The “induced demand“ case against YIMBYism is wrong
An empirically false, wildly illogical objection to housing abundance
Today I want to talk about an idea that I don’t think has much purchase in real-world politics but seems to loom large both on the internet and in academic circles — the idea that new real estate development activity actually causes gentrification through some kind of induced demand phenomenon.
Nathan Robinson in his anti-YIMBY screed explores a scenario in which replacing smaller buildings with larger ones increases housing scarcity even while adding units because “we're luring rich people from elsewhere to our city.” He’s doing a thought-experiment, but clearly one he thinks is plausible, while Amee Chew in a 2018 Shelterforce article states that “numerous studies show that market-rate housing development has price ripple effects on surrounding neighborhoods, driving up rents and increasing the burden on lower-income households.”
Chew’s studies don’t really say that (more on this later) but the studies do say what I also know from personal experience which is that people living through a gentrification process often say this is what’s going on. This is all exacerbated by the fact that many people, as in this recent Washington Post article about Mt Pleasant, define gentrification as the presence of new-built condos with chain store retail on the ground floor. And it is definitely true that if what you mean by “gentrification” is “new buildings and chain stores” then blocking new buildings and chain stores will block gentrification.
But the argument here is supposed to be about prices, displacement, and ultimately people’s living standards.
And the induced demand objection fails on four scores:
It is empirically false, at least most of the time.
Accepting its logic would counsel against all efforts to improve quality of life.
If it were true, it still wouldn’t follow that new construction is bad.
It misconstrues what the YIMBY proposal is in the first place.
As I say, I don’t really believe that people believing in the induced demand argument explains any of the proximate barriers to housing reform. But it does come up a lot online, so I’d like to have a resource for explaining to people what’s wrong with it.
The induced demand intuition
The WMATA Green Line was extended to Columbia Heights in 1999, which meant that by 2003 a young person such as myself looking for a cheap place to live in a walkable neighborhood with good access to mass transit might pick it blindly off Google Maps and Craigslist as a good place to rent a basement apartment.
It was actually a fairly dismal neighborhood at the time, largely because the immediate vicinity of the metro station was a bunch of unsightly vacant lots and construction sites. But if you went to the no-longer-extant Columbia Heights Coffee Shop, all the talk among the gentrifiers was “there’s going to be a Target soon.” And, indeed, a few years later the construction sites were transformed into apartment buildings and retail, so if you go today, the immediate environs of the Metro station feature a Giant, a Target, a Best Buy, a Wawa, a Starbucks, and a bunch of other national chains along with some locally owned businesses. Then the slightly less proximate commercial areas either north on 14th Street or over on 11th Street are thriving ecologies of local businesses.
There is clearly some kind of flywheel effect whereby the new retail amenities make the neighborhood a better place to live, which attracts more (and more affluent) residents, which in turn makes it a more attractive place to locate your business.
Now I think it’s obvious that the start of the Columbia Heights flywheel was the opening of the Green Line, especially given the largest context of falling urban crime during this period. The reason you could find white people in a coffee shop talking about how they were looking forward to the Target opening is that the gentrification cycle was already underway. Neighborhoods normally experience an initial hipster phase of gentrification (typified by dive bars, yoga studios, and independent coffee shops) before the major real estate developments and national chains arrive.
But we can argue intuitions all day — the real question is research.
Empirically, more supply lowers prices
Back when I wrote The Rent Is Too Damn High, I said that the impact of new development on hyper-local prices was theoretically ambiguous and the important thing was to look at city-wide effects that are not.
Fortunately, the past few years have seen a good amount of empirical work on this question. And it turns out that in most cases studied, new units reduce prices even on a very local scale.
Kate Pennington’s recent study of San Francisco is very precise: “I find that rents fall by 2% for parcels within 100m of new construction. Renters’ risk of being displaced to a lower-income neighborhood falls by 17%. Both effects decay linearly to zero within 1.5km.”
Xiaodi Li looked at New York: “For every 10% increase in the housing stock, rents decrease 1% and sales prices also decrease within 500 feet.”
Brian Asquith, Evan Mast, and Davin Reed look specifically at new market-rate housing in low-income neighborhoods in eleven cities and find: “New buildings decrease nearby rents by 5 to 7 percent relative to locations slightly farther away or developed later, and they increase in-migration from low-income areas.”
