Prisoners in Australia methodology

The scope of the statistics in this publication includes all persons remanded or sentenced to adult custodial corrective services agencies in each state and territory in Australia.

Included in the National Prisoner Census are prisoners in the legal custody of corrective services who at the time of the census, were:

• absent on an authorised temporary leave permit except for Victoria and the Australian Capital Territory　
• absent from the correctional facility on a work release permit or program 　
• located in secure wards in a hospital or mental health institution outside the correctional facility administered under Corrective Services departments　
• periodic detainees until 2016 　
• serving post-sentence detention orders.

Excluded from the collection are:

• prisoners who were unlawfully absent from corrective services legal custody, e.g. escapees or prisoners who failed to return from an authorised temporary absence from a correctional facility 　
• prisoners whose legal custody had been transferred to another agency, e.g. police or mental health institutions　
• prisoners who were retained at court at the time of the census are excluded from the data.

The types of correctional facilities and programs where prisoners are held vary between the states and territories.

Included in the collection are:

• gazetted adult prisons in all jurisdictions 　
• periodic detention centres in the Australian Capital Territory until 2016　
• work camps in Queensland, Western Australia and the Northern Territory 　
• cells in court complexes administered by corrective services in New South Wales 　
• transitional centres in New South Wales and Victoria　
• gazetted police prisons in the Northern Territory which are administered and controlled by the Director of Corrective Services　
• police lock-ups under SA Corrective Services department　
• mental health facilities administered under Corrective Services departments

Excluded from the collection are persons held in facilities administered and controlled by other agencies:

• police lock-ups, police prisons and cells in court complexes 　
• immigration detention centres 　
• home detention programs 　
• military prisons 　
• mental health facilities administered under health departments　
• juvenile facilities, including those under the authority of adult corrective services

From 2019, in all states and territories persons remanded or sentenced to adult custody are aged 18 years and over. Persons under 18 years are treated as juveniles in most Australian courts and are only remanded or sentenced to custody in adult prisons in exceptional circumstances. In Queensland, prior to 2018, 'adult' referred to persons aged 17 years and over. From February 2018 onwards, persons aged 17 years are being transitioned from adult correctional facilities into the Queensland juvenile justice system over a two-year period. In 2019, there were no 17 year-olds in Queensland adult correctional facilities.

Persons included in the National Prisoner Census were counted in the state or territory in which they were held in custody regardless of which state or territory imposed the sentence being served.

The statistics in this publication are derived from information held in administrative systems which are maintained by corrective services agencies within each state and territory.

The Australian Bureau of Statistics (ABS) has developed national standards for corrective services statistics to support the comparability of data between states and territories. While efforts have been made to attain maximum comparability between states and territories, some issues with jurisdictional comparability remain due to different legislative and administrative recording practices in the states and territories.

The National Prisoner Census is a census of all persons in the legal custody of adult corrective services in all states and territories as at midnight 30 June of the reference year.

An episode is defined as the period from an offender's latest date of reception into the custody of a corrective services agency for a particular offence(s)/charge(s) until the person is released from custody. When a prisoner breaches parole conditions and is returned to custody, a new episode is deemed to have commenced and a new reception date is provided. Post-sentence detention is deemed to be a continuation of the existing episode.

For the years prior to 2013, Western Australia treated persons who breached parole and returned to custody as continuing a prior episode. From 2013, Western Australia data are based on the standard definition of an episode. From 2017, Western Australia no longer codes breach of parole to ANZSOC and instead provides the original offence for that episode.

The legal status of an offender is determined by the warrant(s) or court order(s) which provide the legal basis for the detention in custody of the offender. Some offenders may have more than one type of warrant issued against them; therefore, it is possible for an offender to have dual status (e.g. under sentence for some offences and awaiting appeal results for others, or under sentence and awaiting deportation). The counting rules for determining the legal status of an offender are as follows:

• If the prisoner has been sentenced for any offence then this takes precedence over any other offence(s)/charge(s) for which the prisoner is unsentenced.
• If the sentenced prisoner has appealed against all of his/her sentences then that prisoner is counted as under sentence.

The aggregate sentence is the longest period that the convicted prisoner may be detained for the current sentenced offences in the current episode. This is also the maximum sentence length for a convicted prisoner for the current episode. Charges pending which are likely to extend the current episode are ignored. Where a new episode is counted when a person returns to prison from a breach of parole, the unexpired sentence of the prior episode can be incorporated into the aggregate sentence for the new episode.

