HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 1 Hate crimes hurt some more than others: implications for the just sentencing of offenders Paul Iganski Lancaster University, UK Spiridoula Lagou H8hurts, UK Author note Paul Iganski, Lancaster University Law School, UK Spiridoula Lagou, H8hurts, UK Correspondence concerning this article should be addressed to Paul Iganski, Lancaster University Law School, Bowland North, Bailrigg, Lancaster, LA1 4YN, UK. E-mail: p.iganski@lancaster.ac.uk HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 2 Abstract An accumulation of research evidence indicates that hate crimes as a category of offence are more serious than similar but otherwise motivated crimes in respect of the greater post-victimization distress reported by victims. Such evidence has been used by advocates of hate crime laws to justify the imposition of greater penalties in the sentencing of convicted hate crime offenders. However, in focusing on the commonalities of the greater level of post-victimization impacts experienced by hate crime victims as a group, the research evidence to date has obscured the diversity of reactions between victims. Consequently, this article expands the evidence in new directions by illuminating the variation in reported victim impacts. The analysis presented uses data from the Crime Survey for England and Wales on reported racially motivated crime and reveals that not all victims report being emotionally affected by hate crime, not all victims are affected the same way, and some victims of racially motivated crime report less of an emotional impact than some victims of equivalent but otherwise motivated crimes. The research findings are used to reason that in any individual case of hate crime the motivating sentiments of the offender provide an unreliable indicator of the harms inflicted upon the victim. Therefore a blanket uplift in penalty in every case of hate crime which rests upon the offender’s mental state— their prejudice, bigotry, bias, or ‘hate’ — cannot be justified if the justification for sentence uplift is to give offenders their just deserts for the harms they inflict. Instead, the justification must rest upon the culpability of the offender for the harms they may or may not actually inflict. Just as there is variation in victim impacts, there will be variation in offender culpability: discretion and flexibility in the mode of enhanced sentencing is therefore necessary to ensure justice for offenders as well as victims. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 3 Introduction: the hate debate In arguments that have been long rehearsed, critics of hate crimes laws in the United States have argued that the additional punishment of hate crime offenders over and above the punishment that can be meted out for a similar, but otherwise motivated crime, amounts to the state criminalizing the expression of certain thoughts, opinions and values (cf. Bruce 2001; Gey 1997; Hurd 2001) as it is such aspects of the offender’s mind that distinguishes a hate crime from a parallel crime. Against this argument, though, supporters of hate crime laws have contended that it is not the bad values or the motivation of the offender (perhaps expressed in things they say while committing the offence) that are being punished. Instead, they propose that the laws impose greater punishment for the greater harms inflicted by hate crimes. The harms most commonly claimed concern the experience of psychological and emotional trauma by victims following a hate crime (cf. Lawrence 1999 & 2006), and also trauma vicariously experienced by those who share the same identity as the primary victim. Accordingly, enhanced sentencing of convicted hate crime offenders simply gives the offenders the just deserts for the harms they inflict. To use the words of Hurd and Moore (2004) on the matter, who cast this argument as the ‘wrongdoing analysis’, the conclusions drawn from the empirical claims about the greater harms inflicted by hate crime are that: “…hate/bias crime offenders deserve more punishment because they are more blameworthy; that they are more blameworthy because they have done greater wrong; and that they have done greater wrong because they have (typically) caused more harm, namely, the psychological traumatization of victims.” (Hurd & Moore 2004, p. 1087) Although Hurd and Moore focussed on the limitations of the empirical claims behind the ‘wrongdoing analysis’, an accumulating body of evidence does now demonstrate that as a group victims of ‘hate crime’ collectively experience a greater degree of negative psychological impacts when compared with victims of parallel crimes. In drawing from a review of some of the evidence, Craig-Henderson and Sloan posited that the range of negative emotions experienced by victims of racist crime are “qualitatively distinct” from the emotions experienced by victims of parallel crimes (2003, p. 482). The consequences for the HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 4 individual victim of this qualitative difference in hate crime are manifest in symptoms of post- victimization distress. Quantitative differences between hate crime victims and victims of parallel crimes in experiencing such symptoms have also been illuminated by a number of research studies — at first in the United States, and more recently in the United Kingdom. Initially, generalizing the findings of this evidence was limited by the use of small samples of victims of hate crimes and comparison crimes (cf. Garofalo 1997) and non-random samples (cf. Herek, Gillis, Cogan & Glunt 1997; McDevitt et al., 2001). Such limitations have now been overcome by the use of large random samples in the secondary analysis of crime victimisation survey data (cf. Iganski 2008; Smith et al., 2012) as is the case with the analysis to be presented in this paper. Overall, the evidence shows that while hate crime victims experience and report the same types of stress and trauma symptoms as victims of parallel crimes, comparisons of samples of hate crime victims and victims of comparison crimes have shown that hate crime victims are more likely to report:  That they were “very much” emotionally affected (Smith et al., 2012, p. 22), that they “rate their crimes as very serious”, and “that the crimes had a ‘great deal’ of effect on their lives in both the short term and long term” (Garofalo 1997, p. 141);  Having a stronger emotional reaction even when controlling for crime type (Botcherby et al., 2011; Iganski 2008);  Having a greater propensity to have unwanted intrusive thoughts about the incident and to “feel like not wanting to live any longer…” (McDevitt et al., 2001, p. 709);  Having a longer period of recovery