# Scottish prison population statistics technical manual

Information on data presented in the Scottish Prison Population Statistics reports from 2019-20 onwards.

## Analytical factors and measurements

Measurements

Factors

### Measurements

This section details the different measures made of the prison population in our statistical releases.

#### Individuals

A count of individuals indicates the number of unique individuals identified in each subgroup or combination of subgroups in the population in the course of an Analytical Period. Individuals are differentiated by their prisoner number.

A result of counting individuals in this way is that, in cases where an individual can occupy more than one subgroup in a given analytical period, a sum down a column or across a row in an analytical table may include some people more than once. For example, where a person spends time in multiple prisons in the course of an analytical period, they will be counted once in each prison they stay in. Adding together the individuals identified in all prisons will therefore exceed the total number of people imprisoned in that period.

Table 2.1 lists the analytical factors where a person may occupy multiple categories over the course of given period.

Table 2.1: Categories where an individual may have multiple values over the course of an analytical period

All Individual measurements relating to:

Measurements of Individuals Arriving:

Measurements of Individuals Departing:

Legal Status

Of No Fixed Abode

Journey Type

Index Offence Group

Local Authority

Journey Type-2

Index Offence

SIMD 2016 Decile Group

Index Sentence

Prison Establishment

#### Average daily population

The average daily population (ADP) over a given Analytical Period across the prison estate is the total number of prisoner days served, divided by the number of days in that period.

A person’s occupancy time within an Analytical Period – termed prisoner days – provides a weight for each prisoner in that analytical period.

When calculating the length of a Stint, some edge cases exist: for example when a prisoner enters and leaves prison on the same day. Since the time of entry and exit are not recorded in these data, this would count as zero prisoner days, and therefore zero weight. To account for this, such stints are assigned a value of 0.25 prisoner days (6 hours) by default.

Note: ADP measurements tend to obfuscate some of the differences between populations. For example, one population where a large proportion of a small number of individuals spend the whole year in prison is indistinguishable in this measurement from another population where a greater number of individuals cycles through prison for a short period of time. This effect is illustrated in Figure 6, where a single prisoner present for 6 consecutive days provides the same average daily population as 6 prisoners present for one day each.

Figure 6: Average Daily Population equivalence between many short-term and few long-term prisoners

#### Arrival

An arrival is when a period of imprisonment begins, and is only counted as such if the arrival occurred on a separate day from any previous departure.

These may be subdivided based on the Legal Status on the first day of an Occupancy Period, e.g. “Arrival to Sentenced”

Counting the number of Individuals arriving provides the number of unique individuals arriving in each subgroup in the analysis.

#### Departure

Departure is counted when an individual is absent from prison for one or more nights.

These may be subdivided based on the Legal Status on the last day of an Occupancy Period, e.g. “Departure from Remand”.

Counting the number of Individuals departing provides the number of unique departing in each subgroup in the analysis.

#### Status transition

A status transition is counted when an individual moves from a period on remand to a period serving a sentence without leaving custody over night. They may be counted alongside Arrivals to Sentenced status or Departures from Remand.

Counting the number of Individuals transitioning provides the number of unique individuals transitioning to or from remand in each subgroup in the analysis.

#### Time Served (Continuous)

All measurements of time served are based on continuous periods in SPS custody. Most commonly this measurement is reported as a factor in banded categories of Time to Departure or Time on Remand. Measurements are taken at Departure or (in the case of time on remand) upon Status Transition.

The measurements provided in prison population statistics reflect both the construction of the database – based on Occupancy Periods – and the distribution of the data.

Time served (in general) has a “long tail” distribution, meaning there are a few individuals in each cohort with very long periods spent in custody, while most will spend much shorter periods in custody overall. As a result of this distribution, the median is used as the more representative measure of central tendency (as opposed to the mean).

Following feedback from users and stakeholders, and to provide better information about the wider distribution, the 90th percentile is also provided in most cases to indicate the minimum stay for the 10% of the population staying the longest.

### Comparator population rates

Throughout the statistical releases, comparisons are made between the prison population and the general population in Scotland.  These are the sources for those comparisons:

#### Populations

The overall Scottish population, and local authority populations, are estimated by National Records of Scotland each year. These provide the basis for comparison and for population rates by local authority.

Where population rates are presented for countries other than Scotland they are drawn from external sources which should be referred to directly.

