UK homelessness: 2005 to 2018

• Homelessness and housing are devolved policies across the UK and each country produces data and statistics according to their own legislative framework, which vary substantially between countries; differences in reporting categories, category definitions and reporting timeframes are key barriers to comparability across all areas of homelessness statistics.

• Trends suggest that the number of people reported to be seeking help for homelessness has been reasonably stable in the UK since around 2013.

• Trends in the number of temporary accommodation placements in the UK are diverging with figures rising in England, variable in Wales, and relatively stable in Scotland and Northern Ireland over the last five years.

• When applicants are found accommodation, Scotland and Northern Ireland rehome almost all applicants in social housing while England and Wales rehome across the social and private rented sector.

• Characteristics of homeless households are relatively similar across the UK despite legislative and reporting differences; the largest categories for households seeking help for homelessness are: single-person households without children; those aged between 25 to 49 years; and males.

• Households seeking help for homelessness with a main household member aged over 60 years have increased in recent years while those with a main household member aged under 24 to 25 years have decreased.

• Trends suggest an increase in the complexity of homeless household needs in recent years, particularly in relation to physical and mental health conditions.

• The reasons cited for loss of previous home are broadly consistent across UK countries and are primarily changing relationships or disputes within households.

• Estimates for rough sleeping based on street counts suggest numbers are increasing in England, Wales and Northern Ireland, but estimates based on homelessness applications suggest numbers are decreasing in Scotland.

Understanding homelessness and estimating how many people are homeless supports the development of social policies to help some of the most vulnerable people in society. Homelessness affects a wide range of people, covering not just people sleeping rough but also those in temporary accommodation, sleeping at friends’ houses or sofa surfing, living in unfit dwellings, or who are soon to be without a permanent home.

Currently, each country in the UK is responsible for producing statistics according to their own legislative framework for housing, which means that definitions of homelessness and the data collected vary between nations. Previous work has found that developing a consistent definition of homelessness is not feasible in the short term and better guidance on current statistics is therefore needed for data users.

This article uses the interactive tool and statistical guidance for homelessness produced by the Government Statistical Service (GSS) Harmonisation Team to compare official homelessness statistics across the UK. It aims to begin building a UK-wide picture of homelessness and to discuss comparability and coherence within currently available data. This article also discusses homelessness housing duties, which describes those who are assisted or housed according to country-specific legislation, households in temporary accommodation, and people who are sleeping rough. Characteristics of homeless households and the reasons for homelessness are also discussed.

This analysis has been written by the Office for National Statistics (ONS) in consultation with the producers of the quoted statistics: the Ministry of Housing, Communities and Local Government; the Scottish Government; the Welsh Government; the Northern Ireland Housing Executive (NIHE); and the Northern Ireland Department for Communities.

GSS Strategy Delivery welcome your views on the content and format of this article, please email us at GSS.housing@ons.gov.uk or email the various statistics producers about their statistics.

This section looks at the number of people approaching local authorities or housing associations for help with homelessness and the different types of legislative duties in place in each country of the UK.

In general, across the UK, a person is regarded as homeless if they lack a secure place in which they are entitled to live or they are not reasonably able to stay where they currently live. Individuals or households who are homeless or threatened with homelessness will be owed a duty to help secure housing, which depends on country-specific legislation and the individual circumstances of the applicant. Legislation is described in detail in the interactive tool for homelessness and the main points are summarised here.

Both England and Wales have recently undergone major changes in legislation for homelessness, in 2018 and 2015 respectively. In both countries, households seeking help for homelessness undergo an assessment to determine if they are homeless or threatened with homelessness¹.Those who are homeless, or those who become homeless following prevention efforts, are owed a relief duty lasting for up to 56 days where local authorities take steps to secure suitable accommodation. Households that remain homeless after 56 days are assessed for whether they are in priority need; those who are in priority need and are unintentionally homeless are owed a main housing duty to secure suitable accommodation.

Priority need refers to legal categories of individuals or groups who are homeless with specific circumstances, such as households with pregnant women or children or those with physical or mental health conditions. Categories are described in detail in the Government Statistical Service (GSS) harmonisation report.

In Scotland, local authorities have a duty to assist all households that are homeless or threatened with homelessness¹ irrespective of priority need, from 2012 onwards. All those who are unintentionally homeless are owed a main housing duty to secure suitable accommodation.

In Northern Ireland, households presenting as homeless undergo legislative tests that decide whether they are unintentionally homeless or threatened with homelessness,¹ eligible² and in priority need. Those who qualify following testing are owed a full³ housing duty to secure suitable accommodation.

In all countries, advice and guidance is available for all households who are not accepted for a legislative housing duty or those who are found to be intentionally homeless.

Findings

Number of households seeking help for homelessness

Owing to substantial differences in legislation, as described earlier, and differences in data recording (see Annex: Data sources), it is not possible to estimate the number of households currently seeking help for homelessness across the UK as a whole⁴. However, trends in individual country data do provide insight into the number of homeless households over time.

Time series data suggest that an increase in prevention-based strategies to tackle homelessness may contribute to a fall in the number of households seeking help with homelessness (Figures 1a to 1d).

England and Wales experienced a drop in the number of decisions made on homelessness applications around 2004, following the Homelessness Act 2002, which placed extra prevention duties on local authorities (Figures 1a and 1c). Conversely, legislative changes widening the number of households owed a homelessness duty in England from 2018 onwards may have increased the number of applications per quarter. England reported an increase from around 20,000 to 25,000 decisions made on applications per quarter prior to the changes, compared with around 60,000 assessments per quarter following the new legislation.

Scotland experienced a drop in the number of decisions made on homelessness applications from 2010, following an increase in prevention activities to prepare for the abolition of priority need in 2012 (Figure 1b).

In contrast, the number of households presenting as homeless in Northern Ireland has been relatively stable over the past 10 years with smaller fluctuations compared with the other UK countries (Figure 1d). Northern Ireland has not experienced a fall in homelessness following the introduction of preventative approaches, as described earlier, but has also not seen substantial increases in homelessness following, for example, welfare reform.

