Abstract
In recent decades, the racial wealth gap has widened with extant literature reporting that Black and Latinx families hold fewer assets than white families. One such asset that receives substantial attention because of its wealth-generating principles is homeownership. Whereas intergroup homeownership inequalities are found throughout the literature, less is known about racialized inequality within groups. Latinxs provide a novel case for exploring how racialized homeownership inequality is structured within an ethnic group. Using data from the American Community Survey, we examine the odds of homeownership and predicted logged home values among Latinxs. We find that the association between race and housing outcomes varies substantially across Latinx groups. Drawing from theories of Latinx racial identity and the future of racial structures, we discuss the implications of our findings for understanding racial inequality among Latinx groups.
Introduction
In a recent study by the Urban Institute (2017) measuring the “powerful wealth-building tool” of homeownership, a disturbing but not unsurprising longitudinal chasm was documented between white and Black and Latinx families. 1 In 1976, 68% of white non-Latinx families owned their home, while 44% and 43% of Latinx and Black families did so, respectively. By 2016, the homeownership gap narrowed marginally for Latinxs but widened for Black families. Nevertheless, both groups reported under 50% and remained nowhere near that of white families. The study underscored the alarming ways that racial-ethnic inequality is a defining characteristic of housing in the United States. However, for Latinxs—an ethnic group shaped by ancestry, immigration, and racial identities and experiences—such broad analytical categories are often not disentangled in ways that reveal intra- and inter-group hierarchies. This is especially true when examining homeownership’s role in the racial wealth gap. In this paper, we examine how ethnicity and race shape Latinx homeownership and housing values in the United States.
The extant literature on housing inequalities identifies vast inequalities along racial and ethnic lines in terms of homeownership (Alba and Logan, 1992; Coulson, 1999; DeSilva and Elmelech, 2012; Kuebler, 2013; Kuebler and Rugh, 2013; Sánchez, 2019; Taylor, 2019), home values (Howell and Korver-Glenn, 2018; Taylor, 2019; Tesfai, 2016), and access to quality housing (Friedman and Rosenbaum, 2007; Taylor, 2019). White non-Latinxs’ access to affordable, quality housing with reasonable mortgage terms sit in stark contrast to the predatory inclusion methods used against Black and Latinx households (Seamster and Charron-Chénier, 2017; Taylor, 2019). 2 However, less is known about how the interaction between race and ethnicity shapes housing outcomes and how the case of Latinxs reinforces or challenges existing anti-Black/pro-white patterns of housing inequality. Latinx groups, who present a full spectrum of racial categorizations and racialized experiences, can be looked to for understanding racial-ethnic interactions and addressing theorizations about the future of racial structures (Bonilla-Silva, 2006; Ray and Seamster, 2016; Seamster and Ray, 2018).
In this paper, we use pooled data from the American Community Survey 2013–2017 to analyze a sample of Latinxs between 25 and 65 years old. Using logit and Heckman regression models, we examine their likelihood of homeownership and predicted home values while controlling for life course characteristics and socioeconomic status. Our findings indicate that both race and ethnicity significantly predict the likelihood of owning a home and their home values. Compared to white Latinxs, all other “race” groups are less likely to own their homes and have lower home values. However, using predicted values, we show that the association between race and housing outcomes varies substantially between ethnic groups for white Latinxs but less so for Black Latinxs, who likely experience a more uniform (discrimination-based) housing market than white Latinxs.
Our findings indicate that racial and ethnic inequality are defining characteristics of housing. Racism, in particular, stands out as a central defining characteristic of Latinx housing, in that when Latinxs are disaggregated, evidence suggests a reinforcement of anti-Black/pro-white racial hierarchies. We also show how racist housing inequality is contextual and may not be fully captured in currently available self-reported data. While our findings clarify the role of racism among Latinx housing status, they also provide additional support for employing measures of “street race-gender” (López, 2014) in quantitative and mixed-method studies of inequality, whereas the difference between how one sees themselves and how others treat them in everyday experiences is underscored. In what follows, we outline how race and racism shape housing outcomes and the debates surrounding Latinx racial identity. We then describe our data and methods before discussing our findings. We conclude with a discussion of the theoretical and practical implications of our findings and highlight potential avenues for future research.
