Empathy and Juvenile Delinquency: A Systematic Review Frances C. Hobden - 10591818
University of Amsterdam
Thesis advisor: Ernst Mulder Word count: 5788
Abstract
In this systematic review, the relation between empathy and juvenile delinquency was examined. First, the relation between juvenile delinquency (a mixed offender group) and empathy, both cognitive and affective, has been studied. Second, cognitive and affective empathy levels were investigated for different types of delinquents (sexual, violent and nonviolent). No significant difference was found between the ‘mixed offender’ group and the nondelinquent group in empathic ability. In contrast, a strong negative association was found when different types of delinquency were separated. Further, a stronger positive relation was found for cognitive empathy and juvenile delinquency than for affective empathy. The findings suggest that empathy levels differ between types of offenders. An overview of the studies examining this relation will be presented in this review.
Empathy and Juvenile Delinquency: A Systematic Review
Juvenile delinquency and recidivism has been a continuous and serious problem for society (James, Asscher, Stams, & van der Laan, 2015). To address this problem, the policy to not only punish juvenile delinquents, but also consider the interest of the child exists in the Netherlands. Further, this policy has the aim to support the reintegration of the juvenile delinquent back into society (van Montfoort & Tilanus, 2007). The term juvenile delinquents applies to children and young adults from 12 to 23 years old (Nederlands Jeugd Instituut, 2012). Instead of using the harsh limit of 18 years old, a person up to the age of 23 years old can be judged as a juvenile delinquent. (Rijksoverheid, 2015). The age limit in youth criminal law was changed because it is thought that adolescents are still developing at the age of 23 and need a more tailor-made punishment (“Adolescentenstrafrecht,” 2014). This law has the aim to reduce the risk of recidivism and to offer young people more perspective in later life. To realise these goals, several correctional programs and youth interventions have been developed the past years.
One of the skills that is focused on in these programs is the empathic ability of juvenile delinquents (Nas, Brugman, & Koops, 2005; Pecukonis, 1990). Cohen and Strayer (1996) have provided a widely agreed upon definition of empathy, which will be used in this review. They view empathy as “the ability to understand and share in another's emotional state or context”. This definition states that empathy can be divided in a cognitive process (the ability to understand someone else's emotions) and an affective capacity (the ability to share and experience some else's emotions).
The reason why empathic abilities are being trained in correctional programs is that empathy is viewed as a factor which may decrease the chance to exhibit criminal behaviour, whilst the lack of it may have a reinforcing effect (Farrington, 1998). Jolliffe and Farrington (2004) conducted a meta-analysis on the relation between empathy and offending. The results
of their study showed that juvenile delinquents exhibit less empathy than juvenile non-delinquents. Nevertheless, they do emphasize that the existing literature should be studied more thoroughly to gain more knowledge on the area.
Furthermore, the empirical evidence for differences in empathy levels between juvenile sexual, violent and nonviolent delinquents appears to be ambiguous. Two opposite theories exist for these differences in empathy levels. On the one hand, it could be expected that sexual delinquents, who use force upon another person in attempting sex relations, have more contact with the victim than violent delinquents, who use force or cause injury to a victim (e.g. assault, robbery and murder). The delinquent can see, hear and experience the victim’s distress and therefore may have a greater opportunity to feel a sense of shame, guilt and experience empathy for the victim (Nagin & Tremblay, 1999). Hence, it could be expected that the intensity of victim contact would promote the expression of empathy. Jolliffe and Farrington (2004) consider this theory as plausible and suggest that an appropriate emotional response from the delinquent would be expected because of the negative emotions expressed by the victim. They also suggest that the same hypothesis applies for violent delinquents, who also have more victim contact than nonviolent delinquents, delinquents who do not use force or cause injury to a victim (e.g. property crimes, fraud and arson). On the other hand, the opposite could also be expected. Castano and Giner-Sorolla (2006) examined the theory of moral disengagement strategy by Bandura (1999). It could be predicted that once a victim is ‘dehumanized’, seen as an object or animal, the delinquent will be less likely to express empathy. In other words, feelings of empathy felt for victims can be undermined when the victim is perceived to be less human. These
contradictory theories should be studied to shed some more light on the subject.
Jolliffe and Farrington (2004) have also shown that cognitive empathy has a negative and stronger relation with offending than affective empathy. This relation depended on the
type of offense. This can be interpreted as that different concepts of empathy may have a stronger correlation with different kinds of delinquency. Other studies that have been conducted on the subject are inconsistent.
Ellis (1982) compared violent and nonviolent delinquents in their empathy levels and found a significant difference. Violent delinquents had lower empathy levels in comparison to nonviolent delinquents. But, he did not divide empathy in cognitive empathy and affective empathy, so the difference remains unclear. Another study conducted by Goldstein and Higgins-D'Alessandro in 2001 found contradictory results. They examined the levels of cognitive and affective empathy in male and female offenders (violent and nonviolent). They also found that violent male offenders scored lower on affective empathy than nonviolent male offenders. However, when age was used as a covariate, the groups no longer differed on affective empathy. Furthermore, no significant results were found on cognitive empathy in the male group and on cognitive and affective empathy in the female group. Juvenile sexual delinquents versus nonsexual (violent and nonviolent) delinquents were examined by Lindsey, Carlozzi and Eells (2001). The results of this study show that sexual delinquents have significantly lower levels of affective empathy than nonsexual delinquents. This difference was not found when comparing sexual delinquents to the non-delinquent group.