Housing duties

Whether a household is owed a main housing duty is typically dependent on a household member being in priority need. However, direct comparison between housing duties in different countries is difficult, primarily owing to differences in when priority need is assessed, if at all.

England and Wales assess priority need after households have flowed through prevention and relief duties and remain homeless. Northern Ireland assesses priority need upon application. Scotland does not assess priority need. Prior to legislative changes in 2018, around half of the applications in England were accepted for a main housing duty based on priority need assessment (52% April 2017 to March 2018). Following legislative changes, local authorities now support households to secure accommodation irrespective of priority need for a minimum of 56 days before a main housing duty is owed. The proportion of applications accepted for a main housing duty in the quarter directly after legislative changes fell to around 10% (April to September 2018).

Similar to England, Wales also saw a drop in the number of main housing duty acceptances following the legislative changes introduced in 2015. Between April 2017 and March 2018, 8% of all outcomes for homeless households were accepted for a main housing duty, as households flowed through prevention and relief duties prior to priority need assessment. However, under current legislation in both England and Wales, households can have multiple outcomes under each duty and so figures are only approximations for main housing duty acceptances.

In Scotland, 73% of applications were accepted as being owed a main housing duty between April 2017 to March 2018. Scotland accepts all applicants who are unintentionally homeless irrespective of priority need, which likely accounts for a higher proportion of acceptances compared with other countries during the same period.

In Northern Ireland, 65% of applications were accepted as being eligible, unintentionally homeless and in priority need, and therefore owed a housing duty. Northern Ireland has seen an increase of 20% in the number of acceptances in recent years from 9,878 (April 2012 to March 2013) to 11,877 (April 2017 to March 2018). The reasons for this increase are not clear but research is currently being carried out by the Northern Ireland Housing Executive (NIHE).

Priority need categories

Although legislative categories for priority need are broadly similar across the UK, the distribution of categories that households are recorded under varies between countries and is likely influenced by the time at which priority need is assessed.

Categories primarily include households with children or pregnant women and households with a vulnerable household member, such as those who are old or young adults, those experiencing violence, or those with physical or mental health needs. Scotland abolished priority need assessment in 2012 and has not been included in this analysis.

A full breakdown of priority need categories is available in the GSS harmonisation report. Categories have been grouped where possible to aid comparisons.

In England, prior to legislative changes in 2018, local authorities assessed all homelessness applications for priority need status. In the period April 2017 to March 2018, the majority of households in priority need were reported as having dependent or expected children, with 72% accepted for this reason and 27% accepted owing to a household member vulnerability⁵ (Figure 2a). However, this distribution may change in the coming years following the introduction of prevention and relief duties before priority need assessment.

In Wales, priority need status is only assessed for households where the local authority was unsuccessful in securing accommodation likely to last for more than six months. Between April 2017 and March 2018, 45% of homeless households who underwent priority need assessment were reported as having dependent or expected children and 54% as having a household member vulnerability under the same categories as England⁵ (Figure 2b).

In Northern Ireland, priority need assessment takes place after a household presents as homeless and is deemed eligible for assistance². Between April and September 2018, 32% of households in priority need were reported as having dependent or expected children and 67% as having a household member vulnerability, including violence⁴ (Figure 2c).

Please note: On 18 September 2019, the subtitle for Figure 2c was corrected from April 2017 to March 2018 to April to September 2018.

Intentionality

All countries assess whether applicants are unintentionally or intentionally homeless. Assessment is at different stages of application in each country and data are not directly comparable. However, the proportions of households that are deemed to be intentionally homeless are consistently low in all countries.

Between April 2017 and March 2018, England reported that 8% of all main housing duty decisions on applications were deemed to be households who were intentionally homeless and in priority need. During the same period, in Wales 4% of homeless households who were unsuccessfully relieved were subsequently found to be intentionally homeless and in priority need; while in Scotland, 4% of decisions on applications were deemed to be intentionally homeless, irrespective of priority need. Between April and September 2018, in Northern Ireland 6% of rejected assessment decisions were deemed intentionally homeless and in priority need.

Conclusions and comparability

Overall, the number of people seeking help for homelessness based on the number of application decisions or number of presentations has been reasonably stable in the UK since around 2013. However, producing a UK-wide figure for the number of people or households who are homeless is not currently possible. Owing to substantial differences in legislation and data collection, alongside within-country changes to data collection, direct comparison of the number of people making homelessness applications and the number accepted for a housing duty is limited. This is particularly evident with acceptance rates for a main housing duty, which are heavily dependent on priority need and prevention activities.

The reasons for main housing duty acceptance, based on priority need category, vary between countries with England accepting primarily for dependent children, prior to legislative changes in 2018, and Northern Ireland primarily for a vulnerable household member. This may be because of differences in legislation or data recording; for example, priority need is a single-choice category, so children may be preferentially chosen over a priority need that could require further assessment.

Changes in recent years mean that relief duties in England and Wales are now broadly similar to each other and more comparable with Scotland. However, data collection methods and differences in the type of information collected still impede comparability. For example, case-level data in England and Scotland, aggregated data in Northern Ireland, and outcome-based data in Wales.

Notes for: Homelessness housing duties

1. Threatened with homelessness within 56 days (England and Wales); within 2 months (Scotland); and within 28 days (Northern Ireland).

2. All countries assess eligibility based on immigration status; Northern Ireland also assesses eligibility based on unacceptable behaviour in a previous Northern Ireland Housing Executive (NIHE) tenancy or antisocial behavioural order.

3. For simplicity and comparison with other countries, full housing duty is referred to from here as “main housing duty”.

4. Seeking help for homelessness” is used here as a general term to include all country-specific cohorts of people reported to be homeless. Full details of each cohort and the data used for each country are provided within the figures and text.

5. Data for England and Wales include vulnerability owing to old age, young age, physical or mental illness, or violence. Data for Northern Ireland combine vulnerable household member and violence categories. Data do not include those homeless in an emergency.

In general, outcomes for homeless households are not easily comparable across the UK. Differences in legislation mean that the data collected and reporting time frame in each country vary substantially.

