Mental illness as a cause of offending behaviour? A critical look at prevalence rates.

Setting the scene

Prevalence rates of mental illness in UK prisons are estimated at 90% (Mental Health Foundation, 2011), with comparisons between the prevalence of these common mental health diagnoses within forensic settings and the general population indicate that mental illness is significantly more prevalent among offender groups than in the general population.  

Personality disorder is defined as general ways of thinking that are “characterized by inflexible and enduring behaviour patterns that impair social functioning … longstanding, inflexible, and maladaptive personality traits which impair social and/or occupational functioning” (APA, 2000).

In DSM-IV-TR, personality disorders are divided into three ‘clusters’.  Cluster A, the ‘odd or eccentric’ personality disorders, comprises of paranoid, schizoid and schizotypal personality disorders.  These are characterised by behaviour that could be mistaken for psychosis, such as paranoia and delusional thinking.  However, these behaviours tend to be persistent and pervasive as opposed to the intermittent cycles of behaviour that are symptomatic of the MI diagnosis of schizophrenia.  Cluster B personality disorders are referred to as dramatic and emotional, and include antisocial, borderline, narcissistic and histrionic personality disorders.  These disorders are characterised by what can be volatile and callous behaviour towards others, as well as impulsivity.  Finally, cluster C diagnoses include the dependent, avoidant and obsessive-compulsive personality disorders.  Known as the fearful personality disorders, these are characterised by behaviours resembling anxiety-related disorders, in the same way as cluster A resembles schizophrenia.

Cluster B personality disorders are the most relevant to examine in relation to offending behaviour (Coid et al., 2006), and these disorders will form the basis of the arguments that follow.  The most common personality disorder diagnosis in forensic settings is that of antisocial personality disorder (ASPD).  This is characterised by a lack of empathy or respect to social norms, and is diagnosed in 50-70% of offenders (Davey, 2008).  ASPD is often co-occurring, or at least co-diagnosed, with narcissistic personality disorder.  Heightened feelings of grandeur and importance are symptomatic of narcissistic individuals, who account for 7.5% of the prison population (Singleton et al., 1998). 

Offenders as 'mad' or 'bad'?

Whether or not mental health diagnoses or personality disorders directly cause criminal behaviour is unclear.  Indeed, the ‘mad versus bad’ debate is often referred to when those deemed to pose no risk to society are released from psychiatric units and go on to commit serious criminal offences. 

Take the case of Stephen Griffiths, for instance.  Calling himself “the Crossbow Cannibal” (BBC, 2010) – a nickname that was also adopted by many members of the British tabloid press - Griffiths was convicted of the murder of three women in West Yorkshire in December 2010.  However, in the years preceding these crimes, he was detained in secure units and youth offending institutions on three separate occasions after threatening a shop worker with a knife at the age of 17, and then holding a knife to the throat of another girl five years later.  Whilst being evaluated in Rampton Hospital, he admitted to wanting to become a serial killer in later life, with a psychiatrist suggesting that Griffiths had “a preoccupation with murder – particularly multiple murder" (Johnston, 2010). 

Griffiths was diagnosed as a schizoid psychopath, and admitted into Rampton Hospital.  However, clauses within the Mental Health Act 1983 made it so that individuals who were sectioned could only be detained if their condition could be alleviated or maintained by medical treatment.  As this was not considered to be the case with psychopathic individuals, Griffiths was released.  Since then, the previous Labour government attempted to close this legislative loophole, but the amended Mental Health Act of 2007 again states that psychopathic disorder is an exception to mental illness, meaning that detention of psychopaths is still difficult to justify.

