Displaced and unsafe: The legacy of settler-colonial racial capitalism in the U.S. rental market

ABSTRACT

Unsafe rental units are disproportionately located in communities of color, resulting in numerous detrimental effects for residents’ health and socioeconomic well-being. Yet, scholars disagree regarding the mechanisms driving this phenomenon. Exogenous capitalism theories emphasize socioeconomic factors while setter-colonial racial capitalism theories emphasize the racist policies and practices that incentivize unequal investment and maintenance. We empirically adjudicate between these mechanisms by merging restricted-access versions of the American Housing Survey, the Rental Housing Finance Survey, and the American Community Survey at a Census Restricted Data Center. Our findings demonstrate neighborhood White proportion is a key mechanism shaping the condition of rental units even when controlling for neighborhood socioeconomic status, property features, and renter demographics. We argue these results support settler-colonial racial capitalism theories and discuss the implications of these findings for future research and housing policy.

KEYWORDS:

• Housing quality
• settler colonialism
• racial capitalism
• rental housing
• neighborhood inequality

Scholars have repeatedly noted unsafe, deteriorating, and uninhabitable units are disproportionately located in communities of color (Garboden & Newman, 2012; Hazekamp et al., 2020; Rosen, 2020) and have detrimental effects on residents’ physical health (Gibson et al., 2011; Hazekamp et al., 2020; Keall et al., 2010), mental well-being (Dunn, 2020; Leventhal & Newman, 2010; Shaw, 2004), residential stability (Desmond, 2016; Rosen, 2020), educational attainment (Massey & Kanaiaupuni, 1993), and economic mobility (Korver-Glenn, 2020; Korver-Glenn & Locklear, 2022). Recognizing the concentration of subpar units in marginalized communities, research investigating the lived experiences of residents in unsafe dwellings has focused on neighborhoods of color (e.g., Desmond, 2018; Desmond & Gershenson, 2017; Garboden & Newman, 2012; Rosen, 2020; Rosen & Garboden, 2022). This work has provided critical insights into renters’ lived experiences. Yet, its narrow spatial focus has curtailed researchers’ ability to disentangle the mechanisms creating these adverse conditions and propose interventions that address their root causes (Goetz et al., 2020; Howell, 2019b; Rendón, 2019; Small, 2015).

The present paper begins to address this gap by examining a representative sample of all U.S. rental units and identifying the neighborhood, property, and household factors that are associated with quality dwellings. To our knowledge, we are the first scholars to merge the restricted access versions of the American Housing Survey, the Rental Housing Finance Survey, and the American Community Survey at a Census Restricted Data Center to create a national portrait of rental units from the perspectives of both renters and landlords. This comprehensive data presents a fuller, more nuanced picture of unit conditions, renter demographics, property characteristics, and neighborhood context, enabling us to disentangle existing hypotheses regarding the factors contributing to subpar dwellings and differentiating the experiences of American Indian/Alaska Native (AI/AN, hereafter),¹ Asian, Black, Latinx, and White residents. We conclude by discussing the policy implications of our findings.

Theorizing urban inequality

For over a century, urban scholars have created theories to explain the persistent segregation of and inequality across U.S. neighborhoods (Hunter & Robinson, 2016; Small, 2015). These theories are nuanced, distinct, and varied. Yet, broadly they can be categorized into two overarching frameworks, which we call exogenous capitalism and setter-colonial racial capitalism.

Exogenous capitalism

Theories pulling from an exogenous capitalism framework perceive neighborhoods as situated in a hierarchy based on socioeconomic status. These theories build off the works of 19th century European philosophers and conjecture residents’ personal socioeconomic status determines their residential options. Often called neighborhood attainment, this theory argues residents with higher socioeconomic status can secure housing units in neighborhoods with high quality dwellings and robust public services (Dantzler et al., 2022). Some of these theories use a functionalist perspective, arguing a neighborhood hierarchy is necessary for maintaining social order and incentivizing socioeconomic mobility (e.g., Park & Burgess, 1925/1967; Zorbaugh, 1929). Others employ a Marxian lens, highlighting the mechanisms of power and exclusion that maintain inequity and curtail democracy (Harvey, 1978; Logan & Molotch, 1987).

However, despite their differences, exogenous capitalism theories agree that contemporary concentrations of subpar housing in racially marginalized neighborhoods is a product of these communities’ socioeconomic status (e.g., Conley, 2001; Gibson et al., 2011). As a result, much of this literature conflates racial and socioeconomic marginalization, assuming that the racial inequality is merely due to the socioeconomic status of racially marginalized residents (Dantzler et al., 2022). Moreover, these theories assume that racially and socioeconomically marginalized residents are concentrated in communities with cheaper, older, poorly constructed housing with more unsafe or unhealthy conditions (Pattillo, 2013). Simply put, residents with fewer financial resources live in cheaper units, are concentrated in poorer neighborhoods, and, as a result, experience more adverse dwelling conditions.

This theoretical perspective is so pervasive throughout the academic literature and colloquial explanations it is often treated as a scientific principle, not a hypothesis requiring empirical evidence. Yet, when empirical evidence is considered, the experiences of AI/AN, Asian, Black, and Latinx residents continue to contradict exogenous capitalism theories. Many scholars posit these contradictions are merely exceptions—a lingering remnant of historical exclusionary practices (e.g., transatlantic slavery, AI/AN displacement, de jure segregation) and individual discrimination (Pattillo, 2013). These scholars argue these discriminatory exceptions do not challenge the overarching logic of exogenous capitalism. However, others disagree, positing that race plays a persistent role in shaping space. They suggest the empirical evidence supports an alternative theory: setter-colonial racial capitalism.

Settler-colonial racial capitalism

Since the origins of U.S. neighborhood inequity, Indigenous and Black scholars have argued White supremacy is the central organizing factor creating spatial hierarchies (Du Bois, 1899/1996; Fleming, 2018; Itzigsohn & Brown, 2020; Ladner, 1973; Ramirez, 2007). In recent decades, these arguments have been broadly categorized under two loosely defined families of thought: settler colonialism and racial capitalism.