from the effects of victimisation (Herek, Gillis, Cogan & Glunt 1997) and that overcoming the incident “had been very difficult” (McDevitt et al., 2001, p. 711);  That “being ‘frightened or scared’ was the primary response in the immediate aftermath of the crime”, (while less likely to report “being ‘angry or mad’ as their predominant response”) (Garofalo 1997, p. 141); Furthermore, in terms of specific symptoms of distress, compared with victims of parallel crimes, victims of hate crime are more likely to report experiencing: HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 5  Higher levels depression (McDevitt et al., 2001) and withdrawal (Ehrlich 1992);  Anxiety (Ehrlich 1992) and nervousness (McDevitt et al., 2001);  Loss of confidence (Ehrlich 1992);  Anger (Herek, Gillis, Cogan & Glunt 1997);  Increased sleep difficulties (Ehrlich 1992);  Difficulty concentrating (McDevitt et al., 2001);  Fear and reduced feelings of safety (McDevitt et al., 2001); These characteristics of post-victimization distress have also been illuminated in depth by a number of studies which have focused solely on hate crime victims without comparison samples of victims of parallel crimes (cf. Barnes & Ephross 1994; Dzelme 2008; Hershberger & D’Augelli 1995; Otis & Skinner 1996; Victim Support 2006). In sum, as a body of evidence, these research findings support the hypothesis that hate crimes hurt more than parallel crimes in respect of the experience of post victimization distress. However, in illuminating the commonality of impacts and the greater severity of stress and trauma symptoms experienced by ‘hate crime’ victims as a group, the research literature on the matter to date has in general obscured the diversity of the experience of impacts among hate crime victims. There have been just a few exceptions to this trend. Craig-Henderson and Sloan (2003) have argued for instance, that the unique reaction by hate crime victims is due to their perception of their experience of victimization as an attack upon the core of their identity. Because of this, black minority victims of racist crime will experience the crime more acutely than white majority group victims because the crime serves as a painful reminder of the cultural heritage of past and ongoing discrimination, stereotyping and stigmatization of their identity group: “When an anti-black racist hate crime occurs it brings all of the dormant feelings of anger, fear and pain to the collective psychological forefront of the victim. This is not the case when whites are the target of racist hate crime” (Craig-Henderson & Sloan, 2003, p. 485). More recently, in employing an intersectional approach to examine how a sample of lesbian, gay, bisexual and transgender people evaluated the severity of their experiences of HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 6 hate crime, Meyer’s (2010) findings from his study involving semi-structured interviews with a non- random sample of forty-four respondents in New York City suggested that middle-class white respondents were more likely than low-income respondents of color to perceive their experiences as severe. This was chiefly because of differing expectations between the groups of the likelihood of hate crime victimization based on different reference groups victims used to interpret their experience. This article expands the research evidence even further by exploring the differential impacts of racist victimization as moderated and mediated by particular demographic and socio-economic characteristics of respondents (in other words, an ‘intersectional approach’) and also by type and frequency of crime victimization. Research design and sample The research involved a secondary analysis of data collected by the Crime Survey for England and Wales (formerly known as the British Crime Survey, but renamed from April 2012 to better reflect its geographic coverage) and obtained from the UK Data Archive 1 . The Crime Survey for England and Wales interviews approximately 46,000 respondents (aged 16 and over) every year. The survey is generally regarded as providing a more complete measure of crime compared with police records given that not all victims report crimes to the police for a well-known variety of reasons. The aim of the analysis reported here was to explore diversity among victims of racist crime in their experience of post victimization psychological impacts. It is the preferred practice, in the reporting of the survey findings by the U.K. Home Office which initiated and reports on the survey, for two years of data to be combined to give a sufficient sample size for fine-grained analysis without a too prolonged reference period. However, given that the analysis that was undertaken was concerned with within-group differences of the survey sample of 1 Material from Crown copyright records made available through the Home Office and the UK Data Archive has been used by permission of the Controller of Her Majesty's Stationery Office and the Queen's Printer for Scotland. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 7 victims of racist crime, rather than with generalizing beyond the sample to the wider population about the incidence or prevalence of crime, consideration of the reference period is not as significant. Hence the analysis combined three sweeps of the survey — from the 2009-10, 2010-11 and 2011-12 surveys (the three most recent data sets at the time of writing) — to enable sufficient sample sizes for logistic regression analyses. All respondents in the survey are asked whether they experienced certain types of crimes in the previous twelve months. Usually, just over one-fifth report being a victim of crime at least once. In its estimates of crime based on the Crime Survey for England and Wales the U.K. Home Office usually excludes reported threats captured by the survey because such incidents do not always meet the criteria of a criminal offence despite being assigned valid offence codes in the survey. However, given that public order offences such as threats constitute a substantial proportion of prosecuted racially aggravated offences in England and Wales the exclusion of threats from the analysis skews the reported picture of racist victimization. Threats are therefore included in the analyses presented in this paper. For each crime they report in the survey, respondents are asked an array of questions. One of the questions is: “Do you think the incident was racially motivated?” (In relying upon the victim’s perception about motivation behind the crime this question is susceptible to over-reporting because some crimes perceived by the victim to have been racially motivated could have been motivated by other reasons. By the same