How Homelessness Is Counted and Measured

Why Counting Matters

Before communities can effectively address homelessness, they need to understand its scope. How many people are experiencing homelessness? Where are they? What are their needs? The answers to these questions shape policy decisions, determine funding allocations, and guide the design of programs and services. Without reliable data, resources may be misdirected, progress cannot be measured, and the experiences of people without housing remain invisible to decision-makers.

Yet counting people experiencing homelessness is one of the most challenging tasks in social measurement. Unlike a census of housed residents, there is no fixed address to visit, no mailbox to deliver a survey to, and no single database that captures everyone. The methods that have been developed represent decades of effort to make the invisible visible — but each comes with significant limitations that are important to understand.

The Point-in-Time (PIT) Count

The most widely recognized method for counting people experiencing homelessness in the United States is the Point-in-Time (PIT) Count, mandated by the U.S. Department of Housing and Urban Development (HUD). Every year, on a single night in late January, communities across the country mobilize thousands of volunteers and outreach workers to count every person they can find who is experiencing homelessness.

The PIT Count has two components. The sheltered count tallies individuals and families staying in emergency shelters, transitional housing programs, and Safe Havens on the designated night. This data is typically gathered from shelter records and Homeless Management Information Systems (HMIS). The unsheltered count, required every other year (though many communities conduct it annually), involves teams of volunteers and outreach workers canvassing streets, parks, encampments, vehicles, abandoned buildings, and other locations where people who are unsheltered may be staying.

The PIT Count is organized at the local level by Continuums of Care (CoCs) — regional planning bodies designated by HUD to coordinate housing and services for people experiencing homelessness. There are approximately 400 CoCs across the country, each responsible for conducting the count within its geographic area and reporting results to HUD.

The data collected during the PIT Count includes not only the total number of people experiencing homelessness but also demographic information such as age, gender, race, ethnicity, veteran status, and whether individuals are part of a family with children or are experiencing chronic homelessness.

The Housing Inventory Count (HIC)

Conducted alongside the PIT Count, the Housing Inventory Count (HIC) provides a companion snapshot of the resources available to people experiencing homelessness. The HIC catalogs every bed and unit dedicated to serving this population, including emergency shelters, transitional housing, rapid re-housing programs, Safe Havens, and permanent supportive housing.

Together, the PIT Count and HIC allow communities and policymakers to compare the number of people experiencing homelessness against the capacity of the local housing and shelter system. This comparison reveals gaps in capacity and helps inform decisions about where to invest in new resources. For example, if a community's PIT Count shows a growing unsheltered population while its HIC shows stagnant shelter capacity, that data can support the case for expanding emergency shelter or investing in permanent housing solutions.

The Annual Homeless Assessment Report (AHAR)

While the PIT Count captures a single-night snapshot, the Annual Homeless Assessment Report (AHAR) provides a broader view. Submitted by HUD to the U.S. Congress each year, the AHAR draws on data from Homeless Management Information Systems (HMIS) across the country to estimate the total number of people who use shelters and transitional housing programs over the course of an entire year.

This longitudinal perspective is critical because homelessness is often a temporary or episodic experience. Many people cycle in and out of homelessness over the course of a year, meaning the number of people who experience homelessness at some point during a twelve-month period is significantly larger than the number counted on any single night. The AHAR helps capture this broader picture, providing data on patterns of shelter use, lengths of stay, and the characteristics of people entering and exiting the homeless services system.

Homeless Management Information Systems (HMIS)

Homeless Management Information Systems (HMIS) are local databases used by CoCs to record and track the services provided to people experiencing homelessness. When a person enters a shelter, receives case management, or accesses other homeless services, their information is entered into HMIS. This creates a longitudinal record of service utilization that can be analyzed at the local, state, and national levels.

HMIS data powers much of the analysis in the AHAR and is increasingly used for coordinated entry systems, which help communities prioritize housing resources for those with the greatest needs. The data collected typically includes demographic information, prior living situation, disability status, income, and service history.

However, HMIS only captures data on people who interact with the formal homeless services system. People who are experiencing homelessness but do not access shelters or services — including many people who are unsheltered, doubled up, or couch surfing — may never appear in HMIS data. Additionally, some programs, such as domestic violence shelters, maintain separate databases for safety and confidentiality reasons, meaning their clients may not be reflected in HMIS.

Limitations of Current Methods

While the PIT Count, HIC, AHAR, and HMIS represent significant achievements in data collection, they have well-documented limitations that affect the accuracy and completeness of our understanding of homelessness.

