One rule for the goose, one for the gander? Wrongfulness and harmfulness in determining reactions to offenders and victims of crime.

People's reactions to offenders and victims of crime follow different rationales. Whereas the punishment of the offender is primarily determined by the severity of the crime (which includes its foreseeable harmful consequences), the actual harm that is experienced by the victim drives the need for his or her support and assistance. With the introduction of the Victim Impact Statement (VIS), in which victims are allowed to express the (harmful) consequences of the crime on their lives, the question is raised whether allowing such victim input during criminal proceedings would influence the offender's sentence. The main goal of the current research is to disentangle how a crime's wrongfulness and harmfulness influence people's reactions to offenders and victims. We show that, whereas people's perceptions of the offender (and the outcome of the trial) are influenced by the severity of the crime, people's judgements related to the victim are more likely to be influenced by an interaction between the severity of the crime and the experienced harm of the crime. That is, in this study no support was found for the argument that the delivery of a VIS would lead to a violation of the proportionality principle.

Previous research has shown that decisions about the punishment of an offender are based upon the severity of the crime rather than the extent of the harm that is experienced by the victim. However, ever since the introduction of the Victim Impact Statement (VIS), in which victims are allowed to express the harmful consequences of the crime as a part of the court proceedings, there has been debate about whether or not this victim ‘instrument’ introduces an irrelevant issue into the sentencing of the offender. Proponents argue that any weight given to the experience of the victim in terms of ‘harm’ would lead to disproportionate sentences and thus to a violation of the proportionality principle. In this study, we examine this issue by disentangling which factors determine people’s reactions to offenders and victims of crime. We predict that, whereas judgements relating to the offender (and his or her punishment) co-vary with the severity of the crime, people’s judgements related to the victim co-vary with both the severity of the crime and the experienced harm.

Punishment of offenders: Empirical desert and the severity of crime

Theories concerning the punishment of perpetrators of crime have been widely debated for centuries. Although they come in many shapes and sizes (for an overview: Tonry, 2011), the key distinction is between the family of deontological theories, most often associated with Immanuel Kant and G.F. Hegel and the utilitarian theories, the heritage of which can be traced to Jeremy Bentham. Whereas utilitarian theories consider the consequences of punishment to be its key motivation, deontological theories view punishment as a necessity, an imperative in the light of the wrongdoing that preceded it. The utilitarian wonders whether punishment will deter the offender from committing crime in the future, or whether the existence of punishment prevents others from doing the same. For the deontologist, these consequences are irrelevant, and in particular in the latter case immoral, because they entail using the offender as a means to the end of crime prevention. What matters is the wrongfulness of the deed (that is, the actus reus) and the motive and mind-state of the wrongdoer (that is, the mens rea). These different rationales also translate into different dimensions upon which the question of the magnitude of punishment is to be decided. For utilitarians, this is proportional to the punishment’s forward-looking, deterrent and preventative effects. For deontologists, it is proportional to the wrongfulness of the crime.

For a long time the debate has been the sole province of ethics and moral philosophy. More recently, investigators have become interested in providing empirical input: how do people actually make decisions about punishment? In particular, the work of Robinson and Darley, and their colleagues, has sought to provide answers to this question. An increasing body of evidence suggests that, even though respondents might explicitly allude to utilitarian theories (Carlsmith et al., 2002), their actual decisions and theories in use are guided by deontological notions of desert (Darley, 2009; Robinson and Darley, 1997, 2007). Importantly, this ‘empirical desert’ (Robinson and Darley, 1997) has been shown to be based upon the severity of the crime in question – its wrongfulness – rather than the extent of the harm the behaviour causes (Alter et al., 2007). Sorting out the wrongfulness and harmfulness of crime can be a tricky matter, given that much of the reason crime is wrong (‘mala in se’) lies in the fact that it is harmful – either in a manifest manner (the damage to property or personal harm) and/or in the symbolic threat it poses to core values of our society (Duff, 2001). The crime of rape is wrong to a large degree because of (the presumption of) the harm it causes, and the extent to which a given crime is more severe (that is, has a higher degree of wrongfulness) is correlated with the harm it is thought to cause. Nevertheless, wrongfulness and harmfulness can be disentangled. An illustration is the intuition that a higher level of punishment is warranted for attempted murder (high level of wrongfulness, lower level of harmfulness) than for negligent homicide (lower level of wrongfulness, higher level of harmfulness), which is indeed also the rule in criminal codes in most, if not all, jurisdictions (for example, Alter et al., 2007).

The importance of experienced harm to victims of crime

Whereas people’s reactions to offenders follow the logic of empirical desert and are guided by the severity of the crime committed, assistance and support to help victims recover from the crime’s impact are guided by the harm that actually emerges. Pemberton (2014) concludes that whereas the offender’s actions and the proportionate reaction to these actions should be measured along the dimension of wrongfulness (‘the yardstick of wrong’), the reaction to the victim’s predicament should take the experienced harmfulness as the point of departure (‘the yardstick of harm’).

Crime’s wrongfulness is in part determined by the harmfulness it usually and/or foreseeably causes, but research in victimology reveals that the experienced consequences of similarly severe crimes display large variation from one case to the next (for example, Winkel, 2007). Although there is a clear correlation between the harm victims experience and the severity of the offence, pre-existing, co-occurring and aftermath factors all influence the experience of victims of crime (Ozer et al., 2003). These idiosyncratic experiences of victims of crime should, as a matter of course, be mirrored in the way they are treated, with emerging consequences determining the necessity of psychological and/or medical treatment (NICE, 2005). A key task of psychological professionals in the aftermath of mass victimization events is to match support and treatment to the extent of harm experienced by the victim.

This is not to say that third-party support and assistance will necessarily track victims’ experience of harm. The research surrounding the justice motive (Hafer and Bègue, 2005; Lerner, 1980) has repeatedly demonstrated that victims’ suffering can give rise to negative derogating and blaming reactions, where paradoxically the extent of enduring suffering can lead to a lack of support and an increase in negative reactions (Aguiar et al., 2008; Correia and Vala, 2003). In addition, recent research has revealed the importance of people’s pre-existing expectations of victims’ reactions: A ‘mismatch’ between the observers’ expectations of the victim’s suffering and a victim’s emotional demeanour negatively influences the observers’ attitude towards the victim (Lens et al., 2014). Where as a rule the (non-verbal) expression of emotions has positive effects on people’s perceptions of the victim’s credibility (also called the ‘Emotional Victim Effect’; Ask and Landström, 2010; Winkel and Koppelaar, 1991), this is constrained by the extent to which the emotional reaction is viewed as a proportional response to the severity of the experienced crime. Lens et al. (2014) showed that, in comparison with victims of severe crimes, victims of less severe crimes who deliver a highly emotional VIS run the risk of being derogated and of being treated with less sympathy.

Mixing rationales? Victim Impact Statements

The reactions to offenders and victims in the aftermath of crime follow different rationales. Whereas the severity of the crime (which includes its foreseeable harmful consequences) determines the punishment of the offender, the actual harm that results drives the need for support and assistance of the victim. The latter is correlated to, but not determined by, the severity of the crime. In addition, people’s reactions are contingent on their expectations of a reasonable response to a crime of a given severity.

Where the responses to offenders and victims are delivered separately, by different agencies and procedures, this difference in rationales can hardly be viewed as problematic. However, this changes when the actual experience of harm (that is, victim-oriented rationale) is used to determine the sentence of the offender, beyond what the consequences the offender should have foreseen.

This issue has become more acute following the introduction of the so-called Victim Impact Statement (VIS) in various jurisdictions. Although the precise form of a VIS can vary from a written statement, which primarily serves a function in awarding compensation, to an oral statement, which may influence the sentence given to the offender (also referred to as a Victim Statement of Opinion), all have in common that they allow victims the right to express the harm they have experienced as a part of the court proceedings (for example, Erez, 2004; Lens, 2014; Lens et al., 2013, 2015; Roberts, 2008). The use of the VIS has proliferated across all common law jurisdictions, as well as in some civil law countries such as The Netherlands (for example, Roberts, 2008).

As a general rule, the academic community is highly sceptical, if not outright opposed, to the influence of a VIS on the offender’s sentence (Ashworth, 1993; Bandes, 1996; Roberts, 2009; Sarat, 1997). The main argument against this practice is that it is seen as an attempt to introduce an irrelevant issue into the sentencing of offenders. Beyond the harm offenders could have foreseen through their actions, it is unclear what bearing the idiosyncratic experience of victims of crime and their opinion on offender’s wrongdoing should have on the sentence (Pemberton, 2014). Any weight given to these matters in the determination of punishment would then lead to disproportionate sentences, co-varying with factors irrelevant to the wrongfulness of the offender’s actions.