Settler colonialism is a set of theories that explains the violence European settlers have wrought against Indigenous peoples (AI/AN people in the U.S.) and land. It identifies the ongoing process of systematically erasing Indigenous peoples through dispossession and elimination of their cultural, social, economic, political, and ecological organization and relationships so that settlers may occupy and extract value from these spaces (Barry & Agyeman, 2020; Coulthard, 2014; Porter & Yiftachel, 2019; Porter et al., 2019; Tuck & Yang, 2012; Wolfe, 2006). Through such violence, colonizers attempt to destroy Indigenous peoples’ access to land, food, autonomy and sovereignty, cultural experiences, and socialization (Fenelon, 2014; Porter et al., 2019; Tuck & Yang, 2012). These theories argue that these processes are ongoing and continue to serve and promote the interests of settler constituents while harming Indigenous peoples (Banner, 2005; Barry & Agyeman, 2020; Coulthard, 2014; Dorries et al., 2019; Dunbar-Ortiz, 2014, 2021; Johnson, 2020; Keeler, 2016; Porter & Yiftachel, 2019; Porter et al., 2019).

Rooted in the Black Radical Tradition, racial capitalism is a line of critical inquiry that explains the mutual constitution of racism and capitalism that predates but was systematized through the transatlantic slave trade’s appropriation of enslaved African people as property (Jenkins & Leroy, 2021; Robinson, 2000). The affluent economy of the U.S. empire was explicitly built upon the raw resources harvested by enslaved labor on colonized lands (Du Bois, 1935/2014). Even after the transatlantic slave trade was banned and lifelong enslavement made illicit, the economy still relied on racialized practices of labor and resource extraction, including convict leasing (Blackmon, 2008), stealing Black land and property, and sharecropping, which persisted well into the 20th century (Du Bois, 1935/2014). In short, racial capitalism argues capitalism requires the extraction of labor and resources from subordinated racial groups and its underlying logics of economic exchange are determined based on a product, service, or location’s usefulness to the White population (Dantzler, 2021; Du Bois, 1935/2014; Robinson, 2000).

Although settler colonialism and racial capitalism have key theoretical, empirical, and foci distinctions, they share complementary perspectives on the root causes of neighborhood inequality. We call the nexus of these two families of thought settler-colonial racial capitalism. Settler-colonial racial capitalism conceptualizes neighborhoods as socially constructed, historically contingent communities whose institutional resources and property values are determined based on the neighborhood’s Whiteness (Dantzler, 2021; Dorries et al., 2019; Goetz et al., 2020; Korver-Glenn, 2021; Lipsitz, 2011; Taylor, 2019). Starting in the 1500s, settlers used treaties and initiatives to devalue AI/AN lands and ways of life (Dorries et al., 2019; Ramirez, 2007; Tomiak, 2017). Simultaneously, settlers also enslaved, subjugated, and dehumanized African peoples, using them as economic tools whose value was based on their ability to serve their enslavers’ desires (Baptist, 2014; Du Bois, 1935/2014). The legal precedents set by the devaluation, dispossession, and displacement of AI/AN and Black people enabled systems of oppression and dehumanization that continue today.

Most notably, the National Housing Act of 1934 institutionalized a national valuation system that assigned higher values to White communities (Faber, 2020; Howell & Korver-Glenn, 2018, 2021, 2022). The federal government categorized homes in White neighborhoods as more valuable, more likely to appreciate, and more stable investments than comparable homes in AI/AN, Asian, Black, and Latinx communities, justifying their racism with the market language of desirability (Taylor, 2019).² Wanting to highlight the systemic injustice of this unquestioned policy, sociologist John McKnight coined the term redlining to describe the physical color and social process used to mark communities of color on federal maps. Giving name to and organizing resistance around the government’s racism led to the outlawing of some industry practices. However, the underlying logic, which elevates White communities as inherently more valuable and justifies the devaluation and displacement of communities of color, has persisted.

The systematic government-subsidized and real estate industry-supported investment in White communities has enabled the construction, maintenance, renovation, and appreciation of White-owned and occupied housing (Connolly, 2014; Korver-Glenn, 2021; Taylor, 2019). Today, this system continues to encourage government officials and real estate industry professionals to devalue and displace AI/AN, Black, and Latinx residents to make way for new “high market value” developments for White settlers (Dorries et al., 2019), to disinvest in communities of color (Connolly, 2014; Korver-Glenn, 2021), and to systematically exclude people of color from low-interest financing that would support property ownership, maintenance, and renovations (Oliver & Shapiro, 2006; Shapiro, 2017; Taylor, 2019; Vale & Freemark, 2012). In short, theories pulling on settler-colonial racial capitalism argue the linchpin of unequal housing conditions is the establishment and maintenance of White neighborhoods through dispossession, disinvestment, and devaluation of AI/AN, Asian, Black, and Latinx people.

Mechanisms perpetuating unsafe units in marginalized communities

Both exogenous capitalism and setter-colonial racial capitalism theories have been used to explain the disproportionate concentration of unsafe housing units in racially marginalized communities (Blatman-Thomas & Porter, 2019; Dantzler, 2021; Dorries et al., 2019; Vargas et al., 2021). Exogenous capitalism conceptualizes the concentration of unsafe housing units in racially marginalized communities as an unfortunate consequence of AI/AN, Black, and Latinx residents’ socioeconomic status while settler-colonial racial capitalism argues this inequality is an inherent part of the (sub)urbanization process that upholds White supremacy. Building off these overarching theories, scholars from each tradition have posited which mechanisms mediate the observed relationship between housing quality and neighborhood racial composition.

Exogenous capitalism mediating factors

Aligning with the overarching exogenous capitalism notion that neighborhood racial inequity is primarily due to historical and contemporary socioeconomic status differences between racial groups, scholars in this tradition posit the higher proportions of unsafe rental units in communities of color can be explained by their neighborhood socioeconomic status, property features, and renter demographics.

Neighborhood socioeconomic status

Historical and contemporary policies have disadvantaged people of color in the labor market, resulting in a lower average socioeconomic status among people of color relative to White people (Lareau & Goyette, 2014; Reardon & Owens, 2014; Waldinger, 1996). Consequently, renters of color with limited income and education are only able to secure lower-quality dwellings located in poorer neighborhoods. In the aggregate, this creates communities of color with more unsafe units, but these racial differences are minimized once neighborhood socioeconomic status is held constant.

Property features

Exogenous capitalism theories argue the overarching relationship between unsafe housing conditions and neighborhood racial composition is also mediated by property features. More specifically, property conditions and management help explain why neighborhoods of lower socioeconomic status have more unsafe units. Exogenous capitalism theories conceptualize (sub)urbanization and residential mobility as a linear process whereby the oldest and cheapest housing is where lower socioeconomic status residents live. Moreover, older units with lower rent payments are assumed to have more unsafe or unhealthy conditions, like the presence of lead (Eisenberg et al., 2020). Thus, property age and annual rent serve as mediating factors explaining the higher number of unsafe units in lower socioeconomic status neighborhoods.

