Homelessness: A Tale of Three Cities

Published in

Hatchbeat

12 min readOct 30, 2018

Homelessness is a crisis: 14 years and counting after then-Mayor Gavin Newsom pledged to “abolish chronic homelessness” in San Francisco, more than 1,000 people still wait each night for a temporary shelter bed while the average homeless population is modeled around 7,500 individuals day by day (based on the point-in-time count)¹.

This could soon change. Though our new mayor has pledged to fix homelessness by following through on her plan, local activists took action sooner to bring Proposition C to the Nov. 6 ballot by citizen petition. Prop C, with consideration of the potential funding, is one of the most salient measures to date to address homelessness in the city.

The business tax measure would draw from the wealthiest businesses in SF, effectively doubling the city’s budget for addressing homelessness thanks to a 28–33% increase in overall business tax revenues.

Prop C is supported by dozens of local and state officials, organizations, and Salesforce CEO Marc Benioff, but opposed by SF Mayor London Breed, State Senator Scott Wiener, Twitter, and Stripe.

Stripe’s CEO Patrick Collison wrote a memo explaining Stripe’s position on Prop C, which emphasized the staggering difference in expenditures on homelessness per capita when comparing SF to Los Angeles and New York City:

“If homelessness was just a question of money, this issue would already be solved. While cities report inconsistently, San Francisco currently spends around $430 per city resident per year on services and programs for the homeless, compared to $260 in New York and $110 in Los Angeles. Yet the problem in our city is worse, and despite increases in spending, has continued to worsen.”

We were curious about such a big difference in spending across cities, so we dove into what these numbers actually mean for each city.

Summary

With more comparable budget numbers, SF spends $293 on homelessness per resident while NY spends around $330.
There are many types of homelessness. Cities have different mixes of people to help. Some needs are more costly to solve than others.
The homeless population has remained somewhat stagnant in SF and NY, peaking at hovering around 7,500 in SF the past few years.
LA is doing the worst at mitigating its homeless problems across all metrics compared in this article (and everything else we saw while researching).
SF’s programs support a higher proportion of chronically homeless adults than NY’s or LA’s.
A majority of the homeless population in NY are homeless family units, with a low number of chronically homeless single adults or people with disabilities.
NY has a right to shelter mandate and operates under a shelter-first approach, so almost everyone in need is sheltered.
NY’s cost per bed per day in shelters are comparable if not slightly higher than SF.
SF invests more in Permanent Supportive Housing (PSH) than NY or LA.
Rental assistance programs are used by SF and NY as both a preventative and problem-solving measure.

The per capita spending numbers are not comparable

If you’ve read a few articles on Prop C you’ve probably seen the figure $380 million for homelessness expenditures thrown around. As explained by Alex Lash of The Frisc:

[SF’s] Department of Homelessness and Supportive Housing’s 2018–19 budget is $284 million.This does not include money that other departments, such as Public Works and Public Health, spend on health care and street cleanups. The true figure, according to the city’s budget analyst, is $380 million.

We agree that this is a better estimate for full costs, but New York and LA’s per resident spending numbers are derived from budgets which do not include these extra costs. LA’s $110/resident comes from the new 2019 budget, which bumps homelessness spending to $440M² (a very small amount of this which does go to street cleaning), NY’s 2018 homeless budget was just under $3B³ (we decided to not include the “street homelessness” spending entry for a conservative comparison). According to most recent numbers⁴ from SF’s Controller’s Office, SF spent $259M on homelessness in 2018. With these numbers San Francisco’s spending per resident on homelessness is actually slightly lower than NY’s: $293 for SF and around $330 for NY.

Homelessness looks different everywhere you go

To get a better sense of how diverse homelessness is across these three cities, we looked at some basic counts and demographics to understand who homelessness services were targeting. Later we’ll see how. All three cities counted between 800–880 homeless people per 100,000 residents in the 2017 point-in-time surveys, but diverged when it came to the subgroups.

Chart 1: Sheltered vs unsheltered homeless individuals per 100K residents

Outright this chart doesn’t look that great for SF or LA. While they have a slightly lower homeless population overall per 100,000 residents, the percentage of people who remain unsheltered in SF or LA is staggering in comparison to NY. New York, which boasts a right-to-shelter mandate, supports one of the largest homeless populations per 100K residents, and is able to shelter most people seeking services.

So, does that mean New York is doing it better? Short answer: not necessarily, but they are at least successfully sheltering everyone. We have to go a few layers deeper to understand each city’s approach and relative measures of success.

