Homelessness and vacant houses

What a leftist talking point gets right and wrong

Lee Carter, the Virginia Democratic Socialist running for governor, got housing Twitter geared up last week by observing that “there are more empty houses than homeless people in Virginia,” indicating that the housing affordability crisis is “driven primarily by artificial scarcity and speculation, not by a genuine supply constraint.”

Carter’s overall housing politics are not so bad, and I think the YIMBY pile-on that resulted from him talking about the vacant houses was probably counterproductive.

But this is a talking point that sometimes gets raised by hardcore NIMBYs as their explanation for why we don’t need to build anything new. And I think you can see Carter himself is torn between one side of him that wants to advocate for more social housing and another side that wants to say we’re just facing a problem of “speculation.”

The basic story is that it is factually true that empty homes exceed homelessness, not just in Virginia but in every state. But well-functioning housing markets require a decent amount of vacancy — the nature of housing is that “just in time” delivery and zero inventory isn’t going to work. Beyond that, when you look around the country at which places have more-severe and less-severe homelessness problems, it’s just not the case that anyone is effectively addressing homelessness by having a low vacancy rate.

Fundamentally, I think a vacancy-centric worldview reveals the limits of a narrow focus on homelessness and “affordable housing” as the lodestar of housing policy. The real goal should be housing abundance, a framework that directly alleviates both homelessness and severe cost burdens, and also makes them easier to eliminate. And a world of housing abundance would have plenty of vacant homes.

Housing abundance, explained

For people to be experiencing homelessness in such a rich country as ours is a tragedy. And it’s also a big problem when we have so many lower-income families experiencing severe housing cost burdens.

These are really welfare state issues, and solving them requires active government provision of money, housing, and social services.

But when you look at the five states with the highest homelessness rate, you see New York, Hawaii, California, Oregon, and Washington. It’s not a question of those five states having unusually stingy welfare states. It’s a question of those states operating in the context of overall housing scarcity, with scarcity contributing to homelessness as well as other problems.

Housing abundance doesn’t just mean that few people will experience homelessness. It should also mean that you don’t see people with low-six-figure incomes talking about how they’re not “really rich.” If you’re making two or three times the median household income in the United States, you ought to be able to get a nice-sized place for a family with a couple of kids to live in without needing special subsidies. In a place with very expensive land that may mean no yard, which won’t be to every family’s taste, but it should be possible.

Housing abundance means that a person who grew up in a given place can reasonably expect to be able to afford a place of his own in his hometown after he gets a job. Probably not a place as big or as nice as his older, higher-earning, and more established parents have — but a place to live.

Most important, housing abundance plays a critical role in economic development. Janna Matlack and Jacob Vigdor find that in markets where the housing supply is allowed to expand in response to increased demand, an increase in the number of high-paying jobs in a city benefits everyone. But where the housing supply is tight, new high-earning residents simply reduce the disposable incomes of existing lower-income ones by increasing rents. With housing abundance, working-class people can move to booming areas to get better-paying jobs. With housing scarcity, working-class people flee tech and finance boomtowns in search of affordable housing.

Housing abundance means vacancies

To have an abundant housing market you need lots of vacancies, for the same reason that a decent supermarket will have plenty of food on the shelf going uneaten. Inventory isn’t the same as waste. And while it’s true that nobody should have to go hungry in the United States, the best solution isn’t to grab random unsold merchandise off the shelves and redistribute it — the best solution would be to ensure that everyone has some money in the context of an overall market goal of abundance.

A well-running housing market is particularly likely to have high vacancies because housing is very irregular.

Each bag of carrots at the store is in some way different from the other bags, but you can ask a person to pick up a bag of carrots (or russet potatoes or yellow onions) and not really care which bag of carrots they buy. But each home is unique. And while summary statistics like square footage and number of bathrooms give you information about a home, they don’t fully capture it. Small differences in location can be important. And consumer preferences can vary widely. You might mind a creaky floor a lot, or you might not. You might mind a drafty window a lot, or you might not. Street noise might bother you a lot, or it might not.