Likewise, in the U.S. context, multifamily and professionally managed complexes are concentrated in lower socioeconomic status neighborhoods and are arguably linked to lower quality housing units (Joint Center for Housing Studies, 2022). Researchers have found multifamily complexes have more undesirable conditions, like polluted air (Price et al., 2006). Research in this vein has also suggested that although professionally managed complexes have economies of scale with which to maintain properties (Newman, 2005), managers at these complexes lack direct control over maintenance due to corporate decision-making policies (e.g., Gomory, 2021; Leung et al., 2020), which could lead to more unsafe and unhealthy conditions over time. Exogenous capitalism theories argue that, together, property age, price, unit size, and management type are key mechanisms explaining the concentration of unsafe housing conditions in lower socioeconomic status neighborhoods.

Renter demographics

Exogenous capitalism theories, specifically neighborhood attainment, stress the impact of renters’ demographics on their ability to obtain safe units. As previously mentioned, residents’ socioeconomic status—both income and education—is seen as the primary factor determining their residential locations (Berry et al., 2022; Krysan & Crowder, 2017). Household income determines the amount of rent residents can afford while educational levels shape residents’ property search knowledge and networks. Likewise, if the renter is paying below-market rent, they might experience higher insistences of poor maintenance and unsafe conditions as the owner and/or management has fewer financial incentives to maintain their unit. Beyond socioeconomic factors, exogenous capitalism theories also conjecture other renter demographics, such as citizenship status, family composition, and racial classification, influence renters’ ability to secure safe units. Historical and contemporary exogenous capitalism theories highlight the concentration of immigrant renters in low quality housing units located within impoverished neighborhoods (e.g., Mundra & Sharma, 2015; Oliveri, 2009; Park & Burgess, 1925/1967). Scholars have also noted that households with children, especially those headed by single women, have a harder time securing quality housing and receiving the needed repairs they request (see Desmond, 2016; Desmond et al., 2013). Finally, exogenous capitalism theories recognize that some individuals of color experience racism when they attempt to secure housing units or ask for maintenance (Garboden & Rosen, 2019; Rosen et al., 2021; Rosenbaum, 1996; Schill et al., 1998).

In short, research rooted in exogenous capitalism theories argues that renter demographics shape the properties they rent and the neighborhood socioeconomic status in which these properties are located. Thus, work in this vein hypothesizes that renter and property-level characteristics are strong predictors of unit safety and explain the relationship between neighborhood socioeconomic status and unit safety, which in turn explains the relationship between neighborhood racial composition and unit safety.

Settler-colonial racial capitalism mediating factors

Settler-colonial racial capitalism theories explaining concentrations of unsafe neighborhood conditions flip the basic tenets of exogenous capitalism on their head. Instead of seeing neighborhoods as relatively stable entities chosen by residents, these theorists argue neighborhood resources and perceived value are socially constructed and mutate depending on the residents’ Whiteness (Barry & Agyeman, 2020; Dantzler, 2021; Porter et al., 2019; Taylor, 2019). This body of work identifies the policies and practices that have elevated White property values, maintenance, and appreciation while devaluing and dispossessing communities of color as the key mechanisms contributing to the concentration of unsafe units in communities of color (Dantzler, 2021; Dorries et al., 2019; Goetz et al., 2020; Korver-Glenn, 2021; Lipsitz, 2011; Taylor, 2019). From this perspective, neighborhood socioeconomic status, property features, and renter demographics are not exogenous factors explaining the correlation between neighborhood race and socioeconomic status. Instead, they are proxies that help capture the ways in which colonial White supremacy manifests in urban and suburban development.

Neighborhood socioeconomic status

Racist housing and labor policies have constrained AI/AN, Asian, Black, and Latinx workers’ ability to receive fair wages for their labor and accumulate generational wealth through asset appreciation (Dorries et al., 2019; Faber, 2020; Lareau & Goyette, 2014; Reardon & Owens, 2014; Waldinger, 1996). Specifically, hyper-appreciation of homes in White neighborhoods has increased White residents’ wealth and, consequently, their educational attainment and household income (Howell & Korver-Glenn, 2021). This results in White communities having higher socioeconomic status as well as more capital to invest in and upkeep properties.

Property features

Racist historical and contemporary housing policies have concentrated multiunit and older housing complexes in communities of color (Eisenberg et al., 2020; Howell & Korver-Glenn, 2021; Joint Center for Housing Studies, 2022; Massey & Tannen, 2015). Yet, ironically, these units are not necessarily cheaper for renters nor managed by professional managers. In fact, the devaluation of communities of color has enabled small-time landlords to accumulate properties with relatively low investments and without the resources or economies of scale to ensure their maintenance (Garboden & Newman, 2012; Newman, 2005). Simultaneously, racist real estate policies and practices filter people into racially segregated communities, limiting renters’ ability to compare units, and contributing to landlords’ ability to charge similar rents across distinct properties and neighborhoods. The relationships between property features and unit quality provide further evidence of a settler colonial racial capitalism housing system.

Renter demographics

Like their conceptualization of neighborhood socioeconomic status and property features, settler-colonial racial capitalism theories conceptualize the relationship between renter demographics and unsafe units as proxies for how racism shapes renter experiences. These theories acknowledge individual-level discrimination across racial, family, class, and citizenship status contribute to renters’ experiences (e.g., Faber & Mercier, 2022; Korver-Glenn & Locklear, 2022). However, they argue that while these individual experiences contribute to the overarching patterns of unsafe units in communities of color, they are only a part of the larger structural conditions that invest more resources into White communities.

Settler-colonial racial capitalism theories emphasize the underlying logics that structure the policies and industry practices that create neighborhood hierarchies and reinforce White supremacy through dispossessing communities of color of their access to safe, habitable housing while extracting resources from them (Barry & Agyeman, 2020; Dantzler, 2021; Dorries et al., 2019; Porter et al., 2019; Taylor, 2019). This pervasive racial hierarchy influences neighborhood socioeconomic statuses, property features, and the experiences of individual renters.

Despite these theoretical explanations, little empirical data has investigated their claims across the contemporary U.S. rental market. We begin to adjudicate between these theories by empirically investigating to what extent the relationship between neighborhood racial composition and unsafe housing conditions can be explained by neighborhood socioeconomic status, property features, and renter demographics.

