Measuring Impact Task and Finish Group: final report and recommendations

2. How homelessness is measured in Scotland

2.1 Brief history of measuring homelessness

a. Scottish Government Homelessness Data

Data is collected by local authorities when households are homeless or threatened with homelessness (called the HL1 data), when providing temporary accommodation (HL2 and HL3) or carrying out Housing Options work (PREVENT1).

HL1, HL2 (and latterly HL3) returns have provided the main sources of regular statistics on homelessness in Scotland since the early 1990s. Where an application is made under homelessness legislation, local authorities record this through electronic case-based returns to the Scottish Government's Homelessness Statistics unit.

HL1 data has been collected since the reporting year 1992-93, and for the first decade data was supplied on paper forms. Electronic data capture was introduced in 2002-03, incorporating improvements in recording household information and repeat homelessness. 2007-08 was the last major revision to HL1 which accounted for the new legislative framework of the 2001 and 2003 Housing and Homelessness Acts.

Mandatory collection of PREVENT1 data started in April 2014. This is case level data collection to monitor the housing options process and homelessness prevention activity undertaken by local authorities. A PREVENT1 approach may also link to an HL1 application. The data collected by each of these returns is listed here [ref 12].

b. Measuring Rough Sleeping 1997-2004

In 2000, the Scottish Executive commissioned Glasgow Homelessness Network (now Homeless Network Scotland) to roll out a common monitoring system used in Glasgow across 57 services in 23 local authority areas. The monitoring programme was intended to measure the scale and demographics of people using services funded through the second phase of the £63m Rough Sleeping Initiative.

This system had two main limitations; firstly, the reach of the system was limited to the services and areas funded by the initiative and therefore not a complete national picture. Secondly, outcomes data was not recorded reliably which meant people’s pathways, or effectiveness of service interventions, could not be established [ref 13].

However, it was the first real-time, common monitoring system on rough sleeping which provided the most detailed picture to date of the demographics and presenting support needs of people using a diverse range of services across multiple areas. The common monitoring system ended with the initiative in 2005, when rough sleeping services were integrated into local authority homelessness and Supporting People strategies.

c. Scottish Household Survey to 2015

A set of questions in the Scottish Household Survey asked adults about retrospective experiences of homelessness, with a particular focus on the last two years. The survey contained specific questions on duration of homelessness, on rough sleeping and on engagement with local authorities. Although the homelessness data from the survey was used by academics to triangulate with other data sources, homelessness related questions were dropped from the Scottish Household Survey from 2015-16.

d. Measuring Rough Sleeping 2004 to date

Beyond the HL1 returns, there has been no national, cross authority or common monitoring of rough sleeping since the Rough Sleeping Initiative. In some larger local authority areas, HL1 data is supplemented by qualitative research at a local level and/or through periodic ‘street-count’ style measures usually led by third sector organisations.

In 2018, HARSAG highlighted that existing frameworks cannot produce the right information in the right way to facilitate joint working or provide a complete picture of rough sleeping in Scotland. And that street counts, while a practical means to gain a point-in-time count across a specific search area, are of limited use for policy development and an unreliable method for giving a full picture of rough sleeping at local authority or national level [ref 14].

HARSAG also noted limitations in how HL1 data measured rough sleeping, including a lack of consensus on whether it represents an overestimate (should an applicant assume a disclosure of rough sleeping will expedite an application) or an underestimate (that it only captures people who present to the local authority and can’t capture the episodic nature of some people’s experience of rough sleeping).

2.2 How homelessness is currently measured

While there is likely to be additional measures at organisational level, there are key methods that draw data from across multiple or all local authority areas. Not all methods are mandatory, and some are undertaken independently using existing data sources.

Mandatory data collection:

a. Scottish Government - HL1-3

The HL1 return has three parts: application, assessment and closure. Data is resubmitted as an applicant moves through stages. HL2-3 returns are used for the monitoring of households accommodated in temporary accommodation under the legislation and this information is returned by local authorities to Scottish Government quarterly.

b. Scottish Government - PREVENT1

Mandatory collection of PREVENT1 data started in April 2014. It is case level data collection to monitor the housing options process and homelessness prevention activity undertaken by local authorities. A PREVENT1 approach may be standalone or may link to an HL1 record if a homelessness application is then made.

c. Scottish Housing Regulator - ARC

The Annual Return on the Charter (ARC) and Annual Assurance Statement are used by social landlords to evidence to the Scottish Housing Regulator (SHR) that they have met the outcomes set out in the Scottish Social Housing Charter [ref 15]. Detailed statements on homelessness, alongside measures on section 5 referrals^[1], tenancy sustainment, rent arrears, evictions, abandonments and void properties are particularly relevant in this context. The Scottish Housing Regulator also regulates local authority’s homelessness services and will take a view on RSL contributions.

Utilising existing data sources:

d. National Records of Scotland - Homeless Deaths

Experimental statistics on deaths of people experiencing homelessness in Scotland, with a breakdown by demographics and geography. Published annually since 2017 although the data remains experimental until it has been assessed against the rigorous quality standards which apply to National Statistics [ref 16].

e. I-SPHERE at Heriot Watt University - Homelessness Monitor

A longitudinal study into the homelessness impacts of economic and policy developments in Scotland. Commissioned and funded by Crisis with a research team led by I-SPHERE at Heriot-Watt University, University of New South Wales and Newhaven Research. The monitor uses existing data sources to assess trends in statutory and ‘core homelessness’ and integrates a qualitative analysis [ref 17]. Core homelessness is a concept which captures the most acute forms of homelessness which are often not captured in statutory homelessness statistics and includes rough sleeping, people living in sheds, garages and other unconventional buildings, sofa surfing, hostels and unsuitable temporary accommodation such as B&Bs.

f. Scottish Government - RRTP Annual Returns

Since 2019 local authorities have developed and implemented Rapid Rehousing Transition Plans (RRTPs), initially returning annual ‘activity and spend’ information to the homelessness unit at Scottish Government. The first annual report for the period 2020-21 was published in May 2022 [ref 18]. In 2023, the format of the monitoring return was updated to focus on impact and is the main way for local authorities to report on the impact of local action relating to priorities set out in their RRTP.

g. Scotland’s Housing Network - RRTP Monitoring

SHN produce annual reports collating local authorities’ core RRTP data, clustered by 5 regional housing options hubs. While not published, the cross-sector membership of the HPSG RRTP sub-group and the housing options hubs review the reports which use published and unpublished homelessness data from the Scottish Government using HL1 and HL3 returns from local authorities, alongside SHR ARC data.

h. Public Health Scotland - Data Linkage

Since 2016, local authorities have been invited to send a subset of their HL1 data to Public Health Scotland who link it to health data which enables an examination of HL1 applicant interactions with health services. 23 local authorities currently participate in this data linkage, which is at exploratory stage but with a significant amount of potential to measure health and wellbeing across the population of people who experience homelessness.

Opt-in data collection:

i. Scottish Government - Housing First Monitoring Framework

Since 2021, data from local authorities is shared with Scottish Government to measure progress in scaling up the Housing First approach. While not mandatory, all local authorities delivering Housing First provide data on household characteristics, referral routes, support provision and tenancy sustainment. It also records positive outcomes in Housing First tenancies, including training and employment [ref 19].

j. Housing Support Enabling Unit – Outcomes Framework

The Housing Support Enabling Unit manages Better Futures, an outcomes monitoring tool used by 40 housing support providers in Scotland. It measures progress toward 5 goals from the person’s perspective: accommodation, health, safety and security, social and economic wellbeing and employment and meaningful activity [ref 20].