Skip to main content
SearchLoginLogin or Signup

Revisiting the Harm of Hate: A Quasi-Experimental Approach using the National Crime Victimization Survey

Holder, E. (2024). Revisiting the harm of hate: A quasi-experimental approach using the National Crime Victimization Survey. Journal of Interpersonal violence,

Published onFeb 07, 2024
Revisiting the Harm of Hate: A Quasi-Experimental Approach using the National Crime Victimization Survey


Early legal challenges to the 1990 Hate Crime Statistics Act were originally excused on the argument that hate crimes “hurt more,” but there remain some empirical gaps on this topic. While many works have concluded that biased offenders cause greater harms to their victims relative to unbiased perpetrators, this effect tends to be sensitive to individual and situational factors like victim and offender characteristics, bias motivation, weapon use, or crime location. This type of confounding has the potential to introduce selection bias in the estimation of victimization harms among biased criminal incidents. With data from the National Crime Victimization Survey (2010-2020), I use propensity scores and inverse-probability weighting to show that, on average, victims of bias motivated offenses are more likely to report later physical and emotional harms despite not suffering greater initial injury in incidence. Findings also demonstrate that the harm of hate varies across different bias motivations, with such crimes directed towards those on the basis of disability, gender, and sexual orientation causing greater short- and long-term individual trauma and damage.

Keywords: Hate crime, bias crime, victimization, NCVS, propensity scores, quasi-experimental

Note: This represents a finalized accepted manuscript to the Journal of Interpersonal Violence.


Do hate crimes hurt more? Over the past two decades, research has been conducted on whether bias motivated victimizations are more injurious than their non-biased counterparts (Benier, 2017; Fetzer & Pezzella, 2019; Ignaski & Lagou, 2015; Lantz & Kim, 2019; Malcom & Lantz, 2021; Mellgren et al., 2021; Pezzella & Fetzer, 2017; Powers & Socia, 2019). Whereas some have shown that victims of bias motivated offenses suffer greater physical and psychological or emotional harm compared to victims of more “general” crimes, other studies have demonstrated that the nature of hate crime incidents are no more violent than “general” criminal incidents (Messner et al., 2004). Yet, these studies are limited in their ability to determine the causal ordering of bias motivation on incidence and later experienced harm more clearly. To further clarify this relationship and better determine the effect of bias motivation and level of victim harm, I use data from the National Crime Victimization Survey and propensity score models to control for selection bias and confounding. In providing greater methodological rigor on this matter, I seek to provide a finer degree of empirical clarity for the sake of not only legal implications for hate crime policies but also for practicality’s sake in facilitating evidence-based responses to victims of hate crime.


While a handful of states began to pass policies targeting identity-motivated criminal offending in the 1980s, hate crime laws began to more fully capture the public’s attention with the 1990 Hate Crime Statistics Act. With this bill, the Federal Bureau of Investigation (FBI) defines any hate crime as a criminal offense motivated by bias or prejudice on select social characteristics including race, religion, ethnicity, sexual orientation, gender identity, and mental/physical ability (FBI, 2019). Though there remains some scholarly debate in defining “hate crime,” the federal definition, which is often shared across states as well, remains a dominant source utilized in many scholarly assessments of bias crime in the United States, and is likewise used other databases like the National Crime Victimization Survey (NCVS). Indeed, since that time, scholarly attention has been directed to understanding the correlates of hate crime offending and victimization, focusing on differences between individual and communities alike (Gladfelter et al., 2017; Green et al., 1998; Holder et al., 2022; Lyons, 2007; McDevitt et al., 2002; Messner et al., 2004; Piatkowska et al., 2019; Stacey et al., 2011). However, much research has been conducted on the social context on the diffusion of the hate crime legislation and its public support (Jacobs & Potter, 1998; Jenness & Broad, 1997; Jenness & Grattet, 2001).

Indeed, opponents of hate crime legislation originally argued that hate crime penalty enhancements violated free speech and punished intent beyond the criminal code, akin to “thought crimes” (Gellman, 1992; Weinstein, 1992). Whereas offenders are punished for their intent to cause harm associated with the crime, adding an additional layer of intent (or motive) is unconstitutional according to those who oppose such policy, serving to regulate freedom of speech and expression protected by the First Amendment. Yet, in considering such a constitutional challenge to state-level hate crime penalty enhancements, the Supreme Court in Wisconsin v. Mitchell (1993) sanctioned the usage of such enhancements as constitutionally permissible on the grounds that victims of identity-based crime typically suffer greater harm than victims of non-bias motivated crimes. Still, policy decisions surrounding identity-motivated criminality have been contentious, with a general lack uniformity among hate crime policies and punishments across the United States.

Yet, a great deal of empirical work has been completed to both understand the diffusion of various hate crime policies across the country as well as individual support for such policies. Research into the former indicates the key mechanism of “framing” in garnering public support for hate crime penalty enhancements and recognition by law, much of through constructing an image of enhanced degree of victimization among those targeted by bias motivated offenses (Grattet et al., 1998; Jacobs & Potters, 1998; Jenness & Broad, 1997; Jenness & Grattet, 2001).

Indeed, Jacobs and Potters (1998) originally noted, “Despite the lack of empirical evidence, the belief that bias crimes cause more physical harm than other crimes in the same offense category has practically become dogma” (p. 81). Much of the latter research has been focused on understanding the dynamics of personal support for hate crime policies, namely additional punishment through penalty enhancements (Cabeldue et al., 2018; Cramer et al., 2013, 2014, 2017; Malcom et al., 2023; Saucier et al., 2006; Steen & Cohen, 2004) While individual support is linked to demographics and political affiliation (with some scholars embedding the criminalization of biased criminality as “identity politics” – see Jacobs & Potters, 1998 and Perry, 2001), there is some evidence that individuals are more supportive of increased punishment if they perceive that hate crime victims suffered an enhanced level of harm, and preparators of hate crime are often thought of more negatively, thus leading more punitive outcomes (Cabeldue et al., 2018; Cramer et al., 2010, 2014, 2017). Criminal justice practitioners, including police and prosecutors, are guided by this perception too (Ignaski, 2001; Nolan & Akiyama, 1999).

Thus, understanding the true nature of the hate crime-harm relationship is paramount to policy changes and criminal justice interventions. Given this, extant scholarship has focused on whether hate crimes are not only more harmful in both their immediate physical harm as well long-term emotional costs but also whether biased victimizations share correlates with other harmful criminal behavior, such as homicide or assault (Benier, 2017; Klein & Allison, 2018). Overall, data indicate that victims of hate crime experience a range of negative emotions and physical harm following victimization including depression, anger, sadness, and even feelings of suicide (Benier, 2017; Cramer et al., 2018; Herek et al., 1997; Ignaski, 2001; Levin, 1999; Williams & Tregidga, 2014). Such reported feelings overlap and are likewise symptoms of post-traumatic stress disorder (PTSD), a commonly experienced mental disorder following criminal victimization but one that has been especially documented among victims of identity-based harm (Szymanski & Balsam, 2011). Additionally, part of this direct and vicarious trauma is to do with the fact that being targeted is associated with individual and community traits that are uncontrolled or immutable (Benier, 2017; Ignaski, 2001; Perry & Alvi, 2012). Victimization data also indicate that hate crimes are disproportionately violent in nature (Kena & Thompson, 2021), though data from official police sources (e.g., the FBI) indicate otherwise (FBI, 2019).

