-- Factors Associated With Experts' Opinions Regarding Criminal Responsibility in the Ne

7/28/2019 -- Factors Associated With Experts' Opinions Regarding Criminal Responsibility in the Ne

1/13

Factors Associated with

Experts Opinions Regarding

Criminal Responsibility in The

Netherlands

Marko Barendregt, Ph.D.,*,y Eline Muller, M.A.,y

Henk Nijman, Prof.z and Edwin de Beurs, Ph.D.y

In many jurisdictions, offenders need to have freely chosen

to commit their crimes in order to be punishable. A mental

defect or disorder may be a reason for diminished or total

absence of criminal responsibility and may remove culp-

ability. This study aims to provide an empirically based

understanding of the factors on which experts base their

judgements concerning criminal responsibility. Clinical,

demographic and crime related variables, as well as

MMPI-2 profiles, were collected from final reports concern-

ing defendants of serious crime submitted to the observationclinic of the Dutch Ministry of Justice for a criminal respon-

sibility assessment. Criminal responsibility was expressed

along a five-point scale corresponding to the Dutch legal

practice. Results showed that several variables contributed

independently to experts opinions regarding criminal

responsibility: diagnosis (Axis I and II), cultural back-

ground, type of weapon used in committing the crime,

and whether the defendant committed the crime alone or

with others. In contract to jurisdictions involving a sane/

insane dichotomy, the Dutch five-point scale of criminal

responsibility revealed that Axis II personality disordersturned out to be mostly associated with a diminished

responsibility. MMPI-2 scores also turned out to have a

small contribution to experts opinions on criminal respon-

sibility, independently of mere diagnostic variables. These

results suggest that experts base their judgements not only

on the presence or absence of mental disorders, but also on

cultural and crime related characteristics, as well as dimen-

sional information about the defendants personality and

symptomatology. Copyright# 2008 John Wiley & Sons, Ltd.

Behavioral Sciences and the Law

Behav. Sci. Law 26: 619631 (2008)

Published online in Wiley InterScience

(www.interscience.wiley.com) DOI: 10.1002/bsl.837

*Correspondence to: Marko Barendregt, P.O. Box 13369, 3507 LJ, Utrecht, The Netherlands.E-mail: [email protected] Institute of Forensic Psychiatry and Psychology, Location Pieter Baan Centre, Utrecht, TheNetherlandszRadboud Universiteit, Nijmegen, Academic Centre of Social Sciences, The Netherlands, and Division ofOrtho- and Forensic Psychiatry (OFP) of Altrecht Mental Health Institute, Den Dolder, The Netherlands

Copyright # 2008 John Wiley & Sons, Ltd.


2/13

INTRODUCTION

In many jurisdictions, in order to be punishable offenders must not only have

committed the criminal act of which they are accused (actus reus), but they must also

have freely chosen to do so (mens rea). Criminal responsibility, thus, is a necessary

component for punishment and diminished responsibility or insanity may be a

reason for acquittal or a mitigated sentence. Many papers addressing criminal

responsibility focus on theoretical issues such as what personality aspects or

psychiatric disorders forensic experts ought to include in evaluating a defendants

criminal responsibility (e.g., Fine & Kennett, 2004; McSherry, 2004; Hannan,

2005). Despite these discussions, there is hardly any consensus on what constitutes

grounds for (diminished) responsibility (Murrie & Warren, 2005). In contrast,

clearly less empirical research has examined what forensic experts actually doconsider important aspects in evaluating criminal responsibility. Such empirical

research is important in order to understand the meaning of the concept of criminal

responsibility, as it reveals on what experts base their judgements.