This means that even with the rise of great real estate apps (Zillow, Trulia, Redfin, etc.) the business of shopping for housing is pretty laborious and inefficient. People want to see places for themselves before signing a lease or a contract, and in general, the market just doesn’t clear that quickly. The fact that moving is a huge hassle also slows things down.

Bottom line — the fact that a relatively large share of an area’s real estate is vacant at any given time is not primarily an indication of “speculators” or “artificial scarcity.”

It’s actually the opposite. A market with a very low vacancy rate is probably dysfunctional. Empty nesters aren’t downsizing (people don’t call families like this “speculators,” but that’s what they are), new people have trouble moving in, new stuff isn’t being built, or housing scarcity is so severe that whenever new units come onto the market, people snatch them up really fast rather than shopping around.

This is why even though it’s true, mathematically, that empty units exceed homelessness in every state, it’s not as if low-homelessness states have accomplished that by having low vacancy rates.

Homeless and vacancies

Here’s a chart Marc did plotting the state-level homelessness rate against the state-level vacancy rate — as you can see, to the extent that these are correlated, states with less homelessness have more empty units, not fewer.

The outlier in terms of high-homelessness rates is D.C., which has the homelessness problems of an expensive city rather than those of a typical state. But if you toss D.C. from the sample, the correlation is still -0.379 — the low-homelessness states didn’t eliminate vacancies and the high-homelessness states don’t have especially high vacancy rates.

At an MSA level, the margin of error on the city-specific vacancy estimates gets big; the homelessness data is poorer; and I would not take these figures to the bank. But the relationship is, again, negative.

The issue is that to the extent that homelessness and vacancy rates have anything to do with each other, they are both symptoms of housing scarcity. A city where housing is scarce will have a low vacancy rate. A city where housing is scarce will also be one where dwellings are hard to afford without subsidy, so people who suffer economic misfortune or problems in their personal life end up unhoused before they can make it to the front of the line for housing assistance.

Now the vacancy-centric narrative is adjacent to a point that’s important and correct that I’ve written about at both Slate and Vox — the best way to address homelessness is to get the people experiencing homelessness into homes.

The unhoused are often suffering from a variety of problems in life and it can be tempting to get trapped by a “root causes” mentality that wants to focus on addiction, mental health, or employment services rather than housing. But the evidence points strongly in favor of what’s called a “Housing First” approach for two reasons. One is that people sleeping rough on the streets experience a lot of costly problems via encounters with law enforcement and emergency rooms. The second is that it’s easier to address people’s underlying difficulties when they have a safe place to sleep at night. So by far the most humane and cost-effective thing to do for people experiencing homelessness is to get them housed.

To the extent that the observation “vacant units exist” supports the intuition that there’s simply no reason for people to be out on the streets, that’s correct. What’s wrong is to see homelessness as conceptually distinct from larger housing supply issues.

We live in a (highly litigious) society

Leftist thinking on housing policy often goes awry by abstracting too much away from the particular institutional details of the American legal system.

For example, you could say “people are homeless while other people own seasonal homes that are empty for most of the year so capitalism is the real problem!”

But neither the state of Virginia nor any other state or local government anywhere in the country can confiscate people’s vacation houses and give them to the homeless. You can buy vacation houses and give them to the homeless. You can even use eminent domain and force the sale of vacation houses. But you still need to pay for them. This means that any attempt to acquire existing structures and repurpose them as social housing is going to benefit from a lower market price of housing. Similarly, any effort to give low-income families rental subsidies will benefit from a lower market price of housing. And if you want to build new social housing or homeless shelters from scratch, you need to run the same gauntlet of litigation and NIMBY protest as any market-rate development.

In other words, while there are a lot of viable paths forward, they all require first changing the rules to greatly reduce people’s ability to exclude new construction from their neighborhood.

If you do that, housing supply problems ease and you get closer to housing abundance. Then operating inside a context of housing abundance, you can give rental assistance vouchers to the needy. Or you can give cash assistance, knowing that it can be used for rent. Or you can give grants to non-profits to build affordable housing. Or you can directly build government housing. Or you can acquire units and use them as social housing. Whatever you want to do is more doable in a context of abundance and harder in a context of scarcity. And abundance probably means more vacant units rather than fewer.