In light of the mix of physical and psychological traumas associated with identity-based victimization, some scholars have supplied theoretical frameworks in which to understand this enhanced victimization, usually in support for more severe penalties for bias motivated criminality (Ignaski, 2001; Levin, 1999; Perry, 2001). Specific to hate crime, many have argued that hate crime offenders do not simply seek to cause harm to the individual victim but strike against the community in which they are perceived to belong. Ignaski (2001), for instance, notes on the “waves of harm” generated by hate crime that not only include the individual victim but their neighborhood group, their group beyond the neighborhood, other targeted communities, and societal norms and values, with hate crimes being perceived as “message crimes.” Perry and Alvi (2012), too, note on the “in terrorem” effects of hate crime and some empirical works comparing the two (terrorism and hate crime) have found their likelihood of occurrence is often related (Deloughery et al., 2012; Mills et al., 2017).

To be sure, others have also noted on the widespread and vicarious harm that is associated with bias crime and the social control mechanism in which it seeks to impose upon people with marginalized identities. Perry’s (2001, 2002) “doing difference” framework stresses this perspective against hate crimes of many different bias motivations (e.g., racial, religious, gender, sexual orientation) in which violence motivated by bias seeks to reinforce structural arrangements and normative identities. Hate crimes should hurt more because they are intended to punish perceived deviant identities that violate hegemonic norms, and victims experience a greater degree of harm due to the fact that the base motivation of their victimization is, in whole or partially, rooted in uncontrollable components of one’s identity. Herek’s (1990) framing of “heterosexism” and anti-LGTQIA+ violence is similar, that such crimes are meant to be oppressive and uphold cultural ideologies of sex and sexuality. With this, studies have clearly demonstrated the “in terrorem” effects on not only victims but their communities as well (Bell & Perry, 2015; Benier, 2017, Ignaski, 2001; Noelle, 2002; Perry & Alvi, 2012). The social and cultural context of hate crime, too, evidence the fact that many hate crimes are intended to send clear messages as to “deviance” of the identified community. Racially motivated hate crimes, for instance, often occur in White communities featuring in-migrations of racial minorities, and victimization rates of hate crimes against Arab and Muslim communities spiked following the terrorist incidents of 9/11 (Disha et al., 2011; Gladfelter et al., 2019; Grattet, 2009; Green et al., 1998; Lyons, 2007). Hate crimes and other violence motivated by one’s perceived sexuality or gender conformity are likewise linked to particular social and cultural contexts, such as increased gay visibility or growth of anti-trans rhetoric and legislation (Brightman et al., 2023).

Others, though, have considered hate crime victimization in an alternative lens that accounts for the fact that some hate crime incidents might not be inherently attached to societal power dynamics or modes of oppression. Much of this perspective is founded upon research that shows that a meaningful portion of hate crime offenders are not necessarily motivated by personal animus but are perhaps discriminatory actors, nonetheless (Martin, 1995; Messner et al., 2004). Analyses of police hate crime investigation files, for example, highlight that many perpetrators of hate crime are not necessarily guided by any personal hostility towards a particular victim or group, but that victims are selected on the basis of established stereotypes as an “easy target” for victimization (Bell, 2002; Levin & McDevitt, 1993; Martin, 1995; McDevitt et al., 2002; Messner et al., 2004). As Lawrence (1994, 1999) has argued, the discriminatory model distinguishes a “bias crime” by the pure characteristics of the victim; it is irrelevant as to why the victim was selected but that the offender did so anyway; such a perspective would be easier to prove in a court of law. McDevitt et al.’s (2002) “thrill seeker” characterization of hate crime offenders fits this mold, with offenders often being young, in groups, intoxicated, and choosing an “easy” target for victimization. While such a perspective of victim vulnerability has received some theoretical advancement in hate studies, particularly with the work the Chakraborti and Garland (2012), the label of “vulnerability” for victims of hate crime has received pushback for its portrayal of victimhood among marginalized individuals and communities. Yet, unlike the animus model of offending and victimization, this discriminatory perspective does not seem to provide any proposition in whether hate crime victims should or do suffer at greater severity.

Regardless of framework, studies directed to understanding whether hate crimes are more violent tend to show that such victims suffer greater psychological and physical distress relative to other types of crime, but these findings are mixed across studies (Fetzer & Pezzella, 2019; Ignaski & Lagou, 2015; Lantz & Kim, 2019; McDevitt et al., 2007; Mellgren et al., 2021; Meyer, 2010; Pezzella & Fetzer, 2017). For example, many studies utilizing victimization data, such as the NCVS or other similar datasets, tend to show that victims not only experience higher levels of physical and psychological harm, but these events are also linked to more immediate serious injuries (Benier, 2017; Fetzer & Pezzella, 2019; Ignaski & Lagou, 2015; Tessler et al., 2021). When using police-reported data, most commonly the National Incident-Based Reporting Program (NIBRS) database, studies are less clear in whether hate crimes “hurt more” (Lantz & Kim, 2019; Messner et al., 2004; Malcom & Lantz, 2021; Pezzella & Fetzer, 2017; Powers & Socia, 2019).

For instance, while Fetzer and Pezzella (2019) demonstrated the increased risk of personal harm from bias-motivated offenses using the NCVS, finding that biased victimization was associated with an increased likelihood of physical and psychological harm, Pezzella and Fetzer (2017) found this relationship was sensitive to bias motivation when using NIBRS. They found hate crimes against White victims, for example, were more injurious than those against Black victims. Also using NIBRS, Powers and Socia (2021) likewise found that Black-on-White hate crimes were more violent than the opposite dyad, White-on-Black hate crimes. Regardless, the authors found, on average, bias-motivated offenses featured both higher odds of minor and major physical injury than both non-biased interracial and intraracial crimes. In comparing the situational correlates of hate crime with more “general” crimes with official data, Messner et al. (2004) found a greater level of empirical support for their “versatility” model of hate crime offending using NIBRS, with hate crimes sharing a meaningful degree of situational correlates with other types of non-biased criminal incidents. Studies focusing on victim-provided data (whether through quantitative or qualitative designs) nonetheless show a significant level of experienced victimization harms associated with bias motivated offenses but often lack comparison groups in which to determine if such harms are more severe than non-biased crimes.

Also complicating the relationship between bias and harm is the fact that such injuries are often moderated by other situational factors and may operate as confounders to both the outcome of interest (personal injury) as well the bias motivation itself. As previously described, studies indicate that the motivation of the offender is often related with certain bias motivations, often on the basis of race or sexuality, being more likely to end in greater injury (Ignaski & Lagou, 2015; Mellgren et al., 2021; Meyer, 2010; Pezzella & Fetzer, 2017; Powers & Socia, 2021; Tessler et al., 2021). Using data from a sample of Swedish individuals, Mellgren et al. (2021) found important distinctions in the emotional reaction to biased victimization among perceived identities. For example, victims of animus towards sexual orientation or gender identity were more likely to report worrisome feelings post-victimization but not those whose harm was due to their national background or religion; all groups, though, reported equal feelings of anger and avoidance following their victimization. Pezzella and Fetzer (2017) reported higher likelihood of serious injury for anti-White hate crimes and those crimes categorized as motivated by anti-lesbian bias. However, with official data from NIBRS, Stacey (2011) found that sexual orientation motivated hate crimes were no more injurious than racially motivated incidents after controlling for other salient situational correlates. Using victimization data from Wales, Williams and Tregidga (2014) found that victims of anti-gender, disability, or transgender hate crimes suffered higher rates of psychological distress net of individual, situational, and community controls.