How Race and Racism Shape (Latinx) Housing Outcomes
In the literature of racial inequality, homeownership receives a significant amount of attention for its contribution to the racial wealth gap (Famighetti and Hamilton, 2019; Herbert et al., 2013; Kuebler, 2013; Markley et al., 2020). Unlike renting, owning a home is expected to provide stable housing that appreciates over time and may provide a financial cushion in times of economic recession. Although homeownership is not the only driver of the wealth gap (Darity Jr. et al., 2018), it remains an important aspiration among Black and Latinx families (McCabe, 2018). It also represents the most commonly owned asset among Black and Latinx immigrant households (however, the proportions pale compared to that of white non-Latinxs; Painter II and Qian, 2016a, 2016b). 3
Explanations for the causes of racial homeownership inequality vary, though it is best understood through a systemic approach that acknowledges how race/racism are central organizing principles of the society and its distribution of resources (i.e. racial projects; Bonilla-Silva, 1997, 2006; Carmichael and Hamilton, 1967; Crenshaw, 1991; Davis, 1983; Feagin, 2006; Omi and Winant, 2014). Racist projects are intimately tied to private property—stemming back to the continent’s colonization, and in the case of the United States, the Indian Removal Act of 1830 and the Homestead Act of 1862 (Omi and Winant, 2014). The US invasion of Mexico in 1846–1848, in between these acts, yielded mass transfers of property in the West and Southwest into white (non-Latinx) hands and allowed US settlers to expand the colonization of once-indigenous lands (Feagin, 2006; Gómez, 2008). The use of racist projects to organize unequal housing continued through the New Deal (Katznelson, 2005) and subsequent patterns of suburbanization and financing.
Throughout the mid-century, redlining and local restrictive covenants segregated Black, Latinx, and Asian families apart from white, and often wealthy, neighborhoods (Massey, 1990; Rothstein, 2017; Taylor, 2019). Predatory lending to these same segregated neighborhoods exacerbated racialized housing inequality (Taylor, 2019) and later underpinned the housing marking collapse during the Great Recession and Foreclosure Crisis (Massey et al., 2016; Rugh and Massey, 2010). Inequalities persist today with an increasing frequency of evictions (Desmond, 2016), appraisals (Howell and Korver-Glenn, 2018; Perry et al., 2018), and methods of upselling and steering (Besbris and Faber, 2017).
Previous findings indicate that most Latinxs are less likely to own their homes compared to whites (Cortes et al., 2007; Coulson, 1999; Sanchez-Moyano, 2020). This is related in part to the large immigrant population among Latinxs who likely lack the start-up capital for purchasing homes (Painter II and Qian, 2016a, 2016b; Tesfai, 2016), as well as their settlement patterns in immigrant destinations (Painter and Yu, 2014; Sánchez, 2019). Previous research also identifies differences across Latinx groups (Aja et al., 2019; Cahill and Franklin, 2013; De La Cruz-Viesca et al., 2016; Martinez, forthcoming). Nevertheless, a broad look at Latinxs shows how race and ethnicity structure their likelihood of homeownership (Rosenbaum, 1996).
