applicabilit y and usefulness for understanding state variations regarding sexual orientation as a bias category in hate crime law. Gays and lesbians have historically been viewed and treated as deviants, but hate crime law inclusion points to a positive, dependent co nstruction. Th e chapter conclude s with the specific research questions for this dissertation. Chapter 4, Research Design and Methodology, explains the mixed methods being utilized, specifically content analysis and time series analysis, to address three hypotheses rela ted to the research qu estions. The hypotheses address the two primary propositions of the Social Construction Framework, target groups and feed forward effects. The hypotheses propose that either a positive construction or high level of po litical influence is associated with target groups seeking hate crime protections, and also that the inclusion of a bias category in state crime law will have a positive impact upon the target group. Findings are presented and interpreted in c hapter s 5 an d 6 so as to provide answers for the research questions. Results indicate that a positive dependent construction is associated with target group inclusion in state hate crime law and that such inclusion fails to impact social change. In addition, politic al influence may be a key variable for dependently constructed groups seeking policy change. The results are discussed in the final chapter with respect to previous studies, consideration of limitations and contributions, and suggestions for future researc h. Primary contributions are theoretical, but practical implications suggest that additional hate crime laws or bias categories may fail to influence positive change.
13 C HAPTER II LITERATURE REVIEW 1 Introduction This chapter expand s on the topics presented in Chapter 1 : hate crime laws and their impacts, dynamics of sexual orientation in the U.S., and social construction. The purpose is to explain the previous research and gaps in the literature that led to the research questions presented in this study. The chapter begin s by explaining the scope of the problem ; that is, why hate crime laws matter. General statistics are provided and issues with under reporting are discussed. The consequences of hate crimes for victims and their communities will be expl ained, with a specific focus on the LGBT community. Next, the intended purpose and actual impacts of hate crime laws will be discussed, including a summary of arguments both in support for and opposition to hate crime laws. The symbolic intent of hate cr ime laws will be an area of focus in that section. Then, the literature review will move to the history and development of hate crime laws in the U.S., including a specific section on sexual orientation as a bias category. Finally, there will be a discuss ion of the complicated social and dynamic political status of gays and lesbians in the United States and how that relates to hate crime laws. Scope of the Problem While only 1 2 % of crimes in the U.S. are hate crimes, research has shown that hate crimes are distinct from non bias crimes in several important ways and, therefore, worthy of specific attention from legislators and the criminal justice system (Wilson, 1 Portions of this chapter may also appear in a forthcoming article in the Wiley Encyclopedia of Crime & Punishment (Ed.) Wesley Jennings
14 2014; Cogan, 2002) Hate crimes are different in their motivation and consequences. The key distinction between a hate crime and a non hate crime is the biased or prejudiced motivation of the perpetrator (Gerstenfeld, 2013; Jacobs & Potter, 1998) Criminal prosecutors must prove that the offender acted out of prejudice or bias toward the victim because of their real or perceived membership in a specific social group (McPhail & Jenness, 2005/06). Prejudice is a severely negative and hostile attitude held about a person simply because they belong to a particular group and are assumed to hold all the negative qualities that group is stereotyped to have in this society (Allport, 1954) so everyone has prejudices (Allport, 1954; Jacobs & Potter, 1998) Humans l earn from a very young age that they belong to some social groups and not to others, and learn to associate specific traits and characteristics with the groups to which they do not belong (Allport, 1954; Gerstenfeld, 2013) Prejudice emerges when an indiv idual internalizes a negative stereotype and applies it to others without concern for its accuracy (Allport, 1954). When prejudice is acted out in a violent or criminal manner, a hate crime occurs (Gerstenfeld, 2013) The bias of the perpetrator can take many, sometimes overlapping, forms (Wilson, 2014). Most of the reported hate crimes in the U.S. are racially motivated and committed against persons. According to the Uniform Crime Report (UCR), r acial bias accounted for 48.3 % of hate crime incidents in 2012, followed by sexual orientation bias (19.6 %) and religious bias (19%) Of the anti racial incidents in 2012, 66.1 % were committed against blacks/African Americans. Of the religiously biased crimes, 59.7% were anti Jewish, a nd of the sexually biased crimes, 54.6 % were anti male homosexual. These
15 general statistics about hate crime have held true for over two decades. Nearly 60 % of reported hate crimes in 2012 were committed against persons, primarily consisting of simple assault, aggravated assault, and intimidation. The crimes against property (37.9%) were categorized as vandalism/damage/destruction and the remaining incidents were defined as crimes against society ( U.S. Department of Justice, 2013 ). Hate crimes are severely under reported though (Gerstenfeld, 2013) From 2003 2006 only 46% of known hate crimes were reported to the police; that percentage decreased to only 35% from 2007 2011 (Sandholtz, Langton, & Planty, 2013). To account for the lack of reported hate incidents, the National Cr ime Victim Survey (NCVS) and data gathered by advocacy groups such as the National Coalition of Anti Violence Programs (NCAVP) are often utilized (Gerstenfeld, 2013). Together the NCVS and UCR provide information regarding hate crime trends over time. Fr om 2007 to 2011 there was an average of 259,700 nonfatal hate crime victimizations each year in the United States, an increase from the annual average of 195,000 victimizations from 2003 2009. Another significant recent change is the number of religiously biased crimes. From 2007 2011, 21% of hate crime victimizations were perceived to be motivated by religious bias, more than double the 10% reported from 2003 2006. A key characteristic of hate crime that has not changed is the high number of violent vic timizations. The vast majority of hate crime victimizations from 2003 2011 were violent, more than 80% from 2006 2009 and about 92% from 2007 2011 (Sandholtz, Langton, & Planty, 2013). Research has revealed two primary reasons for underreporting, either choose not to report or police officers fail to identify a crime as hate motivated. Victims of hate crimes may be ashamed or embarrassed by their experience fear retaliation, or
16 simply lack knowledge of hate crime laws (Gerstenfeld, 2013) Immi grants may lack adequate language skills or fear deportation (Whitlock, 2012). Many victims are concerned about police bias and fear being blamed by law enforcement for their victimization, what Herek, Cogan, & Gillis ( 2002 ) refer to as re victimization. LGBT victims who are still closeted in some, or all, areas of their life have the additional concern of having to out themselves in order to rep ort the crime (Herek, Cogan, & Gillis, 2002). One study found that incidents of abuse were so common place for LGBT individuals, that they felt no need to report even serious incidents (Browne, Balashi, & Lin, 2011). Perhaps most concerning is the fact that a substantial portion of anti gay hate crimes are committed by police officers, anywhere from 8% (NCAVP, 201 1) to 25% (Berrill, 1992), which further decreases the likelihood of those victims making an official report (Mogul, Ritchie, & Whitlock, 2011). One promising study did find that base hate crime reporting rates substantially increase when the unmatched co unt technique (UCT) is used instead of conventional self report surveys (Rayburn, Earleywine, & Davison, 2003). The existence of a hate crime statute is only the first nece ssary step in acknowledging and prosecuting bias motivated crimes. Law enforcement agencies and criminal prosecutors must then interpret the law and determine how to enforce it (Bell, 2002; Boyd, Berk, & Hamner, 1996) The discretion afforded police officers, prosecutors, and judges, along with personal prejudices and a history of insti tutional bias against minorities, make it difficult for bias motivated crimes to be reported, charged, and sentenced as hate crimes (Chakraborti & Garland, 2009; Maroney, 1998; Martin, 1996; Whitlock, 2012) In some jurisdictions police officers, and enti re departments,
17 have ignored hate crime laws and refused their enforcement (Whitlock, 2012) Other times well meaning officers who are poorly trained or insensitive to minority victims have failed to properly report details of crimes that could have led t o hate crime prosecutions (Maroney, 1998) When a hate crime report reaches the prosecutor it is unlikely to be prosecuted at all, and if it is, will not likely be prosecuted as a hate crime (Gerstenfeld, 2013; Maroney, 1998; McPhail & Jenness, 2005 /06) In many states the sole determination of bias motivation falls with the judge during the sentencing phase, further increasing the chance that personal and institutional biases will prevent an offender from being convicted of a hate crime ( Jacobs & Potter, 1998; Maroney, 1998). Gays and lesbians are disproportionately targeted for hate crimes. While comprising approximately 3.5% 5.6% of the U.S. population ( Gates & Newport, 2013 ) they represent anywhere from 13 22% of hate crime victimizations annually (Wilson, 2014). Research to date indicates that anti gay crimes are a widespread problem (Gerstenfeld, 2013). Using a sample of over 2000 gay and lesbian adults, Herek, Gillis & Cogan (1999) found that 1 in 4 gay men and 1 in 5 lesbians had been a victi m of an anti gay crime during their adulthood. Similarly, Herek (2009) found that about 20% of a national probability sample of gay, lesbian and bisexual adults had been the victim of a crime based on their sexual orientation. Perry (2001 ) claimed that a nti gay hate crimes are often the most brutal of all hate crimes, which is supported by findings that anti gay hate crimes are more severe than other hate crimes ( Dunbar, 2006 ) and are more likely to involve threats and physical violence, rather than prope rty crimes ( Herek, Gillis, Cogan, & Glunt, 1997 ). In general, LGBT victims of violence are much more likely to be victims of sexual assault compared to heterosexual victims, who are more likely to be
18 victims of general assault and shootings (Cramer, McNie l, Holley, Shumway, & Boccellari, 2012). It is difficult to grasp the true extent of hate crimes in the U.S. because of compounding issues regarding lack of victim reporting, voluntary compliance for police agencies, and differences in interpretation and definitions of laws. Early estimates tended to overstate the problem by ignoring these problems and using demonstrated unreliable nation crime data to make their point that hate crimes were an increasing problem (Jacobs & Potter, 1998). For example, Tor res (1999) relied solely upon Uniform Crime Report reporting and zero report ing were a known problem, and yet Torres did not qualify his conclusion. In 1999, 12, 122 agencies covering 85% of the U.S. population submitted hate crime reports to the UCR and 84% of those reported zero hate crimes (Nolan, Akiyama, & Berhanu, 2002). T he accuracy of zero reports is nearly impossible to determine, but scholars have noted how suspicious it is that most zero reports come from the Southern region of the U.S., which has a history of racial discrimination and race based violence, as well as t he second highest overall crime rate in the county (Levin & McDevitt, 1993). Consequences of Hate Crimes Hate crime law proponents have asserted that hate crime victims suffer more than victims of non biased crimes (see Lawrence, 1994; Levin, 1999 ) Seve ral studies have examined the physical and psychological impact that hate crimes have on victims This research found that hate crimes can be uniquely psychologically damaging when
19 compared to non hate crimes, resulting in higher levels of depression, anx iety, and anger (Herek, Cogan, & Gillis, 2002), as well as greater fear and loss of a sense of safety ( McDevitt, Belboni, Garcia, & Gu, 2001 ). LGBT victims of hate crimes, specifically, suffer significantly higher levels of acute stress and anxiety compar ed to heterosexual victims (Herek, Cogan, & Gillis, 2002) Some have concluded that hate crimes are more physically damaging than similar non biased crimes (see e.g., Levin & McDevitt 1993), but no strong empirical research has yet supported that claim ( Gerstenfeld, 2013). Hate crimes affect by sending a message to that community that they are unwelcome and will be targeted (Chakraborti & Garland, 2009). Hate crimes increase segregation, promote social isolation, and lead minority group members to modify behavior in attempts to avoid victimization (Bell & Perry, 2012; Dharmapala & Garoupa, 2004; Herek, 1990; Lynch, 2008; Perry, 2009) Recent research has evaluated the impact of hate crime rates on LGBT adolescents and found a significant relationship between high rates of anti gay hate crime assaults and suicidal ideation, suicide attempts, and marijuana use among sexual minority youth in Boston (Duncan & Hatzenbuehler, 2014; Duncan, Hatzenbuehler, & Johnson, 2014). N o connection was found between general violent crime rates and suicide or drug use, or for heterosexual youth and hate crime rates, adding strength to the findings. These two studies were based on a survey of public school youth in Boston and add s upport to the contention that hate crimes negatively impact minority communities and minority group members. Purpose and Impact of Hate Crime Laws The need for hate crime laws and their purpose in the United States is still a matter of debate among scholars, commentators, and members of the criminal justice
20 system. The original intent of hate or bias crime laws was to increase attention to the problems of vulnerable and marginalized minority groups and to send a message that bias motivated violence would not be tolerated (Chakraborti & Garland, 2009). The symbolic purpose of hate crime laws may be the most comm on argument in favor of them (Gerstenfeld, 2013). Hate crime laws promote a more tolerant society by sending a clear message that bigotry and prejudi ce are unacceptable and by placing minority groups on the same plane as majority groups (Franklin, 2002; Grattet & Jenness, 2001, see also Gerstenfeld, 2013). Mason (2007) further contended that hate crime laws allow for social justice. Proponents also ha ve argued that the unique damage caused by hate crimes can be properly addressed with criminal sanctions and that minority groups need special protection because the judicial system itself is sexist, racist, and homophobic (Cogan, 2002; Maroney, 1998). Ev en James J. Jacobs, a recognized opponent of hate crime laws, acknowledged that they could improve community relations with law enforcement (1992). Furthermore, adding new categories such as gender or disability recognizes the distinct vulnerability of th ose groups (Jenness, 2002/2003). Early critics argued that criminalizing bias would only exacerbate social problems, increase resentment of minorities, and draw attention to intergroup conflicts (e.g. Gerstenfeld, 1992; Jacobs & Potter, 1998). Many co mmentators argue that c riminal laws cannot solve social ills and the prejudice that leads to hate crimes cannot be resolved with a prison sentence (Franklin, 2002; Jacobs & Potter, 1998; Maroney, 1998). Prisons unlikely to reform a racist or bigot (Gerstenfeld, 2013, p. 82). To make matters worse, hate crime laws are being
21 institutionalized within a criminal legal system that has well known and ongoing issues with bias toward, and harassment of, minority groups (Jacobs & Potter, 1998; Jenness & Grattet, 2001; Maroney, 1998), so they will likely disempower, rather than embolden, minorities (Gerstenfeld, 1992). The positive symbolic intent of hate crime laws can also be used as a criticism. Passing a hate crime la w is an easy way for legislators to respond to bias motivated violence and give credence to minority constituencies without spending any money or resources (Gerstenfeld, 1992; Jacobs & Potter, 1998; Maroney, 1998). Hate crime laws may even inspire complac ency because policy makers can point to the laws and feel satisfied that nothing more needs to be done to combat bigotry (Gellman, 1991). Perhaps the most common criticism of hate crime laws is that they criminalize constitutionally protected freedoms of speech and thought by specifically punishing the motive of the offender (Gerstenfeld, 2013). A key assumption of a hate crime is that there is a message being sent fr om the offender to the victim. Constitutional scholars have argued that despite the vile nature of that message, the government cannot punish contrary or unpopular beliefs, and should, instead, punish the behavior and the effects of the crime on the vi ctim and his or her community This argument was sustained by the U.S. Supreme Court when i Wisconsin v Mitchell (1993) (Lawrence, 1994). No research exists to verify the positive impacts of hate crime laws, and in fact, data exists which indicates the laws may be having negative effects upon minoritie s (Gerstenfeld, 2013). Because hate crime laws are institutionalized within a criminal legal system full of its own biases, is it easy to understand how racial minorities, specifically
22 African Americans, are overrepresented as offenders of hate crimes, ju st as they are of non hate crimes (Franklin, 2002; Maroney, 1998). When hate crime laws are written and implemented it is done without concern for the social group dynamics that initially prompted their development (Franklin, 2002; Maroney, 1998). In the eyes of the law majority group members, such as white, heterosexual males can just as easily be viewed as victims of a hate crime as minority group members. One New York prosecutor reported stories of white males who would embellish their victimization t o try to gain a hate crime conviction against a minority individual (Maldonado, 1992/1993). Appare ntly, the inequalities and prejudices the laws were meant to address are being replicated and the original intent of the laws is lost Data that indicate hat e crime laws have any deterrent effect upon potential offenders is nonexistent despite arguments that they are designed to be preventative (Whitlock, 2012). Rouse (2010) argued that the best means for preventing hate crimes is to improve incident reporting s tatutes, specifically suggesting that the scope of the Hate Crime Statistics Act be broadened and that a more efficient reporting system be developed. Rouse found that some bias motivated crimes, such as the hanging of effigies and nooses on private prope rty, are not captured by hate crime laws in many states. Rather than endlessly expanding the criminalization of conduct, Rouse argued that improved reporting and increased attention to the problem could provide a deterrent effect. Similarly, Glaesar (200 5) suggested that hate crime incidents need to be widely publicized and that politicians need to vilify hate Other suggestions for prevention include the implementation of
23 educational and communit y building programs, like those designed by the Anti Defamation League (Whitlock, 2012). Development of Hate Crime Law in the United States Scholars have argued that hate crimes, and the prejudice which fuels them, have a deep rooted history in North Amer ica (e.g. Mogul, Ritchie, & Whitlock, 2011 ; Petrosino; 1999) The near genocide of two nations of Native Americans in the 1800s, the enslavement of Africans during the same time, the lynching of black Americans in the 20 th century, and the internment of Japanese Americans during World War II have been discussed as prime examples of how crimes motivated by hate and bias against racial and ethnic minorities have not only occurred in the U.S. for centuries, but have often been sanctioned by the state and sup ported by public officials (Levin, 2002; Petrosino, 1999) Hate crimes against persons who challenge traditional norms of gender and sexuality are as equally rooted in U.S. history as those against persons o f other races and ethnicities. C olonial and rel igious authorities for example, accused Indigenous peoples of sodomy and sexual deviance and used these behaviors as justification for the elimination of their cultures and the domination of their lands. African slaves were accused of being hypersexual a nd predatory, and immigrants were labeled as sodomites who threatened the country with sexual deviance (Mogul, Ritchie, & Whitlock, 2011). Ethnic intimidation laws, anti lynching laws, and the federal Civil Rights Act are often referred to as the precurso rs to the currently recognized category of hate crime law in the United States (Levin, 2002; Whitlock, 2012) During the second half of the 20 th century civil rights groups advocates, and anti violence projects began to draw attention to the problems of prejudice and violence against minorities. The anti
24 hate crime movement, specifically, drew from and grew out of the broader anti discrimination movements that began with Black civil rights groups in the 1960s which fought for increased p olitical and social opportunities and legal recognition (Grattet & Jenness, 2001). In the 1970s discrimination movement to include them, as well. All of the movements were concerned with both expand ing civil rights and combating discriminatory violence (Grattet & Jenness, 2001). In the 1980s the right movement emerged to advocate on behalf of crime victims, particularly violent crime victims, who were unable to acquire the services and ass istance they needed. The anti hate crime movement was then successful in combining the forces of the civil rights groups with the more motivated violence (Jenness & Gr attet, 2001). Although the reports and statistics about the prevalence of hate crimes were inaccurate and unreliable throughout the 1990s (Jacobs & Potter, 1998), hate crimes nd considerable social activism (Jenness & Grattet, 1996, p. 130). Research is unable to confirm that there was a corresponding increase in violence against minorities or an s ensitivity and attention to the problems of historically discriminated groups that helped convince policy makers to formulate a new category of criminal law (Jacobs & Henry, 1996; Jacobs & Potter, 1998; Jenness & Grattet, 2001). Supporting organizations i ncluded the Anti Defamation League (ADL), Southern Poverty Law Institute, National Gay & Lesbian Taskforce, and many others, that developed reports and compiled
25 statistics and made hate crimes a viable and legitimate social problem. Social movements and c hate crime movement, which in turn proved to be crucial to the development of hate crime in the U.S & Jenness, 2001, p. 671). No direct link between the actual rate s of hate crimes and h ate crime law implementation has been found ( Jacobs & Henry, 1996) ; rather, one study found that hate crime policy arises largely out of party competition, attention to the issue, and the strength of interest groups (Haider Markel, 1998). Haider Markel (19 98) developed an index of state hate crime scope and coverage and tested models of social regulatory policy that accounted for interest groups, political party competition, hate crime rate, and bureaucratic strength. The results indicated that interest gro ups have the most influence over the scope of hate crime policy, while party competition has the strongest influence over both the scope and coverage of hate crime laws. This finding was challenged by Jacobs and Potter ( 1998 ) who observed no relation bet ween political party/ideology and hate crime laws. Neither the hate crime rate nor the strength of law enforcement in the state had an influence on the scope and coverage of hate crime laws. One outlier, the state of Oklahoma, pointed to the additional i mportance of political entrepreneurs, a factor not included in the study. The state of California enacted the first statute to specifically criminalize hate or bias motivated violence in 1978 (Cal. Penal Code Â§ 422.55; Grattet, Jenness & Curry, 1998) Ove r the next two decades hate crime developed into a unique category of criminal law. Forty five states now have some version of a hate or bias motivated crime statute, including statutes that criminally sanction hate crimes, statutes providing civil
26 remedi es to victims of hate crimes, as well as statutes requiring the reporting and/or collectio n of hate crime statistics (Gerstenfeld, 2013 ; Smith & Foley, 2010 ). Thirty of the forty five states with hate crime laws currently include sexual orientation as a b ias category; fifteen of those also include gender identity, or protection for transgendered or transsexual persons ( Anti Defamation League, 2013; Gillis, 2013 ) Some states include other categories such as gender, disability, age, and/or political affili ation (Anti Defamation League, 2013; Smith & Foley, 2010) Gender and disability have not been readily accepted as necessary categories at the state level, though scholars have argued for their inclusion (Jenness, 2002/03; Weisburd & Levin, 1994 ). Other scholars argue a need for the extension of hate crime laws to cover the elderly and the homeless because of their extreme vulnerability and to address the historical lack of police response (Garland, 2011) The lack of consistency in the content and cover age of hate crime laws, as well as large variations across jurisdictions regarding the implementation and enforcement of the laws are major concerns for advocates and critics alike (see Jacobs & Potter, 1998; Maroney, 1998; Meyer, 2014). At the state leve l, hate crime law went through a process of homogenization and domain expansion during the 1990s during which state legislatures increasingly adopted laws that followed the same model or form, while also adding new or unique bias categories that distinguis religion were already institutionalized in federal discrimination law when hate crime laws were introduced in the legislative arena, so those were the first categories to be included and continue to be the only categories present in all state hate crime laws (Grattet & Jenness, 2001; Jacobs & Potter, 1998). Less established bias categories, such as sexual
27 orientation, gender, and disability have required the influence of policy advocates and the p ower of interest groups to be added to state laws (Jacobs & Potter, 1998; Jenness & Grattet, 2001). In contrast, collective action was not required for gender or disability to be added to federal laws because they were already recognized categories (Gratt et & Jenness, 2001). Social movement representative s for gays and lesbians had to repeatedly provide federal testimony in order to convince Congress that anti gay violence was similar to other forms of hate crime and was worthy of inclusion. Wh time that the federal government recognized sexual orientation as a legal category (Jenness, 1999). The government however, was not perce ived as taking the issue of anti gay and lesbian violence seriously Consequently watchdog organizations, civil rights groups, and LGBT resource and community centers responded by starting anti violence projects aimed at the prevention of violence and the education of key players like attorneys and policy makers. The anti violence projects included victim assistance, crisis intervention, street patrols, and educational campaigns. The activism of these groups was crucial in making anti gay and lesbian vio lence visible and credible scholars and policy makers (Jenness, 1995). In 1988, only 11% of state hate crime statutes included sexual orientation. By 1998, 50% of state hate crime statutes included sexual orientation, and today, the percentage has risen to 66% (Grattet & Jenness, 2001). The increasing regularity with which sexual orientation was being added to state laws (Grattet & Jenness, 2001) seems to have stopped. The last time a state amended or passed a new
28 hate crime law to include sexual orient ation was in C o lorado in 2005 (Human Rights Campaign, 2014). The timing of policy adoption is a key factor in determining the form and content of a state hate crime law (Grattet, Jenness, & Curry, 1998). Most of the original state laws passed in the 1980s included only the traditionally accepted bias categories of race, ethnicity, and religion. It was not until the following decade when the gay rights movement gained momentum and political power t hat sexual orientation became a more readily accepted category (Jenness, 1995). In most cases, the longer a state waited to adopt a hate crime policy, the more pressure it received and the more likely it was to adopt a statute that includes not only the r eadily accepted categories of race, ethnicity, and religion, but also the debated categories of sexual orientation, gender identity, and/or disability. Some states such as California, that passed their original hate crime statutes without much external p ressure have updated or amended their laws to include previously unconsidered or contested categories like gender identity and sexual orientation (Jenness & Grattet, 2001). There are several states, however, where this finding does not seem to apply. I n the same year (1981) for example, the bordering states of Oregon and Washington both passed a hate crime statute. Washington included sexual orientation as a category in its original law, but Oregon added it in 2007. In Idaho, which borders both Oregon and Washington, a state hate crime law was passed in 1983, but it does not include sexual orientation as a bias category. Though policy diffusion literature often includes geographic proximity as a key variable, it is rarely found to be significant in det ermining the adoption of criminal justice policies (Bergin, 2011), and was actually found to be a negative indicator of hate crime policy diffusion (Soule & Earl, 2001).
