Homelessness

There is a distinction between literal homelessness (sleeping outside of conventional shelter) and housing precarity (lack of stable shelter). This matters for how we measure homelessness. Perhaps the most common measure of homelessness is what’s referred to as a Point-In-Time (PIT) count – a count of how many individuals are unhoused on any given night. The Department of Housing and Urban Development (HUD) estimates these counts through Continuums of Care (CoCs) by literally counting the number of those sleeping in homeless shelters and outside of shelters on a single night in January. While useful, these counts are snapshots in time. They don’t account for how people move in and out of homelessness over time. For this reason, counts of people who have ever experienced homelessness over the past year are much larger than PIT counts and probably larger than most people would guess. Most stints in homelessness are brief which means we usually underestimate the true prevalence of homelessness. About ¼ of individuals and 5% of families experience homelessness for more than a year.¹ 

Federal homelessness assistance is largely routed through organizations called Continuums of Care (CoCs). These are constellations of government and nonprofit organizations (typically county-wide) tasked with operating coordinated homelessness systems. Key elements of these systems include:

• Emergency shelters (hopefully with a coordinated entry system)
• Transitional housing: temporary housing between shelters and permanent housing, often with eligibility requirements
• Rapid rehousing: short-term subsidies that expire quickly and are designed to get people back on their feet and into housing right away 
• Permanent supportive housing (Housing First): getting people into long-term stable housing with wrap-around services if needed

Functional zero is a term that is sometimes used. It refers to the point at which the homeless service system can prevent homelessness wherever possible and place people into permanent housing.² 

Many people think of homeless individuals as older and unemployed men. In reality, the homeless population is much more diverse than this. A uniquely and sobering American problem is that families with children make up about 46% of people experiencing homelessness on any given night (34% of the population experiencing homelessness throughout the year). The rest are individuals without children who make up the vast majority of the unsheltered. 

Those who transition in and out of homelessness are more likely to have medical and behavioral health problems, but these individuals also comprise a much smaller portion of the homeless population than many realize. Chronic homelessness is defined as (1) experiencing homelessness for over 1 year (continuously or at least 4 episodes over the past 3 years) or (2) having a diagnosable substance use disorder, serious mental illness, developmental disability, PTSD, cognitive impairment, or a chronic disease/disability. The chronically homeless account for about 1/4 of the PIT counts. Homeless individuals have much higher rates of disability (49% were reported to have a disability in 2019), including those due to mental illness or substance use disorder. Yet this also means that roughly half of individuals and 3/4 of families experiencing sheltered homelessness over the course of a year do not. 

Among adults without children experiencing homelessness, the best estimates of prevalence of certain conditions are as follows:

• 12.7% have a mental illness
• 11.4% have major depression 
• 23.1% have a personality disorder 
• 37.9% have alcohol dependence 
• 24.4% have drug dependence 

PIT counts, because they are more likely to incorporate unsheltered homelessness, likely overstate the true prevalence of these conditions among the full population of those experiencing homelessness. And it’s obviously worth noting that people’s conditions can improve when they exit homelessness and enter stable housing. 

Veterans are another notable subgroup of people experiencing homelessness. They tend to be older, male, and living on their own. Overall, men are more likely to experience homelessness, but women comprise about 26% of the homeless population.

Young people comprise a large and often invisible share of those experiencing homelessness. Infants under the age of 1 are the age group with the highest risk of homelessness in the United States. Similarly, 2.9 million children face eviction every year and those under the age of 20 are the age group at highest risk of facing an eviction. Black renters with children face an eviction rate of over 25% which is more than double for any other group. 290,000 children use shelters at least once throughout the year. Homelessness also reveals extreme disparities in our society. Black and Native American individuals are especially at risk for homelessness. Black families make up about 52% of families at shelters and about 40% of overall homeless, despite making up roughly 11% of the population. ³

A record number of people – over 653,000 people – were found to be experiencing homelessness on a given night in January 2023. That represents a 12% increase since 2022.⁴ But this only captures homelessness experienced on a given night – which underestimates the true prevalence of homelessness experienced throughout the course of a year. That number is much higher – somewhere around 1.6 million people experience homelessness over the course of a year. These are painful reversals from progress that was underway before the COVID-19 pandemic. Homelessness had dropped a bit over the 2010s, especially chronic and veteran homelessness, which declined by 26% and 50%, respectively.⁵ 

In New Jersey, 10,267 people were experiencing homelessness on January 24, 2023 – a 17% increase from the previous year. In Mercer County, that number was 714 – a 34% increase from the prior year. In both New Jersey and Mercer County, homelessness had generally been declining throughout most of the 2010s. See data from Monarch Housing Associates for more information.

Widespread homelessness did not really exist before the 1980s. Beginning in the 1970s, the federal government retreated from public housing provision and instead began assisting a fraction of the population in need through programs that would become the Housing Choice Voucher (HCV) program. At the same time, communities all across the country began underbuilding housing, particularly due to exclusionary zoning and land use policies that made it impractical or illegal to build more affordable types of housing.

