Exploring Administrative Data
Administrative data is collected by government bodies and service organizations as part of their daily operations, such as program participation records, service delivery logs, and users’ demographic information. This data proceeds from records or registries of multiple institutions and organizations and is “used to extrapolate the number of people experiencing homelessness” (OECD, 2025).
Although primarily gathered for internal management and administrative purposes, administrative data can also be a powerful resource for understanding the scope of homelessness, identifying trends, and shaping more effective policy responses. However, the effectiveness of this method depends on the availability of existing datasets and the consistency with which institutions collect and update information relevant to homelessness.
With a longitudinal approach, administrative data can support capture-recapture estimation techniques, which help estimate the true size of the homeless population by comparing overlaps across datasets.
Administrative data is information gathered by service providers or government agencies in the course of their routine activities.
Introduction to Administrative Data Sources
Administrative data refers to information collected by service providers and government agencies as part of their routine operations, including shelters, healthcare, justice, and social services. This also includes records of program participation, services delivered, and client demographics.
Some of the first steps to developing and using administrative data on homelessness are to:
- Identify key data sources among service providers, public systems, and NGOs.
- Standardize data considering variables like date and type of service use, housing status, demographics, duration and history of homelessness, and exit destination (if applicable). This would include harmonizing data sets, adding any additional indicators, and creating consistent identifiers across data.
- Develop a data collection framework with common definitions and categories, standardized forms, and data entry procedures.
- Create data sharing agreements across govenment agencies and organizations and ensure compliance with privacy laws
The data becomes especially valuable when organizations work together to create a shared database that brings together information from multiple sources. By pooling data, service systems can form a completer and more accurate picture of homelessness in a given area. This collaboration helps reduce duplication of services, improve individual outcomes, and strengthen system-wide planning and accountability.
With data aggregation, it is crucial to use the data for planning and decision-making. As Thomas and Mackie (2020) state, “the operational integration of data we wish to highlight is combining of data from multiple sources to expand institutional knowledge beyond its own boundaries in support of operational decisions” (p. 78). Accordingly, stakeholders should examine trends (Who experiences long-term homelessness? What are common elements?), map service usage and gaps, and identify needs (for example, how many shelter beds, housing units, or case managers are required?).
The subsequent steps involve:
- Clean data to ensure accuracy, consistency, and adherence to data standards.
- Report and communicate findings to create transparency, accountability, and advocacy momentum. This step can include publishing regular reports for policymakers, funders, and the public, using dashboards or interactive tools to visualize data, and disaggregating results by population group to promote equity.
- Monitor outcomes and improve interventions by tracking outcomes for individuals (housing stability, returns to homelessness), evaluating program effectiveness, and comparing over time progress toward goals.
- Establish regular meetings with oversight committees to review and update data sharing agreements and quality control measures. Incorporate citizen and non-government stakeholder participation, including people with the lived experience of homelessness at each point of review.
Using administrative data in this way offers several benefits. It allows organizations to identify patterns over time, track progress, and inform resource allocation more effectively than relying on periodic surveys or point-in-time counts alone. Shared systems also support better coordination among agencies, which can lead to more responsive and person-centered services.
Additionally, the OECD (2025) highlights that, due to the nature of administrative data, this method “may enable researchers to identify people at risk of becoming homeless; data can be used to improve outreach and direct service provision to prevent homelessness before it occurs” (p. 33).
An important challenge is, as Thomas and Tweed (2021) indicate, “those who are not engaged with services – or on a sporadic basis – may be missed,” since people must have some contact with the services to be “visible” in Administrative data (p. 182). Errors in records or inaccuracies during data entry into databases can lead to inaccurate results, ultimately affecting the reliability of the analysis.
Additionally, determining a data governance structure, data sharing agreements, Memorandums of Understanding, client consent processes and forms, and establishing a database can be time consuming.
Leveraging administrative data at scale requires careful and deliberate planning since there are ethical and legal elements to take into consideration. In the process, it is essential to determine appropriate data protection measures, as the privacy and security requirements can vary depending on how the data will be used and the local or national laws governing the sharing and processing of personal information. According to Thomas and Mackie (2020), some of the approaches to ensure the security of data are:
- “Processing of data, i.e. aggregation,
- anonymize or de-identify data, or
- use personal data with added measures to ensure that disclosure risks are minimized, e.g. a “split file” method of sharing data” (p. 85).
Finally, administrative data alone often fails to capture the complexity and full scope of homelessness. In a report published in 2023, the government of Brazil acknowledged that its analysis based on administrative data—which draws from various official sources—provides only a “partial diagnosis of the homeless population”. They emphasized the need to complement this information with data from the national census and other statistical instruments to obtain a more comprehensive understanding.
Participation of Impacted Communities and Partners
- Establish and implement informed consent processes that respect the rights of people whose data is being collected.
- Involve citizen and non-government stakeholder participation, including people with lived experience of homelessness at each step to ensure that principles of inclusive, community-owned data are incorporated.
Standards and Safeguards for Administrative Data Systems
- To protect data privacy and security by carefully treating people’s personal information.
- To ensure consistent data definitions; agencies must agree on shared terminology and standards.
- To create the technical capacity to manage, integrate, and analyze large datasets securely and reliably.
Collaborative Data Systems for Greater Impact
- Administrative data becomes most effective when organizations collaborate to build a shared database that integrates information from multiple sources.
