Homelessness is an incentives problem

The correlation between high absolute rents in large cities and homelessness is being used to promote “build more housing” as a solution to homelessness. But the analysis does little to differentiate between different types of homelessness, which require different approaches. I think there are 3 broad categories of homelessness:

1. Chronic - long term homeless, cannot afford area rent

2. Intermittent - cycles in and out of homelessness, can barely afford area rent

3. Temporary - homeless due to one-off circumstances like job loss, can usually afford area rent

As far as I know, currently existing frameworks on homelessness do not define subgroups based on whether they can afford area rent. If the proposed solution is “build more housing”, then ability to pay rent must be part of the discussion.

Who is helped by building more housing?

Of these subgroups, it seems clear that chronic homelessness is the source of most complaints: encampments, public drug use, erratic or violent behavior, piles of garbage, bicycle chop shops, rampant retail theft, biohazardous waste, fires. However, a solution like building more housing is unlikely, even in the best case scenario, to lower area rents by more than 20%. More damning, just 17% of San Francisco’s homeless reported even sporadic employment¹. 87% are unemployed. How much rent can an unemployed person afford?

Clearly, a solution that focuses on housing will not reduce chronic homelessness without a significant and permanent subsidy from taxpayers. That’s not to say that we shouldn’t build more housing (yes, it should help groups 2 and 3), but more housing will not fix the issue of visible homelessness.

Let’s do a thought experiment if you’re not convinced. Say the average rent for a studio apartment in San Francisco is $2,150 per month. Suppose that drops 20% to $1,720 due to every YIMBY fantasy coming true. Will those currently living in tent encampments suddenly be able and willing to pay for their own housing?

A quicker fix

If proponents of the “homelessness is a housing problem” theory truly believe homelessness is caused by the high cost of housing, then, instead of suggesting “build more” as the method of bringing housing prices down (which would take years at best), they would support a more direct solution that can be implemented cheaply today. Why not call for bussing the homeless to areas of the country where rent is already low?

“The homeless may not be willing to move, and it would be bad to force them,” one might argue. But any homeless person who refuses to move from a high cost of living area to a lower cost of living area is demonstrating via revealed preference that their main goal is not accessing housing they can afford.

If affordable housing is not the primary goal of the homeless, why do policymakers think they can solve homelessness by lowering housing prices?

Affordable housing is not what the chronically homeless are seeking

If the goal is affordable housing, zero homeless people would move from lower cost areas into the world’s most expensive housing markets like San Francisco, New York City, Los Angeles. And yet, they do. By San Francisco’s 2022 Point in Time Count Report, at most 27% rented or owned in SF when they became homeless:

While it’s true that 71% of those surveyed in the same report claim they were “living in San Francisco at the time they most recently became homeless”, the report does not define “living in San Francisco.” The figure is based on self-report alone and likely includes stints in motels or couch surfing. In other words, it offers no evidence that these individuals were ever able to afford area rent.

That last point is the relevant one. Should the taxpayers in expensive cities permanently subsidize housing for anyone who chooses to come set up a tent? While there may be an argument for housing subsidies at the federal level, it doesn’t make fiscal sense for any of the housing units to be located in expensive real estate markets. Why house just one person in San Francisco when the same funds can house 20 in a cheaper market?

Zero sum city budget

In an ideal world with unlimited resources, perhaps one could argue that we should permanently subsidize housing for any homeless person who moves here, even at today’s estimated cost of about $1.2 million per unit.

But that’s not our current reality. San Francisco faces a budget deficit of $291 million over the next fiscal year. What services are San Francisco taxpayers willing to sacrifice in order to pay to house homeless populations from all over the country? Why should that burden fall on San Francisco taxpayers?

Better incentives

Many government benefits and incentives already require applications with proof of residency. Qualifying for a one-time rebate of $9500 on the purchase of an electric vehicle, for example, requires proof not just of San Francisco residency, but is restricted to specific zip codes only. Does it make sense to be so strict on a one-time rebate for an action that benefits all San Franciscans (trading in an older gas car for a non-polluting EV) and yet have zero prior residency requirements for much more expensive ongoing benefits like free housing or cash assistance? Especially when cash assistance has been cited by multiple homeless people on video as the reason they chose San Francisco?

A few simple measures would help reduce our chronic homelessness problem:

• Requiring proof of prior rental/ownership in San Francisco for benefits like permanent housing and cash assistance.

• Consequences for public drug use — e.g., mandated community service (picking up litter, cleaning human waste from the street, painting over graffiti), disqualification from government benefits, fines, ban on visiting certain locations and other consequences like those found in the much lauded Portugal model.

• Asset forfeiture — seizing of any property that was likely stolen. This would reduce the profitability of shoplifting/property crime.

• Nonprofit funding contingent on results — there are significant issues with diverting public funds to nonprofits, but one way to limit taxpayer exposure to ineffective nonprofits is to evaluate nonprofits yearly based on one criterion alone: provable exits from homelessness. Not contacts made, not meals provided, not number of needles or amount of foil given out. Exits from homelessness. Fund only the nonprofits that perform well on this metric.

Given that San Francisco cannot afford to house every homeless person who chooses to come here, it makes sense to make San Francisco a less desirable destination for the homeless. Incentives are powerful, and we are currently providing exactly the wrong ones. Luckily, the correct incentives should be less costly than our current model.

1. https://hsh.sfgov.org/wp-content/uploads/2022/08/2022-PIT-Count-Report-San-Francisco-Updated-8.19.22.pdf, p 39