What Is a Housing Shortage, Anyway?

There’s a wonderful comic, originally from Saturday Morning Breakfast Cereal, that illustrates what it feels like to go through the process of learning about a complex topic. Even better, it doubles as an explanation of much of what’s wrong with the internet:

(Source: Saturday Morning Breakfast Cereal.)

This isn’t an insult directed at anyone: it’s not that the people on Mount Stupid are stupid at all. It’s that when a subject is complex and nuanced, everyone learning about it must at some point traverse Mount Stupid in their journey to knowledge.

Housing markets are a prime example of a subject with a “Mount Stupid” in its learning curve. They’re something about which a whole lot of people know a little…yet they “know” it very confidently and are eager to share. Meanwhile, the people who know a lot tend to be keenly, humbly aware of how little they know.

Witness the fact that the wonky, deep-in-the-weeds (and excellent) housing blog Construction Physics recently published a blog post by Brian Potter titled, “Is there a housing shortage or not?”

Like, how are the experts even asking a question like that? You can walk up to the average American on the street and say, “What do you think about the housing shortage?” and they’ll probably nod sagely and have an opinion. So how on earth is a question that seems so foundational—how many homes are we short?—so contentious?

In fact, “shortage” is quite ill defined. And the number of homes we “need” is a moving target, despite figures confidently bandied about (like the 3.5 million for California—by 2025!—that Governor Newsom made part of his election campaign). It’s profoundly unclear what number, built where and how, will satisfy a particular policy goal.

And this has important implications for how we talk about the housing crisis. It’s analogous to other areas of planning, like traffic, where projections are not only invariably wrong, they’re often a deeply harmful distraction from the conversations we ought to be having.

Projections turn our minds away from holistic, systems thinking, and toward monomaniacal solutions, whether that’s “build build build build build build build build” or “build a carefully calibrated amount of only the right stuff in only the right places.” They make our thinking more linear than it ought to be. They push our policy debate toward pronouncements from Mount Stupid.

In reality, we need to be talking about systemic responses that don’t require us to have a functioning crystal ball, or even be remotely right about the future.

(Source: Unsplash.)

There’s No “Shortage” in Plain English…and That Fact Doesn’t Really Matter

The most obvious, plain-English interpretation of “there’s a housing shortage” is a numerical shortage: “There are literally not enough homes for everyone.” This is flatly not true in the vast majority of communities. Yet you’ll see people misleadingly use this definition of “shortage” to argue that building homes should not be a priority.

Take the factoid/meme that there are far more vacant homes in America (or Your City’s Name Here) than there are homeless people. By now, that meme has circulated through my social media feeds, and probably yours, at least a dozen times. It’s a classic case of Shouting From Mount Stupid. The claim falls into the category of technically true falsehoods: There are more vacant homes than homeless people, but that doesn’t mean what it’s usually implied to mean, which is that we don’t need to build anything, we just need to get people into the homes that exist.

There’s not really any good way to do that. The unstated assumption of the “vacant homes” factoid as it’s usually deployed is that it reflects homes that could easily be inhabited but are simply being held empty by their owners on a whim or for convenience. A lot has been written about the reasons that is overwhelmingly false, and I’m not going to turn this piece into a rant on the “vacancy truther” debate. Suffice it to say that many homes that show up in statistics as “vacant” aren’t meaningfully “vacant and available for someone to live in.” They may be newly built, in between tenants or owners, undergoing renovation, used seasonally, or some other qualifier, or they’re in either a location or condition where the demand doesn’t exist to fill them. No one has found a policy—a vacancy tax or other incentive—that will cause more than a small fraction of these homes to become inhabited.

The deeper truth is that “just enough homes for every person” is a terrible policy goal, anyway. Every market needs some slack in it to function effectively. As a general rule, the more slack—the more vacancy—the more and better options buyers and tenants have when looking for housing.

We want people to have choice: to be able to move cross-country, move out of a parent’s house or in with a significant other, get out of a bad relationship, live closer to a new job, or any number of other reasons that an individual might want to switch homes—and to not have a hell of a time finding a place to live when these circumstances arise.

On examination, it should be obvious that we need there to be more homes than households in order for this kind of choice to exist. So a logical question is, how many more?

“Shortage” as a Functional, Not Literal, Description

When I say “housing shortage” (which you don’t have to dig very hard to find in my writing and social media posts) I’m not talking about absolute numbers of homes. What I mean, and what I think most people who use the phrase mean, is we don’t have enough homes to avoid bad things happening.

The “bad things,” of course, are in abundant evidence. Take California, the state with the worst housing crisis: working-class people are leaving the state by the hundreds of thousands. Homelessness has surged. An eviction in San Francisco or L.A. often means having to leave the city altogether. Record numbers of 35-year-olds live with their parents. Service workers commute two hours or more to jobs in places like Silicon Valley. Apartment showings are attended by dozens of prospective tenants, some of whom wave wads of cash at the landlord.