(Academic) proponents of the VIS (for example, Roberts and Erez, 2004; Pemberton and Reynaers, 2011) do not dispute this. Instead they argue, first, that there is as yet no evidence that a VIS leads to disproportionate and/or more severe sentences (Erez and Rogers, 1999; Roberts, 2009); and, second, that the benefits for victims lie elsewhere. Rather than the influence upon the sentence (Lens, 2010, 2015; Roberts and Erez, 2004), improvements in their experienced treatment, perception of voice and/or control are key (Lens, 2014; Lens et al., 2015).

The present research

Whereas those opposed to the VIS argue that its use will lead to the application of the victim’s idiosyncratic experience of harm to a subject to which this is, or should be, irrelevant (the sentence of the offender), proponents suggest that a VIS will guide the treatment and experience of the victim but not have an impact on the offender’s sentence.

The current research intends to shed light upon this matter.[1] The main hypothesis driving the research is that judgements relating to the offender (and his or her punishment) co-vary with the severity of the crime (operationalized here as the type of crime: rape versus threat), whereas people’s judgements related to the victim co-vary with the experienced harm (operationalized here as the victim’s emotional response in his or her VIS), even when they are tasked to make these judgements at the same time, as part of one procedure. One rule for the goose, and one for the gander? Given that the reactions to victims are also influenced by expectations based upon the severity of the crime, we furthermore expect an interaction effect of the severity of the crime and the experienced harm on respondents’ reactions to victims of crime. The following two hypotheses are tested in this research:

H1: The reaction to the offender co-varies with the severity of the crime, but not with the victim’s experienced harm (Hypothesis 1).

H2: The reaction to the victim co-varies with the interaction between the severity of crime and the victim’s experienced harm (Hypothesis 2).

Method

Participants and design

A total of 214 university students (68 men and 141 women, 5 missing), with ages ranging from 18 to 46 years (M = 21.62, SD = 3.06), voluntarily participated in the study, and were randomly assigned to one of four conditions defined by a 2 (crime severity: high vs. low) × 2 (experienced harm: high vs. low) factorial design.

Materials and procedure

The participants were seated at separate tables in a lecture hall. They were told to work on the experimental task quietly and individually. The participants were given written instructions informing them that they were about to read a scenario and that they subsequently had to answer a number of questions. They were assured that there were no ‘right’ or ‘wrong’ answers and that the results of the experiment would be treated confidentially. A brief background section stated that a crime had taken place, that a suspect had been arrested, that the criminal trial would take that place, and that the victim would have the opportunity to make a VIS on the consequences of the crime. The first paragraph of the VIS was the same in all scenarios: the victim stated that she had been cycling home after a night out with friends when she saw someone (the suspect) standing at the side of the road. She stated that she had recognized this person from the club she and her friends had been to, and, thinking that he was having trouble with his bicycle, she had stopped to help him.

Crime severit

In the high crime severity condition (that is, rape), the victim continued her statement by declaring that the suspect had immediately started to touch her inappropriately. After making it clear to him that she did not want this, the suspect became violent and eventually raped her. In the low crime severity condition (that is, threat), the victim continued her statement by declaring that the suspect had immediately started to swear at her and had threatened to hurt her. In this scenario, the participants were told that no physical violence had been inflicted. To reduce possible confounds, in neither of the two scenarios was any additional information about the victim, the suspect or the trial given to the participants.

Experienced harm

Experienced harm was manipulated by assigning participants to a scenario in which the victim experienced either serious consequences or no consequences of the crime, as expressed in the VIS. The content of the VIS was based on a previous study of victims’ emotional reactions to violent crimes (Lens et al., 2010, 2013). In the high experienced harm condition, participants read the following:

‘This crime has turned my whole life upside down. I can’t sleep any more, I barely eat and I constantly feel anxious. Anxious that this will happen to me again or that I will meet the offender again. I am afraid to be out on the street on my own. I constantly feel tense, restless and sad. I am not who I once was, I am no longer the enjoyable friend or the fun, spontaneous daughter…. Why did this have to happen to me? What did I do to deserve this? I am disgusted by the offender! What possessed him? I did not even know him. I am so mad. Very often at the wrong people. Then I have to take it out on the person standing closest to me. It is just so unfair, I have become a completely different person.’

In the low experienced harm condition, participants read the same text, yet with the victim stating that she had not experienced any consequences at all (beginning the VIS with ‘This crime didn’t change my life at all’). Having read the scenario, the participants were asked to digest the scenario for a while and subsequently turn the page to answer a number of questions.

