Bunce giving a talk

Calculating the impact of HUD’s homeless formulas

It took a pandemic for the federal government to actually consider homeless people in determining how much homeless aid to give out. Without congressional action, a decadeslong formula problem remains.


Eric Smith will finally have a place to call home after years of sleeping in his car and on his aunt’s sofa because of an obscure change in how the federal government calculates homeless aid.

For decades, Washington has determined how much taxpayer money to give states, counties and cities for their homeless residents, using a formula that actually has nothing to do with homelessness. But the COVID-19 pandemic, and its billions in emergency aid, forced lawmakers to rework the old, politically popular homeless formula, which spread the wealth regardless of need.

In Los Angeles, the change has meant nearly $120 million more in aid and a chance for Smith and thousands like him to fundamentally change their lives.

“I'm not going to take that for granted,” said the 44-year-old Smith, who until recently worked in a Los Angeles group home for people with autism and mental disabilities. “Once I get my own place, I could grow more and do more.”

Eric Smith selfie
Eric Smith takes a selfie in the hotel room provided to him as part of California’s Project Roomkey in Norwalk, Calif., in November 2020. He’ll soon move into an apartment that will be partially paid for with CARES Act homeless aid funds. (Source: Eric Smith)

When Congress passed the Coronavirus Aid, Relief, and Economic Security Act in March, it included $4 billion for one of the government’s main homeless funding programs. Given that the cash infusion was nearly 14 times greater than last year, lawmakers also mandated a change in how the payments were calculated, because anyone who knew anything about the funding formula knew it had nothing to do with homeless people.

But the fix was a one-time-only change and, barring any new action by Congress, the government will revert to using the misdirected formula for homeless aid next year.

“If we're going to try and address homelessness, we should use a formula that prioritizes homelessness,” said Alex Visotzky, legislative affairs manager for the Los Angeles Homeless Services Authority, one of the groups that advised the Department of Housing and Urban Development on ways to rework the formula.

HUD agreed, and in June released a new formula for calculating Emergency Solutions Grants that factored in an area’s homelessness and measures of poverty — a dramatic change from the old formula that was heavily influenced by the number of prewar homes in a given area.

The change had a profound impact on which parts of the country got homeless aid.

An analysis by the Howard Center for Investigative Journalism found that most of the CARES Act homeless funds went to a minority of places with the most severe problems. And 62% of the 362 cities, counties, states and territories eligible for these funds received relatively smaller amounts than they would have without the formula change.

California, which has over half of all unsheltered homeless people in America, benefited the most. Cities and counties up and down the state, such as Oakland, San Jose, Long Beach and San Diego, got millions more in funding.

Rust Belt cities and counties saw the biggest relative decreases, the Howard Center analysis found. Five of the 10 largest decreases occurred in western Pennsylvania. Pittsburgh took the largest hit, dropping 82%, followed by other cities in the Midwest.

Applying the revised formula over the last decade, the Howard Center analysis also found that states such as California, Florida, Oregon, Washington and Colorado could have gotten millions more in homeless funding. Los Angeles city and county combined would have received $119 million more since 2011; New York City would have gotten an additional $131 million.

Experts say that amount of money, if directed toward permanent housing, could have helped thousands of people move off the streets.

“The more homeless shelters you build, the more homeless people in shelters you see. And the more affordable housing you build, the more people that are housed you will see. You get exactly the results that you pay for,” said Sam Tsemberis, founder of Pathways to Housing, a homelessness advocacy and training group.

One-time wonder

Congress made the CARES Act homeless relief funding available in two rounds of payouts during the spring and summer. After HUD took a 1% cut to train service providers, that left $3.96 billion for communities trying to avoid a worst-case scenario for their homeless populations in the pandemic.

In the interest of speed, the first round of grants, totaling $1 billion, went out under the old formula, said Todd Richardson, an economist who leads HUD’s research division. The department could have made double that amount available under the law, but chose not to in a tacit acknowledgment of the old formula’s shortcomings.

Comparing Los Angeles and Chicago illustrates those problems.