token, though, it is also susceptible to under-reporting because in some crimes victims might not be aware that they were targeted because of their ‘race’ or ethnicity). While respondents from each ethnic group reported some racially motivated crimes, minority ethnic group respondents were more likely than whites to report such crimes (Table 1) (survey respondents are asked to identify their ethnic group from a list of ethnic group categories used in the 2001 and 2011 U.K. censuses). Overall, just over one-in-fifty crimes captured by the survey were perceived by victims to have been racially motivated. However, the reported rate of racist crime victimization for minority groups was much higher: one-in-seven crimes experienced by Asian respondents, for HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 8 instance, were believed by them to have been racially motivated, compared with just over one in a hundred for whites. (Table 1 about here) Racist language used by the offender was the most common reported indicator for respondents that an incident was racially motivated, along with the victim suggesting that they had been targeted because of their ‘race’ or country of origin (Table 2). (Table 2 about here) The core of the analysis reported in this paper was applied to the 441 incidents of crime believed by victims to have been racially motivated and reported by 398 respondents across the three sweeps of the survey used for the analysis (a number of respondents reported more than one discrete separate racially motivated crime and therefore completed more than one victim form, hence the difference between the number of incidents of crime and the number of victims. Victims of a series of related racially motivated crimes, the same crime under the same circumstances probably committed by the same people, are asked to complete only one victimization module for the most recent incident in the series). How ‘hate’ hurts in different ways: evidence from the Crime Survey for England and Wales To set the context for the results of the analysis to be reported, it is instructive to briefly review the major characteristics of the sample of victims of racially motivated crime, and the sample of incidents of racially motivated crime used in the analysis, in comparison to otherwise motivated crime reported in the survey. To take victims first, when comparing victims of racially motivated with otherwise motivated crimes (Table 3) victims of racially motivated crime were more likely to be in a younger age group, from a minority ethnic group, in a household with a low household income, and to be unemployed. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 9 (Table 3 about here) Statistically significant differences are evident for all these categories of respondents when comparing racially motivated crimes with otherwise motivated crimes. The apparent greater likelihood for males to be victims of racially motivated crime than otherwise motivated crime is just below, but on the margins of, statistical significance (Z= 1.89). To turn to the crimes, when comparing incidents of racially motivated and otherwise motivated crime (Table 4), it is evident that racially motivated crimes were more likely to involve interpersonal offences of violence with and without injury, and threats, and be part of a series of repeat victimization. As a group, victims in incidents of racially motivated crime were more likely to report that they had an emotional reaction following the incident and the differential between victims of racially motivated crime as a group and victims of parallel crime as a group widens further when the extent of the reported emotional reaction is considered. Victims of crimes perceived to be racially motivated were twice as likely as victims of otherwise motivated crimes to state that they had been affected “very much”. (Table 4 about here) The data therefore support the assertion that victims of racially motivated crime — as a group — experience greater negative impacts than victims of otherwise motivated crime. However, it is also apparent that not all victims of racially motivated and otherwise motivated crime report being emotionally affected, or report the same post-victimization effects. In the case of the latter, those victims who report having an emotional reaction are asked for every crime they report, “Which of these reactions did you personally have?” and are given a list of ten options (anger, annoyance, anxiety/panic attacks, crying/tears, depression, difficulty sleeping, fear, loss of confidence/feeling vulnerable, shock, and other) from which they can choose as many as apply. For each type of HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 10 emotional reaction higher proportions of victims in crimes they believed to be racially motivated reported the symptoms, apart from the reaction of “annoyance”, when compared with victims of parallel crimes. The pattern of difference that can be observed, however, obscures variations in emotional reaction among victims of racist hate crime. First, in just under one in ten incidents, victims reported experiencing no emotional reaction. On this measure then, they were less affected than those victims of parallel crimes who did report having an emotional reaction. Second, the different emotional reactions comprising the list for multiple selection by respondents might be grouped into three separate categories of reaction: externalized reactions, internalized reactions, and a third group which combines both externalized and internalized reactions. Anger and annoyance can be considered to be externalized reactions. Anger is a natural response, a basic human reaction when a person feels wronged, or when they have been threatened or attacked. It can be manifest in feelings of retaliation, rage, revenge and aggression towards the source of the wrong doing (we include annoyance as a similar type of natural response). One quarter of the victims who perceived incidents of crime to be racially motivated who reported having an emotional reaction reported solely such an externalised reaction of anger, or annoyance, or both. By contrast, anxiety and panic attacks, crying and tears, depression, difficulty sleeping, fear, loss of confidence and feelings of vulnerability, and shock, are arguably all symptoms of hurt turned within the person who has been wronged (or internalized) rather than against the wrong doer. Almost one-in-five victims of crimes perceived to have been racially motivated reported such an internalized reaction, almost twice the proportion for victims of otherwise motivated crimes. Anger and annoyance can sometimes be positive responses when such emotions provide a force for change or activism