Not only this, but the harm experienced by hate crime victims may also be confounded by factors like co-offending or weapon use. With NIBRS, Lantz and Kim (2019) found that, while victims of bias crime were more likely to suffer greater physical harm, the presence of co-offending significantly mediated this relationship. Also, some studies surprisingly show that biased victimizations are less likely to involve any type of weapon use (Malcom & Lantz, 2021; Pezzella & Fetzer, 2017). Other studies show that hate crimes are also likely to involve offenders that are unknown to the victim, multiple offenders, and the presence of drugs and alcohol among the offender or offenders, which may operate to obscure the relationship between offender bias motivation and victim harm (Fetzer & Pezzella, 2019; Lantz & Kim, 2019; Malcom & Lantz, 2021; Messner et al., 2004; Pezzella & Fetzer, 2017).

Current Study

While the current empirical literature on the violent nature of hate crime offending generally suggests that hate crimes are typically more injurious, such research also points to victim, offender, and situational factors may distort the empirical reality of experienced harm from biased victimization. To address this, I return to the essential question within hate studies of whether hate crimes “hurt more” via a quasi-experimental design. In this domain, using quasi-experimental methodologies are suitable given that bias motivation serves as an appropriate “treatment condition” in which to estimate applicable treatment effects, i.e., whether victims of crimes motivated by bias endure greater physical and psychological harm than those not receiving the treatment. Such an approach also opens the door to more effectively establish causality by ruling out alternate explanations in the form of selection bias and confounding factors. As the previous section has discussed, mathematically estimating personal harm from hate crime victimization is complicated by the presence of individual and situational confounders not only to experienced trauma but also the bias motivation itself. As the reviewed literature demonstrates, such confounding includes factors like bias motivation, location, weapon presence, and victim and offender characteristics. Through random assignment via an experiment, however, one can more specifically entertain the causal pathways between hate and personal harm. Of course, an experiment for this topic is not possible, but quasi-experimental designs that simulate randomization offer a novel approach to this issue and is employed in this study. Accordingly, I seek to provide greater empirical clarity on hate crime victimization through this separate methodological approach.

Data and Methods

Data from this study is drawn from 11 years of the NCVS, 2010 though 2020. The most recent legislation, the 2009 Sheppard-Byrd Act, expanded the definition and serves as the latest national policy to do so, meaning that using data following this expansion is the most encompassing of all potential bias motivations. Administered semi-annually to 40,000 households in the United States including Puerto Rico, the NCVS contains information from all members in a household over the age of 12 on whether they were crime victims in the past six months, including roughly 100,000 individual surveys. Survey estimates are taken from a stratified, multistage cluster design. The sampling design includes creating primary sampling units (PSUs) of counties, groups of counties, or large metropolitan areas. From each stratum, groupings of households are selected through Census address files and interviewed by BJS agents. Data is then gathered on multiple domains including the nature of the victimization incident, perceived offender(s) characteristics, experienced harm (both emotion and physical), and whether they reported the incident to the police.

Since 2003, respondents have been able to record whether they believed their victimization was motivated by their perceived social identities including race, religion, ethnicity, sexuality, gender, and disability. Technically, the NCVS definition of a “hate crime” is shared with the official FBI definition, but the NCVS only considers an incident as bias motivated if three conditions are met: (1) Whether the offender used hateful language, (2) whether hate symbols were present during the incident, or (3) whether a law enforcement officer confirmed that the incident was a hate crime; Estimates from the NCVS show that most hate incidents are evidenced by hateful language (Kena & Thompson, 2021). It is from these incidents that I sampled whether an incident was bias motivated, which is the strictest definition of “hate crime” obtainable via NCVS reporting standards. Others studying this issue have chosen official data as well, namely NIBRS. It is important to note, however, that comparisons between police-reported data and victimization reports or alternate databases tend to find some level of convergence between the two (Holder, 2022; Ruback et al., 2018).

Dependent variables

There are three outcome variables in this study: Incident physical harm, emotional harm, and later physical harm. The first is a binary variable indicating whether the victim was seriously injured in their encounter and encompasses being stabbed, shot, having broken bones, internal bleeding, or being knocked unconscious relative to minor injuries of bruises, cuts, scratches, swelling, or chipped teeth. While this measure does fail to potentially capture a broader range of the incident physical injury, such a dichotomous approach in differentiating between minor and serious physical injury has been utilized in prior studies of this nature (Fetzer & Pezzella, 2019; Lantz & Kim, 2019; Pezzella & Fetzer, 2017). The latter two measures, however, reflect reported harm within a month of the crime. Emotional harm taps the negative emotions experienced following victimization and is a 0-7 scale (Cronbach’s alpha=.816) with higher scores being those who experienced the greatest emotional harm. This includes feeling worried, angry, sad, vulnerable, violated, mistrust, and unsafe. Finally, later physical harm is also a 0-7 scale (Cronbach’s alpha=.854) with similar coding that include experienced bodily harm such as headaches, loss of sleep, eating issues, stomach aches, fatigue, bleeding, and muscle pains.1

Confounding variables

This study includes confounding variables that have the potential to create selection bias in estimating a treatment effect that includes individual and situational factors. In propensity score estimation (described below), the paramount objective is identifying covariates that confound the treatment effect at hand, i.e., confounders. Several types of confounders exist when estimating the causal nature of interventions or treatments, including potential confounders or “true” confounders (Leite, 2017). Controlling for potential confounders means only including factors that potentially covary with the outcome, but the presence of selection bias in treatment effect estimations relates to the issue of selection into particular treatment conditions, and in this study, pertains to whether an incident was motivated by offender bias. On the other hand, true confounders are those covariates that affect both treatment assignment and the outcome among observational data. While arguments exist for the inclusion of either of these confounders, I include both potential and true confounders in this analysis which includes not only factors that covary with treatment (that an incident was hate crime) but the three primary outcomes as well.

First, location assesses whether an incident occurred near one’s own home or outside the home; the reference category are those incidents occurring at or near the respondent’s home (0=home, 1=outside of home). Urban indicates whether the incident occurred in an urban location rather than a suburban or rural location (0=rural/suburban, 1=urban), with official data indicating hate crimes are more frequent in such locations (Holder et al., 2022; Wilson & Ruback, 2003). From a routine activities perspective, incidents in rural areas have the potential to be more violent in nature because of the lack of proximity to others that might intervene or report the crime, and there is some research to indicate meaningful differences in hate prevalence between these ecological settings (Holder et al., 2022; Osborne & Swartz, 2021; Spano & Nagy, 2005). Weapon indicates whether a weapon was used and includes no weapon, handgun, other gun, knife, sharp object, or blunt object (0=no weapon, 1=weapon). While research tends to show that hate crimes are less likely to involve weapon use, it is nonetheless rational to assume that the presence of weapons is correlated with the outcomes under examination (i.e., a potential confounder), which too are important to consider as covariates for adequate treatment effect calculations.