The expected time to serve is the period of imprisonment which a convicted prisoner is expected to serve and, in most cases, refers to the time between the date of reception for this episode and the earliest date of release.

Date of reception is defined as the date the prisoner was received into prison in a state or territory for the current episode. Counting rules for persons returning to custody from an escape attempt or parole are as follows:

• If a prisoner is recaptured and returned to custody following an escape, the date of reception for the current episode is the date the prisoner was received into prison prior to the escape, except in Tasmania. In Tasmania, if a prisoner is recaptured and returned to custody following an escape, it is treated as the start of a new episode and the date of reception for the current episode is the date the person was returned to custody.　
• If parole is revoked resulting in an offender returning to custody, the date of reception is the date the person was returned to custody. That is, a new episode is deemed to have commenced.

The minimum term is the period that must be served before the prisoner is eligible for release from custody to parole, and the difference between the maximum and minimum term is the period that will be served on parole if the prisoner is released at his or her earliest eligibility date. While parole is generally granted at the earliest eligibility date, prisoners may be denied parole for some or all of the period up to the expiry of their maximum term. For both fixed and maximum-minimum sentences, the period actually served in custody may be less than the stated time to serve where administrative mechanisms such as sentence remissions are applied.　

The time a prisoner is expected to serve in custody depends upon the sentence(s) originally handed down, the system of remissions and the forms of parole available in the various states and territories and whether any time was spent in custody prior to reception (for example, time on remand or in police custody). The rules governing date of release are complex and differ between the states and territories. 　

Expected time to serve is not calculated for prisoners sentenced to an indefinite term or to life where no minimum term has been fixed.　

For New South Wales:　

• Under current legislation, a court setting a term of imprisonment is required to first set a 'Sentence' period and then may set a 'Non-Parole' period. A court may decline to set a 'Non-Parole' period. These sentences are referred to as 'Sentence/Non-Parole' and ‘Fixed' sentence types respectively.　
• Those prisoners with a sentence of three years or less, being a sentence that has a non-parole period, are automatically released from custody at the expiry of the non-parole term. Those with a non-parole term greater than three years may be released by the State Parole Authority at any time after serving the non-parole term.　

For Victoria:　

The calculation of the expected time to serve is based on either the non-parole period set by court or by the aggregate sentence and the following:　

• If a court sentences an offender to be imprisoned in respect of an offence for the term of his or her natural life or a term of two years or more the court must, as part of the sentence, fix a period during which the offender is not eligible to be released on parole, unless it considers that the nature of the offence or the past history of the offender make the fixing of such a period inappropriate. 　
• If a sentence of less than two years but not less than one year is imposed, the court may set a non-parole period. 　
• The non-parole period must be at least six months less than the term of imprisonment and must be in respect of the aggregate sentence that the offender is liable to serve under all the sentences imposed. Time spent in detention prior to the commencement of the sentence counts toward the time to be served under sentence if the sentencing judge so orders. 　
• Pursuant to section 74 of the Corrections Act 1986, the Adult Parole Board of Victoria may release an offender upon the expiration of the non-parole period but release on that date is at the discretion of the Board.　

For Queensland:　

• With the exception of sentences of indefinite length, earliest release date is calculated based on the date the prisoner is eligible for parole pursuant to the Corrective Services Act 2006 (CSA) and the Penalties and Sentences ACT 1992. The CSA prescribes two types of parole – court ordered parole and board ordered parole. Different parole eligibility date calculations apply depending on the offence for which the person has been convicted and the length of the term of imprisonment given by the sentencing court. Parole is the only form of early release available to prisoners in Queensland. Mandatory minimum terms of imprisonment also apply for particular offences in Queensland.　
• Fixed periods of imprisonment apply for prisoners sentenced to a term of imprisonment of life before the prisoner is eligible to apply for parole. These range between 15 and 30 years, depending on the offence.　

