This post originally appeared on Greater Greater Washington, but it's applicable to the San Francisco Bay Area as well. Questions about gentrification, displacement, and local culture are pressing down hard on the low-income communities around the Bay. Our region is unique, but it is not facing these problems alone.

It’s often assumed that new development pushes out poor people. But it seems as though development is actually slowing or halting the shrinkage of poor neighborhoods and drawing new low-income families into the city. At least, that’s what a first analysis of Census data shows between 2009 and 2014.

Gentrification is not one thing [1]. Instead, it is a collection of different changes that happen to a neighborhood, sometimes at the same time.

The first and most obvious is changing demographics: wealthier people come in and fix up buildings, or open higher-end shops and restaurants. Sometimes, white people move into a historically majority-minority or immigrant neighborhood. This change doesn’t necessarily lead to displacement if there are enough affordable homes and rents don’t rise.

There also can be indirect residential displacement, which occurs when home or rental prices go up so much that new poor people can’t move in, but not so much that the existing poor residents have to leave. Some changes can lead to direct residential displacement, where poorer people are pushed out due to rising rents, taxes, or redevelopment evictions.

Of these, direct residential displacement is the worst. It robs people of their homes through no fault of their own, tears at neighborhood cohesion, and, given the racial element often present in gentrification, reinforces segregation and oppression. The insidiousness of direct displacement should make us careful not to confuse it with other kinds of gentrification, or we may prescribe the wrong policies to prevent or solve it.

Building more homes mitigates direct displacement, says data

Given the significant amount of housing growth in DC [2], I wanted to test the effect that building more housing has on displacement. Using the 5-year American Community Survey data for DC’s Census tracts [3], I found that two things are associated with the departure of low-income people from a neighborhood: the density of the neighborhood (the denser the place in 2009, the more likely people were to leave), and the level of poverty in a neighborhood (the higher the poverty rate in 2009, the more likely poor people were to leave).

There are many reasons why this might be the case. Impoverished neighborhoods might shrink because of low quality-of-life or low income mobility. Dense places might see their low-income populations leave because demand to live there outstrips new housing supply.

However, my analysis also found that when there was new development in a neighborhood, more low-income people moved into the neighborhood. New homes seem to stabilize neighborhoods that might otherwise shrink due to too-high poverty or being more popular because of higher density.

However, because median rents rise when higher-rent homes come on the market - those new homes skew the data upward, after all - some housing activists across the United States wring their hands at the idea of new homes. Activists in San Francisco especially have fought tooth-and-nail against any new housing [4], especially luxury housing [5], in the face of a mounting displacement crisis in the city and region.

How this looks on the ground

This cause-and-effect is not clear-cut, of course, but the balance of evidence points to new housing working more often than it does not. Three neighborhoods exemplify the three situations: dense, wealthy areas without much new housing; poor areas without much housing; and areas with new housing.

Census tract 42.01. All images from the US Census Bureau. Image by US Census Bureau.

One such dense area (tract 42.01) is on the western edge of U Street NW, bounded by Florida Avenue, 16th Street, S Street, and 18th Street. In 2009, it was the fourth-densest place in the city. Though it had a poverty rate of about 16 percent (slightly lower than the District average of 18 percent), its median household income of $90,000 placed it well above the District median.

Still, it saw minimal housing construction through 2014. Though its population swelled by 7 percent during that time, its low-income population dropped by almost 61 percent. Today, its poverty rate is 5 percent. This would suggest that newer, wealthier residents outbid lower income residents for the limited amount of homes, forcing rents up and pushing lower income residents out.

Kenilworth/Eastland Gardens (tract 96.01) is a good example of a poor neighborhood that lost people. In 2009, almost half its residents lived in poverty. Median household income was just $22,300 per year. The total number of units didn’t budge, and neither did rents. Still, the neighborhood saw a huge drop in its low-income population: one third of low-income residents, or about 360 people, left. The overall population dropped by less, just one fifth. This confirms that stable, affordable rents in a low-income neighborhood do not alone make it livable for its residents.

By contrast, West End/western Dupont Circle (tract 55) saw a veritable housing boom. Between 2009 and 2014, it added over 1,000 homes and saw its population grow by almost a quarter. A not-insignificant proportion of that growth, about 10 percent, came from new low-income residents. The low-income population actually grew in an area many see as entirely unaffordable. Though the demographics changed and skewed more wealthy, low-income people were not left behind.

Results like this are a cautious endorsement of the idea that to fight direct and indirect displacement, we need to build more homes.

Of course, these results are still preliminary. First of all, the analysis shows that the number of people paying more than a third of their income in rent increases as the number of new homes increases. Though this might be new college graduates stretching their income for a studio, it might also be due to impoverished residents choosing to stay despite rising rents. Second, using my methodology there is no way of knowing whether a poor resident may move only to be replaced by another poor resident, hiding that kind of displacement in the data.

Additional research, which I plan to publish alongside a Cornell PhD student, will refine these models and look deeper into DC’s history. We would also like to analyze eviction data to follow research from Toronto [6], which found that evictions precede big demographic changes. In other words, we want to find out if direct displacement in the decade or two leading up to DC’s housing boom.

Still, the data is a nod in the direction of encouraging new homebuilding as a major part of the toolkit policymakers deploy when attempting to address housing inequities and the threat of gentrification-driven displacement.

Works Cited

[1] Jonathan Neeley, “Gentrification. What Does That Word Even Mean?!,” Greater Greater Washington, June 29, 2017.

[2] John Ricco, “DC’s Apartment Boom Continued in 2016. Here’s What That Means for Your Rent.,” Greater Greater Washington, February 3, 2017.

[3] “2009 American Community Survey 5-Year Estimates,” American Community Survey (Washington, DC: U.S. Census Bureau, 2010); “2014 American Community Survey 5-Year Estimates,” American Community Survey (Washington, DC: U.S. Census Bureau, 2015).

[4] Noah Arroyo, “Proposition I: Mission District Housing Moratorium,” San Francisco Public Press, October 5, 2015.

[5] John Wildermuth, “8 Washington Condo Project Loses Big in S.F.,” SFGate, September 6, 2013.

[6] Antony Chum, “The Impact of Gentrification on Residential Evictions,” Urban Geography 36, no. 7 (October 3, 2015): 1083–98, doi:10.1080/02723638.2015.1049480.