The Math Behind Homelessness: A Public Health Crisis

Published on October 21, 2020

By Kurt Waltenbaugh

In Minnesota, winter is coming.

We have had what is most likely our last burst of warm weather – a beautiful fall weekend spent removing screens and air conditioners, preparing gardens for winter, planting bulbs for spring, storing lawn mowers and prepping snow throwers.

It’s also notable what is not on the fall chore list for our families: thinking about where to live, for example, or where to shelter from the lethal temperatures on the horizon.

For most of us, a safe shelter is not on our minds. Yet, on any given night in 2019, nearly 600,000 people in the United States were homeless. That number has climbed even higher during 2020, due to the dual impact of the public health pandemic and corresponding economic crisis.

Minnesota’s share of homeless in 2019 was around 8,000. Five thousand of those were in the urban core of Minneapolis-Saint Paul, including 2,000 living “unsheltered” – on the street, in a vehicle, or tent. By the summer of 2020, the number living unsheltered had grown by nearly 50%, approaching an estimated 3,000 people. Every night. Almost every public park or green space has provided tent space during 2020. One encampment alone – Powderhorn Park, just a few short blocks from the site of George Floyd’s murder – accounted for 550 tents.


So who is affected? As you might imagine, the housing crisis does not affect every community equally. Black Minnesotans are 2.5x more likely to be homeless than white. The highest rates of unsheltered homeless are BIPOC – especially native Americans, Pacific islanders, and Black residents. Overrepresented are both adult and teen LGBTQ individuals.

Financial Costs

While risk varies dramatically by individual, the costs of homelessness are born by every one of us. Taxpayers and charitable donations fund temporary shelter beds and other forms of support. Public health spending is often the largest component. One study (2017) found annual costs exceeding $35,000 per person for the homeless population, with 50% of that cost attributable to the lack of stable shelter. Paying for housing had a net savings of $4,800 per person after covering the $12,800 spend on the living space.

People who are homeless age faster, mirroring clinical conditions of those 15-20 years older than their physical age, and the clinical costs add up quickly. One study estimates than 1/3 of all emergency visits are attributable to the homeless population, averaging five visits per person annually, at a cost of $18,500. Stable housing resulted in a drop of 59% in overall medical spend.

Health Impacts

The effect of homelessness on health is dramatic, both in terms of improved health and quality of life today, along with increased longevity overall. In this northern population, researchers observed a 17.5 year reduction in lifespan – attributable to smoking, infectious disease, and weather related (cold) stress. These findings are consistent across the developed world, with mortality reductions ranging from 15 to 30 years compared with a surrounding control population.

The Math Behind the Need

The most effective way to solve the short-term problem is to provide “housing first” – to get people off the streets and into stable housing, to start with the basic physiological needs of food, water, shelter, and rest.

Numerous studies have demonstrated a positive return-on-investment for stable housing, including a substantial reduction in taxpayer-funded costs. Using Santa Clara County, California, as its baseline, an April 2019 Strong Town Media article explains the “math” behind housing needs and shows that the “benefits of a comprehensive government-run housing first program outweigh the costs to taxpayers.” Regardless of the math, costs and returns are often not well-aligned: the financial incentives to solve the problem are not aligned with the entity currently bearing the costs.

This is where using data to identify the impact can help, by shining a light on the inequities.

What’s Working Today

Bucking the trend, a Canadian experiment gave a one-time $7500 CAD cash grant directly to 50 homeless individuals, following a cohort of 115 people for 12-18 months to measure the impact. The work paid surprisingly large dividends, improving several measures of risk including both housing itself and food insecurity, with a corresponding 39% decrease in amounts spent on alcohol, tobacco, and illicit drugs. In addition, grant funds were spent on transportation (bicycle or auto repair), technology (computer access), entrepreneurial activity, and savings to build cash reserves.

As any economist will tell you, the individual marginal utility of a cash grant is always higher than that of provided services (e.g. food benefits or housing voucher) as each person can use the cash to address the need most urgent to them, as opposed to what an outside organization believes they need.

“People very much know what they need, but we often don’t equip them with the intervention or the services that really empowers them with choice and dignity to move forward on their own terms,” said Claire Williams, the CEO and co-founder of Foundations for Social Change in this recent reporting from CNN. The pro-forma savings due to avoided services (primarily shelter bed days and healthcare spend) were more than enough to pay for the cost of the program.

The Future

Imagine a future when state and local governments team up with health insurers to fund investments in housing and cash payments to the homeless, with the program “paying for itself” in reduced medical and governmental spending.

For the math to work – and for interventions like these to have the best chance for success – those in healthcare and public health need accurate information about what’s at stake. The models in Carrot Health MarketView, for example, are built on data points that include many social determinants of health (SDoH) and can alert payers, providers and policy makers about the risk for homelessness in their communities and among their patients and members. 

Does your organization already assist in your community? Is it enough? How could you do more? Please share your story. I would love to feature your work in a future post!


Blog image sourced from Jon Tyson via Unsplash.

Related Posts


Local and Individualized SDoH Data Is Critical to Advancing Health Equity


How Health Plans Can Improve the ROI of In-Home Test Kits


Q&A on the Carrot Community Committee (C3)