Data and methods

Data sources

To investigate the relationship between neighborhood racial composition and unsafe housing conditions, we combined three U.S. Census Bureau surveys—the American Housing Survey (AHS), the Rental Housing Finance Survey (RHFS), and the American Community Survey (ACS). Together, these data provide a nationally representative snapshot of rental units from the perspectives of both tenants and landlords.

Sponsored by the Department of Housing and Urban Development (HUD) and conducted by the U.S. Census Bureau, the AHS surveys residents on their unit condition, cost, and demographics.³ Also sponsored by HUD and administered by the Census Bureau, the RHFS surveys landlords of AHS renter-occupied units on their property’s financial, mortgage, and maintenance characteristics.⁴ The publicly accessible versions of the AHS and RHFS are not linkable to help protect respondents’ anonymity. However, we applied for and were granted access to the rental unit identification numbers through a Federal Research Data Center (RDC). To our knowledge, we are the first researchers to link these datasets and investigate renters’ housing experiences from multiple vantage points.

We also combined our linked AHS and RHFS dataset with neighborhood and metropolitan level characteristics from the ACS. To match the AHS and RHFS timeframe, we pulled neighborhood and metropolitan demographics from the ACS 2013–2017 5-year summary files. Using the Census classifications, we defined neighborhoods as census tracts and metropolitan areas as all the counties with economic ties to a central city as measured by commuting patterns. Combining these three datasets enabled us to investigate housing quality across renter, property, neighborhood, and metropolitan characteristics.

Housing quality

Drawing on residents’ AHS reports, we defined housing quality as the number of unsafe or unhealthy conditions within the unit. We considered exposed wires, blown fuses, broken electrical outlets, broken furnace, broken toilets, water supply interrupted, sewage failure, rodent infestation, cockroach infestation, foundation damage, roof damage, broken windows, unstable exterior walls, unstable floors, and mold as unsafe or unhealthy conditions.⁵ The majority of the sample (55%) reported no unhealthy or unsafe conditions. A quarter reported one unsafe or unhealthy condition, a tenth reported two unsafe or unhealthy conditions, and the remaining tenth of the sample had three or more unsafe or unhealthy conditions. To reduce the influence of this rightward skew, we capped the number of unsafe or unhealthy conditions at four, ensuring our presented results are conservative.⁶

Neighborhood racial composition

We operationalized neighborhood racial composition as the proportion of the census tract that identified as non-Hispanic White.⁷ Given that White presence has a gradual influence on neighborhood perceptions and resources (Howell & Emerson, 2018; Krysan & Bader, 2009), we operationalized neighborhood White proportion as continuous. That said, in our descriptive table, we provide an overview of neighborhood distinctions by comparing White neighborhoods (census tracts with 50% or more non-Hispanic White residents) to communities of color (census tracts with less than 50% non-Hispanic White residents).

Neighborhood socioeconomic status

We defined neighborhood socioeconomic characteristics as the census tract’s median household income, owner occupied proportion, and vacant proportion. Owner occupancy and neighborhood vacancy are defined with distinct denominators to ensure they capture different phenomena. Specifically, the census tract’s owner-occupied proportion is the proportion of occupied households who own their dwelling while the tract’s vacancy is the proportion of all housing units that are unoccupied. To adjust for the rightward skew of the tract’s median household income and vacancy rate, we used natural logarithm transformations.

Property features

Using information provided by landlords and managers on the RHFS, we controlled for the property’s size, age, average annual rent, and management type. We operationalized property size as a binary variable differentiating single-family homes from multiunit buildings. We define the property age as the year construction started on the complex.⁸ Average annual rent was calculated as the total amount of potential rent divided by the number of units and transformed using the natural logarithm.⁹ Finally, management type was operationalized as a binary variable differentiating between units that are managed by their owner or unpaid family member versus a paid employee or management company.

Renter demographics

We used residents’ self-reports to define their racial classification, family composition, citizenship status, socioeconomic status, and unit cost. We conceptualize respondents’ self-reported race as a proxy for their exposure to individual-level racism in the rental market, not as a category of natural or biological difference. Given this, we followed the empirical findings of Howell and Emerson (2017) and grouped respondents into five categories: AI/AN, Asian, Black, Latinx, and White.¹⁰ To capture the intersectional nature of family composition, we created a variable that includes gender, marital, and parental status. Specifically, this variable has six categories: Childless Couples, Childless Man, Childless Woman, Coupled Parents, Single Father, and Single Mother.¹¹ We defined citizens as U.S. born and naturalized residents. To capture renter socioeconomic status, we created an index including renter’s highest level of completed education¹² and annual household income. This is a standardized index, meaning that negative values represent below average socioeconomic status while positive values indicate above average socioeconomic status.¹³ Finally, to capture whether rental discounts or subsidies explain unit condition, we calculated the ratio between tenant payments and the property’s potential rent as reported by the landlord.¹⁴ Values of one indicate renters who paid market rate rent. Values less than one suggests the renter has a discount or subsidy.

Statistical models and metropolitan controls

To empirically disentangle which neighborhood, property, and renter characteristics mediate the relationship between neighborhood racial composition and housing quality, we estimated ordinary least squares regressions. To account for RHFS’s sampling frame, we used a random effects (multilevel) model that nested housing units within metropolitan areas. This approach allows us to account for the fact that the RHFS randomly selects households within set metropolitan areas. Since the metropolitan areas are not randomly selected and are home to multiple rental units each, a hierarchical model allows us to adjust our standard errors to reflect the nonrandom clustering of the data. In short, this modeling approach enables us to estimate whether the inequities observed within our sample are likely present across rental units throughout U.S. metropolitan areas. Additionally, we controlled for two metropolitan area characteristics to account for the influence that metropolitan level differences might have on observed inequities. These two control variables are single family home proportion and the total metropolitan population.¹⁵ We estimated all models with Stata’s multilevel ordinary least squares regression command: xtreg.

Results

Before we investigated the factors concentrating subpar housing units in communities of color, we utilized our novel data to provide a comprehensive overview of U.S. renters and their units.

U.S. renters: Who they are and where they live

U.S. renters live in a wide variety of housing units across dramatically divergent neighborhoods and metropolitan areas. Some renters are young, single professionals dwelling in Manhattan’s luxurious high-rise complexes. Other renters are families living in century-old single-family farm homes where their closest neighbor is a mile away and their county’s tax base struggles to afford educational and infrastructural services. Still other renters are retired couples residing in newly built townhomes located in gated suburban communities on the fringe of booming Sun Belt cities.