England reports quarterly data on the outcomes for all homeless households, irrespective of priority need, at the end of an initial period of help to find secure accommodation. The most recent quarter (October to December 2018) reported that 42% of households were found accommodation for a period of at least six months, of which 44% were found accommodation in the social rented sector and 24% in the private rented sector. However, case-level data for England is experimental, and a substantial proportion were also recorded under “Not Known” (21%). This percentage is likely to decrease as the data collection system becomes more established. Around 13% of households were recorded as being lost to contact during the initial homeless relief period.

Wales reports all data by outcome of application, both quarterly and annually. For the most recent quarter (October to December 2018), a comparable proportion (43%) of homeless households were successfully relieved, as was reported in England during the same period. Between April 2017 and March 2018, of the households that were successfully relieved an equal proportion were housed in the private rented sector and the social rented sector.

Scotland reports outcomes for homeless households irrespective of priority need in both their biannual and annual reports. Between April 2017 and March 2018, a large majority were found accommodation (69%), of which almost all were secured accommodation in the social rented sector (93%). Around 14% of households were either lost to contact or the outcome was not known.

Northern Ireland reports outcomes for households accepted for a housing duty over a six-month period. Between April and September 2018, the majority of households were reported as being on the waiting list (78%), which includes households placed in temporary accommodation. Of those found accommodation, a large majority (98%) were housed in the social rented sector.

Conclusions and comparability

There is limited comparability across outcomes for homelessness applicants owing to differences in legislation and reporting time frames. However, in general, when applicants are found accommodation, this tends to be in the social rented sector: Scotland and Northern Ireland rehome almost all applicants in social housing while England and Wales rehome across the social and private rented sector.

This section looks at the number of households living in temporary accommodation as an indicator for homelessness across the UK.

Temporary accommodation may be provided while an assessment decision is being made or while homeless households are waiting for longer-term accommodation. Eligibility for temporary accommodation differs across the UK: in England, Wales and Northern Ireland, temporary housing is largely provided for applicants who are in priority need; in Scotland, all applicants are entitled to temporary accommodation. Non-priority need households or households that have not been accepted for a housing duty may also be in temporary accommodation arranged by themselves, but they are unlikely to be captured within currently reported data.

England and Wales report a snapshot of the total number of households accommodated at the end of each quarter. Scotland reports both a snapshot of the total number of households accommodated at the end of each quarter and the total number of placements over the year. Snapshot data from 31 March of each year are used here for analysis. Northern Ireland reports the total number of accommodation placements over a six-month period (households may have multiple placements). The first available report from April to September 2018 has been used here for analysis, but data are not directly comparable with data from England, Scotland and Wales. In addition, Northern Ireland reports a snapshot figure for January 2019.

Findings

Analysis of the data suggests that trends in the number of temporary accommodation placements in the UK are diverging with figures rising in England, variable in Wales, and relatively stable in Scotland and Northern Ireland in recent years (Figures 3a to 3c). Households with children in temporary accommodation are rising in both England and Scotland and are relatively stable in Wales.

Total households

England has seen a steady increase of 60% in the number of households in temporary accommodation, from 50,430 (2012) to 80,720 (2018). Over the same period, the number of temporary accommodation placements in Scotland has remained stable, from 10,750 (2012) to 10,933 (2018). Conversely, Wales saw a drop in the number of households in temporary accommodation, from 2,770 (2012) to 1,875 (2016), and has subsequently reported an increase up to 2,052 (2018). However, legislative changes in Wales from 2015 placing a greater emphasis on prevention and delaying priority need testing until later in the application process may influence this time series. Data for the total number of placements in temporary accommodation in Northern Ireland are available from April 2013 to March 2014 onwards; figures have remained largely stable, with 2,989 placements in April 2013 to March 2014 and 3,024 placements in the year April 2017 to March 2018.

Households with children

The number of households with children¹ in temporary accommodation has been steadily rising in England, from 37,190 (2012) to 61,610 (2018), and falling in Wales, from 1,250 (2012) to 798 (2018). As noted, legislative changes in Wales may influence this trend. Over the same time period, Scotland initially saw a drop in the number of households with children in temporary accommodation, from 3,487 (2012) to 2,479 (2014), followed by an increase to 3,349 (2018). Time series data for households with children have not been published for Northern Ireland.

The proportion of households in temporary accommodation with children reported in March 2018 was similar in Wales (39%) and Scotland (31%), but it was substantially higher in England (76%) (Figure 4). Northern Ireland also reported a similar figure to Scotland and Wales for the percentage of placements recorded as families (42%) between April and September 2018.

In England, Wales and Northern Ireland, these figures are likely to be influenced by priority need consideration for temporary accommodation. Time spent in temporary accommodation may also influence proportions, for example, if different household types are found accommodation at different rates. Without priority need assessment, Scotland might be expected to accept more single people, thereby reducing the proportion of families with children in temporary accommodation relative to the overall pool.

Type of accommodation

Across the UK, there is a large range in the type of temporary accommodation including public sector and local authority or housing association stock; private sector accommodation, either leased or nightly-paid; and alternative forms of accommodation such as hostels, refuges, and bed and breakfast-style accommodation. All countries collect data on the type of temporary accommodation in which households are placed, though categories and definitions are not aligned.

Some accommodation types are intended for short term (days or weeks) and other types are intended for longer-term stays. The length of stay can therefore influence the proportion of households in different types of accommodation. In addition, the proportion of families with children in temporary accommodation (Figure 4) may influence the type of accommodation in which households are placed.

The largest category for households in temporary accommodation in England and Wales is private sector accommodation, either leased or nightly-paid. In England, more than half of temporary accommodation placements in March 2018 were in either nightly paid, privately managed accommodation (26%) or private sector accommodation leased by a local authority (31%), compared with placements in social housing stock (19%). In Wales, 38% placements were in private sector accommodation and 29% of placements were in social sector accommodation (March 2018). In Northern Ireland, a similar proportion of placements were in private sector lets (37%) over the period April to September 2018. Conversely, in Scotland, the majority of temporary accommodation placements are in social sector accommodation (59% in March 2018).

