Here’s What We Actually Know About Market-Rate Housing Development and Displacement

A stack of books that illustrate how much we actually know about new construction, which is overwhelmingly geared toward the luxury market, and its connection to displacement.
Photo by Evan Bench via flickr, CC BY 2.0

Rent control. It’s on the ballot in California this November [here are some lessons from its lossas tenant campaigns pick up steam across the country and revive an old refrain: “The rent is too damn high!” The real estate industry’s biggest argument in opposition? Rent control will hurt new construction. And, as developers would have us believe, the only way to pull ourselves out of our dire housing shortage would be by building new construction.

But this unquestioning reliance on new construction—a code phrase used by developers to signify for-profit building—is deeply flawed.

For one, for-profit new construction is overwhelmingly geared toward the luxury market. But it’s lower-income households who face the most severe affordable housing shortfalls. While our high-end stock has steadily grown, since 1990 on balance we’ve lost over 2.5 million affordable units renting for under $800. To what? In large part, rent increases.

Secondly, new construction takes decades to depreciate down to rents that are actually affordable to most renters. “Trickle down” isn’t happening fast enough. [See: “Trickle Up Housing: Filtering Does Go Both Ways.”]

Even worse, however, new construction actually fuels displacement in the short term, even when no already existing housing is knocked down. Why? Numerous studies show that market-rate housing development has price ripple effects on surrounding neighborhoods, driving up rents and increasing the burden on lower-income households. Many residents in communities transformed by gentrification can already attest to the connection between for-profit development, rising living costs, and the mass exodus of lower-income residents. Maybe this won’t play out in Malibu, or a sparse neighborhood with very few low-income folk, but otherwise the above effects are widespread in our cities.

We need to talk about market-rate construction, and displacement. Here is the what the research says:

