Are We Really Counting America’s Homeless Families?

a policy commentary
January 2018

Right now, a collection of advocates and local government officials across the United States are preparing to spread out in their counties, communities, and neighborhoods to count the number of homeless Americans. When they return to their offices afterwards, they will combine the number of “sheltered” and “unsheltered” individuals, and send those numbers to Washington. In December 2017—eleven months after the previous “point-in-time” count—the U.S. Department of Housing and Urban Development (HUD) released the figures from last January’s effort, indicating only a slight uptick in the nation’s homeless population from 2016 to 2017. 1

This annual count is incredibly powerful because the data are used to make scores of policy decisions each year including:

  • Which homelessness prevention programs have shown the most promise?
  • Which communities deserve the most federal funding?
  • How should we adjust the nation’s social safety net to ensure hard-pressed Americans have a safe and decent place to live?

At first glance, this seems like a fairly sensible way to go about the messy work of measuring an important social and economic indicator.

But, of course, every method has its drawbacks. How, for example, would the people surveying a local community count a family huddled on that particular night with relatives in an over-crowded single-bedroom apartment?

What if a mother were so fearful that her children would be taken from her (authorities sometimes seize the children of adults unable to provide them with shelter) that she purposefully avoids being counted? 2

If only there were an alternative way to measure homelessness—even if just to see if the basic figures make intuitive sense. The good news is an alternative measure does exist.

A More Comprehensive Picture

Since 1987, schools have been required under federal law to monitor homelessness among public school students. Importantly, the U.S. Department of Education (ED) requires schools to include families that have “doubled-up” among the homeless. That’s critical because “doubling-up”—namely having a family bounce from one over-crowded apartment to another because a family has no place of their own—is an enormous if hidden portion of America’s homeless population. But while ED tracks this data point, HUD’s “point-in-time” survey doesn’t.

This seemingly small distinction in who to count sheds light on a crucial misperception about homelessness in America. For members of the general public, the term “homeless” often conjures up the image of a veteran living under a bridge, or a mentally disturbed individual panhandling near a bus stop. But in an era where 63 percent of all renting families earning below 50 percent of the poverty level spend more than 70 percent of their monthly income on housing costs, “family homelessness” is a pervasive yet often invisible challenge.3

Here’s the bottom line: the picture painted by the Education Department’s measure of homelessness is a far cry from the comparably mild scenario you might draw from the Department of Housing and Urban Development. ED’s homeless numbers reveal a more serious picture requiring a more involved response than the conclusions one might surmise from HUD’s data. And the discrepancy is vast.

  • HUD reported that, at the last count, there were 460 homeless people in families in St. Louis, MO. But St. Louis’s public schools reported to the Department of Education that 5,451 children were identified as homeless in 2016—more than 1000 percent more.
  • San Diego, CA reported 1,714 homeless people in families to HUD last year, but the San Diego City School District alone reported 7,082 homeless students.
  • Tampa, FL counted 533 homeless people in families in their HUD count, but Tampa’s schools found 3,311 homeless students in the district.

Data from schools counted 1,065,794 homeless students during the 2010/2011 school year. Four years later, the same survey found 1,263,323 homeless students—a 19% increase. The U.S Interagency Council on Homelessness (USICH), however, uses HUD’s number when claiming that “we’ve seen steady reductions of people experiencing homelessness” including a “23% reduction in family homelessness” from 2010-2016. 4

Just think what that means. HUD’s 2016 Annual Homeless Assessment Report to Congress counted 64,197 homeless family households. 5 ED counted 1.2 million homeless children. The math is simple: By HUD’s figures, the average homeless household includes 20 children! More crucially, these homeless children and their families are truly invisible to many of the federal agencies that are tasked with working directly on the issue of homelessness. This is not merely an issue of opposing numbers – this is hundreds of thousands of American children struggling with housing instability whose plight remains unacknowledged and unseen.

And this leads to perhaps the most significant concern. Even while acknowledging that “homelessness increased for the first time in seven years” between 2016 and 2017, HUD claimed that family homelessness had declined. Few advocates on the ground would have predicted that sort of statistical improvement for a simple reason—none had seen much improvement in the families they work with every day. Most, in fact, thought things had gotten worse. Taken together, these red flags should be prompting us to ask a more fundamental question: Does HUD’s single-night count even begin to provide an accurate picture of how the challenge of homelessness is evolving?

The Big Glitch

The discrepancy between the figures the Department of Housing and Urban Development collect and those collected by the Department of Education could simply be characterized as a difference in methodology. But, in fact, the distinction reflects a much deeper obstacle. Simply put, some see homelessness as a condition that begins and ends with housing. A banner at one recent national conference of homeless advocates read “Housed people are no longer homeless.” Rapidly rehousing a homeless family does not guarantee that the reasons for their housing, economic, and family instability will be resolved once they have a roof over their head.

Many of those on the front lines see homelessness as a symptom of the more comprehensive problem of poverty. A growing portion of the nation’s population spend more than 50% of their income on housing costs. Two-thirds of renting families below the poverty line receive no housing assistance at all.6 In other words, a family of four that earns less than $13,000 a year is spending more than $9,000 on housing. Can you feed and clothe your children on $4,000 a year? Pay for gas? Sign your child up for soccer? Visit a dentist?

If a poor family becomes homeless—because of rent arrears and eviction, because of domestic violence, because of unmanageable health or mental health issues, or because of the loss of a job—they often end up cycling in and out of shelters or in a string of temporary stays at the house of a relative or friend—until the root cause of their family’s instability is addressed. It’s one thing to want to ensure that no one lives on the streets. It’s entirely another to aspire to get at the root of the problem by providing families with stability and opportunity. Here’s the real issue: homelessness is much bigger than either (1) chronic homelessness or (2) veteran homelessness. Family homelessness is a huge and underappreciated part of American poverty. But, in too many cases, the challenge is obscured by the prevailing statistics.

A New Approach

Perhaps at this point we should be asking—why are we trying to measure family homelessness the same way we measure single adult homelessness? This is not to suggest we abandon HUD’s approach altogether, as it serves as an important measure for at least homeless individuals and for current shelter capacity. But we need different tools and a different approach to accurately count family homelessness in America. ED’s figures provide a valuable starting point.

If there’s anything nearly everyone agrees on, it’s the idea that we should be vigorous in measuring which social safety net programs work—and which ones don’t. The metrics matter, and when one fails to offer insight into what’s happening on the ground, it needs to be supplemented.

Everyone who cares about poverty in America—and certainly those who are focused on homelessness—wants to use the nation’s resources as efficiently as possible. Let’s own what’s happening on the ground, and mold the solutions to the reality. An honest accounting and straight-forward definition of who is homeless is the right place to start.

2 As reported in ICPH May 2017 publication “Taken Away: The Prevalence of Homeless Children in Foster Care,” ICPH’s work in New York City showed that over 850 children—about 5% of all children in foster care—were removed from their families because of homelessness; furthermore, 2/3 of those children were < age of 5.
3 Calculated from AHS data pulled through AHS custom data tool
4 Revising and Strengthening the Federal Strategic Plan to Prevent and End Homelessness Participation Guide, USICH, August 2017.
6 American Housing Survey, 2013. Accessed:—ahs-2013.html. As cited by: