Transparency data

Transparency Data: Updated estimates of children with a parent in prison

Published 10 September 2026

Applies to England and Wales

1. Executive Summary  

This publication provides updated estimates of children with a parent in prison following previously released estimates in July 2024. These figures are published as transparency data using a revised methodology to provide a more accurate estimate of the number of children identified. 

Parental imprisonment is recognised as an adverse childhood experience, and children may face more challenges compared with their peers who have not experienced having a parent in prison. In its 2024 Labour Party manifesto, the government committed to effectively  identify  children and provide  support  to children affected by parental imprisonment. This followed the Prisons Strategy White Paper (2021) and Farmer Review (2017) which committed to investigating data sharing on prisoners and their families. Following the manifesto commitment, the government is taking action with pilots set up by the Ministry of Justice (MoJ) to trial the operational recognition and support of children with a parent in prison, working closely with local authorities.

From October 2023 to October 2024, an estimated 219,000 children had a parent in prison in England and Wales, with an estimated 2% of school-aged children experiencing parental imprisonment. This is an increase on our previous estimate of 192,000 children for a 12-month period from October 2021 to October 2022. Both estimates have been rounded to the nearest thousand.

The methodological improvements have enabled a more robust and reliable estimate. Previously, estimates were calculated by identifying prisoners with parental status and applying an average multiplier of 1.77 children per prisoner. In contrast, the current methodology directly identifies and counts children with a parent in prison from the available data, making it more accurate instead of relying on an assumed average.

Methodological improvements provide greater confidence in the latest estimate; however, it remains subject to certain assumptions, primarily the small adjustment for undercount. An increase in children is also expected to correspond to the increase in the prisoner population between 2021 and 2024.

The vulnerability of this population is highlighted by the finding that at least half (53%) of children identified were at risk of harm from the imprisoned parent. Recognition of who these children are, as well as an early offer of family support is important for prevention of harm. Only around half (48%) of identified children had an associated address from MoJ records or HMRC Child Benefit where linked. It cannot be assumed the child lives at the same address as for the parent in prison.

For children identified and with address records (48% of children), when ranked by proportion of children with a parent in prison, many of the local authorities with high proportions of children with a parent in prison were in northern urban areas, and coastal towns. These findings are indicative and should be interpreted with caution given lack of full coverage of all children’s addresses. However, these findings are backed by data from the Office of the Children’s Commissioner also indicating higher proportions of children with a parent in prison in similar areas. The reasons behind this finding are likely to be complex, partly driven by deprivation but requiring further investigation.

While these are early indicative findings, this work aims to ultimately enable all affected children to be identified and offered support. To this end, future work may include further data linkage, improved information extraction from free-text records, or better data collection. Without address data, information cannot be passed to local authorities to identify and support children.

Around a third (32%) of children were recorded on prison visitor lists. This is comparable to findings from the charity Children Heard & Seen, which reports two thirds (67%) of children the charity support do not visit prison at all. The findings show there was a difference in visits of children to female prisoners (31%) versus male prisoners (50%). There may be a range of reasons for this difference, such as distance being a barrier to visiting female prisons, family circumstances, or care arrangements.

The estimates presented here have been produced using administrative data held by His Majesty’s Prison and Probation service (HMPPS) and His Majesty’s Revenue and Customs (HMRC), by linking information from prison records with Child Benefit data.

A related publication providing descriptive statistics on a Welsh cohort of children affected by parental imprisonment and the nature of families affected has been released alongside this data.

1.1 Limitations  

When interpreting these estimates, readers should consider the limitations and assumptions made which are:

• the definition of parental status is broad and unless stated otherwise, references to “parent” include legal guardians and others with wider parental responsibilities (section 2.2)

• some children may not be included in either HMPPS data systems or HMRC Child Benefit claimant data. A validation exercise (section 3.3) indicated coverage of all children was around 85% and an associated coverage adjustment has been applied to derive the final estimate of 219,000 children. This final estimate is rounded to reflect uncertainty and, whilst not comparable to the previous estimate due to differences in methodology, the increase in numbers of children with a parent in prison reflects a corresponding increase in prison population

• our location data covers a subset of the total number of identified children, where an address was available, and provides an indication of geographical distribution only. Location (address) data is essential for enabling information on children to be passed to local authorities to enable families to be offered support

2. Introduction

2.1 Background

Children of people in prison are sometimes described as serving a “hidden sentence”. Children affected by parental imprisonment can experience numerous difficulties, as evidenced across various health, educational and social outcomes. It is therefore vital to work across government to enable children who have experienced parental imprisonment to achieve and thrive. Research consistently shows that children with a parent in prison have poorer educational outcomes than their peers, including lower attainment, higher rates of exclusion, and reduced likelihood of progressing into further or higher education. Children may also encounter disrupted relationships with teachers and peers, often compounded by stigma and secrecy surrounding parental incarceration.

