Find 'em, catch em', put 'em in prison: how to reduce the crime rate

During the 1990s the United States experienced a significant drop in the incidence of most categories of crime. The assault rate in America dropped by more than one-third, burglary rates more than halved, robberies fell by two-thirds and car theft fell by three-quarters.

In Australia, by contrast, burglary and car theft fell only marginally during the 1990s while assault and robbery rates went up. Indeed, the International Crime Victim Survey of 17 countries shows Australians are more at risk than the citizens of most other developed countries. Australia ranks second highest overall (behind England and Wales) on the rate of victimisation, and we score higher than any other country on so-called 'contact crimes' such as robbery and assault.

It would be useful to know how the US managed to reduce criminality during the 1990s.

Explaining crime rate variations

Many different factors have been identified as possible causes for the dramatic fall in the American crime rate. These include the strong economy, the ebbing of the crack-cocaine epidemic and the reduction in the number of young males in the population as the 1960s birth-bulge matured.

Some of these explanations are more convincing than others. It is difficult to see why the strong economy should have reduced crime, for example, when crime increased during other periods of rising prosperity in America (most notably the 1960s). Furthermore, Australia's economy grew even more strongly than America's during the past decade, but crime rates here did not fall.

What we are looking for, then, is something distinctive about America in the 1990s. Two policy changes in particular are worth investigating.

One was the move in New York City and a number of other major centres to what has been called 'Broken Windows' policing. This strategy called for increased police surveillance of behaviour in public places in order to lower the level of official tolerance of relatively minor infractions such as vandalism, vagrancy and graffiti. The theory of Broken Windows held that serious crimes thrive in areas where nobody seems to care about the little things. Thus, by taking action on the small incivilities, the bigger and more serious anti-social behaviours should decline as well.

Economic rationality and criminal behaviour

The other major policy change that occurred in America in the period under review was that more offenders ended up in prison. This policy, too, was driven by an academic theoretical literature, although in this case the theory was more economic than sociological.

In the 1960s, the Chicago economist Gary Becker suggested that crime, like any other 'business', involves a rational calculation by individuals of likely costs, benefits and risks flowing from a given course of action:

"When other variables are held constant, an increase in a person's probability of conviction or punishment if convicted would generally decrease, perhaps substantially, perhaps negligibly, the number of offenses he commits ... a person commits an offense if the expected utility to him exceeds the utility he could get by using his time and other resources at other activities."

Becker reasoned that we can reduce the rate of criminal behaviour by changing this calculus. This can be done by raising the likelihood that offenders will be caught (which means spending more resources on policing) and/or by increasing the severity of the punishment (for example, by spending more resources on imprisoning offenders who are caught and convicted). Becker suggested that the former strategy is generally more cost-effective than the latter, although the relative effectiveness of detection and punishment rates in deterring crime depends on the extent to which potential criminals are risk-averse (risk-avoiders will be deterred more by the prospect of getting caught, while risk-lovers will be deterred more by the prospect of severe punishment).

The economic theory of crime that has developed out of Becker's original paper recognises that different individuals break the law for different reasons, that not all law-breakers are rational utility maximisers, and that different offenders will weigh risks and benefits in different ways. Nevertheless, it claims that, at the margin, more individuals will be deterred from engaging in criminal behaviour if the chances of being apprehended and/or the severity of punishment are increased: 'We can reduce crime in our community by increasing the probabilities of capture and conviction and the severity of the penalties.'

From theory to evidence

This theory has been tested in various countries, including Australia. Most studies find that the probability of detection and punishment does indeed exert a significant influence on rates of criminal behaviour. Reviewing the evidence, Don Weatherburn concludes: "There is now plenty of evidence suggesting that punitive policies do indeed reduce or help constrain the growth in crime. In many instances they provide the only viable short-term option for dealing with it."

Two studies have been carried out in Australia. The first, nearly 20 years ago by Glenn Withers, compared the effects of employment, poverty, education, demographics and even television output with the effect of clear-up rates and imprisonment rates and concluded: "The major reliable determinants of crime rates were found to be committal and imprisonment rates." More recently, Withers' study was repeated by Phillip Bodman and Cameron Maultby using data for the period 1982 to 1991 with a similar result. Robbery, burglary, car theft and fraud all varied significantly with both the number of crimes solved or clear-up rate (a proxy for the probability of getting caught) and the average length of prison sentences (the severity of punishment), although interestingly, the study also found a significant effect for unemployment rates too.

