You can download a PDF version of this article here

The situation for the mentally ill in NSW prisons has been described as terrible, except when compared with almost everywhere else. We are not the worst place for overcrowding and under-treating. I have recently been in California, where there was a serious proposal to release 30,000 prisoners because the state could no longer afford to house them.
Although the topic I was sent was, ‘Should Sentencing Fit the Crime?’, perhaps we should look at those whom we have in prison and ask, instead, ‘Should Sentencing Fit the Criminal?’

Are our prisons the new psychiatric hospitals?
I propose to start with the question, ‘Are our prisons the new psychiatrist hospitals?’, by looking back to the Penrose hypothesis from the 1930s to see if the population previously housed in psychiatric hospitals are now housed in prisons, to review what we know about mental disorder amongst prisoners and whether or not treatment makes any difference to the likelihood of committing a crime or of returning to prison. Having worked within prisons for so long, I will finish up with a few comments on what I think is wrong with our system, and with some suggestions on how we could improve the situation.

The Penrose hypothesis
I want to mention LS Penrose because he is still very influential. A Google search will reveal quite a few scholarly articles about his theory, even in the past year. An article was published in the Nordic Journal of Psychiatry earlier this year. He was the sort of person who would make you proud to be English. He was a Quaker, an ambulanceman in World War I and read mathematics at Cambridge. As he was interested in psychology, he studied Medicine and made some important discoveries such as the inheritance of mental retardation and the genetic abnormality in Down’s Syndrome. He worked out the ‘Penrose Method’, a ratio based on the square root of a population – used to allocate the seats in the European Parliament.
He is probably best known for the ‘Penrose hypothesis’, published in the British Journal of Medical Psychology in 1939, which holds that in any population there is a fixed number, about 1 % of the population, who require institutional care. He based the theory on a comparison of the number of prison inmates and psychiatric hospital populations of 18 European countries during the 1930s, including the Baltic states and Nazi Germany. He found an inverse relationship between the numbers of prisoners and psychiatric hospital patients.
His conclusion is still widely believed, because longitudinal studies in a number of countries have found that, as psychiatric hospital populations went down, prison populations went up. However, his study is mainly of historical interest, because it was performed near the high tide mark of psychiatric hospital populations, shortly before the advent of effective treatment for psychotic illness, and long before the social changes that led to the steady rise in prison populations in some countries.

Penrose replicated in 2004
Despite the many longitudinal studies citing Penrose, his original cross-sectional study had not been replicated until recently, when my colleague Matthew Large found data from the original 18 countries (now 16) and also in 158 countries worldwide – all done in one night thanks to the miracle of the Internet. We obtained the prison populations, psychiatric hospital beds, GDP and crime rates. Our study was published this year in the journal Psychology and Psychotherapy, research and practice (PAPTRAP), the successor to the British Journal of Medical Psychology.
These are the findings for the 18 European countries in 1939 and 2004.
Beds v Inmates in the 18 European countries
1939 2004
R = -0.61 R = 0.1
R2 = 0.37 R2 = 0.01
P = 0.007 P = 0.71
The statistic is a correlation, or the closeness of the data points to a line of best fit between the data points. Re-analysis using modern statistical methods confirms that Penrose was correct in 1939, but that there is no association in 2004 in the same European countries. However, the situation in 158 low and middle income (LAMI) countries is quite different. In 2004, there is actually a positive correlation between the numbers of prisoners and psychiatric hospital patients, rather than the inverse relationship suggested by Penrose. Moreover, it is very significantly related to GDP. In other words, it is very expensive to keep people in any kind of institutional care. As national income increases, countries are more able to afford to open both kinds of institutions. Also, some countries, mainly former British colonies, seem to be better at it, or perhaps inherited these institutions in working order on gaining their independence.

The ratio of prisoners to patients
There is a wide variation in prisoner to patient ratios in advanced countries. In the EU, Canada and Japan, there are more psychiatric hospital beds than prison beds. There are variations in the way hospital beds are counted, for example, beds for people with intellectual disabilities might be within institutions or, as in NSW, might be counted separately. Even so, the differences are very great, with Britain, Australia and New Zealand having more than twice the number of prisoners, and in the USA nine times as many prisoners as psychiatric inpatients. The differences appear to be due to differences in our tolerance for locking people up and also in our willingness as a community to bear the expense. The United States has a staggering number of prisoners, almost 25 percent of the world’s official prisoners.