Attack is a binary measure that indicates whether the offender attacked the victim. According to official data, many acts of hate actually occur in the form of threats and vandalism, though victimization data suggests otherwise (Cheng et al., 2013; Kena & Thompson, 2021). Nonetheless, it is important to control for whether incidents were actually carried out or attempts. Self-protection indicates if a victim took self-protective actions in the incident. Lastly, offender and victim characteristic are included to measure whether the incident had multiple offenders, the offender’s race and sex, and whether the offender was using drugs or alcohol. Offender sex was coded with females as the reference category (0=female, 1=male), offender race is coded with non-White offenders as the reference (0=non-White, 1=White), and drugs/alcohol was coded in whether the victim identified that the offender was using drugs, alcohol, or a combination of both (0=no drugs/alcohol, 1=drugs/alcohol present). The age, sex, and race of the victim are also included as research indicates their relevance to not only the presence of any bias motivation in a crime but also the level of harm experienced by the victim. Victim age represents a continuous measure, victim sex is used to measure whether the victim was female (0=male, 1=female), and victim race indicates if the victim is White (0=non-White, 1=White). Prior research indicates not only that certain victim characteristics are associated with the presence of bias motivation (true confounding) but also experienced harm, with White persons and women self-reporting greater levels of physical harm (Pezzella & Fetzer, 2017; Powers & Socia, 2021). Finally, series victimization includes whether the victim has experienced a “series” of victimizations per NCVS guidelines. If the victim reported experiencing more than five incidents in the past six months (1=yes, 0=no), they are considered as a “series” event, thereby raising the likelihood of future victimization as well.

Analytical strategy

The current analysis has two main objectives: First, I examined the effects of hate crime victimization on incident and later physical and psychological (emotional) injuries. Second, I examined whether these effects varied across perceived offender bias motivations. The analytic strategy combines propensity score estimation along with inverse-probability weighting in which to simulate randomization into the treatment group of importance (Leite, 2017; Rosenbaum & Rubin, 1985). I used logistic regression with the above specified covariates to calculate propensity scores for each observation (or victim) to determine the conditional probability of assignment to a particular treatment given a vector of observed covariates, also defined as e(X)=P(Z=1X)e(X) = P(Z = 1|X). In other words, with estimated propensity scores e(Xe(X), each observation has a unique calculated value that one’s victimization was bias motivated, ranging from 0 to 1.

In this study, I chose to estimate the average treatment effect on the treated (ATT), which examines the mean differences in groups conditional that such observations were assigned to the treatment condition following propensity score estimations. Because the substantive interest surrounds hate crime victimization and harm, the ATT allows me to estimate any potential harm “gained” from an individual being assigned to a biased victimization when the observed incident was without offender bias. To calculate the ATT, I use inverse-probability weighting rather than matching or stratification to preserve the overall sample which can be lost when using other estimation methods. King and Nieslen (2019) refer to issues of matching as the propensity score matching paradox, showing that achieving adequate covariate balance in matching requires drastic data pruning thereby decreasing the power of the method in making causal inferences. The use of matching also introduces the potential of decreasing effective sample sizes, and weighting procedures, like used here, preserve original sample sizes pending sufficient common support following weighting. Diagnostic for the propensity scores to determine whether the estimated propensity scores created comparable treatment and control groups are available in Figure 1 and Table 1, which show sufficient common support and covariate balance respectively.

Figure 1: Propensity Score Common Support by Biased and Non-biased Victimizations

Table 1: Covariate Balance Across Hate Crime Victimizations before and after Weighting on the Propensity Score (n=11,119)*


Original Sample

Weighted Sample

Mean Hate Crimes


Mean non-Hate Crimes


Standardized Difference

Mean Hate Crimes


Mean non-Hate Crimes


Standardized Difference

Open location




























Multiple offenders







Male offender







White offender




























Female victim







White victim







Victim age







Series incident







*Total sample size (n=11,119) is only for emotional harm and later physical harm. For serious injury, the total sample size (n=1,756) is restricted due to cross-tabulated missingness between variables. Further analysis from Table 2 showed no meaningful differences in treatment effect outcomes when restricting models to those 1,756 cases, meaning comparisons across models are valid in terms of sample selection.


Summary statistics for the original and weighted sample are presented in Table 1 as well as standardized mean differences to determine covariate balance. By examining the standardized mean differences before and after weighting, further analysis supports that the covariates are sufficiently balanced in the sample as indicated by not only the reduction of such differences after weighting but also that standardized mean differences fall well below .25 and .01 for some covariates. Along with a generally sufficient level of common support post-weighting, the estimation of the ATT from the weighted sample should be mostly unbiased. By also analyzing the information from Table 1, one can too see the potential of confounding variables in the estimation of hate crime, with data (from the original sample) indicating that hate crimes more often involve incidents with multiple offenders, strangers, or the with the offender being under the influence of drugs and/or alcohol. Not only this, but this sample also suggests hate crime is more likely to include weapon usage by the offender, contrary to prior research on the topic. This may be due to the case that I am using a wider window (11 years of data) that might be inclusive of a greater variety of incidents and increase the probability that more incidents involve weapons.

Table 2: Average Treatment Effect on the Treated (ATT) by Outcome




Serious injury (n=1,756)

.011 (.040)


Emotional harm (n=11,119)

1.112 (.128)


Later physical harm (n=11,119)

.598 (.104)



Notes: Beta coefficients presented with survey-adjusted linearized standard errors in parentheses.

Next, treatment effects (ATT) are presented in Table 2. Generally, these findings demonstrate that hate crimes “hurt more,” but with some nuance in incident versus later harm. First, the results indicate that those in the sample who perceived their victimization to have been motivated by bias were, on average, no more likely to be seriously injured than non-biased victimization. While the relationship is nonetheless positive, it lacks overall statistical significance even after weighting on the estimated propensity scores. However, hate crime victims did experience greater average levels of emotional harm following their victimization than victims of non-biased criminal encounters. On average, victims of bias crime reported a little over one-point higher indicator of emotional injury (b=1.112; p<.001) than victims of non-biased crimes. Finally, Table 2 results indicate that had victims of non-biased crimes perceived their victimization to have been bias-motivated, they would have experienced greater later physical harm, by roughly half a point higher (b=.598; p<.001).

To further understand this matter, I present plotted treatment group means across the outcomes of interest in Figure 2. While this highlights the lack of substantive difference in serious injury among victims of biased and non-bias crimes, it nonetheless elucidates the prior estimated ATT. This data shows that, on average, victims are experiencing more than one-and-half times (2.792÷\div1.680) these summated emotional harms as a relative “gain” (because the ATT was calculated rather than an ATE) by experiencing a hate crime. In this sample, estimates likewise indicate that had victims of non-biased crimes fallen prey to perceived biased offenders, they would have also experienced at least one more later health-related issue including headaches, difficulty sleeping, eating issues, stomach problems, fatigue, high blood pressure, and/or muscle-related issues like soreness. Otherwise, Figure 2 data show, on average, non-biased victimization results in less than one of those issues.

Figure 2: Predicted Means of Victimization Harm by Injury Type and Bias

The fact that hate crimes, in the sample, are not more violent in their incidence raises questions on the “harm” of hate, but also necessitates further analysis to determine whether these findings may be sensitive to other features of the victimization. Past studies indicate that the offender’s bias motivation differentially impacts the experienced biased victimization (Mellgren et al., 2021; Meyer, 2010; Pezzella & Fetzer, 2017; Stacey, 2011; Williams & Tregidga, 2014). In Table 3, I provide the distribution of hate crime incidents among bias motivation types and present ATTs for these primary bias motivations recognized by the NCVS: Race, religion, ethnicity, disability, gender, and sexual orientation. Regarding the effect of bias on serious injury, only those crime motivated by anti-disability bias had significantly greater incident harm, with a relatively small effect size (b=.325; p<.01). And because there is no difference in scaling among dependent and predictor variables, estimates (in both their magnitude and direction) are comparable across models. In considering the level of emotional harm, we see statistical significance across all bias motivations. Victims of gender-motivated crimes, however, tended to suffer the greatest emotional distress (b=1.805; p<.001), but those based on disability (b=1.489; p<.001) or sexuality (b=1.602; p<.001) bias were similar. Likewise, crimes motivated by anti-disability (b=1.333; p<.001) or anti-gender (b=1.314; p<.001) sentiments caused greater later physical harm as well. Overall, assessing such differences in the harms of hate crime highlight that while all hate crime had particular long-term consequences, that victims of bias crime on the basis of disability, gender, or sexual orientation are especially harmful.