There are some more studies along these lines, but I don’t want to inundate you with them because I do agree that the impact is theoretically ambiguous and someday, someone, somewhere will find the opposite result.
It is striking, though, that induced demand theory doesn’t yet have its Card & Kruger paper that changed the game on the minimum wage. There is no Arin Dube of induced demand theory who’s published a series of rigorous empirical papers demonstrating that there’s something weird in this market. Instead you get this from Chew:
Studies show that market-rate housing development is linked to the mass displacement of neighboring low-income residents (Davidson and Lees 2005, 2010; Pearsall 2010). Such displacement occurs even when low-income housing is not directly demolished and destroyed to make way for new development—because it operates through indirect and exclusionary means, such as “price shadowing” (Davidson and Lees 2005, 2010). Market-rate housing production causes significant price impacts in surrounding neighborhoods, raising area rents and real estate taxes (Oliva 2006; Pearsall 2010; Zuk and Chapple 2016).
So I checked the links.
The two Davidson and Lees papers are concerned with a definitional question in urban geography. According to them, the traditional meaning of gentrification in the literature is something like “rich people move into old homes and renovate them.” They argue that we should extend the definition to include things like new mixed-use housing and retail developments in underused industrial spaces. They assert but do not demonstrate that “price shadowing” occurs in these cases, and then argue that this sort of phenomenon ought to count as gentrification.
Zuk and Chapple actually say: “At the regional level, both market-rate and subsidized housing reduce displacement pressures, but subsidized housing has over double the impact of market-rate units.”
Pearsall says that cleaning up brownfields raises prices in nearby Census tracts, which is interesting and raises the question of whether “make people live near toxic waste” is really the affordable housing strategy we want.
Last we get to Oliva, who at last actually finds induced demand. He looks at the Inner Harbor project in Baltimore which transformed a largely derelict stretch of waterfront into a tourist attraction featuring a convention center, an amazing aquarium, a cool park, and a bunch of hotels and restaurants. Note that in this case promoting economic development was the intended goal of the initiative (Baltimore is poor) and the point of his study is that it was a limited success story — home values did rise, but “this impact has been far more pronounced on the prices of properties located within a short distance from the water even decades after the initial projects on the waterfront were started.”
This is all good empirical work. But note that I don’t think anyone has ever argued that building an aquarium will reduce housing cost burdens. The Inner Harbor is (by design) a tourist attraction. A couple of weeks ago, I took my kid to the Maryland Zoo in Baltimore, then we drove down to the Inner Harbor, got some takeout from Shake Shack, walked around waterfront, and bought a couple of bottles of water somewhere. This is what a successful tourist attraction is supposed to accomplish — you bring in retail dollars and tax revenue. We’ll come back post-pandemic and see the aquarium. It’s a nice place to visit. And Oliva shows that if you build a nice place to visit in the middle of a city, then the immediately adjacent homes become more desirable. Similarly, per Pearsall, if you clean up a brownfield the adjacent homes become more desirable.
New infill housing appears to empirically neutralize the induced demand by simultaneously inducing supply. But conceptually there’s a question here — is it bad to make neighborhoods better places to live?
Better neighborhoods are better
Here’s the problem. Consider the following argument:
Constructing new buildings will bring new retail amenities to the neighborhood.
New amenities will make the neighborhood a more attractive place to live.
Because the neighborhood is now more attractive, prices will rise.
Rising prices will displace some existing residents.
Therefore we shouldn’t allow new buildings to be constructed.
In place of (1) you could put all kinds of things:
Constructing a new park
Reducing the crime rate
Renovating the local high school
Improving bus service to downtown
Fixing the potholes
In other words, you could imagine this kind of logic becoming an infinite cycle of bad urban policy. Defund police will lead to more murders? Well, that’s good for housing affordability. Contractors are spending tons of money on giant subway stations for no good reason? Well, that’s good because it creates construction jobs without the gentrification-inducing impact of improving mass transit service.
This is all wrong, wrong, wrong. We simply cannot accept “make sure the neighborhood sucks” as our affordable housing strategy. How do you improve transportation access to a neighborhood without displacing people? You need to expand the housing supply. How do you improve schools without displacing people? You need to expand the housing supply. How do you reduce crime without displacing people? You need to expand the housing supply. How do you improve the retail amenities in Baltimore without displacing people? Well, you need to expand the housing supply!