Shared facilities

Temporary accommodation with shared facilities describes placements where washing facilities, toilets and/or cooking facilities are shared with other individuals or families and are not self-contained. These placements may be in bed and breakfasts, hotels, hostels or refuges. Legislation in England, Scotland and Wales outlines the unsuitability of bed and breakfast accommodation with shared facilities for those in priority need (Wales) or households with pregnant women and dependent children (England and Scotland). The Northern Ireland Housing Executive (NIHE) also specifies that bed and breakfast accommodation should be used for the shortest time possible.

If an individual or family is homeless owing to risk of violence, bed and breakfast accommodation is important for safeguarding when there is a shortage of more suitable accommodation or prior to being moved to longer-term accommodation. Similarly, shared facilities can be beneficial to encourage support links in women’s refuges.

Snapshot data is available for all countries for shared facilities. In March 2018, England reported that 8% of households in temporary accommodation were in bed and breakfast-style accommodation, compared with 12% in Wales and 11% in Scotland (Figure 5). Northern Ireland reported that 3% of households in temporary accommodation were in bed and breakfast accommodation in January 2019.

Of the households in temporary accommodation with children, 9% of placements in England, 18% of placements in Wales and 2% of placements in Scotland were in either bed and breakfast or hostel accommodation (March 2018). In Northern Ireland, 17% of children in temporary accommodation were in either bed and breakfast accommodation (1%) or hostel accommodation (16%) in January 2019, though hostel accommodation arranged by the NIHE can provide families with self-contained, private facilities.

Length of stay

All countries report data for the length of time households spend in temporary accommodation, but the time frame and the cohort of applicants for which data are reported are different. It is therefore not possible to produce a UK-wide estimate for how long households are staying in temporary accommodation with currently available data.

England and Scotland report the length of time spent in temporary accommodation for households who have left (Figure 6). During the year from April 2017 to March 2018, the majority of temporary accommodation leavers in both countries had been there for less than six months: 58% and 64% in England and Scotland respectively. However, when using data for leavers, figures may be biased towards households that have been moved on or those that have had a short stay, particularly if more households enter temporary accommodation than leave. For example, households that have been in temporary accommodation for a long time may be more likely to remain compared to short-stay households.

Wales and Northern Ireland report a snapshot figure for households (Wales) or placements (Northern Ireland) in temporary accommodation. Wales reported that 68% of households had been in temporary accommodation for less than six months as of March 2018, and Northern Ireland reported 39% of placements were for less than six months as of January 2019 (Figure 6).

Major cities

A disproportionate number of households in temporary accommodation in England are reported to be housed by London local authorities (69%, March 2018), relative to the proportion of the population in England that are housed in London (approximately 16%²), though placements may be located outside of London. The second biggest contribution to the total number of households in temporary accommodation in England came from Birmingham (2.5%, March 2018).

In 2018, Glasgow and Edinburgh housed 20% and 13% of all households in temporary accommodation in Scotland, respectively, and Cardiff housed 31% of all households in temporary accommodation in Wales. A breakdown of the data for temporary accommodation placements by city is not currently available for Northern Ireland.

Conclusions and comparability

Across the UK, trends for the number of households in temporary accommodation are diverging with figures rising in England, variable in Wales (quarterly data at the time of writing indicate a potential rise), and stable in Scotland and Northern Ireland.

Legislative differences across the UK as to whether priority need is taken into account for temporary accommodation influences the ability to compare data, particularly for households with dependent or expected children. At present, trends suggest the numbers of households with children in temporary accommodation are rising in both England and Scotland and are stable in Wales. In 2018, there was a substantially larger proportion of households with children in temporary accommodation in England compared with other countries.

The type of temporary accommodation used varies between countries, with a larger proportion of private sector placements in England and Wales and a larger proportion of social sector placements in Scotland. Around 1 in 12 households in temporary accommodation are reported to be in bed and breakfast-style accommodation with shared facilities across the UK as a whole. Aligning definitions and reporting categories for accommodation type would improve comparisons in this area.

It is not possible to accurately estimate how long households are staying in temporary accommodation across the UK with currently available data, primarily owing to differences in data collection methods. At present, the majority of stays reported by all countries are less than one year in length. However, it is not possible to fully determine potential biases towards short stays.

Overall, differences in reporting time frames, such as snapshot data or placements over time, and reporting categories, such as types of accommodation and length of stay, are key challenges to making direct and robust comparisons between UK temporary accommodation data.

Notes for: Temporary accommodation

1. In all countries, “children” refers to children under 16 years of age. In Scotland, figures also include pregnant women.

2. Source: Estimates of the population for the UK, England and Wales, Scotland and Northern Ireland; Office for National Statistics.

This section looks at the characteristics of people seeking help for homelessness across the UK with respect to the type of households seeking help, the age and gender of the main household member, and the support needs for homeless households.

Demographic characteristics are influenced by the time at which information for homeless households is collected and the cohort included in the reporting. For example, demographic data based on households that have been accepted as being owed a main housing duty will show a priority need bias. Where possible, data collected with and without priority need assessment have both been used for analysis.

Household type

All countries report data on the type of households applying for help with homelessness, but collection methods are legislation dependent. Categorical breakdown of household type is not the same across countries and has been grouped here where possible to allow for comparison. Although data from each country is not directly comparable, trends in demographics are evident across the UK.

England reports the household type for main housing duty acceptances and, from April 2018 onwards, the household type for homeless applicants irrespective of priority need¹. Wales reports the household type for outcomes both with and without priority need assessment. Scotland reports the household type for homeless applicants. Northern Ireland reports the household type for households presenting as homeless, before assessment, and those accepted for a main housing duty, which includes priority need assessment.

In general, without priority need assessment, the largest category of households reported to be applying for help with homelessness across the UK were single people, ranging from 49% in Northern Ireland (April to September 2018) to 71% of households under homelessness relief duties in Wales (April 2017 to March 2018) (Figures 7a to 7e). Just under a third of households applying for help in all countries were families, either single parents or couples with children².