Looking across the entire renting population, we observe nearly half of U.S. renters are White, a third are college educated, and a third married (see Table 1). One in four renters live in single-family homes and a third live in properties managed directly by the unit owner. On average, renters live in neighborhoods where nearly half the homes are owner occupied, the median income is approximately $60,000, and they pay $1,100 in housing costs each month. For those familiar with U.S. demographics, it is likely evident that renters are less likely than the entire U.S. population to be White, college educated, and married. This is due to the longstanding federal homeownership subsidies targeted at the middle class and disproportionally available to White families that have made homeownership financially advantageous for most White, college-educated couples (Oliver & Shapiro, 2006; Shapiro, 2017; Taylor, 2019; Vale & Freemark, 2012). That said, it is important to note that White, college educated, married couples still make up a substantial proportion of U.S. renters. This diversity in renter backgrounds and residential neighborhoods enables us to empirically disentangle the mechanisms contributing to higher concentrations of subpar housing in communities of color.

Table 1. Descriptive statistics of U.S. renters.

	All Renters	White Communities	Communities of Color
Dependent Variable
Number of Unsafe Unit Conditions	0.80 (1.12)	0.67 (1.03)	0.96 (1.19)
Neighborhood Racial Composition
White Proportion	0.51 (0.28)	0.74 (0.13)	0.25 (0.15)
Neighborhood SES
Median Income	57,590 (27,380)	65,130 (28,460)	48,790 (23,140)
Owner Occupied Proportion	0.46 (0.22)	0.53 (0.21)	0.36 (0.21)
Vacant Proportion	0.10 (0.07)	0.09 (0.08)	0.10 (0.07)
Property Features
Multiunit Complex	0.75 (0.43)	0.74 (0.44)	0.76 (0.43)
Year Built	1973 (31)	1973 (33)	1973 (29)
Average Annual Rent	13,700 (7,721)	13,770 (8,504)	13,630 (6,693)
Professionally Managed	0.71 (0.45)	0.69 (0.46)	0.73 (0.44)
Renter Demographics
Racial Classification
White	0.48 (0.50)	0.67 (0.50)	0.25 (0.44)
AI/AN	0.01 (0.12)	0.02 (0.13)	0.01 (0.10)
Asian	0.08 (0.27)	0.07 (0.25)	0.09 (0.29)
Black	0.23 (0.42)	0.14 (0.45)	0.32 (0.47)
Latinx	0.20 (0.40)	0.10 (0.30)	0.32 (0.47)
Family Composition
Childless Couples	0.16 (0.37)	0.17 (0.37)	0.16 (0.37)
Childless Man	0.25 (0.43)	0.27 (0.44)	0.23 (0.42)
Childless Woman	0.30 (0.46)	0.33 (0.47)	0.27 (0.44)
Coupled Parents	0.15 (0.36)	0.12 (0.33)	0.18 (0.39)
Single Father	0.02 (0.13)	0.01 (0.11)	0.02 (0.15)
Single Mother	0.11 (0.32)	0.09 (0.29)	0.14 (0.35)
Citizen	0.85 (0.35)	0.91 (0.29)	0.79 (0.41)
Socioeconomic Status	0.14 (0.94)	0.20 (0.89)	0.07 (0.99)
Educational Attainment
Less than High School	0.05 (0.22)	0.03 (0.17)	0.08 (0.27)
Some High School	0.11 (0.32)	0.08 (0.28)	0.15 (0.36)
High School	0.25 (0.44)	0.25 (0.43)	0.26 (0.44)
Some College	0.18 (0.38)	0.19 (0.39)	0.16 (0.37)
Associates	0.12 (0.33)	0.13 (0.33)	0.12 (0.32)
Bachelors	0.19 (0.39)	0.21 (0.41)	0.16 (0.37)
Graduate Degree	0.09 (0.29)	0.11 (0.32)	0.07 (0.26)
Annual Household Income	46,770 (52,800)	49,120 (56,470)	44,020 (48,040)
Annual Rent and Utility Cost	13,260 (11,910)	13,190 (13,020)	13,340 (10,480)
Metropolitan Area Characteristics
Single Family Home Proportion	0.59 (0.10)	0.61 (0.10)	0.58 (0.10)
Total Population	4,649,000(4,390,000)	3,554,000 (3,812,000)	5,927,000 (4,668,000)
Number of Respondents	4,000	2,100	1,800

All counts, averages, proportions, and standard errors are rounded based on the Census Restricted Data Center protocols. These rounding protocols make it seem as though the number of respondents in each neighborhood type does not sum to the total number of respondents. However, in the non-rounded tables this is not the case. We list the White category as the first racial classification because we are explicitly drawing attention to Whiteness as the key factor shaping renters’ experiences of unsafe housing.

Renting in White neighborhoods verses communities of color

Aligning with previous studies, we observed unsafe and unhealthy dwelling conditions are more common in communities of color than White neighborhoods. To visualize this correlation, we divided neighborhoods into a binary classification: White communities (census tracts where more than 50% of the residents identify as non-Hispanic White) and communities of color (census tracts where less than 50% of residents identify as non-Hispanic White).

As seen in Table 1, slightly more than half the sample lives in White communities and they have, on average, 0.67 unsafe or unhealthy conditions. This is approximately two thirds the number of problems observed in units located in communities of color—which have an average of 0.96 unsafe or unhealthy conditions. As expected, White communities are more affluent, with a higher proportion of owner-occupied dwellings and slightly fewer vacant units. Likewise, renters in White neighborhoods are more likely than renters in communities of color to be U.S. citizens (91% compared to 79%), have college degrees (32% compared to 23%), be childless (77 compared to 66%), have higher household incomes (an average of $5,000 more annually), and identify as White (67% compared to 25%) or AI/AN (2% compared to 1%). White communities are also, on average, located within smaller metropolitan areas.

Yet, contrary to expectations, the rented units themselves are more comparable across White neighborhoods and communities of color than often portrayed. In both neighborhood types, units are 45 years old on average, approximately a quarter of them are single family homes, and about 70% of the units are professionally managed.¹⁶ Renters also pay very comparable amounts, with renters in communities of color paying $13 more per month than renters in White neighborhoods. This is a substantively inconsequential difference. Yet, it is critical to point out that this contradicts the prevailing assumption—that renting in communities of color is cheaper than renting in White neighborhoods. Instead, the opposite is true. This fact complicates the common exogenous capitalism theories regarding why units in communities of color are more likely to have unsafe or unhealthy conditions. As discussed above, exogenous capitalism theories presume historical and contemporary segregation concentrates older, cheaper, multifamily, professionally managed properties in communities of color. This literature often suggests these socioeconomic and infrastructural factors explain the observed inequity across neighborhoods. However, in using data on the properties themselves, we highlight these assumptions are inaccurate.