against the source of the wrong doing. But they can also be destructive emotions when internalized, perhaps due to the person’s inability to redress or overcome the wrong inflicted, and they can become linked to other internalized reactions, especially in the long term. Hence a third category of emotional reactions can be established combining both externalized and internalized HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 11 reactions for those respondents who reported both types of reaction. This was the largest category of emotional reaction with almost half (47.3%) of victims of crimes perceived to have been racially motivated reporting a syndrome of both externalised and internalized reactions compared with under a third (29.4%) of victims of parallel crimes. Although the data indicate that hate crimes hurt more in the respect that as a victim group, victims of racially motivated crime are more likely to report experiencing an emotional reaction to the crime compared with victims of parallel crimes, and they are more likely to experience internalized reactions, or a combination of externalized and internalized reactions, it is clear that not all victims react in the same way. Logistic regression was used to estimate how much the risk of experiencing the different types of post victimization reactions is increased or reduced according to demographic and socio-economic characteristics of victims and type and frequency of crime victimization. The aim of the logistic regression analysis was to explore the association between victim and victimization characteristics and the risk of experiencing the particular emotional reactions while controlling for the characteristics of the variables included in the analysis, rather than producing a model of best fit (as the purpose of the analysis was to expand understanding of the variation in impacts of hate crime upon victims rather than seeking to fully model an account for that diversity). The sub-sample of respondents who reported having no emotional reaction after experiencing a crime they perceived as being racially motivated was too small to undertake further analysis for all the variables of interest therefore the findings reported here are concerned with the three categories of reaction: externalized reactions, internalized reactions, and the syndrome of symptoms of both externalized and internalized reactions. The analysis showed some clear associations between victim and victimization characteristics and the risk of experiencing particular types of emotional reaction. The observations here are confined to those findings that were statistically significant (p ˂ 0.05). A summary of the significant associations is provided in Table 5. (Table 5 about here) HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 12 Externalized reactions alone The odds of reporting the externalised reactions of only anger or annoyance or both (Table 6) were over three times greater for male victims of crime compared with female victims. Young people in the 16-29 age group had higher odds than the older age groups in reporting such externalised reactions alone. The odds of reporting externalized reactions were greater for white victims compared with non- white victims. When considering the type of crime, the odds of reporting externalized reactions were significantly greater for victims in incidents of property crime compared with incidents of personal crime or threats. And the odds of reporting externalized reactions alone were also significantly greater for victims who had experienced single separate incidents of crime compared with victims who had experienced a series of incidents or those who had experienced two or more separate and unrelated incidents. Internalized reactions alone When examining reported reactions of the internalized symptoms alone — anxiety and panic attacks, crying and tears, depression, difficulty sleeping, fear, loss of confidence and feelings of vulnerability, and shock — much of the pattern is the mirror opposite of the pattern for externalized symptoms of anger and annoyance in that female victims of incidents of racially motivated crime had over twice the odds of reporting internalized symptoms alone compared with male victims. The mid-age category of 30-59 had the greatest odds of reporting internalized symptoms. And the odds for victims in incidents of personal crime involving violence or threats for reporting internalized symptoms alone were greater when compared with victims of incidents of property crime. Syndrome of symptoms of both externalized and internalized reactions The odds of reporting both externalized and internalized symptoms combined were greater for female victims compared with males, for non-white victims compared with whites, victims in the youngest age category 16-29 compared with the other age groups, victims in the lowest household income category, and victims of a series of incidents. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 13 Implications for just outcomes for offenders The analysis presented in this paper demonstrates that as a group, victims of racist crime collectively experience greater emotional and psychological harms compared with victims of parallel crimes. The same appears to be the case for other groups of hate crime victims. Analysis of CSEW data has shown that disabled victims of hate crime as a group also report more damaging post-victimization psychological impacts than victims of parallel crimes (Coleman et al., 2013; Nocon et al., 2011) and the same is the case when victims of different types of identity-based crime are considered as one aggregate group (Botcherby et al., 2011; Iganski & Lagou 2014; Smith et al., 2012). The greater psychological and emotional damage is arguably inflicted by the message sent by the offender, by the sentiments conveyed by their crime: it is the offender’s expressed values that appear to hurt. Frederick Lawrence has argued that victims of ‘race-hate crime’ experience attacks as a form of racial stigmatisation and that an incident “carries with it the clear message that the target and his (sic) group are of marginal value” (Lawrence, 2006: 3). In essence, it is the values of the attacker striking at the core of the victim’s identity, which hurt more. However, advocates of enhanced sentencing in cases of hate crime do not argue that the offender should be punished more severely for their bad character as manifest in the values expressed in their crime. Neither, and consistent with the application of criminal sanction in the liberal polity (Jacobs 2013), do they propose that the enhanced sentence should serve to require virtuous character beyond encouraging critical reflection by the offender