Thus, whilst the concept of empathy as a cognitive and affective construct is evident and the relation between empathy and delinquency has been studied (e.g. Farrington, 1998; Jolliffe & Farrington, 2004), it should be examined more thoroughly as a consequence of the contradicting results. Moreover, the empirical evidence of the relation between empathy and different kinds of delinquency (sexual, violent and nonviolent) is still not clear. Although, the meta-analysis conducted by Jolliffe & Farrington (2004) did provide very insightful results on the matter, more research should be done. First, they studied both juvenile and adult delinquents. This systematic research will only focus on juvenile delinquents, to gain more
knowledge on this specific group. Furthermore, studies should be repeated to update or adjust scientific knowledge on the matter. Also, in light of the focus on empathy training in youth interventions, the inconsistency of the empirical research is of great concern.
Therefore, this systematic review will summarize the findings from studies that have explored what kind of relation exists between empathy and juvenile delinquency. First, the question “What kind of differences exists in empathy levels between juvenile delinquents and juvenile non-delinquents?” will be answered. Based on the literature, it is hypothesized that juvenile delinquents show less empathy in comparison to juvenile non-delinquents. Studies which have examined both cognitive empathy and affective empathy and studies that solely examined cognitive empathy or affective empathy will be analysed. Second, this review will also focus on the question “What kind of differences exist between sexual, violent and nonviolent juvenile delinquents in empathy levels?”. It is expected that cognitive empathy has a stronger relationship than affective empathy with juvenile delinquency and there are no speculations for de differences in type of offense. Studies that examined cognitive- and affective empathy will be explored.
Methods
The review was based on a systematic search in the database Web of Science. The search strategy used terms to identify studies that examined empathy and juvenile
delinquency: empathy AND juvenile OR adolescent AND delinquen* OR offend*. No date or language restrictions were used, but only articles were selected in document types.
Furthermore, the related records of the meta-analysis of Jolliffe and Farrington (2004) and the selected articles were consulted in order to trace more recent, related studies in the database of Web of Science.
This research only examines delinquents, with recorded criminal offenses, instead of only a report of antisocial behaviour or a psychiatric diagnosis to imply an increased chance
to exhibit criminal behaviour (e.g. callous unemotional traits, psychopathy or conduct
disorder), because this could bias the findings. Moreover, the selected studies have only used juvenile delinquents up to the age of 23 to be able to investigate this specific group. Also, due to the small number of studies on this specific area, no limitations were set for the publication date. Further, the present study is based on only those studies which have used questionnaires to assess empathy, with a high reliability and which have measured the definition of empathy designed by Cohen and Strayer (1996). This decision was made upon the reason that most studies use a questionnaire to asses empathy, instead of alternative methods, and to be able to compare the studies more reliable (Jolliffe & Farrington, 2004). The studies could measure cognitive empathy or affective empathy. Table 1 in the appendix presents all the
questionnaires used in the studies, with a description and the reliability.
A coding protocol was developed to record the important methodical and substantial features of the studies. These included the age of the participants, the definition used for delinquent or offender, the type of offense committed and empathy scales which were used. This information was not given in all studies, but was coded for as far as possible. Besides, the citation report from the Web of Science Core Collection was also coded for, in order to interpret the importance of the studies. Using this protocol, twenty studies have been selected. Altogether, they had 1431 delinquents and 1262 control non-delinquents participating.
Results
What kind of differences exists in empathy levels between juvenile delinquents and juvenile non-delinquents?
The first six studies which will be discussed have examined the relation between empathy and juvenile delinquency. Combined they had 360 juvenile delinquents and 259 control non-delinquent juveniles. However, sample size did differ a lot between studies. The maximum number of participants used in the study was 206 (Mak, 1991), whilst the
minimum number was only 40 participants (Kaplan & Arbuthnot, 1985). Furthermore, the participants had an age ranging from 12 up to 19 years, which also varied between studies. For example, some studies had a wide range of age, 12-18, (e.g. Bush, Mullis, & Mullis, 2000), but Kaplan & Arbuthnot (1985) a very small one; 13-15. Every study has a control group with non-delinquent juveniles, but the studies do differ in what type of empathy they examined and in type of questionnaire used to assess empathy. Two studies examined both cognitive- and affective empathy, two studies examined cognitive empathy and two studies measured levels of affective empathy. Table 2 in the appendix shows an overview of the studies which have investigated empathy in juvenile delinquents in comparison to juvenile non-delinquents.
Four studies have examined cognitive empathy. Hudak, André and Allen (1980) and Chandler and Moran (1990) both have used the Hogan Empathy Scale (HES) to measure cognitive empathy. Hudak et al.(1980) had 32 delinquents participating in the study and 32 control non-delinquents, with an age range of 15 to 18 years. Chandler and Moran (1990) had 60 delinquents and only 20 control non-delinquents participating, with an age range of 14 to 17 years. Whilst the first used both male and female delinquents, the second only had male participants in the study participating. The results of both studies show that no significant differences were found between the delinquents and the non-delinquent juveniles on cognitive empathy.
Bush et al. (2000) and Lee and Prentice (1998) also examined cognitive empathy in juvenile delinquents. The two studies used the same questionnaire to assess cognitive
empathy, the Interpersonal Reactivity Index (IRI). The Perspective Taking subscale (IRI-PT) is used to assess cognitive empathy, which both studies did. (Davis, 1980). They did differ in sample distribution. Lee and Prentice (1998) had 36 delinquents participating in their study and 18 control non-delinquents, the two groups were matched for age and had a mean age of
16.05 years old. Bush et al. (2000) had 103 delinquents participating in the study and 66 control non-delinquents, with an age range of 12 to 19 years old. The studies also had different distributions in gender. Bush et al. had 76 male participants in the delinquent group and 33 male participants in the control group. Lee and Prentice (1998) had solely male participants in the study. Although the studies have a lot of differences, they have the same outcome. The findings of both studies support the findings of the studies discussed earlier (Hudak et al., 1980; Chandler & Moran, 1990). No significant differences were found between the delinquent- and the control group on cognitive empathy.
