The Best Evidence Yet for the “Housing Musical Chairs” Theory
Pre-order your copy of our upcoming book, Escaping the Housing Trap: The Strong Towns Response to the Housing Crisis, written by Charles Marohn and Daniel Herriges!
(Source: Unsplash/Allec Gomes.)
One of the most powerful analogies for understanding how the housing market works is the children’s game of musical chairs. I’ve invoked this analogy many times since first encountering it in a 2017 viral video by Sightline Institute. You know the game: participants dance in a circle to music until the music stops, and then they rush to find a chair and sit down. The catch: there are more players than chairs. So, inevitably, someone loses.
A housing market characterized by a shortage of homes relative to the demand to live somewhere—which is the case in a large number of American cities and neighborhoods—works something like this. People move a lot in their lives. We move because of new jobs or changed relationships. We move for better rent, a better location, because we want more space, because we can, sometimes because we must. In any case, people are moving all the time, and in aggregate, these moves play out like a giant game of musical chairs in which you win a desirable “chair” (home) not by being fast, but by being able to spend more money than others who want it.
The “musical chairs” aspect of the market means something else important. Every time a player sits in a new chair, they vacate an old one. Every time someone moves into a new home, they leave an old home available for someone else to move into. That person, in turn, must have moved out of somewhere. And so on.
This sequence of associated moves is what housing scholars call a “migration chain.” These chains quickly come to connect rich and poor households, as well as people in vastly different neighborhoods. You can think of it kind of like the famous Six Degrees of Separation theory, a mathematical intuition which says that any human on earth is connected to any other through a chain of no more than six acquaintances.
The crucial implication is that more housing options for anyone means, in some sense, more housing options for everyone. You can benefit directly from the creation of a home you’re never going to live in, thanks to the ripple effect that begins with the addition of a “chair” to the game of musical chairs.
New Research Bolsters an Intuitive Model
A mounting body of empirical evidence captures this basic aspect of the housing market “in the wild,” so to speak. As the number of such studies increases, the methodology has also gotten more rigorous and comprehensive. The latest such study, out of Sweden, is the most impressive one yet.
You can find a summary and a link to the paper, by Gabriella Kindström and Che-Yuan Liang, here. Some of its findings and graphs are also well summarized in this X/Twitter thread by urban economist Stephen Hoskins.
🚨⛓️ It's Filtering Friday! 🚛🚨
— Stephen Hoskins 🔰🏗️🧦🪩 (@GeorgistSteve) December 30, 2023
New working paper observes the moving chains of all Swedes 1990-2017 & finds:
🏗️ New buildings trigger moving chains
💸 Which quickly benefit poor people
🌇 Rental apartments filter fastest
🏘️ Even luxury housing helps poor people get larger homes pic.twitter.com/JZ40CPEYlb
The central finding, one previously reached by studies in the U.S. and Finland, is that new market-rate housing construction triggers a migration chain which quickly reaches low-income households. This is true even when the initial occupants of brand-new buildings have well-above-average incomes. In Kindström and Liang’s study, the initial occupants of new buildings had incomes 127% of Sweden’s average. But by the third round of the migration chain, the average income had halved to just over 60%. In other words, a new, relatively high-end housing unit in Sweden triggers a chain of moves which, in just a few steps, results in a significantly lower-income household being able to move into a vacated home.
In fact, when we look at the incomes of those who moved in each round of the moving chain, we find that as early as the first round after the initial moves, those in the poorest 25% of Swedish households represent a plurality of households moving to new homes. This is true even though the largest share of in-movers to the new buildings themselves (round zero) come from the highest income quartile: almost 40%.
In plain language: building housing that will be occupied by (relatively) rich people quickly frees up different housing for (relatively) poor people.
Kindström and Liang also find that the mix of incomes in Swedish apartment buildings tends to even out over time: new buildings are more expensive, while 30-year-old buildings are occupied by a representative mix of households across the income spectrum. And they produce data suggesting that in Swedish cities which build more housing, households across the entire income spectrum live in higher-quality housing (measured in terms of available space in square feet).
Kindström and Liang do contextualize their findings in some of the specific features of the Swedish housing system, including widespread rent regulation, a large stock of social housing, and relatively mild income inequality. It is likely, for example, that in costly U.S. cities, the incomes of those moving into brand-new buildings are much higher than a mere 27% above national average, and that might affect how the migration chain plays out. But the general slope of the migration chain toward lower-income households is consistent with what U.S. studies find.
