For a long time, I thought gentrification was the hottest of urbanism’s hot-button issues. That may still be true. But it has a new (and related) challenger—upzoning, or changing the zoning of an area to allow for higher density.
For years, some urban economists and market urbanists have been making the case that the key challenge facing cities—especially pricey superstar cities and tech hubs—is a lack of housing supply. There are many culprits in this shortage. They include strict land-use regulations and building codes, politically connected NIMBYs, and other factors, but the end result is the same. A lack of housing supply results not only in higher housing prices, but in increased sorting and displacement, which sharpens inequality and segregation. It even limits innovation and productivity—not just in the affected cities, but across the U.S. economy as a whole.
Two weeks ago, I wrote about a new paper by two economic geographers, Michael Storper and Andrés Rodríguez-Pose, who question some of the theoretical and empirical claims made from this YIMBY perspective (or what they call the “housing as opportunity” school of thought). Specifically, Storper and Rodríguez-Pose argue that simply increasing the supply of housing through upzoning is likely to add more housing for high earners with no evidence that it would “filter” down as cheaper housing for less advantaged residents. That in turn would only exacerbate the ongoing sorting process that draws more educated and advantaged people to affluent cities and pushes the less advantaged out of them. The ultimate result would be even worse spatial inequality between leading and lagging places.
My article on Rodríguez-Pose and Storper’s paper resulted in a near-immediate barrage of critiques on the internet. David Schleicher of Yale Law School and many others laid into the paper on Twitter. Over at City Observatory, Joe Cortright questioned both the theory and the evidence underlying Storper and Rodríguez-Pose’s findings (and, in a side shot, my decision to write about them):
Rodríguez-Pose and Storper sidestep these nuts and bolts issues of how to fix zoning so that it isn’t exclusionary, in favor of a knocking down a straw man claim that upzoning is somehow a cure for inequality, (an argument that no one seems to be making). In the process, they (and by extension, Florida) lend credence to the NIMBY-denialism about the central need to build more housing in our nation’s cities if we’re to do anything to meaningfully address affordability.
(For the record, I have long been a critic of restrictive zoning and building regulations and NIMBYism, going so far as to dub the latter “the New Urban Luddism.”)
On Friday evening, I received an email from three of Storper’s colleagues at UCLA, Michael Manville, Michael Lens, and Paavo Monkkonen. They wrote a detailed essay replying to key claims in the Storper–Rodríguez-Pose paper, and they shared it with numerous scholars and journalists, which sparked another round of discussion online.
Their response begins by saying that although Storper and Rodríguez-Pose are esteemed geographers, their paper “badly misses the mark.” Furthermore: “It ignores much of the research on the topic, misstates or misunderstands the research it does cite, presents misleading and oversimplified analyses, and advances an argument that is internally inconsistent.” Like Cortright, the authors accuse Storper and Rodríguez-Pose of setting up a strawman argument:
For the record, we agree with [Storper and Rodríguez-Pose] that building market rate housing will not by itself eradicate inequality, or revive declining regions. We also agree that building market rate housing will not, by itself, get everyone in expensive regions properly sheltered. But as far as we can tell everyone agrees with that. Many people (us included) think that more housing in expensive places in necessary for fighting inequality and increasing affordability, but no one we are aware of thinks it is sufficient, i.e. that all we need to do is build more housing.
Manville, Lens, and Monkkonen light into the data and methodology that Storper and Rodríguez-Pose used to make their argument. For example, they used the percent change in developed land area in a region as a proxy for regulatory stringency. But Manville, Lens, and Monkkonen note that this metric “cannot discern between an overall regime of light regulation and lots of development, and a regime of heavy regulation that limits development while pushing it outward.”