29 Grattet and Jenness (2001) determined that hate crime laws evolved similarly in the le gislative and judicial realms through homogenization and expansion. The definition of what constitutes a hate crime was agreed upon by the courts and legislatures and new categories were added under the established framework. Once a body of law is establ ished and the parameters for its applicability accepted, it becomes easier to add new provisions that meet the understood criteria without external pressure or debate (Jenness, 1999). T he premise, however, appears not to be true for the addition of sexual orientation to state hate crime laws. As previously noted, social and political activism was required for anti gay and lesbian violence to be taken seriously. There are still 15 states in which sexual orientation has not been added as a category in hate crime law and 5 states (Indiana, Wyoming, Georgia, South Carolina, and Arkansas) that still have no hate crime law (Anti Defamation League, 2013; See als o Figure 4.1 ). Deciding which bias categories to include in hate crime can be a contentious politica l process (Jacobs & Potter, 1998) and sexual orientation is the one that has been the subject of the most debate (Gerstenfeld, 2013). While people tend to agree with hate crime laws in general, they are less likely to do so when sexual orientation is sugg ested for inclusion (Johnson & Byers, 2003). There are several reasons for social and political opposition to the inclusion of sexual orientation as bias category in state hate crime law. Anti gay sentiment and anti gay ideology is still accepted and pro moted in several religious, social, and political institutions in the U.S (Whitlock, 2012). Those who maintain heteronormative and anti gay beliefs think that including sexual orientation in state laws sends a message that the government supports the same sex relationships to which they are opposed. Some opponents have tried to argue that sexual orientation is a
30 choice and not an immutable characteristic like race or ethnicity and therefore is not worthy of inclusion (Gerstenfeld, 2013). Lyons (2006) fou nd that survey respondents were sensitive to the racial status of crime victims but revealed no evident sympathy for gay and lesbian victims, which may indicate that the high level of stress and anxiety those victims experience is based on an accurate unde rstanding of societal response. Mason (2007) noted that images of deep moral failure can preclude an emotional response to a queer community had to convince society and the criminal justice system that they are u ndeserving of prejudice or violence As a result, violence against gays and lesbians is not conside red a hate crime in 15 of the 45 states that criminalize bias motivated behavior. Hate crime policy adoption has followed a similar pattern identified by Schneider (2006) for state incarceration levels. Schneider (2006) found that states do not exhibit Jenness, and Curry (1998) might refer to it, but have gone through periods of conv ergence and divergence (homogenization and differentiation), including times of synchronous change (diffusion) during which many states move in the same direction at the same time. Social construction theory expects that states will change together becaus e legislators will learn (Schneider, 2006). In fact, one of the primary reasons given for the wide adoption of hate crime laws is that it allows legislators an opportunity to appear both tough on crime and considerate of minority interests (Jacobs & Potter, 1998). State level variations and internal factors, however, prevent perfect synchronization or complete homogenization from occurring, despite the influence of nationa l political and social forces (Grattet,
31 Jenness, & Curry, 1998; Schneider, 2006). This study hypothesize s that the key state level differences preventing equitable adoption of sexual orientation in state hate crime laws center s on the social construction of gays and lesbians. Social and Political Status of Gays and Lesbians Gays and lesbians have a dynamic social and political status, or social construction, that is the subject of continual debate and manipulation by competing policymakers, the media, and the courts (Schneider & Ingram, 1993; 2005). Social (Schneider & Ingram, 1993, p. 335). Social constructions are composed of two key elements, public opinion and political power (Weiss, 2006). Gays and lesbians are negatively constructed in comparison to the dominant sexual orientation in the world heterosexuality. For a n egative construction to exist there must be a positive one for it to be constructed against (DiAlto, 2005). Heterosexuality is privileged status in social and welfare polic y (Carabine, 1996; Herek; 2006). Heterosexuals and heterosexual relationships are advantaged in numerous public policies including healthcare and tax codes, societal customs and institutions such as marriage and child rearing, religion, and the legal syst em (Carabine, 1996; Herek, 1990, 2006). 317). Even without empirical analysis, logic suggests that heterosexuals are po sitively constructed in
32 American politics and policy design. Expectations of heterosexuality fuel discrimination, stigmatization, and violence against the LGBTQ community (Herek, 1990). Nationally, gays and lesbians are increasingly gaining civil rights, but first they had to construct a pos itive social identity (Schneider & Ingram, 2005). Research show s that there was a significant shift in the final decade of the 20 th century in which mass information about homosexuality and increased awareness of gay and lesbian persons combined with a decrease in the effects of moral traditionalis m to allow for greater support of gay rights and policies (Brewer, 2003). But state by state the exact social construction and status of gays an d lesbians differs greatly. The majority of states (33) do not yet issue marriage licenses for same sex couples, and in only 22 states can same sex couples legally protect their families through second parent adoption (Human Rights Campaign, 2014). Furth ermore, in 33 states an LGBTQ individual can legally be fired from their job simply because of their sexual orientation or gender identity (Human Rights Campaign, 2014). A t the same time, 30 state hate crime statutes that include sexual orientation seem t o acknowledge that gays and lesbians are a minority group in need of enhanced support from the government because of the existence and potential for acts of violence and harassment. So, like hate crime laws, the status and perception of the gay and lesbia n community across states is varied and wide ranging. Gays and lesbians are a popular target of policy makers. Politicians have used them, and other minority groups, as scapegoats during economic downturns and to drudge up votes during election season (Sc hneider & Ingram, 2005). Over the past two decades gays and lesbians have been the targets of numerous, sometimes competing, social policies at both federal and state levels. The courts have played a key role by
33 overturning federal and state prohibitions to same sex marriage, as well as contributing to gay policy (Human Rights Campaign, 2014) but gays and lesbians continue to be second class citizens in many ways and continue to be targeted by law enforcement and politician s as deviant For instance, in the summer of 2013 police in Baton Rouge, Louisiana gained media attention for attempting to enforce an unconstitutional anti sodomy law. At least a dozen men were arrested for agreeing to consensual sex with undercover off because the law was still on the books, it was their duty to enforce it, regardless of validity (Bennett Smith, 2013). The politics of sexuality are an attempt for previously disadvantaged and negativ ely constructed groups to reconstruct themselves (Schneider & Ingram, 2005). Gay and lesbian politics are similar to other policy areas and typically follow a model of interest group politics in which elite values, interest group resources, and past publi c policies are key variables. When gay and lesbian policy issues are placed on the ballot by opponents, however, then it will likely follow a model of morality politics in which religion, education, partisanship and party competition are key variables (Hai der Markel & Meier, 1996). Mucciaroni (2011) argued that same sex policies rarely qualify as morality politics, even though that is often assumed to be the case, because politicians frame their discussions to avoid questions of morality and instead focus on potential negative consequences for society. Nevertheless, unpopular minorities are often harmed by populist features of democracy like citizen initiatives (Pappas, Mendez, & Herrick, 2009), particularly sexual minorities. Resistance to hate crime laws often develops from opposition to the inclusion of sexual orientation as a bias category (Gerstenfeld, 2013)
3 4 and attitudes about sexual orientation can significantly impact support for state hate crime laws (Johnson & Byers, 2003). There are several fac tors that scholars often find are related to public attitudes about gays and lesbians and related policies. These factors include religion and moral traditionalism (Estrada & Weiss, 1999; Haider Markel & Meier, 1996; Herek, 2006; Olson, Cadge & Harrison, 2006; Wald, Button & Rienzo, 1996), education (Lewis, 2011; Herek & Capitanio, 1996), gender (Herek, 2002; Herek & Capitanio, 1999; Kite & Whitley, 1996), race (Herek & Capitanio, 1995; Lewis, 2003), age (Lewis, 2011), political representation (Haid er Mark el, Joslyn & Kniss, 2000), political opportunity structures ( Wald, Button & Rienzo, 1996), and contact with homosexual individuals (Lewis, 2011). Many of these factors are related to liberal versus conservative political ideologies, as well (see Brewer, 2 003; Lewis, 2003, 2011). Liberal political ideologies are typically associated with pro gay attitudes and support for policies such as same sex marriage, while conservative political ideologies are frequently associated with anti gay attitudes and strong opposition to anti discrimination policies and other laws that could benefit gay and lesbian Americans (Frank, 2009; Green, 1999). When examining the diffusion of criminal justice policies specifically, however, political ideology has rarely been found to be a significant factor (Bergin, 2011). Hate crime laws have generally been a bi partisan policy area as they encompass both conservative concerns for law enforcement and liberal concerns for minority rights (Jacobs & Potter, 1998). Christian conservat ives view any progress made by gay civil rights groups as threatening and perceive support for the gay community to mean that they are now the ones being marginalized and persecuted (Linneman, 2004). R eligion and conservatism
35 for example, are combined in the influential and powerful religious right, led by organizations such as Focus on the Family and interest groups like the Christian Coalition and frequently mentioned in the literature as being directly associated with anti gay attitudes and opinions an d with strong political opposition to policies that appear to be beneficial to gays and lesbians (Brewer, 2003; Frank, 2009; Green, 1999; Haider Markel & Kaufman, 2006; Lax & Phillips, 2009). Religious conservatives led the political fights against repeal to argue against marriage equality at both federal and state level s (Frank, 2009; Herek, 2006). Summary Several factors have been examined regarding the adoption of state hate crime laws, but there is little that can be assumed to be true of every state or jurisdiction in the country. Nationally, media attention and interest group politics played a major role in applying external political pressure to states. Research to date also indicates that the takes when it was finally adopted. In five states it appears that external pressures are not sufficient, as they still fail to have a hate crime law. The 45 states that do have laws lack consistency in their included bias categories. Even states that passed their laws at the same time fail to include the same social groups. So the question remains, what factors determine the state variations for sexual orientation as a bias category? Perhaps the social and political status of gays and lesbians at the state level can provide an answer. No research to date has examined the impact of hate crime laws specifically upon the status or public opinions of the social groups in cluded. I t remains un certain whether
36 the effects have been positive, negative, or otherwise because there is no data regarding the feed forward effects of hate crime statutes upon the social construction of target groups. If the laws are fulfilling their purpose of increasing tolerance and equality, then the impact on social groups included as bias categories should be positive. If the reverse is true and the critics are correct that the laws actually increase social conflict and disempower minorities, t hen the impact on included minority groups should be negative. This dissertation represents an attempt to fill that gap in the literature and assist in understanding whether or not hate crime laws are fulfilling their acclaimed symbolic purpose. Sociologi cal, institutional, and constitutional/legal perspectives have almost exclusively been applied to the study of the development, growth, and diffusion of hate crime laws in the U.S. Valerie Jenness and Ryken Grattet, in particular, have extensively utilize d theories of social movements, social problems, social construction, and institutionalization in order to understand how hate crime laws came to be and the process through which they have been accepted, adopted, and differentiated by state and federal gov ernments (see e.g., Jenness, 1995; Jenness & Grattet, 1996; Grattet, Jenness & Curry, 1998; Jenness & Grattet, 2001). Several other scholars, including critics, have also written from social constructionist and/or institutional perspectives (e.g. Jacobs & Henry, 1996 ; Martin, 1996 ). The social construction of hate crime as a social problem and as a developing policy domain/category of criminal law is a common theme in the literature, yet no scholars to date have utilized the increasingly popular policy d esign and process theory, the Social Construction Framework, to analyze hate crime laws. The
37 following chapter presents the Social Construction Framework, explains its applicability to the study of hate crime laws, and presents two research questions.