Chicago is a smaller city with just over 5,000 homeless people, compared to the estimated 41,000 homeless in Los Angeles. Despite the difference, Chicago regularly pulls in significantly more funding than Los Angeles in part because it has a larger stock of homes built before 1940.

While homelessness has decreased nationally since 2007, in California it has increased to about 151,000 sheltered and unsheltered homeless people. That’s due, in part, to a statewide shortage of affordable housing, which drives up prices and pushes out low-income people, like Eric Smith.

The formula’s flaws were well-known to anyone who worked on homeless policy, but the problem persisted because there was not enough political will to push through reform, Capitol Hill and HUD staffers said. When Congress drafted the $2 trillion CARES Act in March, however, it mandated that HUD create a new way of estimating need in order to ensure better distribution of the unprecedented amount of Emergency Solutions Grants, or ESG.

“We have known for some time that the ESG formula, while it's effective at getting money to many poor communities, doesn't target great to homelessness,” Richardson said in an interview. “We believed that we needed to ensure that the vast majority of the funds were targeted as effectively as possible to the places that have the most severe homeless needs.”

To fix the formula, HUD adopted four basic metrics to assess need during the pandemic: total homeless count, as determined by HUD’s biennial Point-in-Time homeless census; total count of unsheltered homeless people; the rate of renter households whose family income is less than half the area’s median family income; and the rate of the very low-income renter households, who are living in overcrowded conditions or without a full kitchen or plumbing.

As a result, Youngstown, Ohio, saw one of the largest relative decreases from the CARES Act’s new homeless funding formula, according to the Howard Center analysis. The city is part of Mahoning County, which counted 148 homeless people within its borders in 2019. By comparison, Los Angeles County had an estimated 59,000 in 2019 and an estimated 66,000 homeless people in January 2020.

Youngstown received just over $1 million in the first round of CARES Act funding. In the second round, with three times the amount of money going out compared to the first, it received $657,102.

Colleen Kosta, who coordinates homeless services for Mahoning County, said she understands the magnitude of homelessness is far greater in states like New York and California, but she also worries that any change in funding would strip her community of precious resources.

“As a small community and a small city, we're already getting significantly less money,” Kosta said. “While we don't have a large unsheltered (homeless) population, we still have a population of people that we're trying to assist and make sure they're safe and safely housed.”

Had the new formula been used years earlier, the Howard Center’s analysis showed, 81 smaller cities and counties nationwide that received direct homeless aid from HUD over the last decade would have fallen below the eligibility threshold.

Rust Belt states like Pennsylvania, Ohio and Illinois would have taken the biggest hits, while California, Florida and New York would have seen the largest funding increases. Using the untargeted formula meant that these coastal states with America’s highest homeless counts lost out on millions of relief dollars.

The birth of a bad habit

Retired economist Harold Bunce was surprised to learn that the 45-year-old formula for fighting urban decay was being used to calculate homeless aid when the Howard Center tracked him down in September.

“They don’t use any direct estimates of homelessness when they distribute the funds?” asked Bunce, who retired as head of HUD’s economic affairs office over a decade ago.

Bunce was a newly minted PhD in the 1970s when he started working at HUD. Analyzing the community development block grant formula was one of his first projects for the department.

Bunce giving a talk
Harold Bunce, then deputy assistant secretary for economic affairs at the Department of Housing and Urban Development, speaks at a housing market summit in Washington, D.C., on Nov. 17, 2006. Early in his HUD career, Bunce worked on allocation formulas that are still in use today. (Source: Ricardo Vargas / National Archives and Records Administration)

The Ford administration created the community development block grant program in 1974 as an outgrowth of the urban renewal movement of the 1950s and 1960s. The program allows recipients to build sewers and parks, tear down old buildings and otherwise redevelop a city.

The formula is actually two in one. The first takes into consideration population, poverty rates and overcrowding. The second considers poverty rates, the number of housing units built before 1940 and “growth lag” — slower population growth compared to the national average or a population decline. A locality receives funds based on whichever formula awards it the most money, and just under half of the homeless grant recipients fall under the second, which considers prewar housing.