I don’t think it’s realistic to promise the Xiaodi Li finding of price declines within a 500-foot radius (that’s really small) everywhere you look. But just logically, there is a finite number of people who would like to live in Baltimore. It’s just not possible that a broad upzoning of the city will, by sparking an improvement in the downtown retail mix, generate a citywide infinite upward price escalator.
The logic of the price escalator
That’s fine for Baltimore, but what about New York? Maybe if you tear down all the brownstones in Brownstone Brooklyn and replace them with large modern apartment towers they’ll all be bought up by Russian billionaires seeking second homes or laundering money.
One question to ask is why exactly is this bad.
Suppose it were the case that upzoning leads to new construction which leads to new amenities which leads to higher prices which leads to even more construction. In that case, upzoning is generating a ton of new construction jobs and property tax revenue. Admittedly right now, new construction doesn’t contribute that much to the tax base because of tax credits. But as the pace of construction increased driving ever-higher demand with it, you could start ratcheting up the taxes. With revenue flowing in, you could cut sales tax, build new subway lines, give teachers raises, turn the parks into outdoor palaces, and otherwise make all your civic dreams come true.
Now is it actually true that cities can unleash an infinite torrent of construction jobs and tax revenue by upzoning? Almost certainly not. But should cities try? Of course they should!
The right model for thinking about this is industrial policy rather than housing. There are communities where people work in factories building mobile homes. Since the homes come off the assembly line and get shipped out of town, it’s true that the mobile home factory doesn’t increase local housing supply and improve affordability. Instead, it has all the normal benefits of any kind of factory whether it makes cars or jet engines or computer chips or refrigerators — it provides blue-collar jobs and tax revenue.
If for some reason people wanted to buy the mobile homes and then instead of living in them, stack them up in huge vertical towers near to the factory, that wouldn’t change the economics at all — it’s still a great source of jobs and tax revenue.
Build away.
What YIMBY is
Last but by no means least, this entire conversation is based on a misconception about what the YIMBY proposal is.
The way high-cost metro areas currently work is this:
No building is allowed in desirable inner-ring suburbs.
No building is allowed in the most expensive parts of the city.
Along a relatively tiny gentrification frontier, a handful of politically well-connected developers can get permits for projects.
Because the permitting process is highly politicized, various activists, neighbors, and politicians try to shake the developer down for side-concessions.
Prices go up and up because this is just way too little new housing to meet demand.
Because leftists like “activists” and don’t like profit-seeking business people, they focus on (4) and see the YIMBYs as siding with the bad capitalists against the virtuous activists.
The actual YIMBY proposal, however, is to change steps 1-3 of the process so that by-right construction is possible throughout the in-demand parts of the city. Actual developers do not lobby for this change, because even though the developers like to beat the activists in step (4), the whole market niche the developers occupy is built around their expertise in navigating the permitting process. You don’t need a “developer” to convert your garage into an ADU, you need a contractor. And if you drive out to a rural area, there’s no such thing as a “developer;” there are just housebuilders.
The YIMBY is anti-anti-developer in the sense that he does not think that obsessively harping on the bad thing that Developer X did with regard to Project Y is a constructive approach to the housing supply. But YIMBY is not about individual projects, neighborhood-level price impacts, or siding with developers. It’s about a systematic approach to making decisions about housing policy, regional housing abundance, and creating a construction boom that makes The Developer a much less relevant character to metropolitan housing supply.
And last but by no means least, while YIMBY is about housing costs to the extent that it addresses the absurdity that middle-class teachers in Silicon Valley can’t afford to live anywhere near their schools, it’s not a comprehensive solution to poverty.
Poor people need more money
Poor people struggle to afford housing for the same reason they struggle to afford shoes and laptops and yoga classes — they don’t have any money.
YIMBY can ameliorate housing cost burdens by making the housing market more like the market for shoes and laptops and yoga classes — a market where typical households don’t have a big problem affording basic versions of the thing. That’s how the market for housing works in Buffalo and Boise and Bentonville and it can work like that in Boston, too. But YIMBY can’t abolish poverty. For that, low-income people need more financial assistance — a more expansive welfare state — which I strongly support, along with every sensible person.