  • Studies show that market-rate housing development is linked to the mass displacement of neighboring low-income residents (Davidson and Lees 2005, 2010; Pearsall 2010). Such displacement occurs even when low-income housing is not directly demolished and destroyed to make way for new development—because it operates through indirect and exclusionary means, such as “price shadowing” (Davidson and Lees 2005, 2010). Market-rate housing production causes significant price impacts in surrounding neighborhoods, raising area rents and real estate taxes (Oliva 2006; Pearsall 2010; Zuk and Chapple 2016). These price impacts have resulted in higher housing cost burdens for low-income residents, as well as their displacement (Davidson and Lees 2005, 2010; Pearsall 2010). In fact, a study of displacement in New York City based on a survey of 18,000 housing units found that most displaced households were forced to move due to cost considerations; in contrast, low-income residents who managed to remain in gentrifying neighborhoods overwhelmingly lived in public housing or rent stabilized units insulated from price dynamics (Newman and Wyly 2006, 29, 41, 43). Rent burdens rose considerably in gentrifying areas, so that only 1 out of 15 poor renters remaining in these New York City neighborhoods rented in the unregulated market (40-1).
  • The influx of higher-income residents, whom market-rate developments are typically geared toward, is itself associated with the displacement of vulnerable groups from the same area. Studies in London, Sydney, and Melbourne using longitudinal census data found that increases in high-income and professional households in a neighborhood were correlated with greater losses or displacement of low-income, family, and working-class households, as well as elderly, disabled, and unemployed residents, from that community (Atkinson 2000a, 2000b; Atkinson et al. 2011). One study found that in neighborhoods with an influx of higher-income residents, working-class residents moved at three times the rate compared to in other areas—and usually out of the neighborhood (Atkinson 2000a, 159).
  • Location matters in predicting the pathway of gentrification. Gentrification is more likely for poor neighborhoods that border rich neighborhoods (Kolko 2007; Guerrieri et al. 2013). A study of over 27 metro regions in the U.S., including Los Angeles, found that out-migration of poor residents and in-migration of richer residents was 64 percent more likely for neighborhoods within half a mile of an existing rich neighborhood, compared to those further from the nearest rich neighborhood (Guerrieri et al. 2013, 59). Again, this is likely due to price effects: housing prices in poor neighborhoods that bordered or were within a mile of rich areas appreciated by a significantly higher amount than prices in poor neighborhoods further away (51, 56). Housing booms do not affect prices in all neighborhoods equally; in fact, poor neighborhoods that start out with low housing prices and are near richer neighborhoods experience the largest price increase effects (46).
  • Unfortunately, in our market-based housing system, proximity to transit stations is a risk factor for gentrification. Numerous studies show that neighborhoods within half a mile of a transit station experience significant housing price and rent increases (Immergluck 2009; Pollack et al. 2010); loss of affordable units (Chapple and Loukaitou-Sideris 2017); increased share of high-income households and decreased share of low-income households (Dominie 2012; Chapple and Loukaitou-Sideris 2017); and increased prices of commercial properties (Weinberger 2001; Debrezion et al. 2007). Moreover, plans for transit investment can drive up property values and housing costs even before construction begins due to real estate speculation, as the plans become known (Knaap et al. 2001; Immergluck 2009).
  • Likewise, new higher-end commercial amenities and big box retailers also add to displacement pressures, again, largely due to the overall marketization of housing in the U.S. and lack of sufficient protections against rising costs. Such commercial development contributes to rising property values, as well as the influx of white and more affluent residents, heightening displacement through competition and rising rents (Zukin 2009). The arrival of large, national retailers has been linked to net job and business loss, as well as decreases in retail wages (Dube et al. 2007). But even smaller-sized yet upscale boutiques contribute to the displacement of local stores and services that long-time, lower-income residents rely on—notwithstanding boutique owners’ purported sensitivity to community identity and racial solidarity (Zukin et al. 2009).
  • Some academic studies have contested whether gentrification in fact causes displacement. However, whether studies detect displacement very much has to do with how they measure, and define, gentrification. For instance, one famous study often cited to prove gentrification does not cause displacement relied on survey data that did not count residents who had doubled-up, moved out of the city, or became homeless (Freeman and Braconi 2004; Newman and Wyly 2006). Even so, though it failed to count the displaced, the study still admits class change was occurring in gentrifying neighborhoods, though if not through direct ‘displacement,’ through ‘replacement’ and probable exclusionary displacement (Freeman and Braconi 2004). And even this study found that gentrification in New York City harmed low-income households by increasing their rent burdens: the researchers reported the average rent burden for poor households in gentrifying areas was 61 percent, compared to 52 percent for poor counterparts in other neighborhoods; and that rents for unregulated apartments in gentrifying neighborhoods increased an average of 43 percent from 1996 to 1999, compared to 11 percent for rent stabilized apartments (50-1). In contrast, a finer analysis of the same New York City survey data by other researchers, that carefully considered place and motive, succeeded in uncovering evidence of gentrification-fueled displacement and migration flows, with rent increases, landlord harassment, and condo conversion emerging as key reasons for moves (Newman and Wyly 2006).
  • Real estate interests and some scholars [and many activistsargue that unaffordable housing costs are primarily due to a shortage in housing supply, and that any increase in supply—including luxury development—will ultimately help depress rents. While there is some evidence new housing production does eventually help lower median rent in the neighborhoods where construction occurred compared to other areas, these effects take decades to surface (Zuk and Chapple 2016; Rosenthal 2014). Worse, by the time such price effects register, large numbers of low-income residents have likely already been pushed out: as one study of construction in the Bay Area found, the increased cost burdens which market-rate production puts on low income residents are far more immediate than any long-term decrease in rents (Zuk and Chapple 2016). And even if median rent is eventually, somewhat, lower than in areas without construction, who is to say that the median rent is actually affordable? In the above study, researchers noted median rents of all areas might still be out of reach for low-income households. During the decades analyzed, significant displacement had already occurred and median rents were hiked up by gentrification. In contrast, the production of subsidized housing had more than double the impact on eventually reducing rents at a regional level, compared to market-rate units. Thus, the production of non-market rate housing matters deeply. [Editor’s note: There is also an argument that production of moderate-income housing is relevant as well.]

In sum, luxury development that centralizes a concentration of higher-income residents in a lower-income surrounding community puts neighboring poor residents at risk of displacement due to the impact on increased living costs. Both luxury development itself and the influx of higher-income residents are linked to higher housing cost burdens for low-income residents, as well as displacement, because of their price effects on the real estate market. Furthermore, even upcoming development can set off real estate speculation and price increases before construction begins. Place matters, and proximity to richer neighborhoods as well as massive capital investment, whether in the form of private development projects or transit infrastructure, are risk factors for gentrification.

Stability for renters should be valued. Housing instability is bad for health and worsens poverty. Even without gentrification, U.S. neighborhoods experience high endemic levels of displacement and eviction when it comes to low-income families, who face dire intergenerational consequences. Gentrification uproots low-income families to relatively far-flung and less-resourced places, with added political, social, and health impacts.