Consideration of the ‘mad versus bad’ debate, however, has led to the formation of specialist dangerous and severe personality disorder (DSPD) units.  One of these, The Peaks, is based at Rampton Hospital, and comprises of seven 10-bedded wards which house patients with severely disordered personalities who have been detained under the Mental Health Act 1983.  This group of patients includes those classified as psychopathic, and are held because of their high propensity for violence, in the interests of public protection.  Hogue et al. (2007) advocate the use of integrated clinical approaches with DSPD patients, and argue against the view that psychopathic individuals are untreatable, suggesting that there is “little empirical support for this stance”.  In addition, treatment programmes in DSPD units, such as the Peaks, are tailored specifically for each offender in order to focus on and address the precise issues that have led to them becoming such high-risk individuals.  This is a significant move in the right direction when compared to the manual-based offending behaviour programmes used within HM Prison Service, where a 'one size fits all' approach to rehabilitation is currently used.

Maniglio (2011) argues that depression and low self-esteem can be a cause of criminality, and cites the example of sexual offending in support of his claims.  His claims provide support to the findings of the widely cited paper by Seto and Lalumiere (2010), which reported significantly lower levels of self-esteem among a sample of sex offenders than in a matched sample of offenders who did not commit sexual crimes.  In addition, it has previously been reported that some anti-depressant drugs reduce deviant sexual desire (Greenberg and Bradford, 1997).

O’Kane and Bentall (2000) studied the role of delusions and hallucinations in relation to crime and observed that paranoid delusions and command hallucinations were most commonly implicated in violent behaviour.  This was the kind of defence used by Peter Sutcliffe, who killed 13 women between the years of 1975-1980.  Sutcliffe’s defence was that he was told to kill prostitutes by a voice coming from a headstone whilst he was working as a gravedigger, and that he believed this was the voice of God.  This is just one example of command hallucinations, when combined with intense religiosity, being a significant contributing factor for violent crime.

Fazel et al. (2010) argue that substance abuse is by far the greatest risk factor related to mental illness and criminality.  They found that schizophrenic patients were three times more likely to commit violent crime if they had a co-morbid substance abuse problem than if they did not, supporting findings reported by Elbogen and Johnson (2009).

In terms of personality disorder, it could easily be argued that criminality is actually one of the diagnostic criterion for ASPD, as demonstrated by criteria A(1), “failure to conform to social norms with respect to lawful behaviours as indicated by repeatedly performing acts that are grounds for arrest”, and A(4), “irritability and aggressiveness, as indicated by repeated physical fights and assaults” (APA, 2000, as cited in Davey, 2008).

Narcissistic personality disorder has been studied widely in relation to its links with aggression.  A key concept in this field was proposed by Baumeister et al. (1996), and is referred to as the threatened egotism hypothesis.  According to this theory, narcissistic individuals require positive evaluation and reinforcement of their own subjective feelings of superiority.  Failure to receive such reinforcement will leave the narcissistic individual tense, and the behaviours that are likely to be displayed in these situations are aggressive and violent acts.  Carrying out such acts then allows the individual to feel powerful, especially if the act is carried out against another person (Maples et al., 2010).  This view is also expressed by Davey (2008), and offered as a possible cause of criminal behaviour by Calhoun et al. (2001).  In addition, Martinez et al. (2008) found that ego threat was not a necessary factor preceding violence in patients with narcissistic personality disorder.  They suggest that a narcissist’s craving for reinforcement of their feelings of power and grandeur are constant, meaning that acts of violence and aggression can occur at any time and without explicit provocation.  Maples et al. (2010) studied trait and pathological narcissism in relation to four types of aggression (laboratory-based, reactive, proactive, and relational).  They found that pathological narcissists (i.e. those meeting DSM-IV-TR criteria for diagnosis of narcissistic personality disorder) were more likely to express all four types of aggression.

Prevalence estimates and research in this area provide a strong argument for a causal link between mental illness and criminal behaviour.  However, all of the studies presented above have used associations between mental health diagnoses and crime to argue that illness is a precursor to the offending act.  However, from a philosophical standpoint, these studies do not present causal links at all, but merely “probabilistic inferences” (Anckarsäter et al., 2009, p.343) that diagnosis may be a contributory factor in offending behaviour.  Mackie (1974) introduced the notion of INUS conditions when referring to cause and effect studies.  It was suggested that what most people consider a ‘cause’ of another thing is in fact and unnecessary, but sufficient, factor in the resulting effect.  This theory of INUS conditions gives rise to the concept of risk factors, which is a particularly common term used in psychology and other social sciences. 