In short, renters in communities of color are more likely to reside in units with unsafe or unhealthy conditions. Yet, the neighborhood, property, and household factors proposed as explanations for this correlation are not all correlated with neighborhood racial composition—suggesting settler-colonial racial capitalism explanations that centralize structural racism’s role in upholding White supremacy might more accurately reflect the empirical data. To further disentangle the factors that mediate the relationship between neighborhood racial composition and unsafe unit conditions, we conducted multiple regression models.

Disentangling the factors contributing to unsafe and unhealthy rental units

As previously outlined, exogenous capitalism theories presume the relationship between neighborhood racial composition and unsafe conditions is mediated by socioeconomic inequality (e.g., Conley, 2001; Gibson et al., 2011). Conversely, settler-colonial racial capitalism theories see racist, dispossessive policies directly creating and sustaining racially unequal neighborhoods (e.g., Barry & Agyeman, 2020; Dantzler, 2021; Dorries et al., 2019; Porter et al., 2019; Taylor, 2019). We empirically adjudicated between these theories by controlling for neighborhood socioeconomic status, property features, and renter demographics.

Neighborhood socioeconomic status

As expected by both perspectives, the relationship between White neighborhoods and unsafe and unhealthy housing conditions is mediated by neighborhood median income (see Table 2, Model 2). However, exogenous capitalism theorists perceive this mediation as demonstrating that socioeconomic status is the main driver of the observed inequity while settler-colonial racial capitalism conceptualizes this mediation as one mechanism by which racist policies shape neighborhoods. To adjudicate between these explanations, we turn to the other measures of socioeconomic status—owner occupancy and vacancy rates. Neither of these measures mediate the observed relationship between White neighborhoods and unsafe and unhealthy housing conditions. In fact, higher owner-occupancy rates result in more unsafe conditions. This suggests the mechanisms shaping the relationship between White neighborhoods and unit conditions cannot be solely explained by socioeconomic status.

Table 2. Coefficients from regression models predicting number of unsafe unit conditions.

	Model 1	Model 2	Model 3	Model 4
Neighborhood Racial Composition
White Proportions	−0.59 (0.07)*	−0.09 (0.02)*	−0.11 (0.02)*	−0.10 (0.02)*
Neighborhood Socioeconomic Status
Median Income, Transformed		−0.13 (0.02)*	−0.09 (0.03)*	−0.07 (0.03)*
Owner Occupied Proportion		0.06 (0.02)*	0.05 (0.02)*	0.03 (0.02)
Vacant Proportion, Transformed		0.02 (0.02)	0.02 (0.02)	0.02 (0.02)
Property Features
Multiunit Complex			−0.04 (0.02)*	−0.02 (0.02)
Year Built			−0.18 (0.02)*	−0.17 (0.02)*
Average Annual Rent, Transformed			−0.03 (0.02)	−0.04 (0.02)
Professionally Managed			0.01 (0.02)	0.01 (0.02)
Renter Demographics
Racial Classification (Ref. White)
AI/AN				0.38 (0.13)*
Asian				−0.13 (0.06)*
Black				−0.05 (0.04)
Latinx				0.02 (0.05)
Family Composition (Ref. Childless Couple)
Childless Man				−0.10 (0.05)*
Childless Woman				−0.09 (0.05)
Coupled Parents				0.16 (0.06)*
Single Father				0.35 (0.12)*
Single Mother				0.22 (0.06)*
Citizen				0.01 (0.02)
Socioeconomic Status, Transformed				0.01 (0.02)
Unit to Property Cost Ratio, Transformed				−0.01 (0.02)
Metropolitan Area Characteristics
Single Family Home Proportion	0.15 (0.40)	−0.01 (0.04)	0.01 (0.03)	0.01 (0.03)
Total Population, Transformed	0.00 (0.00)	0.06 (0.04)	0.06 (0.04)	0.05 (0.04)
Constant	1.02 (0.29)	0.00 (0.03)	0.00 (0.03)	0.01 (0.04)
Within	0.0156	0.0244	0.0570	0.0730
Between	0.0173	0.0314	0.0708	0.1390
Overall	0.0220	0.0290	0.0558	0.0761
Number of Respondents (Metros)	4,000 (250)	4,000 (250)	4,000 (250)	4,000 (250)

*p-value < .05.

All counts, coefficients, standard errors, and R² values are rounded based on the Census Restricted Data Center protocols for model disclosures. We use White racial classification as the reference category because we are explicitly drawing attention to Whiteness as the key factor shaping renters’ experiences of unsafe housing.

Property features

In the third model of Table 2, we introduced property features. These features do not explain the inequality between White neighborhoods and communities of color—quite the opposite. Controlling for property features magnifies the relationship between neighborhood White proportion and safe housing units. This contradicts exogenous capitalism’s hypothesis that the better maintained units in White neighborhoods are due to White communities’ higher quality housing stock. Instead, as hypothesized by settler-colonial racial capitalism, the neighborhood racial composition has a stronger influence on unsafe conditions than unit features or cost.

Renter demographics

Finally, Model 4 of Table 2 introduced renter demographics. Contradicting the expectations of exogenous capitalism, renters’ socioeconomic status does not mediate the relationship between neighborhood racial composition and unit condition nor is it even a statistically significant predictor of unsafe or unhealthy unit conditions. Likewise, citizenship and whether renters are paying full market value for their units has no influence on unit conditions. Households with children do experience more adverse conditions than their childless counterparts but the difference does not explain the correlation between the neighborhood’s Whiteness and unsafe and unhealthy conditions. Finally, even individual-level racism does not explain the relationship between neighborhood racial composition and unit condition. These findings support settler-colonial racial capitalism’s notion that inequality between White neighborhoods and communities of color result from racist, dispossessive policies which financially incentivize landlords to maintain properties in White communities while extracting rent from deteriorating homes in communities of color (e.g., Barry & Agyeman, 2020; Dantzler, 2021; Dorries et al., 2019; Porter et al., 2019; Taylor, 2019).