Dependent or expected children are major reasons a household may be accepted for a main housing duty, based on priority need (see Section 3: Homelessness housing duties). As a result, demographic data reported for households that have undergone priority need assessment show a greater proportion of families compared with data reported before or without priority need assessment¹. In England, families were the largest category of homeless households among those who underwent priority need assessment (Figure 7b, April 2017 to March 2018); in Wales and Northern Ireland, single-person households without children were the largest category but with a smaller majority than figures reported for households that had not undergone priority need assessment (Figures 7d to 7e; April 2017 to March 2018 in Wales and April to September 2018 in Northern Ireland).

Age

All countries publish data on the age of the main household member, either with or without priority need assessment, where applicable. Banded age categories vary between countries but have been grouped together to produce comparable age bands where possible.

Across the UK, young adults aged under 25 years accounted for around a quarter of households applying for help with homelessness or being accepted for a housing duty (Figures 8a to 8e). Between April 2017 and March 2018, in England 21% of main housing duty acceptances were aged 16 to 24 years³ and in Scotland, 25% of applicants were aged 16 to 24 years. For households under homelessness relief duties in Wales, just over a quarter (28%) had an applicant aged 24 years or under between April 2017 and March 2018. Northern Ireland reported that 32% of single-person applicants were young adults aged between 16 to 25 years (April to September 2018).

Above the age of 25 years, age breakdown is only available for England and Scotland. In both countries, between April 2017 and March 2018 the majority of acceptances or applicants were aged between 25 and 44 years (England) and 25 and 49 years (Scotland) (Figures 8a to 8c).

Young adults

In England, Scotland and Northern Ireland, the last five years show a downward trend in the number of young adults applying for help with homelessness or being accepted for a main housing duty (Figures 9a to 9c), where they are the main applicant of the household.

In England, main housing duty acceptances among 16 to 24 year olds have decreased by 29% from 16,820 acceptances (April 2012 to March 2013) to 12,020 acceptances (April 2017 to March 2018 (Figure 9a). As data represent accepted households only, young adults in this instance would be accepted on the basis of a priority need, such as being a care leaver or having a physical or mental health need.

However, independent of priority need, Scotland and Northern Ireland show a comparable reduction in applications from young people. In Scotland, applications from 16 to 24 year olds have decreased from 12,494 (April 2012 to March 2013) to 8,884 (April 2017 to March 2018), a decrease of 29% (Figure 9b). Northern Ireland reports the age of single-person applicants and has also seen a fall in number among those aged 16 to 25 years: from 4,189 single-person applicants (April 2012 to March 2013) who were young adults, to 3,001 (April 2017 to March 2018), a fall of 28%, similar to those seen in England and Scotland (Figure 9c).

Time series data for young adults in Wales are available from April 2015 onwards only but conversely show an increase in the number of outcomes under relief duties for which the applicant is aged 16 to 24 years, from 2,031 (April 2015 to March 2016) to 3,153 (April 2017 to March 2018). However, as households may have multiple outcomes, this may not reflect a trend in applicants.

Older adults

In England, Scotland and Northern Ireland, the last five years show an upward trend in the number of older adults or pensioners applying for help with homelessness or being accepted for a main housing duty (Figures 9a to 9c), where they are the main applicant of the household.

England has seen an increase in the number of main housing duty acceptances for those aged over 60 years, from 1,800 (April 2012 to March 2013) to 2,500 (April 2017 to March 2018; Figure 9a). Northern Ireland has seen a similar increase in the number of pensioner households presenting as homeless, from 1,875 (April 2012 to March 2013) to 2,445 (April 2017 to March 2018 (Figure 9c).

Scotland has also seen a slight increase in older homelessness applicants, though to a lesser degree than England and Northern Ireland, from 1,278 (April 2012 to March 2013) to 1,391 (April 2017 to March 2018) (Figure 9b).

Time series data specifically for older adults in Wales is not currently available.

Please note: On 18 September 2019, the subtitle for Figure 9b was corrected from "Percentage change in main duty acceptances by age of main applicant, England, April 2012 to March 2013 compared with April 2017 to March 2018" to "Percentage change in applications by age of main applicant, Scotland, April 2012 to March 2013 compared with April 2017 to March 2018".

Please note: On 18 September 2019, the subtitle for Figure 9c was corrected from "Percentage change in main duty acceptances by age of main applicant, England, April 2012 to March 2013 compared with April 2017 to March 2018" to "Percentage change in presenters by age of main applicant, Northern Ireland, April 2012 to March 2013 compared with April 2017 to March 2018".

Gender

All countries report the gender of the main household member, but the time of reporting and the cohort of applicants differ. In particular, whether countries report the gender for all applicants or single applicants only varies between countries and limits the opportunity for direct comparison. However, in general, more male homeless applicants than female homeless applicants are reported across all countries, before or without priority need assessment (Figure 10). Among single parents only, the majority of applicants are female.

England reports the gender for single-adult applicants and single-parent applicants. Between October and December 2018, among single adults the majority of applicants were male (62%); conversely, among single parents, the large majority of applicants were female (90%).

Wales reports the gender for all homeless applicants and does not differentiate between single-adult or multiple-adult households. Similar to Scotland, between April 2017 and March 2018, a small majority of male applicants was reported for homeless households under a relief duty (56%).

Scotland collects information on gender for all homeless applicants and for single adults and single parents. Across all homeless applicants, between April 2017 and March 2018, Scotland reported a small majority of male applicants (55%). Among single adults, between April 2017 and March 2018, the majority of applicants were male (69%). However, as with England, among single parents , between April 2017 and March 2018, the large majority of applicants were female (80%).

As with England, Northern Ireland reports the gender for single applicants only and shows a majority of single-male applicants before priority need assessment (64%).

Applicant support needs including physical and mental health

The amount and type of information on support need characteristics varies substantially between countries. Information may be reported under priority need vulnerability categories or stand-alone support need categories, depending on legislation.

Priority need vulnerability categories

Information on priority need vulnerability categories is available for England and Wales only. Between April 2017 and March 2018, physical and mental health needs were key vulnerabilities reported in both countries. In England, of the households accepted for a main housing duty owing to vulnerabilities, 29% were for physical health reasons and 38% for mental health reasons. A similar distribution was seen in Wales, with 20% of those accepted for a main housing duty for physical health reasons and 32% accepted for mental health reasons.