Dependent measures

Perceptions and judgements of the victim

Expectancy violation

In line with previous work (Ask and Landström, 2010; Lens et al., 2014), the participants were asked to assess to what extent the consequences matched the consequences they would expect from a rape/threat victim (1 = did not match at all, 7 = matched completely).

Veracity judgements

In line with previous work (Ask and Landström, 2010; Lens et al., 2014), the participants were asked to make a dichotomous veracity judgement, indicating whether or not they believed that the victim had suffered the consequences as indicated in the VIS, and to assign a confidence rating to the certainty of the expressed consequences (1 = absolutely unsure, 7 = absolutely sure).

Attitude towards the victim

Two items regarding the observers’ attitude towards the victim were adapted from a study by Aguiar and colleagues (2008). As a measure of victim derogation, the participants were asked to indicate their general impression of the victim on a seven-point scale, ranging from 1 (extremely negative) to 7 (extremely positive) (with the lower range of scores indicating victim derogation). Furthermore, they were asked to indicate their feelings of sympathy for the victim on a seven-point scale, ranging from 1 (no sympathy at all) to 7 (very strong sympathy).

Blameworthiness of the victim

As a measure of victim blaming, the participants were asked to rate the extent to which they found the victim was to blame (1 = no blame at all, 7 = full blame). Furthermore, participants had to indicate the extent to which they thought that the victim could have prevented the crime from occurring (1 = absolutely not, 9 = absolutely).

Victim Impact Statement

Participants were asked to make two dichotomous judgements, indicating (1) whether or not they thought that the described consequences of the crime (that is, experienced harm) should be taken into account when determining the punishment of the offender (yes or no), and (2) whether or not they thought that the described consequences, as stated by the victim in the VIS (that is, experienced harm), should have an influence in the criminal justice proceedings (yes or no). For both judgements, participants were also asked to give ratings on 9-point scales, ranging from 1 (absolutely not) to 9 (absolutely). Finally, participants were asked to rate the extent to which they thought the VIS is a good instrument to ‘measure’ the consequences of the crime for the victim (1 = absolutely not, 9 = absolutely).

Perceptions and judgements of the offender

Blameworthiness of the offender

Participants were asked to rate the extent to which they found the offender was to blame on a seven-point scale (1 = no blame at all, 7 = full blame).

Punishment of the offender

Participants were asked to rate the extent to which they thought the punishment of the offender should be severe (1 = not severe at all, 7 = extremely severe).

Results

Manipulation checks[2]

As a check of the effectiveness of the ‘experienced harm’ manipulation, participants were asked to rate the extent to which the victim indicated in her VIS that the consequences of the crime were either negative or positive, ranging from 1 (completely negative) to 7 (completely positive). An independent-samples t-test showed strong significant differences between groups (t(167.19) = −21.23, p < .001). As expected, participants in the high experienced harm condition indicated that the consequences of the crime for the victim were (completely) negative (M = 1.46, SD = 1.04), whereas participants in the low experienced harm condition scored exactly in the centre of the scale (M = 4.09, SD = 0.70), indicating that the victim suffered no consequences at all. Hence, the ‘experienced harm’ manipulation was successful.

As a check of the effectiveness of the ‘crime severity’ manipulation, participants were asked to rate the extent to which they found that the specific type of crime (that is, rape/threat) should be punished in general, ranging from 1 (no punishment at all) to 7 (extremely harsh punishment). An independent-samples t-test showed significant differences between the high crime severity condition (M = 6.24, SD = 0.84) and the low crime severity condition (M = 4.51, SD = 1.17; t(199.04) = −12.38, p < .001). As expected, participants judge rape to be a very severe crime, whereas threat was judged to be mildly severe. Hence, the manipulation of ‘crime severity’ was successful.

Perceptions and judgements of the victim[3]

All means and standard deviations for the below stated variables are displayed in Table 1.

Table 1.

Means (and standard deviations) of perceptions and judgements of the victim as a function of condition.