For South Australia:　

• The Truth in Sentencing legislation implemented in August 1994 determines the way in which release dates are calculated. This legislation requires prisoners with an aggregate sentence of five years or more to formally apply to the Parole Board for release on parole. It enables the Parole Board to release prisoners with an aggregate sentence of five years or more at its discretion and provides directions for the judiciary to take the abolition of remissions into account when ordering sentences. Prisoners with a non-parole period (NPP) and an aggregate sentence of less than five years are paroled automatically. 　
• Release dates for prisoners are calculated as follows: where a prisoner has not had a NPP, the earliest date of release is the aggregate sentence end date. Where a prisoner has a NPP and an aggregate sentence of less than five years, the earliest date of release is the end date of the NPP. Where a prisoner has a NPP, and an aggregate sentence of five years or more, the earliest date of release is the earliest date the prisoner can be released by the Parole Board. If this date has expired and no further release date has been set by the Parole Board, the earliest date of release becomes the aggregate sentence end date, which in the case of Life or Other indeterminate sentences would be unknown.　

For Western Australia:　

Expected time to serve is determined differently depending on whether a parole term has been specified and the length of sentence as follows:　

• For sentences where parole is specified the expected time to serve is calculated between the sentence start date and the Earliest Eligibility Date (EED)　
• For non-parole sentences of less than 12 months if imposed prior to 1 July 2017 or 6 months if imposed thereafter, the calculation of the expected time to serve is based on half of the maximum sentence as the prisoner may be released on Short Term Parole at the EED after serving 50% of the sentence.　
• For non-parole sentences equalling 12 months if imposed prior to 1 July 2017 or 6 months if imposed thereafter, the prisoner must serve the full term and the expected time to serve is calculated up until the sentence expiry date (maximum date)　
• For non-parole sentences of more than 12 months if imposed prior to 1 July 2017 or 6 months if imposed thereafter, the prisoner may be eligible for release on a Re-Entry Release Order (subject to approval by the Prisoners Review Board), and the expected time to serve is calculated up until the RRED.　
• If the EED or RRED have passed, any release date set by the Prisoners Review Board is taken into account. If the Prisoners Review Board has not set a release date but has set a review date, that review date is used. If there are no such dates, sentence expiry date is used.　

For Tasmania:　

• The calculation of expected time to serve is based on the totality of all sentences less remissions which may be granted on eligible sentences. A remission of the whole or any part of a sentence is not to exceed 3 months, exceed one-third of the total period of imprisonment or reduce the sentence below three months.　
• A prisoner may be eligible for remission of more than one sentence during an episode of imprisonment.　
• The calculation of earliest date of release does not take into account eligibility for parole. The actual date of release from custody may be earlier than the calculated earliest date of release if a prisoner is eligible for parole. The actual date of release from custody may also be later than the calculated earliest date of release, based on remission eligibility and how much is granted.
• If a court fails to make an order in relation to parole the prisoner will be ineligible for parole on that sentence.　

For the Northern Territory, Expected time to serve is calculated based on their non-parole period (if this date has elapsed the EDR becomes the full-term expiry date), a fixed release date or an earliest release date based on a suspension of sentence after a set period of time.　

For the Australian Capital Territory prior to 2009, prisoners sentenced in the Australian Capital Territory and who were held in New South Wales prisons were subject to New South Wales calculations for date of release. As of 30 June 2009, Australian Capital Territory prisoners were no longer held in New South Wales prisons.

While all states and territories have access to indeterminate sentences as a sentencing option, the degree to which that sentencing option is utilised varies considerably across states and territories, particularly for Homicide and related offences. This can impact on mean and median aggregate sentence length and expected time to serve data as prisoners with an indeterminate sentence are excluded from these calculations. 

Mean and median sentence length data (both aggregate sentence and expected time to serve) presented in this publication represent only prison sentences being served by prisoners on the night of 30 June of the reference year. They are not representative of all prison sentences imposed by the Criminal Courts. The mean and median sentence length of prison sentences being served by prisoners on the night of 30 June will be longer compared to the mean and median sentence length of all prison sentences handed down by the Criminal Courts. This is a consequence of the fact that prisoners sentenced to longer prison terms are more likely to be enumerated in the Prisoner Census. Prisoners that are sentenced to shorter prison terms over the course of a financial year are more likely to have completed their term by the 30 June, whilst prisoners that are sentenced to longer prison terms are more likely to remain in prison (including prisoners sentenced to imprisonment in previous financial years who are still serving their term). For sentence length data relating to all defendants given a prison sentence by Criminal Courts during a financial year, please refer to Criminal Courts, Australia.

Mean and median aggregate sentence length data relate to sentenced prisoners with a maximum-minimum, fixed term or fine default sentence. They exclude prisoners with an indeterminate, life with a minimum or post-sentence detention sentence.

