Advances in digital technology that some analysts ascribe to a “Tech Boom 2.0”—advances such as cloud computing, mobile computing, and automated, data-driven decision-making tools—are dramatically reshaping how housing is bought and sold by homeowners and investors, operated by landlords, and inhabited by all of us. As Joe Shaw of the Oxford Internet Institute points out, new technologies—from the invention of the surveyor’s chain, in the early 17th century, to the digitization of real estate practice enabled by so-called property technology, or “proptech”—inevitably reshape how we “practice ownership, use, and exchange of the earth.”1 And no wonder: in advanced economies the value of the residential real estate sector is substantial (just over $27 trillion in the US in 2018), yet the industry has been slow to adopt Tech Boom 2.0 tools, making it ripe for “disruption” by venture-capital-backed start-ups (as the taxi and hotel industries were before them). Such digital transformations speak to housing as both a crucial site of capital accumulation and a fundamental vector of social inequality. The contentious power relations that define housing in market societies, (particularly since the 2008 financial crisis2), together with growing tensions about the role of digital platforms in urban life, make the political terrain of proptech impossible to ignore.
The March 2019 lawsuit brought by the US Department of Housing and Urban Development (HUD) against Facebook begins to highlight the contours of this terrain. HUD alleges the social media giant’s Ads Manager platform—widely used by mortgage lenders, real estate agents, and real estate listing services—facilitates violations of the Fair Housing Act (FHA). A signal achievement of the civil rights movement, the FHA outlaws housing market discrimination against members of protected classes.3 According to the government’s complaint, Ads Manager enables advertisers to target what kinds of Facebook users will—or will not—see an ad according to myriad user attributes that correspond to protected classes under the FHA, such as “Hispanic Culture,” “foreigners,” “moms of grade school kids,” “Christian,” and “childfree.” Two points particularly stand out in HUD’s complaint: first, the platform made available “a map tool to exclude people who live in a specified area from seeing an ad by drawing a red line around that area”; and second, Ads Manager operates in discriminatory ways—even when this is not the advertiser’s intention.
These points crystallize the emerging unequal consequences of mediating housing access through digital platforms. The map tool speaks to how 20th-century mechanisms of racialized exclusion such as redlining are being transferred into 21st-century technologies, ensuring that long-standing practices of discrimination by real estate actors can continue. Meanwhile, Ads Manager’s propensity to narrow the target audience for ads in biased ways shows how platforms are not simply neutral utilities. Instead, such internet platforms actively shape content in ways that can (and do) mimic calculated discrimination.
As digital platforms intervene in housing markets, they stand to reinforce long-standing social and spatial hierarchies of (dis)advantage that generate lived experiences of exclusion, displacement, and dispossession for people of color, the poor, and women. In this essay, I discuss digital transformations of housing as they have emerged in recent legal actions against Facebook and other platforms. It is vital to embed such transformations in their wider (social, political, and historical) contexts, and to interrogate their consequences and whose interests they serve. Without an explicit focus on housing justice, proptech is likely to serve the interests of people and places already benefiting from property-led accumulation, undermining the interests of propertyless subjects and marginalized places.
Disparate Treatment, Disparate Impact: Both/And
The history of housing in the United States is inseparable from systematic, finance-mediated underdevelopment of Black geographies through formal and informal practices of racial segregation. Restrictive covenants, credit rationing based on race, and predatory lending, alongside racial steering, blockbusting, and land contracts, created and exploited a dual housing market in which Black Americans pay more for worse housing than white Americans. Keeanga-Yamahtta Taylor explains how the housing market forged in 20th-century America “conflated blackness with economic risk and deteriorating property values,” producing racialized geographies in order to extract value (known as the “race tax” in the Black community).4 In the lead-up to the 2008 crisis these patterns of segregation enabled predatory lending to Black borrowers,5 who remain at a disadvantage in the post-2008 mortgage market.6
The ability for housing and mortgage advertisers to use Facebook to exclude specific groups (a capability the platform has since limited) points to how digital technologies can actively enable the persistence of these processes. As Laura Forlano recently argued, such technologies “act on our biases by replicating them and distributing them into the background of everyday life, thereby reinforcing and even exacerbating existing structural inequalities.” While Ads Manager’s erstwhile “ethnic affinities” tool sounds quite overt when imagined from the perspective of an advertiser, Forlano’s point about biases as background resonates when considered from the perspective of someone looking for housing. The data Facebook amasses about its users, and its subsequent capacity to classify users according to 10s of thousands of unique categories, means no two people looking for housing in the same market will find themselves encountering an identical set of search results or targeted advertisements (long gone are the days when one could tell a friend, “It’s the third result down when you Google it”). Despite its material effects (what geographer Stephen Graham termed, more than a decade ago, “software-sorted geographies”7), and unlike the overt discrimination of pre-FHA housing markets, this personalization is largely invisible.
