Does Prison Work?
Peter Saunders and Nicole Billante
Click here for PDF version

A dramatic fall in the American crime rate over the past decade could hold some lessons for Australian penal policy.

During the 1990s the United States experienced a significant drop in the incidence of most categories of crime. Table 1 below shows that 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. Apart from homicide (where rates in America probably reflect the accessibility of handguns), Americans are today considerably less at risk of becoming victims of serious crime than Australians. 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 Ôcontact crimesÕ such as robbery and assault.1

It would be useful to know how the US managed to reduce criminality during the 1990s, for Australian policymakers might then be able to adopt some of the same remedies.

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.3

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, AustraliaÕs economy grew even more strongly than AmericaÕs during the last decade, but crime rates here did not fall.

Some of these explanations, moreover, may have some validity yet have few if any policy implications. A fall in the number of young males in the population, for example, may have helped depress the crime rate given that most crimes are committed by this demographic group, but there is little that the Australian government might do to emulate such an effect.

What we are looking for, then, is something distinctive about America in the 1990s that Australia might be able to copy. Social phenomena like crime are rarely generated by single causes, so we should not expect to find a single Ômagic bulletÕ explanation, but the multiplicity and complexity of causation does not rule out the possibility that just one or two key changes produced a major effect. 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 ÔBroken WindowsÕ policing. Following the theory originally developed by James Q. Wilson and George Kelling, 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. Rather than letting these things go while concentrating on the big crimes, 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.

We have suggested in an earlier paper that the evidence does suggest that implementation of the Ôbroken windowsÕ theory in New York City does account for some of the decline in the incidence of crime there over the last ten years.4

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.

Economic rationality and criminal behaviour

In the 1960s, the Chicago economist Gary Becker suggested that crime, like any other ÔbusinessÕ, involves a rational calculation by individuals of likely costs, benefits and risks flowing from a given course of action. He believed that sociological, psychological and cultural theories of criminality were missing an essential point, which is that individuals will commit more crime if they calculate that the likely cost to them will be outweighed by the likely benefit:

When other variables are held constant, an increase in a personÕs 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.5

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).6

The economic theory of crime that has developed out of BeckerÕs 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: ÔWe can reduce crime in our community by increasing the probabilities of capture and conviction and the severity of the penalties.Õ7

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: ÔThere 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.Õ8

Two studies have been carried out in Australia.9 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: ÔThe major reliable determinants of crime rates were found to be committal and imprisonment rates.Õ10 More recently, WithersÕ 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.11

Charles MurrayÕs analysis of the potency of penal policy

In 1997, Charles Murray published an essay in the UK entitled Does Prison Work?.12 In it, he showed that crimes reported to the police13 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: ÔIn 1955, crime began to get safer in England.Õ14

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 crimeÑsomething the UK failed to do.

MurrayÕs argument was simple: ÔFalling use of imprisonment [is] to blame for rising crimeÕ. His argument drew on the economic theory of crime, but unlike Becker and his followers, Murray emphasised that prison works both via its deterrence effects and through what criminologists call its ÔincapacitationÕ effect (that is, by taking offenders out of circulation, it prevents them from committing further crimes).15

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

ÔLesson 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.Õ16

Reactions to Murray

Not surprisingly, MurrayÕs argument attracted widespread criticism. Jock Young spoke for many when he charged him 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.17

This is true, but we are still left with the stark statistics on which Murray based his case. In the US, crime rates rose when the rate of imprisonment per crime was falling, and they fell when the rate of imprisonment per crime began to rise. In the UK, where the rate of imprisonment per crime continued to fall into the early 1990s, the crime rate continued to escalate unabated.

Of course, correlations like these do not demonstrate causationÑcorrelations have to be explained. Equally, however, correlations like these cannot simply be ignored. It may well be true that many other factors influence the crime rate, that offenders do not always stop to work out the risk of being caught and punished before they commit crimes, and that prison does little or nothing to reform offenders and in some cases may even make them worse. None of this, however, demonstrates that Murray was wrong to argue that the rate of crime and the incidence of punishment are closely associated.