Mean and median expected time to serve data relate to sentenced prisoners with a maximum-minimum, fixed term, life with a minimum or fine default sentence. They exclude prisoners with an indeterminate or post-sentence detention sentence.

Time on remand data in this publication refers to time on remand to date as at 30 June of the reference year and not the total time spent on remand. Time on remand is influenced by a number of factors, particularly the time it takes for a case to come before a court. Mean and median time on remand is not representative of the total time spent on remand for all prisoners.

In 2017, the methodology for the calculation of time ranges for Aggregate sentence length, Expected time to serve and Time on remand was updated, following a review of the process and is now considered more accurate. The impact on the data was minimal.

The ABS conducts an ongoing program of quality assurance to monitor and improve the quality of Indigenous status data in corrective services agencies. While the ABS has published Aboriginal and Torres Strait Islander status data in Prisoners in Australia for a number of years, ongoing quality assurance is required to better understand the level of accuracy over time.

The quality of the Indigenous status information collected and recorded in corrective services agencies is assessed against the ABS Standard Indigenous Question (SIQ). The SIQ is based upon self-identification by the individual who comes into contact with corrective services agencies. The SIQ requires individuals to be asked “Are you of Aboriginal or Torres Strait Islander origin?” The answers to the SIQ can be ‘No’, ‘Yes, Aboriginal’, or ‘Yes, Torres Strait Islander’. If the offender is of Aboriginal and also of Torres Strait Islander origin, both responses should be identified. If an offender does not supply an answer to this question, or is not asked, the Indigenous status field should be recorded as 'not stated'.

For custodial corrections, current indications are that corrective services agencies in all states and territories, except Western Australia, ask prisoners about their Aboriginal or Torres Strait Islander status in line with the SIQ as described above. In Western Australia, in some instances Indigenous status may be recorded from information previously collected.

Some people in custody have their Indigenous status recorded as unknown on the information systems of corrective services agencies, as their status has not been able to be obtained. 

The court of sentence/remand data refer to whether the offender was sentenced or remanded to custody by the Higher Courts, such as the Supreme Court, the District or County Court or the Lower Courts such as the Magistrates' or Children's Courts. The rules adopted for coding the level of court are: 

• Where an episode comprises orders/sentences imposed by more than one court level, the level of court relating to the MSO is recorded.
• An appeal court is recorded when it has altered the length of sentence of the MSO, except Western Australia.
• For prisoners not under sentence, the level of court which has issued the most recent remand in custody warrant is recorded.
• Prisoners held under a Department of Immigration order or under the authority of the Parole Board are recorded as 'other'.

National corrections statistics are compiled in order to maximise comparability across states and territories. Although differences have been overcome through the introduction of national standards, some legislative, interpretive and processing differences remain.

For New South Wales:

• Data for the Kariong Detention Centre were included for the first time in 2015. Prior to 2015 the Centre, although still operating under the authority of Corrective Services NSW, housed juvenile prisoners as the Kariong Juvenile Correctional Centre and these prisoners were excluded from this collection.
• Periodic detention is a form of custody or order where a sentenced prisoner is required to be held in custody on a part-time basis. New South Wales abolished periodic detention on 1 October 2010, and as such, the number of periodic detainees has been decreasing since then. Prior to and including 2016, there were a small number of periodic detainees in New South Wales and these were included in prisoner totals (where relevant). From 2017 onwards, these prisoners were out of scope for the collection and have been excluded from prisoner totals.
• In 2012, New South Wales introduced improved coding procedures for sentenced and unsentenced prisoners whose most serious offence/charge fell within ANZSOC Division 16 Miscellaneous Offences, specifically those related to import / export regulations. As a result of these changes, caution should be exercised when comparing data from 2012 onwards for Miscellaneous offences for New South Wales with previous years.
• In 2014, New South Wales improved their quality assurance procedures for coding prisoners with life sentences. As a result, a higher proportion of prisoners are being coded correctly to Life with a minimum sentence, rather than to Life with an indeterminate term. This may impact on mean and median expected time to serve. Caution should be taken when making comparisons of data prior to 2014 with subsequent years.

For Victoria:

• In July 2005, the Children and Young Persons (Age Jurisdiction) Act 2004 came into force in Victoria. This Act changed the definition of a child from a person under 17 years of age to a person under 18 years of age. As a consequence, the definition of an adult also changed for corrective services data to a person aged 18 years and over. Data prior to 2006 includes persons aged 17 years.
• Data for Victoria for 2015 were provided for the morning of 30 June (i.e. the midnight muster from 29 June) due to data anomalies on the census date.