The few empirical studies that have addressed factors involved in assessing

criminal responsibility consistently show that experts opinions regarding criminal

responsibility and verdicts of insanity are related to Axis I psychiatric diagnoses;

notably, subjects with psychotic disorders were more likely to be found insane than

subjects without psychotic symptoms (Lymburner & Roesch, 1999; Warren, Murrie,

Chauhan, Dietz, & Morris, 2004; Rice & Harris, 1990; Cochrane, Grisso, &

Frederick, 2001). Additionally, Warren et al. (2004) found a negative relationbetween Axis II personality disorders and opinions supportive of insanity: subjects

with a personality disorder were less likely to be found insane than subjects without a

personality disorder. In contrast, Janofsky, Dunn, Roskes, Briskin, & Rudolph

(1996) did not find any relationship between diagnosis and verdicts of insanity.

Besides diagnoses, age and gender of the defendant (Folino & Urrutia, 2001; Rice

& Harris, 1990; Warren et al., 2004), the number of prior convictions and the

number of prior arrests (Folino & Urrutia, 2001; Lymburner & Roesch, 1999; Rice

& Harris, 1990) have been found to be associated with criminal responsibility. The

largest study among more than 4000 subjects, however, did not find a relationship

between criminal responsibility and prior convictions (Warren et al., 2004). Also, ina Canadian sample racial differences have been found, in the sense that minority

groups had a smaller chance of being acquitted on the basis of insanity than the white

population (Warren et al., 2004).

The type of the index offence seems to be related to criminal responsibility (Folino

& Urrutia, 2001; Rice & Harris, 1990; Warren et al., 2004). None of the offence

types, however, were significantly related to insanity once diagnostic opinions were

simultaneously considered (Cochrane et al., 2001; Warren et al., 2004). This latter

finding suggests that the type of offence does not directly affect experts opinions

regarding insanity, but that individuals charged with different crimes have different

psychopathological profiles, and therefore also have different rates of insanity.

Also, defendants who committed a crime against a related person were

more likely to be found insane than those who victimized unrelated persons

(Lymburner & Roesch, 1999; Folino & Urrutia, 2001). To the current authors

knowledge, no studies up till now have investigated whether offence related

Copyright # 2008 John Wiley & Sons, Ltd. Behav. Sci. Law 26: 619631 (2008)

DOI: 10.1002/bsl

620 M. Barendregt et al.


3/13

characteristics (e.g. weapon choice) are related to the judgement about criminal

responsibility.

Finally, Rogers and McKee (1995) found significant differences for some clinical

scales of the Minnesota Multiphasic Personality Inventory-2 (MMPI-2) betweengroups characterized by type of criminal responsibility. However, there was no

consistent pattern in their results and Rogers and McKee warned against

overinterpretation. Moreover, they hypothesized that the elevations on some of

the MMPI-2 scales are indicative of underlying psychopathology instead of being an

independent factor for criminal responsibility.

In short, the overall picture provided by the earlier studies suggests that several

clinical, demographic and crime related variables seem to be relevant for criminal

responsibility evaluations. Yet, it is important to keep in mind that criminal

responsibility is essentially a legal and normative construct and therefore its

assessment will probably differ greatly between jurisdictions and also may tend tochange over time (Niveau & Sozonets, 2001; Warren, Rosenfeld, Fitch, & Hawk,

1997). Moreover, most of the previous research has exclusively reported on

univariate associations. As it may be expected that many variables are mutually

related, cumulative values of these significant effects may be questioned once all

variables are simultaneously considered.

Current study purposes and objectives

The overall purpose of the present study is to provide an empirically based

understanding of the concept of criminal responsibility as it is used in Dutch forensic

practice. The first aim is to find out whether the results on the predictors for criminal

responsibility reported in other jurisdictions can be replicated in a Dutch population

of clinical responsibility evaluatees.

Second, since it may be expected that the independent variables are mutually

related, univariate analyses may result in overreporting of significant effects.

For a proper understanding of criminal responsibility, such overreporting should be

avoided. The second aim of the study is to test a multivariate model with the mostparsimonious set of variables contributing to the explanatory power of the model.

Finally, in line with Rogers and McKee (1995), it is hypothesized that

relationships between MMPI-2 scales and criminal responsibility are mediated by

underlying psychopathology. It is investigated whether these scales provide

additional explanatory power above diagnostic information.

