A growing body of evidence supports California YIMBY’s founding goal: more housing for more people means a more inclusive California. If this state is truly open to all, it must permit enough housing for all—regardless of income. A new working paper by Xiaodi Li at NYU isolates and measures the price effects of new market-rate development in New York City and finds that, in fact, pricey new high-rises do bring rents down in the immediately surrounding neighborhood. The big takeaways:
- For every 10% increase in the housing stock, rents decrease by 1% within the 500ft vicinity. (Quick math exercise: what would happen with a 100% supply increase?)
- Market-rate housing does produce “amenity effects” (e.g. new restaurants, light poles, you name it) but this does not raise prices nearly as much as the supply increase lowers them.
- New market-rate housing typically follows price increases, rather than preceding them. But new luxury housing can decrease rents even for middle-class housing nearby.
Most growing American cities are undergoing dramatic, painful changes that reflect an intuitive narrative: higher-income job growth attracts new residents to poor, disinvested urban neighborhoods, and when new housing is built for them, rent keeps rising, charcuterie boards replace taco trucks, rinse and repeat. In a word, gentrification. While this reflects lived experiences of displacement and instability, empirical research has managed to parse complex variables into a causal framework that is less intuitive when everything seems to be happening all at once.
Most research so far has found evidence for new housing supply reducing rents in a broader regional housing market, but data on local impacts in the neighborhood where the new supply is built has been scarce. This has been harder to research for a few reasons: as noted, new housing isn’t simply built in a vacuum. Developers typically follow patterns of high-income demand, as Boustan et al (2019) found for new condominiums in central locations. However, as these researchers noted, this development pattern is endogenous—there was already demand from high-income prospective residents drawn to the area. Amenity effects are also complicated. Research on Low Income Housing Tax Credit (LIHTC) developments found that middle-income households raised property values when they moved to new LIHTC housing in lower-income neighborhoods, and lowered them in higher-income neighborhoods. Not surprisingly, renters can see new development as a threat, even when they recognize that new supply is needed.
Li had to develop some novel methods to crack this chicken-and-egg conundrum.
First, Li gathered data on new high-rise construction, rent and sales prices, and new restaurants from 2003-2013 in New York City. Li controls for the endogeneity of price signals that precede new development by also including building permit approvals in the data, which correlates more closely with rapid home value appreciation than building completion. Many buildings simply take longer to build than planned due to a variety of unforeseen circumstances, so their impacts on the local housing market may be delayed. Li also restricts measurement to rentals within 500 feet of new high rises, since the data unsurprisingly shows faster price appreciation in neighborhoods where developers choose to build.
With all these controls in place, the data is illuminating. Per the initial regression results: “Rents for rental buildings within 500 feet of completed new high-rises decrease by 1.6% one year after the completion significantly and persistently.” More surprisingly, this result is much less prevalent 1000 feet away, suggesting that new market-rate housing lowers rents locally far more than elsewhere.
It’s not just a one-time effect, either. Year over year, rents within a 500-foot radius of new high rises continued to decrease, a trend positively correlated with the amount of new housing built. Overall, a 10% increase in local housing stock resulted in a 1% annual rent decrease within the 500-foot buffer zone—with the caveat that this effect was much smaller the closer the building was to the Empire State Building. In the outer boroughs, rents decreased as much as 2% for every 10% increase in housing.
Moreover, Li finds that filtering is real: these new high-rises decreased rents for mid-tier rental housing within the same census tract. We’ve previously covered research from Evan Mast at the Upjohn Institute attempting to model how this works: as higher-income households move into new luxury housing, they vacate older housing that middle-class households can move into. In the medium-term, this filtering effect doesn’t extend to low-end housing, which is why there is a broad economic consensus that subsidies are essential for meeting the housing needs of poor households.
Finally, Li considers one alternative explanation for this effect: do new high rises simply make a neighborhood less desirable? According to the author, the data shows that “blocked views do not explain this negative impact for mid-rise and low-rise rental buildings.”
Sometimes, true beauty lies in taking the long view.