Perceptions and judgements of the victim	Condition
	High crime severity		Low crime severity
	Experienced harm high (n = 49)	Experienced harm low (n = 53)	Experienced harm high (n = 50)	Experienced harm low (n = 62)
	M (SD)	M (SD)	M (SD)	M (SD)
Expectancy violation	5.55 (1.02)a	2.52 (1.31)b	3.73 (1.40)c	3.50 (1.23)c
Veracity judgement	3.55 (0.80)a	2.40 (0.87)b	3.00 (0.81)c	2.56 (0.84)b
Victim derogation	3.63 (1.25)a	5.21 (1.30)b	3.04 (.97)c	4.60 (1.23)d
Sympathy	5.41 (1.17)a	4.79 (1.30)b	3.80 (1.33)c	4.02 (1.29)cd
Victim blaming	1.50 (0.90)a	2.27 (1.54)b	1.82 (1.21)ab	1.84 (1.30)ab
Victim prevention	3.69 (2.02)a	3.04 (1.91)a	3.04 (2.18)a	2.84 (1.82)a
VIS punishment	5.61 (2.28)a	3.98 (2.08)b	5.40 (1.71)a	4.26 (1.84)b
VIS influence	4.57 (2.21)a	3.98 (2.07)a	4.33 (1.94)a	4.20 (1.79)a
VIS measurement	5.59 (1.98)a	3.88 (1.91)b	4.67 (1.76)c	4.65 (1.71)c

Note: Means with a different subscript differ significantly with all p’s < .05.

Expectancy violation

An ANOVA on expectancy violation (reverse scored) revealed a Crime Severity × Experienced Harm interaction, F(1, 208) = 66.31, p < .001, ηp2 = .24, a main effect of Experienced Harm, F(1, 208) = 90.44, p < .001, ηp2 = .30, and a main effect of Crime Severity F(1, 208) = 5.92, p = .016, ηp2 = .03. Participants in the high crime severity – high experienced harm condition were least likely to experience expectancy violation, whereas participants in the high crime severity – low experienced harm condition were most likely to experience expectancy violation. The mean scores in both low crime severity conditions were more subtle, indicating that participants in these conditions have less strong expectations about a victim’s reaction after experiencing threat.

Veracity judgements

A Pearson Chi-square test showed significant differences in veracity judgements of the suffered consequences, as indicated in the VIS, χ2 (3, N = 197) = 72.15, p < .001. Participants in the high crime severity – high experienced harm condition scored highest on this measure (96 percent), followed by the low crime severity – high experienced harm condition (64 percent), the low crime severity – low experienced harm condition (24 percent), and the high crime severity – low experienced harm condition (22 percent). Furthermore, results from an ANOVA on confidence ratings revealed similar effects, as the Crime Severity × Experienced Harm interaction, F(1, 209) = 9.71, p = .002, ηp2 = .04, and the main effect of Experienced Harm, F(1, 209) = 48.02, p < .001, ηp2 = .19, were significant, but the main effect of Crime Severity was non-significant, F(1, 209) = 2.92, p = .089. In short, the experienced harm as expressed by the victim is an important determinant in participants’ veracity judgements.

Attitude towards the victim

An ANOVA on victim derogation revealed a main effect of Experienced Harm, F(1, 209) = 90.14, p < .001, ηp2 = .30, a main effect of Crime Severity, F(1, 209) = 13.37, p < .001, ηp2 = .06, and no Crime Severity × Experienced Harm interaction, F(1, 209) = 0.00, p = .947. Mean scores for all groups differed significantly, all t’s > 2.58, all p’s < .05. Interestingly, whereas participants in both the high crime severity – high experienced harm and the low crime severity – high experienced harm conditions had a moderately negative general impression of the victim, participants in the high crime severity – low experienced harm and the low crime severity – low experienced harm conditions had a positive general impression of the victim.

Furthermore, an ANOVA on feelings of sympathy for the victim revealed a Crime Severity × Experienced Harm interaction, F(1, 209) = 5.68, p = .018, ηp2 = .02, a main effect of Crime Severity, F(1, 209) = 46.04, p < .001, η2 = .18, and no main effect of Experienced Harm, F(1, 209) = 1.32, p = .251. Participants in both the high crime severity – low experienced harm and the high crime severity – high experienced harm conditions displayed higher levels of sympathy for the victim than the participants in both the low crime severity – low experienced harm and the low crime severity – high experienced harm conditions, all t’s > 3.17, all p’s < .05. The difference between the high crime severity – low experienced harm and the high crime severity – high experienced harm condition was also significant, t(99) = 2.51, p = .014).

