This section explores additional methodologies used to estimate and understand homelessness. It highlights approaches such as the Capture-Recapture method, analysis of homeless applications, surveys, and service-based data. Each method offers unique insights and can complement traditional counts to build a more accurate and nuanced picture of homelessness.
About Capture-Recapture
Capture-recapture, is a statistical method used to estimate the size of populations experiencing homelessness. The method involves conducting two or more separate counts (or “captures”) of the same population over time and comparing the results (OECD, 2025). In the second count, researchers track how many individuals were already recorded in the first. By analyzing the overlap between the counts and applying statistical formulas, they can estimate how many people were likely missed altogether. Typically, the counts utilized are spaced out across a year or longer to provide meaningful comparison points.
Capture-recapture relies on other enumeration efforts to work – it does not replace traditional counts but builds upon them. The method requires little in terms of additional resources beyond time and careful coordination between data collection efforts.
While advances in data collection, digital recording, and linkage methods enhance the potential of this method, they also amplify ethical concerns, particularly the difficulty of integrating multiple data sources without compromising data protection and individual privacy (Bird & King, 2018). Additionally, because the estimate is based on a sample (e.g., people “captured” in two or more counts), there is always a chance of a statistical margin of error.
Also known as Multiple Systems Estimation (MSE) and originally developed in ecology to track animal populations, the method has a long-standing history in scientific research and is now widely used in social policy and public health fields.
Capture-Recapture in The Netherlands
Statistics Netherlands (CBS)
The Statistics Netherlands (CBS) uses a combination of administrative registers and statistical estimation methods to monitor homelessness. The different administrative sources include people registered with a postal address at a shelter, those on welfare benefits without a fixed address, and people identified through probation services. CBS addresses the data gaps by applying the capture-recapture method, overlapping multiple datasets to estimate the total homeless population.
Homeless Applications
Homeless applications are formal requests made by individuals or families to receive housing support or assistance due to homelessness or risk of becoming homeless. These applications help local authorities or service providers assess eligibility for emergency shelter, temporary housing, or long-term support services.
The application typically includes the applicant’s household, such as family members who usually live with them, or others expected to do so. However, depending on the country or city, the institutions may also accept joint applications from more than one person, treating their circumstances together unless separate consideration is needed.
Individuals can make homeless applications with the local councils, either by phone or in person. The council has a legal duty to accept the application and provide temporary accommodation while they assess the situation if people are homeless (lack safe and secure housing) or at risk of becoming homeless within the next two months. The process includes an interview to review housing history, income, health needs, and household composition.
In England, a homeless application must be made to a local authority by someone who is homeless or at risk of homelessness, or by someone acting on their behalf, such as a professional, friend, or relative. The requests can be submitted to any local authority and to any department, not just housing services, regardless of local connection. Councils must ensure applications can be made at any time, including outside office hours.
Homeless Household Surveys
Surveys are a key tool for gathering in-depth, person-level data on households and individuals experiencing homelessness. Often conducted alongside or after street counts and censuses, surveys add context and detail that numbers alone cannot provide. For example, while a street count may identify how many people are experiencing homelessness on a given night, a survey can uncover the reasons behind their homelessness, the duration of their housing instability, and their access to services.
Surveys can also complement administrative data, homeless applications, or service requests, helping to validate or expand findings from those sources. In some cases, survey data is used to inform policy or program design by highlighting unmet needs or barriers to housing.
Resources and International Examples
Bonakdar, A. et al. (2023). “Child protection services and youth experiencing homelessness: Findings of the 2019 national youth homelessness survey in Canada.” Children and Youth Services Review, Vol. 153, https://doi.org/10.1016/j.childyouth.2023.107088
European Union Agency for Fundamental Rights (FRA) (2021). “Roma in 10 European Countries”, https://fra.europa.eu/en/publication/2022/roma-survey-findings
Departamento Administrativo Nacional de Estadística – DANE (2019). Cuestionario para Habitantes de la Calle. Colombia. https://microdatos.dane.gov.co/index.php/catalog/663/download/12969
Gaetz, S. et al. (2016). Without a Home: The National Youth Homelessness Survey. Canadian Observatory on Homelessness Press, https://homelesshub.ca/sites/default/files/WithoutAHome-final.pdf
Instituto Nacional de Estadística – INE. (2022). Encuesta a las personas sin hogar. España. https://www.ine.es/prensa/epsh_2022.pdf
OECD (2025). OECD Monitoring Framework to Measure Homelessness. OECD Publishing, Paris, https://doi.org/10.1787/3e98455b-en.
Service-Based Methods
Service-based methods count or survey homeless individuals at locations like shelters, soup kitchens, drop-in centers, health clinics, housing services, and outdoor sites through outreach teams. Broader sampling increases accuracy, though key locations alone, like shelters and soup kitchens, can capture 90–95% of the urban homeless population over 30 days, often more reliably than national censuses (Peressini et al., 2010).
This data can reveal patterns in service use, identify frequent participants, and highlight needs within the system. Unlike one-time counts, service-based data is often collected continuously, offering a more dynamic and real-time view of homelessness.
The method has limitations to take into consideration. Individuals who avoid services altogether, including youth, undocumented people, or those experiencing stigma or trauma, may be undercounted by this method. Additionally, individuals using multiple services may be counted more than once unless data is cross-referenced or de-duplicated.
Combine Methods
Service-based surveys conducted after Street counts can help identify unsheltered homeless individuals missed during the outdoor enumeration, improving accuracy (Tsai and Alarcón, 2022).
Considerations
Some of the services used by homeless individuals, also serve the general public. Yet, excluding facilities that serve only a small portion of the homeless risks omitting people, potentially reducing the accuracy and reliability of the count (Peressini et al., 2010).
In Practice
The U.S. Census Bureau uses Service-Based Enumeration by coordinating with facility administrators to count people at service sites through in-person interviews or paper records collected on Census Day.