For Queensland:

• From 2018, Queensland extracted data directly from the courts which has led to coding changes. This has impacted on some Most Serious Offence coding, most notably Theft and related offences, thus caution should be taken when comparing Most Serious Offence data prior to 2018. These changes are expected to have a marginal impact at the Division level.

For South Australia:

• In 2013, South Australia introduced improved coding procedures for life sentences, resulting in a shift of sentence types from Life with an indeterminate term to Life with a minimum sentence. This may impact on mean and median aggregate sentence length and expected time to serve. Caution should be taken when making comparisons of data prior to 2013 with subsequent years.
• From 2014, South Australian data shows the impact of legislative changes from the Correctional Services (Miscellaneous) Amendment Act 2012 and changes to administrative procedures. This has resulted in an increase in South Australian prisoners held for ANZSOC 1513 Breach of suspended sentences and a decrease in those held for 1522 Breach of parole.

For Western Australia:

• In 2013, Western Australia implemented the use of the standard definition of an episode to their data. The impact of this change has been an increase in the proportion of prisoners with a most serious offence of ANZSOC 1522 Breach of parole, and an increase in the proportion of prisoners with prior imprisonment.
• From 2017, breach of parole is no longer coded to an ANZSOC code in Western Australia. When a prisoner breaches parole in Western Australia, the original offence will be reported under the relevant ANZSOC code. Consequently, the number of breaches of parole has decreased to zero for 2017. Caution should therefore be exercised when comparing offence type data from 2017 onwards with previous years.
• From 2016 to 2017, Western Australia reviewed the classification of prisoners relating to Life indeterminate and Life with a minimum. Previously, all prisoners serving a life sentence were classified as Life indeterminate however the review found that a minimum period for the sentence was specified by the judge in the majority of cases. Therefore, prisoners were reallocated from Life - indeterminate to a Life with a minimum sentence type. This change has impacted expected time to serve as Life with a minimum is included in the calculation of this variable. Caution should be taken when making comparisons of data from 2016 onwards with previous years.
• In 2017, Western Australia changed the way most serious offence/charge was selected for both sentenced and unsentenced prisoners. For both sentenced prisoners and unsentenced prisoners, offence/charge is determined by the application of the NOI, whereas previously offence ranking was based on ANCO. For unsentenced prisoners, Western Australia now report data in line with the counting rules which stipulate that only charges for a prisoner's current episode should be considered in the selection of a most serious charge. Caution should therefore be exercised when comparing offence type data from 2017 onwards with previous years.

For Tasmania, Most serious offence/charge data has been determined by the application of the NOI as the longest sentence cannot be attributed to a single offence. 

 For the Australian Capital Territory:

• Prior to 2009, persons sentenced to full-time custody by the Australian Capital Territory were usually held in New South Wales prisons. A new prison facility to house Australian Capital Territory persons sentenced to full-time custody, the Alexander Maconochie Centre, began taking prisoners on 30 March 2009. As at 30 June 2009, all Australian Capital Territory prisoners held in New South Wales prisons had been relocated to the new Alexander Maconochie Centre.
• From 1 July, 2016, periodic detention was no longer a sentencing option in the Australian Capital Territory. For operational reasons some prisoners housed at the ACT Periodic Detention Centre have been recorded against the Alexander Maconochie Centre. Counts for the former location in table 34 will therefore be understated and counts at the latter location will be overstated. From 2017, periodic detainees are not included in the data as this sentencing option no longer exists, and all ACT periodic detention orders have ended.
• Due to reporting methods, prior imprisonment in Australian Capital Territory includes both episodes of imprisonment under sentence and on remand. Thus, caution should be taken when comparing prior imprisonment in the Australian Capital Territory with other jurisdictions.

A post-sentence detention order is an order which subjects an offender to extended detention in prison following the completion of a custodial sentence. Post-sentence detention orders are made by a court where an offender has a history of serious offending, usually involving sexual or violent offences, and it determines that there is an unacceptable risk that the offender will commit further similar offences if released from prison. Prior to 2015, post-sentence detention data were included in the sentenced prisoner totals. From the 2015 publication onwards, data for post-sentence detention orders are reported in separate categories. For these:

• Post-sentence detention data are currently not applicable (as it is not available as an order) for the Australian Capital Territory.
• As post-sentence detention is deemed to be a continuation of the existing episode (see Counting rules), a detainee will only be recorded as having prior adult imprisonment if there was an earlier episode prior to the episode extended by post-sentence detention.