Criminal responsibility in The Netherlands

In The Netherlands, if the court suspects the presence of a psychiatric disorder, it has

a number of options. The court can order either a mono-disciplinary assessment (by

a psychologist or a psychiatrist) or a bi-disciplinary assessment (by a psychologist and

a psychiatrist). These assessments are usually done in a regular house of detention

and are relatively low intensive, involving one or two interviews with the defendant.

Copyright# 2008 John Wiley & Sons, Ltd. Behav. Sci. Law 26: 619631 (2008)

DOI: 10.1002/bsl

Experts opinions regarding criminal responsibility in The Netherlands 621


4/13

A third option is to order an intensive multidisciplinary clinical assessment, in which

case the defendant is assessed on an inpatient basis during a seven week observation

period in a forensic hospital. Reasons for the court to order an intensive clinical

assessment as opposed to a less intense assessment include the severity of the crime,the severity of the assumed psychopathology, the safety level, and potential societal

disturbance or media attention for the case in question. As a result, the population of

the Pieter Baan Centre covers the more severe criminological and psychiatric cases,

but cannot be seen as representative for the entire forensic population whose mental

status is assessed.

The Dutch regulations concerning criminal responsibility are formulated in

Article 39 of the Penal Code:

Not punishable is he who commits a criminal act for which he cannot be held

responsible because of a defect or disorder of his mental capacities.

This formulation entered the Penal Code in 1886 and, probably due to its abstract

phrasing, is still considered appropriate today, despite the vast developments that have

taken place in psychiatry. The term defect refers to developmental disorders or brain

damage that are in principle life-long, while disorder refers to other mental illnesses

of a more changeable nature and duration. A mental defect or disorder at the time of

the crime may be a reason to remove culpability of the defendant, partly or entirely, in

The Netherlands. It is generally agreed that the criminal act must be due to the defect

or disorder in order to remove culpability (Beekman & van Mulbregt, 2004).

The situation in The Netherlands is somewhat different from other (most notablythe American) jurisdictions (see van der Leij, Jackson, Malsch, & Nijboer, 2001, for

an extensive overview of the Dutch legal system as far as mental health assessments

are concerned).

First, in Dutch forensic practice (although not literally according to the Penal

Code) a five-point scale is used for indicating the degree of criminal responsibility

ranging from complete responsibility, slightly diminished responsibility, diminished

responsibility, and severely diminished responsibility to total absence of respon-

sibility. Although a degree of diminished responsibility is often a reason for reduced

punishment, there is no mandatory proportionality between the degree of

responsibility and punishment (e.g. the length of a prison sentence). Usually, areduced punishment is combined with a measure such as compulsory psychiatric

treatment.

Second, a diminished degree of responsibility is not restricted to a list of

sanctioned diseases and disorders. Thus, in contrast to US jurisdictions (Reid,

2001), in The Netherlands antisocial personality disorder may be a reason for

diminished responsibility.

Third, it is the investigating judge who normally decides whether or not a criminal

responsibility assessment is required, although the defence and prosecution may ask

the judge for such an assessment.1 As a consequence, the assessment is essentially

neutral. The court discloses the final reports to both defence and prosecution.

1 In principle, the prosecution may also decide on an assessment, a possibility that entered the Dutch legalpractice a few years ago and is still very rarely used.


DOI: 10.1002/bsl



5/13

METHOD

Assessment site

All subjects were admitted to the Pieter Baan Centre (PBC) in Utrecht, The

Netherlands, for a multi-disciplinary pre-trial criminal responsibility assessment

(Mooij, 1991). The PBC is the official forensic psychiatric observation clinic of the

Dutch Ministry of Justice with the legal status of a house of detention. The PBC

administers about 90% of all clinical forensic assessments in The Netherlands.

All defendants are evaluated during a seven week period by a multidisciplinary team

consisting of a social worker investigating the life-history and social background of the

defendant, a group worker observing and describing the activities of the defendant

within the clinic during the observation period, a psychologist, and a psychiatrist. The

latter two experts carry the final responsibility for the conclusion concerning criminalresponsibility and the judgement about DSM-IV diagnoses if any. They also advise

about whether forensic treatment would be appropriate in their opinion.