Without an explicit focus on housing justice, proptech is likely to serve the interests of people and places already benefiting from property-led accumulation.
Yet platforms do not only transmit social bias: despite framing themselves as neutral conduits,8 platforms are more than utilities facilitating the interaction of, say, landlords and prospective tenants. Defining features of the platform technology—including what Nick Srnicek describes as their “privileged access to record,” in exchange for mediating between user groups9—can also create new modes of inequality.
Data collection is an inescapable condition of using platforms: the content of posts, clicks, likes, and reposts, and even how long a user pauses while scrolling through a news feed, all provide data points for platform operators. Indeed, the information people share as they interact on Facebook is deeply valuable because of how it supports advertising targeted by location, age and gender, language, interests, behaviors, and social connections: as its guide to digital advertising states, the fact that “People on Facebook share their true identities, interests, life events and more” is what gives the platform an advantage over other online advertising tools. It is also central to how Facebook can actively shape disparate housing market experiences that reinforce existing patterns of advantage and disadvantage.
The platform economy giant constitutes a crucial piece of what Marion Fourcade and Kieran Healy term the “information dragnet,” by which stores of data are amassed, enabling the “process of sorting and slotting people” to extract profits.10 Such classificatory systems also reflect structural biases in society and involve issues of control over and access to information, (mis)representation, and inclusion and exclusion.11 Ads Manager relies on Facebook users’ “sharing” to generate data on their attributes and behavior, which is subjected to predictive analytics to classify users and their likelihood of engaging with a given ad.
These categories are the basis for creating eligible audiences for ads, selecting the users who actually view ads, and differentially pricing ads according to the groups that view them (e.g., charging more for the same ad to show up in the feed of women versus men). Facebook will not show ads to users deemed unlikely to engage, even if the advertiser wants them to see it. HUD’s lawsuit argues that because pages visited and liked, apps used, and purchasing habits all vary in identifiable ways according to protected class, Facebook’s targeted advertising “inevitably recreates groupings defined by their protected class.” As a result, it may produce housing market outcomes indistinguishable from intentional discrimination, for example by systematically showing members of protected classes ads for more costly loan products, in ways that carry economic punishment.
The suit against Facebook reveals how platforms can both serve as conduits for disparate treatment by landlords, real estate agents, and lenders, and automate housing market inequalities through how opportunities are (not) made visible.
Tourists vs. Tenants: The New Dual Housing Market?
Short-term rental platform Airbnb provides another illustration of how the dynamics of platform technology can contribute to housing inequality. Not long before HUD’s case against Facebook, the city of New York sued Metropolitan Property Group, accusing the brokerage of using fake host accounts on Airbnb to operate illegal hotels. The individuals named in the suit legally leased apartments in dozens of buildings, listed them on Airbnb under profiles that looked like typical hosts, and created 18 corporations to route the income generated to themselves. The scheme generated over $20 million in revenue before the city’s recent crackdown. This case points to how Airbnb, itself notorious for skirting regulations (such as zoning rules against hotel lodgings in residential areas) and taxes (starting with hotel taxes), also facilitates illegal behavior. In addition to cons like that operated by Metropolitan Property Group, there is, as David Wachsmuth’s research documents, the phenomenon of “ghost hotels”: listings for multiple private rooms that are in fact all in the same apartment or building.12
What makes these examples compelling is not the insight they offer into fraud tactics, but what they show about how platform dynamics spill over into local housing markets. Airbnb was initially touted as a way for people to earn a bit of extra money by renting out (“sharing”) a spare room or their home while they were out of town. But as Paul Langley and Andrew Leyshon discuss, the imperative to up-scale—to “rapidly and consistently add users,” is at the center of the platform business model.13 For Airbnb, this means more and more listings, especially for entire homes, coming on to the platform, and those listings being available more consistently than would be allowed by hosts renting homes during the odd trip out of town. Drawing more and more rooms or properties onto Airbnb draws them out of what must now be called the long-term rental market to distinguish it from the short-term rental market Airbnb has created and normalized.