Aboriginal and Torres Strait Islander rates and non-Indigenous rates were revised in Table 18 and 20 using updated numbers based on the 2016 Census published in Estimates and Projections, Aboriginal and Torres Strait Islander Australians, 2006 to 2031.  

The Federal Offenders collection presents statistics about prisoners serving sentences for at least one federal offence as at 30 June of the reference year, for all states and territories in Australia. The summary also includes parolees serving sentences for at least one federal offence as at 30 June for all states and territories apart from Victoria. A range of information is presented on the demographic characteristics of the federal prisoners and parolees (age, sex, country of birth and Indigenous status).

The addition of the Federal Offender collection to Prisoners in Australia, provides a more complete picture of corrective services in Australia by identifying those prisoners who are sentenced to imprisonment, or are on parole, for federal offences but are being managed by state and territory corrective service agencies.

The scope of the statistics in this collection includes all offenders, either in custody in an adult correctional institution as a sentenced prisoner, or on parole (in the community), with at least one offence against Commonwealth/federal legislation in the current episode.

This differs to that of the Prisoner Census collection. The Prisoner Census includes all prisoners, both sentenced and unsentenced, with either state offences, federal offences or a combination of both. The Prisoner Census also excludes parolees.

Statistics for prisoners with a federal offence are presented by the state or territory in which they were held (for prisoners). Statistics for parolees with a federal offence are presented by the state or territory that has responsibility for them (for parolees).

The principal counting unit for the Federal Offenders collection is the offender. For this collection, an offender includes those that are sentenced (in prison) and those on parole. A sentenced federal prisoner is a sentenced prisoner in custody in an adult corrective services institution with at least one offence against Commonwealth/federal legislation in the current episode. A federal parolee is an offender on parole, in the community (i.e. not in custody) who is under the authority of the Corrective Services agency, with at least one offence against Commonwealth/federal legislation that has an associated Commonwealth parole order, in the current episode.

The collection counts those federal offenders in custody at an adult correctional facility (as a sentenced prisoner) and those with a parole order (in the community) at midnight 30 June each year.

The ‘current episode’ for an offender refers to an episode that is current on 30 June (of the reference year). The ‘current episode’ for an offender may contain only state offences; or only federal offences; or a combination of both state and federal offences. For this collection, only those offenders with one or more federal offence(s) in the current episode are in scope.

The inclusion of Victoria federal prisoner data was made possible by extracting information using a different method to other states and territories.

The Census and Statistics Act 1905 provides the authority for the ABS to collect statistical information, and requires that statistical output shall not be published or disseminated in a manner that is likely to enable the identification of a particular person or organisation. The requirement means that the ABS must ensure that any statistical information about individuals cannot be derived from published data. To minimise the risk of identifying individuals in aggregate statistics, a technique is used to randomly adjust cell values and summary variables. This technique is called perturbation and was applied to the Prisoner Census collection for the first time for the 2015 reference period. Perturbation involves small random adjustment of the statistics and is considered the most satisfactory technique for avoiding the release of identifiable statistics while maximising the range of information that can be released. These adjustments have a negligible impact on the underlying pattern of the statistics. 

After perturbation, a given published cell value will be consistent across all tables. However, the sum of the components of a total will not necessarily give the same result as the published total in a particular table. As such, proportions may add to more or less than 100%. Readers are advised to use the published totals rather than deriving totals based on the component cells. Cells with small values may be proportionally more affected by perturbation than large values. Users are advised against conducting analyses and drawing conclusions based on small values.

Perturbation has been applied to data presented in this publication, including historical data from 2005 onwards. Previously, a different technique was used to confidentialise the data and therefore there may be small differences between historical data presented in this issue of the publication and those published in previous issues.

Data on persons serving post-sentence detention orders have not been perturbed.

Imprisonment rates enable comparison of prisoner populations across states and territories at a point in time, as well as over time. Prisoner rates are expressed per 100,000 adult population, in accordance with international and state and territory practices. 

Rates for the adult prisoner population are calculated using the estimated resident population (ERP) for each of the states and territories, and total Australia (see National, state and territory population) as at 31 March of the relevant calendar year. All estimates and projections for the Australian Capital Territory exclude Jervis Bay Territory. Estimates and projections for Australia exclude the external territories of Christmas Island and the Cocos (Keeling) Islands, with the exception of table 22. For table 22, data on country of birth excluding external territories are not available. As a result, imprisonment rates by country of birth may be slightly lower than would be the case if these data were available.

