You are here: Home Publications Publications archive (pre-2011) 1999 Review of community-based sentences in New Zealand The Cost-effectiveness of Community-based Sentences

Archived publications

You are currently in the publications archive, and may not be accessing the most up to date information. The more recent publications can be found in the publications A - Z.


The Cost-effectiveness of Community-based Sentences

Fiscal costs
Breaches of community-based sentences
Reviews of community-based sentences
Combined sentences
Other considerations
Effects of the availability of community-based sentences on the use of imprisonment

A claim often made for community-based sanctions is that, to the extent that they are used as alternatives to imprisonment, they save money, and that the savings achieved equal the difference between the average sentence costs per annum for each offender. The first thing to note in respect of this claim is that different types of community-based sentences are likely to have quite different costs depending on the duration, level of intensity and enforcement of the sentences, as well as on the range of supportive services (including non-government services) provided to control, supervise and support the offenders in their assigned work, programme, or code of behaviour. This will be dependent to some extent on the selection of offenders for the particular sentences.

Fiscal costs

The latest costs for the Department of Corrections associated with community sentences and imprisonment in New Zealand are as follows:

Table 24: Costs of sentences to the Department of Corrections#205.

Sentence type

Per annum


Per month


Average sentence

Average sentence cost

Imprisonment (average)





Maximum security



Medium security



Minimum security



Periodic detention










Community programme





Community service





In terms of assessing the cost effectiveness of various sentences, however, there are certain considerations that make the average annual costs per offender of community-based sentences not directly comparable with the average annual cost per offender of imprisonment or the cost of one type of community sentence comparable with another. These considerations include:

  • the extent to which offenders are diverted to particular community-based sentences from fines or diversion rather than prison (usually called net-widening);
  • the rates of breaches and the rates of re-sentencing to more costly sentences;
  • the use of community-based sentences in combination with each other; and
  • the effect of the different sentences on reoffending.


Net-widening is the term used to describe the phenomenon of increasing the number of people under the control of the correctional system through the introduction of a new sanction. This may occur even though it was the intention that the new sanction should be imposed on some offenders in lieu of a more severe sanction. It is also used to describe increasing the severity of sanctions for offenders. This can happen when a new sanction is introduced to reduce the use of a more severe sanction and instead it reduces to a greater extent the use of a sanction of lesser severity. Often it happens when non-custodial sanctions are introduced to replace imprisonment and they come to be used, instead, as substitutes for more traditional sanctions such as probation or fines. Meanwhile, the use of imprisonment goes on as before.

A hypothetical scenario that would illustrate this effect could start with a situation where 20% of convictions result in a sentence of imprisonment and 80% result in probation. A new sanction of intensive supervision is then introduced as a less severe and less costly alternative to imprisonment for some offenders currently receiving imprisonment. The intended result is to reduce the use of prison sentences by say 10%.

However, research has shown that the introduction of community sanctions as alternatives to imprisonment does not usually bring about the expected reduction in custodial sentences.#206. Instead, to a much greater extent, they become alternatives to other alternatives to custody. In terms of the hypothetical scenario, this means that the percentage of offenders receiving imprisonment decreases to a small extent only and intensive supervision is mostly imposed on offenders who previously received probation, which is a less severe and intrusive disposition than intensive supervision. The net result is therefore an increase in sentence severity due to those probationers now serving intensive supervision. Contrary to the original intention, it is the less severe sanction of probation rather than the more severe one of imprisonment which is mainly diminished by intensive supervision.

This starting point, intended result and actual result are illustrated below:

  • Starting point
  • Imprisonment 20%

    Probation 80%

  • Intended result: a reduction in imprisonment
  • Imprisonment

    Intensive supervision

    Probation 80%

  • Actual result; a reduction in probation
  • Imprisonment

    Intensive supervision

    Probation 72%

    The process does not stop here. Although 8% of the offenders who would previously have received probation will be placed under intensive supervision, there is likely to be no real decrease in the overall use of probation. The missing 8% of probationers will be replaced by a selection of those offenders who used to receive a less severe sanction of, say, a fine. The process can continue to have this domino effect in the direction of less severe sanctions so that it eventually involves those offenders who previously were diverted from the correctional system and received no sanction, thus increasing the overall number of people in the correctional system. This can be illustrated as follows:

    d) The net widening effect


Two things are happening that were not intended. First, imprisonment rather than one of the new community sanctions continues to be used for crimes that are at an intermediate level. Second, the more severe community sanctions are being used for some of the least serious offences, where they are disproportionately onerous, instead of being imposed, as intended, for offences of medium gravity.