38 CHAPTER III SOCIAL CONSTRUCTION FRAMEWORK Introduction Social constructions derive from the same source as prejudice and bias the human need to categorize, label, and separate self sociological and psycho that results in the exaggeration and stereotyping of other groups is often referred to as chneider, 2005, p. 3). There remain asis for the very large differences in social constructions of deservedness, trustworthiness, honesty, and proclivity toward criminality that distinguish popula This chapter prov ides an overview and explanation of a public policy process and design theory, the Social Construction Framework, which can aid in our understanding of the social construction of sexual orientation and its relation to state hate crime laws. Previous appli cations of the theory in studies of criminal justice and gays and lesbians are discussed, along with operationalization of the key variables, criticisms, and the application of the framework to hate crime laws. The chapter concludes with the research ques tions for this study and explanation of their importance and relevance. Framework Overview and Assumptions Schneider and Ingram (1993) developed a specific theory of social construction as it applies to public policy design, the Social Constructio n Framework (SCF), that is considered to be one of the leading theories of the policy process (Pierce, Siddiki, Jones, Schumacher, Pattison, & Peterson, 2014). The SCF seeks to explain why social groups
39 are advantaged or disadvantaged in the political pro cess, how those differences in advantage impact public policy, and how policy can impact the future of those social groups politically, economically, and socially (Schneider & Ingram, 1993). The theory maintains two primary propositions: target population s and feed forward effects (Pierce et al., 2014). Target populations are the groups or individuals chosen for behavior se in political power (Ingram & Schneider, 1991, p. 335). In politics, there is a continual struggle to gain and maintain acceptance of particular constructions of target groups, as well as the consequences of those constructions. Government exploits the values held and shared among citizens every time it singles out a social group or category of persons for policy change (Schneider & Ingram, 2005). The feed forward effects of policy design refer to the consequences of the content and intent of a given policy (Schneider & Sidney, 2009). Policy designs influence opportunity structures and send messages to target groups about how they are viewed and should expect to be treated by the government. Punitive policy designs, for example, restrict benefits and /or allocate burdens to the target group, which sends a signal to that group that they are undeserving of government resources and are unlikely to benefit from participation in the political process (Ingram, Schneider, & deLeon, 2007). Feed forward effe cts influence the social status and political participation of target populations, and can be either instrumental or symbolic in their impact. Positive policy designs will increase the participation, opportunity, and status of the target group, while nega tive
40 policy designs will likely result in less political participation and power within the target group (Ingram, Schneider, & deLeon, 2007). construct target populations in p ositive and negative terms and distribute benefits and deLeon, 2007, p. 93). The actions and decisions of policymakers as both influencing and responding to the public opini ons of social groups ha ve been documented outside of the SCF framework as well (e.g. Brewer, 2003; Mucciaroni, 2011). For instance, Mucciaroni (2011) discussed how politicians respond to calls from religious conservatives to pass punitive policies regardi ng sexual orientation and fo und that they typically avoid framing the issue as one of morality politics. Legislators rely on social constructions when making policy and benefit politically from developing policies that respond to the socially accepted construction of a target group. Constructions are used as precedent for future policy making decisions, especially once they are embedded in law, making them difficult to change (Ingram, Schneider, & deLeon, 2007). Social constructions range from posit ive to negative and target groups vary in degree of political power from weak to strong (See Figure 3. 1). This results in a four fold classification system, or matrix, within which target groups fall (Schneider & Ingram, 1993). Advantaged gh levels of political power resources and enjoy positive social construction as deserving people important in the political and social deLeon, 2007, p. 101). Examples of Advantaged groups include homeowners, members of the military, small business owners, and the elderly. Contenders have substantial
41 political power and resources but are negatively constructed as undeserving of government aid because they are considered u ntrustworthy or morally suspect. Big business, labor unions, the wealthy, minorities, and the Religious Right generally fall under the Contender category. Dependents include those groups that have little political power but are positively constructed as deserving sympathy and aid, for example, mothers, the disabled, children, and the poor. Deviants have neither a positive social construction nor significant political power, such as criminals, drug addicts, gang members and terrorists (Ingram, Schneider, & deLeon, 2007; Schneider & Ingram, 1993). Figure 3.1 : Social Construction matrix Based on Ingram, Schneider, & deLeon, 2007
42 The four fold classification also explains why some groups are more often the target of burdensome policies and why others typi cally receive more beneficial treatment. Those with high levels of power and positive social constructions are most likely to receive benefits from government policies because they have influence and control over the decision making process. Those groups without power and with negative social constructions will find thems elves the targets of burdensome or punishing policies because they lack adequate representation or resources to gain beneficial treatment and can easily be made into scapegoats for ambiti ous policymakers. Advantaged groups often resent government support or economic aid for dependents and deviants (Schneider & Ingram, 2005), which makes it even more difficult for those groups to gain positive status. The types of policies that are aimed at a p articular target group can help identify which area of the matrix a group fits in (Schneider & Ingram, 1993). Gays and Lesbians in the SCF Matrix The current political climate and evolution of public policy in the gay and lesbian policy arena make it es pecially suited for a test of the Social Construction Framework. previous design choices or unintended consequences of policy can sometimes lead to change. Negat ive constructions are especially difficult to change because the punitive treatment oriented or rehabilitative policies (Ingram et al., 2007, p. 109). Gays and lesbians have been subject to negative constructions and punishing policies throughout the 20 th century and into the present (Mogul, Ritchie & Whitlock, 2011; Schneider & Ingram, 2005). Currently, applications of the SCF to issues of sexual orientation are
43 limited and outdated. Not since the 1990s have scholars examined the social construction of non heterosexuals, when four studies explored constructions associated with gay men and gay inmates with HIV/AIDS (Donovan, 1993, 1997; Hogan, 1997; Schroedel & Jordan, 1998). Two of the studies identified gay men as deviants (Donovan 1993, Hogan, 1997) and the others designated them as contenders (Donovan, 1997; Schroedel & Jordan, 1998). Ingram, Schneider, & deLeon (2007) described gay activist groups as 103). Social movements and interest groups, as previously noted, are an important factor in changing the social construction of gays and lesbians from strongly negative to more positive. 2 Gays and lesbians still struggle to find appropriate strate gies for resisting negative attitudes and opinions about them, but have recently gained substantial political power (Schneider & Ingram, 2005). S tudies have shown that target groups are dynamic and capable of moving within the SCF matrix (e.g. Weible, Si ddiki, & Pierce, 2011) but it often takes several decades for a true shift in social construction to occur (Pierce et al., 2014). For instance, Hogan (1997) examined the social construction of target populations involved with prison based AIDS policy and found that several of the groups were able to improve their status over a period of five or more years, resulting in more lenient policies and better care of HIV positive inmates. Gay inmates, however, were unable to remove themselves from deviant status In another example, Weible, Siddiki, & Pierce (2011) quantitatively plotted the movement of 10 target groups related to the Lake Tahoe Water Basin policy arena and showed how they improved their constructions by working in increasingly collaborative set tings over a 17 year period. 2 See section on Development of Hate Crime Laws in Chapter 2
44 The Social Construction Framework is a unique public policy theory because it attempts to explain not solely the process or design of a public policy, as many theories do, but the entire cyclical process of how a policy comes t o be, who it targets and why, and how the design of the policy impacts both future policy making and the status of the targeted group (Schneider & Sidney, 2009). The feed forward effects of policies are important but historically absent of consideration in the public policy literature (Schneider & Sidney, 2009). The idea that policies create politics is often attributed to Theodore Lowi (1979) and the designation of target groups can be traced back to the (Schneider & Sidney, 2009). Policy scholars and institutional theorists, though, have only recently begun to focus on these elements. Typically, the focus of policy scientists has been on policy designs and policy analysis, but Schneider and Ingram n ote the importance of addressing democratic institutions. Criticisms of the Framework edited book, Deserving & Entitled (2005), and its contributing authors are cited fairly extensively throughout this chapter (e.g., DiAlto, 2005; Nicholson Crotty & Meier, 2005). A book review of that text pointed out a repeated assumption of the authors that was never em pirically verified. Schneider and are damaging to democracy and undermine the effective political participation of citizens negative
45 consequences for democracy assumed to result from negative construc tions have not yet been verified empirically. Another critic ism of the book and the theory is that the exact mechanisms through which target groups can change their constructions are not identified, and that examining that link is essential for establishing the framework as a more predictive policy theory (deLeon, 2005) DeLeon also noted confusion around the directionality of dependent and independent variables b social constructions and that social constructions impact policy. An earlier critique of the framework from Lieberman (1995) also referred to this element of the framework as potentially taut ological. Schneider, Ingram, and deLeon (2007) and Schneider and Sidney (2009) however, have since clarified that policy design is viewed as both an independent and dependent variable in the Social Construction Framework. This facet of the framework is essential to understanding social construction in the public policy process. Dozens of scholars have contribu ted to the development of the SCF and answered the authors call to consider both normative and empirical causes and implications of policy design in the U.S. Some key examples are reviewed in the next section. Previous Applications of the SCF The majority of app lications of SCF to date have relied on narratives and historical essays to determine and analyze the social construction of target groups (e.g., Colvin, 2010; DiAlto, 2005; Donovan, 1993, 1997; Hogan, 1997). An analysis of 111 applications of the theory from 1993 2013 revealed that 62% of them were purely qualitative in nature. Additionally, only 11% of the applications examined criminal
46 justice policies, 12% examined state level policies, and 19% examined both propositions of the theory ( Pierce et al., 2014). Of the nine ap plications in criminal justice, six utilized quantitative methods and analysis, such as logistic regression of cross sectional survey data (Brucker, 2007), and multivariate analysis of state regulations, incarceration rates, and polit ical representation (Owens & Smith, 2012). Previous SCF research exploring criminological issues favors quantitative methods. In fact, Nicholson Crotty and Nicholson Crotty (2004) asserted that their study relating negative constructions to state expend 248). The authors measured social construction with four factors, two indicating power (previous state policy and black political representation) an d two indicating social construction (state punitiveness and welfare benefits). Their results showed that states with the most negative constructions of criminals spent demonstrably less on inmate health, jeopardizing efforts to prevent and treat HIV/AIDS and tuberculosis. The topics st udied in criminal justice applications to date include drug use (Brucker, 2007; McKenna,), drug policy (Owens & Smith, 2012; Yates & Whitford, 2009), inmate health (Nicholson Crotty & Nicholson Crotty, 2004), drunk driving laws (Houston & Richardson, 2004), state incarceration rates (Schneider, 2006) and private public partnerships in the prison system (Schneider, 1999). The SCF and hate crime law have yet to be studied. Operationalization of Social Construction A consistent ma nner or model for determining the social construction of a target
47 & Ingram, 1993, p. 3 35). The authors of the SCF emphasize interpretive and qualitative methods and point scholars to text based sources for the determination of so cial constructions. Schneider and construction processes in p olicymaking, policy designs and policy impacts requires interpretive research methods [that] are empirical, take language and other texts/artifacts 115). Thus, most scholars to date have conducted ca se studies and historical analyses, and have relied upon media events, interviews, and legislative histories when making determinations of social construction (Pierce et al., 2014). For example, DiAlto (2005) conducted historical analysis of the changing construction of Japanese Americans from the 1890s to the present and discussed the role of the media, the courts, and public policy in ch anging that construction from strongly negativ e to overwhelmingly positive Operationali zation of Political Power type of political power is being measured. The SCF allows for any of the three recognized forms of political power to be utilized (Pierce et al., 2014). The first face of power is the most commonly utilized and includes direct influences over policy such as wealth, votes, and the mobilization of political actors ( Bachrach & Baratz, 1962 ). The second face of power involves a less visible type of influence ; the ability to control information and keep issues off the policy agenda, for instance, and operationalization of
48 and rat 4). Quantitative Applications Schneider and Sidney (2009) claimed that policy design theory, like the SCF, a and of 115). Despite the heavy qualitative and normative focus of the SCF, a few scholars have accepted the challenge and applied quantitative techniques to determine the social construction of t arget groups and to visualize their movement from one construction to another. Nicholson Crotty and Nicholson Crotty (2004) were the first to attempt a quantitative assessment and utilized factor analysis to determine a social construction score based upo n four factors, two indicating power and two indicating construction. The indicators used in the study were specific to state expenditures for inmate health and not generalizable to a study about hate crime laws. As Ingram, Schneider, and deLeon (2007) w rote set in specific contexts have generalizable or repeatable measures for social con struction. The first, and only known, attempt to quantify the SCF matrix was done by Weible, Siddiki, and Pierce (2011) when they showed the movement of target populations through the matrix quadrants by plotting their so cial construction and power at thre e different points in time. Determination of social construction and power was based on questionnaires in which respondents scored each target group on a scale of 0 100 regarding their influence (power) and evaluation (construction). The matrix was the n
49 created using multi dimensional scaling with each axis ranging from 0 100. This formulation of the matrix did not allow for negative constructions (<0) because that would not have made sense in the context of their study. Schneider (2006) used time seri es analysis to examine state incarceration rates and potential feed forward effects of historical policies. She found that states tend to move their policies and incarceration rates in the same direction at the same time, but do not converge on a common l evel of incarceration. The lack of convergence was attributed to the feed forward effects of historical policy positions and to the internal characteristics of states th at ensure their distinctiveness over time. Application of the SCF to State Hate Crime Laws T his work represents the first study to utilize the SCF t o examine state hate crime laws. In this study, empirical analysis explores the construction of gays and lesbians and the relationship of that construction to the inclusion of sexual orientati on in state hate crime laws. The inclusion of sexual orientation in a state hate crime law points to a dependent status for gays and lesbians D ependents are viewed with sympathy or pity, and are typically targeted with inadequate or symbolic policy designs bolstered by the heavy rhetoric of legislators. Because dependents lack political power, legislators have little incentive to provide them with mea ningful solutions or benefits (Ingram, Schneider & deLeon, 2007). Hate crime laws allow legislators to pay lip service to minority groups without spending any money, providing an easy way for the government to respond to the problem of bias motivated violence (Gerstenfeld, 1992; Jacobs & Potter, 1998). Hate crime scholars and advocacy groups like the National G ay and Lesbian Task Force have pointed to the unique vulnerability, invisibility, and marginalization of the LGBTQ
50 community as evidence that they deserve to have added protections afforded to them through hate crime laws (Cogan, 2002; Mogul, Ritchie, & Wh itlock, 2011). The use of paternalistic terms and the implementation of purely symbolic policies are typical when dealing with dependent groups ( Ing ram, Schneider, & deLeon (2007). In addition, critics have argued that hate crime laws may actually disemp ower and patronize minorities, which would only reinforce a dependent status (Gerstenfeld, 1992). Research Questions 1. Is the social construction of gays and lesbians related to the inclusion or exclusion of sexual orientation in state hate crime statutes? This research question attempts to address the target group proposition of the SCF. The SCF posits that policy designs often reflect and perpetuate social constructions because policy makers respond to the public opinions being expressed in accepted socia l constructions of target groups In order to be included as a protected category in state hate crime law, gays and lesbians would need to have a social construction that corresponded to the policy design. As previously discussed, the inclusion of sexual orientation as a bias category in state hate crime laws points to a dependent construction for gays and lesbians. Dependent groups lack political power, but are generally viewed positively. Empirical confirmation of the social and political status of ga ys and lesbians in states both with and without sexual orientation in their hate crime laws can help explain what internal state factors are allowing for, or preventing, the addition of sexual orientation as a bias category in state hate crime laws.
51 2. Does the addition of sexual orientation to a state hate crime law have an impact on the social construction of gays and lesbians in that state? The second research question addresses the feed forward proposition of the SCF. The SCF posits that policy designs have consequences for target groups, politics, justice, and democratic institutions. Government can either change or reinforce constructions through policy choices. Policies are not the most important determination of target group cy is the dynamic element through which governments anchor, potential impact of inclusion in hate crime laws on gays and lesbians is an interesting and relevant research q uestion. Summary This study accepts the challenge to expand the range of methods that can be used to study public policy designs and their impacts by quantitatively determining social construction, mathematically plotting target groups on the SCF m atrix, and examining feed forward effects of policy design through time series analysis. By using mixed methods on a large sample size, this study provide s a replicable method for determining soc ial construct ion and plotting target groups. This study further contributes to the growing body of work that recognizes the wide applicability of social construction to the public policy process by adding to the limited amount of research that has examined criminal justice policies at the state level and that takes into account both of the primary propositions of the framework, target gro ups and feed forward effects.
52 The Social Construction Framework can help explain why sexual orientation is excluded i n some state hate crime laws, and also allows for consideration of the social and political impact of that inclusion on gays and lesbians as a target group. First, the social construction of gays and lesbians is empirically determined through content anal ysis of newspaper articles Then, quantitative analysis is used to determine if there is a relationship between the construction of gays and lesbians in a state and the inclusion or lly, a change to the social construction of the target group following the hate crime policy change is examined through time series and regression analyses. The following chapter present s and discuss es the research design, methods, hypotheses, and analyse s that are used to a ddress the research questions.
53 CHAPTER IV RESEARCH DESIGN AND METHODOLOGY Introduction This chapter presents the research design and methodology. Three hypotheses are explained, the two key variables are operationa lized, and a thorough discussion of the sampling technique is included. The chapter conclude s with a brief explanation of the analyses utilized. Multiple Time Series Design This study utilized a multiple time series design with mixed methods. Two phases of data analysis were conducted; first, content analysis was employed to determine the social construction of gays and lesbians in each state, and second, quantitative data analyses were used to test the hypotheses. Time series quasi experimental design, according to Rossell (1975 76), is the preferred approach for analyzing public policy impact. Multiple time series designs improve upon standard longitudinal designs by allow ing the experimental group to be compared with a control group, by fixing the po licy intervention point in time, and by Gottman, McFall, & Barnett, 1969, p. 300). A time series design allows for a pre /post policy comparison of social construction, as well as analysis of the intervening policy variable -the addition of sexual orientation to a state hate crime law. A quasi experimental design further allows for comparison between states that have sexual orientation as a bias category and states that do not. The control/comparison states help account for the national trend towards more positive public opinion of gays and lesbians that has occurred since the 1990s.
54 The multiple time series design is characterized by 1) periodic measurement on a variable 2) an intervening quasi experimental variable somewhere in the series, and 3) a control group that has not received the intervening variable and which can be compared to the experimental group, and can be visually modeled (Campbell & Stanley, 1966): O O O O O X O O O O O -----------------O O O O O O O O O O observation ; in this study each observation is a monthly score for either social construction or political influence in this case, the addition of sexual orientation to a state hate crime law. The first group experienced the intervention and the second group, the control group, did not. Hypotheses The following three hypotheses address the two key propositions of the Social Const r uction Framework discussed in Chapter III : target populations and feed forward effects. The public opinion and political power of the target group is the focus of the first two hypotheses. I t is currently unknown which variable of social construction i s the key to the inclusion of sexual orientation in state hate crim e law. Some evidence suggests that social movements and interest groups were instrumental in getting sexual orientation included in federal hate crime laws and to drawing attention to bias motivated violence at state and local levels Accordingly, political power may play an important role in social construction If political power is necessary, then gays and lesbians will likely hold Contender status in states with sexual orientation in t heir hate crime laws. But, normative
55 criticisms regarding the symbolic intent of hate crime laws use terminology and descriptions that match a Dependent status for target groups included as bias categories. If that is the case, then a positive public opi nion would be the key element, not political power. Therefore, the first two hypotheses address the two variables, social construction and political power, separately. H 1 The social construction of gays and lesbians will be more positive in states that ha ve sexual orientation included as a bias category in hate crime law than in states that do not have sexual orientation included in hate crime law. H 2 Gays and lesbians in states that include sexual orientation will have more political power than gays and l esbians in states that do not include sexual orientation as a bias category in hate crime law. The third hypothesis addresses the potential feed forward effects of the policy design on the target group, gays and lesbians. Public policy can reinforce exist ing constructions or influence change. T he addition of sexual orientation to a state hate crime law is assumed to represent a positive policy design intended to improve the social construction of gays and lesbians. H 3 The social construction of gays and l esbians will be more positive after the passage of a state hate crime law which includes sexual orientation as a bias category compared to the same co nstruction before the law was passed.
56 Operationalization This study examines social construction via the news media. The use of newspaper articles for the analysis makes the construction of media events a key element in the methodology. Three normative orders guide journalists in the determination of what is news: 1) Professional norms of objectivity, fairne ss, and balance, 2) Democratic norms about the role of the press in politics, and 3) Economic norms about efficiency and profitability (Bennet t 1996; Boykoff & Boykoff, 2004). When reporting on political issues, however, perfect objectivity is nearly imp ossible because journalists first and foremost want to build a story line. To do so they identify players in their story who hold the most power, or ability to influence the development and outcome of the story. knowledge, common sense, and decisions about what will make a good story. Journalists make personal decisions about what to include or exclude from a story and often interject this way, political news is socially constructed (Bennett, 1996). Social Construction News articles are a social construction that reflect the political ideologies of both the journalist and the intended reader (Bennett, 1996) and can thus be used as a measure of the social construction of a political target group. News articles about gay and lesbian individuals, events, issues, and policies provide a measure of the social construction of gays an d lesbians in each state. Social constructions are intimately tied to the mass media deserve what kind of government treatment are supported by rationale that is provi ded
57 about the world and existing groups (DiAlto, 2005, p. 84). Policy makers o ften embrace the constructions and beliefs perpetuated by the media and embed them in policy design, institutionalizing them and giving them added credibility and legitimacy (DiAlto, 2005; Nicholson Crotty & Meier, 2005). The news media fails to provide n eutral constructions (Bennett, 1996), but yields the point of the view of the powerful men and women who own and control coverage This is especially problematic for minority groups as control of the media falls into fewer and fewer hands (Gamson, Croteau Hoynes, & Sasson, 1992). Political Power The political power of gays and lesbians is determined by the number of front page news articles about gay and lesbian policy issues that appear in each paper. The ability to control information, influence publi c opinion, and keep controversial policies off the legislative agenda is referred to as the second face of power (Bachrach & Baratz, 1962; Doan, 2011). This study use s the second face of power in its operationalization of the political power variable in t he SCF. The second face of power may be more readily recognized by scholars as political influence. Weible, Siddiki, and Pierce (2011) also utilized the second face of power when they operationalized political power as influence in their study water polic y target groups. Throughout the remainder of this dissertation, the terms power and influence are used interchangeably. The media and political actors prov iding news stories exhibit the second face of power when they determine what information is provid ed to the general public, control
58 how stories are told, and influence and reinforce social and political values. The news media is both influenced by and help shape political power (Cook, 2006). News stories, like social constructions, are a product of p olitics and also have feedback effects that influence future politics (Bennett & Livingston, 2003). Sociologists tend to infer that power or status exists wherever resources exist. W hen a social problem dominates the newspapers the assumption is that it h as powerful sponsors (Gibson, 2011). Journalism research reveals that these assumptions are not off base. Political power and influence is not equally distributed, so advocacy and interest groups must work to build relationships with journalists in order to gain legitimacy and access to credible publicity. News rarely simply happen s ; it often has to be created. Journalists have incentives to get their articles on the front page and to get their stories not just from events that are occurring, but from w hat political actors have to say about it (Yoon & Boydstun, 2014). Government actors and interest groups both utilize the media to gain support for their positions because news coverage influences policy debates (Yoon & Boydstun, 2014). Interest groups get their issues in the news by gaining the respect of journalists. Organizations that are judged by the media to be credible and knowledgeable gain epistemic authority and are able to not only provide their viewpoints on a developing story, but also able to bring stories to the journalists for publication (Motion & Weaver, 2005). Editorial decisions influence which stories are featured and journalistic norms determine that only players with power or credibility are allowed to inform those stories (Benne t, 1996; Motion & Weaver, 2005). Agenda setting studies also show that attention given to a policy issue by political elites can directly impact media attention to the issue.