Table 3: Average Treatment Effect on the Treated (ATT) by Outcome and Bias Motivation

Bias motivation

(Mean; SD)

Serious injury

Emotional harm

Later physical harm


(.722; .011)








(.762; .023)








(.562; .048)








(.793; .021)








(.757; .016)







(.725; .024)






***p<.001; **p<.01; *p<.05

Notes: Beta coefficients presented with survey-adjusted linearized standard errors in parentheses. Sample sizes for serious injuries/emotional and later physical harm for each motivation include 1,711/10,832 (Race); 1,657/10,524 (Religion); 1,647/10,463 (Ethnicity); 1,659/ 10,513 (Disability); 1,675/10,611 (Gender); 1,660/10,536 (Sexuality).

Discussion and Conclusion

The legal lineage of hate crime polices are rife with tensions regarding the true nature of the injuries suffered from crimes motivated by offender bias. While scholars have astutely noted on the “in terrorem” effects of hate crime victimization and far-reaching harms suffered by individuals and communities, whether this level of injury is beyond what is experienced from more “general” crimes is empirically complex. In this paper, I sought to further consider whether victims of hate crime experience greater harm than those who experienced other non-biased events. Analyses on this issue have yet to consider the potential of selection bias, with greater empirical rigor needed to clarify the causal relationship between bias motivation and victim injury. This issue is important to bear in mind not only for any legal ramifications, that hate crimes policies are legally permissible based on the idea that biased criminal offenses are more harmful, but for tangible solutions by criminal justice agents and other practitioners in addressing hate crimes. By using a quasi-experimental design, propensity scores and inverse-probability weights, to diminish the influence of selection bias, I find that while the harm of biased offending is limited in the short-run (i.e., incidental harm), that victims experience significantly greater long-term emotional and physical harm. From this, I would nonetheless argue that hate crimes are more injurious than general crimes.

More specifically, I estimated the average treatment effect on the treated (ATT) in determining the extent of harm resulting from biased offending. Unlike an average treatment effect, which is used to calculate outcome differences between treatment and control groups, the ATT is used to estimate the level of harm that victims of perceived non-biased crimes would have experienced if their offender has been motivated by hate. In doing so, I found that hate crimes are no more serious, with respect to incidental serious injury, than non-biased motivated crimes. However, I likewise found that if an individual had, instead, experienced a hate crime, they would have suffered from significantly greater emotional and later physical harm, such as headaches or high blood pressure. Thus, while these results tend to not support the argument that hate crimes are perhaps more violent in nature in terms of immediate physical harm, hate crimes are nonetheless associated with attendant later trauma. In this sense, one cannot reject the underlying argument to hate crime legislation that such identity-motivated crimes and victimizations are more harmful. The findings from this analysis do, in fact, point to the fact that persons experiencing identity-based violence do suffer at a greater degree relative to their peers whose victimization was not motivated by bias.

My findings were also sensitive to offender bias motivations, with some past research indicating that identity-based harms are more severe depending on the targeted victim group (Pezzella & Fetzer, 2017). In doing this, I found only victims of anti-disability hate crimes experienced greater serious injury during the perceived biased incident. Relative to race and religion, hate crimes directed towards disabled persons, in both a physical and mental sense, have received less empirical or theoretical attention. While some past works have demonstrated the heightened levels of harassment and other violence disabled individuals experience (see Hughes et al., 2012 or Jones et al., 2012 for adults and children, respectively), criminal justice interventions appear relatively rare (Macdonald et al., 2017).

Understanding this phenomenon through a lens of vulnerability, it is not surprising the increased risk disabled persons face in not only being victim to bias motivated crimes but also in receiving serious scholarly or practical consideration that might be used to either prevent or mitigate the increased harms imposed by anti-disability violence, but recent research indicates the evidence on such interventions is currently mixed (Mikton et al., 2014; Roulstone et al., 2011). The flip side of vulnerability, of course, is resilience, and it could be the case that the variation in reported injuries is likewise attributable to personal resilience uncaptured in the data. Recent research has pointed to in-group empathy as potential source of positive victim help-seeking behavior and, especially among the LGBTQIA+ community, and that such intergroup empathy is not only emphasized following hate crime victimization but is also linked with lower feelings of victim blame and greater intentionality to help hate crime victims (Patterson et al., 2019, 2023). This does shed light, though, on a larger issue of post-treatment bias in which other potential outcomes of the treatment that may affected the dependent measures are not appropriately addressed like individual differences in vulnerability or resilience or the context in which the crimes occur, but the data from which the sample for this study was drawn (the NCVS) contains limited information in such potentially corrupting information.

I also found that gender-based harms were more pronounced in terms of both later emotional and physical harm. Whether these effects varied by respondent gender was unexplored, but such an effect might be driven by crimes directed towards the transgender and gender non-conforming community and whether such persons are at a heightened risk of hate crime. Studies already indicate that transgender people experience discrimination and violence in various forms throughout their life, and that transgender people are likewise more likely to experience feelings of mental illness, depression, and suicidal ideations as a result of peer, family, other social reactions to their gender nonconformity (Connolly et al., 2016; Rothman et al., 2011; Stotzer, 2009). Such a finding is especially important within the timing of this article, in which several states are embarking on political campaigns that make more difficult the ability of trans and gender-nonconforming teenagers and adults to seek gender-affirming care, even despite evidence that such medical care is linked with lower feelings of depression or actions of self-harm (Lee et al., 2023; Tordoff et al., 2022).

Paying particular attention to later experienced emotional trauma, these effects experienced by hate crime victims in this sample again mirror symptoms of PTSD. Much work coming from scholars of trauma and violence have long documented that identity-based victimization is linked with PTSD among various populations but has been especially studied among LGBTQIA+ victims whose victimization was motivated by bias against their perceived identity (Cramer et al., 2012; Szymanski & Balsam, 2011; Wiginton et al., 2023). It is perhaps not surprising, then, that these results emphasize the increased risk of emotional harm for crimes motivated by bias against gender non-conformity and sexuality. Understanding these harms through the lens of trauma and PTSD is valuable to the extent that it frames solutions in the form of help seeking and responses to victimization, especially considering trauma-informed care. Healthcare providers and victim response services, then, can not only work within the paradigm of trauma-informed care but also practice cultural sensitivity towards marginalized populations that might already be hesitant to come forward with their victimization out of fear of secondary victimization or practitioner stereotyping. Indeed, trauma is intersected with culture, gender, race, location, or language, and culturally sensitive trauma-informed care is already a well-established domain with traumatology (Brown, 2008).

It is important to make note that my findings contrast with some previous studies. Past work on this issue, comparing biased and non-biased victim harm, has, instead, indicated that hate crimes are more harmful in their initial incidence (Fetzer & Pezzeall, 2019; Lantz & Kim, 2019; Malcom & Lantz, 2021). Because of this, my findings should be interpreted not as an attempt to discredit these past works, but instead fit within a larger field of evidence on the variety of harms associated with biased offending and victimization. But while these findings might conflict with other work on the topic, they still generally overlap with other theoretical and empirical work that emphasize the deleterious effects related to bias crime. As Perry (2001) argues, hate crimes are often committed with the intention to not only victimize the individual but the larger community from which the victim is perceived to belong to as well. Understood within this lens of power and vulnerability, that hate crimes are already perpetrated on the most vulnerable and marginalized persons in society, it would make sense that victims of such crimes are affected in such a long-lasting negative manner (Chakraborti & Garland, 2012).