Referring to the American criminologist, John DiIulio, Murray anticipated many of the criticisms which were levelled against him:

John DiIulio, weary of hearing the critics of prison repeat that ÔIncarceration is not the answer,Õ got to the heart of the matter: ÔIf incarceration is not the answer,Õ he asks, Ôwhat, precisely, is the question?Õ

If the question is: ÔHow can we restore the fabric of family life and socialize a new generation of young males to civilized behaviour?Õ then prison is not the answer. If the question is, ÔHow can we make unemployable youths employable?Õ prison is not the answer. If the question is ÔHow can we rehabilitate habitual criminals so that they become law-abiding citizens?Õ prison is only rarely the answer.

But, if the question is ÔHow can we deter people from committing crimes?Õ then prison is an indispensable part of the answer.18

Trends in Australia and New Zealand

Murray compared US and UK data over time and found a clear pattern. But what happens if we extend his analysis of crime and prison trends to other Anglophone countries?19 In particular, does MurrayÕs 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. As Figure 1 above shows, 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 (Figure 3 opposite) while staying flat in Australia (Figure 2 below). Thus, although the number of prisoners increased in all three countries over the last 40 years, the probability of ending up in prison for a serious offence declined quite dramatically.

Australian penal policies

Law and order policy in Australia has been very different from that in the US. During the period examined, the guiding principle in Australia has been that imprisonment should only be used as a last resort.20 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 Ôto the lowest levels that are consistent with public acceptanceÕ.21

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 Ôlaw and orderÕ, 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 22 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 too. The result is that the number of prisoners has increased (up 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 AustraliaÕs 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 (see Figure 2). Between 1964 and 1986, when the number of serious crimes per head of population increased by 428% (from 596 to 2,553 per 100,000), the number of prisoners per 100,000 population actually decreased, from 72 to 69. To paraphrase Murray, crime got a lot safer in Australia through the 1970s and 1980s.

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. Figure 3 shows how 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.

The graphs for New Zealand and for England and Wales represent a much milder and lagged version of the same pattern. In both countries, 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. This would seem to support MurrayÕs contention that a rising crime rate can be reversed if you start locking more people up.

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.23 This does seem to lend considerable credence to MurrayÕs 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, all of this is only evidence of co-variationÑit is not evidence of causation. While the statistics we have examined look consistent with MurrayÕs hypothesis that Ôprison worksÕ, 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? Such a factor would need to have varied consistently with crime rates over the last 40 or 50 years across all four countries, just as imprisonment per crime appears to have done.

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 peopleÕs 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. We saw earlier that the number of serious crimes per head of population increased by 428% between 1964 and 1986 while the number of prisoners per 100,000 population remained roughly constant. It turns out, however, that 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 dramaticallyÑfrom 194 officers per 1,000 serious crimes in 1963 to 54 officers per 1,000 serious crimes in 2000.24

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 ÔworksÕ, but detection and conviction too.25

Conclusion

The evidence reviewed here is consistent with Charles MurrayÕs 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 MurrayÕs thesis, these patterns are certainly consistent with it.

There may be all sorts of good reasons why we would still choose not to follow the American example and increase our rate of imprisonment, but in resisting such a policy, we should recognise that we may be paying a price in terms of higher crime. Whether by taking offenders out of circulation, or by deterring people from comm

itting 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.

Statistical appendix

It is difficult to compile longitudinal reported crime rates for Australia, as can be collected for New Zealand, England and Wales, and the United States This has been an ongoing problem due to a lack of standard classification for offences. Each state has historically had its own classification for offences so what may be included as assault in one state may not be in another. However, in 1964, the first attempts were made to report on national recorded crime statistics for the offences of homicide, serious assault, robbery, rape, breaking and entering, and motor vehicle theft.