59 Front page news coverage is a particularly potent symbol of the second face of power, the ability to influence and control information, because stories chosen for the front page represent editorial decisions about what is the most important news of the day and which stories deserve the most attention (Gibson, 2011; Polumbaum, 2000). Desk, the chief editor determines what lands on the front page (Gibson, 2011). Additionally, according to 50% of newspaper editors surveyed in a study of front page news, reader/community interest is the most important factor influencing the decision to place a reflect and respond to what the community feels is important, interesti ng, and/or relevant The frequency of front page news articles covering gay/lesbian topics is used as the indicator of political power because gay and lesbian interest groups, as well as politicians, must be viewed as credible sources with news worthy issues in order to gain media attention. If their stories land on the front page, whether good news or bad, it is a strong indicator of their importance and influence. Granted, it would have been ideal if t he exact proportion of gay and lesbian front page articles could have been determined for each paper, but given the temporal scope of the study, sheer volume of articles under consideration, and availability of the data, that was simply not feasible. Sam ple Out of the total 45 states that have a hate crime statute, t hree sets of paired states were selected based on the timing of hate crime law adoption that previous research indicated to be a key factor in the determination of the form and content of sta te hate
60 crime laws (Jenness & Grattet, 2005). The states chosen in each pair passed their original hate crime statutes within one or two years of each other. One state in each pair later amended its hate crime statute to include sexual orientation and th e other did not. The second state acts as a point of control and comparison for the states that include sexual orientation in their laws (See Table 4.2). One state in each pair includes sexual orientation in its hate crime statute and one does not hav e sexual orientation in its hate crime statute, which makes the sample purposive and non random. States were included or excluded based on the existence of a statute that criminally sanctions or provides sentence enhancements for crimes based on hate, pre sexual orientation (See Figure 4.1 ). The determination of a state criminal statute and verification of the date of legal adoption was based on a study conducted by Grattet, Jenness, and Curry (1998), a 2010 report by th e Congressional Research Service, data compiled by the Anti Defamation League, and a search of state statutes on Westlaw and state government websites.
61 Figure 4.1 State Hate Crime Laws The final sample of states and demographic characteristics from 2000 U.S. Census data are listed in Table 4.1. The census categories included were chosen based on their empirical relation to attitudes and politics regarding gays and lesbians, as previously discussed in chapter four. The categories and their definitions ar e determined by the U.S. population aged 18 e percent of the population determined to be living in a same sex household. This category is not completely accurate because the 2000 Census failed to explicitly ask about sexual orientation or permit respondents to indicate that they were in a same sex relationship. Rather, the Census Bureau derived this statistic based upon responses in which two members of the household stated they were in a relationship, and also happened to be of the same sex.
62 reported adherence to one of the following conservative religions: Orthodox, Evangelical Protestant, Catholic, or Muslim. Table 4.1: State Demographic Frequencies State Female Young White Education Gay Partners Religion Colorado 49.6 25.4 82.8 32.7 0.6 2 8.61 Oklahoma 50.9 23.4 76.2 20.3 0.4 46.25 Texas 50.4 25.7 71.0 22.9 0.6 45.99 Alabama 51.7 23.5 71.1 19.1 0.5 44.22 Tennessee 51.3 24 80.2 26.9 0.5 40.57 North Carolina 51 25.1 72.1 22.5 0.5 29.87 Mean 50.75 24.69 74.22 23.23 0.52 40.22 From ea ch state, two newspapers were chosen as the units of observation from which social construction and political influence were determined. When using newspapers, four potential validity concerns must be considered : prestige, circulation, ideology, and timin g (Earl, Martin, McCarthy, & Soule, 2004 ; Franzosi, 1987 ). These issues are discussed in the following paragraphs. The prestige press includes papers that are nationally and internationally read, such as The New York Times and Washington Post, and journali sm literature consistently distinguishes the prestige press from local press (Boykoff & Boykoff, 2009). To account for prestige, only local or regional daily papers are included in the research To account for circulation, the study attempted to utilize the largest daily newspapers in each state, based on circulation provided by NewsBank, as well as a list of the Top 100 Newspapers in the U.S. compiled media relations firm, Burrelles Luce The Top 100 list was compiled from the Audit Bureau of Circulation s for a six month period in 2006 07 (Burrelles Luce 2007). Seven of the twelve (58%) newspapers in the sample are included on the list,
63 which indicates that the papers are some of most widely read and utilized news sources in the U.S. and provide the most representative sample possible of the actual social construction of gays and lesbians. To account for ideology, one liberal and one conservative paper in each state was chosen for a total of 12 units of observation. Distinguishing between and purposely including newspapers of different political viewpoints is i nstrumental because the news media present distorted partial views of events based on a range of ideologies (Sargent, 2009). News editors are aware of the economic constraints of efficient and pro fitable news delivery (Bennett, 1996). Consumers of the news media seek out papers, channels, and magazines that reflect their personal attitudes about politics and social issues. Gentzkow and Shapiro (2006) conducted an extensive analysis which confirme d that 1) and that this demand drives media positioning. Gerber, Karlan, and Bergan (2009) discovered a clear difference in the way that right and left leaning papers covered the news. Consequently, both liberal and conservative papers from each state were included in order to ensure that the social construction of gays and lesbians is determined based on opinions along the political spectrum, and not j ust from one end or the other. endorsement of either the Republican (conservative) or Democratic (liberal) Presidential candidate in the election closest to the passage of the sta te amendments that added sexual orientation as a protected category. Presidential endorsements by newspapers are
64 the democratic process (Porter, 2004). For the pap ers in Texas, North Carolina, Tennessee, and Alabama ideology was determined by endorsement of Bush or Gore in the 2000 Presidential election. A list of endorsements for that election was compiled by Democracy in Action, a non profit firm that provides te chnological support and training to progressive non profit organizations (Appleman, 2000). For Colorado and Oklahoma, the 2004 election provided the determination of political leaning (See Table 4.2). Difficulties arose in identifying two large newspape rs that represented opposite ends of the political spectrum in Texas and Oklahoma. Texas is large and seemingly diverse, but the major daily newspapers across the state maintained support of their former Governor, George W. Bush, in both the 2000 and 2004 elections. The paper chosen as the liberal leaning news source for Texas, the San Antonio Express News, was p is considerably more centrist, or left leaning, than that of the Dallas Morning News. Oklahoma is a strong newspapers have consistently endorsed republican candidat es. An examination of the editorial sections of the papers, however, revealed that Tulsa World has editors and readers which lean ever so slightly more to the center of the political spectrum than those of The Oklahoman, ( See Table 4.2)
65 Table 4.2: Newspaper Sample based on Passage of Statute and Political Leaning State Original Year of Passage/Updated SO included? Liberal leaning paper Conservative leaning paper TN 1989/2000 Yes Chattanooga Times Knoxville News Sentinel* NC 1991 No Charlotte Observer* Winston Salem Journal TX 1993/2001 Yes San Antonio Express News* Dallas Morning News* AL 1994 No Anniston Star Mobile Press Register CO 1988/2005 Yes The Daily Camera The Denver Post* OK 1987 No Tulsa World* The Oklahoman* *Paper listed as one of Top 100 papers in the U.S. (Burrelles Luce 2007) All relevant articles from each paper were drawn from the NewsBank database of local, regional, and national news sources in th e United States. Keywords searched included: gay, lesbian, sexual orientation, same sex, Pride, and LGBT. The best and most complete results were obtained by running the search terms twice; once to search Headlines and once to search Lead/First Paragraph. This approach guaranteed that all articles covering gay and lesbian issues w ere collected. Temporal scope of the article search limited the sample of articles to ones published between 5 years pre and post implementation of the amended state hate crime statutes. Only news articles, not editorials or columns, were included for analysis. Methods This study utilized mixed methods; content analysis to determine social construction and quantitative analysis to test the hypotheses.
66 Content Analysis Qua ntitative content analysis of news articles determined positive or negative constructions of gays and lesbians. This type of content analysis is referred to as assertion, or thematic, because it attempts to identify attitudes and opinions, or themes, abou t a specific topic (Krippendorff, 2004). Coding The coding framework for this study is located in Appendix A. Each sentence of every article was coded as positive (+1), neutral (0), or negative ( 1) related to its representation and coverage of gay and lesbian policies and social issues. The codes were then summed in order to determine if the article overall was positive, negative, or neutral. Articles were coded randomly to decrease problems with validity related to testing and instrumentation; meani ng that articles were randomly selected from different papers and from different time periods so that no pattern related the source or time period would impact coding. In order to ensure coding reliability, 15% of the articles were additionally coded by a graduate research assistant and the results were compared and checked for at least 85% consistency. Potential threats to validity with a multiple time series design primarily concern testing, instrumentation, and maturation. Maturation was addressed by explicitly considering both liberal and conservative news sources for analysis; th is approach helped temper the generally more positive opinions and attitudes about gays and lesbians that were occurring nationally over the time period covered in this study Testing and instrumentation problems were addressed by randomly coding the sample of articles
67 rather than in chronological order, so that changes or improvements in coding would not be associated with previously discussed chronological changes, ideologi cal perspectives, or state by state differences. Variables Social construction was coded as negative, neutral, or positive ( 1, 0, +1), based es. Whether or not the state had sexual orientation in its hate crime law was also coded as a dummy variable (0 = no, 1 = yes). Analysis The first two hypotheses examined the two primary components of the Social Construction Framework and were analyzed i n the same way. Chi square statistics and odds ratios were calculated to determine whether or not associations exist between the variables and the inclusion or exclusion of sexual orientation in state hate crime laws. The third hypothesis was tested and examined in two ways. First, using time series analysis and second, through the development of a quantitative Social Construction Framework matrix. As previously noted, time series analysis was chosen because it allows for a pre /post policy comparison of social construction, as well as analysis of the intervening policy variable, the addition of sexual orientation to a state hate crime law. Perhaps most importantly, the analysis allow s for easy visualization of the changes in social construction in each state over the entire course of the study. Social construction and political power scores for each state are calculated monthly so that changes over time can be statistically and visually verified. Regression analysis determines the impact of
68 the policy intervention and t tests were manually calculated to determine statistically significant difference s in the slopes of the time series both before and after the policy intervention. In order to view pre and post law differences, frequencies were calculat ed and plotted onto the SCF matrix in order to visually determine if the inclusion of sexual orientation in the state hate crime laws impacted the placement of gays and lesbians. Each state was plotted on its own matrix for easy evaluation. Both social c onstruction and power were calculated in order to correctly place gays and lesbians in the SCF matrix and allow for reliable comparison to state hate crime statutes, and between the experimental and control groups. Power was determined by the percentage o f front page articles both pre and post crime law. Social construction was determined by the frequency of both negative and positive articles, again pre and post passage of the law. The negative percentage was subtracted from the positive in order to come up with a single term that could be plotted on the SCF matrix. Summary T his study utilize s a multiple time series design to analyze the potential impact of a policy intervention. The key va ri ables of the Social Constructi on Framework are operationalized via the news media. Content analysis and quan t itative time series analysis are used to test three hypotheses that address the primary propositions of the SCF and attempt t o answer the research questions. The following two chapters present the results. The first Results chapter present s the findings for research question 1 and the second Results chapter present s the findings for research question 2.
69 CHAPTER V RESULTS I The findings for the fi rst research question are presented in this chapter. First, sample frequencies and missing data are reported. Then the results for the first two hypotheses which correspond to the target group proposition of the Social Construction Framework are presente d and interpreted. Sample Descriptives The final sample included 5023 articles; 2639 (49.9%) were from conservative sources and 2651 (50.1%) were from liberal sources, providing a strong representation across the political spectrum. Similarly, the final sample included 2671 (53.2%) articles from states without the bias category and 2353 (46.8%) from states with sexual orientation in their law, providing a fairly balanced sample. The exact number of articles drawn from each state var ied and are listed in Table 5 .1. The data collection period for each set of paired states was determined by the implementation date of the hate crime law with sexual orientation. Table 5 .1: Sample Frequencies State Count Frequency Data Collection Period Colorado 634 12% 20 00 2010 Oklahoma 1064 20.1% 2000 2010 Texas 1523 28.8% 1996 2006 Alabama 734 13.9% 1996 2006 Tennessee 328 6.2% 1995 2005 North Carolina 1007 19% 1995 2005 Total 5023 100% Missing Data During certain time periods, newspaper coverage i s denoted as
70 News Bank database, and in several other data w ere simply unavailable (Table 5 .2) Several a ttempts to fill those g aps by utilizing Lexis Nexis, Google News, and newspaper websites to find articles during the missing time periods were u nsuccessful. Table 5.2: Missing Data State Newspaper Data Collection Period No data Colorado Daily Camera 2000 2010 1/2000 5/2/2001; 4/21/2008 7/8/2008 Alabama The Anniston Star 1996 2006 1/1/1996 9/30/1998 Alabama Press Register 1996 2006 10/1/2001 1/10/2003 North Carolina Winston Salem Journal 1995 2005 1/24/1996 10/27/1997 1/1/1995 1/23/1996 Research Question 1 Target Groups The first research question regarding the relation ship between the social construction of the target group and the inclusion of sexual orientation as a category in state hate crime laws was addressed through categorical data analysis. Specifically, the first two hypotheses are tested using the entire sample of states and articles. The states with sexua l orientation in their laws are compared to the states without it in order to determine if there is an association between the two variables, social construction and political power, and the inclusion or exclusion of the sexual orientation bias category. Crosstabs and chi square are used to analyze the first two hypotheses because the data are categorical. Then, the target group in each state is plotted on a quantified SCF matrix so that the changes in construction can be visualized and the association be tween construction and bias category inclusion can be further developed.
71 H 1 The social construction of gays and lesbians is more positive in states that have sexual orientation included as a bias category in hate crime law than in states that do not have sexual orientation included in hate crime law. There is no significant difference between the two groups of states when determining the likelihood of neg ative or positive news articles, indicating that social construction is not associated with a target gr Table 5 .3 provides crosstabs with z test results indicatin g the nature of the association between sexual orientation in hate crime law and the social construction of gays and lesbians. Findings show that gays and lesbians fail to have more positive news articles in states that include sexual orientation than in states without. Similarly, gays and lesbians in states without sexual orientation in hate crime laws are not associated with more negative news coverage. These results indicate that the hypothesis is not confirmed because the states with the bias category do not have more positively coded articles, or a more positive social construction of gays and lesbians, than s tates without the bias category. The null hypothesis cannot be rejected, however, because there are significant results. The significant results involved neutral articles. W hen sexual orientation is not in cluded in state hate crime law there are significantly fewer neutral articles than expected and when sexual orientation is included there are significantly more neutral articles that expected. Chi square results show that the relationship is significant ( x 2 = 16.888, p = .000) and that there is a low, but still significant association ( V = .058, p =.000) between the inclusion/exclusion of sexual orientation in state hate crime law and
72 crime law, the odds of neutral news articles covering gays and lesbians ar e 1.35 times higher than in states without sexual orientation in their hate crime law. Table 5.3: Social Construction Crosstabs The significant difference between the two groups of states re lated to neutral news coverage may indicate that less extreme opinions, or more balanced views, are presented in the news coverage of states that provide hate crimes protections to gay s and lesbians. While no causality can be inferred, it may be that the work on behalf of the target group that occurs before the policy change, and/or the resulting change in policy, results in more neutral views of gays and lesbians as a target group. H 2 Gays and lesbians have more political power in states that include sexual orientation than in states that do not include sexual orientation as a bias category in hate crime law. For 104 articles in the sample the page numbers could not be determined bec ause the information was missing in NewsBank. Consequently, 2.1% of the articles in the sample cannot be included in the power variable. This missing data occurred in four of the six states and is noted in Table 5 .6. M ultiple imputations failed to impac t results and Does state have SO in hate crime law? Social Construction Total Negative ( 1) Neutral (0) Positive (+1) No Count 1111 406 154 2671 Expected 1075.2 460 1135.8 Std. Residual 1.1 2.5* .5 Yes Count 911 459 982 2352 Expected 946.8 405 1000.2 Std. Residual 1.2 2.7* .6 Total 2022 865 2136 5023
73 were determined to be unnecessary because the percent missing is low, so the results presented do not represent imputed data. Table 5 .4 provides crosstabs with z test results indicating the nature of the association between sexual orientation in hate crime law and the likelihood of front page news articles covering gays and lesbians. When s exual orientation is not included in state hate crime law, there are significantly fewer front page articles than expected and when sexual orientation is i ncluded, there are significantly more front page articles tha n expected. Chi square results show that the relationship is significant ( x 2 = 53.123, p = .000) and that there is a low, but still significant association ( V = .104, p =.000) between the inclusi on/exclusion of sexual orientation in state hate crime law and political power. front page news articles covering gays and lesbians are 1.92 times higher than in st ates without sexual orientation in their hate crime law. These results provide support for the hypothesis a nd indicate that political influence as operationalized by front page news coverage, is associated with target groups that are included as bias cat egories in state hate crime law. Table 5.4: Power Crosstabs Does state have SO in hate crime law? Power (front page) Total 0 1 No Count 2407 228 2635 Expected 2324.8 310.2 Std. Residual 1.7 4.7* Yes Count 1933 351 2284 Expected 2105.2 268.8 Std. Residual 1.8 5.0* Total 4340 579 4919
74 Combined, the results of the first two hypotheses leave much room for speculation about the status of gays and lesbians in the states in which sexual orientation is included. High levels of political power and a negative construction are equated with the Contender category while a positive social construction and low political power would point to Dependent status, and the results of chi square analysis do not clearly point to either construction. The target group may not have a more positive stat us in the states with the bias category than in the states without, as hypothesized, but that does mean that they fail to have a positive construction. One possible explanation is that all the states in the sample exhibit a positive construction and that the target group is dependently constructed throughout. The results showing a significant difference in front page news coverage, t hough, points to political influence as the defining characteristic of the target group, which is associated with Contender or Advantaged status. Further analysis in the following section will clarify and confirm the construction of gays and lesbians in each state in the sample. The SCF Matrix A quantitative method for plotting the pre and post law frequencies was developed so the social construction of the target group in each state could be graphed and examined visually ( See Figures 5.1 5 .6 ). The Social Construction Framework strongly embraces normativity, and only one previous study has attempted to quantify the matrix to show the exact placement and movement of target groups Weible, Siddiki and Pierce (2011 ) provided the first quantified matrix when they plotted the movement of groups related to the Lake Tahoe Basin policy arena by utilizing survey responses on a scale from 0 100. The nature of the data in this study requires the use of a slightly
75 different scale in which negative social constructions can be accounted for. Social construction is placed on the horizontal axis and power on the vertical axis. Social const ruction ranges from 100 to +100, with 0 being the mid point. The range of values for political power varies slightly because it is not possible to have less than 0 front page articles, or a negative amount of political power. So, the range for the power axis is 0 100, with 50 as the mid point Table 5.5: Social Construction Scores State SO Law First Year Implementation Last Year Colorado Yes 43 13 24 Oklahoma No 7 2 33 Texas Yes 1 27 19 Alabama No 15 44 9 Tennessee Yes 18 29 22 North Caroli na No 18 20 14 A plot point, or score, for social construction was calculated by subtracting the percentage of negative articles from the percentage of positive articl es for the first year of the series, the policy implementation year, and again for the last year of the series for each state so that the movement of the target group can be viewed over the course of the period under study (See Table 5 .5). The power or influence, score represents the frequency of f ront page articles, again, for the first y ear of the time series, the policy implementation year, and the final year of the se ries for each state (See Table 5 .6). These two points plotted on x y axes provide a visual representation of gays and lesbians as a target group in the SCF matrix.