The Farmer Review on “The Importance of Strengthening Prisoners’ Family Ties (2017)” pointed to significant gaps in official data on the impact of parental imprisonment, noting that there was not a process within the Ministry of Justice (MoJ) to centrally record the number of children affected. This is not an isolated issue; other countries are also grappling with use of data to understand the scale of children affected by parental imprisonment. Other approaches rely on partial and fragmented data rather than a full, accurate picture. To date, this publication is the most comprehensive estimate for England and Wales.

Strengthening family ties, where it is safe and appropriate to do so, is a vital aspect of the work of MoJ and HMPPS to help support family cohesion and improve outcomes for children and families. It also enables positive reengagement with society upon release where maintaining a relationship with a parent who has been imprisoned is safe, appropriate and in the child’s best interests. Following the manifesto commitment in 2024, government is taking action. Pilots have been started to trial MoJ led operational recognition of children with a parent in prison and Department for Education (DfE) led support.

2.2 Definition of a parent

There is no single formal definition of children affected by parental imprisonment. However, in 2020, the Council of Europe provided some international standards and guidelines on children with imprisoned parents recommending a broad, rights-based understanding that encompasses:

• children of imprisoned parents whose rights and wellbeing are affected

• children of imprisoned caregivers, not only biological parents

• children affected at all stages of the criminal justice process

A wide range of stakeholders were consulted on definitions of parental imprisonment. The definition remains broad to capture as many scenarios as possible where children could be identified. These children include:

• prisoners’ child(ren)

• prisoners’ current or prisoners’ ex-partner’s child(ren)

• prisoners who are linked by address – where a prisoner has previously shared an address with a child(ren)

The last category is filtered to account for children from each prisoner’s wider social network (for example grandparents, aunts, uncles), where the child has lived with the prisoner. This enables identification of more children. For children’s records, those aged under 18 as of 1 October 2024 were retained.

3. Methodology

Estimating the number of children with a parent in prison is methodologically complex and is fully described in the Technical Guide section 4. However, the process can be broadly summarised in three steps:

1) Extracting information on prisoners’ social relationships, including their children, using a combination of structured data and Artificial Intelligence (AI) analysis of case notes.

2) Linking these records to HMRC Child Benefit data.

3) Adjusting the resulting estimates to account for children who may not have been identified by comparison with independent data sources.

3.1 Data sources

All analysis is based on a cohort of individuals who spent any amount of time in prison between 1 October 2023 and 1 October 2024 referred to here as the prisoner cohort^{[footnote 1]}. This includes all individuals in custody, whether they were on remand, sentenced, on recall or held for a civil offence^{[footnote 2]}.

Information on prisoners’ social relationships was extracted from five HMPPS prison and probation data sources for the 154,620 individuals in the cohort. The data sources and data fields are fully described in section 2 of the Technical Guide.

3.2 Use of Artificial Intelligence (AI)

Information about family and social relationships is recorded both in structured data fields and in unstructured free-text records, such as NOMIS prison officer and NDelius probation officer case notes. Large Language Models (LLMs) were used to identify and extract relevant information from these free-text sources. LLMs are a form of AI trained on large volumes of text and can process and analyse human language to perform specific tasks.

The estimates presented in this publication are based on a refined methodology developed from lessons learned during our previous publication. Two key insights emerged from previous learnings.

First, linkage between the predominantly male prison population and HMRC Child Benefit claimant records was less effective than anticipated because Child Benefit is often claimed by the child’s mother or another family member who acts as the primary carer. In 2025, 85% of claimants were women compared to 15% of men. The methodology was adapted to extract information on prisoner social relationships including their children.

Second, recent advances in LLMs now enable highly accurate extraction of information from case notes and the efficient processing of large volumes of unstructured text. The methodology was adapted to use the latest models for more accurate and more complete coverage of information about prisoners and their children.