When the homeless relief grant program was created in 1987, HUD lacked a means to accurately count homeless people in America. So it used the tool it had on hand: the community development formula. But repurposing that formula for the homelessness program meant that communities with a large number of old homes often got more homeless relief dollars.

The Senate Appropriations committee acknowledged this flaw in a 2001 report that said, “The CDBG formula has no real nexus to homeless needs.” Yet the program carried on relatively unchanged until the COVID-19 pandemic.

The emergency solutions grant program traditionally has been a relatively small part of the government’s overall homeless relief aid. In the 2020 budget, for example, it received only $290 million, compared to HUD’s bigger homeless grant program, involving Continuums of Care, which received over $2.3 billion. These HUD-approved entities coordinate homeless services at the regional and local level. But even in that grant program, the flawed emergency solutions formula plays a role in how much money some localities can receive.

Igor Popov is the chief economist at Apartment List, a website where renters can find a place to live. In his 2017 Stanford University doctoral dissertation, Popov detailed how a community’s stock of pre-1940s housing is a terrible indicator of economic outcomes. In an interview with the Howard Center, he called its use one of HUD’s “original sins.”

“This was not some terrible misunderstanding of what, you know, drives homelessness,” Popov said, adding that old housing was just a variable “that really stuck.”

He pointed to the Boston suburb of Newton, Massachusetts, as a stark example of how wealthy communities and their residents often have money to preserve historic houses, maintaining their prewar housing stock.

Compare that to a Midwestern city like Topeka, Kansas, which has more homeless people but fewer homes built before 1940. Since homeless relief funds are tied to old housing stocks, Newton earns more homeless aid than Topeka.

A more perfect formula

Most complaints about the homeless funding formula have been parochial, of the “my city didn’t get their fair share” ilk, rather than a desire to overhaul an ineffective system, said a senior congressional staffer.

But when it came time to write the CARES Act, legislators were in total agreement that the formula needed to be changed, he said, adding it was just a matter of deciding which variables to use.

Even though lawmakers have not updated the formula in the past, the CARES Act change was less contentious because it was a one-off and no community lost money since the aid funds were in addition to 2020’s regular allocation.

The CARES Act "broke every program in government," said the staffer, who spoke on condition he not be named. He said he was looking "forward to finding out which flexibilities that we've applied should be taken forward as a general rule."

But changing to the new formula permanently will be an uphill battle. As Bunce put it: “Once something is locked in, it's hard to change it in Washington.”

Part of the reason for that is how homelessess is counted, a key metric in the CARES Act formula. It’s a notoriously difficult task and the current method, HUD’s Point-in-Time count, is a flawed data-collection tool, according to experts.

The count is conducted at least biennially, but annually in some jurisdictions. It generally happens on a night in January, when volunteers and service providers walk around to count and interview people experiencing homelessness. They also count those in homeless shelters to report totals for sheltered and unsheltered people.

This method generally undercounts the total homeless population and leaves out those who are in unstable housing situations — couchsurfers like Eric Smith who can’t afford a home of their own.

Richardson said HUD would not use homelessness counts in a reformed formula because that data is collected by the same organizations that would stand to benefit if HUD tied additional funding to higher homeless numbers. HUD avoided this problem in the CARES Act because it used data that was collected in 2019 before service providers knew higher counts would garner larger stimulus funds.

Despite the counting conundrum, HUD does not lack ideas for how to remedy the funding process, and Richardson estimated the department could devise a new formula in under a year.

But getting such a permanent change passed in Congress would be tough. Lawmakers don’t typically vote to cut their constituents’ funding, and the CARES Act formula would have increased homeless grants to just 11 states and the District of Columbia over the last decade, while cutting funding everywhere else, according to the Howard Center analysis.

Homeless experts like Visotzky, the Los Angeles official, say the solution to the formula problem is more homeless funding overall to get people into housing and prevent others from falling into homelessness.

”The more we have a scarcity of resources, the more we have to fight over these formulas,” Visotzky said. “We need to use this opportunity to really rethink the way we do a lot of things.”

"It would be a shame to have made all of this progress,” he said, “and not learn from it.”


Read Other Stories