The Swedish study is notably different from previous studies of housing migration chains in two ways. One, the data set is huge: the entire Swedish population from 1990 to 2017.
Two, this data makes it possible to look at the incomes of the actual households that moved into and out of buildings. The most-cited migration chain study in the U.S. is a 2019 one by Evan Mast, conducted using Postal Service change-of-address data. But Mast’s study is only able to demonstrate that the migration chains created by the construction of a new building quickly reaches low-income neighborhoods. In theory, this does not prove that low-income residents of those neighborhoods were the ones actually able to trade up to better housing or locations.
Why Do So Many Find This Evidence Unconvincing?
The Swedish study makes bold and excessively provocative claims: “New housing for the rich benefits the poor.” “New homes trickle down to the poor.” “It is not important to build homes directly affordable for low-income residents; they will reap the benefits of more housing space anyway through ripple effects.”
The use of the words “trickle down” is unfortunate here, at least to my American ears. I’m not sure how it’s perceived in Sweden, but here, the phrase is indelibly associated with a specific set of Reagan-era fiscal policies known as “trickle-down economics.” It lends this research an ideological association that it does not deserve, one that will likely raise the hackles of many readers who are suspicious of a certain kind of free-market dogma.
To be very clear, the notion that high-end housing benefits lower-end households through migration chains is not the same thing as trickle-down economics: not at all. The latter refers to a set of theories that giving money back to the richest people (typically through tax cuts) will result in their investing in jobs and innovation that produce overall economic growth.
The idea that when new housing is built, the people with the most ability to pay will tend to get their pick of it first, and the people with less ability to pay will get what’s left over, is not trickle-down economics; it is simply an observation about the nature of a market economy. You don’t have to be a hardcore free-market ideologue, either, to recognize the core insight of these migration-chain studies: that when there are new options in a market that has some fluidity (because people do move), those new options ultimately result in more options for everyone than before.
This is not really something that requires belief in a sort of crude market fundamentalism to accept. It’s something that anyone who has not just played musical chairs, but been part of any sort of game or competition involving finite resources, ought to understand. When there is more of that finite resource available, most players, on average, are going to do better.
When a Trend Is Not a Guarantee
There’s something else going on here, though, that I think explains the resistance to this idea. Or at least to its policy implication, which is that we should embrace housing supply—even when it consists of relatively unaffordable new buildings—in order to alleviate cost and scarcity pressures across the entire market spectrum.
It’s something that Strong Towns Founder Chuck Marohn and I have tried to address throughout the writing of Escaping the Housing Trap, our upcoming book (coming spring 2024 from Wiley & Sons) about the Strong Towns response to the North American housing crisis. We worked not just to make abstract observations about the functioning of the housing market as a system, but to acknowledge the very personal and direct ways in which real people think about and relate to the abstraction we call the “housing market.”
For nearly everyone, our relationship to housing is both deeply personal and intensely local. The stakes are about as high as the stakes of anything involving local government or the built environment can be. And many, many people feel a deep sense of precarity when it comes to housing—for themselves, and for the communities they care about.
When faced with precarity, we become risk averse. We don’t want assurances that statistically things will be better for people like us under Policy X or Y. Rather, we want guarantees that our fundamental interests will be safe. People respond better to such guarantees. And evidence that markets work to improve options for people in the abstract—but maybe not for you, or for me, or for your cousin or son or daughter or friend—don’t meet that desire for certainty.
Most people who aren’t housing policy people aren’t that interested in knowing that new housing supply will create lower rents or an available vacancy for someone, somewhere, in theory. They want to know how their own neighborhood or family or community will be affected.
Housing is a complex system, and in a complex system, local effects can and often do run in a contrary direction to global trends. We see this across all sorts of contexts. Sometimes, factories close and workers get laid off even in a good economy. Towns stagnate and decline into poverty even in a region that is prospering; neighborhoods within a given city do the same.
Hyper-local concerns are more likely to animate the risk-averse opponents of new housing development, whereas theoretical or regional concerns are more likely to motivate its supporters. This disconnect is obvious whenever specific development proposals become politically controversial. YIMBYs will come out to public meetings and cite things like the research findings out of Sweden, making an argument about the importance of supply in general terms. It's in fact more common for someone to come out and say, “I want this building to be built because we need housing as a city or as a region,” than to come out and say, “I want this building to be built in my neighborhood because I really look forward to looking at it on my morning commute,” or even because they personally think they might go live there.