All data was used for analytical purposes only, providing overall estimates relating to prisoners and their children in aggregate form. The analysis was not used to identify individual prisoners or children.

3.3 Linking to HMRC data

The HMPPS data has been linked to HMRC Child Benefit data to create a linked MoJ-HMRC data share. This is to provide more complete knowledge on children identified in the structured and unstructured data sources. The full social network of each prisoner was examined as Child Benefit is claimed by the legal guardian, so in some cases the claimant may not be the parent themselves under eligibility rules. Child Benefit data was chosen as it has very high coverage, with 90% of all children eligible for Child Benefit being claimed for, and provides accurate information on children connected to a claimant, their ages, and their residence.

3.4 Coverage of children identified

The data sources linked together enable a count of children with a parent in prison. The estimates were further adjusted to account for children who could not be identified through structured data, unstructured case notes, or HMRC data matching. This undercount was measured by comparing children identified against two independent data sources:

• Operation Paramount in the Thames Valley region. Operation Paramount is a police-led initiative that connects children of imprisoned parents with third-sector organisations that can provide support. It has been running for several years and is taken to be a robust source of information on children with a parent in prison for that area

• the Children’s Commissioner for England’s one-off school census. Schools were asked (amongst other questions) to provide an estimate of the numbers of their pupils with a parent or carer in prison in summer term 2024, this was used to sense check the extrapolation for undercount from Thames Valley region to nationally

A detailed analysis comparing a sample of children from the linked MoJ-HMRC data with Operation Paramount suggested the majority of children (and their associated siblings under 18) had been identified (see Technical Guide section 4.6 for more detail). This analysis was used to derive a modifier of 1.2 extrapolated for national undercount. Using the count of 182,724 children and the modifier indicates that there were 219,000 children with parents in prison between 1 October 2023 and 1 October 2024. This figure represents the best estimate after accounting for missing data and potential under-identification, with uncertainty reflected through rounding to the nearest thousand.

3.5 Comparability with previous estimate

The 23/24 estimate of 219,000 children with a parent in prison is around 14% higher than the estimate of 192,912 in 2021/22. The prison population has increased by 12% from 78,324 (30 June) in 2021 to 87,360 (28 June) in 2024, as a result the number of prisoners’ children might also be expected to increase.

The methodological improvements have enabled a more robust and reliable estimate. Previously, estimates were calculated by identifying prisoners with parental status and applying an average multiplier of 1.77 children per prisoner. In contrast, the current methodology directly identifies and counts children with a parent in prison from the available data, making it more accurate instead of relying on an assumed average.

As a consequence of these methodological changes, the estimates in this publication are not directly comparable with previous figures. Using the current approach, a ratio of 2.44 children under the age of 18 was found for every prisoner who is a parent, which is substantially higher than the 1.77 ratio used in the earlier publication.

4. Findings

The figure of 182,724 children identified is unadjusted for undercount and will be used to provide insights for the remainder of the report.

4.1 Data sources overview

Full descriptions of these data sources are detailed in the Technical Guide.

4.2 The count and percentage of children identified in each data source

Some children appear in multiple data sources hence the counts do not sum to 182,724. Similarly, the percentages in each column do not add up to 100% as the same child will appear in multiple data sources.

Source	Description	Count of identified children	Percentage of total identified children	Percentage of total identified children solely in this data source
OASys Child at Risk	Information on children who may be at risk of serious harm from the offender.	97,277	53%	7%
OASys full assessments	Detailed assessments carried out throughout an individual’s time in prison.	94,484	52%	8%
OASys BCST (Basic Custody Screening Tool)	Conducted within 72 hours of arrival into custody. Asks prisoner about caring responsibilities (yes/no).	1,779	1%	<1%
Prison case notes	Notes recorded in free text by prison officers.	17,100	9%	1%
Prison visitor list	Individuals approved to visit prisoners.	59,212	32%	7%
Probation case notes	Notes recorded in free text by probation officers.	47,872	26%	0%
Probation contact list	Key contacts associated with individuals under probation supervision.	8,711	5%	2%
HMRC Child Benefit	Government benefit provided for those responsible for bringing up a child.	67,146	37%	14%

In line with findings from the previous publication, no single source contains details on all identified children. The unstructured data sources were generally more informative but also more difficult to process. Note that the findings reflect success of information extracted on names, ages and addresses of children (to reflect current methodology) rather than a yes/no binary indication of parental status (old methodology). This meant the BCST was less useful for the current methodology.