76 Tab le 5.6: Power Scores State SO Law First Year Implementation Last Year N Missing Data % Colorado Yes 18 27 27 603/625 3.5 Oklahoma No 5 5 7 964 0.0 Texas Yes 11 9 19 1398/1443 3.1 Alabama No 3 9 4 679/715 5.0 Tennessee Yes 13 28 16 287/288 0.3 North C arolina No 5 14 5 988 0.0 These political power/influence scores show a clear difference between the policy intervention states (CO, TX, and TN) and the control group states (OK, AL, and NC). The intervention states have consistently higher power score s than the control group states, except for one instance in which Texas and Alabama both had a score of nine. A SCF matri x for each state was created The pre law status of gays and lesbians is marked with a blue diamond, the implementation year is noted with a green triangle, and the post law status is indicated with a red square. Each set of paired states is discussed, and then a summary of the re search question results is presented. Colorado/Oklahoma In Colorado, gays and lesbians were in Dependent st atus throughout the time per iod under study (See Figure 5.1 ). From the first year to the last year of the series, the target group did not increase their power or construction enough to move into a different quadrant. It makes sense that a group construc ted as dependent were included in the state hate crime law, since dependent groups often have protective policies aimed at them. A protective policy design would also reinforce a dependent construction, keeping gays and lesbians in Colorado in the depende nt category following the policy change. In Oklahoma during the same time period, the target group moved from Deviant to
77 Dependent status (See Figure 5.2 ) The social construction of gays and lesbians in the stat e increased fairly dramatically between t he first and last year of the series, yet sexual The target group was primarily deviant throughout the study, though, and exhibited low levels of political influence when compared to the targe t group in Colorado. Figure 5.1 Colorado Matrix Figure 5.2 Oklahoma Matrix
78 Texas/Alabama Prior to implementation of a new state hate crime law that explicitly listed sexual orientation as a protected category, gays and lesbians in Texas maintained a slightly deviant status (See Figure 5.3 ) In the year of the policy implementation, 2001, the target group held a dependent status due to an increase in their social construction. After the policy change there was an apparent backlash of public opinio n that negatively impacted the social construction of gays and lesbians in the state, but not enough to push them back into the Deviant category. During the same period, gays and lesbians in Alabama experienced little change in their status (See Figure 5. 4 ) Both prior to 2001 and for many years afterward, they remained in Deviant status, except for the year in which the policy change occurred in Texas. Since the target group had a negative construction and little political power, it is not surprising th at the policy change failed to occur in Alabama. Just as in Colorado, findings show a Dependent status associated with target group inclusi Figure 5.3 Texas Matrix
79 Figure 5.4 Alabama Matrix Tennessee/North Carolina Tennessee implemented a hate crime law that included sexual orientation in 2000. Before that time gays and lesbians in the state held Deviant status (See Figure 5.5 ) A temporary positive shift in their social construction and an increase in political po wer occurred during 2000, possibly allowing for the policy change to occur. After implementation however, the target group quickly returned to the deviant category The passage of a hate crime law with sexual orientation in a state in which gays and les bians are negativ ely constructed is unexpected, but the findings reflect what SCF theorists would expect to occur in the rare situation when a positive policy is aimed at a Deviant group; the policy change had a small window of opportunity and was sure to experience slippage. During the same period in North Carolina, gays and lesbians had low political power but positive social construction, placing them squarely in Dependent status (See Figure 5.6 ) Based on the SCF, the sexual orientation law was much m ore likely to have passed in North Carolina than in Tennessee.
80 Figure 5.5 Tennessee Matrix Figure 5.6 North Carolina Matrix Summary of H 1 and H 2 Results Plotting the precise social construction of the tar get group in each state helps further explain the results of the chi square analyses. It is now evident that in two of the states without sexual orientation included in their hate crime law, Oklahoma and North Carolina, gays and lesbians nevertheless gained or maintained a positive social constructio n during the time period under study. This finding certainly explains why there was not a significant difference in the amount of positive news articles between the
81 intervention and control group states. Furthermore, there is one state, Tennessee, in whi ch gays and lesbians appear to have a mostly Deviant status, yet sexual orientation was added to the state hate crime law in a year in which they temporarily held Dependent status. This finding also explains the lack of a more positive social construction in states with the bias category than in states without. Overall, there is a clear association between a Dependent construction and inclusion as a bias category in state hate crime law. In all three states in which the policy change occurred, the target group was dependently constructed at the time of the policy implementation. In two of the three states in which the policy change failed to occur, the target group was primarily deviant. The potential importance of political influence is a unique finding In the intervention states the target group had significantly higher levels of political influence than the control group. In fact, when looking closely at the political power scores and the matrix placement of the target group in each state, it appears there may be a threshold for political influence at which the inclusion of sexual orientation in a state hate cri me law becomes possible Dependent groups are not associated with significant levels of political power or i nfluence in the SCF literature, y et political influence scores in this sample are a clear distinguishing factor between the states that did achieve policy change and the states that did not. G ays and lesbians in Tennessee for instance, maintained a high level of political influence throu ghout th e time period under study, with an exceptionally high jump in the year of the policy i mplementation (See Table 5.6). The sharp and temporary increase in construction for gays and lesbians in Tennessee in the year 2000 may represent a small
82 was able to take advantage of because of a pre existing level of political influence. Even in the years in which gays and lesbians were considered to be deviant in Tennessee, they still maintained a higher level of p olitical influence than their positively constructed neighbors in North Carolina. North Carolina gays and lesbians had low levels of political influe nce both pre and post policy change. The target group in North Carolina was likely unable to take advant age of any similar window of opportunity because they lacked a sustained level of political influence in the state. If poli tical influence is the key factor associated with target groups in state hate crime law it w ould explain why the policy change fail ed to occur in Oklahoma or Alabama despite increases in social construction in those states. It may be that sustained levels of political influence are necessary for positive policy change, which w ould explain why the target group in North Carolina was u nable to take advantage of a window of opportunity as the targ et group in Tennessee did, and w ould also allow for the temporary low power score in Texas to occur without preventing policy change. Research Question 1 Summary The research question asked if the social construction of gays and lesbians is related to the inclusion or exclusion of sexual orientation in state hate crime laws. The answer appears to be and lesbians held Depende nt status at the time of the pol icy implementation. This provide s strong support for a direct association between hate crime law protections and a specific positive social construction, dependency. However, it is clear that a dependent status is not suff icient for the inclusion of sexual orientation to state hate crime law
83 because gays and lesbians in North Carolina maintained a dependent construction throughout the study, but that state has never added sexual orientation as a bias category to its hate cr ime law. Similarly, at the end of the time period under study, the target group in Oklahoma had increased its construction enough to move into the dependent category, yet that state has not added the bias category to its hate crime law. High levels, and possibly sustained levels, of political influence may also be necessary in order for target groups to achieve positive policy change Even within the Dependent category that is not associated with significant levels of political power or influence, there may a threshold of influence needed in order for pos itive policy changes to occur. This is an important finding for social construction theorists because Dependent groups are not considered to have significant levels of political power or influence.
84 CHAPTER VI RESULTS II This chapter presents and interprets the findings f or the second research question regarding feed forward effects of policy design. The second research question asked about the potential impact of the inclusion of sexual orien tation in state hate crime law upon the construction of gays and lesbians. The third hypothesis addresses this research question and is evaluated with time series and regression analyses. H 3 The social construction of gays and lesbians is more positive af ter the passage of a state hate crime law which includes sexual orientation as a bias category compared to the same cove rage before the law was passed. For this hypothesis, the term social construction refers to the overall construction and placement of ga ys and lesbians in the SCF matrix, so both the social construction and political power /influence variables are analyzed for each state. First, the entire sample of six states was analyzed in order to determine the significance of the policy intervention, the addition of sexual orientation to three of the six state hate crime laws. Then, to tease out the state level differences, time series are presented for each state in the sample. The model for the time series was an autoregressive order 1, or AR(1), to account for the influence of each previous month on the next month in the series ( McDowall, McCleary, Meidinger, & Hay, 1980). A time variable was calculated with the earliest date in the sample, January 1995, as the zero point, using the following formu la: (Year First year of the series) + (Month 1)/12 The resulting continuous variable was used to run regression analyses for the
85 entire sample with social construction and political power as dependent variables. The independent variable, the policy int ervention, was coded as 1 = yes, 0 = no. Regression results fail to support the assumption that the policy intervention would have a positive impact upon the social construction and political power of the target group ( See Table 6.1 ). Model fit statistic s show that the model has good explanatory power. F values specifically indicate that the null hypothesis must be rejected. The alternative hypothesis presented in this study, however, is not supported. The inclusion or addition of sexual orientation as a bias category to the hate crime laws in Colorado, Texas, and Tennessee did not result in a significant improvement in the status of gays and lesbians. In fact, the poli cy intervention had a negative, though just barely non significant, impact on both s ocial construction ( p = .06) and political influence ( p = .07) in the sample states Table 6.1 Regression Dependent Variable Intervention Female Social Construction Political Power Model Fit Coeff. t Coeff. t Coeff. t Coeff. t Adjust.R 2 F Social Const ruction 0.469 1.89 0.613 3.60* ----0.256 3.16* .069 15.0* Political Power .0197 1.78 0.491 6.51* 0.054 3.44* ----.097 21.4* *p Of all the demographic variables collec ted from the 2000 U.S. Census (S ee Table 4 .1), only the percentage of female persons in the population had a significant impact upon the model. me series becomes a unit of analysis. Because of the resulting small sample size ( n = 6) and the constant nature of the demographic data, only one of these variables could be added to the model at a time. Analysts frequently remove non significant variab les from time
86 series regression analyses in order to determine the most parsimonious model ( Wagner, Soumerai, Zhang, & Ross Degnan, 200 2). Unlike previous research would suggest (e.g., Herek, 2002), a higher percentage of females in the state population d id not have a positive effect upon the status of gays and lesbians in the sample states. Female populations had an unexpected negative and statistically significant effect on both the social construction and political influence of the target group The d ependent variables were included in the regressions as independent variables, as well, because of the known association between the two predicated in the Social Construction Framework. Results show that power has a negative impact on social construction i n the sample states, and likewise, social construction has a negativ e impact on political power. The negative relationship is indicative of either a Contender or Dependent status for the target group, which is what the previous literature suggests the act ual construction of gays and lesbians might be. In this study, the relationship may also have b e e n caused by high numbers of negatively coded front page news articles and indicates that the media portrayal of gays and lesbians plays an important role in t heir social construction. In order to tease out further details and specifically test the hypothesis regarding the impact of the policy intervention in the 3 states in which it occurred (CO, TX, and TN), t scores were calculated to compare the pre policy and post policy slopes of the series. Again, a continuous time variable was calculated that would allow for regressions to be run. T scores are then manually calculated with this formula:
87 Determining the standard error for two different slopes required calculation of the following formula: Each set of paired states is presented together for the purposes of visual examination and quantitative comparison. Time series analysis in SPSS allows for one dependent variable to be plotted at a time, so the social construction and political power score for each state are presented as separate time series. For this s tudy, the data are scored and plotted monthly so that changes over the 9 10 year period for each state can be clearly viewed. Each series has more than the minimum 50 data points recommended for modeling a time series ( Chatfield, 1980). Each series was e xponentially smoothed to account for wide variation around the mean and allow for easier examination of the data. The first step is to visually examine the time series, and then second, to test for statistically significant changes in the dependent variab le (W agner et al., 2002). In each of the figures below, the red line is the observed monthly score for the variable, the light blue line is the smoothed time series, and the dark blue line is what the simple non seasonal model predicts would happen to th e time series based upon the pre policy scores. The black line in the middle marks the month in which the policy intervention was implemented and the purple dashed lines indicate the upper and lower confidence bounds. Colorado/Oklahoma To help account fo r missing data, the Colorado and Oklahoma time series are run from January 2001 through December 2010. In June 2005, a new hate crime law that
88 included sexual orientation as a bias category for the first time was implemented in Colorado. The hypothesis s tates that the policy intervention should result in an increase in the social construction and political influence of gays and lesbians in the state. Visual examination of the social construction series in Colorado and Oklahoma reveals similar trends in both states. Prior to the policy intervention in Colorado in June 2005, the social construction of gays and lesbians was declining. After the intervention, the observed scores were primarily positive and the model predicted that the slope would remain a t zero None of the post policy data points exceeded the upper confidence interval, suggesting that there may not be statistical support fo r the hypothesis (See Figure 6.1 ). Social construction did increase after the policy intervention (b 1 = .215), thou gh, when compared to the slope before the policy change was implemented (b 0 = 0.373), and the difference is statistically significant, t (118) = 2.497, p the hypothesis is supported. In the comparison state of Oklahoma there was no policy change. The observed scores show a similar time series, however, in which there was a negative dip in social construction between July 2003 and Jul y 2004, and a primarily positive trend leading up to and following the policy intervention in the neighboring s tate of Colorado (See Figure 6.2 ). It appears that the social construction of gays and lesbians in Oklahoma after June 2005 was more positive th an the model predicted, possibly indicating a regional or national trend that impacted both Colorado and Oklahoma. There were several positive national events that occurred during the post policy time period that may account for the increase in both state s. During the 2008 Presidential campaign Barack Obama was the first Presidential candidate in history to endorse civil unions. In 2009 the federal Hate
89 Crimes Prevention Act was passed that included the name of a gay victim in its title, Matthew Shepare d, and added gender identity to the list of federal bias categories. And gay policy, was ruled unconstitutional by federal courts and repealed by Congress. These national events present threat s to the validity of the positive result in Colorado and make any definitive conclusions about causal order difficult. Figure 6.1 Colorado Social Construction Time Series
90 Figure 6.2 Oklahoma Social Construction Political power in Colorado was fairly high throughout the time series (See Figure 6.3 ) with points exceeding the upper confidence level both prior to and after the policy intervention. Comparison of the pre and post policy slopes for power in Colorado fails to confirm the hypothesis. The po litical power of gays and lesbians in Colorado after the policy intervention (b 1 = .363) is not higher than before the policy intervention (b 0 = .213). The difference is not statistically significant, t (118) = 3.8272. In the comparison state of Oklahom a, the time series for power is similar, but power scores are much lower (Figure 6.4 ) As previously discussed in C hapter 5, t he sustained level of political influence maintained by the target group in Colorado may have been a key factor in their ability to achieve positive policy change.
91 Figure 6.3 Colorado Power Time Series Figure 6.4 Oklahoma Power Time Series Texas/Alabama The Texas and Alabama time series ran from January 1996 through December 2006. In September 2001, a new hate crime law that included sexual orientation as a bias category for the first time was implemented in Texas. The hypothesis states that the policy intervention should result in an increase in social construction and political
92 influence of gays and lesbians in the state. Visual examination of the social construction series in Texas (See Figure 6.5 ) shows a slight upward trend prior to the policy intervention with two months that exceeded the upper confidence bounds. After the policy intervention, though, the social const ruction of gays and lesbians was overwhelmingly negative, including a month that dipped below the lower confidence level. The social construction time series for Texas does not support the hypothesis. The social construction of gays and lesbians in Texas actually decreased after the policy intervention ( b 1 = 0.99) compared to their construction before the policy change ( b 0 = .241). But, the difference is not statistically significant, t (130) = 1.205. There may have been a backlash in the state in which moral entrepreneurs attempted to re assert their negative opinions of gays and lesbians. In the comparison state of Alabama the time series is quite similar during the pre policy period showing a steady level that hovered around a zero mean (See Figure 6.6 ). The post policy period was more severely negative in Alaba ma than in Texas, though, with eight months that fell below the lower confidence level for the series.
93 Figure 6.5 Texas Social Construction Time Series Figure 6.6 Alabama Social Cons truction Time Series The political influence of gays and lesbians was high in Texas both before and after the policy change occurred. In both Texas and Alabama the series exhibit several months in which the power score exce eded the upper confidence level but the mean was slightly higher in Texas. Comparison of the pre and post policy slopes f or political
94 influence in Texas supports the hypothesis (See Figu re 6.7 ). The political power of gays and lesbians in Texas after the policy intervention (b 1 = .1 52) is higher than before the policy intervention (b 0 = 0.174), and the difference is statistically significant, t (130) = 3.822, p The power time series for Alabama also shows an increase, indicating that there are other factors beside the policy change influencing the status of the target group. In both Texas and Alabama there was a noticeable increase in political influence that occurred a full two years after the policy change. It is likely associated not with a change in hate crime law, then, but with the 2004 presidential campaigns and the national debate about a constitutional amendment to ban same sex marriage that was occurring during the same time. Fi gure 6.7 Texas Power Time Series
95 Figure 6.8 Alabama Power Time Series Tennessee/North Carolina The Tennessee and North Carolina time series run from January 1995 through December 2005. In July 2000, a new hate cri mes law which included sexual orientation as a bias category for the first time was implemented in Tennessee. The hypothesis states that the policy intervention should result in an increase in social construction and political influence of gays and lesbia ns in the state. In Tennessee, the status of gays and lesbians remained fairly consistent throughout the time period included in the analysis (See Figure 6.9 ) The social construction of the target group did not show a positive increase after the policy intervention ( b 1 = 0.301) when compared to the construction prior to the policy intervention ( b 0 = .176). In fact the status of gays and lesbians in Tennessee was fairly negative throughout the time series, with several months that fell below the lower confidence level. The pre and post policy slopes are not significantly different, t (130) = 3.184, and the hypothesis is not supported. In the comparison state of North Carolina,
96 the social construction of gays and lesbians is much more dynamic, and app ears to be more positive than the construction of gays and lesbians in Tennessee (See Figure 6.10 ) These results, though not indicative of a specific cause and effect relationship, raise questions about the symbolic intent and actual impact of hate crime laws. Figure 6.9 Tennessee Social Construction Time Series Figure 6.10 North Carolina Social Construction Time Series
97 The political power of the target group in Tennessee, again, remained fairly stable throughout the time series, until a series of po sitive spikes occurred between May 2003 and May 2005 (See Figure 6.11 ) The power of gays and lesbians did increase after the policy intervent ion, th ough the impact was delayed by three years. The post policy coefficient ( b 1 = .280) is more positive than the pre policy coefficent ( b 0 = .072), and the difference is statistically significant, t (130) = 3.118, p Again, the increase in political influence occurred in both states, was far from immediate, and may be attributed to the 2004 presidential campaigns and the corresponding debate about same sex marriage (See Figure 6.12 ) Fig ure 6.11 Tennessee Power Time Series
9 8 Figure 6.12 North Carolina Power Time Series Summary of H 3 Results T he hypothesis is not supported by the regress ion and time series results. Regression results for the entire sample show that the policy intervention had a negativ e, rather than the hypothesized positive, impact upon the social construction of gays and lesbians in the sample states. The negative impact was not s tatistically significant, but still warrants concern There is a negative relationship between the two k ey variables, as well, and state level analysis further explains this finding. In two of the three intervention states, Texas and Tennessee, the social construction of the target group decreased while political power increased in the years following the p olicy intervention. In the third state, Colorado, social instruction increased while political power slightly decreased in the post policy intervention time period. These findings also provide further explanation of the results for the first two hypothes e s. Since social construction failed to improve in all three of the intervention states following the policy change, as
99 hypothesized, it is not surprising that the first hypothesis f ailed to be supported. T he social construction of gays and lesbians in t h e policy intervention states f ailed to be more positive than in the comparison states where the intervention did not occur and the time series results help explain this Political power was found to be higher in the intervention states, however, and the statistically significant increase exhibited in both Texas and Tennessee likely contributed to that result. Of all the demographic factors that could have influenced the time series, only the percent female population was found to be significant, and in t he opposite direction expected. A higher percentage of female population had a statistically significant negative impact on the social and political construction of gays and lesbians in the sample states. This is an unexpected finding that cannot be expl ained by this study. Previous research has consistently shown that women are more supportive and tolerant of gay and lesbian individuals and relationships than men but that difference does not appear to translate into positive changes for gays and lesbia ns as a social group. There were some significant results that prevent rejection of the null hypothesis. In Colorado there was a significant increase in social construction following the policy change. In Texas and Tennessee there was a significant inc rease in political influence in the years following the policy change. These findings indicate that changes in the social construction of gays and lesbians in the states vary, and that there are likely factors other than hate crime policy change influenci ng how and to what degree that construction changes. In Tennessee, for example, the state sodomy law was repealed in 1996. While this was a significant policy change for the gay community, the time series shows no correspon ding increase in political infl uence during that time. Furthermore, the increase
100 in front page news coverage that di d occur in Tennessee was fully three years after the change to the state hate crime law and was likely due to the 2004 national election and corresponding political disco urse surrounding a possible constitutional amendment to ban same sex marriage. Unfortunately, this study cannot explain why front page news coverage, and thus political influence, increased during the national gay marriage debate, but not when the state s odomy law was repealed, nor when sexual orientation was added to the state hate crime law. It could be though, that the latter laws which indicate gays and lesbians were moving out of deviant status were not perceived to be politically or socially threate ning, while same sex marriage which indicates a move into advantaged status would be a more contentious issue for those in power. In the comparison states that did not experience a policy intervention, only one exhibited a status change for gays and lesbia ns. In Oklahoma the target group had a substantial increase in their social construction over the course of the time series. The increased social construction in Oklahoma obviously cannot be attributed to a change in state hate crime law, which leaves ro om for speculation about its origins. The state of Oklahoma did not pass any pro gay laws during the time period under examination, and has not done so since. The target group exhibited a dependent status in the final year of the time series, 2010, but n o policy changes in that state would indicate that gays and lesbians have been treated as dependents. Oklahoma does not provide hate crime protections, legal recognition of same sex households, anti discrimination guarantees, or any other positive policy design aimed at gays and lesbians. Research Question 2 Summary The research question asked simply if the policy change would impact the
101 descriptive, but regression r esults show that the policy intervention failed to have a significant effect upon the target group. Each state in which the policy intervention occurred showed changes in the social construction and/or political power of gays and lesbians during the post policy change time period. But, the changes were not consistent and were not all positive, as hypothesized. The comparison of slopes produced some significant results, but those cannot be definitively attributed to the policy change because of timing and other unaccounted for factors and national events Results Conclusion Results of this study indicate that a dependent construction of gays and lesbians is necessary, but not always sufficient, for the inclusion of sexual orientation in state hate crime l aw. In all three of the intervention states the target group held dependent status at the time of the policy implementation. Gays and lesbians also appear to be solidly dependently constructed in the state of North Carolina, however, yet no policy interv ention has occurred there. In Tennessee, the target group exhibited a temporary increase in social construction during the year in which the policy change was implemented, but then returned to deviant status in the following years. Both of those states f ailed to behave as expected. Although a dependent status points to low political power and a positive social construction, political power was the variable found to be significantly associated with the inclusion or exclusion of sexual orientation in stat e hate crime law. Political power was operationalized through front page newspaper coverage, indicating that the abilit y to gain media attention is essential for target groups seeking hate crime protections. This
102 finding also indicates that there may be a critical level of power needed, even within a dependent construction, in order to produce policy change. Comparison of the level of the power in North Carolina, a solidly dependent state for gays and lesbians in this study, shows that front page news co verage in that state was less frequent there than in Colorado, Texas, and Tennessee where the policy change did occur (See Table 5.6). This is a significant finding for social construction theorists, and one that could not have been revealed through purel y normative or qualitative means. The following chapter include s a discussion of the results in the context of existing literature, policy implications, the limitations of the study, and suggestions for future research.