It can be concluded that all studies discussed found similar results when examining the differences in cognitive empathy levels between delinquents and non-delinquents. None of the studies found a significant difference between the delinquent group and the non-delinquent control group on cognitive empathy level.
Four studies explored the relation between affective empathy and juvenile delinquency. All studies used different measures to assess affective empathy. The study conducted by Bush et al. (2000) also examined affective empathy, using the IRI-EC scale. The findings show the same as for cognitive empathy, they did not find any significant differences between the delinquent and the non-delinquent group on affective empathy.
Lee and Prentice (1998) used the QMEE to assess affective empathy in their research. They found the same results as they did when examining cognitive empathy. They did not find significant results supporting their hypothesis that the delinquent group would have lower levels of affective empathy in comparison to the non-delinquent group.
The study conducted by Mak (1991) had 103 delinquents, 63 male, and 103 control participants with the same sex distribution. Both groups had an age range from 13 up to 18 years old. He used the Impulsiveness–Venturesomeness–Empathy Scale (IVE) to measure empathy. Mak (1991) did not find a significant lower result for the male delinquents on
affective empathy than de male non-delinquents, but he did find this result for the female delinquents.
The last study which has examined the differences in empathy levels in delinquents and non-delinquents is a study conducted by Kaplan and Arbuthnot (1985). The study had 20 delinquents participating, 10 male, and the same size for the control group. All participants were 13 to 15 years old. The researchers used two methods to assess affective empathy. First, they used the Duggan’s Measure of Empathy, which is an unstructured task (Duggan, 1978). Second, they used the Index of Empathy for Children and Adolescents (Bryant, 1982). They did not find any significant differences between the two groups in affective empathy levels using the IECA, but they did with the unstructured task. The delinquents performed more poorly in comparison to the non-delinquent group on the task.
Overall, the majority of the studies did not find a significant difference in affective empathy levels for both the delinquent and the non-delinquent group. The only study that did find a significant difference was the study conducted by Kaplan and Arbuthnot (1985). However, this difference was found on the unstructured affective empathy task and not on the questionnaire. This task will not be taken into consideration in this review, because this task does not have a high reliability and is not a questionnaire, which is coded for in the protocol.
What kind of differences exist between sexual, violent and nonviolent juvenile delinquents in empathy levels?
The next 14 studies have examined different types of offenders and their levels of empathy. Combined, they had 1071 delinquents and 923 control non-delinquents
participating. The studies vary a lot in design. For example, Jolliffe and Farrington (2004) conducted a meta-analysis which focused on violent, nonviolent and sexual offenders with a control group and both cognitive and affective empathy, whilst Hunter, Figueredo, Becker,
and Malamuth (2007) compared sexual and nonsexual (violent and nonviolent) offenders in levels of affective empathy. Furthermore, sample size differed a lot between studies. In the study conducted by Stams et al. (2008) 75 delinquents and 579 control non-delinquents participated, while Owen and Fox (2011) had no control group and 80 delinquents in their study participating. Moreover, some studies have examined cognitive or affective empathy, some have studied both and one study has studied cognitive- and affective empathy
separated, but presented them together in their findings as general empathy. All studies used reliable and sometimes overlapping questionnaires to assess empathy. Table 3 in the
appendix shows all the studies with their coding variables.
Milojevic and Dimitrijevi (2014), Robinson, Roberts, Strayer, and Koopman (2007) and Ellis (1982) compared violent delinquents and nonviolent delinquents in levels of cognitive empathy. Milojevic and Dimitrijevi (2014) had 43 delinquents and 47 control participants, age 15 to 17. Robinson et al. (2007) had 64 delinquents, age 14 to 18 and 60 control participants matched for age and Ellis (1982) had 331 delinquents participating and 64 control participants, all had an age from 12 to 18. Every study had only male participants. Both Robinson et al. (2007) and Milojeciv and Dimitrijevi (2014) used the IRI-PT to assess cognitive empathy, but had different findings. The first found that violent delinquents had lower scores in cognitive empathy level than nonviolent delinquents and the latter found no differences between the two groups. Ellis (1982) did not use the IRI-PT to measure cognitive empathy, but the HES. He found significant differences between the two groups. Violent delinquents scored significantly lower on the HES than nonviolent delinquents.
One other study also examined violent and nonviolent delinquents, but also
researched sexual delinquents with a control group. Jolliffe and Farrington (2004) conducted a meta-analysis, consisting 35 studies, with both juveniles and adult delinquents and studied their levels of cognitive empathy. They found a significant association between low cognitive
empathy and offending for violent delinquents, but this relation was relatively weak for sexual offenders. They add that this relationship was stronger for juveniles than for adults and that this relation was stronger for cognitive empathy than for affective empathy.
Three studies have examined sexual delinquents compared to a control group. All studies have used a similar measurement to assess cognitive empathy, the IRI-PT, but have very small sample sizes and all the participants were male. Moriarty, Stough, Tidmarsh, Eger, and Dennison (2001) had only 15 delinquents participating and 49 control non-delinquents, aged 14-17. Burke (2001) had 23 delinquents participating and 23 control participants (age 13-18) and Varker and Devilly (2007) used 16 delinquents and 16 control delinquents, aged 13 to 20 years. All studies have produced the same finding. They all came to the result that the sexual delinquents showed lower levels of cognitive empathy in comparison to the non-delinquent control group. However, this difference was not found to be significant in the study conducted by Moriarty et al. (2001), which can possibly be explained by the small sample size. A small sample size may give biased results (Hoyle, Harris, & Judd, 2002).