The higher revenues to be earned from renting properties to tourists and travelers rather than to tenants, along with the lower legal/regulatory burden, have spawned commercial-scale Airbnb operators with multiple listings (and the fraud tactics discussed above). In New York, Wachsmuth estimates, Airbnb has removed thousands of rental units from the market for city residents, and is responsible for about 16 percent of the city’s increase in rent between 2014 and 2017.14 The growth of entire-home listings was especially high in predominantly Black and Latinx neighborhoods subject to gentrification pressures (such as Bedford-Stuyvesant), where the difference between the rents of long-time residents and potential Airbnb income was greatest.15 The resulting loss of housing and rent increases are more likely to affect Black residents, while hosts in such neighborhoods—and therefore the economic beneficiaries of this process—are predominantly white.16
In the case of platforms like Airbnb, the rush to “scale” creates new speculative opportunities for property owners and affects housing opportunities by pitting tenants against tourists. But this process also maps onto existing patterns of racialized capital accumulation via the housing market, in which communities of color are destabilized for the economic benefit of white landlords.
Digitizing Housing Justice
Digital platforms can clearly work to reinforce social and spatial hierarchies of housing (dis)advantage and reproduce ideology that privileges homes as vehicles for speculation. In addition to “uploading” long-standing strategies of racial and social exclusion, core features of platform operation (such as data collection, classification, and up-scaling) serve as novel mechanisms of housing inequality, albeit mechanisms that only work because of structural injustices within the US housing market.
Rather than embracing a dystopian technological determinism that forecloses progressive digital transformations of housing, we must look to movements that are using platforms and other tools of Tech Boom 2.0 to support long-standing struggles for housing justice. This includes Housing Data Coalition, which uses New York City public data to create tools to demystify property ownership, and support tenant rights; the Anti-Eviction Mapping Project’s digital storytelling about dispossession and narratives of resistance made with community partners in San Francisco, Los Angeles, and New York; and Wem Gehört Berlin (Who Owns Berlin), which crowdsources data from tenants to document property ownership and investment patterns. By looking at how platform technologies can be used from and for the margins, it is possible to generate more radical urban futures.
- Joe Shaw, “Platform Real Estate: Theory and Practice of New Urban Real Estate Markets,” Urban Geography (2018). ↩
- See, e.g., Renee Tapp, “Renters’ Revolt: Revisiting City of Quartz to Understand Los Angeles’s Housing Crisis,” City, vol. 23, no. 1 (2019). ↩
- Defined in terms of race, color, religion, national origin, sex, disability, and familial status. ↩
- Keeanga-Yamahtta Taylor, “Back Story to the Neoliberal Moment: Race Taxes and the Political Economy of Black Urban Housing in the 1960s,” Souls: A Critical Journal of Black Politics, Culture, and Society, vol. 14, nos. 3–4 (2012). ↩
- Jacob S. Rugh Len Albright Douglas S. Massey, “Race, Space, and Cumulative Disadvantage: A Case Study of the Subprime Lending Collapse,” Social Problems, vol. 62, no. 2 (2015). ↩
- Jacob William Faber, “Segregation and the Geography of Creditworthiness: Racial Inequality in a Recovered Mortgage Market,” Housing Policy Debate, vol. 28, no. 2 (2018). ↩
- Stephen D.N. Graham, “Software-Sorted Geographies,” Progress in Human Geography, vol. 29, no. 5 (2005). ↩
- Tarleton Gillespie, “The Politics of ‘Platforms,’” New Media & Society, vol. 12, no. 3 (2010). ↩
- Nick Srnicek, Platform Capitalism (Polity, 2016), p. 44. ↩
- Marion Fourcade and Kieran Healy, “Seeing like a Market,” Socio-Economic Review, vol. 15, no. 1 (2017). ↩
- See, e.g., Safiya Umoja Noble, Algorithms of Oppression: How Search Engines Reinforce Racism (NYU Press, 2018). ↩
- David Wachsmuth et al., “The High Cost of Short-Term Rentals in New York City,” a report from the Urban Politics and Governance research group at the School of Urban Planning, McGill University, January 30, 2018. ↩
- Paul Langley and Andrew Leyshon, “Platform Capitalism: The Intermediation and Capitalisation of Digital Economic Circulation,” Finance and Society, vol. 3, no. 1 (2017). ↩
- Wachsmuth et al., “The High Cost of Short-Term Rentals in New York City.” ↩
- David Wachsmuth and Alexander Weisler, “Airbnb and the Rent Gap: Gentrification through the Sharing Economy,” Environment and Planning A: Economy and Space, vol. 50, no. 6 (2018). ↩
- Murray Cox and John Morris, “The Face of Airbnb, New York City: Airbnb as a Racial Gentrification Tool,” InsideAirbnb.com, March 1, 2017. ↩