about how the expression of their particular values is consistent with participation in civil society. Instead, in applying the principle of retributive desert, they simply propose that the severity of the punishment should be proportionate to the severity of the criminal conduct. From this desert perspective, the imposition of more or less criminal sanction than the severity of the criminal conduct would be unfair (von Hirsch, 1998: 172). If bad values, or bad character, are excluded from the calculation of severity, criminal sanction should be determined by the relative harmful outcomes inflicted. Yet the evidence presented in this paper of the disparate outcomes in the post-victimization impacts experienced by hate crime victims raises a problem for the just sentencing of offenders if an indiscriminate uplift in penalty would be applied in all cases. Fundamentally, in taking the analysis beyond the type of findings about the mental trauma of hate crime that have already been published to date, the evidence HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 14 presented in this paper demonstrates that there is considerable variation among victims of racist crime when it comes to their reported reactions. It is evident that not all victims react the same way, and some victims of racially motivated crime report less post-victimization distress than some victims of parallel crimes. The analysis shows therefore that victims of racially motivated crime are not a homogeneous group in terms of the emotional impact of the crimes (and neither are victims of parallel crimes for that matter). Although the evidence has yet to be established, given the already known similarities between different identity groups of hate crime victims in terms of the post-victimization trauma they experience, it is likely that there will also be considerable variation in the impacts felt among victims who are targeted because of their religion, sexual identity, or a disability they might have, for instance. This lack of uniformity in the impacts of hate crimes upon victims exposes a potential problem of justice in sentencing outcomes for offenders. Arguably, to justify a blanket uplift in penalty, every victim of hate crime must experience greater harms than every victim of parallel crime. The analysis presented in this paper shows that this is certainly not the case. If an indiscriminate penalty uplift would be applied, some hate crime offenders would potentially receive a higher sentence on the basis of the greater harms they are assumed to inflict even in instances where their victims experience less harm than victims in some cases of otherwise motivated but equivalent crimes where the offenders will receive lower penalties. The logical process for fair sentencing then, to avoid potential injustice for the offender, would arguably be for the victim impact to be demonstrated on a case-by-case basis to the satisfaction of the court that the impact of the hate crime on the victim is greater than would be the case in a parallel crime. Without evidence of the gravity of the hurts experienced by the victim to enable a calculation to be made of what proportionate sentence should fairly be imposed upon the offender, injustice will potentially prevail. Injustice will also prevail if in the absence of evidence of the nature and gravity of the impact the motivating sentiments of the offender are simply used as a proxy for the victim impact given that the variation in reported post-victimization distress among victims demonstrated by the analysis in this paper indicates that the motivating impulses of the offender serves as an unreliable proxy for the harmful outcomes inflicted. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 15 There are two potential solutions to this problem, but neither is satisfactory. First, it might be argued that to avoid injustice for offenders, discretion and flexibility in sentencing is necessary, with penalty enhancement applied only on a case-by-case basis where it is demonstrated that the impact of the hate crime is greater than would be the case in a parallel crime. The problem with this approach, however, is that it could compound the injustice already experienced by victims by further traumatizing them through the requirement to relive and recount their felt consequences of victimisation in the pursuit of justice for the offender. Second, it might be argued that a hate crime not only hurts the individual victim targeted as potentially each hate crime has many victims in respect of the vicarious, or terroristic, impacts they can have upon targeted communities (Iganski 2001), and such impacts could be used to determine sentencing. As is the case for the harmful outcomes experienced by the individual hate crime victim, it could be reasoned that for the just treatment of offenders in respect of determining their sentences, evidence of any such vicarious impacts must be clearly understood and presented to the court. This would conveniently avoid the problem of further victimisation for the primary victim who could be absolved from unfolding their own suffering before the court. However, to date, evidence about the vicarious impacts of hate crime stands little more than assertion as scholars have so far not convincingly demonstrated such impacts, apart from some small scale qualitative (Noelle 2001 & 2009) and non-generalizable research findings (Perry & Alvi 2012). There is little prospect at the moment then for reliable evidence about the vicarious impacts of hate crime to be brought before the courts, or to reliably inform the understanding of the courts to inform sentencing. Furthermore, it is likely that when such evidence might be produced reliably there is likely to be as much variation in the vicarious impacts as there is in the impacts experienced by individual victims. A way forward in adhering to the principle of retributive desert in the sentencing of hate crime offenders in light of the evidence of the disparate impacts of hate crime provided in this paper, is to step beyond the concern with the actual harmful outcomes of offenders’ actions to focus instead on the culpability of offenders for unleashing risks of harms (Alexander 2011) irrespective of the harmful outcomes that actually occur. Given that the evidence that hate crime as a category of crime inflicts greater harm than otherwise motivated crime is now well established, the culpability of the offender HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 16 can be the primary consideration in deciding sentencing outcomes, not the actual harmful outcomes inflicted. To illustrate this, four different offender types might be distinguished with two different sentencing