An equivalent breakdown of household member vulnerabilities is not available for Northern Ireland. However, households recorded as homeless under the category of “accommodation not reasonable” include physical health and mental health reasons, accounting for 59% and 12%, respectively, between April 2017 and March 2018.

Wider support need categories

England and Scotland report wider information on support needs identified for households presenting as homeless that are separate from legislative priority need categories. Categories have been aligned where possible to allow for comparison.

Physical health and mental health conditions were primary support needs identified for homeless households from April 2017 to March 2018. Physical health accounted for 14% of total support needs in England (October to December 2018) and 20% in Scotland (April 2017 to March 2018). Mental health accounted for 22% of total support needs in England (October to December 2018) and 31% in Scotland (April 2017 to March 2018). Drug and alcohol dependency were also primary support needs identified for homeless households, accounting for 10% in England and 15% in Scotland, though there may be some overlap between these categories, which could overestimate these figures.

Trends in priority needs or support needs over time

Time series data are available for priority needs only in England and Wales (not support needs) and support needs in Scotland. Time series data are not available for Northern Ireland.

Data suggest an increase in the complexity of homeless applicant needs in recent years, particularly in relation to physical and mental health conditions among household members.

The proportion of main housing duty acceptances owing to a priority need vulnerability has increased slightly in England over the last five years by 8%, from 14,090 (April 2012 to March 2013) to 15,250 (April 2017 to March 2018).

Specifically, there has been an increase of 23% in the number of acceptances owing to physical health needs and a 34% increase owing to mental health needs over the past five years. As only one priority need category is reported per application, irrespective of whether households fall under multiple priority need categories, this may be influenced by changes in the choice of priority need category under which households are recorded.

In Wales, the number of households reported as having either a physical or mental health vulnerability as a priority need category has been steadily rising since legislative changes in 2015. From April 2015 to March 2018, physical health as a priority need category has increased by 76% and mental health needs by 111%, which could indicate an increase in the complexity of applicant needs. For comparison, during the same time frame, the number of outcomes reaching priority need assessment stage increased by 38%. However, numbers for priority need categories are small (less than 400) and may be influenced by data quality issues between April 2015 and March 2016 following new legislation. In addition, different household groups, such as those with children or those with vulnerable members, may be rehoused at different rates prior to priority need assessment.

Scotland has seen an overall increase of 22% in applications where at least one support need was recorded, from 11,025 (April 2012 to March 2013) to 13,504 (April 2017 to March 2018). Similar to England, there has been an increase of 30% for applicants reporting physical health needs and 58% for applicants recording mental health needs over the same period.

Conclusions and comparability

A central challenge to comparing the demographics of households seeking help for homelessness is the cohort reported by each country and the existence of priority need assessment bias. Moreover, reporting categories are not the same and must be grouped for comparison, which limits the information available for analysis. Despite these challenges, there are UK-wide trends in demographics for homeless applicants.

Single-person households without children make up the largest category of homeless applicants before or without priority need assessment. With priority need assessment, the majority of households accepted for a housing duty are families with children. However, data with and without priority need assessment cannot be directly compared, particularly in England.

The proportion of young people applying for help with homelessness is similar in all UK countries. In addition, trends in the age of the main household member over time are similar across the UK: England, Scotland and Northern Ireland all show a downward trend in applicants aged 24 to 25 years and under and an upward trend for older adults aged 60 years and over. Reasons for these trends are not currently understood. Comparability could be improved here with all countries reporting the same age categories and reporting the age for all applications, rather than subsets of applications.

In general, the most recent data for all countries indicate that most homeless applicants are male, without or before priority need assessment, except when looking at single parents only, when the majority are female. However, there are differences in the cohort of applicants reported, such as single person, single parent or total applicants, which make direct comparisons a challenge.

Physical and mental health conditions emerge as primary priority need vulnerabilities or support needs among homeless households and, in general, analysis of the data suggests an increase in the complexity of household needs in recent years. Analysing changes in household needs over time is difficult because of legislative changes affecting the information available, reporting categories and the cohort included in data output. For this reason, it is not fully understood whether this is reflective of changing trends in the incidence of medical conditions and how this affects whether people become homeless or way different demographic groups flow through the housing and homelessness system.

Notes for: Demographics

1. Based on different datasets and legislation and cannot be compared; see the Annex for more information.

2. Families in Northern Ireland may not always include children under the age of 18 years and can include parents and children who are over the age of 18 years or adult siblings and other relatives living together.

3. 16 to 17 year olds who are accepted for a main housing duty in England are care leavers.

This section looks at the reasons why people seeking help for homelessness have lost their previous home.

Findings

Reporting categories for the reason for loss of last home are not currently aligned across UK homelessness data, making direct comparisons a challenge. However, key themes are evident including changes to relationships or sharing conditions and loss of rented accommodation.

Changes to relationships and households

Relatives or friends no longer willing or able to offer accommodation, relationships breaking down, and disputes within households are, collectively, primary reasons for loss of home across the UK, though reporting categories vary between countries.

No longer willing or able to offer accommodation may include situations where there has been a dispute within a household but will also include instances where there is no longer enough space in the home for everyone to stay.

Between April 2017 and March 2018, England and Wales reported that around a quarter of homeless applicants cite relatives or friends no longer willing or able to offer accommodation as the reason for loss of home (England, 27%; Wales, 28%¹). During the same period, 25% of applicants in Scotland cite a comparable category, “asked to leave”, as the reason for loss of home.

A separate category for breakdown of a relationship with a partner (non-violent) was cited by applicants in England (6%) and Wales (11%) between April 2017 and March 2018. A similar proportion of applicants in Northern Ireland over the same period cited marital relationship breakdown (10%).

Scotland and Northern Ireland both report a wider category of non-violent family or household disputes, though marital relationships may be included within this category for Scotland. Between April 2017 and March 2018, 18% of applicants in Scotland cited this as the reason for loss of home, while 21% cited this in Northern Ireland.

