Ockham's Razor Ockham's Razor Radio National   

Email Ockham's Razor
Tape Sales
Tuning In Details
About Ockham's Razor
Story Index - Transcripts
Ockham's Razor Home

ABC Science
 All in the Mind
 The Buzz
 Health Report
 In Conversation
 Science Show
 The Lab
 Health Matters
 Science News

Radio National Home

Broadcast Sunday 15 June 2003 
with Robyn Williams

The Science of Crime Control


Professor Peter Grabosky from the Australian National University's Regulatory Institutions Network, looks at whether science can inform crime control. Many programs have been introduced overseas with mixed success and given that the Australian government, both Commonwealth and State, spend billions of tax dollars each year on criminal justice and crime prevention, he challenges them to identify what does work, what is cost effective and then to invest our tax dollars there.


Robyn Williams: It’s a crime; or is it? Can you be scientific about lawlessness in the same way, for instance as you can scientifically examine ill health or its treatment? Because if you can, then many things follow. Not least how to prevent criminality, or at least, lessen the damage.

Well Peter Grabosky is at the Research School of Social Sciences at the Australia National University in Canberra, and he’s been looking at the possibilities and so far, success.

Peter Grabosky: Few areas of public policy are as passionately contested as crime and what to do about it. Unfortunately, solutions to crime are more often grounded in ideology than in science, and much debate on crime can better be described as a dialogue of the deaf, or a religious debate, than as rational discourse.

Imagine if a majority of students at a local primary school were suddenly stricken with encephalitis. Would we look to ideology, to folk remedies, or to religion for a treatment, or for measures to prevent a reoccurrence? Hardly. You can bet that we would look towards evidence-based medicine for an appropriate response. So why can’t we do the same thing for crime?

Why is it that new pharmaceutical drugs must undergo the most rigorous testing for safety and effectiveness before they can be put on the market, but programs and policies for crime control are often implemented as an act of faith?

In other words, can science inform crime control?

The good news is that we have begun to make a start. Social scientists in Australia and elsewhere in the world have begun to embrace evidence-based crime control, and have brought the most rigorous research techniques available to bear on the question of ‘What works?’

The randomised controlled trial is often referred to as the ‘gold standard’ of evaluation. It involves assigning subjects randomly to treatment and control groups, and comparing the outcomes recorded for each group. Because participants are randomly assigned, significant differences in outcome are attributable to the treatment as opposed to any other factor. This is the basic method used to test new medical procedures and pharmaceutical drugs for safety and effectiveness. You can bet that the safety and effectiveness of any prescription drug available in Australia has been tested in a randomised controlled trial.

A number of randomised controlled trials have revealed positive outcomes, real achievements in crime prevention and control.

For example, peri-natal home visits to pregnant teenage girls from disadvantaged backgrounds by specially trained and qualified paediatric nurses not only result in higher birthweight babies, but also reduce the risk that these babies will suffer abuse or neglect. Moreover, 16 years later, these children were less likely to be arrested than were their counterparts in the control group.

The celebrated Perry Preschool program, established in Michigan for children from disadvantaged backgrounds over 40 years ago, provided preschool enrichment classes, combined with weekly home visits by program staff. This has also been evaluated in a randomised controlled trial. Long-term follow ups revealed that program participants have significantly lower juvenile and adult arrest rates, but also significantly higher rates of high school completion, tertiary education, employment and earnings. In addition to its proven effectiveness, the Perry program has passed a cost/benefit analysis with flying colours. Total benefits have been estimated at three times the program costs.

Woolfenden and her colleagues reviewed eight separate randomised controlled trials to determine the impact of parenting and family support programs on young people’s conduct disorder or delinquency. Family and parenting interventions significantly reduced the time spent by juvenile delinquents in institutions. There was also a significant reduction in the risk of a juvenile delinquent being re-arrested, and in their rate of arrests one to three years later.

Enhanced street lighting is another crime prevention strategy that has been evaluated, and demonstrated to work. Professors David Farrington of Cambridge University, and Brandon Welsh of the University of Massachusetts, prepared a systematic review of 13 separate studies conducted in the United Kingdom and the United States. They reported that enhanced street lighting reduces crime by about 20%.

While most of the systematic reviews are based on evaluations done in the Northern Hemisphere, researchers in Australia have undertaken one such series of experiments. I refer to the experiments in restorative justice done by the Australian National University. People charged with middle-range property or violent crimes, or with drink driving in Canberra were randomly assigned either to special restorative justice conferences or to court appearances in the usual way. Restorative justice, which entails victims and offenders, together with their families and friends, meeting to discuss the crime and its consequences, was shown to produce no significant difference in later reoffending rates, compared to court for the property and drink driving offenders, but a big drop in reoffending by violent offenders, about 40% less than those who went to court. Of arguably greater significance was the fact that victims of crime were more satisfied with conferences than with conventional court proceedings.

Of course the flip side of success stories is, What doesn’t work? This is an important question, because some solutions to crime problems are not solutions at all. They are either ineffective, achieving nothing, or counterproductive, doing more harm than good. Let me give you two examples:

In the United States, beginning in the late 1930s, a state-of-the-art delinquency prevention program was established in the Boston area. The Cambridge-Somerville Youth Study was everybody’s dream. Keeping kids from poor neighbourhoods out of jail, providing them with mentors, counselling and recreational opportunities to enable them to keep themselves on track and make something of their lives; sounds great, doesn’t it?