outcomes. The first type, the successful instrumental bigot, is fully cognizant of the risks of their actions and intends to inflict particular hurts which are in actuality suffered by the victim. The second type, the unsuccessful committed bigot, is also fully cognizant of the risks of their actions and also intends to inflict particular hurts. However, in this case the victim has the resilience or some other good fortune not to suffer the intended hurts. The disposal of both of these offenders should be exactly the same as it would be unfair for the second offender to benefit from the good fortune of targeting a resilient victim when they fully intended to inflict hurts upon them. In both cases the same enhanced penalty proportionate to the risks unleashed, but not necessarily inflicted, would be just in terms of the principle of retributive desert. It is likely, however, that such offenders will be in a small minority as most offenders will not be committed bigots or fully aware of the depth of risk of their actions given that most people will not be inclined to consult the scholarly evidence of the type cited in this paper of the particular hurts inflicted by acts of hate crime. There is a third type of offender then, the unfortunate expressive bigot, who intends to inflict some hurt, perhaps in an emotional outburst in the heat of the moment, or perhaps to have a laugh at the other person’s expense, or perhaps to go along and get along with friends, but who is not aware of the full depth of hurt they inflict upon their victim. Rather than an enhanced penalty, and in keeping with the growing repertoire of sentencing practices (Tonry, 2006), a rehabilitative (Iganski et al., 2011) or some other form of therapeutic intervention aimed at helping the offender manage their emotions in situations of heightened stress and to also begin to address the personal and social contexts for their offending will be more appropriate and just. As too will helping the offender to appreciate the full impact of their actions, perhaps involving reparation as well as retribution with a restorative justice component (Shenk, 2001; Walters & Hoyle, 2010; Gavrielides, 2012; Walters, 2014) or some other form of community mediation (Iganski et al., 2014), would be more appropriate, and more just, for the enhanced element of the sentence over and above the sentence for the predicate offense. The same would apply to a fourth type of offender, the fortunate expressive bigot, who also intends to inflict some hurt, but is not aware of the full depth of hurt they could have inflicted upon their victim had HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 17 they been less resilient. An intervention with such an offender which seeks to raise their empathy for the victim, so that they might think twice before acting the same way again, would be more just in light of their lesser culpability than the infliction of added deprivation. Conclusion The evidence is now well established that hate crimes as a category of offence are more serious than similar but otherwise motivated crimes in respect of the greater post-victimization distress experienced by victims. However, the research evidence to date has obscured the diversity of impacts experienced by victims. This paper has taken the evidence in a new direction by presenting data to illuminate the variation in reported victim impacts. Such variation raises potential problems, however, for retributive desert justifications of enhanced penalties for hate crime offenders. If penalties are to be fairly proportionate to the harms inflicted they would need to be decided on a case-by-case basis with the evidence of the harms inflicted brought before the courts. Such an approach would mitigate against any blanket sentence enhancement in the case of hate crime over and above the predicate offense. The culpability of the offender, however, for unleashing the risk of harm irrespective of the actual harm that occurs, can fairly serve to inform sentence enhancement. Irrespective of the outcome of their actions, different offenders will have different awareness of the full depth of harm they might inflict and different intent to inflict such harms. Flexibility in sentencing, with rehabilitative and reparative dimensions for those with lesser culpability is therefore appropriate. References Alexander, L. (2011). ‘Culpability’, in J. Deigh and D. Dolinko (eds) The Oxford Handbook of Philosophy of Criminal Law, Oxford: Oxford University Press. Barnes, A. and Ephross, P.H. (1994). The impact of hate violence on victims - emotional and behavioural responses to attacks. Social Work, 39(3), 247-251. Botcherby, S. Glen, F. Iganski, P. Jochelson, K. & Lagou, S. (2011). Equality groups’ perceptions and experience of crime. Analysis of the British Crime Survey 2007-08, 2008-09 and 2009-10. Briefing Paper 4, London, UK: Equality and Human Rights Commission. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 18 Bruce, T. (2001). The New Thought Police. New York, NY: Three Rivers Press. Coleman, N., Sykes, W. & Walker, A. (2013). Crime and Disabled People. Baselinestatistical analysis of measures for the formal legal inquiry into disability-related harassment. Research Report 90: Manchester, UK: Equality and Human Rights Commission. Craig-Henderson & Sloan, L.R. (2003). After the hate: helping psychologists help victims of racist hate crime. Clinical Psychology: Science and Practice, 10(4), 481-490. Dzelme, I. (2008). Psychological Effects of Hate Crime. Riga, Latvia: Latvian Centre for Human Rights. Ehrlich, H. (1992). The ecology of anti-gay violence. In G. M. Herek & K. T. Berrill (Eds.), Hate Crimes. Confronting Violence Against Lesbians and Gay Men (105-112). Newbury Park, CA: Sage. Gavrielides, T. (2012) Contextualizing restorative justice for hate crime. Journal of Interpersonal Violence, Vol. 27, No. 18: pp. 3624-3643. Gey, S.G. (1997). What if Wisconsin v. Mitchell had involved Martin Luther King Jnr? The constitutional flaws of hate crime enhancement statutes. George Washington Law Review, 65, 1014-70. Garofalo, J. (1997). Hate crime victimization in the United States. In R.C. Davis, A.J. Lurigio, & W.G. Skogan (Eds,). Victims of Crime (2 nd ed., pp. 134-145). Thousand Oaks, CA: Sage. Hershberger, S.L., and D’Augelli, A.R. (1995). The impact of victimization on the mental health and suicidality of lesbian, gay, and bisexual youth. Developmental Psychology, 31, 65-74. Herek, G.M., Gillis, J.R. and Cogan, J.C. &Glunt, E.K. (1997). Psychological sequelae of hate crime victimization among lesbian, gay, and bisexual adults: Prevalence, psychological correlates, and methodological issues. Journal of Interpersonal Violence, 12(2), 195-215. Hurd, H. (2001). Why Liberals should hate ‘hate crime legislation’. Law and Philosophy, 20, 215-32. Hurd, H.M. Moore, M.S. (2004). Punishing hatred and prejudice. Stanford Law Review, 56(5), 1081- 1146. Iganski, P. (2001). Hate crimes hurt more. American Behavioral Scientist, 45(4), 626-638. Iganski, P. (2008). Hate Crime and the City. Bristol, UK: Policy Press. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 19 Iganski, P., Smith, D., Dixon, L., Keilinger, V., Mason, G., McDevitt, J., Stelman, A., Bargen,J., Lagou, S. and Pfeffer, R. (2011) Rehabilitation of Hate Crime Offenders. Glasgow, UK: Equality and Human Rights Commission. Iganski, P., Ainsworth, K., Geraghty, L., Lagou, S. and Patel, N. (2014) Understanding how ‘hate’ hurts: A case study of working with offenders and potential offenders, in N. Chakraborti & J. Garland (eds.) Responding to hate crime: the case for connecting policy and research, Bristol, UK: Policy Press. Iganski, P. and Lagou, S. (2014). The personal injuries of hate crime, in Hall, N., Corb, A., Giannasi, P. and Grieve, J. (eds.) The Routledge International Handbook on Hate Crime, (pp. 34-46) London: Routledge. Jacobs, J. (2013). The liberal polity, criminal sanction, and civil society, Criminal Justice Ethics, Vol. 32, No. 3: 231-246. Lawrence, F. M. (1999). Punishing Hate. Bias Crimes under American Law. Cambridge, MA: Harvard University Press. Lawrence, F. M. (2006). The hate crime project and its limitations: evaluating the societal gains and risk in bias crime law enforcement. Working Paper no. 216, Washington, DC: The George Washington University Law School. McDevitt, J., Balboni, J., Garcia, L. and Gu, J. (2001). Consequences for victims: a comparison of bias and non-bias motivated assaults. American Behavioral Scientist, 45(4), 697-713. Meyer, D. (2010). Evaluating the severity of hate-motivated violence: Intersectional differences among LGBT hate crime victims. Sociology, 44(5), 980-995. Nocon, A., Iganski, P. and Lagou, S. (2011). Disabled people’s experiences and concerns about crime, Briefing Paper 4. Manchester, UK: Equality and Human Rights Commission. Noelle, M. (2001). 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. Noelle, M. (2009). The psychological and social effects of antibisexual, antigay, and antilesbian violence and harassment. In P. Iganski (Ed.), Hate Crimes. The Consequences of Hate Crimes (pp. 73-105), Westport, CT: Praeger. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 20 Otis, M.D. and Skinner, W.F. (1996). The prevalence of victimization and its effect on mental well- being among lesbian and gay people. Journal of Homosexuality, 30, 93-122. Perry, B. and Alvi, S. (2012). ‘We are all vulnerable’: The in terrorem effects of hate crimes. International Review of Victimology, 57-71. Shenk, A.H. (2001) ‘Victim-offender mediation: the road to repairing hate crime injustice’, Ohio St. J. Disp. Resol. 185, pp. 185-217. Smith, K., Lader, D., Hoare, J. and Lau, I. (2012). Hate crime, cyber security and the experience of crime among children: Findings from the 2010/11 British Crime Survey. London, UK: Home Office. Tonry, M. (2006) Purposes and functions of sentencing. Criminal Justice Ethics, 34: 1-53. Victim Support (2006). Crime and prejudice. The support needs of victims of hate crime: a research report. London, UK: Victim Support. Von Hirsch, A. (1998) Proportionate sentences: a desert perspective, in A. von Hirsch and A. Ashworth (eds.) Principled Sentencing. Readings on Theory and Policy. (2 nd edition) Oxford: Hart. Walters, M. and Hoyle, C. (2010) Healing harms and engendering tolerance: the promise of restorative justice for hate crime. in N. Chakraborti and J. Garland (eds.) Hate Crime: Concepts, Policy, Future Directions, Cullompton: Willan. Walters, M. (2014) Hate Crime and Restorative Justice, Oxford: Oxford University Press. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 21 Table 1 Proportions of incidents of crime perceived to be racially motivated, by ethnic group Row percentages England & Wales, adults aged 16 and over % of incidents Unweighted base ALL INCIDENTS 2.3 35521 Ethnic group White 1.2 32450 Mixed *5.8 393 Asian or Asian British *14.3 1445 Black or Black British *6.1 785 Chinese or other *12.7 448 All minority ethnic groups *11.1 3071 Source: Crime Survey of England and Wales, 2009-10, 2010-11, 2011-12 Variables: racemot (recoded to exclude ‘don’t know’/refused), ethgrp2 *p ˂ 0.05 HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 22 Table 2 Indicators of perceived racial motivation of crimes Column percentages England and Wales, adults aged 16 and over % of incidents ALL INCIDENTS BELIEVED TO BE RACIALLY MOTIVATED Because of victim’s race/country of origin 49.4 Racist language used 49.0 Because of offender’s race/country of origin 18.1 Because it has happened before 18.1 Because some people pick on minorities 9.5 Because offence only committed against minorities 8.3 Other 11.1 Don’t Know 0.4 Unweighted base 441 Source: Crime Survey of England and Wales, 2009-10, 2010-11, 2011-12 Variables: yracemoa – yracemoi, racemot (recoded to exclude ‘don’t know’/refused) HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 23 Table 3. Characteristics of victims of racially motivated crimes compared with victims of otherwise motivated crimes Column percentages England & Wales, adults aged 16 and over Victims of otherwise motivated crime Victims of racially motivated crime Sex Male 51.1 *57.9 Female 48.9 *42.1 Unweighted base 27159 398 Age 16-29 32.8 *39.1 30-59 53.1 54.6 60+ 14.2 *6.2 Unweighted base 27159 398 Ethnic group White 88.7 *43.9 Non-white 11.3 *56.1 Unweighted base 27112 396 Total household income Less than £20.000 24.3 *35.9 £20.000-£39.999 24.3 22.8 £40.000 or more 28.8 *16.9 Not stated/not enough info 22.6 24.4 Unweighted base 27125 397 HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 24 Respondent’s employment status Employed 67.4 *61.6 Unemployed 4.8 *9.4 Economically inactive 27.8 29.0 Unweighted base 27069 396 Source: Crime Survey of England and Wales, 2009-10, 2010-11, 2011-12 Variables: sex, agegrp, ethgrp3, tothhin3 (recoded), remploy, racemot (recoded to exclude ‘Don’t know’/refused) *p ˂ 0.05 HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 25 Table 4. Characteristics of incidents of racially motivated crime compared with incidents of otherwise motivated crime Column percentages England & Wales, adults aged 16 and over Otherwise motivated incidents Racially motivated incidents Offence type Burglary/theft 41.7 *7.6 Criminal damage 19.5 15.7 Violence with injury 9.9 11.7 Violence without injury 8.7 *25.1 Other theft/attempted theft from the person 5.2 *0.9 Threats 15.0 *38.9 Unweighted base 35147 441 Summary offence