- Data use depends on clear governance structures. These should define who can access, share, and use the data—and under what conditions.
Case Studies: Centralizing Data on Homelessness
População em situação de rua: diagnóstico com base nos dados e informações disponíveis em registro administrativo e sistemas do Governo Federal
In March 2023, the Brazilian Ministry of Human Rights and Citizenship (MDHC) released the report “Homeless Population: Diagnosis Based on Data and Information from Administrative Records and Federal Government Systems,” which draws on available data from national registries and systems.
Given that it used data from multiple sources and organizations, the report examined the social assistance network, access to employment, education, income, and legal documentation for people experiencing homelessness. Additionally, the document highlighted the significance of expanding health services and the need for more structural solutions, identifying the Housing First program as a priority strategy recommended by the MDHC.
Housing Finance and Development Centre (ARA)
In Finland, data on homelessness is compiled annually by the Housing Finance and Development Centre (ARA) through a nationwide municipal survey. The approach combines administrative data and localized reporting to produce a detailed snapshot of homelessness. In Finland, a person is considered homeless if they do not have their own rented or owned home and are living on the streets, in temporary shelters, dormitories, hostels, welfare-type housing, or other institutions, or temporarily housed with friends or family (OECD Country Note, 2024).
Municipalities must report the number of homeless individuals based on local registers and services, including but not limited to social welfare and housing service registers, municipal rental housing providers, the Social Insurance Institution (Kela) register, and the Digital and Population Data Services Agency. The results are disaggregated by sociodemographic characteristics, locations, and household composition.
Censimento Permanent della Popoliazone e delle Abitazioni
Since 2018, Italy’s Permanent Census of Population and Housing, managed by the National Institute of Statistics (ISTAT), integrates administrative records with sample surveys to provide annual demographic data, including on people experiencing homelessness. The ISTAT adopted a definition of homelessness based on Regolamento Anagrafico, which refers to those “who are registered with the civil registry at a national residential address and at the address of NGOs/associations that provide assistance to people experiencing homelessness, and those who elect their domicile at the municipality where they usually reside” (OECD Country Note, 2024). Municipalities assign “fictitious addresses” or use NGO addresses for the administrative registration of people without fixed residences.
The use of Administrative Data in the country reduces costs and offers localized insights, but it undercounts certain populations, like undocumented individuals or people who are recently experiencing homelessness and have not processed their applications.
CHAIN: Combined Homelessness and Information Network
In London, data on rough sleeping is collected through CHAIN (Combined Homelessness and Information Network), a multi-agency system funded by the Mayor of London and managed by Homeless Link. It is the most detailed administrative database on street homelessness in the UK since it gathers information from outreach services (who are responsible for creating new records), assessment and reconnection programs (e.g., No Second Night Out), homeless hostels and supported accommodation services, and specialist providers.
CHAIN was designed to support coordinated service delivery by allowing different organizations to share real-time information about individuals experiencing homelessness.
The data collected is published in visualization tools that are updated quarterly and annually. While individual-level data is restricted, aggregate data and analyses on trends in London are regularly published by Homeless Link and made publicly available online.
Homeless Management Information System (HMIS)
The United States gathers administrative data on homelessness through the HMIS, a standardized local data system used by communities to collect and manage information on individuals and families experiencing homelessness. This system is used by service providers (shelters, outreach teams, and housing programs), so data quality and completeness can vary by provider and geography.
The HMIS was mandated by the HEARTH Act (2009), which requires all offices and institutions receiving federal homelessness funds to operate an HMIS that captures unduplicated counts of people experiencing homelessness. Participation in HMIS is mandatory for the Continuum of Care (CoC) Program and Emergency Solutions Grant (ESG) recipients.
The HMIS Data Standards Manual was updated in June 2025. You can also visualize the guidelines provided for different stakeholders (HMIS Lead, worker doing a data entry, vendor, etc.) and the system standards in the HUD website.
Additional Resources
This collection of resources contains examples, research papers, and toolkits to provide more information about Administrative Data:
OECD (2025), OECD Monitoring Framework to Measure Homelessness, OECD Publishing, Paris, https://doi.org/10.1787/3e98455b-en.
Thomas, I. & Mackie, P. (2020). The Principles of an Ideal Homelessness Administrative Data System: Lessons from Global Practice. European Journal of Homelessness. Volume 14(3). pp. 63-85. https://www.feantsaresearch.org/public/user/Observatory/2021/EJH_14-3/EJH_14-3_A3_web2.pdf
Thomas, I. & Tweed, E. (2021). The Promises and Pitfalls of Administrative Data Linkage for Tackling Homelessness. European Journal of Homelessness. Volume 15(3). pp. 175-186. https://www.feantsaresearch.org/public/user/Observatory/2021/EJH_15-3/Final/EJH_15-3_A12.pdf
Culhane, D. (2016), “The Potential of Linked Administrative Data for Advancing Homelessness Research.” European Journal of Homelessness, Vol. 10 (3). https://works.bepress.com/dennis_culhane/209/
Yue, D., Pourat, N., Chen, X., O’Masta, B., Huynh, M., & Xin, K. (2020). How to Identify Homelessness Using Administrative Data. Health Services Research, 55 (Suppl 1), 139–140. https://pmc.ncbi.nlm.nih.gov/articles/PMC7440552/