Recent data show that in New Zealand extending the range of community-based sentences did lead to a substantial increase in the use of these sentences. Although imprisonment rates for some offences did decrease, these decreases were much smaller than the overall increases in the use of community-based sentences. In particular, there are indications that community service is now being used where previously (in the 1980s) a monetary penalty would have been imposed. This suggests that net-widening has occurred.#207.

Breaches of community-based sentences

Even if savings are made by diverting some offenders from prison to community-based sentences, breaches of community-based sentences reduce the savings to the criminal justice system of those sentences. For example, the net savings in the number of prison beds would be the number of persons diverted, less the number of persons sentenced to imprisonment for breaching the alternative sentence or committing new crimes while not incapacitated by a prison sentence. In New Zealand periodic detention has a breach rate of about 20% (see table below) and in 1997 18% of these cases resulted in imprisonment (in 1988 it was as high as 26%) with an average custodial sentence length of 1.9 months.#208. The 1997 prison census revealed 50 inmates whose major offence was breach of periodic detention out of a total of 4,935 sentenced inmates.#209. Breaching the conditions of a supervision sentence or a community service sentence is not an imprisonable offence and there is no offence of breaching community programme. However, a review of these sentences can result in a sentence of imprisonment being imposed.

Where net-widening occurs, breaches can produce a net increase in imprisonment. If 30% of offenders sentenced to a particular community-based sentence are diverted from prison and 70% from fines, diversion, or another community sentence for which imprisonment is not available as a sanction for breach, then the proportion of that 70% who have the sentence revoked and are imprisoned represent a new demand for prison beds, the costs of which may exceed the savings made from the 30% diverted from imprisonment.

The table below shows the trends in convictions for breaches of community-based sentences over the last decade. As noted above, in New Zealand a breach of a community-based sentence can lead to a review of the sentence by the court and the offender can be re-sentenced on the original charge for which the community-based sentence was imposed. This type of breach of sentence resulting in a sentence review rather than a prosecution for breach is not included in the data below.

Table 25: Number of convicted cases involving breaches of community-based sentences, 1989 to 1998#210.

Sentence type











Periodic detention

No of sentences

No of cases involving a breach

Breach rate (per 1000 sentences)

































No of sentences

No of cases involving a breach

Breach rate (per 1000 sentences)
































Community service

No of sentences

No of cases involving a breach

Breach rate (per 1000 sentences)































Periodic detention is the community-based sentence that has the highest breach rate (over 5 times that of community service and 3 times that of supervision). This is likely to be because of the different characteristics of the offenders receiving that sentence and the fact that it is closely supervised and breaches will therefore nearly always be detected. It should also be noted that community service is imposed with the consent of the offender and will generally require the employing authority to report any breach in order for it to come to the attention of the system and action to be taken.

Convictions for breach of periodic detention increased substantially over the decade. The increase in the period between 1989 and 1991 is likely to have been partly due to an increase in the use of the sentence. The increase in breaches between 1992 and 1996, however, occurs despite the use of periodic detention decreasing, then levelling off. It therefore appears that the proportion of offenders breaching periodic detention increased, or else the likelihood of periodic detention wardens laying informations in court for such breaches increased, or both. In 1997 and 1998 the number of periodic detention sentences increased and the breach rate declined.

As the proportion of offenders who receive imprisonment for breach of periodic detention has declined there has been an increase in the use of periodic detention as the sentence for this offence (in 51% of cases in 1995 compared to 29% of cases in 1983).#211.