Subjects

The population from which the sample was drawn consisted of defendants needing

pre-trial multi-disciplinary criminal responsibility assessment. Because of the low

number of female defendants admitted to the institution (less than 10%), it was

decided to include only male subjects in the current study. A total of 885 maledefendants who cooperated in the assessment were admitted to the PBC between

1 January 2000 and 31 December 2005 for an assessment. Each subject was screened

by the teams psychologist for capacities and willingness to complete self-report

questionnaires. The major reasons for not administering questionnaires included

unwillingness to cooperate in the test-assessment, poor intellectual functioning, the

presence of a severe psychosis and language problems. As a result of this screening,

the MMPI-2 was administered to 249 (28%) defendants.

The sample was composed of 27 (11%) defendants considered by the team to

have complete responsibility, 52 (21%) somewhat diminished responsibility,

135 (54%) diminished responsibility, 24 (10%) severely diminished responsibilityand 11 (4%) complete absence of responsibility. Because the latter group was to

small for statistical analyses, the defendants with complete absence of responsibility

were excluded from the analyses, resulting in a final sample of 238 subjects. An

alternative could have been to include those defendants with complete absence of

responsibility in the severely diminished group. However, because of the special legal

status of those defendants with complete absence of responsibility (no punishment

versus mitigated sentence) we found combining these groups unwarranted as they

may be expected to be qualitatively dissimilar. Potential consequences of this choice

will be discussed.

Material

The dependent variable (experts opinion regarding criminal responsibility) was

collected from the final reports in which experts express their opinions according to


DOI: 10.1002/bsl



6/13

the aforementioned five-point scale. Besides the experts opinion on criminal

responsibility, 14 variables as well as the MMPI-2 profiles of the defendants were

collected from the final reports. The 14 variables comprise three categories of

information: clinical (presence/absence of DSM-IV Axis I psychiatric symptoms,presence/absence of Axis II personality disorder, presence/absence of Axis I

substance abuse), demographic variables (age of defendant at time of assessment,

number of earlier convictions, age at first conviction, first offender, cultural

background), and crime related variables (type of index crime, attempted murder,

relationship between the perpetrator and the victim (i.e. acquainted or not

acquainted with the victim), type of weapon used, crime scene at offenders home or

elsewhere, offence committed with others or alone). Axis I substance abuse has been

scored on the basis of the official diagnosis in the final report. In Dutch forensic

practice, however, substance abuse is usually assessed only when judged clinically

relevant for the index crime. As a result, there may be some underreportingcompared to actual substance abuse (e.g. nicotine abuse). For first offenders, the age

at the time of the assessment has been taken as a proxy for age at first conviction. For

all defendants charged with murder or manslaughter, the variable attempted

murder indicated whether the charge involved attemptedmurder/manslaughter (i.e.

not lethal to the victim) or actual murder/manslaughter (i.e. lethal to the victim).

The MMPI-2 (Hathaway & McKinley, 1989) is a self-report questionnaire

consisting of 567 true or false questions designed to detect psychopathologic

symptoms in psychiatric patients. Included in this study were T-scores on three

validity scales as well as the basic clinical scales. The validity scales are L (lie,

unsophisticated lying), F (low frequency, indicating the tendency to answer

affirmative to items rarely endorsed by normal people; F also indicates

psychopathology), and K (correction, sophisticated lying). Clinical scales include

Hs (hypochondrias), D (depression), Hy (hysteria), Pd (psychopathic deviation), Mf

(masculinityfemininity, measuring the subjects identification with traditional

gender roles), Pa (paranoia), Pt (psychasthenia, indicating inter alia compulsions,

obsessions, abnormal fears and difficulties in concentration), Sc (schizophrenia), Ma

(hypomania), and Si (social introversion).