Boys were matched to another of similar age, social background, temperament, and biological somatotype, then randomly assigned either to the program or to the control group. Unfortunately, an evaluation by Powers and Witmer and a follow-up by Professor Joan McCord produced some surprising results. Participants in the program were more likely to be convicted of serious street crimes; they had died an average of five years younger, and they were more likely to have received a medical diagnosis as alcoholic, schizophrenic, or manic depressive.

Here’s another example: Scared Straight. The basic model of Scared Straight has young offenders taken on a tour of a prison and sternly lectured to by inmates, who urge them to ‘make something of your lives and don’t turn out like us.’ Scared Straight programs had a strong political resonance in the United States, where they were adopted enthusiastically in over 30 locations.

Anyhow, nine separate Scared Straight programs were evaluated in randomised controlled trials. A review of these was conducted by Anthony Petrosino of the American Academy of Arts & Sciences, and his colleagues. Their analysis showed that participants in the Scared Straight programs were more likely to re-offend than were controls. This is pretty powerful evidence, and I would certainly not hesitate to advise any Minister or public servant who may be enamoured of the Scared Straight model to forget about it.

By now, you can begin to appreciate the contribution that rigorous method can make to crime control. To further the cause of evidence-based social policy, including crime control, a group of scholars from around the world established the Campbell Collaboration in the year 2000. Modelled on the Cochrane Collaboration, which publishes systematic reviews on the effects of medical interventions, the Campbell Collaboration will prepare, update, and disseminate systematic reviews of the literature on what works in three areas of public policy: education, social welfare and crime control.

Systematic reviews have explicit criteria for including or excluding studies (generally only the most rigorous studies are included). They are based on extensive searches for eligible evaluation reports from all over the world. Campbell reviews are based on careful extraction and coding of key features of studies, and include a structured and detailed report on the methods and conclusions of the review. As new evaluations are conducted, the Campbell Collaboration will continually update and electronically disseminate their systematic reviews.

So if you want to know whether programs for reducing school truancy work or not, or what are the most appropriate interventions for the homeless mentally ill, you can visit the Campbell website.

Reviews of some 30 different types of criminal justice interventions will also be published over the next couple of years. In addition to those on street lighting and Scared Straight programs that I have already mentioned, they will include programs for domestic violence, correctional boot camps, hot-spot policing, electronic monitoring of people on probation or parole, and prison based drug treatment programs, among many others.

The good news for Australia is that we too are represented on the Campbell Crime and Justice Steering Group and can nominate categories of crime control programs for systematic review, and the Campbell Crime and Justice website is hosted by the Australian Institute of Criminology.

It may be important to ask more than just whether a program works, or not. Why for example, does participation in a Scared Straight program increase the risk of re-offending? One may surmise that the young participants perceive the criminals who lecture them as role models. Or they may think to themselves, ‘I’m tough, I can handle this.’ In any event, bringing a group of delinquents together for any purpose may help reinforce their delinquent self image; indeed, it may inspire a degree of competition to see who can be the ‘baddest’.

We know that crime control interventions can have unintended consequences. But not all of these are adverse. Some interventions produce unanticipated benefits. The systematic review of research on street lighting revealed that areas with enhanced lighting at night also experienced less crime during daylight hours than did control locations. ‘What’s going on here?’ you might ask. Well, it has been suggested that street lighting does more than discourage crime through increasing surveillance of potential offenders by improving visibility, and through increasing the number of people on the street at night. The installation of new lighting may give a signal that there is increased community investment in the area, greater pride, cohesiveness and informal social control, 24 hours a day.

There is also the question of differential treatment effects. We all know that some people suffer adverse effects from the same drugs that work splendidly for others. The same thing can happen with crime control. Here’s an example: The effect of mandatory arrest for domestic violence was the subject of a randomised controlled trial in Minneapolis some years back. Initial findings suggested that arrest was a deterrent; it was less likely to lead to subsequent abuse. Policies of mandatory arrest were embraced around the world as the best way to respond to domestic violence. Unfortunately, subsequent replications of the Minneapolis Experiment were inconclusive. Then, closer analysis of the data revealed that mandatory arrest did have a deterrent effect on Caucasians who were employed, but was more likely to lead to subsequent abuse by unemployed African-American men.

Such divergent findings may be unsettling to lawyers and other who embrace the principles of the rule of law and equal justice. But they are important to know.

Australian governments spend billions of our tax dollars each on criminal justice and crime prevention. With what effect? What return are we getting on this investment? Applying scientific method and a bit of economics, can answer these questions.

Some social programs simply don’t work. Others may even have harmful effects. But there are those programs that not only do work, they even pay for themselves, by helping people to become healthy, productive taxpaying citizens rather than a drain on our social welfare and criminal justice systems. I challenge our governments, Commonwealth, State and Territory, to identify what does work, what is cost effective, and to invest our tax dollars there.

Robyn Williams: Put your money where it matters. The future of crime control. Peter Grabosky is a Professor at the Research School of Social Sciences at the Australian National University in Canberra.

Next week, Ockham’s Razor comes from Adelaide, where Bill Hall wants to tell you about ‘kitchenalia’, the science and technology in the kitchen, and why you may want to collect it.

I’m Robyn Williams.


Professor Peter Grabosky
Research School of Social Sciences,
Australian National University,
Canberra ACT 0200

More information:

Main Campbell Website

Campbell Crime and Justice Website

Story Index

Navigate the Radio National Website...

Search Radio National...

Choose a program...


Schedule   Tune In   Contact Us   About