Also, the evidence provided directly contributes to the debate surrounding hate crime legislation. By showing the increased risk of harm from biased victimization, these findings can be used to support hate crime penalty enhancements on the grounds of the increased harm to the victim, a retributive position. However, this work has little to say on other potential aims of hate crime legislation such as rehabilitation, incapacitation, or deterrence. If the aim of hate crime policy is to punish offenders on the basis of victim harm, then this work supports that notion by evidencing greater harm among biased victims. Yet, considering such victim harm is a dual-edged sword and should be used to push for greater hate crime prevention efforts and victim reporting services too. Similar services are offered to those of domestic violence or sexual assault, and criminal justice actors could likewise direct resources to hate crime victims knowing their increased risk of emotional and physical damage. Unfortunately, while there have been specific developments in the services of hate crime victims specifically across the country and internationally, empirical scholarship on these service providers is relatively limited. Much of this may be due to the issue of victim reporting to the appropriate authorities or other help-seeking behavior. Hate crime victims, usually coming from already marginalized communities and positions, are often hesitant to report their victimization for a handful of reasons including interpersonal factors or their perspective of police legitimacy and authority (Chakraborti, 2018; Lantz & Wenger, 2022; Myers & Lantz, 2020; Zaykowski, 2010; Wiedlitzk et al., 2018). Thus, the lack of inclusive practices for hate crime victims may speak to larger issues of criminal justice-community relations which then affect victim help seeking and attendant knowledge regarding hate crime victim services.

On the other hand, some have advocated for shift away from criminal justice intervention in addressing hate crimes. Indeed, Western social control institutions (particularly in the US) are founded upon the ideal that crimes are an offense to the state rather than the individual. Accordingly, justice in such contexts is retributive and offender-focused, often ignoring any material and symbolic losses experienced by the victim. Not only this, but such an approach has the potential to further alienate the offender from their respective community, shaming rather than reintegrating, perhaps increasing the likelihood of recidivism and continued hate victimization. Instead, a handful of scholars have advocated for restorative justice practices in dealing with hate crimes (Coates et al., 2006; Gavrielides, 2012; Kayali & Walters, 2021; Umbreit et al., 2003; Walters, 2014). Ranging from different approaches like formal apologies, group conferences, or individualized dialogue, applying restorative justice to hate crime has been limited, but with some promising results in addressing the harms associated with biased offending (Coats et al., 2006; Gavrielides, 2012; Umbreit et al., 2003). While the effect on reoffending has yet to be evaluated, future studies may be directed towards to expanding the usage of restorative justice or other reintegrative methods and any attendant success.

These findings should be taken in light of their limitations. First is that the analysis was restricted to only reported hate crimes with sufficient evidence and not all instances where the victim thought the offender was biased. This is a relatively stringent legalistic definition of “hate crime,” with some criticism, that might have impacted findings. Such an approach is related more so to the animus model of hate crime, but others might include, or compare, relative risk of harm of animus versus discriminatory hate crimes. Any person who perceived their victimization was bias motivated, but without the presence of the material evidence, might also experience elongated harms but are not considered as “real” hate crimes by NCVS standards. Another limitation within the data is the lack of information regarding bias motivated homicide, which is only available through official sources. Given the fact that many hate crimes end in homicide, and that research shows some qualitative differences in identity-based killings relative to “average” homicides, future studies might work include likelihood of death to their scholarship. Further, readers might further consider that these results may not be generalizable to findings from other data sources. While the NCVS and official sources (e.g., UCR or NIBRS) share commonality in their definitions of “hate crime,” the NCVS drastically over-estimates the incidence of hate crime in the US at a national scale, despite the fact the two sources seem to move in tandem over extended periods of time (Holder, 2022; Kenna & Thompson, 2021). In other words, using official data with this same methodology may yield different results and is worthy of further empirical investigation.

Another limitation to consider was the common support in the data following propensity score estimation. While covariate balance remains the crucial determinant of propensity score quality, common support among estimated propensity scores between the treated and untreated are necessary to make reliable comparisons across individuals in these groups. While I argued that the common support reached in this analysis was sufficient, there is no formal statistical test or procedure to make this determination, leaving it to researcher interpretation. Poor common support can also indicate a misspecified model in the estimation of propensity scores, and propensity score development is limited only to observable information contained in the dataset (Garrido et al., 2014; Leite, 2017). My decision to use IPWs and an estimated ATT (rather than ATE) were partially driven by the fact that common support was strongest using these methods. Additionally, the usage of IPWs can introduce extreme weights that also indicate propensity score model misspecification, but the estimated weights from this study did not include any notable outliers. And compared with matching, IPWs do not require each treated observations be paired with an untreated case but is more so used to achieve adequate covariate balance, which was completed in this study (Leite, 2017).

Despite these limitations, future research should continue to investigate the unique forms of harm associated with biased criminal offending. While hate crimes hurt more, as the study shows, it may be necessary to understand the potential for moderating effects, especially for race, gender, sexuality, and more. While it may be that hate crimes do hurt more, they hurt more for others. Whereas research has generally indicated that hate crime harm may be moderated by weapon use, co-offending, and more (Lantz & Kim, 2019; Malcom & Lantz, 2021; Mellgreen et al., 2021), victim characteristics and their relation to bias offending have been empirically ignored. While some research shows that the extent of hate crime harm may diverge on the bias motivation of the offender (e.g., anti-race, religion, sexuality, etc.), a more nuanced approach is needed to better understand how intersecting victim identities may contribute to variety of experienced harms or also the risk of hate crime in the first place (Meyer, 2010). Such an intersectional approach is needed to understand how the gendering of race (or racialization of gender), for instance, might produce different victimization outcomes and associated physical and emotional trauma. While this study found little difference between the experienced incident harm between biased and non-biased crimes, such a finding may vary based on these intersecting dynamics and is worthy of continued empirical investigation.


Austin, P. C. (2011). An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behavioral Research, 46(3), 399-424.

Bell, J. G., & Perry, B. (2015). Outside looking in: The community impacts of anti-lesbian, gay, and bisexual hate crime. Journal of Homosexuality, 62(1), 98-120.

Benier, K. (2017). The harms of hate: Comparing the neighbouring practices and interactions of hate crime victims, non-hate crime victims and non-victims. International Review of Victimology, 23(2), 179–201.

Brightman, S., Lenning, E., Lurie, K. J., & DeJong, C. (2023). Anti-transgender ideology, laws, and homicide: An analysis of the trifecta of violence. Homicide Studies, 10887679231201803.

Brown, L. S. (2008). Cultural competence in trauma therapy: Beyond the flashback. American Psychological Association.

Cabeldue, M. K., Cramer, R. J., Kehn, A., Crosby, J. W., & Anastasi, J. S. (2018). Measuring attitudes about hate: Development of the hate crime beliefs scale. Journal of Interpersonal Violence, 33(23), 3656-3685.

Chakraborti, N. (2018). Responding to hate crime: Escalating problems, continued failings. Criminology & Criminal Justice, 18(4), 387-404.

Chakraborti, N., & Garland, J. (2012). Reconceptualizing hate crime victimization through the lens of vulnerability and ‘difference’. Theoretical Criminology, 16(4), 499-514.