So what is there to do? Rent control must be paired with any strategy of new construction and investment in order to prevent displacement. On a practical level, rent control would stop the hemorrhage of remaining affordable units now—provided allowed rent increases are appropriate to low-income renters’ finances—and include strong protections against eviction and landlord harassment. Rent control would also preserve and potentially even recover the affordability of tens of millions of homes nationally, working on a scale unrivaled by Section 8 vouchers and any new construction.

Construction trickles, but rent control works instantly. Rent control costs the public little. And while Section 8 follows prices set by the market, thus doing little to stop rents from increasing overall, rent control would make sure landlords get a fair return but cannot rent gouge. Section 8’s targeted subsidies, supposedly more “efficient” because they help only a few of the neediest, can perversely reward landlords who impose large rent increases. But rent control’s more universalist approach, covering all renters, better protects the public good.

Finally, as bitter a pill as it may be to swallow, we cannot rely on the private market to provide the new construction we need. Our housing market is broken. Most renters now pay unaffordable levels of income on rent. But for-profit housing cannot meet most renters’ needs, and that’s by design: when profit determines pricing, the housing needs of low-income folks never matter as much as the demand of a few rich individuals at the luxury end.

Instead, we must massively expand non-profit finance, development, and construction of social and public housing. We must protect land and housing from the vagaries of the market by creating community land trusts, cooperative housing, and mutual housing on a large scale. Other wealthy countries have done it. Sweden addressed its dire postwar housing shortage with hundreds of thousands of cooperatives and an even more massive boom in public housing construction. Thanks to these policies, along with with strong rent regulations, a much larger swathe of its population enjoys extremely low housing costs than in the U.S.

We can start by pooling our own money into cooperative banks, to finance these nonprofit housing schemes. As much as this country’s administration, headed by a tax-evading slumlord-in-chief, is gutting all our safety nets—they can’t stop us from doing that.

If you’re interested in research on this topic, see also our researcher roundtable, “What We Don’t Know About Displacement and Development,” and “Why Voters Haven’t Been Buying the Case for Building.”

Amee Chew has a doctorate in American Studies & Ethnicity and is a Mellon-ACLS Public Fellow at PolicyLink.


  1. Actually, Zuk and Chapple write: “Consistent with the LAO Report, we find that new market-rate units built from 2000 to 2013 significantly predict a reduction in the displacement indicator from 2000 to 2013” and “What we find largely supports the argument that building more housing, both market-rate and subsidized, will reduce displacement. ” This seems to be precisely the opposite of what you claim.

    More broadly, I agree that rent control might not create housing shortages if it is combined with a “massive boom in public housing construction.” But since the same progressives who want to spend more money on that also want to spend money on lots of things that are more politically popular, I doubt that this will occur in our lifetimes, EVEN IF progressives take over municipal governments in high-cost cities.

  2. Here is what I took from the Zuk and Chapple study:

    – Zuk and Chapple found market-rate housing production in San Francisco was associated with higher housing cost burdens for low-income households. These burdens – and median rents – spiked in the short term after market-rate construction.

    – For the purposes of comparability to the LAO report that they critique, Zuk and Chapple adopted that report’s same methods of defining displacement. But, Zuk and Chapple admit this probably underestimates displacement.

    As I discuss above, whether studies can even detect displacement that is occurring depends on how they measure it. Your work, and others’, has tried to argue that gentrification doesn’t really cause displacement. Many of these studies have been flawed. First, those populations who are displaced and most harmed by gentrification – low-income people of color – typically fall off the map of any measures of “area median income” (AMI) or “median rent.” By AMI estimates, the people of color in areas targeted by gentrification – areas with rapid price appreciation – tend to be extremely low-income, at 20-30 percent of AMI (because places that start out poorer are prone to the biggest price increases). If you don’t look at this population you might not see displacement. It appears Zuk and Chapple use a definition of low-income at 50-80% of AMI – again this blurs seeing who is affected by gentrification. (Meanwhile, as Zuk and Chapple point out, median rent may eventually go down, but this doesn’t matter for the most marginalized folks’ cost-burdens, and isn’t enough to mean displacement didn’t happen.)