The breach rate for supervision declined from 1993 to 1996, increased in 1997, and then fell again the following year. The number of convictions for breaching community service in 1998 was nearly triple the number in 1989. The increase was most apparent between 1989 and 1993 (as was the breach rate) but decreased over the next 5 years. This pattern is similar to the use of community service as a sentence over the same period except for an increase in the number of sentences in 1998.

The proportion of offenders prosecuted for breach of a particular sentence may tell us as much about the degree of tolerance towards violations or the effectiveness of the mechanisms for dealing with breaches as it does about true compliance with the sentence. With community service it is the community sponsor ("employing authority") that oversees the offender (although statutory responsibility for overall supervision rests with the community probation service). Some sponsors, at least in the past, have avoided any enforcement role in the event of the offender not fulfilling the sentence requirements and have chosen not to inform the probation service of breaches because they have no wish to appear in court as a witness against the offender.#212.

Reviews of community-based sentences

There are also the costs associated with the community probation service actioning reviews of community-based sentences where there are difficulties over compliance. These are different from actions for breach of sentence. The reviews involve applications to the courts and may result in changes to the conditions of a sentence, cancellation of a sentence, or a new sentence.

Combined sentences

The extent to which community-based sentences are used concurrently with each other increases the cost of their imposition. In New Zealand the two community-based sentences of periodic detention and supervision can be imposed concurrently. In 1998 supervision was imposed as the most serious sentence in 5,331 cases. It was imposed in a further 3,735 cases in combination with periodic detention. To add a sentence of supervision to one of periodic detention significantly adds to the cost of the total sentence (see table 24).


Any full cost effectiveness analysis of the different community and custodial sentences needs to take into account any differential impact that the sentences may have on re-offending. In other words, in terms of later criminality by convicted and sentenced offenders, do community-based sentences diminish or increase crime compared with comparable groups of offenders sentenced to imprisonment or to monetary penalties? If community-based sentences produce higher re-offending rates than is the case with the other sentences, then there is an additional cost involved with community-based sentences.

Reconviction rates provide the only viable means of assessing the effectiveness of particular sentences in preventing re-offending but are limited in their usefulness in making comparisons for two principal reasons. One is that reconviction rates are only an approximate measure of reoffending, since they record only offending that has been successfully detected and prosecuted. The other is that reoffending is influenced more by other factors than by the sentence. Factors closely associated with reconviction are age, gender, ethnicity, and criminal history.#213 Different types of sentences tend to be given to different sorts of offenders with widely different probabilities of reoffending.

The table below is based on a study of reconviction rates for offenders in the two years following the conviction date of their first 1991 proved case or, for offenders sentenced to imprisonment, within two years of their estimated release date. All offenders who had one or more proved cases involving an imprisonable offence in 1991 were included in the data-set (about 77,000 offenders). The figures show the predicted probability of reconviction for each sentence type calculated from a statistical model of reconviction rates, using current and past offending and demographic characteristics of offenders as predictor variables. This is compared with the actual reconviction rate. #214.

Table 26: Actual percentage of offenders reconvicted within two years compared to the percentage predicted, by sentence type, for cases finalised in 1991

Sentence Type

Predicted reconviction percent

Actual reconviction





Periodic detention



Community care/programme



Community service






Monetary penalty



Other sentence



No sentence



Reconviction rates for most of the sentences are very close to the rates that were predicted for each group on the basis of the criminal histories of the offenders (e.g. the seriousness and frequency of past convictions) who received that sentence, along with their age, gender, and ethnicity. In other words, the fact that community service had a significantly lower reconviction rate than the other community-based sentences is accounted for by the type of offenders receiving that sentence rather than the type of sentence per se.

The results of the modelling indicated that the most important variables for predicting recidivism were the criminal history and the demographic group (age, gender, ethnicity) of the offender. None of the sentence types were highly significant predictors of recidivism. Community programme and prison had no significant effect on recidivism rates, relative to monetary penalties, once other factors (such as differences in criminal history and current offending) were taken into account. Periodic detention, community service, and supervision appeared to be associated with an increased risk of recidivism, relative to monetary penalties, although this may mean that people sentenced to those community-based sentences have other characteristics that increased the likelihood of re-offending that could not be measured statistically. Previous sentences of imprisonment or any community-based sentence appeared to increase the risk of recidivism. The model also indicated that the seriousness of the current offence, despite being a key factor in determining which sentence is imposed, has relatively little impact on recidivism following completion of the sentence.#215.