#### Ethnicity

The base population by ethnicity for 2011-12 is drawn from the 2011 Census. For subsequent years, ethnic group sizes are drawn from the pooled sample of Scottish Surveys Core Questions (SSCQ). Due to the SSCQ's methodology, population rates are based only on populations aged 16 or over. Using SSCQ data in this way also makes it possible to control for the increasing age (into adulthood) of those youngest minority ethnic populations when the census was taken in 2011.

The SSCQ estimates of ethnic group population sizes have an associated margin of error. The calculation of this margin of error is explained in the SSCQ technical report (section 1.3). The upper and lower confidence bounds provided with the prisoner population rate statistics indicate a 95% confidence that the true population rate lies between those two values. This is like saying if the SSCQ were performed 100 times, the estimated population of the different ethnic groups in Scotland would fall within those bounds in 95 of the 100 runs.

Data from the SSCQ is only available up to the calendar year 2019. For subsequent years these data have been reused. This will be subject to revision when the results of the census are made available.

### Factors

#### Prisoner age

Prisoner age is calculated based on dates of birth and the details of the Analytical Period.

In the first publication from this source, prisoner ages were based solely on the middle date of the analytical period, i.e. 1st October for each financial year. This has since been revised. Calculating ages in this way meant that a number of young people arriving after 1st October who had birthdays between that date and their arrival date were incorrectly identified as being under 16 years of age.

Prisoner ages are calculated as at 1st October where they arrived prior to that date (including those arriving in a previous analytical period). For those arriving after the 1st October, their age on arrival is used.

#### Age Bands

Age bands are based on Prisoner Age and grouped into five-year age bands except for the following bespoke categories:

• Under 16
• 16-17
• 18-20
• 21-22
• 23-24
• 75 and over

#### Gender

The collection of Prisoner Gender is determined by SPS recording policy.

#### Age groups

Division by age into young people (under 21) and adults.

Prisoner Age in this case is banded to separate individuals classified as Young Offenders (20 and under).

#### Age-Gender

Age-Gender are based on a combination of Gender and Age Groups.

#### Marital status

Self-reported relationship status.

Where the number of individuals in a category is small it is necessary to combine these groups for the purposes of reporting. This is likely to mask variations within aggregated groups.

The combined categories are listed in Table 2.2.

Table 2.2: Self-Reported Marital Status Categories

Reporting Category

Source Category

Single

Single

Married/Partnered

Married

Common Law

Civil Partnership

Divorced/Separated

Divorced

Separated

Dissolved Civil Partnership

Surviving Partner

Widowed

Surviving civil partner

#### Ethnicity

Self-reported ethnic identity.

Categories are provided by the harmonised principles developed for the census 2011 by the Government Statistical Service and National Records of Scotland. Where the number of individuals in a category is small it is necessary to combine these groups for the purposes of reporting. This is likely to mask variations within aggregated groups.

The combined categories are listed in Table 2.3.

Table 2.3: Self-Reported Ethnicity Categories

 Reporting Category Source Category White Scottish, English, Welsh, Northern Irish or British Irish Gypsy or Irish Traveler Polish Any other White background Asian, Asian Scottish or Asian British Pakistani, Scottish Pakistani or British Pakistani Indian, Scottish Indian or British Indian Bangladeshi, Scottish Bangladeshi or British Bangladeshi Chinese, Scottish Chinese or British Chinese Other African, Caribbean or Black African, Scottish African or British African Other African Groups Caribbean, Scottish Caribbean or British Caribbean Black Other Mixed or multiple ethnic groups Mixed White and Asian Mixed White and Black African Mixed White and Black Caribbean Mixed Other Other ethnic groups Arab, Scottish Arab or British Arab Middle Eastern Other Other Ethnic Origin

#### Disability

Self-reported disability.

#### Sexual Orientation

Self-reported sexual orientation.

Where the number of individuals in a category is small it is necessary to combine these groups for the purposes of reporting. This is likely to mask variations within aggregated groups.

Non-heterosexual categories are combined into a single category: Lesbian, Gay, Bisexual and Other (“LGB & Other”). Those declining to respond are retained separately.

#### Armed forces

Self-reported history of employment in the armed forces. These data are provided as the two-fold variable "Veterans" indicating past service in any of the UK armed forces, and as the four-fold variable indicating which of the services the individual served in.

Service Number is collected and checked by SPS where possible, but is not extracted by Scottish Government analysts.

#### Legal status

One of Untried (U), Convicted Awaiting Sentencing (A) or Sentenced (S).