Cheng, W., Ickes, W., & Kenworthy, J. B. (2013). The phenomenon of hate crimes in the United States. Journal of Applied Social Psychology, 43(4), 761-794.

Coates, R. B., Umbreit, M. S., & Vos, B. (2006). Responding to hate crimes through restorative justice dialogue. Contemporary Justice Review, 9(1), 7-21.

Connolly, M. D., Zervos, M. J., Barone II, C. J., Johnson, C. C., & Joseph, C. L. (2016). The mental health of transgender youth: Advances in understanding. Journal of Adolescent Health, 59(5), 489-495.

Cramer, R. J., Chandler, J. F., & Wakeman, E. E. (2010). Blame attribution as a moderator of perceptions of sexual orientation—based hate crimes. Journal of Interpersonal Violence, 25(5), 848-862.

Cramer, R. J., Clark III, J. W., Kehn, A., Burks, A. C., & Wechsler, H. J. (2014). A mock juror investigation of blame attribution in the punishment of hate crime perpetrators. International Journal of Law and Psychiatry, 37(6), 551-557.

Cramer, R. J., Laxton, K. L., Chandler, J. F., Kehn, A., Bate, B. P., & Clark III, J. W. (2017). Political identity, type of victim, and hate crime‐related beliefs as predictors of views concerning hate crime penalty enhancement laws. Analyses of Social Issues and Public Policy, 17(1), 262-285.

Cramer, R. J., McNiel, D. E., Holley, S. R., Shumway, M., & Boccellari, A. (2012). Mental health in violent crime victims: Does sexual orientation matter?. Law and Human Behavior, 36(2), 87.

Cramer, R. J., Nobles, M. R., Amacker, A. M., & Dovoedo, L. (2013). Defining and evaluating perceptions of victim blame in antigay hate crimes. Journal of Interpersonal Violence, 28(14), 2894-2914.

Cramer, R. J., Wright, S., Long, M. M., Kapusta, N. D., Nobles, M. R., Gemberling, T. M., & Wechsler, H. J. (2018). On hate crime victimization: Rates, types, and links with suicide risk among sexual orientation minority special interest group members. Journal of Trauma & Dissociation, 19(4), 476–489.

Deloughery, K., King, R. D., & Asal, V. (2012). Close cousins or distant relatives? The relationship between terrorism and hate crime. Crime & Delinquency, 58(5), 663-688.

Fetzer, M. D., & Pezzella, F. S. (2019). The nature of bias crime injuries: A comparative analysis of physical and psychological victimization effects. Journal of Interpersonal Violence, 34(18), 3864–3887.

Gavrielides, T. (2012). Contextualizing restorative justice for hate crime. Journal of interpersonal Violence, 27(18), 3624-3643.

Garrido, M. M., Kelley, A. S., Paris, J., Roza, K., Meier, D. E., Morrison, R. S., & Aldridge, M. D. (2014). Methods for constructing and assessing propensity scores. Health Services Research, 49(5), 1701-1720.

Gellman, S. (1992). Hate crime laws are thought crime laws. Annual Survey of American Law, 4, 485-507.

Gladfelter, A. S., Lantz, B., & Ruback, R. B. (2017). The Complexity of Hate Crime and Bias Activity: Variation Across Contexts and Types of Bias. Justice Quarterly, 34(1), 55–83.

Green, D. P., Strolovitch, D. Z., & Wong, J. S. (1998). Defended neighborhoods, integration, and racially motivated crime. American Journal of Sociology, 104(2), 372–403.

Hate crime statistics, 2019. (2019). Federal Bureau of Investigation.

Herek, G. M., Gillis, J. R., Cogan, J. C., & Glunt, E. K. (1997). Hate crime victimization among lesbian, gay, and bisexual adults: Prevalence, psychological correlates, and methodological issues. Journal of Interpersonal Violence, 12(2), 195–215.

Hughes, K., Bellis, M. A., Jones, L., Wood, S., Bates, G., Eckley, L., McCoy, E., Mikton, C., Shakespeare, T., & Officer, A. (2012). Prevalence and risk of violence against adults with disabilities: A systematic review and meta-analysis of observational studies. The Lancet, 379(9826), 1621–1629.

Holder, E. (2022). The convergence of Hate: An analysis of UCR–NCVS hate crime trends, 2003 to 2020. Crime & Delinquency, 00111287221131001.

Holder, E., Edwards, B., & Osborne, D. (2022). An exploratory study on the structural and demographic predictors of hate crime across the rural-urban divide. Victims & Offenders, 1-23.

Iganski, P., & Lagou, S. (2015). Hate crimes hurt some more than others: Implications for the just sentencing of offenders. Journal of Interpersonal Violence, 30(10), 1696–1718.

Jacobs, J. B. & Potter, K. (1998). Hate crimes: Criminal law & identity politics. Oxford: Oxford University Press.

Jenness, V. & Broad, K. (1997). Hate crimes: New social movements and the politics of violence. New Brunswick: Aldine Transaction.

Jenness, V. & Grattet, R. (2001). Making a hate crime: From social movement to law enforcement. New York: Russell Sage Foundation.

Jones, L., Bellis, M. A., Wood, S., Hughes, K., McCoy, E., Eckley, L., Bates, G., Mikton, C., Shakespeare, T., & Officer, A. (2012). Prevalence and risk of violence against children with disabilities: A systematic review and meta-analysis of observational studies. The Lancet, 380(9845), 899–907.

Kayali, L., & Walters, M. A. (2021). Responding to hate incidents on university campuses: benefits and barriers to establishing a restorative justice programme. Contemporary Justice Review, 24(1), 64-84.

Kena, G. & Thompson, A. (2021). Hate crime victimization, 2005-2019. Bureau of Justice Statistics.

King, G., & Nielsen, R. (2019). Why propensity scores should not be used for matching. Political Analysis, 27(4), 435-454.

Klein, B. R., & Allison, K. (2018). Accomplishing difference: How do anti-race/ethnicity bias homicides compare to average homicides in the United States? Justice Quarterly, 35(6), 977–1003.

Lantz, B., & Kim, J. (2019). Hate crimes hurt more, but so do co-offenders: Separating the influence of co-offending and bias on hate-motivated physical injury. Criminal Justice and Behavior, 46(3), 437–456.

Lantz, B., & Wenger, M. R. (2022). Are Asian victims less likely to report hate crime victimization to the police? Implications for research and policy in the wake of the COVID-19 pandemic. Crime & Delinquency, 68(8), 1292-1319.

Lawrence, F. M. (1994). The punishment of hate: Toward a normative theory of bias-motived crimes. Michigan Law Review, 93, 320-381.

Lawrence, F. M. (1999). Punishing hate: Bias crimes under American law. Cambridge: Harvard University Press.

Lee, J. Y., & Rosenthal, S. M. (2023). Gender-Affirming Care of Transgender and Gender-Diverse Youth: Current Concepts. Annual Review of Medicine, 74.

Leite, W. (2017). Practical propensity score methods using R. Sage Publications.

Lyons, C. J. (2007). Community (dis)organization and racially motivated crime. American Journal of Sociology, 113(3), 815–863.

Macdonald, S. J., Donovan, C., & Clayton, J. (2017). The disability bias: understanding the context of hate in comparison with other minority populations. Disability & Society, 32(4), 483-499.

Malcom, Z. T., & Lantz, B. (2021). Hate crime victimization and weapon use. Criminal Justice and Behavior, 48(8), 1148-1165.