But, again, everything is like this. Better schools don’t abolish poverty. Nor do nicer parks or safer streets or having a cool aquarium downtown. But refusing to relax building restrictions doesn’t magically conjure a more expansive welfare state into existence. On the contrary, crippling the economy makes welfare state expansion harder. And the theory that we should try to keep neighborhoods affordable by having them full of dilapidated buildings with no retail amenities is morally deranged. Slums full of concentrated poverty are bad (something everyone used to acknowledge until a surge in demand for perverse anti-housing arguments) and having new businesses open up is good, just like it was good to build the Green Line. What’s bad is when you don’t allow the housing supply to expand, which would allow more people to enjoy the benefits.
Honestly the whole "YIMBY policy causes induced demand and so would backfire" is one of the most egregious examples I've seen from the left side of the spectrum of starting with a conclusion (new housing construction is bad because of gentrification) and then backfilling a justification for it in order to feel like they know what they're talking about. Because in their worldview anything that involves private for-profit enterprises must be treated with suspicion at best and hostility at worst, so anything that whiffs of pro-market ideas must be bad. I think Noah Smith had a good post on this specific strain of left-NIMBYism, which is an interesting/weird part of the broader anti-growth coalition with right-NIMBYism and "Left"-NIMBYism of wealthy areas and suburbs of progressive cities that's basically just right-NIMBYism with a blue-colored mask.
I think a lot of this dovetails nicely with yesterday's post about how the public is largely just small-c conservative: people just tend not to like change, and will come up with reasons why ex post facto. I've certainly been guilty of this in the past myself on certain things, for what it's worth. I split my time between two large, famously progressive, and famously unaffordable West Coast metro areas that attract people largely for lifestyle reasons (my family lives in one, and I go to college in the other), and it's shocking how much of the complaints I see about new developments in the city boil down to some variation of "I liked this place more when it sucked, we should make things shittier so people stop wanting to come here", in which case...why do anything to improve the place, or any place? Shouldn't we try to find a way to make things better for everyone, or barring that, as many as possible?
Great piece. Part of the confusion in the discourse comes from two distinct problems contained within the phrase "housing shortage". The economic problem is that we need to increase the supply of homes in high-demand cities to reduce prices and rents. But the policy problem is that we need to change how our urban planning institutions are designed and function to end a systemic shortage, which is causing that economic problem.
This is why that, while it is true that building more homes in a neighbourhood *does* decrease local housing costs as the papers you linked showed, it doesn't *feel* like that to the majority of people. And that's because the systemic shortage - the rationing of development by inflexible zoning and discretionary, case-by-case permits - isn't changed if permits for one house, ten houses, or a thousand houses are handed out by local government.
"Shortage" in this housing policy problem sense is much closer to the permanent "shortage economy" of the former Eastern Bloc than the typical sense in which economists use "shortage" to mean a temporary blip of undersupply as prices and quantities adjust.
In, say, 1980s East Germany, there was a systemic shortage of cars because production was rationed by planners' control of permits. There was "induced demand" for cars, in that every car which was built was immediately sold, and there were no unsold or "vacant" cars which as an surplus put pressure on sellers to produce and sell cars quickly and cheaply to a high standard. Second hand cars traded at the same prices as brand new cars and as assets could be treated as dependable stores of wealth, as it seemed no matter how many cars were made, the shortage of cars did not end.
Building extra cars within this systemic shortage did make East Germans better off. Stopping or reducing the production of cars because the cars only go to better off/politically dependable groups would have been bad, and made everyone worse off. But the systemic shortage of cars was only ended when the institutional framework which rationed the production of cars through case-by-case permits was dismantled. East Germans were made far better off when they were able to participate in an economy where car companies could produce as many cars as they liked, and which acquired all of their inputs (steel, tyres etc) through simple purchases rather than permits. The economist János' Kornai's work on these shortage economies explains all of this in really clear language.
For housing, the policy problem is the same. Institutionally, we need far more "by-right" urban planning, where developers who propose something that complies with the rules legally must be granted permission to build, rather than jumping through dozens of hoops. These rules then need to be really generous and allow lots of things to be built, not just single family homes. This is how you permanently improve the long-term outcomes for everyone, by changing the institutional conditions to permanently increase total production, while haggling over whether to build 10 or 15 apartments on a site only makes a tiny and temporary improvement.