Convicted Awaiting Sentencing is sometimes shortened to “CAS”.

In the 2-fold factor, Untried and CAS together are termed “Remand” (R), as distinct from "Sentenced" (S).

Time within an Occupancy Period is subdivided into different statuses based on the Legal Status Hierarchy.

#### Index offence

The offence category for which the longest sentence has been received over a period.

For prisoners on remand, the index offence is determined based on the average sentence associated with offence types from SG Criminal Proceedings.

For further information, see Index Offence derivation.

#### Index offence group

The Index Offence for which the longest sentence has been received over a period, grouped into higher-level categories. Groupings are provided in Table 2.4.

For further information, see Index Offence derivation.

Note: Offences in categories outside of group 1 (Crimes of Violence) may entail violence against their victims. Common Assault, for example, covers a wide range of behaviours that can include lower-level physical violence.

Table 2.4: Index Offence Higher-level Groupings

 Group Group Name SG50 Labels 1 Non-sexual crimes of violence Murder and culpable homicide Causing death by driving dangerously Serious assault and attempted murder Common assault Robbery Domestic Abuse (Scotland) Act 2018 Other non-sexual violence 2 Sexual crimes Rape & attempted rape Sexual assault Causing to view sexual activity or images Communicating indecently Threatening to or disclosing intimate images Indecent photos of children Crimes associated with prostitution Other sexual crimes 3 Crimes of dishonesty Housebreaking Theft by opening lockfast places Theft from a motor vehicle Theft of a motor vehicle Shoplifting Other theft Fraud Other dishonesty 4 Damage and reckless behaviour Fire-raising Vandalism Reckless conduct 5 Crimes against society Crimes against public justice Weapons possession (not used) Drugs - Supply Drugs - Possession Other crimes against society 6 Antisocial offences Threatening and abusive behaviour Racially aggravated conduct Drunkenness and other disorderly conduct Urinating etc. Community and public order offences 7 Miscellaneous offences Coronavirus restrictions Environmental offences Licensing offences Wildlife offences Other misc. offences 8 Road traffic offences Dangerous and careless driving Driving under the influence Speeding Unlawful use of vehicle Vehicle defect offences Seat belt offences Mobile phone offences Other road traffic offences

#### Index sentence

The longest single sentence being served during a period in custody.

Categories in banded index sentence length analysis are provide in Table 2.5.

For further information, see Index Sentence derivation.

Table 2.5: Index Sentence bands

 Index Sentence Band Low (days) High Description Untried - - Remand CAS - - Remand 0<=3 months 1 92 3<=6 months 93 184 6<=12 months 185 366 12<=18 months 367 548 18<=24 months 549 731 2<=4 yrs 732 1461 4 < 10 yrs 1462 3651 10+ yrs 3652 ∞ Life/OLR - - Any life sentence supersedes any determinate sentence or recall from the date it was first passed Recall - - Recalls supersede any determinate sentence for categorisation Unknown - - While the occupancy period has an associated sentenced warrant, the information on sentence duration is missing (Missing) - - There is no associated sentence information associated with this occupancy period

#### Prison establishment

CellWise data covers a period of transition in the prison estate, and includes historic information about populations in prisons that have since closed.

Prisons, or “Establishments”, included in the cellWise data are listed in Table 2.6 along with their approximate opening and closing dates.

The small capacity of Community Custody Units (CCUs) Bella and Lilias has led us to combine the two facilities in our statistical outputs to maintain the anonymity of the people staying there.

Table 2.6: Prisons in the scope of Scottish Prison Population Statistics

 Prison Establishment Opened Closed HMP YOI Aberdeen 1890 2014 HMP Addiewell 2008 HMP Barlinnie 1880 HMP YOI Cornton Vale 1975 2023 HMP Dumfries 1863 HMP YOI Edinburgh 1920 HMP Glenochil 1966 HMP YOI Grampian 2014 HMP Greenock 1907 HMP Inverness 1902 HMP Kilmarnock 1999 HMP Low Moss 2012 HMP Open Estate (Castle Huntly) 1946 HMP Perth 1812 HMP Peterhead 1888 2013 HMP YOI Polmont 1911 HMP Shotts 2012 HMP YOI Stirling 2023 CCU Bella 2022 CCU Lilias 2022

#### Of no fixed abode

Self-reported status. On arrival the individual is unable to, or declines to, provide a home address.

As with other geographical data, this measure is only available for Arrivals to custody in each Analytical Period.