The case of Latinx housing inequality is particularly useful for understanding how racism shapes housing because of the varied ideologies underpinning racialization of and racial identification among Latinxs. Anti-Blackness and anti-indigenous ideologies inform how Latinx groups identify, both in the United States and their countries of origin (Darity et al., 2005; Sue, 2013; Telles, 2012). These ideological stances are reinforced by material resources, social status, and racialized experiences (Haywood, 2017; Hordge-Freeman and Veras, 2020; Logan, 2010; Rivera, 2006). Among Latinxs, racialization and categorization may also stem from racial schemes held by individuals (Roth, 2012). Two schemas in particular, the basic nationality and pan-ethnic nationality schemes, where nationalities and ethnicities are used in place of racial categories, may explain why previous research identifies inequality across nation-origin Latinx groups. However, Latinxs can also be racialized by others as a distinct racial group outside of whiteness/Blackness (Celia Olivia Lacayo, 2017; Genova, 2006), a complex process anchored in anti-Blackness itself (discussed below). In the home buying process, real estate actors rely on stereotypes related to citizenship, income, family composition, and financial knowledge to exclude prospective Latinx homebuyers and submit them to harsher, more predatory financing terms (Howell and Korver-Glenn, 2018). Racial identity and racialization are also a product of the racial contexts in which Latinx groups find themselves (Roth, 2012). For example, Cuban-American identity is shaped in part by their residence (or lack thereof) among other Black and Latinx groups (Aja, 2016; Hay, 2009; Newby and Dowling, 2007). The resulting “mismatch” between racialization and identity among Latinxs creates issues for “measuring” race and racism using secondary data analysis.
The racialized structure of Latinx housing inequality is important to study because of its implications on the future of racial stratification. Theorists of race and racism have broadly posited three possible futures of inequality. First, the tri-racial order, or Latin Americanization thesis, argues that post-1965 immigration modified a white-Black binary with a middle group of “honorary whites” (Bonilla-Silva, 2002). This theorization places groups, including various Latinx subgroups, within three “loosely organized racial strata,” equally shaped by social race (Bonilla-Silva, 2006, 2002). For instance, Afro-Latinxs may be racialized in a larger “collective Black” category alongside African Americans, Afro-Caribbean, and racialized South Asian groups like Filipinos and Vietnamese. While white assimilated Latinxs are categorized as white, their light-skinned co-ethnics, including “most Cubans and segments of the Mexican and Puerto Rican communities,” are considered “honorary whites” (Bonilla-Silva, 2002: 5, 2006). These distinctions also reflect differences in the racial and colonial contexts in which groups were incorporated into the U.S. racial strcuture, and how U.S. intervention in the Global South shaped the migration paths of many Black, Latinx, and Asian groups. Others have predicted that the future of racialized inequality is instead a strengthening of racialized boundaries that reinforce a Black/non-Black binary of inequality (Warren and Twine, 1997; Yancey, 2003). More recent literature suggests instead that examining racial inequality obscures how race operates as master status of material distribution (Seamster and Ray, 2018).
In light of these particularities of the Latinx case and extant literature on homeownership inequality, we hypothesize that both race and ethnicity shape housing tenure and home values among Latinxs (hypothesis 1), but that the ideologies undergirding racial identity obscure the extent to which racism shapes Latinx housing outcomes (hypothesis 2). Building on previous literature on the role of anti-Blackness in shaping Latinx identity, we also expect that Black Latinxs are least likely to own their home and have the lowest home values (hypothesis 3), but that their experiences are more similar to each other across ethnic groups when compared to other racialized Latinxs (hypothesis 4).
Data and Methods
Data for this research come from the Integrated Public Use Microdata Series American Community Survey (IPUMS-ACS) for the years 2013–2017. The ACS is a random sample of US households that accounts for approximately 1% of the national population. The data cover an extensive period following the foreclosure crisis and, because of the large and representative sample size, we can account for ethnicity and racial identity. Additional data at the Public Use Microdata Area (PUMA)-level (geographical units of approximately 6500 people) were merged into the full dataset to account for geographic differences in Latinx housing outcomes. We use a sample of head-of-households between the ages of 25 and 65 who identify themselves as “Hispanic” on the ACS. 4 The analytic sample includes 450,561 individuals with complete information.