One other research examined sexual delinquents (Lindsey, Carlozzi, & Eells, 2001). They compared sexual delinquents to nonsexual delinquents with a control group. The nonsexual delinquents were defined as violent and nonviolent delinquents. The sample existed of 81 male delinquents and 27 male control non-delinquents, with an age range of 13 to 18. The IRI-PT was used to measure cognitive empathy. All three groups differed
significantly in scores on the PT subscale. Sexual delinquents scored the lowest on the perspective taking scale and non-delinquents the highest.
The next study discussed is the last study which has examined cognitive empathy levels in sexual offenders. This study is a literature study conducted by Nangle, Hecker, Grover and Smith in 2003. They conclude that cognitive empathy is just one factor which may influence juvenile sex offenders but that “despite the almost universally held belief
among therapists that efforts to enhance empathy are vital to treatment, the association between cognitive empathy and sexual offenses is not all that clear.” (p. 81).
Overall, the association between cognitive empathy and different types of offending is still inconsistent, but most studies conclude that sexual, violent and nonviolent offenders differ in cognitive empathy levels (e.g. Ellis, 1982; Lindsey et al., 2001; Robinson et al., 2007). Only one study found no significant differences between violent delinquents and nonviolent delinquents (Milojevic & Dimitrijevi, 2014) and one study found a weak association for sexual delinquents in comparison to non-delinquents (Jolliffe & Farrington, 2004).
The next four studies examined affective empathy in violent and nonviolent
delinquents. Markus and Gray (1998) and Lardén, Melin, Holst and Långström (2006) used the IECA to measure levels of affective empathy. Markus and Gray (1998) had 101 juvenile delinquents participating, age was not given and with no control group. Lardén et al. (2006) had 58 delinquents and 58 control non-delinquents participating in the study. Fifty percent of the sample was male and all participants were between 12 and 18 years old. Both found no significant results. Violent delinquents did not significantly score lower on the IECA in comparison to the nonviolent delinquents and the control group. Robinson et al. (2007) used both the IECA and the IRI-EC and did find a significant difference, violent delinquents had lower levels of affective empathy than nonviolent delinquents. The last study which also investigated cognitive empathy in violent and nonviolent delinquents and was discussed before, used the IRI-EC to examine affective empathy (Milojevic & Dimitrijevi, 2014). Their findings showed that violent delinquents, nonviolent delinquents and the non-delinquent control group did not differ in affective empathy levels.
Two studies also examined violent and nonviolent delinquents with a control group, but also had sexual delinquents participating in the study (Stams et al., 2008; Jolliffe and
Farrington, 2004). Stams et al. (2008) had 75 delinquents participating, 60 male, and 579 non-delinquents in their control group, 340 male. The range of age was 10 to 20 years old and they used the IECA to examine affective empathy. All delinquent groups did not significantly differ from the control group in levels of affective empathy. Jolliffe and Farrington (2004) included 14 affective empathy studies in their meta-analysis. Juvenile violent delinquents had lower affective empathy levels compared to nonviolent delinquents and non-delinquents. Lower affective empathy levels were also found for sexual delinquents, but were very weak.
Three studies did not examine violent and nonviolent delinquents, but compared sexual delinquents with a control group of non-delinquents (Moriarty et al., 2001; Burke, 2001; Varker & Devilly, 2007). Although all studies found positive correlations for cognitive empathy, they found different results for affective empathy. Moriarty et al. (2001) and Burke (2001) have found differences in affective empathy between sexual delinquents and the control group. Sexual delinquents scored lower on the IRI-EC compared to the control group. However, the findings of Burke (2001) were not significant. Contradicting results were found by Varker and Devilly (2007). They concluded that no difference exists in affective empathy levels between sexual delinquents and non-delinquents.
Hunter et al. (2007) also studied affective empathy levels with the IRI-EC in sexual delinquents, but did not have a control group in the study. 184 male delinquents participated in the study, with an age range of 12 to 18 years. The results of their study show that affective empathy is negatively associated with sexual delinquency.
One other study examined affective empathy and different types of offense. Lindsey et al. (2001) studied sexual delinquents and nonsexual delinquent with a control group. They used the IRI-EC to measure affective empathy. Their findings show that the total scores of the three groups differ significantly on the IRI, but not on the EC-subscale, which measures affective empathy.
The last study discussed has examined violent delinquents and nonviolent delinquents (Owen & Fox, 2011). The researchers aimed to investigate affective- and cognitive empathy separately, but the initial attempts to divide empathy were abandoned. The authors decided that there was too much overlap between them, so they have measured general empathy instead. They have examined 80 delinquent males, 13 to 21 years old with the empathy quotient (Baron-Cohen & Wheelwright, 2004). The results of the study show that nonviolent delinquents scored significantly higher on empathy compared to the violent delinquents.
All considered, it can be stated that no definite answer can be given to the question if affective empathy levels differ between types of delinquents. Markus and Gray (1998), Lardén et al. (2006) and Milojevic and Dimitrijevi (2014) did not find significant differences in affective empathy levels in violent and nonviolent delinquents, but Robinson et al. (2007) did. Furthermore, the studies examining sexual offenders are also inconsistent in their findings. For example, Moriarty et al. (2001) do find a significant difference between sexual delinquents and the control group, but Varker and Devilly (2007) do not. For general
empathy, the results show that nonviolent delinquents had higher levels in empathy than violent delinquents (Owen & Fox, 2011).