Malcom, Z. T., Wenger, M. R., & Lantz, B. (2023). Politics or prejudice? Separating the influence of political affiliation and prejudicial attitudes in determining support for hate crime law. Psychology, Public Policy, and Law, 29(2), 182.

McDevitt, J., Levin, J., & Bennett, S. (2002). Hate crime offenders: An expanded typology. Journal of Social Issues, 58(2), 303–317.

Mellgren, C., Andersson, M., & Ivert, A.-K. (2021). For whom does hate crime hurt more? A comparison of consequences of victimization across motives and crime types. Journal of Interpersonal Violence, 36(3–4), NP1512-1536NP.

Messner, S. F., McHugh, S., & Felson, R. B. (2004). Distinctive characteristics of assaults motivated by bias. Criminology, 42(3), 585-618.

Meyer, D. (2010). Evaluating the severity of hate-motivated violence: Intersectional differences among LGBT hate crime victims. Sociology, 44(5), 980–995.

Mikton, C., Maguire, H., & Shakespeare, T. (2014). A systematic review of the effectiveness of interventions to prevent and respond to violence against persons with disabilities. Journal of Interpersonal Violence, 29(17), 3207-3226.

Mills, C. E., Freilich, J. D., & Chermak, S. M. (2017). Extreme hatred: Revisiting the hate crime and terrorism relationship to determine whether they are “close cousins” or “distant relatives”. Crime & Delinquency, 63(10), 1191-1223.

Myers, W., & Lantz, B. (2020). Reporting racist hate crime victimization to the police in the United States and the United Kingdom: A cross-national comparison. The British Journal of Criminology, 60(4), 1034-1055.

Noelle, M. (2002). The ripple effect of the Matthew Shepard murder: Impact on the assumptive worlds of members of the targeted group. American Behavioral Scientist, 46(1), 27-50.

Nolan, J. J., & Akiyama, Y. (1999). An analysis of factors that affect law enforcement participation in hate crime reporting. Journal of Contemporary Criminal Justice, 15(1), 111-127.

Osborne, D. L., & Swartz, K. (2021). Routine activity theory and farm equipment theft: A macro-level approach. International Journal of Rural Criminology, 6(1), 1-18.

Paterson, J. L., Brown, R., & Walters, M. A. (2019). Feeling for and as a group member: Understanding LGBT victimization via group‐based empathy and intergroup emotions. British Journal of Social Psychology, 58(1), 211-224.

Paterson, J. L., Walters, M. A., & Hall, L. (2023). Ingroup empathy, help, and blame after anti-LGBT+ hate crime. Journal of Interpersonal Violence, 08862605231200212.

Perry, B. (2001). In the name of hate: Understanding hate crime. Routledge.

Pezzella, F. S., & Fetzer, M. D. (2017). The likelihood of injury among bias crimes: An analysis of general and specific bias types. Journal of Interpersonal Violence, 32(5), 703–729.

Piatkowska, S. J., Messner, S. F., & Yang, T. C. (2019). Understanding the relationship between relative group size and hate crime rates: Linking methods with concepts. Justice Quarterly, 36(6), 1072-1095.

Powers, R. A., & Socia, K. M. (2019). Racial animosity, adversary effect, and hate crime: Parsing out injuries in intraracial, interracial, and race-based offenses. Crime & Delinquency, 65(4), 447-473.

Rosenbaum, P. R., & Rubin, D. B. (1983). The central role of the propensity score in observational studies for causal effects. Biometrika, 70(1), 41-55.

Rosenbaum, P. R., & Rubin, D. B. (1984). Reducing bias in observational studies using subclassification on the propensity score. Journal of the American Statistical Association, 79(387), 516-524.

Rosenbaum, P. R., & Rubin, D. B. (1985). Constructing a control group using multivariate matched sampling methods that incorporate the propensity score. The American Statistician, 39(1), 33-38.

Rothman, E. F., Exner, D., & Baughman, A. L. (2011). The prevalence of sexual assault against people who identify as gay, lesbian, or bisexual in the United States: A systematic review. Trauma, Violence, & Abuse, 12(2), 55-66.

Roulstone, A., Thomas, P., & Balderston, S. (2011). Between hate and vulnerability: Unpacking the British criminal justice system’s construction of disablist hate crime. Disability & Society, 26(3), 351-364.

Ruback, R. B., Gladfelter, A. S., & Lantz, B. (2018). Hate crime victimization data in Pennsylvania: A useful complement to the Uniform Crime Reports. Violence and Victims, 33(2), 330-350.

Saucier, D. A., Brown, T. L., Mitchell, R. C., & Cawman, A. J. (2006). Effects of victims' characteristics on attitudes toward hate crimes. Journal of Interpersonal Violence, 21(7), 890-909.

Spano, R., & Nagy, S. (2005). Social guardianship and social isolation: An application and extension of lifestyle/routine activities theory to rural adolescents. Rural Sociology, 70(3), 414-437.

Stacey, M. (2011). Distinctive characteristics of sexual orientation bias crimes. Journal of Interpersonal Violence, 26(15), 3013-3032.

Stacey, M., Carbone-López, K., & Rosenfeld, R. (2011). Demographic change and ethnically motivated crime: The impact of immigration on anti-Hispanic hate crime in the United States. Journal of Contemporary Criminal Justice, 27(3), 278-298.

Steen, S., & Cohen, M. A. (2004). Assessing the public's demand for hate crime penalties. Justice Quarterly, 21(1), 91-124.

Stotzer, R. L. (2009). Violence against transgender people: A review of United States data. Aggression and Violent Behavior, 14(3), 170-179.

Szymanski, D. M., & Balsam, K. F. (2011). Insidious trauma: Examining the relationship between heterosexism and lesbians’ PTSD symptoms. Traumatology, 17(2), 4-13.

Tordoff, D. M., Wanta, J. W., Collin, A., Stepney, C., Inwards-Breland, D. J., & Ahrens, K. (2022). Mental health outcomes in transgender and nonbinary youths receiving gender-affirming care. JAMA Network Open, 5(2).

Umbreit, M., Lewis, T., & Burns, H. (2003). A community response to a 9/11 hate crime: Restorative justice through dialogue. Contemporary Justice Review, 6(4), 383-391.

Walters, M. A. (2014). Hate crime and restorative justice: Exploring causes, repairing harms. Oxford University Press.

James Weinstein (1992) First amendment challenges to hate crime legislation: Where's the speech?. Criminal Justice Ethics, 11(2), 6-20.

Wiedlitzka, S., Mazerolle, L., Fay-Ramirez, S., & Miles-Johnson, T. (2018). Perceptions of police legitimacy and citizen decisions to report hate crime incidents in Australia. International Journal for Crime, Justice and Social democracy, 7(2), 91-106.

Wiginton, J. M., Murray, S. M., Baral, S. D., & Sanchez, T. H. (2023). Targeted violence as a risk factor for posttraumatic stress disorder among cisgender gay, bisexual, and other men who have sex with men in the United States. Journal of Interpersonal Violence, 08862605231169755.

Williams, M. L., & Tregidga, J. (2014). Hate crime victimization in Wales: Psychological and physical impacts across seven hate crime victim types. British Journal of Criminology, 54(5), 946–967.

Wilson, M. S., & Ruback, R. B. (2003). Hate crimes in Pennsylvania, 1984–99: Case characteristics and police responses. Justice Quarterly, 20(2), 373-398.

Wisconsin v. Mitchell, 508 U.S. 487 (1993).

Zaykowski, H. (2010). Racial disparities in hate crime reporting. Violence and Victims, 25(3), 378-394.

No comments here
Why not start the discussion?