While there are those who devote political energies to opposing housing on more abstract grounds, often environmental or a sort of philosophical anti-development stance, these people are few. Far more numerous are those who engage in the political process specifically to say, “I don't want this in my backyard,” and they really do mean “my.” Often, their concerns are essentially individual, quotidian fears: I will struggle to find parking. I will lose peace and quiet. There will be more shade on my vegetable garden in the afternoons.
Under this same umbrella we find concerns about neighborhood gentrification, about the displacement or adverse effect on vulnerable community members that can result from development that contributes to a rapid change in the character of a place. These concerns are not self-centered in the same way as the classic “NIMBYism” of parking concerns and the like, but the shared thread is that they are typically local, not regional or universal in nature. People don’t express alarm that a new “luxury” apartment building on the corner is going to transform their metro area; their protest signs and public testimonies are always about the fear that it’s going to transform their neighborhood in a way that could harm people they care about.
Whether these fears are warranted or not, recognize that they are real and sincere. And recognize that they are not resolved, particularly at the specific, local level, by research like the findings out of Sweden we discussed above.
Such research basically says to the housing insecure, “Statistically, people like you will probably tend to do better” in a world in which that luxury apartment building goes up than in one in which it doesn’t.
Over the past century, it has become increasingly embedded in our assumptions about housing that the role of government is to guarantee stability and permanence in our neighborhoods. You even hear from people who essentially think of their neighborhood’s zoning as a contract: they believe a promise has been made to them that the kinds of buildings around their home (or, implicitly, the kinds of people living in them) will not change, and that if such change is allowed, that the local government has somehow breached that contract.
This is not true in a legal sense. And it's not viable in a practical sense: we cannot have living, thriving cities with neighborhoods that do not change and evolve. We simply cannot. This is a cultural challenge that needs to be overcome.
This is the basic tension. People want guarantees. What the empirical research points to is that the way we guarantee good housing options to the greatest number of people is to have an abundant supply of housing, and to maintain that abundance through a steady rate of new building that keeps up with demand.
Sometimes those new buildings will be unaffordable to many. But the research on migration chains gives us a compelling reason to think that those who cannot immediately live in them will nonetheless benefit, and won’t have to wait decades to do so.
And yet that narrative does not and cannot answer people's hyper-local concerns, which are rooted in the fact that change of any sort brings uncertainty and risk. The central argument we have thus come to make at Strong Towns, and that we make in Escaping the Housing Trap, is that the way we deliver housing abundance and meet those concerns is through broad-based incremental development. It's not enough to count housing units, to say it’s good enough that a lot of apartments are being built somewhere.
To the extent that we can allow new supply to take on a broad range of forms in a very broad range of neighborhoods, that’s what we need. We must ease the barriers to adding inherently less expensive forms of housing such as the missing middle and smaller units in existing neighborhoods, and we must allow small-scale incremental developers into the game. We can achieve the supply surge we need in a way that answers some—not all—of people's concerns about predictability and neighborhood continuity.
We will not and cannot satisfy those who believe they are entitled to an absolute guarantee that their block or neighborhood won't change in a way that makes them uncomfortable. That existing housing won't ever be torn down and replaced, or that the demographics of their neighborhood won't change over time. Certainty cannot be the goal of housing policy. An abundance of options in a fluid market, though, should be a primary goal.
Daniel Herriges has been a regular contributor to Strong Towns since 2015 and is a founding member of the Strong Towns movement. He is the co-author of Escaping the Housing Trap: The Strong Towns Response to the Housing Crisis, with Charles Marohn. Daniel now works as the Policy Director at the Parking Reform Network, an organization which seeks to accelerate the reform of harmful parking policies by educating the public about these policies and serving as a connecting hub for advocates and policy makers. Daniel’s work reflects a lifelong fascination with cities and how they work. When he’s not perusing maps (for work or pleasure), he can be found exploring out-of-the-way neighborhoods on foot or bicycle. Daniel has lived in Northern California and Southwest Florida, and he now resides back in his hometown of St. Paul, Minnesota, along with his wife and two children. Daniel has a Masters in Urban and Regional Planning from the University of Minnesota.