The above findings are consistent with a 1984 study of recidivism within one year following the imposition of either a sentence of community service or of periodic detention.#216 The analysis found an overall reconviction rate of 38% for the community service group and 59% for the periodic detention group. However, reconviction rates varied significantly when the samples were disaggregated. For example, the highest reconviction rate following a sentence of community service (74%) came from a subgroup who had not received additional probation and who were 17 years old or younger at the time of their first conviction. The lowest rate (25%) came from the group who had not received probation and who were 23 years or more at the time of their first conviction. Similarly, the highest reconviction rate following periodic detention (80%) came from a subgroup who had not received additional probation and who were 17 years or younger at the time of their first conviction. The lowest rate (25%) came from the group who did not receive probation and who were 23 years or more at the time of their first conviction. #217.

The study concluded that "reconviction rates depended to a great extent on factors other than the actual sentence". The analysis

showed that when certain of these factors were taken into account when comparing the sentences of community service and periodic detention, in many cases there was no significant difference in the relative reconviction rates. Specifically, there was no difference in the reconviction rates of community service and periodic detention groups when extreme groups were compared - the highest risk group and the lowest group.#218.

In the case of moderate risk groups, there was a difference in reconviction rates between the two sentences. However, this did not necessarily mean that those who posed a moderate risk of re-offending would be less likely to be re-convicted if the sentence of community service rather than periodic detention was imposed. A different interpretation was favoured for 3 reasons. Firstly, as reconviction rates varied greatly in relation to factors besides the sentence given, the validity of an apparent difference between sentences was questionable. The second reason was that the variables utilised were not fully inclusive of all factors that may impact on individuals' reconviction rates. For example, employment, family, and education details were excluded from the analysis. Thirdly, the study did not examine post-sentence circumstances that might have made a difference to reconviction rates.#219.

The table below shows reconviction rates after one and two years, for sentences imposed in 1995. Reconviction rates for custodial sentences of one year or less are included as a comparison. As already noted, the most important variables for predicting recidivism were the criminal history and the demographic group (age, gender, ethnicity) of the offender. None of the sentence types were highly significant predictors of recidivism.

Table 27: One year and two year reconviction rates for people sentenced in 1995#220


One year reconviction rate

Two year reconviction rate(2)

Prison for one year or less(1) 0.63 0.80
Periodic detention 0.60 0.73
Community programme 0.47 0.63
Community service 0.37 0.51
Supervision 0.45 0.61
Monetary 0.30 0.41
Other 0.39 0.51


  1. Reconviction rates are within one or two years from the estimated date of release from prison. People given a custodial sentence of longer than one year in 1995 were excluded from the table as in most cases, all or part of the one and two years after the prison sentence was imposed was spent in custody which limits the opportunity for reoffending and hence does not allow a valid comparison with reoffending rates for non-custodial sentence types.
  2. Two year reconviction figure excludes data for 6% of cases where there was less than two years between the person's estimated date of release from prison and when the data on reconvictions were extracted.


Re-offending increases in significance when it also leads to the fast-tracking of offenders through the range of available penalties towards imprisonment. This arises when there is a general perception among the judiciary that when one type of penalty has been tried on an offender and he or she re-offends, then that penalty has failed and a more severe one must be imposed. In other words, when there is repeat offending, sentences become progressively more severe with each new offence, causing offenders to escalate up the penalty scale. Fast-tracking then occurs when offenders start receiving penalties out of proportion to the gravity of the current offence. Thus, an offender sentenced to community service relatively early in his or her offending history, for a comparatively minor offence, may be sentenced to periodic detention for a subsequent offence of similar severity rather than community service again or a fine, which might have been the penalty if it had been a first offence. A further offence is then punished with imprisonment. In comparison, someone who receives a fine at the early stage may receive community service for the subsequent offending. In reality persistent offenders do not progress step by step up a ladder of sentences since they neither consistently offend at the same level of seriousness, nor offend at a consistently increasing level of seriousness. Rather they will engage in a wide range of offending of varying seriousness and type which will result in a variety of penalties before they end up in prison (unless at any stage they commit an extremely serious offence). Nevertheless, "fast-tracking" may be the generalised trend, with sentences increasing in severity if offending were to remain at the same level.