type Property crimes 61.2 *23.3 Personal crimes of violence and theft/attempted theft from the person 23.7 *37.8 Threats 15.0 *38.9 Unweighted base 35147 441 Incident type Series 34.8 *50.2 Single 65.2 *49.8 Unweighted base 35147 441 One incident or two or more separate incidents for the same victim HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 26 1 55.3 *85.7 2 or more 44.7 *14.3 Unweighted base 35147 441 Emotional reactions Yes 85.1 *91.8 No 14.9 *8.2 Unweighted base 35140 441 How much affected Very much 19.4 *39.0 Quite a lot 30.4 33.5 A little 50.2 *27.5 Unweighted base 29871 407 Types of emotional reactions Externalized 44.9 *23.7 Internalized 10.8 *18.5 Syndrome of externalized & internalized 29.3 *49.6 No reaction 15.0 *8.2 Unweighted base 34845 440 Source: Crime Survey of England and Wales, 2009-10, 2010-11, 2011-12 Variables: offence (recoded), pincid, emotreac (recoded to exclude ‘Don’t know’/refused), howaff1 (recoded to exclude ‘Don’t know’/refused), NMatch (Constructed variable that aggregates discrete incidents for each victim for each category of racemot Yes/No), (recoded to exclude ‘Don’t know’/refused), whemota-whemoti (recoded to create the types of emotional reactions groups) *p ˂ 0.05 HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 27 Table 5. Types of emotional reactions Sex Age Ethnic group Total household income Respondent’s employment status Crime type Single or series One incident or two or more separate incidents for the same victim Externalised Male 16-29 White Property crime Single One Internalised Female 30-59 Personal crime, threats 2 or more Both externalised and internalised Female Non- white Less than £20.000 Series HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 28 Table 6. Explanatory factors associated with the reporting of externalised reactions to incidents of crime believed by victims to have been racially motivated Dependent variable: Incidents for which respondents reported experiencing anger or annoyance (or both) following the crime. Variables B-coefficient Standard error p-value Odds ratio Confidence interval Constant -3.79 0.45 0.000 0.02 Sex Female Male 1.13 0.20 0.000 3.10 2.08 - 4.63 Age 16-29 0.80 0.20 0.000 2.23 1.50 - 3.33 30-59 0.000 60+ 0.01 0.38 0.973 1.01 0.48 - 2.12 Ethnic group White 0.55 0.18 0.003 1.73 1.21 - 2.46 Non-White Total household income Less than £20,000 0.591 £20,000 - £39.999 0.28 0.25 0.271 1.32 0.80 - 2.18 £40,000 or more 0.31 0.27 0.239 1.37 0.81 - 2.30 HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 29 Not provided 0.21 0.24 0.388 1.23 0.77 - 1.98 Respondent’s employment status Employed 0.113 Unemployed -0.25 0.31 0.434 0.78 0.42 - 1.45 Economically active 0.38 0.22 0.087 1.47 0.95 - 2.28 Crime type Personal 0.002 Threats 0.12 0.21 0.575 1.13 0.74 - 1.71 Property 0.86 0.26 0.001 2.36 1.41 - 3.94 Single or series Single 0.49 0.18 0.008 1.62 1.14 - 2.32 Series One incident or two or more separate incidents for the same victim 1 0.60 0.29 0.37 1.83 1.04 - 3.21 2 or more Unweighted base 438 Categories in italics were those used as reference categories. Where categories of variables are statistically significant at the 95% level (p ˂ 0.05) values are highlighted in bold. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 30 Table 7. Explanatory factors associated with the reporting of internalised reactions to incidents of crime believed by victims to have been racially motivated Dependent variable: Incidents for which respondents reported experiencing any emotional reactions of (and any combination of) anxiety/panic attacks, crying/tears, depression, difficulty sleeping, fear, loss of confidence/feeling vulnerable or shock, following the crime. Variables B-coefficient Standard error p-value Odds ratio Confidence interval Constant -3.39 0.44 0.000 0.03 Sex Female 0.80 0.21 0.000 2.22 1.48 - 3.32 Male Age 16-29 0.000 30-59 1.14 0.23 0.000 3.13 1.98 - 4.95 60+ 0.37 0.51 0.472 1.44 0.53 - 3.92 Ethnic group White Non-White 0.12 0.20 0.546 1.13 0.76 - 1.66 Total household income Less than £20,000 0.925 £20,000 - £39.999 -0.16 0.28 0.566 0.85 0.50 - 1.46 HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 31 £40,000 or more -0.15 0.31 0.619 0.86 0.47 - 1.57 Not provided -0.02 0.27 0.945 0.98 0.58 - 1.66 Respondent’s employment status Employed 0.671 Unemployed -0.18 0.42 0.664 0.83 0.37 - 1.89 Economically active 0.16 0.24 0.512 1.17 0.73 - 1.89 Crime type Personal 0.88 0.30 0.003 2.41 1.34 - 4.33 Threats 0.95 0.28 0.001 2.58 1.50 - 4.44 Property 0.002 Single or series Single Series -0.32 0.20 0.114 0.73 0.49 - 1.08 One incident or two or more separate incidents for the same victim 1 2 or more 1.22 0.25 0.000 3.37 2.05 - 5.54 Unweighted base 438 Categories in italics were those used as reference categories. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 32 Where categories of variables are statistically significant at the 95% level (p ≤ 0.05) values are highlighted in bold. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 33 Table 8. Explanatory factors associated with the reporting of both externalised and internalised reactions to incidents of crime believed by victims to have been racially motivated Dependent variable: Incidents for which respondents reported experiencing any emotional reactions of (and any combination of) anger, annoyance, anxiety/panic attacks, crying/tears, depression, difficulty sleeping, fear, loss of confidence/feeling vulnerable or shock, following the crime. Variables B-coefficient Standard error p-value Odds ratio Confidence interval Constant -0.19 0.30 0.522 0.82 Sex Female 0.45 0.16 0.006 1.56 1.14 - 2.14 Male Age 16-29 0.315 30-59 -0.03 0.17 0.879 0.97 0.70 - 1.36 60+ -0.53 0.36 0.138 0.59 0.29 - 1.19 Ethnic group White Non-White 0.41 0.15 0.008 1.50 1.11 - 2.02 Total household income Less than £20,000 0.001 HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 34 £20,000 - £39.999 -0.86 0.22 0.000 0.42 0.28 - 0.65 £40,000 or more -0.09 0.22 0.694 0.92 0.59 - 1.42 Not provided -0.19 0.20 0.359 0.83 0.56 - 1.24 Respondent’s employment status Employed 0.032 Unemployed 0.16 0.27 0.566 1.17 0.69 - 1.98 Economically active -0.45 0.19 0.019 0.64 0.44 - 0.93 Crime type Personal 0.03 0.22 0.874 1.04 0.67 - 1.59 Threats -0.16 0.21 0.432 0.85 0.57 - 1.27 Property 0.490 Single or series Single Series 0.50 0.15 0.001 1.65 1.22 - 2.22 One incident or two or more separate incidents for the same victim 1 2 or more -0.23 0.22 0.298 0.79 0.51 - 1.23 Unweighted base 438 Categories in italics were those used as reference categories. HATE CRIMES HURT SOME MORE THAN OTHERS DOI: 10.1177/0886260514548584 35 Where categories of variables are statistically significant at the 95% level (p ˂ 0.05) values are highlighted in bold.