In addition to mental illness only being an INUS risk factor for offending, it must be bore in mind that not all patients with MI diagnoses and personality disorder go on to commit crimes.  If mental illness was a direct cause of criminality, it would be expected that a much higher proportion of the psychiatric population would be convicted of criminal offences.  Personality and circumstantial factors of the majority of psychiatric patients who do not commit offences should be the focus of more research, the findings of which could inform intervention strategies and programmes aimed at reducing offending among groups with additional risk factors.

Philosophical considerations

Upon trying to interpret prevalence rates, experts are faced with two options.  The statistics can be taken at face value, and direct links between mental ill health and crime can be deduced, or they can be viewed critically, and ultimately challenged.

Early criminological theories, such as the biological positivism (Lombroso, 1876), could be used to counter the use of prevalence rates as evidence for mental ill health being a cause of criminality.  Essentially, biological positivist criminology assumes that offenders are born to be criminals. 

Biological positivism was the predominant criminological theory for much of the late 1800s and early twentieth century.  However, the theory does not allow for findings emanating from the age-crime curve, which clearly shows that rates of offending dramatically decrease between the ages of 20 and 25 years.  Surely, if a propensity to criminal behaviour was inbuilt, then a stable rate of offending would be observed throughout the lifespan.  A reasonable explanation for why this stable offending rate is not observed is that young adults tend to start families in their mid-to-late twenties, meaning that opportunities to offend are dramatically reduced.  As this reduction in opportunity is evidently effective in reducing offending, it can be concluded with some confidence that biological positivism cannot adequately explain criminality.

The example of biological positivism is effective in highlighting potential problems in using prevalence of mental disorder, especially personality disorder, as an explanation and cause of crime.  Biological positivism served a function in society as it enabled the majority of society – those who were not convicted criminals – to feel in some way superior to offenders.  This function is still served today, albeit indirectly, through the use of prevalence rates and mental health diagnoses within forensic settings.

In conclusion ...

Reported prevalence rates of mental ill health in prisons and other forensic settings may be misleading in addressing the question of whether or not mental illness or personality disorder actually cause offending behaviour.  Indeed, some classifications of disorder, such as ASPD, lend themselves well to the offender population, making diagnosis of this disorder among violent offenders virtually inescapable.  The previous government’s attempted changes to the Mental Health Act 2007 indicated a shift towards viewing offenders, particularly violent offenders, as mentally ill.  Although this still differentiates the offending population from the rest of society, it does at least offer some hope for treatment.  The formation of DSPD units, such as those described by Hogue et al. (2007) provide much needed optimism for personality disordered offenders, in addition to giving an outwards impression to the general public that violent and dangerous individuals can potentially be treated.  There would appear to be a need to look at root causes of offending in prison offending behaviour programmes instead of purely looking at the symptomatology of particular offenders.  For example, it is necessary to look at the bigger picture when working with an offender with, for example, narcissistic personality disorder, in order to establish why this particular person has offended as opposed to someone with narcissistic personality disorder who has not.

There does appear to be some link between mental ill health and offending behaviour.  However, as argued in Anckarsäter et al. (2009), it is suggested that MI diagnoses should be viewed as risk factors for criminality, as they alone, at least in the majority of cases, do not directly lead to offences being committed.  It is concluded that examination of the wider circumstances leading to criminality should be carried out, and the humanizing of offenders should be encouraged to give the best possible opportunity for successful rehabilitation.

I hope this has given some food for thought.  Remember to comment on the post, debate the material and think about these issues to form your own views.  If you have any issues, questions or comments, just get in touch by email or Twitter.

All the best,

Craig