Recent research to determine the relative influence of various statistical factors on sentencing  #221. does show that the probability of receiving a particular community-based sentence is strongly influenced by the type of any community-based sentence served in the past. The research also shows a degree of escalation taking place with sentencing. That is, the previous sentence increases the risk of the same sentence or a more serious sentence being imposed, all other factors being equal. For example, the probability of receiving a periodic detention sentence is higher if the most recent or previous sentences are periodic detention or community service. Having a previous periodic detention sentence increases the probability of receiving a prison sentence and decreases the probability of a community service sentence or monetary penalty.

If an offender has already served a community service sentence then they are more likely to get periodic detention or supervision or community service again as the next sentence, but less likely to get a monetary penalty. Similarly, once a periodic detention sentence has been served the less serious community service is less likely than periodic detention or imprisonment to be the next sentence, even when the seriousness of the offence and criminal history of the offender are taken into account.

However, it needs to be noted that although the likelihood of imprisonment is relatively higher for offenders whose most recent sentence was periodic detention, the use of imprisonment for this group of offenders has decreased considerably since 1983. In that year 32% of offenders whose last sentence was periodic detention received a custodial sentence as their next sentence and in 1995 this had decreased to 16%. There has also been a decline in the use of imprisonment in general for offenders with several previous convictions. Thus, the impact of fast-tracking has been reduced in practice by a decrease in the use of imprisonment for these offenders.#222.

Other considerations

There are three other considerations that complicate cost comparisons: transaction costs, marginal costs, and costs to the wider community.

The first of these considerations is that, in addition to the additional sentence costs of the net-widening process that shifts those who might have been fined to a community sentence and then shifts some of those to prison because of a breach of sentence, this process creates transaction costs. These are new expenses for prosecutors, the courts, and the probation service in administering each of those violations and transfers.

The second consideration is that imposing a community sentence on an offender who would otherwise have received 6 months or 1 year's imprisonment does not save the prison system the average cost of an inmate serving six months or a year's prison sentence. When the marginal costs of 50 to 100 inmates in a prison system of 5,000 inmates (spread over 19 institutions) are compared with the marginal costs of placing that number of offenders in a small programme then the potential cost savings are diminished. Unless the reduction in the number of inmates reaches a level which enables a prison or a prison unit to be closed or not opened, the only savings from diverting that number will be incremental costs for food, laundry, supplies, and other routine items. The major costs of payroll, administration, debt servicing, maintenance, and depreciation will not be affected to any great extent. In other words, savings in the prison system will only accrue when the alternative sanctions reach a threshold of usage that allows institutional changes in costs of imprisoning.

Third, there are the costs of crimes (often not "cleared" by the police) that will occur if offenders are assigned to a community sentence rather than imprisoned. This specifically relates to crimes committed in the community while the offender is on the sentence rather than recidivism in general, which is discussed above. It is likely that averting crimes through incapacitation produces savings to the larger community not possible with alternative sanctions, although estimates of the level of savings have been contested, particularly in terms of the supposed number of crimes prevented for each inmate confined and the dollar amount that that saves.#223.

A full cost-effectiveness analysis of community-based sentences would also have to take into account such factors as the benefits of the work done by offenders on periodic detention or community service. A recent Department of Corrections publication states that the estimated value of the work done by detainees is $18 million per annum.#224. There are also the benefits to offenders of attending programmes while on supervision or community programme. These might include the learning of new skills.

