Combining multiple administrative register sources takes homelessness research to a new level

Introduction: The Secondary Use Act and its implications to homelessness research

Important seeds were planted in 2019 in the field of data protection. These seeds are now about to grow into new fruits for the field of social sciences. We are talking about analyzing joint administrative register-based data. While disciplines like health sciences and economics have already been using large adminis-trative register-based datasets as common practice for quite some time, the social sciences are now start-ing to tap into their full potential, especially with marginalized populations. This is particularly relevant in homelessness research, where the integration of data from different administrative sources can be seen as essential for progress.

In 2019, Finland enacted the Act on the Secondary Use of Social and Health Data (the so-called Secondary Use Act). The primary use of register data refers to the purpose for which the data was collected, for ex-ample, the use of data collected in a social welfare client register in social work client work. Secondary use is any other use, such as the use of register data for statistical purposes, scientific research, teaching, management, or supervision of services.

Before the Secondary Use Act, authorizations for secondary use of social and health data had to be re-quested separately from each controller. The Act harmonized the practices for creating datasets combin-ing different registers and centralized the authorization process. Findata, the Social and Health Data Per-mit Authority, was established to manage the authorization process and the secure processing of the data.

In the field of homelessness research, the Secondary Use Act and Findata makes it possible to merge multiple register sources in much more feasible way than before. When studying hard-to-reach popula-tions (both for research and often also for services), the integration of multiple register sources is particu-larly important. For instance, previous studies (e.g., Kauppinen et al., 2015) have demonstrated that rely-ing solely on traditional national register sources, like the Population Register, leads to inadequate and limited identification of the homeless population. This is primarily due to delays in or lack of updating ad-dress information, resulting in many homeless individuals being inaccurately recorded as housed, and vice versa.

Local social and health services often encounter homeless people face to face, even on a daily or monthly basis. As a result, the client registers of these services provide up-to-date information on homeless peo-ple and their circumstances. However, these registers have rarely been used in scientific studies, and it’s particularly rare to find studies that use a combination of local and broader national data sources. Each register contains different types of information and, when combined, they can provide a more compre-hensive understanding of the overall situation of homeless people and shed light on the pathways into and out of homelessness. Crucially, local registers provide more reliable and timely information about the homeless than large datasets, precisely because they record face-to-face interactions.

Two register-based homelessness studies underway

Two research projects starting in the spring of 2024 will pilot work aimed at unlocking the potential of the Secondary Use Act described above for the homelessness field. An economics researcher from the Y-Säätiö and a social work researcher from the University of Turku have been working together for more than a year before the projects start, going through the complex data permit process and Findata’s long waiting time. Now, in early 2024, the combined datasets will be anonymized and made available, and the actual analysis work will begin

The projects will analyze homeless people’s housing histories and their use of services and social benefits in the years leading up to beginning of their homelessness. The focus will be on how and where homeless people have lived, their household composition, the social and health services they have used, and the social benefits they have received. Of particular interest are the timing of different changes in relation to each other and to the transitions into and out of homelessness periods, as well as the gaps and overlaps between different forms of support.

One research project considers individuals whose homelessness and service needs were addressed in the SAP (Selvitä-Arvioi-Palveluohjaa) group in Espoo and SAS group (Selvitä-Arvioi-Sijoita in the ASTU – Asumisen tukijaos office) group in Helsinki in 2018-2019. The study analyses housing, benefit, and service pathways from 2016 to 2020 and also perform service cost calculations. The goal is to evaluate the eco-nomic impact of Housing First, the main approach used to tackle homelessness in Finland. All subjects (over 1000 individuals) were homeless at the beginning of the process, and the set-up compares those who were placed in Housing First in the process and those who remained homeless after the assessment, waiting for housing to be arranged.

The other research project looks at people who were homeless in Turku in November 2022 (almost 400 individuals) according to the annual national homelessness statistics of the Housing Finance and Devel-opment Centre of Finland (ARA, 2023). The comparison group (1:1) consists of adult social work clients who were not homeless in Turku during the same period. The analysis follows a similar approach to that of Helsinki and Espoo, but covers a longer timeframe, including several decades. It primarily focuses on the comparison setting and does not incorporate cost-effectiveness calculations.