The evidence for settler-colonial racial capitalism becomes even stronger when these results are contextualized within the U.S. neighborhood racial landscape. For Black, Latinx, and White renters, unit conditions are shaped by their neighborhood racial composition—not individual experiences of racism. This is possible because these three groups are highly segregated, with most residents living in census tracts where their racial group is the majority (see Table 3). Conversely, AI/AN and Asian renters experience a strong correlation between their individual race and unit conditions. In particular, AI/AN renters have considerably more unsafe and unhealthy conditions than their White neighbors. Although AI/AN people are more likely than any racial group to live in neighborhoods where their own racial group is concentrated¹⁷ and just as likely as Black and Latinx residents to live in neighborhoods of color when they reside in diverse counties, most AI/AN residents live in majority White counties, where they live in AI/AN clusters within majority White census tracts. These patterns are the result of settler colonial processes: historical displacement, allotment sales, and census tract boundaries cutting through Tribal Nation borders and AI/AN urban developments (Banner, 2005; Rossiter, 2012).¹⁸ In other words, U.S. census tracts do not distinguish AI/AN communities from White neighborhoods. As a result, our models can only capture the devaluation of these communities through the experience of individual AI/AN renters. Yet, their adverse experiences combined with their residential distribution provides additional support for the settler-colonial racial capitalism notion that racial composition drives property evaluation and maintenance.

Table 3. Neighborhood racial demographics for all U.S. residents, ACS 2013–2017.

	AI/AN	Asian	Black	Latinx	White
Proportion of U.S. Population	0.01	0.06	0.13	0.18	0.63
Proportion Living in a Neighborhood Where Their Racial Group is Overrepresented	0.88	0.80	0.81	0.77	0.80
Proportion Living in a Neighborhood Where Their Racial Group is the Majority	0.24	0.14	0.41	0.44	0.88
Proportion Living in Majority White Counties	0.64	0.41	0.52	0.38	0.81
Majority White Counties
Neighborhood AI/AN Proportion	0.14	0.00	0.00	0.01	0.01
Neighborhood Asian Proportion	0.03	0.12	0.03	0.04	0.03
Neighborhood Black Proportion	0.06	0.09	0.35	0.10	0.06
Neighborhood Latinx Proportion	0.11	0.11	0.11	0.27	0.08
Neighborhood White Proportion	0.66	0.67	0.50	0.58	0.82
Majority Residents of Color Counties
Neighborhood AI/AN Proportion	0.48	0.00	0.00	0.00	0.00
Neighborhood Asian Proportion	0.04	0.31	0.05	0.07	0.10
Neighborhood Black Proportion	0.07	0.09	0.54	0.10	0.11
Neighborhood Latinx Proportion	0.17	0.25	0.19	0.58	0.23
Neighborhood White Proportion	0.24	0.35	0.21	0.24	0.56

For this table, racial groups exclude multiracial individuals from the numerators and denominators. Neighborhood proportions are weighted by each racial group’s population.

Discussion and conclusion

By combining renter and landlord surveys, we created a novel, nationally representative dataset. This data enabled us to empirically investigate the neighborhood, property, and renter characteristics mediating the relationship between neighborhood racial composition and unit condition. Our results demonstrate that inequality across White neighborhoods and communities of color cannot be explained by neighborhood and renter socioeconomic status—as exogenous capitalism theories have suggested (e.g., Conley, 2001; Gibson et al., 2011; Rosenbaum, 1996; Zavisca & Gerber, 2016). Nor can it be explained by property features or even landlord discrimination against individual renters of color. Instead, we find neighborhood White proportion shapes unit condition, even when controlling for unit cost, neighborhood socioeconomic status, property features, and renter demographics. These results support settler-colonial racial capitalism theories, which argue that a constellation of governmental, industry, and private policies and practices have financially incentivized hyper-appreciation and investment in White settler communities—a process that is contingent on dispossessing communities of color of safe, habitable homes while simultaneously extracting resources from them (Barry & Agyeman, 2020; Dantzler et al., 2022; Dorries et al., 2019; Fleming, 2018; Howell & Korver-Glenn, 2021, 2022; Ladner, 1973; Taylor, 2019). These findings challenge common narratives about subpar rental units and, in doing so, suggest new directions for future research and policy interventions.

Future research: Settler-colonial racial capitalism informing data and theories

By demonstrating settler-colonial racial capitalism is pervasive yet context-specific, these results highlight the importance of designing research that considers the sociohistorical context of communities and individuals. In particular, the distinct residential patterns and experiences of AI/AN renters illuminated by our findings reinforce the critical importance of including accurate measures of AI/AN communities and identity. This includes moving away from census tracts’ settler-colonial roots and exploring alternative geographic boundaries and scales that more precisely reflect AI/AN communities. It also entails collecting data on tribal affiliation, enrollment status, and other indicators of tribal belonging to examine the heterogeneity across regions, tribal structure, and colonial experiences (Huyser & Locklear, 2021; Huyser et al., 2010, 2014), in concordance with Indigenous Data Sovereignty (Kukutai & Taylor, 2016). Beyond more accurate and multilayered data, research should also reconceptualize U.S. land as AI/AN space, occupied by settlers, descendants of transatlantic enslaved people, and AI/AN people alike (Barry & Agyeman, 2020; Dunbar-Ortiz, 2021; Porter & Yiftachel, 2019). Such a reconceptualization centers the ongoing violence, erasure, and exploitation experienced by AI/AN, Black, and Latinx people (Addie & Fraser, 2019; Barry & Agyeman, 2020; Porter et al., 2019; Wolfe, 2006).

Policy interventions: Reshaping the foundations of urban inequity

Building off the pervasive assumption that Whiteness is normative and desirable, policy interventions attempting to address subpar housing conditions in communities of color often encourage relocating residents to majority White neighborhoods or elevating residents’ socioeconomic status (Howell, 2019b). However, our results suggest these approaches fail to address the fundamental factors creating unsafe and unhealthy units. As epitomized by AI/AN renter experiences, living in White neighborhoods is not a panacea for the devastating consequences of White supremacy. Instead, interventions need to reshape the policies and industry practices that incentivize hyper-investment in White communities and dispossession and resource extraction in communities of color. This requires rethinking the monetary policies that perpetuate cyclical investment and disinvestment, introducing fiscal policies that provide reparations to harmed communities, and creating enforcement mechanisms to reduce renter exploitation and dispossession.

Disclaimer

Views expressed in this article are those of the authors and not those of the U.S. Census Bureau. The Census Bureau’s Disclosure Review Board and Disclosure Avoidance Officers have reviewed this information product for unauthorized disclosure of confidential information and have approved the disclosure avoidance practices applied to this release. This research was performed at a Federal Statistical Research Data Center under FSRDC Project Number 2080. (CBDRB-FY21-P2080-R9105).