It also ends up putting cities like San Francisco and Detroit in the same bucket. Detroit didn’t add much new developed land between 1990 and 2010, which is a sign of stagnation. San Francisco also didn’t add much new developed land, but the interpretation is very different: “San Francisco is strictly zoned, but it is also bounded by mountains and ocean—there just isn’t much undeveloped land to build on.” Manville, Lens, and Monkkonen conclude that this measure of regulatory stringency is actually worse than the (admittedly imperfect) alternatives that Rodríguez-Pose and Storper criticize.
They also invoke a number of studies which show that adding housing supply does lead to filtering and better affordability. Although upzoning may have its risks, they write, given the reality of exorbitant housing prices in many cities, it is time to move forward with policies to upzone and add housing, because “the consequences of inaction also matter.”
This is largely a debate about what kind of a good housing is. On the market-urbanist side, there is an idea that a house is basically like a refrigerator (although harder to manufacture), in that there is one basic market. As Manville, Lens, and Monkkonen write:
Rich people also buy refrigerators and groceries and televisions, but these are not noticeably more expensive in Los Angeles or San Francisco than in the rest of the country. One might counter that these goods are easier than housing to produce. But that’s not a counterargument: it is our point. It is the difficulty of producing housing, not the incomes of the people buying it, that makes housing so expensive.
For Rodríguez-Pose and Storper, though, housing in prime neighborhoods in expensive cities is like a special type of fridge that you need a college degree to operate. It’s not just that skilled workers can better afford housing in prime urban locations; the amenity also means more to them.
Storper and Rodríguez-Pose (who declined to comment for this article) suggest that the housing market is more segmented than other markets, and that the crucial issue is really the inequality of labor and wages. Knowledge workers capture a large urban wage premium, while blue-collar manufacturing and service workers fall further behind economically. “In 1950, denser urban areas offered higher wages for both educated and less educated workers,” they write. “Today, when wages are adjusted for density, workers without a college degree have very little advantage from locating in large cities.” So adding more housing in expensive cities, in these scholars’ view, is wholly insufficient to solving the underlying problem of a split labor market and highly unequal wages.
Over on Twitter, Mark Muro of the Brookings Institution and Will Wilkinson of the Niskanen Center picked up on this. Building more housing in superstar cities and tech hubs will not increase migration to thriving cities across the board, but simply cause more knowledge and professional workers to move in, resulting in an even broader gulf between have and have-not regions.
But who's saying that urban upzoning would *reduce* regional inequality? That's what baffles me. Basic economic logic says we should expect restrictive zoning in the most productive labor markets to retard growth in economic concentration/regional inequality.— Will Wilkinson 🌐 (@willwilkinson) May 20, 2019
In fact, in 2017 I wrote about a study by Chang-Tai Hsieh and Enrico Morretti which showed that liberalizing land-use restrictions in pricey cities like New York and San Francisco would lead to substantial increases in employment in these cities with significant productivity gains for the economy as a whole, but significant job losses in Rust Belt metros such as Binghamton, New York, and Muncie, Indiana.
All in all, this is a good and productive debate. I actually think there is more agreement than disagreement among market urbanists and Storper and Rodríguez-Pose. (I often say that urbanism suffers from the narcissism of small differences on steroids.) All agree that more housing has to be built. All agree that adding supply alone will not solve the housing-affordability problem, and that building more must be complemented with more affordable housing. As Yonah Freemark puts it:
Of course, we must recognize — as most YIMBY groups do — that upzoning can only be “one part of the answer,” as Richard Florida argues. When Michael Storper warns that with upzoning, “the market will naturally respond best in areas with the greatest returns on upzoning — mostly places with dense, white-collar employment where high-income people will want to live to be closer to their jobs,” he’s right. Upzoning is a strategy that relies on the market to produce housing. The market is ultimately going to build units that are profitable to a select class of individuals.
Solving the affordable-housing problem is not an either/ or proposition, it is both/and: better jobs, higher wages, more transit, place-based policies to assist lagging cities, and a more robust social-safety net. That’s the inclusive-prosperity agenda we have to work harder to achieve.