Charles Murray's analysis of the potency of penal policy

In 1997, Charles Murray published an essay in the UK entitled Does Prison Work?. In it, he showed that crimes reported to the police in England and Wales had been rising over several decades at the same time as the probability of being apprehended and incarcerated had been falling. Even though the absolute number of prisoners in Britain had increased as crime went up, the likelihood of being locked up if you committed a serious crime had fallen. As Murray put it: "In 1955, crime began to get safer in England."

Murray contrasted this with what had happened in the United States. Not only has the US a substantially higher imprisonment rate per head of population than other Western countries, but it has also since the late 1970s had a rising rate of imprisonment per recorded crime. Murray believes that the decline in crime rates across the US that began in the 1980s and continued into the 1990s was to a large extent the result of the willingness of the US to increase its use of imprisonment to match the escalation in crime-something the UK failed to do.

Comparing US and UK crime and prison trends, Murray drew two basic lessons:

"Lesson 1: When crime is low and stable, it is a catastrophe to stop locking people up . . . Lesson 2: Prison can stop a rising crime rate and then begin to push it down.'"

Reactions to Murray

Not surprisingly, Murray's argument attracted widespread criticism. Jock Young spoke for many when he charged Murray with overlooking the inherent complexity of crime as a social phenomenon and ignoring sociological, psychological and cultural factors implicated in any analysis of international crime rates.
This is true, but we are still left with the stark statistics on which Murray based his case. Of course, correlations like these do not demonstrate causation - correlations have to be explained. Equally, however, correlations like these cannot simply be ignored.
Referring to American criminologist John DiIulio Murray anticipated many of the criticisms that were levelled against him:

John DiIulio, weary of hearing the critics of prison repeat that 'Incarceration is not the answer,' got to the heart of the matter: 'If incarceration is not the answer,' he asks, 'what, precisely, is the question?'

If the question is: 'How can we restore the fabric of family life and socialize a new generation of young males to civilized behaviour?' then prison is not the answer. If the question is, 'How can we make unemployable youths employable?' prison is not the answer. If the question is 'How can we rehabilitate habitual criminals so that they become law-abiding citizens?' prison is only rarely the answer.
But, if the question is 'How can we deter people from committing crimes?' then prison is an indispensable part of the answer."

Trends in Australia and New Zealand

Does Murray's analysis hold for Australia and New Zealand?
In Australia and New Zealand, as in the US and the UK, the prison population has grown in absolute numbers over the last 40 years, although the growth has been nothing like as large as in the US. Imprisonment per head of population roughly doubled in Australia and New Zealand (and in England and Wales) in the 40 years after 1960, with most of the increase coming after the mid-1980s, although the Australian trend has lagged somewhat behind that of the other two countries.

Where Australia, New Zealand and England and Wales all differ from the US, however, is in the rate of imprisonment per crime committed. That rate fell in all three countries until the mid-1980s. Since then it has remained relatively constant, rising slightly in England and Wales and New Zealand in the last few years while staying flat in Australia. Thus, although the number of prisoners increased in all three countries over the past 40 years, the probability of ending up in prison for a serious offence declined quite dramatically.

Australian penal policies

During the period examined, the guiding principle in Australia has been that imprisonment should only be used as a last resort. Throughout the 1970s and into the 1980s Australia actually decreased its imprisonment rate per head of population despite an escalating crime rate. In a book published in the 1980s, David Biles pointed to the deliberate attempt to reduce imprisonment rates "to the lowest levels that are consistent with public acceptance".
The principle of prison as a last resort remains in evidence in sentencing today, and there are ongoing attempts to lower imprisonment rates in Queensland and in Western Australia.

However, the political environment surrounding 'law and order', including penal policy, has changed since the 1980s. In New South Wales, for example, the former Prison Minister, Michael Yabsley, adopted a tough position through the Sentencing Act of 1989 and the NSW Labor Party responded with its own law and order campaign in 1995. Both major parties have been promising tougher penalties for crime in the lead-up to the March 2003 State election, and this pattern is being repeated in other states. The result is that the number of prisoners has increased from 6.1 per 10,000 persons in 1985 to 10.93 per 10,000 persons in 2001.