103 CHAPTER VII DISCUSSION This stud y was designed to examine the interplay of social construction and public policy. The research questions presented regarding the impact of social construction on state hate crime policy, and vice versa, were based on two primary assumptions. First, based on the Social Construction Framework, was the assumption that a positive social construction and/or high political power would be necessary in order for the target group, gays and lesbians, to be included in state hate crime laws, and therefore gays and l esbians in states with sexual orientation would be more positively constructed than in states without. The second assumption was that inclusion of sexual orientation as a bias category in a state hate crime law would represent a positive policy change and have a net positive effect upon gays and lesbians as a target group. Research Question 1 The first research question asked if the social construction of gays and lesbians was related to the inclusion of sexual orientation as a bias category in state h ate crime associated with target group inclusion in state hate crime law. Previous studies and theoretical assumptions support the finding that gays and lesbians were cons tructed as a dependent target group in all three intervention states at the time of the policy implementation. As described by Ingram, Schneider, and deLeon (2007), dependents are viewed with sympathy or pity, and are typically targeted with inadequate or symbolic policy designs bolstered by the heavy rhetoric of legislators. Because dependents lack
104 political power, legislators have little incentive to provide them with mea ningful solutions or benefits. Hate crime laws allow legislators to pay lip servic e to minority groups without spending any money, providing an easy way for the government to respond to the problem of bias motivated violence (Gerstenfeld, 1992; Jacobs & Potter, 1998). Hate crime scholars and advocacy groups like the National Gay and Les bian Task Force have pointed to the unique vulnerability, invisibility, and marginalization of the LGBTQ community as evidence that they deserve to have added protections afforded to them through hate crime laws (Cogan, 2002; Mogul, Ritchie, & Whitlock, 20 11). Mason (2014) further contends that it is necessary for victim groups to construct themselves as worthy of compassion in order for hate crime laws to have a positive impact. The use of paternalistic terms and the implementation of purely symbolic policies ar e typical when dealing with dependent groups. In addition, critics have argued that hate crime laws may actually disempower and patronize minorities (Gerstenfeld, 1992; Meyer, 2014), which would only reinforce a dependent status ( Ing ram, Schneider, & deLe on, 2007) as occurred in Colorado and Texas. The defining characteristic of the target group in the intervention states was not simply a positive construction, however, but a sustained level of political influence as well. The higher power scores observed in the intervention states help confirm the importance of interest groups and policy advocates in the adoption of hate crime policies and the addition of sexual orientation as a bias category ( Grattet & Jenness, 2001; Haider Markel, 1998; Jacobs & Potter, 1998; Jenness, 1995;), since interest groups and social movement representatives seek media attention in order to influence social and political change (Yoon & Boydstun, 2014). Direct political power, such as votes, wealth,
105 and resources, is not associat ed with dependent groups. The SCF allows for the operationalization of all three different forms of political power and the results of this study indicate that the second face of power, the ability to influence, may be the form of power that dependent gro ups need or rely on to gain social and/or political change. The higher levels of political influence exhibited in the intervention states of Colorado, Texas, and Tennessee have theoretical implications. The findings for research question one showed a s ignificant difference statistically through chi square analysis, and visually in the SCF matrices between the intervention states and the control group. In a qualitative study the difference between the two groups ma y have been indistinguishable, but quan titative determination of the placement of the target group in the matrix illustrates that levels of political influence and power within the Dependent category may be a key factor associated with policy change Gays and lesbians in North Carolina were co nstructed as dependents throughout the time period under study, yet sexual orientation compared to the intervention states in which the policy change did occur, however, there is a noticeable difference. In North Carolina the target group never obtained high power scores, as were observed in the intervention states. Common sense and normative considerations of social reality reveal that gays and lesbians are not advantaged in any state in the sample, thus the difference in political power scores is not large enough to put the target group into the Advantaged category and cannot be explained by an error in the quantification of the matrix. It appear s that even within the dependent category that is associated with low levels of political power, the amoun t of influence a target group has is important.
106 Quantifying the SCF matrix revealed how gays and lesbians as a target group are moving through the quadrants and changing their const ruction. This study provide s an update on the current social construction of the gay and lesbian community that had not been examined since the 1990s. Studies in that decade examined the construction of gay men and gay inmates and determined that they we re deviant and moving toward contender status. In addition, Ingram, Schneider, and deLeon (2007) power (pp. 103 4). The results of this study do not provide s upport for that status. In two of the three intervention s tates, Texas and Tennessee, there was an increase in political influence and a decrease in social construction in the years following the policy change, indicative of contender status. But in neit her state did the target group end up in the contender category quantitatively. T he rise of political interest and advocacy groups surrounding the gay and lesbian community also gives indication of increa sed political power, but a historically marginalize d group which represents only 3.5 % 5.6% of the population ( Gates & Newport, 2013 ) is unlikely to maintain hig h levels of political power without the support and influence of existing advantaged groups (Aiken & Musheno, 1994). Rather, this study reveals t hat gays and lesbians are improving their social status and moving into the Dependent category. This finding is supported by theoretical assertions that deviant groups must first gain a positive construction befo re increasing political clout (Ingram & Schneide r, 2005). Tennessee stands out as an anomaly that cannot be fully explained by the data in this study. It is quite unexpected that gays and lesbians appear to have maintained a deviant status in the state throughout the majority of the study because the inclusion of
107 sexual orientation to the state hate crime law does not coincide with the types of policies typically aimed at deviant groups. Deviants represent a permanent underclass that is targeted for punishment, blamed for the ills of society, treated with disrespect, and isolated from society (Ingram & Schneider, 2005; Ingram, Schneider, & deLeon, 2007). Deviants are not targeted by policies that send a positive message about their place in society, such as inclusion in state hate crime protections. Evidently, the target group in Tennessee was able to temporarily shift their construction positively enough to get th e hate crime law passed in 2000 which may be indicative of a small window of op portunity that sometimes appears for deviant groups. Th e social construction literature asserts that when a positive policy change is aimed at deviant group, the win dow of opportunity will be brief and the policy change will reflect punctuated equilibrium and p. 108 ). This observation is confirmed in this study by the quick return to deviant status and the lack of any additional positive policy changes in the state of Tennessee The presidential election campaigns occurring that year may have provided the wind ow of opportunity for Tennessee gays and lesbians particularly considering that the Democratic nominee, then Vice President Al Gore, claimed Tennessee as his home state. members, and that positive policy stance combined with national attention on the state may have provided gays and lesbians with a unique opportunity. The pre existing level of political influence allowed for the target group to take quick political ac tion, but the lack of an existing positive social construction meant that the policy change failed to reflect and reinforce a Dependent status, as it did in Colorado and Texas.
108 In Alabama the target group maintained deviant st atus throughout the time perio d under study, and there is nothing about Alabama laws or politics regarding gays and lesbians to suggest that status is inaccurate. In North Carolina the target group maintained dependent status throughout the time series, but as with Oklahoma, there hav e been no positive policy changes in the state to suggest that gays and lesbians are being treated as a dependent group. Further research is required in order to ascertain why gays and lesbians solidly constructed as dependents have not benefited from tha t status in the public policy arena, but findings in this study suggest that a lack of political influence may be the defining factor. A more thorough qualitative examination of the news articles may reveal whether or not relevant interest groups and poli cy entrepreneurs were active in the intervention states and missing in the control group states. Previous research indicates that a strong interest group party competition, or policy entrepreneur s in Tennessee is likely the mechanism that caused the poli cy change to occur and may be the missing link in North Carolina. It could also be, as Soule and Earl (2001) found, that there was a negative regional diffusion impact in which Oklahoma and North Carolina viewed the policy changes in the ir neighboring states as misguided or irrational, and there fore opted not to follow suit. hate crime law, it did not examine factors related to which bias categories would be included or excluded in a particular state. Soule and Earl (2001) also found that states are more likely to be influenced by hate crime policy adoption in other states when the I n Colorado fo r example, when the hate crime law was amended to include sexual
109 orientation the state legislature was dominated by the Democrats while Governor Bill Owens was a R epublican. According to Soule and partisanship should have increased t he likelihood of other states adopting the same policy change. But as this study shows that did not occur. T likely cannot be extended to the discussion of bias categories. Research Question 2 The second research que stion asked what the impact of including sexual orientation in a state hate crime law would be on the social construction of gays and lesbians. Although there were specific and significant changes to either the social or political status of gays and lesbia ns in the intervention states, regression of the time series shows that the addition of sexual orientation to the state hate crime laws actually had an overall negative impact on the construction of gays and lesbians in the intervention states. Specifical ly, political power decreased in Colorado and social construction decreased in both Texas and Tennessee. The decreases cannot be fully attributed to the change in hat e crime law protections, but provide support to the assertions that hate crime laws a lone fail to solve issues of prejudice and may actually exacerbate social conflict and hostile treatment of minorities (Jacobs, 1993; Jacobs & Potter, 1998; Maroney, 1998). Scholars have contended that the deep social divisions that lead to bias motivated crimes cannot be solved with prison sentences and the results of this study indicate they may be correct (Bronski, Pellegrinia, & Amico, 2013; Franklin, 2002; Gerstenfeld, 1992; Mogul, Ritchie, & Whitlock, 201 1; Whitlock, 2012 ). The symbolic nature and purpose of hate crime laws has been widely discussed by hate crime scholars (e.g. Gerstenfeld, 1992; Jacobs & Potter, 1998; Jenness & Grattet,
110 2005, 2008; Mason, 2014). Critics have suggested that hate crime symbolic and serve no instrumental purpose, although at least one study has confirmed that hate crime laws have more than symbolic effects when implemented (Jenness & Grattet, 2008). While this study was not explicitly testing the assump tion that hate crime laws have a positive symbolic purpose, the results can inform the debate. The construction of gays and lesbians as dependent in all three of the intervention states at the time of the policy change lends support to the contention that hate crime laws are symbolic and will have little to no direct impact because those types of policies are typically aimed at dependent groups who do not have the political power necessary to demand more purposeful and relevant policies. The assumption i n the literature is that hate crime laws send a positive message from government to minority groups that can encourage their democratic participation, but fail to actually prevent or deter violence against them (Ingram & Schneider, 2005; Jacobs & Potter, 1 998; Meyer, 2014). The laws are powerful because they are a symbol and prejudice. The laws tell minority groups that they will be protected and treated as equal citizens. Typically, positive policies that provide benefits to target groups result in an increase in their political participation and influence because gove rnment benefits send the message that the target group is welcome in the democratic process and valued in society. Hate crime laws, however, may fail to provide direct benefits to minority groups. As critics have noted, passing a hate crime law does not require the government to spend any additional resources. Providing criminal redress for bias motivated violence is a worthy goal, but fail to offer the type of direct benefit to target groups that, for example, tax
111 breaks or subsidies do. In this sense the laws are purely symbolic and the results of this study indicate gaining inclusion in a state hate crime law fails to have a clear or consistent positive impact upon the target group, and may actually have negative results. This study did not examine exact changes in the political participation of the target group or changes in victimizations based on sexual orientation, however, so it cannot be determined if those exact intended positive impacts occurred. An early criticism by Gellman (1991) that hate crime laws would allow legisl a tors to be complacent about additional measures to combat bigotry may have been correct. Only in Colorado was the change in hate crime law followed by an increase in social construction and additional positive policy changes regarding gays and lesbians. Colorado is the only state in the sample in which same sex couples are afforded legal recognition, second parent adoption is guaranteed, and discrimination in employment based on sexual orientation is illegal ( Human Rights Campaign, 2014 ). The increase in social construction that occurred in the state after the policy change has evidently been sustained and the target group may even be approaching advantaged status. Although Colorado was the last stat e in the sample to amend its hate crime law to include sexual orientation, gays and lesbians in that state are the only ones that seem to have benefited from a sustaine d, positive social construction The change in hate crime policy cannot be described as the sole cause for the continued improv ements that followed it, because several other national political events regarding sexual orientation occurring during the same time period. When placed in context the lack of clear positive impacts is understandable First, hate crime laws have been institutionalized within a legal system that is historically
112 racist, sexist and bigoted. Criminal laws rarely cure social ills. Placing a bigoted individual in a prison system rife with racist and heterosexist violenc e is un likely to have a positive influence. The criminal legal system exists to punish actions, not to change attitudes and beliefs. To make matters worse, law enforcement are often the perpetrators of bias motivated violence and rely upon social stereot ypes about race, gender, and sexuality when making arrests and reacting to citizens. Without extensive and relevant education and training for all members of the legal system, the lofty goal of social justice for discriminated minority groups will fail to occur. Second, r esults of this study Public policy is at least partially representative of public opinion and can infl uence positive change, but usually serves only to bolster existing cultural stereotypes and power imbalances as it tends to reflect the goals and attitudes of those with the most power and influence (DiAlto, 2005; Ingram, Schneider, & deLeon, 2007). Once a construction has been underlined in policy, it becomes exceedingly hard to alter and will impact future constructions and policies accordingly (Pierson, 1993; Schneider, 2006). This is evident in the sample states in this study. In both Colorado and Te xas, gays and lesbians were dependents when the policy change occurred and remained dependent several y ears afterwards, suggesting that the inclusion of sexual orientation to those state hate crime laws simply reinforced the status of the target group. In Tennessee, however, the target group held deviant status both before and after the inclusion of sexual orientation to the state hate crime law in 2000, indicating that the policy change alone was insufficient to change the negative construction of gays an d lesbians in the state.
113 Finally, i power are correlated ( Czech, Krausman, & Borkhataria, 1998, p.1111), and in this study a specific directional relationship was i dentified Not unexpecte dly, in the context of sexual orientation and hate crime law a negative relationship between the two variables exists. The negative relationship found between social construction and political power makes sense when considering the status of gays and lesb ians over the past few decades and their theorized placement in the SCF matrix. A few studies in the 1990s asserted that gay men and gay men with AIDS were successfully moving themselves from deviant to either dependent or contender status (e.g., Aiken & Musheno, 1994 ; Donovan, 1993, 1997), and hate crime law protection specifically indicates dependency. If the gay and lesbian community has indeed moved, or continued to move, into dependent or contender status, then a negative relationship would not only be expected but required. Dependents have positive constructions but low political power, and contenders have negative constructions but high levels of power (Ingram, Schneider, & deLeon, 2007). Policy Implications There are two possible implications for policy and target groups that can be derived from this study. First, the findings in this study confirm the assertion that public policy, in and of itself, cannot create or sustai n social change. The media, the courts policy makers, and moral entrep reneurs also play a critical role perpetuating social stereotypes and influencing policy change especially ones grounded in history and reinforced through past policies (DiAlto, 2005; Nicholson Crotty & Meier, 2005) The policy response to bias motivated violence is based upon centuries of harassment,
114 discrimination, and violence toward minority groups that cannot be quickly alleviated through criminal laws and the resulting expansion of police powers. Social constructions can be altered, but it will tak e more than one type of policy change and several decades to do so. The results of this study indicate that hate crime laws which include sexual orientation, at best, reinforce existing constructions, and at worst, have a negative impact on the target gro up. Policy makers and political activists may want to consider these findings before advocating for more laws or more bias categories because hate crime laws appear to have little to no influence upon the social construction of target groups. Those who w ield influence after all, decide definitions of crime and the results often have preservation of existing social orders than with the safety of the larger Interest groups or social mo vement representatives for target groups that desire hate crime protections can learn from this study that a positive social status, specifically a dependent construction, along with a sustained level of political influence is necessary for them to gain in clusion in state hate crime laws. If a target group hopes to improve their social construction through policy change, this study also indicates that it will likely take more than one policy and more than one decade to accomplish that goal Deviantly con structed gro ups m ay want to consider avenues other than public policy for achieving social change. Scholars have suggested some alternative methods for dealing with bias and prejudice, including the implementation of restorative justice practices (Whitlock 2012) increased reporting of hate crimes (Rouse, 2010), shaming of the offenders (Glaeser, 2005), civil rights education training, and review of the effectiveness of existing statutes (Boeckmann & Turpin Petrosino, 2002).