#### Local authority

Local authority area of a prisoner's home address on arrival or, for those of no fixed abode, their local authority of residence.

As with other geographical data, this measure is only available for Arrivals to custody in each Analytical Period.

#### SIMD 2016 Decile Group

Area deprivation of home address, based on the postcode of those with a known address in the Geographical Data. As with other geographical data, this measure is only available for Arrivals to custody in each Analytical Period.

Scottish Index of Multiple Deprivation provides a typology of datazones in Scotland. This index is broken down into decile groups, with the first group representing the most deprived 10% of datazone areas and the 10th group representing the least deprived 10% of areas.

Some prisoners are excluded from this analysis:

• Unknown 1 - No match and NFA: The prisoner is of no fixed abode.
• Unknown 2 - No/partial postcode: the postcode data is incomplete and therefore cannot be matched to a datazone
• Unknown 3 – No match: The information provided in the postcode field cannot otherwise be matched to the datazone lookup file.

#### Journey type

Legal Status stages in an individual's journey during a period in prison: Untried (U), Convicted Awaiting Sentencing (A) or Sentenced (S).

Note: the simplified journey types presented here mask the ambiguity in legal status arising as a result of the one-directional Legal Status Hierarchy inherent in the cellwise data construction.

#### Journey type 2

Legal Status stages in an individual's journey during a period in prison: Remand (R), or Sentenced (S).

Note: the simplified journey types presented here mask the ambiguity in legal status arising as a result of the one-directional. Legal Status Hierarchy inherent in the cellwise data construction.

#### Liberation type

Table 2.7 lists the liberation descriptions provided by the prisoner record system, and the way they have been combined into higher level categories.

Some liberation types (e.g. “Sentence Served”) would be expected to follow a period in custody with the Sentenced Legal Status. However, due in part to the one-directional Legal Status Hierarchy inherent in the cellwise data construction, liberation descriptions do not always tally with the apparent legal status at the end of a listed Journey Type. For reference, in Table 2.7 we also provide the expected legal status of individuals leaving prison with different liberation types.

Table 2.7: Liberation types

 Expected Status On Liberation Summary Liberation Type Full Liberation Description Sentenced Sentence Served Lib Sentence Expired On licence/Parole/SRO/HDC Home Detention Curfew Lib On Licence Lib On Parole Lib To Supervised Release Order Appeal/Pardon Lib On Appeal Pardon Fine Paid Lib Fine Paid Remand Bailed To Bail Court/PF Court Lib From Court Procurator Fiscal Ambiguous Immigration/Deportation Lib Immigration Author Lib Dungavel Detention Centre Deported Repatriation Lib To Early Removal Scheme Other Deceased Lib To Mental Hosp Lib To List D Local Authority Discretionary Early Release Gender Reassignment Lib To Interlib Police Interview

People who have died in prison are counted in the “Other” category in Table 2.7. Deaths in custody are the subject of Scottish Prison Service reporting, available on the SPS website.

#### Time to Departure

Overall time in custody for a continuous occupancy period.

Upon Departure during an Analytical Period, we can determine the total continuous time spent by a prisoner in custody. Banding prisoner days from arrival to departure provides the categories for this variable, as follows:

• 0<=3 months - less than or equal to three months
• 3<=6 months - greater than three months and less than or equal to six months
• 6<=12 months - greater than six months and less than or equal to twelve months
• 12<=18 months - greater than twelve months and less than or equal to eighteen months
• 18<=24 months - greater than eighteen months and less than or equal to two years
• 2<=4 yrs - greater than two months and less than or equal to four years
• 4 + yrs - greater than four years

This information should not be taken as an indicator of time served on a specific sentence. Prisoners may leave custody during their sentence for a night or more for a variety of reasons, thereafter returning to complete their time in custody.

#### Time on remand

Time spent with a remand status (untried/convicted awaiting sentence) before the first intersecting sentencing warrant.

Upon Departure from a remand Legal Status, or upon Status Transition to the sentenced population, we can determine the total continuous time spent by a prisoner on remand. Banding prisoner days from arrival to departure or transition provides the categories for this variable, as follows:

• 1 day or less
• 2-7 days
• <=2 weeks
• 3 weeks
• 4 weeks
• 5 weeks
• 6 weeks
• 7 weeks
• 8 weeks
• 9 weeks
• 10 weeks
• 71-140 days
• > 140 days

#### Financial Year

Financial years run from the 1st April to the 31st March.