Crime rates and prisoner numbers
We also examined whether or not there was a relationship between crime rates and the numbers of prisoners in developed countries, and found no relationship. It has been argued that the increase in prison populations might be reducing crime through reduced opportunity. However, the very wide variations in the numbers of prisoners appear to be due to variations in community tolerance for locking people up rather than in response to the number of crimes. This is borne out by the differences in the results of surveys of community perception of crime and actual crime rates.
In summary, an examination of Penrose’s hypothesis in advanced countries today finds that imprisonment rates are due to different thresholds for custodial punishment, rather than due to differing crime rates, and that there is no relationship between the number of psychiatric hospital beds and the number of prisoners.
This next slide shows you the number of prisoners in the United States over the last century. There is a distortion because it doesn’t show the growth in population, but you can see that the United States had a similar number of prisoners to European countries until about 1980, but then came the ‘war on drugs’ and that’s when things really took off, to the point that the number of prisoners could account for another 1 or 2 % of unemployment if the rate of imprisonment in the US were the same as in other advanced countries.
Because we have been happy to join the United States in any wars, including the war on drugs, we have followed the US in the rise in the number of prisoners. The next slide doesn’t look quite as dramatic as the American data because it shows imprisonment rates, rather than the total number of prisoner, but it does reflect a big jump in the total number of prisoners in Australia in the last 20 years.
You’ll see that the proposal to get rid of sentences of under six months is laughable when you look at the steep rise in the bottom line of the graph, which is the number of remand prisoners. You can see that the reduction in the availability of bail, which is a direct reflection of community dissatisfaction with recidivism, is one reason for the steep rise in the number of prisoners.

Prisoners with psychiatric disorders
The next few slides present data on the number of people in New South Wales gaols believed to have psychiatric disorders. You can see that 34% of men and 20% of women have had admissions to psychiatric hospitals. That figure includes admissions to the prison hospital, so that the men outnumber the women in that statistic. But in nearly all other statistics, the female prisoners have far more psychiatric disorders, with more taking psychotropic medication and more reporting a history of previous suicide attempts and deliberate self-harm.
Looking at the diagnoses, we find that prisoners are a very depressed population, with nearly 20% of men and 30% of women reporting depression, by contrast with community rates of three and four percent respectively. It is also a very anxious population, probably because of recent circumstances and withdrawal from stimulant drugs (this is taken from a reception inmate survey). The reported rates of substance use disorder of 50 and 60% is probably an understatement.
Regarding the number of prisoners with psychotic illness, mainly schizophrenia, the disorder causing the most disability, at least 5% have a definite diagnosis and 1.9 percent have a probable psychotic illness. As this study is based on interviews of the prisoners themselves and on a review of all available records, we believe that the figures are accurate. However, any screening instrument based on self-reporting is going to miss many prisoners with psychosis, because they often don’t recognise or report their symptoms, and it is not uncommon to come across people who have been quietly mentally ill for long periods in prison but have not come to the attention of health services.
As the community prevalence of psychotic illness is about 0.5%, a finding that between 5 and 7% of NSW prisoners have a psychotic illness reflects a rate that between 10 and 14 times higher than in the wider community. In that sense, the gaols are a big psychiatric hospital, as these figures translate to between 500-700 prisoners with schizophrenia in NSW gaols. This estimate does not include forensic patients, because anyone found not guilty on the grounds of mental illness would go to a gazetted hospital when a bed becomes available. Prisoners with schizophrenia are not always receiving treatment, especially when they move to lower security gaols away from Sydney, which don’t have the same level of mental health services.

Would prisoners be treated if there were more psychiatric hospital beds?
The question is: would those prisoners be treated in psychiatric hospitals if the beds were available? The answer is not many, and not in recent times. Our de-institutionalisation happened long before the rise in prison numbers. For example, the population of Rozelle Hospital peaked at about 4,000 patients soon after the war and had fallen to about 400 when I worked there as a psychiatric nurse in 1977. By the time I started training in psychiatry in 1988, soon after the much-maligned Richmond Report, the number of beds had fallen to about 180, so you can see that the door was opened long after the horse had bolted. What’s more, many of the patients who were discharged from long-term hospitalisation made their way to the inner-city homeless accommodation such as Matthew Talbot Hostel and Edward Eager Lodge. The survey of the homeless mentally ill performed in 1990 did not find many patients who had gone from long term hospital care to regular prison terms.

Inter-generational trauma
A stronger effect on offending than de-institutionalisation is inter-generational trauma. Most prisoners report some kind of neglect or trauma in early life. It is unusual to come across a prisoner who had a stable and happy childhood and there is usually an obvious explanation for why they became involved in drugs and went into a lifestyle of not respecting the rights of others. There has always been a criminal sub-culture. I’m sure that criminal families date back to the First Fleet. However, the advent of heroin abuse and social security payments in Aboriginal communities from about 1970 seems to have increased the rate of family disruption in some areas. One hears more and more of people who have been brought up by substance-abusing parents in dysfunctional communities.
I am reminded of an apocryphal tale of a boy visiting his father in prison being asked what he wanted to be when he grew up. He replied that he wanted to be the sweeper in 3 Wing. This reflects one toxic effect of prison and criminal subculture, because prisoners often have many children and are, on the whole, poor at parenting. Any measure to reduce the rate of crime and recidivism will have to look at inter-generational neglect and abuse, especially that arising from parental substance-abuse.