Chart 2: Percentage of homeless population by subgroup

Subgroups represented across the homeless population were similar in LA and SF but distinctly different between the east and west coast, with a higher percentage of homeless youth and veterans and high rate of chronically homeless individuals in SF and LA versus a primary family-based homeless population in NY and low rate of chronic homelessness.

Again, 59% of the NY homeless population are part of families, compared to 9% in SF and 15% in LA. (Note that the heads of households also fit into that description.) Of that 59% in NY, about one in four, or 28% of families⁵, were dealing with temporary homelessness though a family member was employed. This is slightly higher than the overall percentage of working homeless individuals in San Francisco, where about 13% are working part-time or full-time¹. Many factors can contribute to the low rate of employment while homeless, and in fact the leading causes of homelessness in SF is from a lost job¹.

There’s a high ratio of chronically homeless individuals in west coast cities

Chart 3: Percent of individuals facing chronic homelessness

What is chronic homelessness?

The Department of Housing and Urban Development (HUD) defines chronic homelessness as “either an individual with a disabling condition who has been continuously homeless for a year or more, or an individual with a disabling condition who has had at least four episodes of homelessness in the past three years.”

Table 1: Point-in-time survey health conditions in San Francisco, 2015

This table demonstrates the challenges chronically homeless individuals face, as they are drastically more likely than the rest of the homeless population to be experiencing drug or alcohol abuse, psychiatric or emotional conditions, physical disabilities, PTSD, chronic health problems, brain injuries, and AIDS/HIV.

What’s the current strategy?

Chart 5: Temporary beds (emergency, transitional, and safe haven) per 100k residents

One of the quickest way to shelter someone who is homeless and unsheltered is to admit them to a homeless shelter. That’s been NY’s approach. To address its right to shelter mandate, NY aggressively increased its temporary bed count between 2010–2017 to shelter its homeless population. The result: NY maintains 909 shelter beds per 100k residents compared to 324 in SF and 165 in LA.

Sheltered people stay in emergency shelters, transitional housing, or stabilization rooms. San Francisco has been putting more emphasis on increasing the number of shelter beds in recent years, with the late Mayor Ed Lee pledging last winter to pull 1,000 people off the streets before his death, and Breed continuing on with Lee’s direction. In the past, circa 2004–2014, we actually decreased the number of shelter beds deliberately, moving in favor of 24-hour crisis shelters.

As Nick Josefowitz points out:

Only in 2015 did shelter building return, but in the form of Navigation Centers. These facilities come with more supportive services and fewer rules than traditional shelters — couples can stay together, pets are allowed, belongings can be stored, the centers are open 24 hours a day, meal times are not strictly set — but they also kick people out after 30 or 60 days and cost twice as much to operate as traditional shelters.

How do the three cities compare for the costs of individual shelter beds?

Before we look at this section we have to acknowledge that each shelter has a different cost of operation, supports varying subgroups and number of individuals at a time, and offers various services (such as hours of operation and accessibility, showers, kitchens, lockers, etc.). We have to also acknowledge that permitted lengths of stay vary, with NY permitting a seeming indefinite period until exiting homelessness and SF and LA maintaining a ~90-day emergency shelter period.

We looked at what was publicly listed in budgets to determine an approximate cost per bed per day in each city and confirmed against numbers reported in various articles. We didn’t find any reports or budgets which were detailed enough for San Francisco, but various articles report around $90 per bed per day during 2017⁶. New York’s cost per bed per day in 2017 is $100 (calculated from the NY budget and corroborated with other sources). LA reimburses shelters $50 per bed per day, a third party study estimated the average costs at $62 per day⁷.

So where is the rest of San Francisco’s money going? Mainly housing.

Chart 6: Total permanent supportive housing beds per 100K residents

What is permanent supportive housing?

The HUD defines permanent supportive housing (PSH) as “permanent housing with indefinite leasing or rental assistance paired with supportive services to assist homeless persons with a disability or families with an adult or child member with a disability who achieve housing stability.”

By looking at the number of PSH beds you can start to understand a big strategic difference between NY and SF/LA. New York is focused on opening up as many shelters as possible. In 2018, 63% of its homeless budget went to shelters³, compared to 12% in San Francisco in 2015–2016⁸. San Francisco had 971 PSH beds per 100k resident in 2017 compared to 332 for NY and 198 for LA. In fact, it has aggressively been building or leasing more the past few years in contrast to NY and LA, which have seen no growth in PSH units.