Undercounting hidden homelessness. People who are doubled up with friends or family, staying in motels, or living in other precarious situations are generally not included in the PIT Count or HMIS data unless they access homeless-specific services. This means millions of people in unstable housing situations are invisible in the most commonly cited statistics.

Rural and suburban gaps. The PIT Count methodology was designed primarily for urban environments where people experiencing homelessness are more concentrated and visible. In rural areas, people may be spread across vast geographic areas, living in cars, abandoned structures, or wooded areas that are difficult for volunteers to reach. As a result, rural homelessness is particularly undercounted.

People avoiding services. Some individuals actively avoid shelters and outreach workers due to past negative experiences, distrust of institutions, mental health conditions, or fear of law enforcement. These individuals are among the hardest to count and often among the most vulnerable.

Weather and timing effects. Conducting the count in January means that weather conditions can dramatically affect results. A particularly cold night may drive more people into shelters (increasing the sheltered count but potentially decreasing the unsheltered count), while milder weather may have the opposite effect. Rain, snow, or extreme cold can also limit the ability of volunteers to canvass effectively.

Definitional exclusions. HUD's definition of homelessness for the PIT Count does not include all people who lack stable housing. People who are doubled up, those in institutions (such as jails or hospitals) who have no home to return to, and people in other precarious situations may not be counted.

Volunteer methodology inconsistencies. The quality and thoroughness of the unsheltered count varies significantly across communities. Some CoCs deploy large, well-trained teams with extensive knowledge of local encampments, while others rely on smaller groups with less experience. This inconsistency makes it difficult to compare results across communities or track trends over time with confidence.

Alternative and Complementary Approaches

Recognizing the limitations of existing methods, researchers and practitioners have developed alternative approaches that complement the PIT Count and provide a more complete picture of homelessness.

School-based identification (McKinney-Vento). Under the McKinney-Vento Homeless Assistance Act, public schools are required to identify and support students experiencing homelessness, including those who are doubled up, living in motels, or in shelters. School districts use a broader definition of homelessness than HUD, and their data consistently shows far more children and families experiencing housing instability than the PIT Count captures. In the 2022–2023 school year, schools identified over 1.1 million students experiencing homelessness — a number that dwarfs the PIT Count's family figures.

HMIS longitudinal data. Rather than relying solely on single-night snapshots, communities can analyze HMIS data over time to understand patterns of shelter use, identify people who cycle in and out of homelessness, and track outcomes for individuals who receive housing interventions.

Administrative data matching. Some communities and researchers are linking data across systems — including homeless services, healthcare, criminal justice, child welfare, and public benefits — to identify people experiencing homelessness who may not appear in any single dataset. This approach can reveal the full scope of system involvement and help target interventions more effectively.

Community surveys and participatory research. Engaging people with lived experience of homelessness in the research process can surface information that traditional counting methods miss. Peer-led surveys, focus groups, and community-based participatory research can provide qualitative context that enriches quantitative data and ensures that the voices of those most affected inform how homelessness is understood and measured.

Why Better Data Matters

The way homelessness is counted directly affects the resources available to address it. Federal funding for homeless services is allocated in part based on PIT Count data, meaning communities that undercount their populations may receive less funding than they need. Conversely, communities that invest in thorough counting methodologies may be better positioned to secure resources.

Better data also enables better program design. Understanding who is experiencing homelessness, for how long, and what services they need allows communities to tailor their responses. For example, data showing a high rate of chronic homelessness might support investment in permanent supportive housing, while data showing many people experiencing brief, crisis-driven episodes might point toward rapid re-housing and prevention programs.

Accurate measurement is also essential for accountability. Communities, funders, and the public need reliable data to assess whether strategies are working, whether progress is being made, and where adjustments are needed. Without trustworthy baselines and consistent measurement, it is impossible to know whether investments are producing results.

Conclusion

Counting people experiencing homelessness is both a technical challenge and a moral imperative. The methods currently in use — the PIT Count, HIC, AHAR, and HMIS — represent important tools for understanding the scope and nature of homelessness, but each has significant limitations. The PIT Count provides a valuable annual snapshot, but it is just that: a snapshot of one night that cannot capture the full reality of a dynamic, complex social issue.

Understanding both what the data tells us and what it misses is essential for anyone seeking to engage meaningfully with the issue of homelessness. When we see a PIT Count number, we should recognize it as a floor, not a ceiling — a minimum estimate that almost certainly understates the true number of people affected. By supporting better data collection methods, investing in complementary approaches, and listening to the experiences of people who have lived through homelessness, we can build a more complete and honest picture of the challenge — and a more effective response.