Effects of the availability of community-based sentences on the use of imprisonment

The evidence available from overseas seems to point to it being unlikely that a sustainable reduction in the prison population can be brought about through expanding community-based sentencing options. Even when such sentences have been specifically designed to reduce imprisonment this appears to be difficult to realise. Studies monitoring imprisonment rates in the United States, Canada, and Britain during the 1970s and 1980s, found that in spite of the development and implementation of a range of alternative community-based sentences, imprisonment rates kept climbing. In the United States and Canada there were also dramatic increases in corrections expenditure and personnel over this period.#225. The explanation for this was that it was because many of those receiving the alternatives would not have gone to prison in the first place and when they were being used as genuine alternatives they were nearly always impacting on the use of the shortest custodial sentences and thus having little impact on prison musters.#226. Both these trends (increasing musters and corrections expenditure) have also shown up in New Zealand.

However, any historical association between a rising prison population or the rate of imprisonment and an expanding system of community sanctions does not mean that the latter provided the means for the former to happen, and certainly does not establish a direct causal link between the two trends. There are likely to be other factors at work which bring about changes in the prison population. Such factors include changes in policing practices (law enforcement), legislation (e.g. penalties), sentencing by the courts (e.g. length of prison sentences or the imposition of imprisonment for particular types of offences), crime rates, and the types and seriousness of offending being dealt with by the courts, the circumstances and criminal histories of offenders being sentenced, and changes to parole eligibility or final release dates for prison inmates. In New Zealand the size of the prison population is particularly sensitive to the number of serious violent offenders prosecuted because they serve by far the longest sentences.

In New Zealand the percentage of convicted imprisonable cases resulting in imprisonment has changed relatively little, fluctuating between 8% and 11% in the period 1982 to 1998. However, the indications are that there has been a relative decrease in the use of imprisonment in that with most offences other than the more serious violent, sexual, and drug dealing offences there was almost twice the likelihood of receiving a prison sentence in 1983 than there was in 1995. This reflects the introduction of presumptions against the use of imprisonment for less serious property offences by the Criminal Justice Act 1985 and the introduction of suspended prison sentences in 1993. This is strong evidence that the availability of community-based sentences has reduced the range of offenders being sentenced to imprisonment.

Nevertheless, the prison population has increased by more than 50% in the last decade (1989-1998). The factors driving the growth of the prison population have been significant increases since 1983 in the average seriousness of cases being prosecuted and in the average number of previous convictions of convicted offenders (fewer first offenders and more very persistent offenders), and increases in sentence lengths for very serious offences. There has been an increasing proportion of violent offenders in the prison population. In particular there has been a significant increase in the length of prison sentences served for serious violent offenders due in no small part to the removal of parole for serious violent offenders and an increase in the minimum non-parole periods for inmates serving life imprisonment and preventive detention.#227. The growth in New Zealand's prison population is examined in detail in The Use of Imprisonment in New Zealand (1998). #228.


There is no evidence to suggest that community-based sentences are generally more or less effective than imprisonment in terms of recidivism, although the specific programme content of any sentence may of course have an impact on recidivism. Therefore, the relative costs of the various sentences become important considerations. While it is the stated purpose of the community-based sentences to act as an alternative to more costly sentences of imprisonment, the overall cost savings from the availability of community-based sentences may be considerably less than what at first appears likely to be the case, for the following reasons:

  • the costs of net-widening, whereby many offenders serving community-based sentences would have incurred a fine had those sentences not been available. This involves a loss of fines revenue as well as the costs of administering the community-based sentences;
  • to the limited extent that community-based sentences replace the use of imprisonment, this mainly impacts on short terms of imprisonment which have relatively little effect on prison numbers. Savings in the prison system will only be realised when the alternative sanctions reach a threshold of usage that either enables some prisons or prison units to be closed down or avoids the need, that would otherwise arise, for extra buildings to be built;
  • the costs of fast-tracking offenders towards imprisonment;
  • the costs of imprisonment or alternative sentences incurred as a result of breaches of the original community-based sentences (and the associated transaction costs);
  • the costs of concurrent sentencing; and
  • the costs of crimes committed because the offenders have not been incapacitated by imprisonment.