Discussion
The aim of this review was to answer the questions “What kind of differences exists in empathy levels between juvenile delinquents and juvenile non-delinquents?” and “What kind of differences exists between sexual, violent and nonviolent juvenile delinquents in empathy levels?”. The results of this review show that no significant differences were found in empathy levels, both cognitive and affective, for delinquent and nondelinquent juveniles. These findings contradict the hypothesis that juvenile delinquents have lower empathy levels than juvenile non-delinquents.
However, when looking at the findings from the studies that examined different types of offenders, other results show. Most studies suggest that sexual, violent and nonviolent juvenile delinquents differ in cognitive empathy levels. The results show that violent delinquents have lower cognitive empathy levels in comparison to nonviolent delinquents. Further, the findings show that juvenile sexual delinquents have lower cognitive empathy levels in comparison to nonsexual (violent and nonviolent) juvenile delinquents and that nonsexual juvenile delinquents have lower cognitive empathy than nondelinquent youth.
The studies that examined affective empathy levels and different type of offense are inconsistent. The majority of the studies show that no significant differences exist between violent, nonviolent and sexual juvenile delinquents, but some do. Two studies concluded that sexual delinquency has a strong negative correlation with affective empathy and the meta-analysis conducted by Jolliffe and Farrington (2004) also found a strong negative association for violent juvenile delinquency in comparison to nonviolent juvenile delinquency and a weak negative association for sexual juvenile delinquency. They included 14 affective
empathy studies and have been cited 189 times in the Web of Science Core Collection, so this should be taken into consideration.
Thus, no definite answer can be given to the question what kind of differences exist between type of offense and empathy levels. But, this review does show some coherence in the literature. In line with the hypothesis, most of the studies discussed do show that sexual, violent and nonviolent juvenile delinquents differ in cognitive empathy levels. Whereas sexual juvenile delinquents have the lowest cognitive empathy levels and nonviolent juvenile delinquents the highest. This review also suggests that affective empathy levels do not significantly differ between sexual, violent and nonviolent juvenile delinquents or when found, that the association is less strong in comparison to cognitive empathy. This finding
confirms the expectation that cognitive empathy has a stronger association with juvenile delinquency in comparison to affective empathy.
Nevertheless, the finding that every study, which has investigated ‘mixed offense’ juvenile delinquents, did not find a significant difference is remarkable. Several possible reasons may account for this finding. First, not taking type of offense into account may be a reason why the results of this review are contradicting the hypothesis. One explanation is that the studies that have explored delinquent youth versus nondelinquent youth do not find significant results because the empathy levels in the delinquent group may differ a lot. It is unclear what the sample distribution was for the delinquent group and it may be that these studies had many nonviolent delinquents participating and just a few sexual delinquents and that their ‘total’ empathy levels did not differ significantly from the control group.
Fortunately, this problem of unknown sample distribution has been tackled and recent delinquency studies mention the sample distribution of their delinquent group more often (e.g. Stams et al., 2008; Moriarty et al., 2001).
Another explanation may be that these studies have had all types of delinquents equally participating in their study, but that the two theories mentioned before (Bandura’s moral disengagement strategy theory and Nagin and Tremblay’s victim contact intensity theory) are both applicable, but for different types of offense. For instance, it could be suggested that Bandura’s moral disengagement strategy applies for sexual juvenile
delinquents. Imaginable, the dehumanization of victims may be necessary for the severity of the offenses committed by sexual delinquents, but for smaller ‘victimless’ crimes such as property crimes this may not be the case. But, the victim intensity theory could possibly be more applicable for violent and nonviolent juvenile delinquents. For example, it may be that violent delinquents express more empathy in comparison to nonviolent delinquents, because they have more victim contact and have a greater opportunity to experience empathy for the
victim. It could be that sexual delinquents have very low levels of empathy, but that violent delinquents do have high empathy levels and that this difference may explain why no significant correlation between empathy and delinquency were found, when delinquents are treated as a whole group.
Further, other variables such as intelligence and social economic status may also influence whether or not a delinquent commits a crime. For instance, the study by Chandler and Moran (1990) did find a significant difference at first, but this difference disappeared when intelligence was controlled for. Also, Lee and Prentice (1998) did not find significant differences for empathy levels, but they did for role-taking, logical cognition and Kohlberg’s moral dilemmas. Further, Mak (1980) did not match the control group for intelligence and SES. This may indicate that other variables should also be taken into consideration when conclusions about empathy levels in juvenile delinquents are being made.
Noteworthy is that these studies are relatively old, the most recent study was conducted in 2000. First, it was thought that the coding of this review may had biased the results, because of the inclusion criteria for measurement. But, this was checked for by removing this in the search and no other recent studies were found that examined mixed offense and empathy levels. So, a possible explanation for this is that this type of research might have been a trend in the ‘research history’ for this type of research. Perhaps,
researchers in the 1980’s and 1990’s did not find type of offense relevant or did not realise that type of offense may be an important variable in research investigating empathy levels. This may explain why type of offense is not a variable in early research, but only in more recent studies. Further research should investigate both different type of offenders and the delinquent group as a ‘mixed offense’ group to learn if this may have biased earlier findings.
