Introduction

Victims of violent civilian trauma may suffer from a wide array of negative psychological and social adjustment problems, such as such as posttraumatic stress disorder (e.g., Orth et al. 2008), depression (e.g., Bargai et al. 2007), somatization, hostility, generalized and phobic anxiety (Norris and Kaniasty 1994; Winkel 2009), eating disorders (Brady 2008), substance abuse (Vermeiren et al. 2003), insomnia (Krakow et al. 2001), and sexual dysfunctioning (Letourneau et al. 1996).

Many European countries, such as the Netherlands (see Wet Schadefonds Geweldsmisdrijven of 1975 and the European Convention on the Compensation of Victims of Violent Crimes of 1983), and the United States (see US Victims of Crime Act of 1984) provide victims of intentional violent crimes the opportunity to apply for financial compensation for these immaterial damages if they are not covered otherwise and are not due to culpability or unlawful behavior on the side of the victim. In spite of this, only a minority of eligible victims applies for state compensation (Alvidrez et al. 2008; McCormack 1991). Furthermore, while legal scholars have theorized about the rationale of state compensation for victims of crime (e.g., Ashworth 1986; Buck 2005; Willis 1984) and investigated differences between jurisdictions (e.g., Greer 1994; Groenhuijsen 2001; Villmow 1991), little empirical knowledge is available on how satisfied victims who do apply are with amount of compensation awarded and, more importantly, which factors are associated with satisfaction ratings. This topic is important to address, since lack of satisfaction may be seen as a kind of secondary victimization (i.e., perceived additional violation of legitimate rights or entitlements by the victim; see Montada 1994; Orth 2002; Orth and Maercker 2004). Despite a dearth of knowledge, a few studies have attempted to uncover correlates of victims’ satisfaction with state compensation.

Victim Satisfaction with State Compensation: A Brief Overview of the Literature

The few studies that have been conducted on violent crime victims’ satisfaction with compensation from the state suggest that satisfaction ratings are associated with white female sex and administrative inconvenience, such quickness of claim resolution, and perceptions of adequate reward (Elias 1984; Newmark et al. 2003; Newmark 2004). The latter findings are in line with the objective of victim compensation schemes to express society’s solidarity with the victim. To fulfill this purpose, compensation should be “subjectively felt by the victim to be adequate” (Wright 1998, p. 89).

Since satisfaction represents a “(…) general evaluation or feeling of favorableness toward the object in question (…)” (Fishbein and Azjen 1975, p. 11), victims’ satisfaction levels may be distorted by affective states (see also Prince 2005). To date, one study has addressed this topic. Orth (2004) investigated whether satisfaction with level of compensation for material damages and pain and suffering (i.e., immaterial damage) was associated with feelings of revenge. Revenge is an affective state in response to harm intentionally caused by others (Collens 1998). Mean satisfaction level in this study was rather low given the scale width on which satisfaction items had to be rated. Both univariate and multivariate analyses yielded significant negative associations between satisfaction with compensation level and current feelings of revenge. Notwithstanding these results, one might speculate that the contribution of affective states to low satisfaction ratings is particularly strong if they reflect the presence of mental health problems. After all, several studies suggest that negative affect is related to increased symptoms of psychological distress (e.g. Crawford and Henry 2004). Preliminary support for this contention is provided by studies on satisfaction with services conducted in medical patient populations.

Mental Health and Satisfaction with Services in Medical Literature

Generally stated, medical research on the association between mental health and satisfaction with services is mainly cross-sectional in nature and has relied on either of two statistical approaches. One set of studies entered mental health as the dependent variable and satisfaction with services as the independent variable into statistical analyses. Many of these studies suggest that satisfaction with services is associated with adverse mental health status above and beyond background characteristics (e.g., Carlson and Gabriel 2001; Garland et al. 2003; Gigantesco et al. 2002). Reversely, a number of other studies entered mental health as the independent variable and satisfaction with medical services as the dependent variable into statistical analyses and found that mental health problems were associated negatively with satisfaction with services above and beyond background characteristics (e.g., Chen et al. 2006; Greenberg and Rosenheck 2004; Hoff et al. 1999; Rosenheck et al. 1997; Ruggeri et al. 2003; Turchik et al. 2010; for an early meta-analysis see Lehman and Zastowny 1983). In line with Switzer et al. (1999), these results may be interpreted to suggest that satisfaction serves as an important indicator of unidentified or unsolved medical or psychological problems. This notion is also relevant with respect to violent crime victims. This group often fails to seek help for psychological problems resulting from the act of victimization (Logan et al. 2005), but may nevertheless seek compensation from the state. Dissatisfaction with state compensation might thus be speculated to indicate that the victim involved suffers from underlying mental health problems unknown to or neglected by victim services.