The mental health of Aboriginal prisoners
We can’t talk about the mental health needs of our prisoners without considering the effect of Aboriginal heritage, because 15% of our prisoners describe themselves as Aboriginal. I was involved in a criminal trial in Alice Springs where they needed four different interpreters because few of the witnesses spoke English. The events were interpreted according to traditional ideas, like the influence of kadaicha men. However, when interviewing people of Aboriginal background in NSW gaols, I don’t find much Aboriginal language or Aboriginal taboos. Being Aboriginal in NSW gaols seems to reflect coming from the lowest socio-economic group, as well as growing up in some of the most dysfunctional communities, such as in Redfern and in some country towns, where a lot of the kids are exposed to neglect and abuse. There is more brain injury, including injury from solvent abuse, and a lot more medical problems, and there might also be more psychotic illness. A recently published study from New Zealand shows a three times higher rate of psychosis among Maori. There are also some interesting data about the effect of childhood neglect and trauma on the risk of psychosis over and above any genetic or any other biological effect, suggesting that early life experiences might greatly increase the risk of psychosis. So, given the high rates of all forms of psychiatric disorder and the lack of community support for a law-abiding lifestyle after release from gaol, it is not surprising that recidivism is higher amongst people of Aboriginal background.

The role of substance-abuse
I have mentioned substance-abuse as being the main reason for offending and also for relapsing back into a criminal lifestyle. We also know that use of cannabis and amphetamine induces psychotic illness and causes depression and anxiety. Another aspect of the abuse of these drugs is that they greatly multiply the risk of violence by people with a psychiatric disorder.

Opiate dependence and its long-term management
It is now recognised that opiate dependence is a chronic and relapsing condition and that we should be thinking about treating opiate addicts long-term, including with Methadone maintenance outside gaol. There was previously an optimism about weaning people off Methadone in the hope that they wouldn’t need it any more, but the risk of relapse and the risk of fatal overdose is so much greater once you have weaned a long-term user off Methadone that we should just accept that we have to treat opiate dependence indefinitely.

Positive aspects of imprisonment
A point worth mentioning is that gaol can actually be a positive experience for some prisoners. It is probably one of the few places where substance abusers can achieve abstinence. There is a lot of talk about drugs being available in gaol, but prisoners are routinely drug-tested with relatively low rates of positive tests. There are severe sanctions for using drugs in prison and the rate of drug use in gaol is probably much lower than it was. Some prisoners receive good health care and gaol might be the only place they get dental care. They are certainly well fed compared with the amphetamine diet of many ex-prisoners in the community. People often come out of gaol heavily muscled because they have nothing to do but train with weights. Considering the potential changes that could be made on a captive population, it is a shame we have not made more of the opportunity to treat the psychological and lifestyle problems of many prisoners. This is reflected in NSW having the second highest recidivism rate in Australia after the Northern Territory. In that sense, the productivity of NSW prisons is low.

Do mental health services have a role in preventing crime?
We know that most crime by people who have a recognised mental illness occurs when they are not treated. Even continuing psychological care means that there might be some opportunities to persuade people not to make the wrong choices. However, we also know that, in community health centres and other primary care situations, there can be a nihilistic attitude to treating substance use, which seems to be one of the main pathways to relapse and recidivism. We therefore need greater coordination between probation and parole services and health care agencies.

Supported housing
A key part of holistic treatment is varying levels of supported housing. The type of housing is crucial to treating chronic mental illness, because the disability arising from mental illness makes it very difficult to properly arrange housing. It is hard to provide continuing medical care and continuing rehabilitation services unless one can find the patient or know that they are in a situation where they are being treated.

Violence by untreated people with a psychosis
Matthew Large and I have shown, for the first time, that most of the serious violence committed by the mentally ill occurs before they are first treated for mental illness. For example, 40 percent of homicides due to psychotic illness worldwide occur before the first treatment. Later in the illness, which is usually life-long, the rate of homicide is much lower. Earlier treatment would, therefore, prevent a significant amount of violence by the mentally ill, and that’s a significant proportion of all violence, as between 6% and 8% of homicides and maybe 10% of non-lethal assaults are committed by people with a diagnosis of schizophrenia.
As an example, I refer you to our recent study of homicide by people with mental disorders in England and Wales from the mid 1950s through to 2007. Homicides by the mentally ill rose dramatically with the rising rates of other homicides until the late 1970s, but fell steadily from that time onwards, while the total homicide rate in the UK continued to rise. This decline was probably due to better treatment.