Race and Ethnicity
Our data include information on self-reported race and ethnicity. Although previous research indicates that most Latinx groups identify as either white, Black, or “other” (Hogan, 2017; Logan, 2010), we include the full set of racial categories in our analysis. We do so to acknowledge the vast heterogeneity of Latinx groups. The categories include white (reference), Black, Native American/Indigenous, Asian, other, and Multiracial. Similarly, we include all ethnicity groups: Mexican (reference), Puerto Rican, Cuban, Dominican, Central American, South American, and Other. 5
Dependent and Independent Variables
The analysis calls for two dependent variables. The first, homeownership, is coded dichotomously where 1 indicates that the head-of-household owns their home and 0 indicates that it is rented. The second dependent variable is home values, which are logged transformed to account for their positive skew. In addition to race and ethnicity, our analysis includes demographic and socioeconomic characteristics. Individual demographic measures include age and its square and gender. Extant literature highlights the role of family structure in explaining homeownership outcomes (DeSilva and Elmelech, 2012). We also include select familial measures: marital status, number of children, and number of family members. Given the large immigrant population among Latinx groups, we code an interaction between immigrant status and years in the country with US-born as the reference category.
Socioeconomic measures include household income, educational attainment, and employment status. Income is coded in thousands of dollars, and the categorical education variable from the ACS is recoded as a continuous variable measured in years. Because immigrants and their children, who make up a large proportion of the Latinx population, often engage in self-employment (Robles and Cordero-Guzmán, 2007), we create a four-category employment measure. The categories include unemployed (reference), not in the labor force, self-employed, and wage worker.
We also account for geographical characteristics that shape housing outcomes. As such, we control for residence in a metro area and region of residence. Additionally, we measure four variables at the PUMA-level: the ratio of median income over median home value, the proportion of single-family homes, the percentage of homeowners, and the percentage of residents who are Black. Finally, we account for survey year in multivariate analyses.
Analytic Technique
The goal of this analysis is to examine the determinants of housing outcomes among Latinx groups accounting for both race and ethnicity. We first discuss the racial and ethnic distribution of Latinx groups in the United States. We then present summary statistics of the independent and dependent variables by ethnic group. To understand how these characteristics shape Latinx housing outcomes, we run a logistic regression and include an interaction term of race and ethnicity, and then a Heckman regression. Finally, we plot predicted likelihoods of homeownership and predicted logged home values, net of covariates.
Because characteristics among Latinx homeowners may explain their subsequent home values, we control for selection into homeownership in the Logit regression. We do so by including three variables: the number of children, the proportion of single-family homes, and the ratio of median income over median home value at the PUMA level. We then estimate the Inverse Mills ratio of the logit model, which accounts for selection into homeownership, and include it in the Heckman regression. The Heckman model also accounts for the number of rooms in a house as a proxy for the size of a home, the percentage of homeowners in a PUMA as a predictor of home values, and the percentage of Black residents in a PUMA.
Findings
The Racial Identities of Latinx Groups in the United States
We begin our analysis by presenting the racial distribution of select Latinx groups in the United States in Table 1. White self-identification represents the largest share of all groups but one group (Dominicans). White identification is highest among Cubans (88.05%) and South Americans (73.58%) and represents a majority among all other groups. Although Dominicans are less likely to identify as white (35.07%), no racial category represents more than half the Dominican population.
Racial Distribution of Latinx Groups.
Source: American Community Survey, 2013–2017. Note: Sample restricted to head-of-households aged 25–65 who indicated Latinx ethnicity.
Black identification among Latinxs is unsurprisingly low. Only Dominicans have a Black population larger than 10%. This value is less than 5% among all groups except Puerto Ricans (7.72%). The low rates may be related to the particular (im)migration histories of each group and the legacies and present manifestations of mestizaje and anti-Blackness operating within groups through state-led migration processes. For example, post-revolution (1958-) Black Cubans did not immigrate to the United States at the same rate as the white upper and middle classes from the island (Aja, 2016; Eckstein, 2009; Prieto, 2009). However, their migration to the United States is fairly substantial in recent years, particularly throughout the Southwest (Dowling and Newby, 2010; Gosin, 2017; Newby and Dowling, 2007). Instead, the low rate of Black identification may reflect the combined racialized experiences and white supremacist-situated ideologies of Latinxs in the United States and Latin America. Because race is experienced and thus contextual, Latinxs who may be racialized as Black immediately in one city may undergo a longer racialization process as Black in another (especially in localities with larger Black non-Latinx populations; see Logan, 2010). Previous findings highlight the racial mismatch between identification and racialization (López, 2014) and how they may vary by locality (Aja et al., 2019). However, identifying as Black among Latinxs cannot be removed from anti-Black ideologies, either from the United States or from Latin America (Darity et al., 2005).