Homelessness is highest where housing costs are highest. Consequently, policies that promote housing affordability prevent homelessness.⁶ We have good evidence that HCVs essentially end homelessness in most cases, even among people with serious  mental illness (when paired with supportive services). Yet, we provide far less rental assistance than is needed. We also provide less rental assistance relative to other developed countries.⁷ 

Those experiencing homelessness face much higher mortality. Drug overdoses, suicides, homicide, and infectious disease are all more prevalent among those experiencing homelessness. Adults experiencing homelessness also die from chronic conditions about 10-15 years earlier than among those in the general population. Elevated mortality is especially acute among the unsheltered population, who face ten times the mortality risk of the general population. Those experiencing homelessness also have worse health in general, including higher rates of disability, mental illness, substance use disorder, and loneliness.⁸ 

The causes of homelessness are not a mystery. They are structural in nature and include factors like poverty, inequality, and especially a lack of affordable housing. These factors explain how many people experience homelessness whereas individual vulnerabilities explain who is at most risk for experiencing homelessness. There is a very strong connection between homelessness and the cost of housing. Homelessness is consistently highest in places with the highest housing costs. Places like Seattle and San Francisco have much higher rates of homelessness than in places like Detroit or Milwaukee, despite having much stronger economies. Homelessness is ultimately a policy choice – individual risk factors are only risks at all because of our social structures that permit homelessness in the first place.⁹ 

A lack of affordable housing drives this crisis. We have entitlement programs like SNAP and Medicaid to ensure access to food and medical care, but no similar entitlement to housing, despite its obvious connection to well-being. Only ¼ of those who qualify for the Housing Choice Voucher (HCV) program, the nation’s largest housing assistance program, actually receive any assistance and a significant portion of these households – in some places, over half of them – are unable to use it because they are unable to find a landlord who will accept it (landlords can legally refuse HCV as a form of payment in most of the country and many others ignore regulations meant to address this).¹⁰ We have a profound shortage of affordable housing – 7.3 million according to the National Low Income Housing Coalition, which also estimates that only 33 homes exist for every 100 extremely low income renting households.¹¹ 

For how obvious the causes of homelessness are, it is perhaps surprising that so many misguided explanations exist. Homelessness is not driven by serious mental illness, drug use, or unemployment. Similarly, nice weather does not encourage homelessness. States and localities that have more generous welfare policies do not create homelessness. Those experiencing homelessness are not migrating to places to be without shelter. Housing prices and vacancy rates really seem to explain it.¹²

The individual-level predictors of homelessness are well-known and include: 

• A previous bout of homelessness
• A previous eviction record
• Doubling-up 
• Public assistance receipt
• Pregnancy or household with young children 
• Child trauma or history with CPS 
• Family disorder 

Behavioral factors like mental illness, substance use disorder, criminal history, experiences of domestic violence, are all major risk factors for homelessness. 

We have solid evidence that the following prevent homelessness: 

• Permanent, deep rental subsidies 
• Rental assistance for those facing eviction 
• Community-based services like HomeBase in NYC 

We also have decent evidence that the following prevent homelessness: 

• Universal screening of veterans
• Permanent, shallow subsidies for those at risk 
• Eviction diversion programs

The following may prevent homelessness, but not enough evidence exists to say so conclusively:

• Case management 
• Flexible cash grants for low-income households 
• Bringing services to high-risk neighborhoods
• Integrating homelessness and anti-poverty services 
• Targeted screenings
• Progressive engagement¹³

Despite the current situation, we have made some progress in reducing homelessness in the past. We halved veteran homelessness from 2009-2019 – a decrease from 73,000 in 2010 to 38,000 in 2018) in large part due to a series of coordinated and concerted efforts like X. This decrease shows what political will, resources, and coordination can do. 

Some places have also demonstrated success in reducing overall homelessness. Through coordinated systems, hotlines, case management, and Medicaid funding, Hennepin County in Minnesota had helped drive steady reductions in chronic homelessness. Similarly, Houston had moved 25,000 into stable shelter over the 2010s – a drop of 63% in homelessness – thanks to a coordinated response across a range of social service providers and streamlined processes. Some places, like Bergen County in New Jersey had virtually ended chronic homelessness by reaching functional zero. 

Of course, these successes were all prior to the COVID-19 pandemic. Homelessness has since been back on the rise, which means we will need more political will to prevent further backsliding. 

We have clear evidence that long-term subsidies for families and supportive housing for individuals with high needs works. The best evidence comes from the Family Options Study, which demonstrated that HCVs overwhelmingly keep people out of homelessness. 4% of the HCV group in the study spent at least one night in shelter compared to 12% in the usual care group. Relative to the usual care group, families with a HCV experienced less food insecurity, school disruption, psychological distress, substance use disorder, and domestic violence; they also were more likely to stay together. 86% of HCV families were still living in subsidized housing 18 months after first receiving assistance.

Permanent supportive housing based on the Pathways Housing First model has also been proven successful. Evidence from NYC shows that HF recipients spent 3% of the study period (6 months) in a state of homelessness compared to 20% for those in transitional housing.¹⁴ Voluntary services also matter for successful HF models and they depend on good implementation. The evidence that HF saves money is not clear, but this is likely besides the point. Reducing and preventing homelessness is well worth the investment whether or not it saves money. More recently, a study found that individuals experiencing homelessness in Vancouver who received an unconditional cash transfer of CAD 7,500 spent 99 fewer days without shelter and 55 more days in stable housing compared to individuals in a control group after a one-year follow up.¹⁵ 

The estimated cost to fully fund the HCV program is around $30.8 million, which is roughly what we spend subsidizing disproportionately wealthy homeowners through the mortgage interest deduction. Income supports, child care, better employment protections, and enforcement against source-of-income discrimination will all help, but homelessness is fundamentally rooted in housing shortages and housing unaffordability.¹⁶ 

Comprehensively eliminating homelessness, however, will require a significant amount of affordable housing construction, which requires confronting the barriers to housing construction – namely, exclusionary zoning and a lack of public financing. At the end of the day, homelessness is a housing problem, so homelessness policy is housing policy.