115 Limitations While the large scope of the study is a strength and the quantitative determination of social construction an important contribution, there are several limi tations that must be considered including issues related to operationalization, sampling, temporal scope, and histo rical threats to validity. Th e sample of states is small, was purposively drawn and primarily represents the Southern region of the United States so results may be limited in their generalizability Replication of the study with more states in different regions of the country would provide more definitive confirmation of the results. T he operationalization of the power variable may not be satisfactorily reliable to some critics, which would limit the strength of the results. Power was operationalized th rough front page news coverage, an indicator of political influen ce specifically, that many scholars are unfamiliar with. Political is not as commonly used in socia 2014; Doan, 2011). Scholars who are not familiar with different forms o f political power and influence or their operationalization may not agree with the operationalization of that variable in this st udy. Social constructions are difficult to transform and target groups often take several decades to successfully change their status (Ingram, Schneider, & deLeon, 2007; Pierce et al., 2014; Weible, Siddiki, & Pierce, 2011). This study spanned no more than 11 years, based upon the understanding among policy scholars that at least 10 years is needed to adequately examine policy change (Sabatier & Weible, 2007). A decade may not be long enough to sufficiently understand changes in the constructions of ta rget groups. Japanese
116 Americans for example, and the combination of policy, the media, and th e courts is credited with what wa s viewed as a quick and dramatic shift (DiAlto, 2005). Gays and le sbians have been trying to change their status for over 40 years, suggesting that the prejudices and negative stereotypes connected to sexual orientation are more deeply entrenched in U.S. politics and culture than those previously attached to Asian ethnic minorities. So while the time span of 10 years was originally thought to be more than sufficient because it provided a large amount of data and met or exceeded suggested temporal scopes for both the study of policy change and the use of time series analy sis, it may have actually been a limitation of the study. There are several factors influencing social constructions and political power on a continual basis, and public policy is only one (Ingram, Schneider, & deLeon, 2007). This study had a limited fo cus on one type of policy -hate crime statutes -and its potential impact upon a specific target group, gays and lesbians. The media also is a widely recognized influence on social constructions and the policy process, which is why it was utilized as the s ource of data for this study (DiAlto, 2005; Gamson, Crote au, Hoynes, & Sasson, 1992 ; Nicholson Crotty & Meier, 2007). This research considered the roles of the media and public policy, but did not include the role of the courts moral entrepreneurs, inter est groups, or any other influences on hate crime policies and social constructions a certain limitation. National news events and policy changes that occurred during the time period under study may also have influenc ed and impact ed the results, particularly for the third hypothesis. Several national events were identified that
117 threaten the validity of the observed positive increases in social construction and political influence that occurred in the intervention states. Two final limitations of this study in clude missing data and the lack of a definite causal order. There is no way to determine how many news articles are missing, so estimating their social and political scores is equally indeterminable. Whether or not the missing data would have changed the results of the study is a lingering and unanswerable question. Because of the manner i n which the data w ere coded and analyzed, the exact causal mechanism that led to or prevented the policy change cannot be determined; neither can the impact of the policy change u pon the target group be fully understood. But the importance of political influence indicates that future research should examine factors related to political power and influence. While the original intention of the study was to analyze public policy as both an independent and dependent variable, the final analyses w ere unable to adequately do so because time series analysis is primarily descriptive and because additional factors, such as national news events and federal policy changes, were unaccounted f or. There are certainly important take aways and contributions made however, by this study to the hate crime and social construction literature Contribution s This study represents the first attempt to examine hate crime laws utilizing the Social Construction Framework. The primary contri bution of th e research is the quantitative determination of the social construction of gays and lesbians at the state level, which had not been previously attempted. Quantifying the SCF matrix allows for several things tha t may not have been possible othe rwise. First, it revealed distinctions in
118 the level of political influence between the target groups that qualitative meth ods would be unable to ascertain. The lower political scores found in the control group states and the sustained levels of political influence observed in the intervention states revealed that political influence may be a causal factor related to the inclusion of a bias category in state hate crime law. Qualitative analysis could not have revealed this key difference between the state s, nor would current normative understandings of Dependent groups have allowed for it to be taken into consideration. This is a particularly relevant finding for SCF theorists who do no t associate political power or influence with Dependent groups. Quan tifying the SCF matrix also provides a replicable method for determining the construction of target groups tha t can be used in future studies, something that the SCF literature currently lacks. Providing a visualization of the movement of the target group in t his study also informs scholars about the actual path toward policy change and an improved social construction that gays and lesbians are following in the U.S. While studies in the 1990s indicated that gays and lesbians were increasing their political pow er and moving toward Contender status as their means for achieving policy change, the results of this research clearly show that they are actually improving their social status and moving into Dependent status. This has implications for how other deviant groups, or previously deviant groups, may achieve policy change and improve their social status. Besides providing a quantified matrix and determina tio n of social construction this study helps to explain wh y gays and lesbians in some states lack specific criminal redress for bias motivated violence against them, and how inclusion in hate crime laws might impact them as a social group. These are important issues to understand because
119 hate crime laws have bee n accepted and institutionalized as an appropriate legal response to a uniquely damaging type of victimization. This study empirically verifies, for the first time, that a positive social construction and a sustained level of political influence is associ ated with target groups who gain, or have gained, inclusion in state hate crime laws. This may be a surprising finding for some hate crime scholars because hate crime laws are often viewed as a means of redress for negatively constructed minority groups, not as a benefit being provided to positively viewed groups. Finally, this study provides strong support for the utilization of mixed methods in the examination of social construction and policy design. Neither purely quantitative nor qualitative methods appear to be able to reveal the entire story of how a target group achieves social and political change or how those changes may impact them in return. Social constructions are highly emotional, symbolic and divisive, and focus on value laden differences placement of target groups in the matrix is somewhat subjective (Ingram, Schneider, & deLeon, 2007) which may be why the majority of applications of the SCF to date have been qualitative This research confirms that the framework can be used quantitatively. The best approach may be to use mixed methods that would allow for consideration of value laden subjects as well as replication and verification of the results. T he deviant status cons istently held by gays and lesbians in Tennessee for example, does not match their inclusion in the state hate crime law, and qualitative analysis of media reports and legislative debates may shed light on how and why the policy change occurred, and then f ailed to have a positive impact. Future studies should utilize mixed methods in order to
120 fill the gaps left unexplained by sole reliance on either quantitative or qualitative forms of analyses. Future Research The results of this study leave many unanswe red questions and provide room for future research regarding the interplay of social construction and hate crime laws, as well as further development of the Social Construction Framework as a tool for understanding public policy processes. Future research needs to further explore the determination of social construction made in this study, and more importantly, the potential importance of political influence for Dependent groups. If political influence is vital for Dependent group s to achieve policy chang e, it c ould repr esent a significant shift in current thinking and understanding of the types of policies aimed at those groups and how they come to be. This study did not specifically consider the role of policy entrepreneurs or the strength of interest gr oups, though it has been identified as an important factor in the inclusion of sexual orientation in federal hate crime law, as well as in sexuality politics more generally. Moral and/or political entrepreneurs have been identified as key players in chang ing or reinforcing both social construction and public policies. Future studies need to explicitly address the presence and involvement of interest groups and policy entrepreneurs. Consideration of those factors, for example, may reveal the exact mechani sm that allowed for the policy change to occur in Tennessee and prevented it from occurring in North Carolina. No internal state factors have been identified that can be consistently associated with the passage of state hate crime laws, but Soule and Earl (2001) introduced the idea that a combination of both internal and external factors must
121 be taken into consideration in order to understand the adoption and diffusion of these policies. No study before this one, though, explicitly looked at state differe nces regarding bias categories. Previous studies only examined the initial adoption of a law criminalizing hate or bias motivated violence. Future research should examine both the determination of bias categories as well as initial adoption of state laws and should take into consideration not only unique internal state characteristics, but also how the state and other jurisdictions respond to external pressures. Scholars have suggested alternative methods for dealing with bias and prejudice, including the implementation of restorative justice practices (Whitlock, 2012) increased reporting of hate crimes (Rouse, 2010), shaming of the offenders (Glaeser, 2005), civil rights education training, and review of the effectiveness of existing statutes (Boeckmann & Turpin Petrosino, 2002). Future research should examine whether any of these alternative methods can more definitively improve the social construction of minority groups than positive, or perceived positive, policy change The existence of a hate crime statute is only the first necessary step in acknowledging and prosecuting bias motivated crimes. Law enforcement agencies and criminal prosecutors must then interpret the law an d determine how to enforce it. Since laws matter only when enforced (Bell, 2 002), future studies of the impacts of hate crime laws should examine the extent to which varying levels of implementation are effecting target groups, and how. Conclusion This research suggests that a positive dependent status is necessary in order for a target group that was previously considered deviant to be inc luded in state hate crime
122 law. Importantly, the target group may need to experience an increase in social construction, as occurred in Colorado, if they hope to continue to improve their overall status and increase the number of positive policies aimed at them. It is evident, though, that political influence also plays a key role because a positive construction is not sufficient in and of itself to create or sustain policy change, as evidenced b y the target group in North Carolina. There is little consensus among judicial, social and political institutions regarding the status of gays and lesbians in the United States, as is partially evidence by the varied and dynamic status of the target gro up in the states in this study. The courts, media, and politicians have not yet combined forces in complete favor of equitable policies and positive treatment of the queer community as they did for Japanese Americans in the mid 20 th century. While the ci vil courts are increasingly ruling in favor of equal rights for LGBTQ individuals and households, the criminal legal system continues to be plagued by homophobia and transphobia. For centuries the criminal legal system has been invested in punishing sexua l deviance, not in the prevention of violence against sexual minorities (Mogul, Ritchie, & Whitlock, 2011). Hate crime laws may be an attempt to address historical imbalances and to prevent acts of bias motivated violence, but they fail to act as a singul ar force that can change attitudes and fail to address the systemic prejudice that plagues the criminal legal system (Maroney, 1998; Mogul, Ritchie & Whitlock, 2011) While laws can shape attitudes, the legal system exists only to punish behaviors (Bronski Pellegrinia, & Amico, 2013), and may, therefore, not be the ideal solution for problems of prejudice, bias, and other damaging attitudes or beliefs.
123 REFERENCES Aiken, J.H., & Musheno, M. (1994). Why have nots win in the HIV litigation arena: Soc io legal dynamics of extreme cases. Law & Policy 16 (3), 267 297. Allport, G. W. (1954). The nature of prejudice. Boston, MA: Addison Wesley Pub. Co Anti Defamation League. (2013). Anti Defamation L e ague state hate crime statutory provisions Retri eved from http://w ww.adl.org/assets/pdf/combating hate/ state_hate_crime_laws_march_2013.pdf Anti Defamation League. (2012). Hate crime laws. Retrieved from http:// www.adl.org Appleman, E.M. (2000). Endorsements by newspapers and magazines. Report by Democracy in Action, Retrieved from http://www.gwu.edu/~action/natendorse5.html Arm, J.R., Horne, S.G., & Levitt, H.M. (2008). Negotiating connection to GBLT experience: ti GLBT movements and policies. Journal of Counseling Psychology August, 1 15. Bachrach, P., & Baratz, M.S. (1962). Two faces of power. The American Political Science Review, 56 (4), 947 952. Bell, J. (2002). Policing hatred : Law enforcement, civil ri ghts, and hate crimes. New York: NewYork University Press. Bell. J.G., & Perry, B. (2012). Outside looking in: The communit y impacts of anti lesbian, gay, and bisexual hate crime. Paper submitted for publication. Bennett, W.L. (1996). An i ntroduction to journalism norms and representations of politics. Political Communication, 13 373 384. independent press: Government control and journalistic autonomy in the po litical construction of news. Political Communication, 20, 359 362. Bennett Smith, M. (2013, July 7). Louisiana police sting targets, arrests g ay men for sex using unconstitutional anti sodomy law. The Huffington Post. Retrieved from http://www.huf fingtonpost.com. Bergin, T. (2011). How and why do criminal justice policies spread throughout U.S. states? A critical review of the diffusion literature. Criminal Justice Policy Review 22 (4), 403 421.
124 Bernstein, M. (2002). Identities and politics: Tow ard a historical understanding of the lesbian and gay movement. Social Science History 26 (3), 531 581. Berrill, K.T. (1992). Anti gay violence and victimi zation in the United States: An overview. In G.M. Herek & K.T. Berrill (Eds.), Ha te Crimes : Confronting violence against lesbians and gay men (pp. 19 45). Newbury Park, CA: Sage. Blitzer, J. (2005/06). Toward an interdisciplinary field of hate studies: Developing a framework. Journal of Hate Studies, 4, 139 149. Boeckmann, R.J., & T urpin Petrosino, C. (2002). Unders tanding the harm of hate crime. Journal of Social Issues, 58 (2), 207 225. Boyd, E. A., Berk, R. A., & Hamner K.M (1996). Motivated by hatred or prejudice : Categorization of hate motivated crimes in two police divisi ons. Law and Society Review, 30 (4), 819 850. Boykin, K. (1996). One more river to cross: Black and gay in America. New York: Anchor Books. Boykoff, M.T., & Boykoff, J.M. (2004). Balance as bi as: Global warming and the U.S. prestige press. Globa l Environmental Change 14 125 136. Brewer, P.R. (2003a). The shifting foundations of public opinion about gay rights. The Journal of Politics 65 (4), 1208 1220. Brewer, P.R. (2003b). Values, political knowledge, and p ublic opinion about gay rights : A framing based account. The Public Opinion Quarterly 67 (2), 173 201. Bronski, M., Pellegrini, A., & Amico, M. (2013). And 20 other myths about LGBT life and people. Beacon Press. Browne, K., Bakshi, L., & Lim Understanding and addressing contemporary lesbian, gay, bisexual and trans safety beyond hate crime paradigms. Journal of Social Problems, 40, 739 756. Brucker, D.L. (2007). Substance abuse treatment partic ipation and employment outcomes for public disability beneficiaries with substance use disorders. Journal of Behavioral Health Services & Research, 34 (3), 290 308. Byers, B., Crider, B. W., & Biggers G.K (1999). Bias crime motiv ation: A study of hate crime and offender neutralization techniques used against the Amish. Journal of Contemporary Criminal Justice,15 (1), 78 96. Burrelles Luce (2007). 2007 Top newspapers, blogs, c onsumer magazines. Retrieved on April 3, 2011 from http://www.burrellesluce. com/top100/2007_Top_100List.pdf
125 Carabine, J. (1996). A straight playing field or queering t he pitch? Centring sexuality in social policy. Feminist Review 54 (Autumn), 31 64. Carney, G. (2010). Citizenship and structured depe nde ncy: The implications of policy design for senior political power. Aging & Society 30 (2), 229 251. Chakraborti, N., & Garland, J. (2009). Hate crime: Impact, causes, and responses Thousand Oaks, CA: Sage Publications. Chatfield, C. (1980). The analysis of time series: An introduction, 2 nd ed. New York: Chapman and Hall. Cogan, J.C. (2002). Hate crime as a category worthy of policy attention. American Behavioral Scientist 46 (1), 173 85. Colvin, R. (2010). Critical incidents, invisibl e population s, and public policy: A case of the LGBT community. Journal of Critical Incident Analysis Fall, 30 40. Cook, T.E. (2006). The news media as a political institutio n: Looking backward and looking forward. Political Communication, 23 (2 ), 159 171. Czech, B., Krausman, P.R., & Borkhataria, R. (1998). Social construction, political power, and the allocation of benefits to endangered species. Conservation Biology 12 (5), 1103 1112. Cramer, R J., McNiel, D. E., Holley, S. R., Sh umway, M & Boccellari A (2012). Mental health in violent crime victims: Does sexual orientation matter? Law and Human Behavior, 36 (2), 87 95. Cronin, S .W., McDevitt, J. Farrell, A. & Nolan, III J.J (2007). Bias crime reporting: Organizational responses to ambiguity, uncertainty, and infrequency in eight police departments. American Behavioral Scientist, 51 (2), 213 231. Dharmapala, D. & Garoupa, N. (2004). Penalty enhancement for hate crimes: An economic analysis. American Law and Economics Review, 6 (1), 185 207. DiAlto, S.J. (2005). From problem minority to model minority: The changing social constructions of Japanese Americans. In (Eds. ) A.L. Schneider & H.M. Ingram, Deserving & Entitled: Social Constructions and Public Policy (pp. 81 103). Albany, NY: State University of New York Press. deLeon, P. (2005). Social construction for public policy. Public Administration R eview, 65 (5), 635 637.
126 Doan, A.E. (2011). "What's wrong with being sexy?" Why political sci ence needs to get serious about sexuality. Symposium: The State of LGBT/Sexuality Studies in Political Science January, 31 34. doi: 10.1017/S1049096510001812. Donovan, M.C. (1993). Social constructions of peopl e with AIDS: Target populations and United States policy, 1981 1990. Policy Studies Review 12 (3/4), 3 29. Donovan, M.C. (1997). The problem with making AIDS comfortable. Journal of Homosexuality 32 (3), 115 144. Dunbar, E. (2006). Race, gender, and sexual orientati on in hate crime vi ctimization: Identity politics or identity risk? Violence & Victims, 21 (3), 323 337. Duncan, D.T., & Hatzenbuehler, M.L. (2014). Lesbian, gay, bisexual, and tran sgender hate crimes and suicidality among a population based sample of sexual minori ty adolescents in Boston. American Journal of Public Health, 104 (2), 272 278. Duncan, D.T., Hatzenbuehler, M.L., & Johnson, R.M. (2014). Neighborhood level LGBT hate crimes and current illicit drug use among sexual minority youth. Dru g and Alcohol Dependence 135 65 70. Epstein, S. (1987). Gay politics, ethnic identity: The li mits of social constructionism. Socialist Review 17 (4), 117 162. Erickson, R.S., Wright, G.C., & McIver, J.P. (1993). Statehouse democracy: Public opinion and policy in the American states Cambridge: Cambridge University Press. Eskridge, W.N. (2003). Dishonorable passions: Sod omy laws in America, 1861 2003. New York: Viking Penguin. Estrada, A.X., & Weiss, D. J. (1999). Attitudes of milita ry personn el toward homosexuals. Journal of Homosexuality 37 (4), 83 97. Frank, N. (2009). Unfriendly f ire: How the gay ban unde rmines the military and weakens America Franklin, K. (2002). Good inten tions: The enforcement of hate crime penalty enhancement statutes. American Behavioral Scientist, 46 154 72. Franzosi, R. (1987). The press as a source of socio historical data: Issues in the methodology of data collection from newspapers. Hist orical Methods, 20 (1), 5 16. Gamson, W.A., Croteau, D., Hoynes, W., & Sasson, T. (1992). Media images a nd the social construction of reality. Annual Review of Sociology 18 373 93.
127 Gan, L., Williams, R. C., & Wiseman T (2011). A simple model of op timal hate crime legislation. Economic Inquiry, 49 (3), 674 684. Gates, G.J., & Newport, F. (2013). LGBT percentage highest in D.C., lowest in North Dakota. State of the States, Gallup, Inc., Retrieved from www.gallup.com. Gentzkow, M., & Shapiro, J.M. (2 006). What drives media slant? Evidence from U.S. daily newspapers. University of Chicago and NBER Retrieved from www.psu.edu Gerber, A.S., Karlan, D., & Bergan, D. (2009). Does the media matter? A field experiment measuring the effect of new spapers o n voting behavior and political opinions. American Economic Journal 1 (2), 35 52. Gerstenfeld, P.B. (1992). Smile when you call me that!: T he problems with punishing hate motivated behavior. Behavioral Sciences and the Law, 10 259 85. Gibson, D.R. (2011). All the news that fits to print: Desk co mpetition for front page space at The New York Times Sociological Forum, 26 (2), 287 305. Gillis, B. (2013). Understanding hate crime statutes and b uilding towards a better system in Texas. American Journal of Criminal Law, 40 (2), 197 226. Glaesar, E.L. (2005). The political economy of hatred. The Quarterly Journal of Economics, 120 (1), 45 86. Grattet, R. & Jenness, V. (2001). The birth and maturation of hate crime policy in the U nited States. American Behavioral Scientist, 45 (4), 668 696. Grattet, R., & Jenness, V. (2005). The reconstitution of law in local settings: Agency discretion, ambiguity, and a surplus of law in the policing of hate crime. Law and Society Review, 39 (4), 893 942. Grattet, R. & Jenness, V. (2008).Transforming symbolic law into organizational action: Hate crime policy and law enforcement practice. Social Forces, 87 (1), 501 527. Grattet, R. Jenness, V. & Curry T.R (1998). The homogenization and differe ntiation of hate crime law in the United States, 1978 to 1995: Innovation and diffusion in the criminalization of bigotry. American Sociological Review 63 (2), 286 307. Green, J.C. (1999). Antigay: Varieties of opposition to gay rights. Retrieved from: http://www9.georgetown.edu/faculty/wilcoxc/green.PDF Green, D. P., Glaser, J. & Rich A (1998). From lynching to gay bashing: The elusive connection between economic conditions and hate crime. Journal of Personality and Social Psychology, 75 (1), 82 92.