Around a third (32%) of children were listed on prison visitor lists, this is comparable to findings from the charity Children Heard & Seen with two thirds (67%) of children supported who do not visit prison at all.

OASys Child at Risk identifies over half of all children. These are children for whom safeguarding - risk of harm by the prisoner to the child - is an issue and indicates a vulnerable child population who are victims or may become victims of their parents’ criminal behaviour. Harm may include sexual offences against children, child abuse or neglect, grooming or exploitation, domestic abuse where children are exposed. This finding aligns with the “Characteristics of children affected by household imprisonment in Wales” publication which shows that family and household vulnerability factors were over five times more common in the cohort of Welsh children impacted by household imprisonment including parental domestic abuse, substance misuse, mental ill health, physical ill health, and learning disabilities.

Further findings presented are:

• a comparison between the annual data and a single time point snapshot during the year (section 4.3)

• children’s age group (section 4.4)

• sex breakdown of prisoners and visits from children (section 4.5)

• an indication of the spatial distribution of children of prisoners (section 4.6)

4.3 Annual estimate versus snapshot estimate

The total number of children identified with a parent in prison seen in our single day estimate based on the date 28 June 2024 is 102,785. This was calculated by counting children who were associated with a person in prison on that date.

4.4 Breakdown of children of prisoners by children’s age group

Age information was available for 80% of all those identified with a parent in prison. Alongside the figure of 182,724 children identified, 35,327 individuals were over the age of 18 when the parent was imprisoned.

Of the children (under 18) with a parent in prison:

• 32,579 are aged 0-4 years (pre-school)

• 77,178 are aged 5-11 years (primary school)

• 71,521 are aged 12-17 years (secondary school)

• 148,699 are aged 5-17 years (school age)

• 60,824 are aged 10-14 years (“Characteristics of children affected by household imprisonment in Wales” publication age range)

The linked data estimates around 1.4% of school-age children (5 to 17 years old) experienced parental imprisonment between October 2023 and October 2024. Taking into account potential undercount, this would relate to around 2% of children with a parent in prison. This percentage is consistent across both primary school (5 to 11 years old) and secondary school (12 to 17 years old) age ranges, as well as the age range used in the “Characteristics of children affected by household imprisonment in Wales” publication (10 to 14 years old). However, at the preschool age (0 to 4 years old) this percentage is lower at 0.9%. This may be due to factors such as under-reporting by prisoners or difficulty establishing family ties with very young children, as well as delayed claiming behaviour leading to lower coverage in HMRC Child Benefit for this age range.

4.5 Sex breakdown of prisoners and visits from children

Information on the sex of prisoners was available in almost all cases (99.8%). Between 1 October 2023 and 1 October 2024 there were 8,707 female prisoners (6%) and 145,913 male prisoners (94%).

Of these, 4,374 (50%) female prisoners were parents (average 2.1 children) and 70,444 (48%) male prisoners were parents (average 2.5 children). These percentages by sex are consistent with the previous publication with 55% and 53% prisoner parents for females and males respectively. This average number of children per prisoner parent is also consistent with findings from the “Characteristics of children affected by household imprisonment in Wales” publication which shows the majority of children living in households of 4 to 5 people.

Analysis of the contact and visitor lists shows that male prisoners more commonly had contact with their children.

Of those prisoners who are parents:

• 31% of female prisoners had either been visited by their child or had them in their contacts list (1,350 individuals)

• 50% of male prisoners had either been visited by their child or had them in their contacts list (34,913 individuals)

The lower proportion of female prisoners with recorded child visits or contacts may partly reflect the smaller number of women’s prisons in England and Wales, which can mean women are held further from family and social networks than male prisoners, making visits more difficult. There may be other reasons for fewer visits.

4.6 An indication of the spatial distribution of children identified as having a parent in prison

Around half of children identified with a parent in prison had available location data (48%). This is as expected given the primary purpose of HMPPS data is to record information on the prisoner rather than their children, and not all children could be matched to an HMRC Child Benefit record. These findings should be interpreted with care and are intended to indicate broad geographic patterns rather than provide definitive local authority estimates.

