New Market-Rate Housing — It Lowers the Rent

May 28, 2021
View of blue trimmed apartment building

The W.E. Upjohn Institute is back with a paper studying the impacts of new market-rate housing development on local rents in low-income neighborhoods. Long-time readers of this newsletter may recall when we covered the working paper by Asquith, Mast, & Reed (2021) here, but now that it has gone through peer review, its results are more compelling than ever.

Key takeaways:

  1. “New buildings decrease rents in nearby units by about 6 percent relative to units slightly farther away or near sites developed later”— even with migration increasing to high-income areas.
  2. Consistent with Mast’s previous paper, this one finds that as more people move into new housing and vacate older housing stock, the availability of housing in the area actually increases.
  3. The “amenity effects” of new housing, i.e. bringing more desirable shops and services to the neighborhood, don’t raise prices more than the increased housing supply lowers them. As a caveat, the authors note that these effects “could be limited because most buildings go into already-changing neighborhoods, or buildings could create disamenities such as congestion.”

Does new market-rate housing increase the average price of housing in a neighborhood as wealthier residents move in? Or does increasing supply to meet demand help to cool off high prices? The evidence from this paper supports the hypothesis that new market-rate development can help to ease displacement and gentrification pressures, rather than making them worse.

This is difficult to measure with a reasonable degree of confidence; cities are not labs where all variables can be controlled and randomized. So how did the authors study the problem?

Because developers do not choose to build new apartments in completely random locations, the researchers measured how rents and migration patterns look in areas where a new building is built, compared to analogous areas without new development at the same time. The areas were defined by a set of broadening, concentric circles around neighborhoods with new housing — one circle with a 250 meter radius, and another with a 600 meter radius (about a half mile).

The new buildings studied in 11 cities were selected only if they were “pioneer buildings,” with no other new development within a 250 meter radius, within the sample years of 2015-16 and 2019. Additionally, the authors only selected large market-rate buildings in lower-income urban neighborhoods, where concerns about gentrification are most acute, and only where enough rental listings were available nearby to study the impact on local rents.

In-migration patterns were also observed “using individual address histories from Infutor Data Solutions.” This data, the authors argue, enables them “to study cheaper segments of the market that may be underrepresented in the Zillow data.”

Even though new developments “could theoretically change local amenities or reputation by enough to instead increase demand and raise nearby rents,” the authors observe that “[their] findings suggest this is not the case.” It’s possible that this is because neighborhoods are already changing when new development is proposed, or there are “disamenities” such as increased traffic congestion countervailing the amenity effects.

The evidence in this paper is conclusive: “new buildings lower rents in nearby buildings by 5 to 7 percent relative to trend”—even as the neighborhood becomes more desirable.

“We find that the concerns about rent increases driven by new market-rate housing are mostly unfounded,” the authors argue. However, they caution against extrapolating too much from the results, since the data inherently has a survivorship bias toward neighborhoods where new development is already allowed.

New development that is prevented in higher-income neighborhoods, with more restrictive land-use controls, is harder to study before it even exists. Nevertheless, gentrification concerns over new development generally get the causation backwards. “While there is a strong observed correlation between new construction and rising rents,” the authors observe, “this appears to be because new buildings are typically constructed in areas that are already changing.”