Blameworthiness of the victim

An ANOVA on victim blaming revealed a Crime Severity × Experienced Harm interaction, F(1, 209) = 5.16, p = .024, ηp2 = .02, no main effect of Crime Severity, F(1, 209) = 0.10, p = .758, and a main effect of Experienced Harm, F(1, 209) = 4.68, p = .032, ηp2 = .02. There was a significant difference in victim blaming only between the high crime severity – high experienced harm condition (M = 1.50, SD = 0.90) and the high crime severity – low experienced harm condition (M = 2.27, SD = 1.54), t(76.30) = 3.03, p = .003. Other conditions did not differ from each other, all t’s < 1.64, all p’s > .103. These low ratings indicate that victim blaming was quite unlikely for all groups of participants. Furthermore, an ANOVA on victim prevention revealed no Crime Severity × Experienced Harm interaction, F(1, 209) = 0.70, p = .405, no main effect of Crime Severity, F(1, 209) = 2.46, p = .118, and no main effect of Experienced Harm, F(1, 209) = 2.48, p = .117. Overall, participants were equally unlikely to think that the victim could have prevented the occurrence of the crime.

Victim Impact Statement

Participants in the high crime severity – low experienced harm condition (48 percent), the high crime severity – high experienced harm condition (56 percent), the low crime severity – low experienced harm condition (59 percent), and the low crime severity – high experienced harm condition (60 percent) were equally hesitant in indicating that the VIS should have an influence on the criminal justice proceedings, χ2 (3, N = 206) = 1.83, p = .608. Furthermore, results from an ANOVA on the scale measure of this item revealed similar effects, as the Crime Severity × Experienced Harm interaction, F(1, 206) = 0.67, p = .412, the main effect of Crime Severity, F(1, 206) = 0.00, p = .968, and the main effect of Experienced Harm, F(1, 206) = 1.73, p = .190, were all non-significant.

A Pearson Chi-square test showed significant differences in whether participants thought that the VIS should be taken into account when determining the punishment of the offender, χ2 (3, N = 208) = 19.94, p < .001. Participants in the low crime severity – high experienced harm condition (82 percent) scored highest on this measure, followed by the high crime severity – high experienced harm condition (75 percent), the low crime severity – low experienced harm condition (64 percent), and the high crime severity – low experienced harm condition (42 percent). Furthermore, results from an ANOVA on the scale item of this measure revealed similar effects, as the Crime Severity × Experienced Harm interaction, F(1, 206) = 0.81, p = .370, and the main effect of Crime Severity, F(1, 206) = 0.01, p = .911, were non-significant, but the main effect of Experienced Harm, F(1, 206) = 25.34, p < .001, was significant. An ANOVA on whether or not the VIS is a good instrument to ‘measure’ a victim’s consequences of the crime (that is, experienced harm) revealed a Crime Severity × Experienced Harm interaction, F(1, 208) = 10.99, p = .001, ηp2 = .05, and a main effect of Experienced Harm, F(1, 208) = 11.75, p = .001, ηp2 = .05, but not a main effect of Crime Severity, F(1, 208) = 0.10, p = .756. Participants in the high crime severity – high experienced harm condition scored highest on judging the VIS a good measure, as compared with the high crime severity – low experienced harm condition, the low crime severity – high experienced harm condition, and the low crime severity – low experienced harm condition, all t’s > 2.43, all p’s < .05. Participants in the high crime severity – low experienced harm condition scored lowest on judging the VIS a good measure, all t’s > 2.16, all p’s < .05.

Perceptions and judgements of the offender

All means and standard deviations for the below stated variables are displayed in Table 2.

Table 2.

Means (and standard deviations) of perceptions and judgements of the offender as a function of condition.

Perceptions and judgements of the offender	Condition
	High crime severity		Low crime severity
	Experienced harm high (n = 49)	Experienced harm low (n = 53)	Experienced harm high (n = 50)	Experienced harm low (n = 62)
	M (SD)	M (SD)	M (SD)	M (SD)
Blameworthiness	5.55 (1.73)a	6.37 (1.12)b	4.66 (1.41)c	5.77 (1.08)d
Punishment	5.61 (1.04)a	5.62 (.89)a	3.32 (1.17)b	3.61 (1.32)c

Note: Means with a different subscript differ significantly with all p’s < .05.

Blameworthiness of the offender

An ANOVA on offender blaming did not reveal a Crime Severity × Experienced Harm interaction, F(1, 209) = 0.66, p = .418, but did reveal a main effect of Crime Severity, F(1, 209) = 16.11, p < .001, ηp2 = .12, and a main effect of Experienced Harm, F(1, 209) = 27.27, p < .001, ηp2 = .07. Participants in the high crime severity – low experienced harm condition scored highest on blaming the offender as compared with the high crime severity – high experienced harm condition, the low crime severity – high experienced harm condition, and the low crime severity – low experienced harm condition, all t’s > 2.79, all p’s < .01. Participants in the low crime severity – high experienced harm condition blamed the offender least, all t’s > 2.80, all p’s < .01.

