Mini-review: Chapters 2 and 3 of Wonder City

In Chapters 2 and 3, Ellsworth tries to argue for supply skepticism- that is, the idea that new housing (or at least the high-end towers that she opposes)* will not reduce rents or housing costs. She has made some effort to review both academic and non-academic writing on these issues. However, her grasp of her own sources is a bit uneven.

She begins by addressing pro-housing studies: studies by Evan Mast, Xioadi Li, and Kate Pennington showing that new buildings reduce housing costs near them.** Ellsworth correctly points out that these studies show fairly small reductions in rent, and do not affect citywide rents. But I think the authors of these articles would agree: they are not trying to claim that a few towers here and there would reduce rents throughout a metro area, just that they would not create destabilizing rent increases in their own neighborhood. So these articles aren’t really relevant to the broader issue that separates Ellsworth from YIMBYs- that is, whether a truly significant level of new construction would hold down rents citywide.

What does Ellsworth cite to show that new housing is useless or harmful?  First, she refers without elaboration to studies by “Freeman, Angotti, and Rypkema.”  (p. 57) Her bibliography includes a study by Lance Freeman about gentrification. This article seems to be about the relationship between gentrification and displacement, rather than the relationship between new housing and either- so I do not know why Ellsworth cites it. (Maybe she discussed it in an earlier draft, and forgot to delete the reference here).

Her reference to Rypkema is apparently a reference to a report by Rypkema’s firm Place Economics defending historic preservation.  She describes the report as follows: “when eighteen new skyscrapers were built in the study area, the population of poor and working-class white and Black families declined- by a lot.” (p. 73)   I think she is referring to a bar graph on page 16 of the report.  The bar graph shows that before the skyscrapers, somewhere between 30 and 40 percent of the population in the block groups containing the skyscrapers earned under $50,000.  (Because the report used a bar graph and not a table, the report does not contain an exact percentage) .  After the skyscrapers were built, that percentage declined to somewhere between 20 and 30 percent. 

Does that mean the population of under 50k-residents declined? Absolutely not. because if the overall population increased, it is quite possible that the number of under $50k households increased, but that the number of richer households increased by even more.  

More broadly, Rypkema rejects Ellsworth’s concerns that New York is becoming too dense.  He writes: “There is little doubt that New York has to add density and that skyscrapers are one of the ways to do that.” (p. 19).  Even though Rypkema agrees with Ellsworth’s vigorous support of historic preservation, he was apparently not an opponent of new housing supply.

Ellsworth also relies on work by Angotti and Maxwell Cabello which (according to her) claims that in Greenpoint and Williamsburg, “where Mayor Bloomberg upzoned and unleashed massive amounts of new construction of luxury housing” (p. 59), black and Latino population declined. 

In fact, the Cabello report does say this—but the executive summary also points out that “areas with wealthy white populations were down-zoned to lower residential densities” (Executive Summary).  In a section on Park Slope, for example, large chunks of Park Slope were rezoned “to prevent out-of-context development and to protect the historic low-rise character of the neighborhood.”   The report itself states that “by down-zoning the affluent, white area of the neighborhood, they created a limited supply housing market that pushed out many low-income residents who lived there…” In other words, rents rose not because there was too much housing but because there was too little.   

Ellsworth cites Jersey City as another example of what not to do, claiming that new housing there led to higher prices (pages 59-60).In support of this, she mentions an 2022 website post by journalist Mary Sparago claiming that Jersey City is “the most expensive city for renters”  and claiming that the average rent there was $5500. I did a Zillow search to verify this, and found that of over 3900 rentals listed on Zilow, only 258 (less than 10 percent) were for more than $5500.  By contrast, in New York City, out of just over 17,000 rentals, 4176 (or about 24 percent) were for more than $5500.  In Manhattan, the percentage rose to over 40 percent (2989 out of 7363). 

I do think Ellsworth has accurately described an article by Professors Andres Rodrieguz-Pose and Michael Storper.  They argue that “the structure of jobs and incomes is what drives housing prices.” (p. 64) – in other words, that demand affects housing costs. 