Although the results from the studies which examined different types of offenders are mostly consistent, some limitations should be mentioned to improve further research. First,
some studies did not use a control group or did have one but not matched for age, gender, intelligence and social economic status. The results may have been biased because some studies did not have a design with a (matched) control group. For example, it is suggested by Jolliffe and Farrington (2004) that empathy should not be investigated solely, but that
intelligence should always be investigated too. It may be that the differences in empathy levels between juvenile delinquents and non-delinquents are an artifact of the strong correlation between intelligence and delinquency or between empathy and intelligence
(Constantine & Gainor, 2001). Future research should always use matched control groups and should also explore the variable intelligence when examining empathy and juvenile
delinquency.
Second, the studies differed a lot in age of the participants and its range. Van Vugt et al. (2012) states that empathy and social behaviour develops over time and that this
development may also influence the correlation between empathy and juvenile delinquency. Thus, it is recommended to have a wide range of ages and divide developmental stages into different groups to reduce the risk that this may distort the findings. All types of offenses should have a wide range of age if possible in future research.
Third, the classification of offenders might be misleading the results. Classifying offender type solely on official records may not be sufficient. Some offenders may be currently convicted for a nonviolent offense, but have a history of violent offenses. These misleading classifications might have biased the results of this systematic review. To avoid this in future research, it is suggested that not only an official record should be used, but also self-report questionnaires about criminal behaviour to identify different types of offenders (Farrington, 1998).
Fourth, it may also be relevant to inquire if the delinquent is a first or a second offender. It can be speculated that first offenders may have higher empathy levels than
second offenders. Not only type of offender may be relevant, but also the number of times a delinquent commits the offense (Jolliffe & Farrington, 2004). Moreover, it should also be noted that delinquents are not all incarcerated. Jolliffe and Murray (2012) argued that incarcerated delinquents are a very selected group of delinquents. Only a small number of delinquents are convicted and only some have to serve time. Future research should take into account that little is known about the non-incarcerated group of juvenile delinquents and empathy.
Lastly, longitudinal studies are needed to investigate juvenile delinquency and its relation to empathy. The presented literature has only examined the correlation between empathy and juvenile delinquency, but has not investigated the causal relations. Longitudinal research may give insight whether low empathy levels may increase delinquent behaviour, or that delinquent behaviour may decrease empathy levels. Also, as mentioned before, it may be that intelligence is an important influence on criminal behaviour and it may be that the studies have examined the correlation between intelligence and delinquency instead of
empathy and delinquency, whereas intelligence and empathy are also highly correlated. Thus, longitudinal research may present more knowledge on the causal relations between empathy and delinquency, when intelligence is also taken into consideration.
Despite these limitations, this review does give some insight in the relation between juvenile delinquency and empathy. The findings show that different types of offenders do differ in empathy levels and that cognitive empathy has a stronger relation to delinquency than affective empathy. But, considering the importance of empathy training in all
correctional settings, future longitudinal research must be conducted. Further research is needed to investigate whether intelligence causes low empathy, which causes offending or if this link of causation may have a different route. For now, the literature does suggest that treatments designed to enhance empathic abilities may be useful in decreasing juvenile
delinquency. These trainings should have its focus on cognitive empathy and should be altered for the type of offense, but should also have the intellectual ability of the offender in mind.
References
Adolescentenstrafrecht: 16- tot 23-jarigen op maat berecht (2014, April 1). De
Rechtspraak. Retrieved from https://www.rechtspraak.nl
Constantine, M. G., & Gainor, K. A. (2001). Emotional intelligence and empathy: Their relation to multicultural counseling knowledge and awareness. Professional School Counseling, 5(2), 131-138. Retrieved from http://www.psycnet.apa.org
Hoyle, R. H., Harris, M. J. &Judd, C. M. (2002). Research Methods in Social Relations. (7th ed.). London, England: Wadsworth, Thomson Learning
Bandura, A. (1999). Moral disengagement in the perpetration of inhumanities. Personality
and Social Psychology Review, Special Issue on Evil and Violence, 3, 193–209.
doi:10.1023/A:1005162526769
Bryant, B. K. (1982). An index of empathy for children and adolescents. Child
Development, 53, 413–425. doi:10.2307/1128984
Burke, D. M. (2001). Empathy in sexually offending and nonoffending adolescent males. Journal of interpersonal violence, 16, 222-233. doi:10.1177/088626001 016003003
Bush, C. A., Mullis, R. L., & Mullis, A. K. (2000). Differences between offender and nonoffender youth. Journal of Youth and Adolescence, 29, 467–478. doi:10.1023 /A:1005162526769
Castano, E., & Giner-Sorolla, R. (2006). Not quite human: Infrahumanization as a
response to collective responsibility for intergroup killing. Journal of Personality and
Social Psychology, 90, 804–818. doi:10.1037/0022-3514.90.5.804
Chandler, M., & Moran, T. (1990). Psychopathy and moral development: A
comparative study of delinquent and nondelinquent youth. Development and
Cohen, D., & Strayer, J. (1996). Empathy in conduct-disordered and comparison youth.