Punishment of the offender

An ANOVA on punishment of the offender did not reveal a Crime Severity × Experienced Harm interaction, F(1, 209) = 0.87, p = .351, and a main effect of Experienced Harm, F(1, 209) = 0.91, p = .341, but did reveal a main effect of Crime Severity, F(1, 209) = 191.61, p < .001, ηp2 = .48. Participants in both the high crime severity – low experienced harm and the high crime severity – high experienced harm conditions found that the punishment of the offender should be more severe than the participants in both the low crime severity – low experienced harm and the low crime severity – high experienced harm conditions, all t’s > 8.68, all p’s < .001.

Discussion

Overall, we found support for the first hypothesis (H1) that people’s perceptions regarding the offender (and the outcome of the trial) are more likely to be influenced by the severity of the crime than by the experienced harm of the crime. Only the severity of the crime had an influence on the punishment of the offender: participants who read about a highly severe crime think that the offender should be punished more severely than participants who read about a crime that is less severe. However, the blameworthiness of the offender was influenced by both the severity of the crime and the experienced harm of the crime: people who read about a victim experiencing no consequences from a highly severe crime scored highest on blaming the offender, whereas participants who read about a victim experiencing serious consequences from a less severe crime were least likely to blame the offender. We examined the correlations between blameworthiness and punishment of the offender per condition and found that there was a significantly positive correlation in all conditions, with the exception of the high crime severity – low experienced harm condition (which was r = .006, p = .966). We do not have a clear explanation for this finding. It might be the case that people judge a victim who does not display (emotional) consequences from a severe crime to be in need of suppressing the consequences of the crime in order to cope with the situation and therefore may even be harmed more than victims who do express the consequences. More research is needed to examine what people take into account when making judgements about the blameworthiness and punishment of the offender.

Finally, we found support for the second hypothesis (H2) that people’s judgements related to the victim are likely to be influenced by both the severity of the crime and the experienced harm. First, people who read about a victim experiencing serious consequences from a severe crime were least likely to experience expectancy violation, whereas participants who read about a victim experiencing no consequences from a severe crime were most likely to experience expectancy violation. Thus, participants who read about a severe crime had stronger expectations about the emotionality of the victim than participants who read about a mildly severe crime. In addition, people tended to have more sympathy for victims in the high crime severity conditions than for victims in the low crime severity conditions, and most sympathy was felt for victims experiencing serious consequences from a severe crime. This same pattern was found for the participants’ confidence in the suffering of the victim. Finally, and in line with previous research (for example, Lens et al., 2014), participants were quite unlikely to blame the victim for the crime or to think that the victim could have prevented the crime.

The items concerning the VIS revealed that participants who read about a victim experiencing serious consequences were more likely to think that the VIS should be taken into account when determining the punishment of the offender than participants who read about a victim not experiencing serious consequences. However, although it seems that ‘experienced harm’ is an important determinant in participants’ judgements about whether the VIS should be taken into account in the punishment of the offender, we found no differences on the more general question of whether people think that the VIS should have an influence on the criminal justice proceedings as such. Furthermore, people who read about a victim experiencing serious consequences from a severe crime scored highest on judging the VIS to be a good instrument to ‘measure’ the victim’s actual experience, whereas participants who read about a victim experiencing no consequences from a severe crime were least likely to consider the VIS a good instrument.