From 2016, all crude imprisonment rate data are based on the perturbed population counts in these tables. Prior to 2016, data for crude imprisonment rates were based on unperturbed data. As per previous years, age standardised imprisonment rates are calculated from unperturbed data.

The formula for calculating the imprisonment rate is as follows:

The formula for calculating the ratio of two imprisonment rates is as follows:

In June 2013, the ABS 'recast' the historical ERP data for the September 1991 to June 2011 period, as a response to a methodological improvement in the Census Post Enumeration Survey. As a result, the rates per 100,000 adult persons in the Prisoner Census were recast and used ERP data based on the 2011 Census of Population and Housing. From 2018, rates per 100,000 adult persons in the Prisoner Census use ERP data based on the 2016 Census of Population and Housing. 

For population estimates and information on the methodology used to produce the ERP, see National, state and territory population. 

The Australian adult ERP used in the calculation of national imprisonment rates includes only persons aged 18 years and over who can legally be sentenced/remanded into an adult correctional custodial facility in each state and territory for that reference period. Prior to 2019, persons aged 17 years and over were included in the ERP for Queensland.

Prior to 2009, imprisonment rate data for the Australian Capital Territory are calculated on the basis of the total number of Australian Capital Territory prisoners held in New South Wales prisons, as well as in the Australian Capital Territory. New South Wales imprisonment rates data are only based on the count of New South Wales prisoners held in New South Wales prisons. For more information about Australian Capital Territory prisoners held in New South Wales prisons, refer to Data comparability and significant events.

Imprisonment rates by birthplace are calculated using country of birth data, this data is available on request from Migration, Australia. Caution should be taken when using these rates as age profiles may differ between countries of births, and using crude rates to examine differences between different countries of births may lead to erroneous conclusions being drawn about variables that are correlated with age. Age standardised rates are unable to be calculated for imprisonment rates by country of birth due to the small number of prisoners born overseas. 

In July 2019, the ABS published the backcast historical population estimates for Aboriginal and Torres Strait Islander persons, along with Aboriginal and Torres Strait Islander population projections in Estimates and Projections, Aboriginal and Torres Strait Islander Australians, 2006 to 2031. Rates have been revised in the 2020 publication with finalised rebased population estimates released using the results of the 2016 Census of Population and Housing. 

The adult ERP used to calculate Aboriginal and Torres Strait Islander imprisonment rates is based on Series B projections for 30 June of the relevant calendar year. This is consistent with the decision made in 2009, following consultation with the National Corrective Services Statistics Advisory Group and other stakeholders. The publication, Estimates and Projections, Aboriginal and Torres Strait Islander Australians, 2006 to 2031, suggests that Series B will be the most appropriate choice for most users.

Age standardisation is a statistical method that adjusts crude rates to account for age differences between study populations. The age standardised rates presented in this publication are based on the direct method of calculation, for more information please see Appendix 1 of the 2010 issue of Deaths, Australia. Age standardisation requires a minimum number of persons in each age group in order to calculate accurate rates, therefore age standardisation rates by multiple data items may not be possible, such as sex by Indigenous status by state. 

There are differences in the age distributions between Australia's Aboriginal and Torres Strait Islander and non-Indigenous populations with the former having a much younger population. In 2001, the proportion of Aboriginal and Torres Strait Islander people aged 40 years and over was 20%, compared with 44% of non-Indigenous people (and 43% of the total Australian population).

Due to these differing age profiles, using crude rates to examine differences between Aboriginal and Torres Strait Islander and non-Indigenous populations may lead to erroneous conclusions being drawn about variables that are correlated with age.

Based on comparisons across age groups, it is known that imprisonment rates decrease in older age groups, i.e. that the imprisonment rate is correlated with age. Further, based on a comparison of overall imprisonment rates between Aboriginal and Torres Strait Islander and non-Indigenous persons, it is likely that the imprisonment rate in the Aboriginal and Torres Strait Islander population will be higher because of the larger proportion of young people in the Aboriginal and Torres Strait Islander population.