NSW gaols
I want to share my experience of working in NSW gaols for the last 15 years. As far as I can tell, there has been a cynical pact between government, the government service – it is not the public service, it’s the government service nowadays – and the prison officers’ union. Between the three of them, the gaols have become less efficient and less effective. It is much harder to see a patient at a clinic. It is much harder to get in and out of a gaol for a professional visit. Prisoners are constantly being moved, which disrupts care and makes it much harder to keep your patient in a therapeutic setting. There is an excessive focus on security at the expense of rehabilitation programs. A lot of the best programs have been virtually closed down, for example, works release and day leave towards release. You have only to look at the fuss created in trying to find release programs for serious sex offenders to realise how difficult it is for Corrective Services to arrange any kind of graded release program.
It is not surprising when you look at the way the prisons are staffed. There is a kind of para-military structure. It seems that there are more and more high-ranking officers and more conflict between the unions and the management over manning levels, which are presented in the guise of occupational health and safety concerns. The pre-occupation with security can also be excessive, as some wings have three and four musters per day while doctors sit idle at clinics unable to see prisoners. The excessive focus on security is one reason for the low productivity of gaols and translates directly to higher recidivism, more victims and more community anger.

Suggestions for improvement
If I were to make any suggestions on how to improve things and reduce recidivism, the crucial area would be the interface between prison and the community. That’s where the Department of Corrective Services and the Parole Service seem to have the greatest difficulty. They are so keen to avoid any kind of embarrassment or mishap that it becomes hard for them to allow people day release or intermediate placement.
Some kind of release planning and step-down care is crucial for disabled patients, such as the 500 or 700 inmates with schizophrenia. What happens to them is that they go through from maximum security, to B classification, to C classification and then get released from prisons like Glen Innes and Kirkonnell, with seven days’ supply of medication and a pension cheque and are expected to organise their own lives. It is quite common to hear of people who have re-offended on their way back to Sydney after release from remote prisons.
There really should be a lot more attention given to the interface between prison and the community and far more graded release to the prisoner’s eventual place of residence. That’s what you try to do for patients discharged from psychiatric hospitals and that’s what we should be trying to do as a routine in gaols.
There should also be a presumption for drug treatment, perhaps to leave gaol to go to a contained rehabilitation program before returning to the community. That would apply to the 60% or so of prisoners who are in gaol because of drug-related problems and who carry the greatest risk of recidivism.
One of the things about the inefficiency of providing therapeutic programs in prison is the limited access to prisoners because of the large amount of time spent in lock downs. One way of getting around this is to introduce a lot more automated programs. The way I envisage it is new kinds of cells with flat screen TVs behind thick perspex. You spend a lot of time in your cell these days, which is often spent talking to your cell mate about past crimes. The amount of time available to many prisoners for face-to-face counselling, or for education or vocational training, or even work is quite limited. If automated programs were available, we could ensure that prisoners had completed the modules necessary to be eligible for parole.
I would suggest that prisoners have some kind of remote control, so that they have to complete an hour or two of relevant programs, say literacy, or drug and alcohol, or good parenting or cognitive therapy for anxiety, or whatever, before they could get their sport or entertainment. Fantastic programs are available for many psychiatric disorders. I would like to be involved in devising programs which could be delivered in this way to every prisoner, so that each would have completed a drug and alcohol course, with every prisoner at least having the information part of any course. The efficacy of automated programs offered over the internet from St Vincent’s is impressive, and this sort of program might increase the productivity of our prisons and reduce recidivism.
I think we need more outside prison schemes. I am really talking about post-release schemes, after detoxification and rehabilitation. Perhaps we should separate the organisation of security from rehabilitation and of custody from parole. Parole officers often have a welfare background and they go to people’s homes and see immediate social work needs. However, they are told in no uncertain terms that their job is supervision and security. Perhaps we need to separate these two roles.
Lastly, there are always calls for more research, but I would call for more interventional research. Although there are nearly 400 published studies on recidivism from Australian jurisdictions, I have not found one good quality randomised control trial of an intervention. The first one I would like to do is on the effect of automatic programs.

About Nielssen

Dr Nielssen graduated in Medicine from the University of Sydney in 1985, obtained a Master of Criminology degree and has specialised in psychiatry. He has worked in the UK and California and for 15 years as a visiting psychiatrist at Long Bay. He holds an appointment at St Vincent’s Hospital. He is a member of the Mental Health Review Tribunal. He also does a clinic at the Matthew Talbot hostel. He has conducted research into related topics such as mental illness in prisons, the epidemiology of homicide and the effects of changes in mental health law.


Scientific Meetings, Dinners and events.


Search Documents and Downloads...