Developmental Psychology, 32, 988–998. doi:10.1037/0012-1649.32.6.988
Davis, M. H. (1980). A multidimensional approach to individual differences in
empathy. JSAS Catalog of Selected Documents in Psychology, 10, 85-86. Retrieved from http://www.citeulike.org
Davis, M. H. (1983). Measuring individual differences in empathy: Evidence for a
multidimensional approach. Journal of Personality and Social Psychology, 44, 113– 126. doi:10.1037/0022-3514.44.1.113
Duggan, H.A. (1978). Second Chance. Boston, Massachusetts: Lexington Books Ellis, P. L. (1982). Empathy: A factor in antisocial behavior. Journal of Abnormal
Child Psychology, 10, 123–134. doi:10.1007/BF00915957
Eysenck, S. B. J., & McGurk, B. J. (1980). Impulsiveness and venturesomeness in a detention center population. Psychological Reports, 47, 1299–1306. doi:10.2466/ pr0.1980.47.3f.1299
Farrington, D. P. (1998). Individual differences and offending. The handbook of crime and
punishment, 241-268. New York, New York: Oxford University Press
Goldstein, H., & Higgins-D'Alessandro, A. (2001). Empathy and attachment in relation to violent vs. non-violent offense history among jail inmates. Journal of Offender
Rehabilitation, 32, 31-53. doi:10.1300/J076v32n04_03
Hudak, M. A., Andre, J., & Allen, R. O. (1980). Delinquency and social values:
Differences between delinquent and nondelinquent adolescents. Youth & Society, 11, 353–368. doi:10.1177/0044118X8001100304
delinquency in juvenile sexual offenders: The mediating and moderating influences of emotional empathy. Journal of Family Violence, 22, 43-54. doi:10.1007/s10896-006-9056-9
James, C., Asscher, J. J., Stams, G. J. J., & van der Laan, P. H. (2015). The Effectiveness of Aftercare for Juvenile and Young Adult Offenders. International journal of offender
therapy and comparative criminology, 1, 1-26, doi:10.1177/0306624X15576884
Jolliffe, D., & Farrington, D. P. (2004). Empathy and offending: A systematic review and meta-analysis. Aggression and Violent Behavior, 9, 441-476. doi:10.1016
/j.avb.2003.03.001
Kaplan, P. J., & Arbuthnot, J. (1985). Affective empathy and cognitive role-taking in delinquent and nondelinquent youth. Adolescence, 20, 323-333. Retrieved from http://apps.webofknowledge.com
Lindsey, R. E., Carlozzi, A. F., & Eells, G. T. (2001). Differences in the dispositional empathy of juvenile sex offenders, non-sex-offending delinquent juveniles, and nondelinquent juveniles. Journal of Interpersonal Violence, 16, 510-522. doi:10.1177/088626001016006002
Nagin, D., & Tremblay, R. E. (1999). Trajectories of boys' physical aggression, opposition, and hyperactivity on the path to physically violent and nonviolent juvenile
delinquency. Child development, 70, 1181-1196. doi:10.1111/ 1467-8624.00086 Nas, C. N., Brugman, D., & Koops, W. (2005). Effects of the EQUIP programme on the moral judgement, cognitive distortions, and social skills of juvenile delinquents.
Psychology, Crime & Law, 11, 421-434. doi:10.1080/1068 316050025570
Nangle, D. W., Hecker, J. E., Grover, R. L., & Smith, M. G. (2004). Perspective
taking and adolescent sex offenders: From developmental theory to clinical practice.
Nederlands Jeugd Instituut (2012). Being Young in the Netherlands. Retrieved from http://www.nji.nl/english
Mak, A. S. (1991). Psychosocial control characteristics of delinquents and nondelinquents.
Criminal Justice and Behavior, 18, 287–303. doi:10.1177/0 093854891018003003
Marcus, R. F., & Gray, L. (1998). Close relationships of violent and nonviolent
African American delinquents. Violence and Victims, 13, 31–46. Retrieved from http://apps.webofknowledge.com
Milojević, S. Z., & Dimitrijevic, A. (2014). Empathic capacity of delinquent convicted minors. Psihologija, 47, 65-79. doi:10.2298/PSI1401065M
Moriarty, N., Stough, C., Tidmarsh, P., Eger, D., & Dennison, S. (2001). Deficits in
emotional intelligence underlying adolescent sex offending. Journal of Adolescence,
24, 743-751. doi:0.1006/jado.2001.0441
Owen, T., & Fox, S. (2011). Experiences of shame and empathy in violent and non-
violent young offenders. Journal of Forensic Psychiatry & Psychology, 22, 551-563. doi: 10.1080/14789949.2011.602096
Pecukonis, E. V. (1990). A cognitive/affective empathy training program as a function of ego development in aggressive adolescent females. Adolescence, 25, 59. San Diego, California: Libra Publishers Inc
Rijksoverheid (2015). Straffen en maatregelen voor jongeren. Retrieved from https://www.rijksoverheid.nl
Robinson, R., Roberts, W. L., Strayer, J., & Koopman, R. (2007). Empathy and emotional responsiveness in delinquent and non-delinquent adolescence. Social
Development, 3, 555–579. doi:10.1111/j.1467-9507.2007.00396.x
L. H., Tavecchio, L. W. C., Hendriks, J., van Schijndel, M. (2008). The relation of punishment- and victim-based moral orientation to prosocial, externalizing, and norm trespassing behavior in delinquent and non-delinquent adolescents: A validation study of the Moral Orientation Measure. Journal of Experimental Criminology, 4, 41–60. doi:10.1007/s11292-007-9045-x
van Montfoort, A. J., & Tilanus, C. P. G. (2007). Jeugdzorg & jeugdbeleid. Amsterdam, Nederland: SWP.