The three smallest racial groups of Latinxs (Asian, Indigenous, and Multiracial) represent less than 10% of any group. Largest among the three are Multiracials, who may be the children of parents from different racial identities or identify through the lens of pan-ethnicity (DaCosta, 2007). Our data also show that Indigenous identification is consistently low among all groups, likely due to “Native American” being viewed as a US racial construct and the likelihood that indigenous Latinxs choose between “white,” “other,” or “Multiracial” as variant expressions of the middle-category, mixed-race-situated mestizaje (see discussion below and throughout). Although less is written about Asian Latinxs, which may reflect the lower rates in our data, Asian Latinxs are represented throughout the United States and challenge existing social norms of race, ethnicity, and migration (Ocampo, 2016; Ropp, 2000).
This general overrepresentation of whiteness stands in stark contrast to values for the five other racial categories. After white, “other” Latinxs are the second largest group among all ethnicities. Almost half of all Dominicans identify as “other” (43.37%), and nearly one-fifth or more of Mexicans, Puerto Ricans, Central Americans, and South Americans identify as “others.” Throughout Latin American and within select groups, ideologies of Latinidad and mestizaje provide nationalistic or pan-ethnic solidarity and identity. They also operate as colorblind distinctions of mixed-race populations, with implications for hidden anti-Blackness and anti-indigeneity. Given the growing number of second-, third-, and fourth-generation Latinxs in the general population, this observation may also reflect a move away from the Hispanic/Latinx label altogether (Lopez et al., 2017). One other possible explanation may be the racial categories themselves, which may not present the full experience of racism and racialization for Latinxs.
Previous research corroborates our observations and further explains their origin in perceptions and experiences with discrimination, skin tone, and a preference for whiteness rooted in white supremacist ideology (Darity et al., 2005, 2010; Golash-Boza and Darity, 2008; Mason, 1997, 2004). This is particularly true among Latinx “others”, the fastest growing segment of the US Latinx population (Golash-Boza and Darity, 2008; Logan, 2010), who may be expressing the collision of racialized hierarchies in Latin America, the Caribbean, and United States rooted in different stages of settler-colonialism and expansion into larger globalized presence (Bonilla-Silva, 2006). Identifying as “other” may also represent an immediate alternative to whiteness and Blackness among Latinx groups navigating racialization in the United States, generational differences, and material markers of social status (Logan, 2010: 200), especially for newer arrivals. Over time, though, the collective “flight to whiteness” (Darity Jr., 2016; Darity et al., 2005) is apparent. These findings support the need for more accurate methods for understanding race among Latinx individuals (López, 2014).
How Latinx Groups Compare
Summary statistics for the study measures are listed in Table 2 by Latinx identification. Less than half (45%) of Latinxs in the United States own their homes. Homeownership is highest among Cubans and Other Latinxs, and lowest among Dominicans, Puerto Ricans, and Central Americans. Although these findings are not disaggregated by race, they can be interpreted based on the previous discussion of anti-Blackness and short-term alternatives to whiteness in racialized identities. Groups in which people identify as Black at higher-than-average rates also have the lowest rates of homeownership. These preliminary findings suggest how race and racism shape the housing tenures of Latinx groups when observed through ethnicity.
Summary Statistics for Study Measures, by Latinx Group.
Source: American Community Survey, 2013–2017. Sample restricted to head-of-households aged 25–65 who indicated Hispanic ethnicity. Note: Mean home values reflect subsample of homeowners (N = 226,469).
