The annual point-in-time counts of New Jersey’s homeless population were conducted on January 24 this year. When the estimated number of homeless people is released by the U.S. Department of Housing and Urban Development (HUD) later this year, we would do well to relate to them with skepticism. The full extent of homelessness, in fact, is not revealed in these estimates. They substantially underestimate the magnitude of homelessness in New Jersey.
HUD’s Annual Homeless Assessment Report to Congress provides a detailed estimate of the number of homeless people in the United States on a single day each year. This is the most readily accessible and consistent data on homelessness. The 2017 counts estimated 8,536 homeless people in New Jersey, 7,122 of whom were sheltered (staying in emergency shelters, transitional housing programs, or safe havens) and 1,414 who were unsheltered (people whose primary nighttime location was a public or private space that is not designated for, or ordinarily used as, a regular sleeping accommodation, such as the streets, vehicles, or parks).
According to the point-in-time counts, the number of homeless people in New Jersey has been decreasing steadily. In 2017, the homeless population in New Jersey declined by 4.6 percent, the sixth successive annual decrease. Since 2011 the homeless population has decreased by 5,601 people, a 40 percent reduction. Moreover, in 2017 New Jersey’s homeless population was slightly less than half what it was in 2007, when 17,314 people were estimated to be homeless. Even more notable, the percentage decline in homelessness in New Jersey between 2007 and 2017 is larger than in any other state in the nation. (Because of changes in methodology Michigan’s numbers are not included in the comparison, although their percentage reduction may be slightly larger.)
Those declining trends would be impressive if they were in fact good estimates of the extent of homelessness in New Jersey. Unfortunately, they are not. Homelessness is far more pervasive than the HUD point-in-time estimates suggest. There is no broadly accepted definition of who counts as homeless. Variation exists among governmental departments in how they count the homeless. HUD’s estimate of 553,742 homeless people nationally in 2017 did not include people and families who were “doubled up” or “couch surfing” (living in someone else’s home). In addition, it excluded homeless people who may be in hospital or in jail on the day of the count.
The U.S. Department of Education’s count of homeless children, on the other hand, is almost 2.5 times larger than HUD’s point-in-time estimates for the nation’s entire homeless population. The Department of Education includes children who are homeless at any time during the year, including those who are living doubled up, staying in hotels/motels, abandoned in hospitals, or awaiting foster care placement. Their counts of homeless children and youth enrolled in public schools have gone up every year since data was first collected in 2003.
But even the Department of Education’s counts may be low. A recent national survey of homeless youth undertaken by Chapin Hall at the University of Chicago suggests that the number of homeless is probably higher. They estimated that approximately one in 10 young adults ages 18 to 25 — a shameful 3.5 million young adults — and at least one in 30 adolescents ages 13 to 17 — about 700,000 — undergo some form of homelessness over a 12-month period. These counts include out-of-school youth, not covered by the Department of Education’s counts.
Second, annual data does a better job capturing the flow of individuals moving in and out of homelessness during the course of the year than the point-in-time counts, which are conducted on a single day. A homeless person may alternate between residing in a homeless shelter, living on the street, or doubling-up over the course of a year. After an eviction, for example, a homeless family may choose to double up before entering a shelter. In addition, the counts of people living in shelters may be a consequence of the availability of beds, indicating the number of people served, rather than the number of people who actually sought shelter.
A 2001 study using data collected from administrative records of homeless services providers found that the annual homeless population in nine U.S. jurisdictions ranged from 2.5 times greater than the point-in-time estimates in Spokane, WA, to 10.2 times greater in Rhode Island. According to a Homeless Assessment Report conducted by Monarch Housing Associates, New Jersey’s emergency shelter, transitional housing and safe-haven projects served a total of 24,519 persons during calendar year 2015. That number is 2.7 times larger than the point-in-time count for sheltered homeless people that year.
If HUD’s 2018 report shows a continuation in the declining homeless trend, we could be led to believe that New Jersey continues to do a good job in reducing its homeless population. We would be wiser, though, to heed the findings of research based on national data, regardless of whether homelessness increases or decreases, and to treat the point-in-time numbers with extreme caution. They substantially undercount the actual number of homeless people in New Jersey.