The Current Study

Given the aforementioned, the purpose of the current study was to explore whether mental health problems and satisfaction with amount of state compensation for intentional violent crime victimization were interrelated. If so, satisfaction may not be an adequate indicator for policy makers to evaluate services provided by victim compensation schemes and should merely be conceived as an indicator of undetected mental health problems. Relying on medical literature, it was expected that mental health problems and satisfaction with compensation would be negatively correlated.

Methods

Procedure

This study used wave 2 data from a larger study into the psychosocial consequences of violent victimization (see also Kunst et al. 2010). Participants were recruited among victims who had applied for compensation with the Dutch Victim Compensation Fund (DVCF) in 2006. Eligible for participation were those who: (1) did not have missing file data on age, gender, and time since victimization (i.e., time elapse in years between victimization and study entry)Footnote 1; (2) were ≥18 years old at time of study entry; (3) indicated they had been awarded compensation from the fund, and (4) had provided informed consent to participate in the current study. Those who had agreed to participate were invited to complete an Internet survey on satisfaction with compensation, mental health status, feelings of revenge, and victim service utilization. The latter two variables were important to assess to enable adjustment for potential confounding in statistical analyses. Revenge is the only (non-pathological) affective state which has been identified as a correlate of victims’ levels of satisfaction with state compensation in previous research (see Orth 2004) and mental health (e.g., Orth et al. 2006), while history of service utilization in victims eligible for state compensation appears to be an important associate of mental health (see for example New and Berliner 2000). Participants who did not have access to the World Wide Web or preferred to fill out the questionnaire on paper could request for a hardcopy version. Remaining background information (age, gender, time since victimization, level of compensation for pain and suffering (0 = €0 and 9 = €9,100), number of claimed material damage categories, and type of violenceFootnote 2) for all victims who had applied for compensation in 2006 was retrieved from the DVCF’s electronic files. The study was approved by the DVCF Committee.

Participants

Two hundred and thirty-five (36.7%) of those who had agreed to participate responded to the request for participation. One hundred and fifty-one respondents fulfilled inclusion criteria. Victims included in the study were more often female (97/151, 64.2% vs. 1,737/3,570, 48.7%, P < .001), had been victimized earlier (mean time since victimization for participants in years was 5.66 (SD = 5.02) vs. M = 4.75 (SD = 3.53) for non-participants, P < .01), had received higher levels of compensation for pain and suffering (M = 2.6, SD = 1.8 vs. M = 2.0, SD = 1.8, P < .001), had sought compensation for more categories of material damages (M = 2.71, SD = 2.15 vs. M = 2.21, SD = 2.17, P < .001), and had experienced sexual violence more often than non-participants (29/151, 19.2% vs. 441/3,570, 12.4%, P < .025). No differences were observed for age (M = 41.2, SD = 14.5 vs. M = 39.5, SD = 15.6), severe physical assault (12/151, 7.9% vs. 403/3,570, 11.3%), moderate physical assault (643/151, 28.5% vs. 1,205/3,570, 33.8%), and theft with violence (43/151, 28.5% vs. 896/3,570, 25.1%).

Measures

Satisfaction with Compensation

Two items were used to measure satisfaction with compensation: one for satisfaction with compensation for material damages (“Are you satisfied with amount of compensation rewarded for material damages?”) and one for satisfaction with compensation for pain and suffering (“Are you satisfied with amount of compensation rewarded for pain and suffering?”). Items were derived from Orth (2004) and had to be rated on a 7-point Likert scale (0 = not at all, 6 = sufficient). Prior to statistical analyses, an average compensation score was calculated by summing item scores and dividing its sum by the number of awarded compensation types (1 or 2). Internal consistency reliability of the compensation scale among those who had received both compensation for material damages and pain and suffering (n = 71) was Cronbach’s α = .74.

Mental Health Status

The 12-item General Health Questionnaire (GHQ-12) was administered to approximate participants’ current mental health status. The GHQ-12 is an excellent screening instrument (Goldberg et al. 1997). Items had to be rated on a 4-point Likert scale (0 = not at all, 1 = same as usual, 2 = rather more than usual, and 3 = much more than usual). An example item is “Have you lost much sleep lately/recently?” A threshold of 11/12 was used to differentiate between victims with and without probable mental health problems (cf. Goldberg et al. 1997), with subjects scoring above this cut off qualifying for probable mental health problems. “Probable” does not indicate that such persons suffer from psychiatric illness, but suggests that they require further attention (Jackson 2007). Internal consistency reliability of the GHQ-12 was Cronbach’s α = .81.

Feelings of Revenge

A Dutch translation of the 3-item revenge scale developed by Orth (2003) was administered to assess feelings of revenge during the past 4 weeks. This instrument is similar to the revenge scale used by Orth (2004). An example item is “How often did you think about doing something to the perpetrator, without actually doing it?” Answers had to be rated on a 6-point Likert scale (0 = never, 5 = very often). Internal consistency reliability of the revenge scale was Cronbach’s α = .97.

Victim Service Utilization

One self-developed item was used to check whether participants had received victim support (“no”, “yes, by a Victim Support Netherlands (VSN) volunteer”, “yes, by a social worker”, “yes, by a clinical psychologist/psychiatrist”, “yes, by another aid administrator”). Those who had received help were coded as yes = 1 and no = 0.

Statistical Analyses

To describe the study sample, means and standard deviations or frequencies and percentages were computed for all background variables. Independent t tests and Chi square analyses, as appropriate, were conducted to examine differences in satisfaction with compensation and background variables between participants with and without probable mental health problems. Next, a hierarchical linear regression analysis was conducted to test whether an observed association between probable mental health problems (as independent variable) and satisfaction with compensation (as dependent variable) would still reach significance after adjusting for feelings of revenge, history of service utilization, and other background variables. In addition, hierarchical logistic regression analysis was used to see whether satisfaction with compensation (as independent variable) would remain significantly associated with probable mental health problems (as dependent variable) in a multivariate model. In both regression analyses, the independent variable of interest was entered on the first step and background variables on the second step. Prior to statistical analyses, data were checked for underlying assumptions. The alpha level was set at .05 in all statistical tests. All statistical analyses were performed using the software package SPSS 16.0 for Windows (SPSS Inc., Chicago, Illinois).

Results

Descriptive Statistics

Descriptive statistics are summarized in Table 1. One hundred and seventeen (74.5%) participants fulfilled the GHQ-12 cut off score for probable mental health problems. Subjects in this group had experienced sexual violence less often than those scoring below the cut off (18/117, 15.4% vs. 11/34, 32.4%), χ² (1, N = 151) = 4.88, P < .05. No differences on any of the remaining background variables were observed. As expected, mean level of satisfaction with compensation for participants with probable mental health problems (M = 3.8, SD = 2.0) was significantly lower than for those without such problems (M = 4.6, SD = 1.3), t(149) = 2.9, P < .01).

Table 1 Sample characteristics by mental health status
Full size table

Regression Analyses

Linear regression analysis yielded a significant negative association between probable mental health problems and satisfaction with compensation, even after adjusting for background variables. None of the background variables correlated significantly with satisfaction in this analysis (see Table 2). The final model explained slightly more than 10% of the variance in satisfaction with state compensation. Logistic regression produced similar results. Satisfaction with state compensation was significantly and negatively associated with probable mental health problems. Inspection of odds ratios suggested that the odds for probable mental health problems would decrease with almost 25% with each unit increase in satisfaction with state compensation. Sexual violence was the only background variable which was significantly and negatively associated with probable mental health problems. Odds ratios indicated that victims of sexual violence were almost twice as less likely to suffer from probable mental health problems than not compared to those with histories of non-sexual violent victimization. The final model explained between 15 and 22.8% of the pseudo variance in probable mental health problems (see Table 3).

Table 2 Linear regression model predicting the variance in satisfaction with state compensation (n = 151)
Full size table
Table 3 Logistic regression model predicting the odds of probable mental health problems (n = 151)
Full size table

Discussion

The present study explored the association between mental health problems and satisfaction with state compensation for intentional violent crime victimization. Relying on studies about the association between mental health and satisfaction with services in medical patient populations, mental health was predicted to influence satisfaction scores and vice versa. In line with expectations and previous studies (e.g., Hoff et al. 1999), results indicated that victims with high levels of self-reported mental problems reported lower levels of satisfaction with amount of compensation than those with low levels. The negative association between mental health problems and satisfaction levels remained significant in multivariate analyses and irrespective of whether satisfaction or mental health scores served as outcome variable.