Whereas homeownership varies substantially across Latinx groups, home values are more similar. Among the sample, average home values range between US$124,100 and US$118,200. Home values are highest among South Americans and Cubans and lowest among Mexicans and Other Latinxs. These patterns likely reflect both residential segregation among Latinx groups (Iceland and Nelson, 2008; Lichter et al., 2016) as well as their proximity to both white and Black neighborhoods (Howell and Korver-Glenn, 2020).
Moving to demographic characteristics, all groups are relatively similar in age and are in their early- to mid-forties. Similarly, the proportion of women within each group is relatively equal across the groups, though Dominicans are a notable exception. Marriage, however, ranges substantially across groups. Whereas half or more of all Mexicans and South and Central Americans are married, less than 40% of Puerto Ricans and Dominicans are married. This may be associated with the gendered composition and migration patterns of each group. Differences in marital rates are important in that they shape household incomes, which in turn could affect rates of homeownership. Across Latinx groups, the number of children and family size do not range substantially and average one child and three family members. More than half of the Latinx population is foreign-born. This rate is significantly lower among Latinx “Others” and likely reflects pan-ethnic identification in later generations (Lopez et al., 2017).
Summary statistics of socioeconomic measures again show the vast heterogeneity of Latinx groups. Household incomes average US$65,000, though South Americans (84.42) and Cubans (79.17) have higher than average incomes. On average, Latinx groups have 11.8 years of education, roughly equivalent to a high school diploma. We note that educational attainment does not seem to have a universal association with household incomes. For example, Dominicans have more years of education than Mexicans but have lower incomes. Given previous findings linking earnings and race/racism, these findings likely reflect the fact that Dominicans have a higher share of Black people compared to other groups. On a final note, we find that although self-employment is typically related to Cubans in the literature, Central and South Americans have similar rates of self-employment to them.
Regression Findings
Results from our regression models are listed in Table 3. In support of hypothesis 1, race and ethnicity significantly shape homeownership and logged home values among Latinx groups. Compared to white Latinxs, the likelihood of homeownership is 37.1% lower among Black Latinxs, 16.4% lower among Indigenous Latinxs, 5.5% lower among Other Latinxs, and 9.9% lower among multiracial Latinxs. Similarly, white Latinxs have higher logged home values than Black, Indigenous, Other, and multiracial Latinxs. Asian Latinxs, however, are not significantly different from white Latinxs in both homeownership odds and predicted logged home values.
Logit and Heckman Regressions Predicting Housing Outcomes Among Latinx Groups.
Source: American Community Survey, 2013–2017. Sample restricted to head-of-households aged 25–65 who indicated Hispanic ethnicity (N = 450,561). Race-ethnicity interactions are only included in Logit regression and not shown in the table. Heckman model predicting logged home values reflects a subsample of homeowners (N = 226,469).
*p < .05; **p < .01; ***p < .001.
Despite restricting our sample to Latinxs, our findings align with previous research and support hypothesis 3, showing that Black households have the lowest homeownership rates, whereas Asian households are no different from white households (Alba and Logan, 1992; DeSilva and Elmelech, 2012). While the latter finding may be related to the small number of Asians in the sample, the former likely reflects the effects of ownership in a racist housing market. Although our findings on home values somewhat mirror previous research, they also show that Indigenous Latinx groups have the lowest home values. Our findings highlight the salience of race/racism in explaining housing inequality among a sample of Latinxs. They show how race operates as a master status shaping the distribution of material resources and that racism is a fundamental cause of homeownership inequality (Omi and Winant, 2014; Ray and Seamster, 2016; Seamster and Ray, 2018).
Regression results on ethnic inequality reveal that Puerto Ricans, Dominicans, and Central Americans are less likely to own their homes than Mexicans, while Cubans, South Americans, and other groups are more likely to own their homes. These findings are not mirrored in the Heckman model. Instead, we show that all Latinx groups have higher logged home values than the reference group. Each group’s housing tenure status is likely related to the group’s racial makeup, while housing values likely reflect the neighborhood they live in (Howell and Korver-Glenn, 2018, 2020; Taylor, 2019). However, each group has encountered different socio-political contexts in the United States and has a unique history of colonial subjectivity to migration that also affect their housing status. For example, compared to Cubans, Puerto Ricans, and Dominicans experienced distinct processes of racialization and othering in the United States that shaped their social and material status in the country (Aja, 2016; Candelario, 2007; Grosfoguel, 2003; Hay, 2009; Roth 2012). These differences in experience and material status are likely reflected in the patterns observed.