Scientific and practical implications

The current study extends previous research in an important regard. Previous studies that examined people’s judgements of a criminal case mainly focused either on the effects of crime severity on people’s perceptions of the offender and the decisions on punishment (for example, Carlsmith et al., 2002; Darley et al., 2000) or on the effects of a victim’s (non-verbal) emotional expression on people’s credibility judgements and perceptions of the victim (for example, Ask and Landström, 2010; Bollingmo et al., 2009; Hackett et al., 2008; Kaufmann et al., 2003; Klippenstine and Schuller, 2012; Mulder and Winkel, 1996; Wessel et al., 2006, 2012; Winkel and Koppelaar, 1991). The only study that examined possible interaction effects of crime severity and the experienced harm of the crime on people’s perceptions of different participants in the criminal trial included a question only about the appropriate punishment of the offender and not about blameworthiness (Rose et al., 2006). The results suggest that these two constructs are not judged in the same manner. Furthermore, this study did not include questions about people’s perceptions regarding the VIS. This is especially important because of the ongoing debate around the acceptance – and desirability – of victim input during the trial. The desirability of enabling victims to deliver a VIS is highly debated on conceptual as well as practical grounds (for example, Buruma, 2004; Erez, 1994; Hill, 2005; Myers and Greene, 2004; Roberts and Erez, 2004). Whereas proponents argue that the delivery of a VIS may promote accurate or effective sentencing outcomes by enabling victims to speak about the experienced consequences of the crime for their lives (for example, Erez and Rogers, 1999; Roberts and Erez, 2004), opponents argue that its delivery may lead to a violation of the proportionality principle, because differences in post-crime reactions by the victim may lead to differences in the outcome of the trial (for example, Buruma, 2004; Luginbuhl and Burkhead, 1995; Myers and Arbuthnot, 1999). This study showed that the punishment of the offender is influenced by the severity of the crime and not by the expression by the victim of the consequences (that is, experienced harm). In other words, this study found no support for the argument that enabling a VIS during a trial might lead to a violation of the proportionality principle.

Finally, the present research highlights that people (at least partially) rely on the emotions they experience as a consequence of being confronted with the emotions of a victim in decision-making. Hence, the idea that people exclusively behave ‘logically’, as often stated in economic and criminological studies, does not find support in the current study.

Limitations

Some limitations of this study should be noted. First, the sample of participants (that is, university students) lacked personal experience of criminal cases, which may limit generalization of the present findings. It could be that professionals involved in the criminal justice procedure (for example, prosecutors, judges) develop more fine-grained schemas for evaluating the severity and the experienced harm of the crime and thus rely less on stereotypical expectancies (see also Ask and Landström, 2010; Lens et al., 2014). For example, Wessel et al. (2006) showed that credibility ratings of court judges, unlike those of lay people, were not influenced by the emotions displayed by the witness. However, other studies demonstrated that even experienced professionals are susceptible to a victim’s post-crime reaction. Frohmann (1991), for example, suggested that a rape victim’s demeanour had a substantial influence on the victim’s perceived credibility in the eyes of prosecutors. Because it remains an empirical question as to whether a more ‘representative’ sample would yield similar findings, we suggest that future research should focus on other groups of participants (for example, judges, prosecutors). Second, only a written depiction of the experienced harm was provided, which allowed us to exclude possible effects from a victim’s non-verbal emotional expression. However, given that previous research has repeatedly shown effects from a victim’s non-verbal emotional expression on people’s judgements of the victim (for example, Ask and Landström, 2010; Nadler and Rose, 2003; Rose et al., 2006; Wessel et al., 2006, 2012; Winkel and Koppelaar, 1991), it is likely that similar, if not stronger, effects would emerge when participants are presented with a video presentation by the victim. Finally, the experimental conditions contained two (separate) dimensions of statement credibility – as operationalized by the 19 Criteria Based Contents Analysis approach: criteria 12 (that is, description of subjective mental states) and criteria 17 (that is, self-depreciation). These two dimensions might each have different and uncontrolled influences on people’s perceptions of the credibility of the VIS. Future studies should contain only one of these dimensions to disentangle these effects.

Conclusion

This article substantiates the finding that people’s reactions to offenders and victims of crime follow different rationales. Whereas the punishment of the offender is primarily determined by the severity of the crime (which includes its foreseeable harmful consequences), the actual harm that is experienced by the victim drives the need for his or her support and assistance. With the introduction of the Victim Impact Statement, the question was raised of whether allowing such victim input during the criminal proceedings would influence the offender’s sentence. The main goal of the current research was to disentangle how a crime’s wrongfulness and harmfulness influence people’s reactions to offenders and victims.

In conclusion, the present research has shown that, whereas perceptions regarding the offender (and the outcome of the trial) are more likely to be influenced by the severity of the crime (Hypothesis 1), people’s judgements related to the victim are more likely to be influenced by an interaction between the severity of the crime and the experienced harm of the crime (Hypothesis 2). Hence, we conclude that no support was found for the argument that enabling a VIS during a trial might lead to a violation of the proportionality principle and that the debate about whether or not to allow victims to deliver a VIS during the criminal proceedings is not as black and white as previously argued. The results of this study give a more nuanced view of which factors are more or less likely to be influenced by the delivery of a VIS.