Age standardised Aboriginal and Torres Strait Islander and non-Indigenous imprisonment rates have been presented in tables 17 and 18 of this publication. Ratios of the rates are also included in these tables. The ratio of crude rates included in table 20 can be compared with the ratio of age-standardised rates in table 18 to obtain an idea of the impact of the different age structures for the two populations on the likelihood of being imprisoned. If the crude ratio is substantially higher than the age standardised ratio, this in an indicator of the extent to which the Aboriginal and Torres Strait Islander population have higher proportions of persons in the age cohorts most likely to be imprisoned.

The standard population against which each population is age standardised is the total Australian Estimated Resident Population at 30 June 2001. The standard population is revised every 25 years; the next revision will be based on final data from the 2026 Census.

For sentenced prisoners in all states and territories except Tasmania, the Most Serious Offence (MSO) is the offence for which the prisoner has received the longest sentence in the current episode for a single count of the offence, regardless of the possible result of any appeals, and regardless of whether the sentence for that offence has actually expired at census date. Where sentences are equal, or the longest sentence cannot be determined, the MSO is determined by applying the National Offence Index (NOI). 

In Tasmania, magistrates and judges have the capacity to provide global sentencing, i.e. a combined sentence for all offences. An individual offence will not have an identified period of imprisonment, so it is not possible to identify the offence with the longest period of imprisonment. The MSO is therefore determined by applying the NOI.

From 2017, for unsentenced prisoners in all states and territories, the Most Serious Charge (MSO) is determined by applying the National Offence Index (NOI). Prior to 2006, all states and territories applied the Most Serious Charge (MSC) for unsentenced prisoners by determining the charge which carries the longest statutory maximum penalty. 

There are some state and territory variations.  In New South Wales and the Australian Capital Territory, if a prisoner is convicted but as yet unsentenced for at least one charge, the MSC is selected from those offences for which convictions have been recorded.

The offence information presented in this issue are based on the Australian and New Zealand Standard Offence Classification, 2011 (ANZSOC). The ANZSOC provides a uniform national framework for classifying offences across Australia for statistical purposes. The classification is a hierarchical structure allowing for varying degrees of detail to be published depending on the level of detail in the source information. 

The first release of this classification was the Australian Standard Offence Classification (ASOC97). Issues of the publication from 2001 to 2009 were based on ASOC97. Prior to 2001, offence data were based on the Australian National Classification of Offences,1985 (ANCO).

In 2008, the ABS released a second edition of the Australian Standard Offence Classification (ASOC08), which reflected changes to criminal legislation since the first edition. In 2011, the offence classification was updated from the Australian Standard Offence Classification (ASOC) to ANZSOC; however, changes were not made to the content of the classification. 

For the 2009 Prisoner Census, all jurisdictions implemented ASOC08 for the provision of offence data with the exception of Queensland and Western Australia which were still supplying offence data on ASOC97. This classification changes affected offence information for both sentenced and unsentenced prisoners, however analysis found that the overall impact to the time series prior to 2009 and to comparability across jurisdictions was minimal. For more information refer to the Technical Note in the 2009 issue of Prisoners in Australia.

For the 2010 Prisoner Census, both Queensland and Western Australia implemented ASOC08.

In this collection, the National Offence Index (NOI) is used to determine the most serious charge for unsentenced prisoners for all states and territories (see Most serious offence/charge). The NOI is a tool which provides an ordinal ranking of all offences in ANZSOC according to the perceived seriousness of each offence. The purpose of the NOI is to enable the representation of a prisoner by a single offence/charge in instances where multiple offences/charges occur for the same prisoner. 

The original NOI was introduced into the Prisoner Census in 2006, and was revised in 2009 to accommodate the changes made in ASOC08. An update to the NOI was published in 2018.  For the 2020 cycle this version of the NOI has been implemented by all jurisdictions except New South Wales and the Australian Capital Territory.

For further information about the NOI refer to National Offence Index, 2018. 

Country of birth information is classified according to the Standard Australian Classification of Countries (SACC), Second Edition. 

Prisoners born in South Sudan may be included in the counts for Sudanese-born prisoners. From 2018, the imprisonment rate for Sudan has been calculated using both Sudan and South Sudan Estimated Resident Population. Prior to 2018 the imprisonment rate was calculated using the Estimated Resident Population for Sudan only and thus caution should be used when comparing prior to 2018. 

Special tabulations may be produced on request to meet individual user requirements. For further information, contact the ABS National Information and Referral Services (NIRS) on 1300 135 070 or email client.services@abs.gov.au.