Race, Profit, and Algorithms: How iBuyers Leverage Neighbourhood Inequality
The housing market has long been a space where financialization, technology, and systemic inequities intersect. In recent years, in the US, a new class of real estate actors—iBuyers—has emerged, promising to streamline the home-selling process using advanced automated valuation models (AVMs). Companies like Opendoor leverage machine learning to assess property values and make near-instant cash offers to home sellers.
While this business model claims to be efficient, a closer look reveals a more complex picture: iBuyers generate higher profit margins in neighborhoods with larger populations of marginalized racial groups. This phenomenon raises critical questions about how algorithm-driven business models interact with historical patterns of segregation, property devaluation, and racial capitalism.
Screenshot from the Opendoor website
The Business of iBuying: Speed, Convenience, and Profit
The emergence of the iBuying industry is closely linked to a technocratic intervention facilitated by Silicon Valley. The intervention focuses on the home-buying process, perceived as old-fashioned by tech sectors. iBuyers appeal to homeowners by offering fast transactions with minimal effort.
This is how the iBuying process works: a seller submits their home details online, receives an algorithm-generated offer, undergoes an in-person inspection, and then closes the sale in a matter of days. iBuyers make money by reselling properties at a markup, charging service fees (typically around 5–6%), and deducting repair costs from the offer price. iBuyers focus on quick resales with minimal intervention, averaging to 3-4 months.
Figure 1 illustrates the spatiotemporal distribution of iBuyer purchases. It shows these have mostly focused on sunbelt regions (map), and that they peaked substantially during the pandemic period.
Figure 1. Spatiotemporal distribution of iBuyer purchases
Racialized Profit Margins
To understand how iBuyers applied their algorithmic business model at the neighborhood scale, particularly along racial lines, I examined property transaction data from 2014–2022, exploring iBuyer profit margins across different neighborhoods. This is what I found: (1) iBuyer profit margins tend to be spatially clustered, meaning that higher profits tend to concentrate in specific neighborhoods rather than being evenly distributed across a city or region and 2) that iBuyers gain more profits when they resell to individuals than to institutions, and 3) that some iBuyers have a statistically significant correlation between their profit margins and the proportion of marginalized racial groups within a census tract. Specifically, 3 out of 4 iBuyers generate relatively higher profits in communities of color compared to predominantly white neighborhoods, even when controlling for individual and neighborhood-level housing and socioeconomic characteristics. These findings suggest that iBuyers' profitability is not merely a function of housing market efficiency but is also shaped by racialized patterns of property valuation, neighborhood investment, and historic discrimination. The disparities in iBuyer profits across different racial compositions highlight the embedded nature of algorithmic decision-making within broader systemic inequalities in housing markets.
The Mechanisms Behind Racialized Profits
Several factors may explain why iBuyers generate higher profits in communities of color. Depressed property values due to historical segregation, redlining, and appraisal bias allow iBuyers to purchase homes at lower prices while still reselling at competitive market rates. Seller vulnerability could play a role, as homeowners in these communities may be more likely to accept below-market offers due to financial strain, lack of savings for repairs, or the need for immediate liquidity. Algorithmic bias is another factor, as iBuyer AVMs incorporate neighborhood data that reflect historical inequalities. Features such as local price trends, housing quality, and even proximity to certain amenities may encode racial disparities, reinforcing patterns of exploitation. Finally, bulk sales to institutional investors remove homeownership opportunities and exacerbate racial wealth gaps, as institutional investors convert properties into rental units.
One way to provide opportunities for homeowners of color to maintain homeownership and contribute to reducing the racial wealth gap would be to offer a range of options for home repairs. Previous research has shown that about a third of the 100 largest cities offer home repair grants, and about half offer home repair loans, thanks mostly to the Community Development Block Grant (CDBG) programs. Cities like Los Angeles, Dallas, and San Jose offer extensive repair programs with budgets exceeding $1 million, whereas smaller cities like Buffalo and North Las Vegas allocate less than $500,000. Most loans are forgivable if the property owner maintains occupancy for 10 to 15 years, offering a critical safety net to lower-income homeowners facing urgent repair needs. Additionally, the U.S. Department of Housing and Urban Development (HUD) recently updated the Federal Housing Administration's 203(k) program, increasing the maximum allowable rehabilitation cost from $35,000 to $75,000 and extending rehabilitation periods. These changes, aimed at making the program more accessible, allow more homeowners, particularly those in historically disinvested neighborhoods, to secure financing for critical repairs.
Race for Profit in Colorblind Racism
The case of iBuying exemplifies the simultaneous processes of devaluation and extraction, illustrating what Scott Markley theorizes as “planned spatial obsolescence” in the U.S. housing market. This concept suggests that value production in one location requires the creation of “anti-value” elsewhere, manifesting as economic loss in segregated or devalued areas. In this context, residential segregation serves as a mechanism that delineates where value accumulates and where economic loss is concentrated. To that end, iBuying serves as a unique example; within the framework of residential segregation and devaluation, these practices extract value from segregated neighborhoods by temporarily inflating property values for resale, which enables short-term profit-taking. However, this process does not necessarily contribute to long-term neighborhood investment or wealth-building for residents. Instead, it often reinforces unequal property relations by enabling capital to extract value while offloading the risk and potential economic loss back onto marginalized communities.
This practice has implications for the role of contemporary technologies and algorithms in a society where deeply racialized practices and policies continue to disadvantage people of color. For example, current regulations related to racial discrimination prohibit the use of identity markers such as race or gender in the development of policies or practices. While these prohibitions are often justified as a means of creating “fair” algorithms or attitudes, the problem is that these race-neutral markers, such as credit score, are often strongly correlated with other socioeconomic characteristics, and machine learning algorithms can pick up on these correlations and use them to increase the accuracy of their predictions. As a result, while iBuyers’ algorithms or business logic may not be harmful in and of themselves, by actively not engaging with a housing market that has been deeply discriminatory, they end up exploiting profits, thanks to the system of unequal housing markets.
———
Wonyoung So is a PhD candidate at the Department of Urban Studies and Planning at MIT. His research focuses on the role of contemporary technologies that perpetuate systemic racism and exclusion in housing in the U.S., with particular interests in rental housing, eviction, and mortgage lending. You can learn more about his work here.