128 Green, D.P., McFalls, L.H., & Smith, J.K. (2001). Hate crime: An emergent research agenda. Annual Review of Sociology 27 479 504. Haider Markel, D.P. (1998). The politics of social regulatory policy: State and federal hate crime policy and implemen tation effort. Political Research Quarterly 51 (1), 69 88. Haider Markel, D.P., Joslyn, M.R., & Kniss, C.J. (2000). Minority group interests and political representation: Gay elected officials in the policy process. The Journal of Politics 62 (2 ), 568 577. Haider Markel, D.P., & Kaufman, M.S. (2006). Public opinion and policy making in the culture wars: Is there a connection between opinion and state policy on gay and lesbian issues? In Public Opinion in State Politics, (Ed.) Jeffrey E. Cohen. S tanford, CA: Stanford University Press. Haider Markel, D.P., & Meier, K.J. (1996). The poli tics of gay and lesbian rights: Expanding the scope of the conflict. The Journal of Politics 58 (2), 332 349. Halperin, D. (1990). One hundred years of homose xuality: And other essays on greek love New York: Routledge. Hayes, A.F., & Krippendorff, K. (2007). Answering the call for a standard reliability measure for coding data. Communications Methods and Measures 1( 1), 77 89. Herek, G. M. (1989). Hate crimes against lesbians and gay men. American Psychologist 44 (6), 948 955. Herek, G.M. (1990). The context of anti gay violence: Notes on cultural and psychological heterosexism. Journal of Interpersonal Violence 5 (3), 316 333. Herek, G.M. (1 review of empirical research with the ATLG scale. In B. Green & G.M. Herek (Eds.), Contemporary Perspectives on Lesbian and Gay Issues in Psychology (pp.206 228). Newbu ry Park, CA: Sage Publishing Herek, G.M. (2002). Gender gaps in public opinion about lesbians and gay men. The Public Opinion Quarterly 66 (1), 40 66. Herek, G.M. (2006). Legal recognition of same sex relationships in the United States. American Psycholo gist 61 (6) 607 621. Herek, G.M. (2009). Hate crimes and stigma related experiences among sexual minority adults in the United States: Prevalence estimates from a national probability sample. Journal of Interpersonal Violence, 24 (1), 54 74.
129 Her ek, G.M., & Berrill, K.T. (1992). Hate crimes: Confronting violence against lesbians and gay men. Newbury Park, CA: Sage Publications. oward lesbians and gay men in the United Sta tes. Journal of Sex Research 32 (2), 95 105. Herek, G.M., & Capitanio, J.P. (1996). Some of my best friends: Intergroup contact, Personality and Social Psychology Bull etin, 22 (4): 412 24. Herek, G.M. & Capitanio, J.P. (1999). AIDS stigma and sexual prejudice. American Behavioral Scientist 42 (7), 1130 1147. Herek, G M., Cogan, J C., & Gillis. J.R. (2002). Victim experiences in hate crimes based on sexual orientat ion. Journal of Social Issues, 58 (2), 319 339. Herek, G.M., Gillis, J.R., & Cogan, J.C. (1999). Psychological sequelae of hate crime victimization among lesbian, gay, and bisexual adults. Journal of Consulting and Clinical Psychology, 67, 945 951. Herek, G.M., Gillis, J.R., Cogan, J.C., & Blunt, E.K. (1997). Hate crime victimization among lesbian, gay, and bisexual adults: Prevalence, psychological correlates, and methodological issues. Journal of Interpersonal Violence 12 195 215. Hogan, N.L. (1997). The social construction of target populations and the tran sformation of prison based AIDS policy. Journal of Homosexuality 32 (3), 77 114. Hong, D. (2009). Tenth annual review of gender and sexuality l aw: Constitutional law chapter: Hate crime r egulation and challenges. T he Georgetown Journal of Gender and the Law, 10, 279 300. Houston, D.J., & Richardson, Jr., L.E. (2004). Drinking and driving in America: A test of behavioral assumptions underlying public policy. Political Research Qu arterly, 57 (1), 53 64. Human Rights Campaign, (2014). Maps of state laws and polic ies. Retrieved on from http://www.hrc.org/resources/entry/maps of state laws policies Ingram, H., & Schneider, A. (1991). The choice of target populations. Adm inistration & Society 23 (3), 333 356. Ingram, H., & Schneider, A.L. (1995). Social construction (continued): Response. The American Political Science Review 89 (2), 441 446.
130 Ingram, H., Schneider, A.L., & deLeon, P. (2007). Social con structi on and policy design. In Theories of the Policy Process, 2 nd ed. (Ed.) Paul A. Sab atier, pp. 93 126. Boulder, CO: Westview Press. Jacobs, J.B. (1992). Rethinking the war against hate crimes: A New York City perspective. Criminal Justice Ethics, 11 (2), 55 61. Jacobs, J.B. (1993). Should hate be a crime? The Public Interest, Fall, 3 14. Jacobs, J.B., & Henry, J.S. (1996). The social construction of a hate crime epidemic. The Journal of Criminal Law and Criminology 86 (2), 366 391. Jacobs, J B. & Potter, K. (1998). Hate c rimes: Criminal law and identity politics New York: Oxford University Press. Jenness, V. (1995). Social movement growth, domain expans ion, and framing processes: The gay/lesbian movement and violence against gays and lesbians as a social problem. Social Problems, 42 (1), 145 170. Jenness, V. (1999). Managing differences and making legislation: Social movements and the racialization, sexualization, and gendering of federal hate crime law in the U.S. 1985 1998. Social Problems 46 (4), 548 71. Jenness, V., & Grattet, R. (1996). The criminalization of hate: A comparison of structural and polity influences on the passage of "bias crime" legislation in the United States. Sociological Perspectives, 39 (1), 129 54. Jenness, V., & Grattet, R. (2005). The law in between: The effects of organizational perviousness on the policing of hate crime. Social Problems 52 (3), 337 359. Johnson, S.D., & Byers, B.D. (2003). Attitudes toward hate crime laws. Journal of Criminal Justice, 31 227 235. Kite, M.E., & Whitley, B.E., Jr. (1996). Sex differences in attit udes toward homosexual persons, behaviors, and civil rights: A meta analysis. Personality and Social Psychology Bulletin 22 236 53. Knopp, L. (198 7). Social theory, social move ments and public policy: Recent accomp lishments of the gay and lesbian moveme nts in Minneapolis, Minnesota. International Journal of Urban and Regional Research, 11 (2), 243 261. Krippendorff, K. (2004). Content an alysis: An introduction to its methodology, 2 nd ed. Thousand Oaks, CA: Sage. Lax, J.R., & Phillips, J.H. (2009). Gay rights in the states: Public opinion and policy responsiveness. American Political Science Review 103 (3), 367 386.
131 Lieberman, R .C. (1995). Social construction (continued). The American Political Science Review 89 (2), 437 441. Levin, B. (1999). Hate crimes: Worse by definition. J ournal of Contemporary Criminal Justice 15 1 21. Levin, J., & McDevitt, J. (1993). Hate Crimes: The rising tide of bigotry and bloodshed New York: Plenum Press. Lewis, G.B. (2003). Black white differences in attitudes towar d homosexuality and gay rights. Public Opinion Quarterly 67 59 78. Lewis, G.B. (2011). The friends and family p lan: Contact with gays and support for gay rights. Policy Studies Journal 39 (2), 217 238. Lewis, G.B., & Rogers, M.A. (1999). Does the public support e qual employment rights for gays and lesbians? In Gay and Lesbians in the Dem ocratic Process: Public Policy, Public Opinion, and Political Representation (Eds.) Ellen D.B. Riggle & Barr y L. Tadlock, (pp. 146 69). New York: Columbia University Press. Linneman, T.J. (2004). Homophobia and hostility: Christian conservative reactions to the political and cultural progress of lesbians and gay men. Sexuality Re search & Social Policy, 1 (2), 56 76. Lynch, A.M. (2008). Hating the neighbors: The role of ha te crime in the perpetration of Black residential segregation. International Jo urnal of Conflict & Violence, 2 6 27. Lyons, C .J. (2006). Stigma or sympathy? Attributions of fault to hate crime victims and offenders. Social Psychology Quarterly, 69 (1), 39 59. Lyons, C.J. (2007). Individual perceptions and the social construct ion o f hate crimes: A factorial survey. The Social Science Journal 45 107 131. Maldonado, M. (1992/1993). Practical problems with enforcing hate crimes legislation in New York. Annual Survey of American Law 555 558. Maroney, T. A. (1998). Note: The st ruggle against hate crime: Movement at a crossroads. New York University Law Review 73 564 620. Martin, S. E. (1995). A cross burning is not just an arson: Police social construction of hate in Baltimore County. Criminology, 33 (3), 303 326. Martin, S. E. (1996). Investigating hate crimes: Case characteristics and law enforcement responses. Justice Quarterly, 73 (3), 455 480.
132 Mason, G. (2007). Hate crime as a moral category: Lessons from the Snowtown case. The Australian and New Zealand Journal of Criminol ogy, 40 (3), 249 271 Mason, G. (2014). The symbolic purpose of hate crime law: Ideal victims and emotion. Theoretical Criminology, 18 (1), 75 92. McDevitt, J., Balboni, J., Garcia, L., & Gu, J. (2001). Consequences for victims: A comparison of bias and n on bias motivated assaults. American Behavioral Scientist, 45, 697 713. McDevitt, J. Levin, J., & Bennett S. (2002). Hate crime offenders: An expanded typology. Journal of Social Issues, 58, 303 317. McDowall, D., McCleary, R., Meidinger, E.E., & Hay, Jr., R.A. (1980). Interrupted time series analysis Sage Series on Quantitative Applications in the Social Sciences, no. 07 021. Beverly Hills, CA: Sage Publications. McPhail, B., & Jenness, V. (2005/06). To charge or not to charge? -That is the questi on: The pursuit of strategic advantage in prosecutorial decision making surrounding hate crime. Journal of Hate Studies, 4 Meyer, D. (2014). Resisting hate crime discourse: Queer and intersectional challenges to neoliberal hate crime. Critical Criminol ogy, 22 113 125. Mogul, J.L., Ritchie, A. J., & Whitlock K (2011). Queer (In)j ustice : The criminalization of LGBT people in the United States Boston, MA: Beacon Press. Motion, J., & Weaver, C.K. (2005). The epistemic struggle for credibility: Rethinki ng media relations. Journal of Communication Management 9 (3), 246 255 really about morality?: Framing opposition to gay and lesbian rights. Policy Studies Journal 39 (2), 187 216. Nat ional Coalition of Anti Violence Programs. (2011). Anti le sbian, gay, bisexual, and trans gender violence in 2010. Retrieved from http://www.avp.org/resources/reports/term/summary Nicholson Crotty, S., & Meier, K.J. (2005). From perception to pub lic policy: Translating social constructions into policy designs. In (Eds.) A.L. Schneider & H.M. Ingram, Deserving and Entitled, (pp. 223 242). Albany, NY: SUNY Press. Nicholson Crotty, J., & Nicholson Crotty, S. (2004). Social construction an d policy implementation: Inmate health as a public health issue. Social Science Quart erly, 85 (2), 240 256.
133 Nolan, J. J., & Akiyama, Y. (1999). An analysis of factors that affect law enforcement participation in hate crime reporting. Journal of C ontemporary Criminal Justice, 15 (1), 111 127. N olan, III, J.J., Akiyama, Y. & Berhanu, S (2002). The Hate Crime Statistics Act of 1990: Developing a method for measuring the occurrence of hate violence. American Behavioral Scientist, 46 (1), 136 153. Olson, L.R., Cadge, W. & Harrison, J.T. (2006). Religion an d public opinion about same sex marriage. Social Science Quarterly 87 (2), 340 360. Owens, M.L., & Smith, A.R. (2012). "Deviants" and democracy: Punitive policy designs and the social ri ghts of felons as citizens. American Politics Research, 40 (3), 531 567. Pappas, C., Mendez, J., & Herrick, R. (2009). The negative effects of populism on gay and lesbian rights. Social Science Quarterly, 90 (1), 150 63. Patterson, C.J. (2006). Lesbia n and gay family issues in the conte xt of changing legal and social policy environments. In (Eds.) Kathleen J. Bieschke, Rup erto M. Perez & Kurt A. DeBord, Handbook of Counseling and Psychother apy with Lesbian, Gay, Bisexual and Transgender Clients, 2 nd Edition. (pp. 359 377). Washington, DC: American Psychological Association. Perry, B. (2001). In the name of hate: Understanding hate crimes. New York: Routledge. Perry, B. (2009). "There's just places ya' don't wanna go": The s egregat ing impact of hate crime against Native Americans. Contemporary Justice Review, 12 (4), 401 418. Petrosino, C. (1999). Connecting the past to the future: Hate crime in America. Journal of Contemporary Criminal Justice 15 (1), 22 47. Phillips, N. D. (2009). The prosecution of hate crimes: The limitations of the hate crime typology. Journal of Interpersonal Violence, 24 (5), 883 905. Pierce, J.J, Siddiki, S., Jones, M.D., Schumacher, K., Pattison, A., & Peterson, H. (2014). Social Construction and policy design: A review of past applications. Policy Studies Journal, 42 (1), 1 29. Pierson, P. (1993). Review: When effect becomes case: Policy feedback and political change. World Politics, 45 (4), 595 628. Pinderhughes, H. (1993). The anatomy of raciall y motivated violence in New York City: A case study of youth in southern Brooklyn. S ocial Problems, 40( 4 ) 478 492. Polumbaum, J. (2000). News for the culture: Why editors put strong men hitting baseballs on page one. Newspaper Research Journal 21 (2), 23 39.
134 American Journalism Review, Retrieved from http://www.ajr.org/Article.asp?id=3750 Rayburn, N.R., Earleywine, M., & Davison, G.C. (2003). Base ra tes of hate crime victimization among co llege students. Journal of Interpersonal Violence, 18 (10), 1209 1221. Sabatier, P.A., & Weible, C.M. (2007). The advocacy coali tion framework: Innovations and clarifications. In Theories of the Policy Process, 2 nd ed. (Ed.) Paul A. Sabatier, pp. 189 220. Boulder, CO: Westview Press. Sandholtz, N., Langton, L., & Planty M (2013). Hate crime victimization, 2003 2011. Bureau of Justice Statistics NCJ 241291. Sargent, L.T. (2009). Contemporary political ideologies Belmont, CA: Wadsworth Cengage. Schneider, A.L. (2012). Punishment policy in the American states from 1890 to 2008: Convergence, divergence, synchronous change, and feed forward effects. The Policy Studies Journal 40 (2), 193 210. Schneider, A., & Ingram, H. (199 3). Social construction of target populations: Implications for politics and policy. The American Political Science Review 87 (2), 334 347. Schneider, A., & Ingram, H.M. (Eds.) (2005). Deserving and ent itled: Social constructions and public policy Albany, NY: State University of New York. Schneider, A., & Sidney, M. (2009). What is ne xt for policy design and social constru ction theory? The Policy Studies Journal 37 (1), 103 119. Schweppe, J. (2012). Defining characteristics of politi cising victims: A legal perspective. Journal of Hate Studies, 10 173 198. Soule, S.A., & Earl, J. (2001). The enactment of state level hate crime law in the United States: Intrastate and interstate factors. Sociological Perspectives, 44 (3), 281 305. Smith, A.M., & Foley, C.L. (2010). State statutes governing hate crimes. Congressional Research Service Report for Congress 7 5700. Stacey, M. (2010). Distinctive characteristics of sexual orientation bias crimes. Journal of Interpersonal Vi olence, 26( 15 ), 3013 3032. Thorp, J. (1992). The social construction of homosexuality. Phoenix 46 (1), 54 61.
135 Torress, S. (1999). Hate crimes against African Americ ans: The extent of the problem. Journal of Contemporary Criminal Justice, 15 (1), 48 63. U .S. Department of Justice. (2013). Hate Crime Statistics, 2012 Uniform Crime Report. Wagner, A.K., Soumerai, S.B., Zhang, F., & Ross Begnan, D. (2002). Segmented regression analysis of interrupted time series studies in medication use research Journal of Clinical Pharmacy and Therapeutics 27 299 309. Wald, K.D., Button, J.W., & Rienzo, B.A. (1996). The politics of gay rights in American communities: Explaining antidiscrimination ordinances and policies. Journal of Political S cience 40 (4), 1152 1178. Walters, M. A. (2011). A General Theories of hate crime? Strain, doing difference and self control. Critical Criminology, 19, 313 330 Warbelow, S. (2010). Equality from state to state. A report by the Human Rights Campaign Ret rieved from www.hrc.org/statetostate. Weible, C.M., Siddiki, S.N., & Pierce, J.J. (2011). Foes to friends: Changing contexts and changing intergroup perceptions. Journal of Comparative Policy Analysis: Research and Practice, 13 (5), 499 525. Weisburd, S.B Recognizing gender based bias crimes. Stanford Law & Policy Review, Spring. White, M.D., & Marsh, E.E. (2006). Content analysis: A flexible methodology. Library Trends 55 (1), 22 45. Wh itlock, K. (2012). Reconsidering hate: Policy and politics at the intersection. A discussion paper for Political Research Associates. Somerville, MA. Yates, J., & Whitford, A.B. (2009). Race in the war on drugs: The social consequences of Presid ential rhetoric. Journal of Empirical Legal Issues, 6 (4), 874 898. Yoon, J., & Boydstun, A.E. (2014). Dominating the news: Government officials in front page news coverage of policy issues in the United States and Korea. Journal of Public Policy 34 (2), 207 235.
136 APPENDIX A CODING FRAMEWORK Negative Indicators ( 1): Discusses or states a religious view against homosexuality or homosexual sex behavior to sins/crimes l ike pedophilia Criticizes supporters of gay and lesbian rights or their tactics, for example States or that marriage, adoption, etc. are for man and woman or husband and wife only, or minimizes/delegitimizes same sex relationships Provides support for any law or policy aimed at removing or denying civil rights of homosexuals, or criticizes attempts to expand rights/protections for gays and lesbians Testimony or actions by legislators, experts, or citizens against the rights of homosexuals Requests action, policy, or lawsuit that will remove or prevent rights for homosexuals Describes acts of discrimination, prejudice, or violence against gays and lesbians Positive Indicators (+1): Calls for support and love of gays or condemns violence and discrimination against gays Dissent again Assertion of equal rights and protections for homosexuals or supports expanding civil rights Supports pro gay groups and organizations such as PFLAG or pro gay events like Pride Criticizes any law or policy that denie s or removes civil rights from homosexuals Criticizes anti homosexual remarks or stances or depicts anti gay groups as prejudiced Testimony or actions by legislators, experts, or citizens in support of rights for homosexuals Promotes visibility and/or acce ptance of same sex relationships Neutral Indicators (0): Places both homosexual and heterosexual behavior in negative terms Discusses views not specific to gays and le sbians or their rights Discusses issues beyond sexuality, such as First Amendment rights Describes procedural issues or implementation issues of a vote or policy
137 Speculation about what the effects of a policy might be Any statement not directly related to pro or anti gay issues or sentiments