The points noted above reduce the chances of achieving net cost savings. They also mean that the difficulties in making any calculation of those savings are formidable and would involve making a considerable variety of assumptions.#229. A true cost-benefit analysis of the impact of community-based sentences would only be possible if sentences were imposed in accordance with strict guidelines. If this were the case it would be possible to determine the savings or additional costs which might result from changes to the guidelines which, for example, caused all cases of a particular offence type to result in one specific sentence rather than another. However, the sentencing menu provided by the Criminal Justice Act 1985 was deliberately established to provide the courts with a range of sentencing options, whereby any particular sentence imposed would be determined on the basis of a very wide range of factors (including aggravating and mitigating circumstances) specific to the individual case rather than by, for example, offence type only. It is not possible, therefore, to assume on the basis of statistical data, what alternative sentences might have been imposed had the sentencing menu been more restricted. The high premium placed on judicial discretion in the sentencing process means that an accurate cost-benefit analysis of the impact of community-based sentences is not possible.


205. These are the full costs to the Department of Corrections and include Public Prisons Service and Community Probation Service costs plus an allocation of Departmental Head Office and infrastructural costs such as computer and phone systems. Community-based sentence costs are made up of personnel costs 52.9%; operating costs 14.4%; programme costs 6.4%; depreciation 3%, capital charges 1.4%; and Departmental overheads 21.9%. Figures (GST inclusive) are provided by Department of Corrections.

206. Ministry of Justice, 1998; Tonry, 1990, p158.

207. Triggs, 1999, pp122; 131-3.

208. Spier, 1998, pp57-8.

209. Lash, 1998, Census of Prison Inmates 1997, p30.

210. Source: Criminal Justice Group, Ministry of Justice.

211. Triggs, 1999, p130.

212. Asher and O'Neill, 1990, p25.

213. Lloyd, Mair, and Hough, 1994; Triggs, 1999, Appendix 2, pp140-5.

214. Triggs, 1999, p142.

215. Ibid, p143-5.

216. Leibrich, 1984.

217. Ibid, p203.

218. Ibid, p204.

219. Ibid, p204.

220. Source: Criminal Justice Group, Ministry of Justice.

221. Triggs, 1999.

222. Ibid, p130.

223. Tonry, 1996, p107.

224. Department of Corrections, Corrections News, June 1999, p1.

225. Chan and Zdenkowski 1986, pp136-8. Bottoms (1987) examined trends in sentencing in England and Wales from 1965 to 1985, during which suspended sentences, the community service order and probation with special conditions (including attendance at day centres) were introduced to reduce prison use. Both the prison population and the proportionate use of custody continued to increase. Hylton (1982) has cited Canadian provincial data from 1962 to 1979 to reach similar conclusions.

226. Vass, 1996, p170. The New Zealand prison census in 1997 showed that 63.4% of the population were serving sentences of more than 2 years (p27).

227. Triggs, 1999, pp46-9 and p105.

228. Ministry of Justice, 1998.

229. There are nevertheless likely to have been some cost savings made in New Zealand in the 1983-95 period as a result of the real decrease in the use of imprisonment (after taking into account the increase in the seriousness of offences and offenders coming before the courts) that followed from a greater use of community-based sentences, even though the majority of those sentences did not replace imprisonment. A crude calculation could be as follows:

  • estimated real reduction in imprisonment of 3.7 per 100 cases, averaging 3 months with release after half the sentence equates to 3.7 x 3/12 x 0.5 = 0.46 inmate/years. At an average inmate cost of $52,000 per annum this equals a cost decrease of $24,000 per 100 cases;
  • estimated real increase in community-based sentences of 14.8 per 100 cases, with an average offender cost of $1,000 equals a cost increase of $14,800;
  • estimated real decrease in fines of 15.7 per 100 cases at an average fine of $500 equals a cost increase (revenue loss) of $7,800;
  • cost saving equals $1,400 per 100 cases. ($24,000 - 14,800 - 7,800).

This does not take into account costs resulting from breaches and reviews of community-based sentences and subsequent re-sentencing, or the additional costs of any crimes committed whilst the offender was in the community rather than in prison.