Local authorities with the highest number of children identified as having a parent in prison generally correspond to larger child populations. The following section discusses the proportion of children identified, rather than absolute numbers. This approach accounts for differences in both the size of the child population and also helps avoid patterns being driven by concentration of prisons in certain regions, particularly the North of England.

• Proportions of children identified in English local authorities:

It is important to note that this section relates to the 48% of children who had address information. In England, indicative findings suggest the majority of local authorities with the highest proportions are located in northern urban areas, and coastal towns. Local authorities considered to be in the North were those within the North East (England), North West (England) and Yorkshire and The Humber (England).

Deprivation rankings (IMD 2025) were compared with the proportion of children with a parent in prison across local authorities in England. The analysis indicated a moderate positive relationship between deprivation and the prevalence of children with a parent in prison. This suggests that children with a parent in prison are more likely to live in more deprived areas; however, the relationship is not strong enough to indicate that deprivation alone drives the observed variation. Other factors are also likely to influence the distribution of this population across local authorities.

For comparison, the independent data collected through the Children’s Commissioner for England’s school census also indicates that the highest proportions of children identified as having a parent in prison are seen in northern urban areas, and coastal towns. However, five London local authorities also had high proportions - not seen in the linked MoJ-HMRC data. A source of this disparity might come from the location data used. The Children’s Commissioner school census uses the address of the school, whereas the linked MoJ-HMRC data uses the address of the child’s home which might account for differences in less residential local authorities, such as central London.

Further work to explore spatial disparities in England requires gathering better data on where the children live. Address data enables families to be contacted so that they can be offered support. While limited and indicative data suggest children in the North of England have disproportionately higher rates of parental imprisonment, the discrepancy for London indicates better data recording on children’s addresses is needed.

Counts and proportions of children with a parent in prison by local authority can be found in the Annexes (Table 1 with associated Map). and should be interpreted with caution as they relate to only children with address information recorded.

• Children outside of England:

England houses the largest population of those serving a custodial sentence and has the most children with a parent in prison in Great Britain and Northern Ireland. Although the prisons in Scotland and Northern Ireland are not covered in this analysis, children of those serving custodial sentences in England are seen in these countries.

Available children address data within Wales suggests there was not the same geographical bias to the proportions of the total child population identified. Similarly to England, the Welsh Index of Multiple Deprivation (WIMD 2025) was compared with the proportion of children with a parent in prison across local authorities in Wales. This yielded a moderate positive relationship between deprivation and the prevalence of children with a parent in prison. The independent data collection of the Children’s Commissioner school census did not cover Wales and could not be used for this analysis.

Findings from the “Characteristics of children affected by household imprisonment in Wales” publication also found a higher proportion of children in the study cohort experienced two or more forms of deprivation (52% vs. 20%).

5. Further development

Significant improvements in methodology since our last publication means a more robust estimate of the number of children with a parent in prison has been possible. However further improvements could provide more detail around the children and families identified.

5.1 Refining the methodology

The matching process confirmed that some prisoners had children, but because some prisoner records matched multiple children, it was not possible to reliably identify the specific child or analyse detailed child and family information. The methodology could be refined to send over better data to ensure a higher match rate with HMRC Child Benefit records.

Further work to refine the definition of parental and caring relationships would improve comparability with other estimates of parental imprisonment and with administrative data sources that use different definitions of parenthood and caring responsibility.

Additional improvements could provide more complete information on children’s address, the duration of contact between prisoners and children, and the nature of those relationships.

5.2 Coverage of data sources

While the case notes in the Prison National Offender Management Information System (NOMIS) and National Delius (NDelius), the prison and probation administrative databases, are the two main sources of free-text information in HMPPS databases, there may be additional data sources that could be used to produce the aggregated estimates. Court data may be more useful in future as systems are updated and more information is recorded digitally.

6. Further information

This publication series is the first time HMPPS data covering the entire prison population for England and Wales has been linked together to produce an estimate for the number of children with a parent in prison.

Please direct any enquiries or feedback to the team at CAPI@justice.gov.uk

7. Footnotes

1. This provides a cumulative prevalence of prisoners and is therefore not comparable with prison population statistics (published as part of the ‘Offender management statistics’ series), which measures a snapshot of the number in prison on one day. ↩

2. Further information on these categorisations is available in MoJ’s 2016 Guide to Offender Management Statistics (PDF). ↩