Violence

Across the UK as a whole, domestic violence is reported in around 1 in 10 cases as a reason for homelessness, though it should be noted that reporting categories and data collection methods are not fully aligned across countries.

Between April 2017 and March 2018, among applicants in England, 12% reported a violent relationship breakdown with a partner as a reason for loss of last home. In Scotland, 13% of applicants cited a violent or abusive household dispute, while 2% cited harassment and 3% cited fleeing non-domestic violence. Similarly, 11% of households under relief duties in Wales cited a violent relationship breakdown with a partner and 2% cited other violence and harassment². Northern Ireland has several categories for violence and harassment for homeless applicants: between April 2017 and March 2018, domestic violence was cited by 5% of homelessness presenters, intimidation by 3% and neighbourhood harassment by 8%.

Rent and tenancy

Loss of rented accommodation, primarily in the private sector, is a key reason for loss of home across the UK but is reported as a reason in England more frequently than in Wales, Scotland and Northern Ireland.

Between April 2017 and March 2018, 27% of households accepted for a main housing duty in England cited loss of private sector tenancy as a reason for homelessness. During the same time period, 11% of applicants in Scotland cited landlord termination of tenancy as a reason for homelessness. Wales and Northern Ireland also report loss of rented accommodation between April 2017 and March 2018, with 16% of households under relief duties in Wales and 15% of households presenting as homeless in Northern Ireland citing this as a reason for loss of home. For more information on private rented sector data, UK private rented sector: 2018 brings together existing data sources to assess comparability and coherence.

Leaving prison, hospitals or care

Leaving institutions including prison, hospitals or care is a less common reason cited for homelessness. England and Wales currently include vulnerable prisoners with a local connection as being in priority need (where applicable).

In England, 2% of applicants were reported to be institution leavers from October to December 2018. Figures reported for Wales between April 2017 and March 2018 were slightly higher, with 13% of outcomes under relief duties recorded as prison leavers (11%) or institution or care leavers (2%), but households may have multiple outcomes. During the same period, 6% of applicants in Scotland were homeless upon discharge from hospital, care, prison or other institutions, while in Northern Ireland 2% of households presenting as homeless were people who had been released from prison or hospital care.

Conclusions and comparability

In general, the reasons for loss of previous home are largely consistent across the UK. However, direct comparability can be challenging owing to differences in reporting categories and definitions. In addition, changes to reporting categories make time series analysis difficult and data users may struggle to identify changing reasons for homelessness in recent years.

Changes to relationships are a central reason for people becoming homeless. Changes could include a marital or familial relationship breakdown involving a dispute or a change to circumstances that means the household can no longer accommodate everyone. Domestic violence is reported in around 1 in 10 cases as a reason for homelessness across the UK as a whole.

Around 1 in 4 homeless households in England report loss of private sector rented accommodation as a reason for homelessness, compared with an average of 1 in 7 across Wales, Scotland and Northern Ireland, though reporting categories for rented accommodation differ slightly across the four countries.  

Notes for: Reasons for homelessness

1. Data for Wales are for homeless households under relief duties:Housing (Wales) Act 2014 - Section 73.

People sleeping rough represent one of the most vulnerable groups in society, and policies to end or reduce rough sleeping are a priority across the UK. However, current methods for data collection and recording of rough sleeping vary between countries and no single method provides a complete picture of the situation.

In general, across UK legislation, rough sleeping is defined as a person bedding down or about to bed down in the open air, an outside shelter or other places not intended for habitation. In all countries, it is the responsibility of each local authority (Housing Executive in Northern Ireland (NIHE)) to provide a measure or estimate of rough sleeping for that area.

In England, data are collected on a single night between 1 October and 30 November. Data are based on either street counts carried out by outreach workers or evidence-based estimates based on spotlight counts and/or information gathered from homeless support agencies, hostels, police, local residents and businesses. Methods record only those people seen or thought to be sleeping rough on a single typical night and do not include everyone in an area with a history of sleeping rough or everyone sleeping rough across October to November.

Wales takes a similar approach to England, collecting information from local agencies over a two-week period in October and conducting a snapshot street count in November. Information on both methods is presented separately to estimate rough sleeping across the country. Northern Ireland also conducts street counts annually in autumn of each year.

Accurately counting or estimating the number of people sleeping rough is inherently difficult given the hidden nature of rough sleeping. There are a range of factors that can impact the number of people seen or thought to be sleeping rough on any given night, including the weather, where people choose to sleep, the date and time chosen, and the availability of alternatives such as night shelters.

Scotland and England both report information on rough sleeping through case-level data for local authority homelessness applications. In Scotland, applicants are asked whether they have slept rough in either the previous three months or the previous night before approaching a local authority for help with homelessness. England reports whether an individual has a rough sleeping support need or is rough sleeping at the time of application.

Findings

In England, since the current methodology was established in 2010, the numbers of people sleeping rough have been steadily rising across London and the rest of England (Figure 11). Over the last five years, the numbers of rough sleepers identified across the whole of England have nearly doubled, from 2,414 to 4,677, though a small decrease was reported between 2017 and 2018. Since April 2018, rough sleeping is also recorded in case-level data for homelessness applications by local authorities. During the most recent quarter, 1,830 people were sleeping rough at the time of approaching a local authority for help, lower than the single night estimate for the same year. However, further work is needed to understand the overlap between the two data collections.

An increase in rough sleeping has also been reported in Wales since the current methodology was established in 2016. One-night snapshot estimates have risen by 12% from 141 (2016) to 158 (2018) and evidence-based (two-week) estimates by 11% from 313 (2016) to 347 (2018).

Northern Ireland has generally reported far lower levels of rough sleeping compared with the rest of the UK and rough sleeping has typically been estimated in Belfast only during street counts. In 2016, 6 people were reported to be sleeping rough on a typical night in Belfast. In 2018, a single-night street count identified 16 people sleeping rough in Belfast and 38 people sleeping rough across Northern Ireland.