Statistical Analyses

Univariate analyses included chi-square statistics or ANOVA for all variables. Two

multivariate stepwise forward multinominal logistic regression analyses were also

conducted. A stepwise regression analysis involves a procedure in which in each step

a variable is entered into the model that best improves the explanatory power of the

model and, consecutively, the exclusion of variables that have been entered before

but may have dropped below significance level. When none of the remaining

variables significantly improve the explanatory power of the regression model, theanalysis is terminated. The thresholds for entering and exclusion were both set to

p .05. First, a regression analysis was conducted to which all variables were

presented for potentially entering the regression model.(Regression 1). This

regression analysis was conducted in order to exclude effects of mutually related

variables and hence overreporting of significant effects based on univariate analyses.


DOI: 10.1002/bsl



7/13

Second, a regression analysis was conducted that started with a basic model

including the two diagnostic variables (Axis I psychiatric symptoms and Axis II

personality disorders) and subsequently presenting all MMPI-2 variables but only

letting enter those MMPI-2 variables that significantly improved the power of themodel (Regression 2). This latter analysis was conducted to test specifically for the

hypothesis that relationships between MMPI-2 profiles and criminal responsibility

are in fact mediated by underlying pathology and that MMPI-2 profiles

consequently do not provide additional information above diagnostic information.

All analyses have been conducted using SPSS 12.0.1; significance levels were set to

.05, two tailed.

RESULTS

The first row of Table 1 presents the frequencies of the different degrees of criminal

responsibility for the total number of subjects. The group of subjects with total

absence of responsibility (N11, 4.4%) has been excluded from subsequent

analyses as their number was too small for statistical analysis.

Table 1 also presents the results of the univariate analyses. The presence of Axis I

psychiatric symptoms was clearly associated with an increasing amount of

diminished responsibility. A total number of 61 subjects (26%) received a diagnosis

on Axis I, ranging through psychotic disorders (N19, 31%), affective disorders

(N14, 23%), paraphilias (N14, 23%), autistiform disorders (N5, 8%),

ADHD (N5, 8%), and other disorders (N4, 7%). The majority of subjects were

diagnosed with an Axis II personality disorder (N186, 78%). Axis II personality

disorders are primarily found in the diminished responsibility group (i.e. the middle

of the responsibility spectrum); almost all offenders (91.9%) who were judged by

experts to have diminished responsibility were diagnosed as having an Axis II

disorder. The cultural background of the defendant was significantly related to the

experts opinion on criminal responsibility. In line with earlier findings, cultural

minorities were increasingly less represented in the more diminished responsibility

groups. The index offence itself did not turn out to differentiate, but (significant and

borderline significant) differences were found for all crime related characteristics.Generally, defendants who offended at their own homes were thought to be less

responsible than those who offended elsewhere. Those who offended alone were

considered less responsible than those who offended together with others. Also

weapon choice turned out to be related to criminal responsibility. In particular, the

use of firearms was associated with more responsibility.

The results of ANOVA tests concerning the MMPI-2 profiles demonstrate that

the mean scores on some MMPI-2 scales differed between the responsibility groups.

Differences were also found for the validity scales F and K. Social introversion was

the only clinical scale that differed significantly between the groups, although the

psychopathic deviation and schizophrenia scales have low p-values as well.The results of the two logistic regression analyses are presented in Table 2. The

first regression analysis (Regression 1) involved an analysis to which all variables

were presented, with the degree of responsibility being the dependent variable.

Regression 1 resulted in a final model including six variables, which was highly

significant (x2(27, N221)242.782, p< .000, Nagelkerke pseudo-R2 .709).