These offences have been the basis for the statistics used in this paper up to 1993 (see endnotes for statistical sources),26 at which time a new national standard classification was introduced. From 1993 onward the new classifications most closely related to the previous offences have been used. There are significant differences between the counting methods of the two sources (victim vs offence based) and their classification of offences (for example, rape vs sexual assault). This means that there is a break in the time series data at that time and as such the increase in crime between 1992 and 1993 should be disregarded.

While there are concerns about the comparability of the statistics between states prior to 1993, the data are still able to show us a general trend in recorded crime over time. Furthermore, recorded crime statistics should always be read with caution given problems arising from changes in police recording practices or in victimsÕ willingness to report offences to the police.

The exact figures used for this paper should therefore be treated with caution, but the overall trends are consistent with previously published accounts of AustraliaÕs crime rates. This also means that the position of AustraliaÕs crime rate relative to that of other countries should not be taken as entirely representative. For an international comparison, the International Crime Victim Surveys (ICVS) are the most appropriate source. While there is some criticism of ICVS data as opposed to the Australian Bureau of Statistics (ABS) crime victims surveys, the international aspect of the survey allows for a valid comparison with other countries. For a discussion of the validity of the ICVS please see Carlos Carcach and Toni Makkai, The Australian Component of the 2000 International Crime Victims Survey (ICVS), Technical and Background Paper Series No 3 (Canberra: Australian Institute of Criminology, forthcoming).

Endnotes

1 See Don Weatherburn, ÔLaw and order bluesÕ, Australian and New Zealand Journal of Criminology 35 (2002), 127-144.

2 John van Kesteren, Pat Mayhew, and Paul Nieuwbeerta Criminal Victimisation in Seventeen Industrialised Countries (the Hague: Ministry of Justice WODC, 2000) 178.

3 For examples of these arguments see Benjamin Bowling, ÔThe Rise and Fall of New York Murder: Zero Tolerance or CrackÕs Decline?Õ, The British Journal of Criminology 39: 4 (1999) 531.

4 See Nicole Billante and Peter Saunders Six Questions about Civility (Sydney: The Centre for Independent Studies 2002) and see George L Kelling and William H Sousa Jr, Do Police Matter? An Analysis of the Impact of New York CityÕs Police Reforms, (New York: The Manhattan Institute, 2001) for a statistical analysis of broken windows policing.

5 Gary Becker, ÔCrime and punishment: An economic approachÕ, Journal of Public Policy 76 (1968), 176.

6 Cathy Buchanan and Peter Hartley, The Economic Theory of Crime (Sydney: The Centre for Independent Studies, 1992).

7 As above, 63.

8 Weatherburn, ÔLaw and Order BluesÕ, 134.

9 Glen Withers, ÔCrime, punishment and deterrence in AustraliaÕ The Economic Record 60 (1984) 176-85; Philip Bodman and Cameron Maultby, ÔCrime, punishment and deterrence in Australia: A further empirical investigationÕ, International Journal of Social Economics 24 (1997) 884-901.

10 As quoted in Buchanan and Hartley, The Economic Theory of Crime, 42.

11 Bodman and Maultby, ÔCrime, punishment and deterrence in AustraliaÕ.

12 Charles Murray, Does Prison Work?, (London: Institute of Economic Affairs, 1997) Murray has applied an economic calculation approach to analysis of various different aspects of human behaviour including welfare dependency (he believes that poor people respond rationally to higher welfare benefits by reducing their effort to find and maintain employment) and family formation (government support for single mothers reduces the cost of sole parenthood and therefore acts as a disincentive to marriage), as well as to crime See, for example, Charles Murray, Losing Ground, (New York: Basic Books, 1984).

13 It is accepted in criminology that there is generally higher actual crime than that which is reported This is borne out by crime victim surveys Nevertheless, although reported crimes do not accurately indicate the actual number of crimes being committed, the rate of reported crime over time does provide a reasonable indication of the rate of change in criminal activity.