    The best studies of displacement are ones that actually track real migration flows of specific residents over time (the studies by Atkinson, and Newman & Wyly, which I cite above actually do this). Zuk and Chapple did not measure displacement this way. Instead, they attempt to proxy displacement by measuring overall demographic shifts. For reasons I won’t get into fully now, that has problems. But in brief. Zuk and Chapple note that since the time period of impact they look at was the Great Recession, the population of low-income households was buoyed up by this external factor, decreasing perceptions of displacement. Additionally, their measure of displacement does not seem to include cases where 1) the overall population stayed the same but that of low-income folks decreased; 2) the overall population declined and the low-income population declined at the same rate. (In case 2, while rich people can easily return, once low-income people leave, they are less able to move back into a more highly priced area.) Other studies of displacement have looked more finely at changing population composition – i.e., tracking the changing number and proportion of higher-income and lower-income folks (see Davidson and Lees).

    So, using a less than ideal measure, Zuk and Chapple did not find much evidence of displacement. But, other studies that I cite above, do.

    – Timing: Contemporaneously to new construction, Zuk and Chapple found median rents and cost-burdens increased. They also found that “anti-displacement” impacts – which they likely over-estimate – lessened with time and became insignificant a decade later. Finally, depreciation of market-rate rents to be affordable to those at 50% of AMI could take 50 years. Assuming underestimations of displacement, it’s conceivable that the weight of cost-burdens on the poorest families could increase over time and eventually push them out, if they didn’t to begin with.

    – Scale Matters: “Anti-displacement” impacts became insignificant at the smaller, block group level. So those closest to the developments are experiencing pressures the study does not seem to fully account for.

    What Zuk and Chapple’s study was designed to show — which it does effectively — is that subsidized housing is much more protective than market-rate production. I think it is designed to more accurately measure impact on cost-burdens than on displacement.

    Finally, the area under study, San Francisco, is covered by rent stabilization, and Zuk and Chapple note how they saw this matter in preventing displacement. What about in cities that don’t have rent control? What about extremely low-income neighborhoods in LA which are getting luxury developments of 1,000 units – market-rate development on a much huger scale that those in Zuk and Chapple’s study?

  3. Of course expensive areas are ones that are expensive and where builders built. Plus market rate requires no subsidy from the public. There is no reason we can’t have massive amounts of both.

  4. We agree that Zuk and Chapple assert that subsidized housing is “protective than market-rate production.” But they also assert that market-rate production is protective as well; they are for both.

    Chew relies heavily on the Atkinson study, which focused on gentrifying parts of Sydney and Melbourne and shows that working-class people are modestly more likely to leave gentrifying areas than other areas. But these two places are notoriously expensive, so of course housing is more likely to become a zero-sum game in such places. (see e.g. ). (The Newman study, which she also cites, seems to be behind a paywall). Of course, when housing is scarce, people with more money are more likely to outbid people with less money, causing displacement.

    This illustrates the vicious circle caused by government regulation: government limits market-rate housing supply, which insures high housing costs, which leads to displacement, which leads to calls for more government limitation of supply. The result is of course San Francisco.

  5. Yes, Section 8 was what *originally* enabled – in fact, ENCOURAGED landlords to endlessly raise rent prices in low income neighborhoods far beyond the actual value of their (usually substandard) apartments. In more recent years, though, the displacement of working class residents has been caused by wealthy foreign residential real estate investors keeping hundreds of thousands of homes vacant year round (93,000 in just the city of Los Angeles alone) while they live overseas. To make matters worse, the wealthy “local” speculators took full advantage of the ensuing housing bubble and kept hundreds of thousands of ADDITIONAL homes vacant, perpetually on the resale market being “flipped” from one speculator to the next for higher and higher prices multiple times within the same calendar year. Now, since China’s new law (limiting the amount of money their citizens are permitted to spend outside their country) went into effect on January 1, 2017, the reselling activity has slowed down dramatically, but we’re still left with hundreds of thousands of vacant single-family and multi-family houses, while equal numbers of U.S. citizens are living in shelters and on the streets. On top of that, the wealthy Chinese meth and fentanyl manufacturing magnates are still getting around China’s new law and are still laundering their dirty illegal drug money in real estate here in the U.S., as well as in Canada, the U.K., and Oceana. New Zealand has already placed a ban on foreign real estate buyers (with the exception of buyers from neighboring Australia), and Vancouver, British Columbia in Canada has enacted a Vacancy Tax and a Speculator Tax.

  6. Pennington’s new research “The Impact of Housing Production on Legal Eviction in San Francisco” effectively rebuts most of your claims, by showing that displacement does not go up in areas with new construction. In fact, it goes down a bit. It is interesting that there are so many studies coming to opposite conclusions here.


Please enter your comment!
Please enter your name here

This site uses Akismet to reduce spam. Learn how your comment data is processed.