Acknowledgments

The authors would like to thank Hajar Yazdiha for providing helpful feedback on a previous version of this manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This research uses data from the Census Bureau’s Longitudinal Employer Household Dynamics Program, which was partially supported by the following National Science Foundation [Grants SES-9978093, SES-0339191 and ITR-0427889]; National Institute on Aging [Grant AG018854]; and grants from the Alfred P. Sloan Foundation. This specific project was also funded by the Russell Sage Foundation and the University of New Mexico ADVANCE program and the UNM Office of the Vice President for Research.

Notes on contributors

Elizabeth Korver-Glenn

Elizabeth Korver-Glenn is assistant professor of sociology at Washington University in St. Louis. Dr. Korver-Glenn’s scholarship, which has been supported by the Russell Sage Foundation and the Swedish Research Council, among other external and internal grants, examines the causes and consequences of racial, class, and gender inequalities in urban contexts and how policy can mitigate these inequities.

Sofia Locklear

Sofia Locklear is an enrolled citizen of the Lumbee Tribe of North Carolina and currently an Assistant Professor at Western University. Dr. Locklear’s research investigates racial inequities, Whiteness, and the racialization of Indigenous people.

Junia Howell

Junia Howell is an is an urban sociologist and race scholar who uses quantitative and qualitative tools to identify and dismantle the specific policies, processes, and practices that uphold White supremacy. Currently, her work focuses on the housing industry and disaster relief. Dr. Howell currently holds a faculty position at the University of Illinois Chicago.

Ellen Whitehead

Ellen Whitehead is an assistant professor of sociology at Ball State University who specializes in family sociology, social inequality, and race and ethnicity. Dr. Whitehead’s research has been supported through grants from the Russell Sage Foundation, the National Institute of Food and Agriculture, and the Aspire Program at Ball State.

Notes

1. Aligning with the terminology recommendations of the National Congress of American Indians, we use American Indian/Alaska Native (AI/AN) throughout the paper in reference to communities whose ancestors lived in the Americas before European settler colonizers invaded the land. This includes people of Tribal Nations who resided in the continental United States (American Indians) and Tribal Nations and Villages of Alaska (Alaska Natives).

2. In the 1930s, Asian and Latinx communities were less common than Black and AI/AN neighborhoods. Thus, less attention has been brought to the role HOLC classifications had on these communities. For an example, see the HOLC map of San Antonio: https://digital.utsa.edu/digital/collection/p16018coll12/id/78/rec/3.

3. The survey is nationally representative with an oversample of the 15 largest metropolitan areas.

4. The AHS is collected from May to September of odd number years and the corresponding RHFS is administered the following year. We used the most recent available data: the 2017 AHS (n = 82,591 households) and the 2018 RHFS (n = 4,330 units).

5. Wires not enclosed in a wall, metal, or plastic covering were defined as exposed. Broken furnaces included instances where the primary heating stopped working during winter months. Broken toilets included toilets that stopped working in the last 3 months. Interrupted water supply was defined as losing running water in the last 3 months. Sewage failures entailed an unusable system in the prior 3 months. Infestations were defined as visual evidence of rodents or insects within the last 6 months. Foundation, roof, and window damage included visible holes, cracks, crumbling, missing shingles, and sagging. Unstable walls and floors were defined as those with cracks or holes at least 4 inches across and deeper than 3 inches or visible buckling, leaning, or sloping. Mold covered an area larger than an 8.5 by 11-inch piece of paper.

6. Our variable mirrors HUD’s classification of inadequate housing units. Like our variable, HUD counts the number and type of plumbing, heating, electric, wiring, and upkeep (defined as external and internal water leaks, holes and cracks in the walls, ceilings, or floors, peeling or broken plaster, and rats) problems, but they create quality categories (e.g., severely inadequate). Although the substantive results of the two measures are comparable, we maintain a continuous variable to reflect how more problems gradually make the unit more unlivable.

7. In supplemental models, we used alternative definitions of racial neighborhood composition to empirically confirm our theoretical assertion that the White proportion drive neighborhood inequality (see also Goetz et al., 2020; Howell & Korver-Glenn, 2021; Howell, 2019a).

8. Most complexes were built in 1 year. However, for the few multiyear, multibuilding projects, we are unable to identify the AHS respondent’s building. To produce conservative results, we use the oldest building as the property’s age. In a couple of cases, owners only report the age of their newest building. For these cases, we used this year to reduce the missing data.

9. Some owners did not report potential or market rate rent. For these respondents, we used the received rent across all units as a proxy for potential rent.

10. Building off Howell & Emerson’s (2017) findings, we classify all respondents into one of five categories. Specifically, Latinx renters are all Hispanic respondents. White renters are monoracial non-Hispanic White respondents. Black renters are monoracial and all multiracial non-Hispanic Black respondents. AI/AN renters are monoracial non-Hispanic AI/AN respondents and multiracial non-Hispanic AI/AN respondents who do not identify as Black. Asian renters are monoracial non-Hispanic Asian and/or Pacific Islander respondents and multiracial non-Hispanic Asian and/or Pacific Islander respondents who do not identify as Black or AI/AN.

11. Couples include married and unmarried romantic partners. Parents refer to primary caregivers of children under 18 years of age.

12. Education was defined as less than high school, some high school, high school (diploma, GED, or equivalent), some college, associate’s degree (diploma or vocational certificate), bachelor’s degree, and graduate degree.

13. Both education and income were standardized and then their mean was calculated. We then used a natural logarithm transformation to adjust for the rightward skew. Supplemental models with educational attainment and annual household income as separate variables produced comparable results. We present the combined measure because it is more parsimonious.

14. Since some landlords include utilities (e.g., water) in rent while others do not, the AHS creates a comparable variable by adding together rent, electricity, natural gas, oil, wood, coal, kerosene, or other fuel, water and sewage, and garbage and trash collection for all respondents. We used the RHFS variable discussed above for the property’s potential rent. We applied a square root transformation to this ratio to adjust for its skew.

15. Single-family homes are all detached single unit dwellings. A square root transformation was used to adjust for the total population rightward skew.

16. Renters in communities of color are slightly more likely to live in multiunit complexes and have professionally managed units than renters in White neighborhoods. Yet, these differences are statistically insignificant and substantively small, especially compared to the notable differences in neighborhood and renter demographics.

17. Concentration is defined as a resident’s own racial group consisting of a larger proportion of their neighborhood than their racial group’s proportion in the U.S. population.

18. See also Denetdale (2016) and Huyser et al. (2018), who find that AI/AN people who live away from their reservation lands and the supports these lands provide experience particularly acute consequences of settler colonialism.