For example, the low likelihood of homeownership among Puerto Ricans may be related to their treatment in the United States as outsiders despite citizenship status and ability to apply for US aid (Aranda, 2007; Gónzalez, 2011). Compared to Mexicans, they have a shorter history in the United States and can have vastly different reasons for migrating to the mainland. However, evidence is clear that both groups experienced discrimination by state-level programs (like the Federal Housing Authority (FHA)) that disparately provided access to home ownership to non-Latinx whites (Feagin, 2006; Rosales, 2011, 2017). Similarly, Dominicans, who are less likely to own their homes than Mexicans, have had a shorter history of migration to the United States. Moreover, their citizenship status likely presents barriers to purchasing a home not experienced by Puerto Ricans. While Central Americans are also affected by hostile US immigration policy, they may also be affected by lower rates of selectivity (Feliciano, 2005) and by systemic violence by deportation and incarceration in the United States (Menjívar and Abrego, 2012). We note that Central Americans migrating from different countries likely have different housing statuses and that these findings obscure some of the heterogeneity within the group.
Notably, Cubans and Other Latinxs are more likely to own their homes compared to Mexicans. Cuban homeownership may be related to racialized government support and high selectivity among early waves of migration (Aja, 2016; Eckstein, 2009; Martinez, forthcoming; Masud-Piloto, 1996). Alternatively, high homeownership among Other Latinxs may be explained by their generational status. Second-, third-, and fourth-generation Latinxs are less likely to identify as Hispanic/Latinx (Lopez et al., 2017), and their incorporation into the United States may play a role in obtaining homeownership.
In addition to race and ethnicity, we find evidence that demographic and socioeconomic characteristics shape housing outcomes among Latinx groups. Age and marital status are highly predictive of homeownership and logged home values, emphasizing that homeownership is an aspect of the life course (DeSilva and Elmelech, 2012). We also find that US-born Latinxs are more likely to own their homes than immigrants who migrated less than 20 years ago. However, immigrants who have lived in the United States for longer than 20 years are more likely to own their homes than the US-born. On average, immigrants are also more likely to have lower home values compared to the US-born—a pattern that diminishes in strength as their time in the United States increases. Like marital status and family size, years in the United States likely reflects the importance of understanding homeownership as a life course event.
In line with previous research, we find that measures of socioeconomic status are positively and significantly related to the likelihood of owning a home and logged home values (Alba and Logan, 1992; DeSilva and Elmelech, 2012; Tesfai, 2016). Homeownership odds increase by 1.1% for each additional thousand dollars in household income and 5.9% with each additional year of education. Similar associations exist with logged home values, though the magnitude is smaller. Notably, self-employed Latinxs, compared to wage workers and those who are unemployed or not in the labor force, have the highest likelihood of owning their home and highest logged home values. This may be related to the higher than average incomes associated with self-employment among select groups (Martinez, forthcoming; Portes and Martinez, 2019).
Geographic characteristics significantly predict housing outcomes. For example, Latinxs residing in metro areas are 20.3% less likely to own their homes than those outside of metro areas. However, residing in a metro area is associated with higher home values. These findings likely relate to housing affordability and availability across different geopolitical spaces. Across the country, homeownership and logged home values vary substantially—a pattern that may be explained by regional histories. For example, homeownership in the West South regional division is likely driven by the large Mexican population that resides in Texas. In contrast, home values in the Middle Atlantic and Pacific regions are driven by expensive cities in New York and California. The final controls, survey years, show that all years are lower than 2013, indicating a worsening outlook for Latinx homeownership.