Relevance to homelessness research and work

Finland has gained international fame for its success in reducing homelessness. Finland’s strategies to address this issue have consistently been ambitious, with the latest government program aiming to eradi-cate long-term homelessness by 2027. While the Finnish method for collecting the annual homelessness statistics is considered of high quality on an international scale, it is important to acknowledge that it is not without challenges (Kaakinen, 2023; Pitkänen, 2010). Both in Finland and worldwide, homelessness statis-tics mostly rely on cross-sectional data, which presents certain limitations and complexities.

The information provided by cross-sectional data alone is not sufficient to meet the comprehensive in-formation needs required to end homelessness. Homelessness is a multidimensional phenomenon ex-plained by structural, systemic, interpersonal and individual factors, which often interact (i.e., Fitzpatrick, Bramley and Johnsen 2013). Individual factors – such as addictions or health problems – interact with struc-tural factors, such as the provision of affordable housing and access to social and health services, and sys-temic factors, such as inefficiencies in allocation mechanisms for social housing or social benefits. These factors contribute to broader regional welfare systems.

Gaining a comprehensive understanding of the vulnerabilities of people experiencing homelessness in Finland, and the complex dynamics between housing, services, and benefits, is crucial in the effort to eradicate homelessness. In simple terms, we need to delve deeper into the specific risk factors that con-tribute to homelessness, the use (or lack thereof) of services and benefits prior to homelessness, and how their usage evolves during transitional periods. Our ongoing projects focus on utilizing new forms of joint administrative data at the individual level on a monthly basis, as well as comprehensive multi-year analyses, to effectively address these questions.

As the Secondary Use Act for research purposes is still relatively new in Finland, the data approval pro-cesses and the analytical work related to our projects represent a methodological pioneering work, espe-cially in the field of social sciences. In the preparatory stages of the data permit processes, we have al-ready learned a lot of valuable information about the possibilities and constraints, both technical and ethi-cal, associated with the use of local administrative register data in particular. Similar projects focusing on various social and health issues are being initiated all over Finland. We look forward to working with oth-ers to improve and facilitate these processes.

Register-based quantitative research has been carried out in various fields for a long time. The merging of different health registers is a process that began long ago and has enabled important scientific advances in the field, in addition to improving the ability to serve health care clients. Researchers in other fields, such as economics, have also benefited from integrated data sources. For example, by cross-referencing registered data such as income statements, business expenses, and bank records, tax authorities can identify discrepancies or inconsistencies that may indicate potential tax evasion. The development of data protection legislation is currently expanding the possibilities for register-based research in the social sci-ences, and along the way we can not only achieve important results, but also follow and influence the epistemological implications associated with the new methodology, and their consequences also for the social sector practices and policies. The fruits are beginning to ripen and will soon be ready to be tasted!

Sources of funding

The research project Register Data Analysis for Addressing Homelessness and Reducing it (RADAR) 2024-2026 is a joint project by Y-Sääriö and University of Turku, Department of Social Research, funded by The Housing Finance and Development Centre of Finland (ARA).

The part of the project located in Turku is also co-funded by the Turku Urban Research Programme through Asunnottomuuden polut Turun seudulla – murroksista kohti sosiaalista, terveydellistä, taloudellista ja teknologista kestävyyttä [Homelessness pathways in Turku region – from organizational changes towards social, medical, economic and technological sustainability 2024-2025], a joint project by University of Turku, Department of Social Research, Turku School of Economics and Satakunta School of Applied Sciences.

The RADAR research project can be followed in the future on the project’s website.

References:
ARA (2023) Asunnottomat 2022. Helsinki: Asumisen rahoitus- ja kehittämiskeskus.
Fitzpatrick, Suzanne, Glen Bramley, and Sarah Johnsen. 2013. “Pathways into Multiple Exclusion Home-lessness in Seven UK Cities.” Urban Studies 50(1):148–68.
Kaakinen, Juha (2023) Kotiin. Selvitysraportti tarvittavista toimenpiteistä asunnottomuuden poistamiseksi vuoteen 2027 mennessä. Helsinki: Ympäristöministeriö.
Kauppinen, Timo, Hannikainen-Ingman Katri, Sallila Seppo & Viitanen Veera (2015) Pienituloisten asuin-olot. Työpaperi 22/2015. Helsinki: Terveyden ja hyvinvoinnin laitos.
Pitkänen, Sari (2010) Selvitys pitkäaikaisasunnottomuuden määrittelystä ja tilastoinnista. Asumisen rahoi-tus- ja kehittämiskeskuksen raportteja 2/2010. Helsinki: Asumisen rahoitus- ja kehittämiskeskus.