DOI: 10.1002/bsl



8/13

Table1.Clinical,dem

ographic,crimerelated,andMM

PI-2variablesbydegreeofcriminalresponsibility

Variable

Complete

responsibility

Slightly

diminished

responsibility

Diminished

responsibility

Severely

diminished

responsibility

x2

orF

df

p

N(%oftotalsample)

27(10.8%)

52(20.9%)

135(54.2%)

24(9.6%)

AxisIpsychiatricsymptoms

1(3.7%)

9(17.3%)

35

(25.9%)

16(66.7%)

29.91

3

.000***

AxisIIpersonality

disorder

11(40.7%)

35(67.3%)

124(91.9%)

16(66.7%)

42.41

3

.000***

AxisIsubstanceu

se

7(25.9%)

14(26.9%)

4

0(29.6)

10(41.7%)

2.02

3

.569

Agedefendant

32.4

31.6

33.8

32.5

.73

3,234

.534

Priorconvictions

3.67

4.27

4.36

4.46

.13

3,234

.945

Ageatfirstoffence

24.52

26.75

26.02

25.00

.37

3,234

.776

Firstoffender

8(29.6%)

15(28.8%)

37

(27.4%)

6(25.0%)

.18

3

.981

Culturalminority

14(51.9%)

25(48.1%)

33

(24.4%)

7(29.2%)

14.27

3

.003**

Ind.off.

7.44

6

.282

(Att.)murder/m

ansl.

19(70.4%)

30(57.7%)

7

1(52.6)

14(58.3)

Rape/indecenta

ssault

4(14.8%)

7(13.5%)

37

(27.4%)

5(20.8%)

Otheroffence

4(14.8%)

15(28.8%)

27

(20.0%)

5(20.8%)

Attemptedmurder(non-fatal)

1(5.3%)

9(30.0%)

25

(35.2%)

5(35.7%)

6.69

3

.082

Acquaintedvictim

18(69.2%)

27(52.9%)

92

(73.0%)

16(69.6%)

6.77

3

.080

Weapon

39.88

9

.000***

None(onlyphysical)

6(23.1%)

10(19.2%)

51

(38.3%)

11(45.8%)

Knife

8(30.8%)

15(28.8%)

33

(24.8%)

8(33.3%)

Firearms

8(30.8%)

18(34.6%)

7

(5.3%)

2(8.3%)

Other

4(15.4%)

9(17.3%)

42

(31.6%)

3(12.5%)

Offenceatoffende

rshome

4(14.8%)

8(15.4%)

45

(33.3%)

8(33.3%)

8.75

3

.033*

Offencewithothers

9(33.3%)

14(27.5%)

11(8.1%)

4(16.7%)

17.18

3

.001***

L

54

54

54

56

.15

3,234

.928

F

54

57

65

65

4.93

3,234

.002**

K

59

56

52

50

3.27

3,234

.022*

1(Hs)

57

58

56

57

.55

3,234

.648

2(D)

55

59

60

59

.76

3,234

.517

3(Hy)

60

62

60

59

.36

3,234

.783

4(Pd)

73

74

77

70

2.39

3,234

.070

5(Mf)

50

54

54

55

1.62

3,234

.186

6(Pa)

67

67

69

69

.70

3,234

.551

(Continues)


DOI: 10.1002/bsl



9/13

Table1.(Con

tinued)

Variable

Complete

responsibility

Slig

htly

dimin

ished

respon

sibility

Diminished

responsibility

Severely

diminished

responsibility

x2

orF

df

p

7(Pt)

59

6

3

62

62

.76

3,234

.518

8(Sc)

61

6

1

65

66

2.16

3,234

.093

9(Ma)

57

5

5

57

61

1.30

3,234

.275

0(Si)

47

4

7

52

53

3.08

3,234

.028*

Notpresentedintable:totalabsenceofresponsibilityN

11(4.4%).Thisgrouphasbeenexcludedfromsubsequent

analyses.

Ind.off.indexoffence,(att.)murder/mansl.(att

empted)murder/manslaughter.M

MPI-2variables(T-scores):L,lie,unsophisticatedlying;F,lowfrequency;K,

correction;Hs,hypochondrias;D,depression;Hy,hysteria;Pd,psychopathicdeviation

;Mf,masculinityfemininity;Pa,paranoia;Pt,psychasthenia;Sc,schizophrenia;

Ma,hypomania;S

i,socialintroversion.

*p

Documents

-- Factors Associated With Experts' Opinions Regarding Criminal Responsibility in the Ne