Imprisonment rates per head of population and per crime committed

While Australia's imprisonment rate per head of population has been growing, the size of the prison population has lagged far behind the growth in the rate of reported crime. In the mid-1960s, Australia locked up approximately 120 people for every 1,000 serious crimes that were committed, but by the 1980s, this figure had fallen to fewer than 30, and it has stayed around this level ever since. Between 1964 and 1986, when the number of serious crimes per head of population increased from 596 to 2,553 per 100,000, the number of prisoners per 100,000 population actually decreased, from 72 to 69.

This is in marked contrast to the pattern in the US where, from the 1980s onwards, both the imprisonment rate per head of population and the imprisonment rate per crime committed increased substantially. This shift in penal policy matches a dramatic change in the crime rate. Thus, we see the crime rate climbing until the 1980s, then flattening following the stabilisation of the imprisonment rate, then falling through the 1990s following the increased rate of imprisonment from the 1980s onwards.

In New Zealand and for England and Wales, the crime rate increased inexorably well into the 1990s, and then began to fall. Imprisonment per crime fell consistently in both countries until the late 1980s/early 1990s, since when it has been rising.

Alternative explanations

The association between crime and the probability of imprisonment seems to vary fairly consistently across the four different countries we have looked at. This seems to lend considerable credence to Murray's belief that the reduced probability of going to prison went hand-in-hand in all cases with an increase in criminal activity, and that tougher penal policies (particularly in the US, and latterly in New Zealand and England and Wales) were associated with a reversal of this trend.

However, while the statistics we have examined look consistent with Murray's hypothesis that 'prison works', they do not prove him right. Is it possible that some factor or set of factors other than the declining use of imprisonment might have been responsible for the shifts in crime rates in the countries we have been considering?
The economic theory of crime discussed earlier emphasises both the severity of punishment and the risk of getting caught. Indeed, Gary Becker believes that the latter may weigh more heavily in most people's calculations than the former. We have seen what happened to imprisonment rates over the last 40 years, but what happened to clear-up and conviction rates?

In his research on Britain, Murray reported that not only the use of imprisonment but also the success of the police in solving crimes, and of prosecutors in securing guilty verdicts, declined relative to the number of crimes being committed from the 1960s onwards. This means that not just imprisonment rates, but clear-up rates and conviction rates were all falling at the same time as the crime rate was rising.

It seems a similar pattern occurred in Australia. Success in clearing up crimes also lagged badly behind the number of crimes being committed during this period. The absolute number of crimes cleared up by the police doubled (from 183 to 364 per 100,000 population), but relative to the number of crimes taking place, it halved. This in turn reflects trends in the number of police officers since the 1960s, for while absolute numbers of police relative to the population increased, the number of police relative to the number of crimes has fallen dramatically - from 194 officers per 1,000 serious crimes in 1963 to 54 officers per 1,000 serious crimes in 2000.
Putting all this together, the probability of going to jail if you committed a serious offence in Australia fell from roughly 1 in 8 in 1964 to 1 in 37 in 1986. This was partly due to a growing disinclination to lock up convicted offenders, and partly to the decreasing ability of the police to clear up crimes. The risk of getting caught more than halved in Australia between 1964 and 1986, and once apprehended, the risk of going to jail then halved again. It is likely that both trends played a part in enabling the growth in crime. This being the case, it is not only prison that 'works', but detection and conviction too.

Conclusion

The evidence reviewed here is consistent with Charles Murray's view that a weakening in the willingness to use prison as a punishment has been strongly associated with an explosion of crime rates. All the countries we have reviewed saw their crime rates rise dramatically as they eased off on imprisonment. Those countries (notably the US) that subsequently increased their use of imprisonment have seen their postwar rise in crime rates stopped, and then reversed. In Australia, where use of prison has not been increased to the same extent, the crime rate has not been curbed with the same success. While not proving Murray's thesis, these patterns are certainly consistent with it. Whether by taking offenders out of circulation, or by deterring people from committing crimes in the first place, the evidence does seem to support the view that prison works.

But this is not the whole story. The economic theory of crime suggests that the risk of getting caught is likely to be as, or more, important in deterring crime as the anticipated severity of the punishment. In Australia, it does seem that the spiralling crime rates of the 1970s and 1980s had as much to do with declining detection and conviction as with declining use of imprisonment. This suggests that penal policy is an important element in the fight against crime, but it is only part of the solution. As economists have been telling us for more than 30 years now, increasing the probability of getting caught appears no less important than increasing the severity of the punishment that follows.