These are human consequences of a housing shortage, and by this use of the phrase, they are themselves the evidence that there is a shortage. People are struggling to find housing. They’re being squeezed. How much construction—and where, and of what form—would stop the squeeze? That’s the whole question, and the answer is elusive, not definitive and knowable.

How Many Homes Do We “Need”?

If you like graphs and charts, go read Brian Potter’s whole post on whether there is a housing shortage or not. It’s fascinating. If you’re more of a normal person who’s not into graphs and charts, here’s the takeaway from Potter: we don’t know how many homes we “should” be building, and most of the metrics we might logically use to project that answer don’t actually seem to tell us very much.

Potter starts with an apparent contradiction between two facts, raised by a recent debate spurred by journalist Kevin Drum. These facts are:

1. Housing starts fell off a cliff in the U.S. after the 2008 crash, and haven’t recovered yet. We’re building less than we historically did for decades on end.

2. Despite this, the number of homes in the U.S. has kept pace with household growth and actually outpaced population growth.

View fullsize

(Click to enlarge. Source: FRED.)

(Click to enlarge. Source: Kevin Drum.)

Potter goes on to explain why you can’t infer a shortage just from construction rates alone. There is no “correct” number of homes we should be building, as a country or in any given city. That number depends on factors such as:

• Household size and composition. In what kinds of groups are people living together?

• New household formation. When do people move out from under their parents’ roof, or move out of a shared apartment with roommates to get their own place, etc.?

• Population growth. This includes immigration and domestic migration.

• The rate of loss of old homes, either to demolition or conversion.

Here’s the thing: all of the above are themselves affected by the housing supply that exists, and by how much new construction is happening! It’s a bunch of interconnected feedback loops:

• If rents are cripplingly high, more people will choose to live with roommates well into adulthood, or extended families will live together who might otherwise not have done so. Household size will creep upwards. 

• Local/regional population growth is of course profoundly affected by housing supply. In a sense, it is possible to choose not to grow: if you simply don’t add homes, you won’t add households, almost tautologically. What will happen is you’ll alter the mix of who stays and who leaves. And the rich (both existing rich, and rich who’ve moved in from elsewhere) will tend to displace the poor.

• What about replacement of old homes? This is affected by a lot of factors, and one interesting thing is that it’s profoundly different from one country to the next. Japan, often cited as a YIMBY ideal, has tremendously high housing starts in part because Japanese homes depreciate to almost worthless over a generation. Meanwhile, many Americans live in lovingly refurbished 100-year-old homes, and as a lover of such houses who grew up in one, I’m not going to sit here and tell you that’s a bad thing!

This stuff is confounding even to the experts. There is not a neat linear relationship even between things you’d expect. The most intuitive measure of a housing shortage—home construction failing to keep pace with population growth—doesn’t tell us very much. (“Hilariously non-predictive” is how Potter puts it.)

One thing that does: vacancy rates. Potter presents evidence that within a metro area, rental vacancy rate tracks pretty convincingly with the rate of increase in rents. When there are few vacancies, landlords hold all the cards, and tenants get squeezed. When there are many vacancies, tenants have more options, and landlords can’t hike the rent as sharply (or at all).

There is still a world of nuance here. This trend relates to average rents across a region, but a regional vacancy rate may not tell us much about the options available to the poorest residents of a place, or within a given neighborhood. It’s also true that the relationship between vacancy and rent increase falls apart when we try to use it to compare one city or region to another. Every place appears to be at least somewhat a unique beast.

But the vacancy thing is pointing at a central truth: The market meets people’s needs when there are options. That’s the ball we should really keep our eye on—not hitting some fixed, target level of housing supply, but making it easy and fast to create more housing options.

A lot of progressives groan at any mention of “the market” because they read, “let the market decide,” and hear, “for-profit companies have our best interests at heart.” But that’s not what I mean. “The market” doesn’t mean “the whims or decision-making processes of market actors” (that is, developers, in this case). “The market” means feedback loops. It means the emergent process by which people change their course of action based on real-time information on what’s working or isn’t. And the shorter and faster those loops are, the better.

The fastest feedback response in housing is things like home sharing or taking in a subletter. This should be really easy and ordinary. Carving more units out of existing dwellings—a separate basement apartment or backyard ADU—should, again, be something that ordinary homeowners can do without financial or logistical strain. We don’t need to leave it to specialized developers. Developing a property to the next increment should be simple and as-of-right. None of this negates the need for larger development, as well.

Our housing policy will be more productive the more it is focused on allowing a thousand flowers to bloom—let more people create more options in more places in more ways—than on gazing into the crystal ball to discern how many homes we need.

---

---