Findings may be interpreted to argue that caution should be taken when estimating victims’ levels of satisfaction with amount of financial compensation by the state. Apparently, satisfaction scores do not necessarily reflect adequacy in terms of meeting victims’ needs for recognition through monetary reward and may also be interpreted to indicate which victims (still) suffer from psychological distress, either due to or not due to the act of victimization. Lawyers and other professionals involved in the evaluation of victims’ requests for compensation may play an important role in advising such victims to seek help. Unfortunately, to date, victim compensation funds have solely been pictured as targets of referral (e.g., Fritsch et al. 2004) and not as a source of referral. Significant in this respect is that a contract concluded between the DVCF and VSN in 2006 aims to improve referral to the fund by VSN volunteers, but not the other way around. Furthermore, it does not regard referral to other mental health services. Inclusion of a short mental health screener in DVCF application forms may ensure that victims with mental health problems still get the help they need and failed to ask for themselves (cf. Kunst et al. 2010). An example of such a screening instrument is the Trauma Screening Questionnaire (TSQ; Brewin et al. 2002). Dekkers et al. (2010) have recently shown that the TSQ is suitable for screening purposes. Relying on the collected data, distressed victims should, however, not be simply referred to VSN. After all, the majority of study participants had received support from a VSN volunteer (see legend Table 1). Presumably, those who apply for compensation with the fund and still suffer from high levels of distress at the start of the application procedure require a level of intervention that surpasses standard victim support.

When considering the study’s results alternative interpretations need consideration as well. Another explanation for the differences in satisfaction between victims with and without probable mental health problems is that the former were truly less satisfied with amount of financial reward than the latter. For example, one might speculate that victims suffering from mental health problems spent more money on overcoming the aftermath of victimization than those without such problems. Particularly if victims’ losses in the probable mental health problems group exceeded the maximum amount of compensation allowed to be rewarded by the fund more often than losses in the other group, differences in level of satisfaction between the two groups may have reflected real variation in discontent. Furthermore, due to its written nature, the application procedure may have neglected victims’ immaterial needs. It might be argued that those in the probable mental health problems group needed an opportunity to comment upon the submission of the claim. If so, this may have biased their satisfaction scores.

The present study was not without its limitations. First, findings were cross-sectional in nature. Cause and effect could therefore not be determined. Second, the study relied on self-report measures to assess mental health problems. Self-report assessments are less reliable than interview-based assessments. Third, study data did not allow adjustment for victims’ ethnic background. As mentioned in the “Introduction” section, previous studies suggest that this variable may be an important confounder in satisfaction research among victims of violent crime applying for state compensation (Newmark et al. 2003; Newmark 2004). Fourth, the psychometric properties of our measures of satisfaction with compensation and feelings of revenge are unknown. Fifth, our sample may have lacked representativeness for two reasons: (1) response rate was rather low and (2) non-response analyses indicated that participants differed from non-participants in several respects. However, many of these differences were likely to be due to differences in sample size. Sixth, effect sizes in this study were rather low. This indicates that many factors determining satisfaction and mental health were not measured. For example, this study did not assess perceived procedural (i.e., victims’ appraisals of the fairness of the procedure; see Lind and Taylor 1988; Thibaut and Walker 1975; Tyler 1990) and interactional justice (i.e., victims’ appraisals of being treated with respect and politeness; see Bies and Moag 1986; Lind and Taylor 1988). Given the purely written nature of the application procedure, this was not deemed appropriate. After all, such procedures do not allow victims to tell their story in person to the decision maker—a key source of differences in appraisals of procedural and interactional justice. This cannot refute, however, that the lack of opportunity to do so may have influenced victims’ satisfaction ratings.

Despite the study’s limitations, it was worthwhile for several reasons. Most important, it was the first study to examine the relationship between mental health problems and satisfaction with amount of compensation for intentional violent crime victimization. A methodological strength was that it took into account the potentially confounding effects of self-reported feelings of revenge and victim service utilization and other claimant characteristics. Another strong feature is that it largely relied on objective file data when collecting background information.

In addition to overcoming this study’s limitations, future studies might consider the investigation of positive outcomes of violent crime victimization, such as experiences of posttraumatic growth (PTG), in relation to satisfaction with amount of reward. Zoellner and Maercker (2006) defined PTG as “the subjective experience of positive psychological change reported by an individual as result of the struggle with trauma” (p. 628). Individuals who have experienced a highly traumatic event need to accommodate pretrauma schema’s about the world, self, and others (Janoff-Bulman 1992) and thereby tend to find meaning in the event itself (Taylor 1983). The outcome of this mean-making process may result in PTG (Joseph and Linley 2005, 2006; Tedeschi and Calhoun 1995, 2004). It would be particularly interesting to test whether satisfaction with compensation facilitates PTG and the other way around and whether PTG or underlying resiliency factors moderate the association between mental health problems and satisfaction with compensation.