14Murray, Does Prison Work?, 1.

15 Economic theorists tend to resist incapacitation as an explanation, for it has nothing to do with the rational cost-benefit utility maximization that lies at the heart of their theory In their work ten years ago for the CIS, for example, Buchanan and Hartley suggested that deterrence probably exerts ten times the effect of incapacitation, although it is in practice very difficult to separate the two in statistical analysis From MurrayÕs point of view, of course, it does no really matter which explanation is the stronger, for both point to the importance of prison in reducing crime.

16 Murray, Does Prison Work?, 16-17

17 See: Jock Young, ÔCharles Murray and the American Prison Experiment: The Dilemmas of a LibertarianÕ in Does Prison Work?: Commentaries, 31-39.

18 Murray, Does Prison Work?, 20.

19 By limiting the analysis to Anglophone countries, we can hold constant some of the cultural variations which Jock Young quite rightly suggests bedevil international comparisons Elsewhere we have pointed to the cultural similarities between the US, the UK, Australia, New Zealand and Canada as compared with other advanced western societiesÑsee Peter Saunders, ÔAustralia is not SwedenÕ, Policy 17:3 (Spring 2001)

20 George Zdenkowski, ÔContemporary Sentencing IssuesÕ, in The Australian Criminal Justice System: The mid 1990s, ed Duncan Chappell and Paul Wilson (Sydney: Butterworths, 1994), 187.

21 David Biles, ÔPrisons and their ProblemsÕ, in The Australian Criminal Justice System: The mid 1980s, ed Duncan Chappell and Paul Wilson (Sydney: Butterworths, 1986), 241.

22 Russell Hogg and David Brown, Rethinking Law and Order (Sydney: Pluto Press, 1998), 39.

23 While due to space limitations we have not included state breakdowns here, there is relative consistency within Australian states It should be noted that imprisonment rates do vary greatly between the states, so while the pattern is roughly the same, the scales are different.

24Jennifer Buckingham, Lucy Sullivan, and Helen Hughes, State of the Nation 2001, (Sydney: The Centre for Independent Studies, 2001), 110-111.