van Vugt, E. S., Asscher, J. J., Hendriks, J., Stams, G. J. J., Bijleveld, C. C., & van
der Laan, P. H. (2012). The relationship between psychopathy and moral development in young sex offenders. Psychology, Crime & Law, 18, 655-667. doi:10.1080
/1068316X.2010.533177
Varker, T., & Devilly, G. J. (2007). Types of empathy and adolescent sexual
offenders. Journal of Sexual Aggression, 13, 139–149. doi: 10.1080/13552600701 661573
Appendix
Table 1
Questionnaires to measure empathy
Scale Description Reliability
HES (Hogan, 1969) 64 item questionnaire
True or False scale
Test-retest r=.84 (Hogan, 1969)
QMEE (Mehrabian & Epstein, 1972)
33 item questionnaire Likert scale
Split-half r=.84
Test-retest r=.83 (Bryant, 1982)
IVE (Eysenk & Eysenk, 1978) 63 item questionnaire
True or False scale
21 items composed from the QMEE
Alpha reliability .65 for men and .64 for women (Eysenk & Eysenk, 1978)
IRI (Davis, 1980) 27 item questionnaire
Likert scale 7-item subscales,
measuring perspective-taking, fantasy, personal distress and empathic concern
Internal reliability from .71 to .77 for the four subscales
Test-retest r=.68 to .79 (Davis, 1983)
IECA (Bryant, 1982) 22 item questionnaire
True or False scale Derived from the QMEE
Test-retest r=.74
EQ (Baron-Cohen & Wheelwright, 2004)
60 item questionnaire Likert scale
Alpha reliability r=.81 (Baron-Cohen & Wheelwright, 2004)
Table 2
Empathy (cognitive and affective) in juvenile delinquents and non-delinquents
Year Authors Sample Age Type of Empathy Measures Findings Citation report
1980 Hudak, André, & Allen 32 delinquents 16 male 32 control 16 male 15-18 15-18
Cognitive Empathy HES No significant differences
were found 5 1985 Kaplan & Arbuthnot 20 delinquents 10 male 20 control 10 male 13-15 13-15
Affective Empathy IECA No significant differences
were found 24 1990 Chandler & Moran 60 delinquents 60 male 20 control 20 male 14-17 14-17
Cognitive Empathy HES No significant differences
were found 43 1991 Mak 103 delinquents 63 male 103 control 63 male 13-18 13-18
Affective Empathy IVE Female delinquents scored
significantly lower on empathy than female non-delinquents 23 1998 Lee & Prentice 36 delinquents 36 male 18 control 18 male M=16.05 Matched for age Cognitive Empathy Affective Empathy IRI-PT QMEE No significant differences were found 51 2000 Bush, Mullis, Mullis 109 delinquents 76 male 66 control 33 male 12-18 15-19 Cognitive Empathy Affective Empathy IRI-PT IRI-EC No significant differences were found 21
EMPATHY AND JUVENILE DELINQUENCY 29
Table 3
Empathy in juvenile delinquents (sexual, violent and non-violent)
Year Authors Sample Age Type of offender Type of Empathy Measures Findings Citation report
1982 Ellis 331 delinquents 331 male 64 control 64 male 12-18 12-18 Violent Nonviolent Cognitive Empathy
HES Violent delinquents
significantly lower than nonviolent on HES 45 1998 Markus & Gray 101 delinquents - Violent Nonviolent Affective Empathy IECA No significant difference found 13 2001 Moriarty, Stough, Tidmars, Eger, & Dennison 15 delinquents 15 male 49 control 49 male 14-17 14-17 Sexual Control Cognitive Empathy Affective Empathy IRI-PT IRI-EC No significant difference found 40 2001 Burke 23 delinquents 23 male 23 control 23 male 13-18 13-18 Sexual Control Cognitive Empathy Affective Empathy IRI-PT IRI-EC Offenders scored significantly lower than nonoffenders 36 2001 Lindsey, Carlozzi, Eells 81 delinquents 81 male 27 control 27 male 13-18 13-18 Sexual Nonsexual Control Cognitive Empathy Affective Empathy IRI-PT IRI-EC
Total scores of three groups were significantly different, but on EC-subscale 27 2003 Nangle, Hecker, Grover, Smith - Literature overview Cognitive Empathy - The association
between empathy and sexual offenses is not all that clear
2004 Jolliffe & Farrington Meta-analysis Juvenile Adults Violent Nonviolent Sexual Control Cognitive Empathy Affective Empathy - The relationship
between low empathy and offending was relatively strong for violent offenders, but weak for sex offenders and was stronger for juveniles than for adults 189 2006 Lardén, Melin , Holst & Långström 58 delinquents 29 male 58 control 29 male 12-18 12-18 Violent Nonviolent Control Affective empathy
IECA No association was
found 26 2007 Varker & Devilly 16 delinquents 16 male 16 control 16 male 13-20 13-20 Sexual Control Cognitive empathy Affective empathy IRI-PT IRI-EC Delinquents showed lower levels of cognitive empathy, but not for affective empathy 8 2007 Robinson, Roberts, Strayer & Koopman 64 delinquents 64 male 60 control 60 male 14-18 15-19 Violent Nonviolent Cognitive empathy Affective empathy IRI-PT IRI-EC IECA Delinquents scored lower on empathy in comparison to control. Violent delinquents scored lower than nonviolent delinquents 25 2007 Hunter et al. 184 delinquents 184 male 12-18 Sexual Affective empathy
IRI-EC Affective empathy was
found to be negatively associated with sexual offenders
2008 Stams et al. 75 delinquents 60 male 579 control 340 male 13-20 10-20 Sexual Violent Nonviolent Control Affective empathy IECA No significant
differences were found 3 2011 Owen & Fox 80 delinquents 80 male 13-21 Violent Nonviolent
General empathy EQ Non-violent offenders
were found to have significantly higher empathy scores compared to violent offender 4 2014 Milojevic & Dimitrijevi 43 delinquents 43 male 47 control 47 male 15-17 15-17 Violent Nonviolent Control Cognitive empathy Affective empathy IRI-PT IRI-EC Delinquents scored significantly lower on PT, but not on EC. No differences between violent and nonviolent group