Professors Michael Manville, Michael Lens, and Paavo Mondroken have responded to the “demand” argument: “[Rodrieguez-Pose and Storper] show that high-income people move to expensive places, but never explain why removing existing restrictions on housebuilding in those places would not help absorb those people, and alleviate the pressure they put on prices.”  In other words, the law of supply and demand is a law of supply AND DEMAND- so if demand matters, it logically follows that supply should matter as well. Manville et. al. also point out that housing prices may drive “the structure of jobs and incomes” if lower-income residents are priced out, which means Rodriguez-Pose and Storper have it backwards.

Ellsworth also cites Pete Saunders’s claim that “There is no shortage of housing in most US metro areas… there is a shortage of housing in areas most attractive to today’s young and affluent urban planners.” (p. 64)

But Saunders’ position is a bit more nuanced. Saunders writes:  “there is a shred of truth to this in many coastal metros, and in a select group of interior metros like Chicago.  They are, in fact, not adding new housing supply relative to the demand for housing. “ He then adds a chart showing the difference between San Francisco (which adds very little housing) and Houston (which adds a lot). In other words, Saunders seems to think that maybe there IS a housing shortage in New York and San Francisco. Saunders goes on to focus on Chicago, which is a much less expensive region than New York.

Ellsworth then misreads Saunders more aggressively, writing that “He would agree that deregulating the core would, as I have argued above, destroy it, transform into it into Dubai, and is therefore a destructive strategy.” (p. 64).  But in fact, Saunders writes that upzoning “in Chicago, and in many cities around the nation, would result in a spurt of housing growth in the most desirable areas—and in the most desirable areas only… Doing so would result in cities made up of contiguous enclaves of affluent, dense, walkable and self-sufficient communities.  They would be beautiful, and they would be made up of the kind of development that is most preferable to me.  Yet they would be adjacent to but wholly disconnected from half-empty expanses of impoverished, working class and middle class neighborhoods.  I can’t reconcile that in my mind. Do we really want that?  Can’t we do better than that?”   This is not a claim that upzoning would “destroy” downtown Chicago but a claim that the upzoning might not improve the city’s poorer areas.

Ellsworth then cites a more scholarly article about Chicago by Yonah Freemark, which she interprets to mean that “upzoning in Chicago resulting in both displacement and increased prices, not a fall in housing prices.” (p. 74).  But Freemark emphasized, in the very first page of his study, that this particular upzoning led to “no new additional housing construction”.  By contrast, Ellsworth’s primary argument seems to be that new construction won’t lower rents.  So there isn’t really any relationship between Freemark’s article and Ellsworth’s conclusion.

Finally, she cites a 2018 blog post by RentCafe writer Nadia Balint saying that “more than 80 percent of the new rental housing that has been built in this country’s overpriced housing market is ‘high end’ and that filtering failed to supply housing to those most in need.”  (p. 74) But this study doesn’t mean what Ellsworth thinks it means.  If very little housing is built, it doesn’t matter whether 80 percent of it is high end or 100 percent is: 80 percent of almost nothing is still almost nothing. ** Nationally, multifamily construction in 2018 ranged between 316,000 and 415,000 per month- far below the one-million peak level in March 1973.  Moreover, if this percentage correlated with high rents we would find that the highest “high end” percentages were in the most costly cities, such as New York.  In fact, in 2018 New York ranked sixth from the bottom (out of 30 or so cities listed) in “high end” housing, and the highest “high end” shares were in Dallas, Fort Worth and Boston: one city that is costly but less so than in New York, and two that are clearly less costly.

*She seems to be ambivalent about new low-rise housing.  On the one hand, she argues for upzoning outer-borough areas dominated by one- and two-family houses (pp. 312-13); on the other hand, she favors a variety of policies that would make such development more difficult, such as expanded design review and expanding rent control to cover newly built housing (pp. 313-15).