Building or leasing permanent housing is way more expensive than shelters, which starts to explain SF’s higher homeless spending per resident. San Francisco spent 45% on PSH in 2016⁸.

At this point it’s clear that LA is under-performing across all metrics when it comes to shelters and PSH units. Homelessness in LA has also rapidly increased over the past few years, during which NY and SF essentially stabilized the size of their homeless populations (see table 2). Keep in mind that LA spends less per capita for homelessness compared to NY and SF. The number used in Stripe’s memo ($110 per resident) was calculated with the 2019 budget, which doubles the spending compared to this year, mostly due to prop HHH (voted in 2016) which allocates money for the building of supportive housing units.

Why is SF building so many more PSH units? This goes back to the high percentage of chronically homeless individuals and those who are facing issues that require ongoing support programs (not to mention the high proportion of unserved low-income families and those who are at-risk of homelessness).

It clearly needs to do more to help at a point in time when 58% of individuals go unsheltered. We can now also see that SF and NY have to solve slightly different problems. While NY’s homelessness skews towards families with transitory homelessness, SF deals mostly with individuals, a sizable amount of which are chronically homeless.

Exits

It’s also interesting to look at the numbers of people exiting homelessness, but unfortunately hard to compare them across cities. Keep in mind that the number of people who were homeless at some point in a year fluctuates as people exit and enter homelessness, so the number of people requiring support annually is overall much higher than the number represented by a point in time count. Cities can only track individuals who pass through their shelter system and counts beyond that must be extrapolated from rendered services and point-in-time counts, which entails a city-wide canvassing effort.

San Francisco publishes open metrics on known exits: 2192 individuals transitioned out of homelessness from mid 2017 to mid 2018, compared against the 7,500 individuals who were accounted for in the point-in-time count.

Source: https://public.tableau.com/profile/hsh.data.performance#!/vizhome/HOMStat_Distrib_TableauPublic/HOMStatStory

It was slightly harder to get numbers for NY, but thanks to the Coalition For The Homeless’ report⁹, we estimated around 7,150 families exited homelessness throughout 2017 out of 15,170 homeless families at a point in time. 1,500 single adults in NY were moved to permanent housing out of 16,200 homeless single adults at a point in time.

What were the main drivers for these exits? In San Francisco it boils down to overwhelmingly two routes: permanent supportive housing and relocation assistance (Homeward Bound program).

So where else does the budget go?

As far as housing is concerned the last category is eviction prevention and rental subsidies, which make up 12% of the budget in SF⁸.

New York’s city-initiated rent subsidies contributed to 63% of their 2017 family exits by helping them to move out of shelters with 30% of family exits involving a move from shelters to NYCHA public housing⁹.

It’s noteworthy to mention that 25% of the surveyed homeless people in the point in in time survey in SF were facing their first-time of homelessness¹. As mentioned earlier, the primary cause for homelessness was job loss for 22% of surveyed individuals. Following that, 59% of the surveyed homeless people in SF mentioned not being able to afford rent as one of the obstacles to finding housing¹.

This is where rental assistance programs come in. The City of SF subsidizes rent for 275 families each year and also gives one-time grants to 800 individuals (both single adults and families)¹⁰. These programs primarily target people who are at-risk of homelessness, though some families and individuals who are already receiving government assistance qualify for rental subsidy programs. This doesn’t make up a large percentage of the budget or number of people who are supported in SF.

Our Take

Having a conversation about various budget allocation strategies across all programs is just as important as looking at the accountability of various programs.
What is the best mix of shelters, permanent housing, and rental assistance programs? Which program is the most cost-effective for each type of homelessness?
SF’s Prop C would double the spending on homelessness with an allocation strategy fairly similar to the current one: invest strongly in permanent housing.
What if we want to change the Prop C budget allocation in the future? Will there be enough momentum and willpower to introduce a new ballot measure? Can the BOS change the charter without voter approval?
Both Liz and Fernand will vote yes on Prop C. Liz is a “hell yes” while Fernand has some questions (that Liz agrees with) around how easy it will be to amend the current spending distribution if it is no longer a fit for future needs and allocation strategies.

This article was edited on 11/4/2018. We updated the SF budget numbers (and source) to use 2018 numbers instead of 2019 ($259M instead of $284M) and the SF homelessness exit numbers after a email exchange with Jeff Kositsky who pointed us to more accurate sources.

Sources