Scotland has seen a downward trend in the number of applicants reporting rough sleeping in the previous three months or the night before approaching a local authority since the year April 2010 to March 2011 (Figure 12). In the year April 2012 to March 2013, 3,096 applicants reported sleeping rough at least once in the last three months compared with 2,682 in the year April 2017 to March 2018, a decrease of 13%. Without priority need assessment, case-level application data in Scotland are likely to capture a larger estimate of individuals sleeping rough than in England where single individuals not falling under a priority need category were potentially less likely to approach local authorities for help prior to legislative changes in place from 2018.

All countries report higher levels of rough sleeping in major cities. London accounts for just over a quarter of all estimated rough sleepers in England (27%, 2018). Similarly, Glasgow accounts for 20% and Edinburgh 11% of people reporting rough sleeping at least once in the three months prior to making a homelessness application. Cardiff represented 41% of rough sleepers identified during the 2018 one-night snapshot count and 29% of the evidence-based estimate, followed by Wrexham, which accounted for 15% and 16% of each estimate, respectively.

England is the only country currently to report demographics for rough sleepers. In 2018, 84% of rough sleepers were male and around 94% were aged over 25 years. Characteristics are largely similar for London and the rest of England. However, outside of London, the majority of rough sleepers are UK nationals (76%) while London reports a majority of non-UK, EU nationals (48%).

Conclusions and comparability

Comparability of rough sleeping across countries is limited, and no clear, UK-wide trends are apparent from the currently available data. Rough sleeping has been falling in recent years in Scotland and rising in England, though the most recent figure for England shows a small decrease from 2017 to 2018. However, given the substantial differences in methodology on which these trends are based, they cannot be directly compared. Moreover, methodology has recently changed in Wales, in 2016, limiting the opportunity for comparison. In general, Northern Ireland has reported far lower estimates for people sleeping rough in Belfast compared with other UK cities, though there has been an increase in the last two years. England introduced a case-level data collection in 2018 that includes rough sleeping, and Wales is currently assessing the feasibility of a similar system, though whether this will include rough sleeping is still to be determined. Aligning data collection for people sleeping rough will help with between-country comparisons but differing legislation and policies will still need to be taken into account.

Across the UK, the Ministry of Housing, Communities and Local Government (MHCLG); the devolved administrations; other government departments; and the Office for National Statistics (ONS) are working to improve the range and quality of the statistics, which feed into the overall evidence base.

Analysts across the Government Statistical Service (GSS) and Northern Ireland Statistics and Research Agency (NISRA) are working together to improve the coherence and accessibility of all housing and planning statistics, including homelessness.

There is a wide range of improvements and innovations underway and the challenge is to continue acknowledging the devolved nature of the statistics, and corresponding policy frameworks, while improving coherence where possible to provide a UK picture.

More information about improvements to homelessness and rough sleeping statistics across the UK can be found in this related article.  

This report uses published government statistics on statutory homelessness, temporary accommodation and rough sleeping from the following data providers: Ministry for Housing, Communities and Local Government (MHCLG); Scottish Government; Welsh Government; Northern Ireland Housing Executive (NIHE); and Northern Ireland Department for Communities.

Annual statistics and time series data up to 2018 are included in the main analyses. Where possible, data beyond this time point have been used to confirm trends, but this was not available for all countries at the time of writing.

This report discusses individuals and households that are homeless. Legislation and data collection are also in place to prevent homelessness among individuals and households who are threatened with homelessness but for reasons of scope, this report does not discuss prevention in detail.

Data collection

The Government Statistical Service (GSS) Harmonisation Team have published an interactive tool for homelessness to help users better understand homelessness legislation and data collection. A brief summary of the data collection methods for each country is provided in the following. Further details can be found within the statistical publications for each country.

England

Until April 2018, local authorities returned aggregated data for homelessness via P1E forms on a quarterly basis. Legislative changes brought in from April 2018 onwards introduced a new homelessness case level collection (H-CLIC) system and significantly changed the variables and information collected for homelessness in England.

For this reason, both P1E and H-CLIC data have been used in this report. In general, time series data and snapshot data for the year April 2017 to March 2018 are based on P1E data. H-CLIC data have been used where possible to increase information available for England and enable more cross-country comparisons. P1E and H-CLIC data represent different cohorts of applicants, legislative duties and information collected and cannot be directly compared.

Ministry of Housing, Communities and Local Government:

Statutory Homelessness in England

Rough Sleeping in England

Scotland

Multiple data collection systems are used to collect quarterly data on homeless households. HL1 collects case-level information about applicants and HL2 and HL3 collect information on temporary accommodation. HL2 is an aggregate, snapshot data collection, whereas HL3 is a placement-level data collection that is linked to the HL1 collection. Scotland does not assess for priority need at any stage of application, which should be taken into account when comparing with other countries. In June 2019 Scotland published data for the year April 2018 to March 2019, but for comparability, data for the year April 2017 to March 2018 have been used here for analysis.

Scottish Government:

Homelessness in Scotland

Wales

Data collection in Wales underwent a substantial change from April 2015 onwards and moved to an outcome-orientated approach. Importantly, each application can be recorded under multiple outcomes as they flow through the homelessness system, and outcomes are therefore not representative of unique households. All of the data in this report for Wales beyond the year April 2014 to March 2015 represent outcomes rather than households and should be taken into account when comparing with other countries. In July 2019 Wales published data for the year April 2018 to March 2019, but for comparability, data for the year April 2017 to March 2018 have been used here for analysis.

Welsh Government:

Homelessness in Wales

National rough sleeper count

Northern Ireland

The Northern Ireland Housing Executive (NIHE) and the Northern Ireland Department for Communities collect aggregated data for homelessness applications and households in temporary accommodation and publish a quarterly Housing Statistics report. A more detailed report, the Northern Ireland Homelessness Bulletin, has been published from March 2019, covering data from April to September 2018, and has been primarily used in this report. As reporting time frames are different in Northern Ireland to other countries, this should be taken into account when making comparisons.

Northern Ireland Housing Executive

Northern Ireland Homelessness Bulletin

Northern Ireland Housing Statistics report

Rough sleeper street counts