25 We shall analyse policing and conviction trends in more detail in future publications

26 Statistics Sources:
Australia - Crime
1964-1987: Satyanshu Mukherjee, Anita Scandia, Dianne Dagger, and Wendy Matthew Source Book of Australia Criminal and Social Statistics 1804 ø 1988 (Canberra: Australian Institute of Criminology, 1989).
1964-1970 ø Homicide: ABS (Australian Bureau of Statistics), Yearbook, ABS Cat. No. 1301.0 (Canberra: ABS, respective years).
1988-1992: John Walker ÔTrends in Crime and Criminal JusticeÕ in The Australian Criminal Justice System: The mid 1990s ed Duncan Chappell and Paul Wilson (Sydney: Butterworths, 1994).
1993- 1994: ABS (Australian Bureau of Statistics), National Crime Statistics, ABS Cat. No. 4510.0 (Canberra: ABS, respective years).
1995- 2000: ABS (Australian Bureau of Statistics), Recorded Crime Statistics, ABS Cat. No. 4510.0 (Canberra: ABS, respective years). Australia ø Prisoners
1964-1987: Satyanshu Mukherjee, Anita Scandia, Dianne Dagger, and Wendy Matthew Source Book of Australia Criminal and Social Statistics 1804 ø 1988 (Canberra: Australian Institute of Criminology, 1989).
1988-1993: AIC (Australian Institute of Criminology) Australian Prison Trends (Canberra: AIC, respective years). 1995: ABS (Australian Bureau of Statistics), Yearbook, ABS Cat. No. 1301.0 (Canberra: ABS, respective years).
1996: ABS (Australian Bureau of Statistics), National Correction Statistics, ABS Cat. No. 4512.0 (Canberra: ABS, 1996) 1997-2000: ABS (Australian Bureau of Statistics), Corrective Services, ABS Cat. No. 4512.0 (Canberra: ABS, respective years) Australia ø Population
1964-1987: Satyanshu Mukherjee, Anita Scandia, Dianne Dagger, and Wendy Matthew Source Book of Australia Criminal and Social Statistics
1804 ø 1988 (Canberra: Australian Institute of Criminology, 1989).
1988-2000: ABS (Australian Bureau of Statistics) Population Estimates Time Series Spreadsheet, AusStats website England and Wales ø Crime
1960-2000: Home Office Summary of recorded crime
1900-2000/2001 (London: HMSO, 2001), http://www.homeoffice.gov.uk/rds/pdfs/100years.xls. England and Wales ø Prisoners 1960-1968 Home Office People in Prison, England and Wales (London: HMSO, 1969).
1969-1979 Home Office Statistics of the Criminal Justice System, England and Wales (London: HMSO, 1979)
1980-1988 Home Office Prison Statistics, England and Wales, (London: HMSO, respective years)
1989-2000 Home Office Prison Statistics, England and Wales 2000 (London: HMSO, 2001), http://www.homeoffice.gov.uk/rds/pdfs/chapter1.xls. England and Wales ø Population
1951-2000 National Statistics Online Time Series Data: Monthly Digest of Statistics: Mid-year estimates of resident population, http://www.national-statistics.gov.uk/statbase/TSDtimezone.asp. New Zealand ø Crime
1950-1972 G T Bloomfield New Zealand: A Handbook of Historical Statistics (Boston:ÊG.K. Hall,Ê1984). 1972-2000 Statistics New Zealand Official New Zealand Yearbook (Wellington, Statistics NZ, selected years) New Zealand ø Prisoners
1950-1998 Statistics New Zealand Official New Zealand Yearbook 1999 (Wellington, Statistics NZ, 1999), http://www.stats.govt.nz/domino/external/pasfull/pasfull.nsf/84bf91b1a7b5d7204c256809000460a4/ 4c2567ef00247c6acc25697a00043f4a/$FILE/Chp10data%20-%2020KB.xls. New Zealand ø Population
1950-1997 Statistics New Zealand Official New Zealand Yearbook 1998, (Wellington, Statistics NZ, 1998)
1998-2000 Statistics New Zealand National Population Estimates, (Wellington, Statistics NZ, 2000) URL http://www.stats.govt.nz/domino/external/web/prod_serv.nsf/3153e23ac69cb3d84c 25680800821fa4/564b6fc27b87b65bcc256b200073a589?OpenDocument. United States ø Crime 1960-2000 US Department of Justice, Bureau of Justice Statistics Sourcebook of Criminal Justice Statistics Online, http://www.albany.edu/sourcebook/1995/wk1/t3120.wk1. United States ø Prisons1960-1996 Critical Criminology compiled from US Department of Justice, Bureau of Justice Statistics, http://www.soci.niu.edu/~critcrim/prisons/pris.pop.
1997- 2000 US Department of Justice, Bureau of Justice Statistics, http://www.ojp.usdoj.gov/bjs/prisons.htm#selected. United States ø Population
1960-1990 US Census Bureau Population Estimates, http://eire.census.gov/popest/archives/ 1990.php?PHPSESSID=80857278f5d778c99faff352f0c1e35e.

Peter Saunders is Director of Social Policy Programmes at The Centre for Independent Studies, and Nicole Billante is a Research Assistant at the Centre. Thanks to Barry Maley, Jennifer Buckingham, David Hitchin and an extremely helpful anonymous external referee for comments on an earlier draft.

Policy is the quarterly review of The Centre for Independent Studies. For more information on subscribing to Policy, click HERE

If you are interested in the Centre's activities and publications, why not subscribe to e-PreCIS, our regular email update on the latest news and events.

(e-PreCIS requires html capable email facilities, such as Microsoft Outlook Express or Netscape Messenger)