Finally, selection into homeownership and neighborhood characteristics significantly predict housing outcomes. The likelihood of being a homeowner is negatively associated with unaffordability and larger families and positively associated with residence in areas with more single-family homes. Logged home values increase approximately 11% with each additional room but are negatively associated with larger Black populations and higher rates of homeownership. The former results reflect how home values are socially constructed based on the racial composition of neighborhoods—a persistent pattern linked to wealth inequality (Howell and Korver-Glenn, 2020; Taylor, 2019).
Predicted Likelihoods of Homeownership
To better understand how race and racism shape Latinx housing outcomes, we included an interaction between race and ethnicity in the Logit regression. Below, we plot the predicted likelihood of homeownership using the interaction between race and ethnicity. Figure 1 clearly illustrates that white Latinxs have the highest likelihood of homeownership, regardless of their Latinx ethnicity. 6 These findings further illuminate the power of whiteness in shaping increased access to material resources (Omi and Winant, 2014). However, the varying degrees to which whiteness matters for different groups is a novel finding. For example, among white Latinxs, predicted homeownership ranges between 30% and 60%. White Cubans and “other” Latinxs have the highest rates of homeownership, while white Dominicans have significantly lower rates. Nevertheless, compared to Black and “other” Dominicans, white Dominicans are more advantaged.

Predicted Values of Homeownership Likelihood by Race Among Latinx Groups.
Figure 2 illustrates predicted logged home values based on results in the Heckman regression. Because the model uses a subsample of homeowners, the number of Latinxs within each racial group decreased substantially. We account for this by excluding the interaction term between racial and ethnic groups. 7 Unlike our homeownership results, home values vary at similar degrees for white Latinxs and Black Latinxs. As noted previously, these patterns likely reflect the racial composition of neighborhoods in which Latinx groups live. Notably, South Americans among all groups have higher home values, while Mexicans have the lowest.

Predicted Logged Home Values by Race Among Latinx Groups.
These findings support hypothesis 2, in that the ideologies undergirding racial identity obscure the extent to which racism shapes Latinx housing outcomes. For some Latinx groups, whiteness may reflect how they are racialized in the United States or their countries of origin (Darity et al., 2005; Golash-Boza and Darity, 2008; Vargas, 2015). For example, elaborate racial schemas exist in Brazil (Hordge-Freeman, 2013, 2015; Telles, 2012, 2014), and the processes of racialization in the United States are highly contextual (Hordge-Freeman and Veras, 2020; Roth, 2012; Soto-Márquez, 2019). Our findings show how these ideologies and experiences are intimately tied to the material realities of resource distribution and how race operates as a master status in housing inequality (Ray and Seamster, 2016; Seamster and Ray, 2018). Moreover, the varied ways that whiteness matters highlight the contextualization of race and racism and merits better measures of race and its experience (López, 2014).
Our findings also illustrate the comparative lack of heterogeneity among Black Latinxs and offer mixed support for our final hypothesis. Black Dominicans and Puerto Ricans are less likely to own their home than other Black Latinxs, though the magnitude of those differences is vastly smaller when compared to white Latinxs. This may be due to their long-time racialization in the United States (they have the largest Black populations of all Latinxs groups analyzed, however under-reported), and their spatial and social proximity to African Americans. The concentration of Black Latinx homeownership rates may again reflect race as a master status and racism in the housing market as a fundamental cause of homeownership inequality (Ray and Seamster, 2016; Seamster and Ray, 2018). Predatory inclusion into homeownership does not necessarily guarantee that houses hold universal wealth-generating potential, but instead are racialized commodities (Markley et al., 2020; Seamster and Charron-Chénier, 2017; Taylor, 2019). Considered alongside the vast heterogeneity of Other and Multiracial Latinxs, findings on Black Latinxs indicate that anti-Black racism, and racism more generally, are necessary approaches for understanding housing inequality among Latinxs.
Whereas homeownership odds are moderately concentrated among Black Latinxs, they range substantially more in terms of home values. This heterogeneity among Black Latinxs likely reflects how home values are socially constructed based on the racial composition of neighborhoods (Howell and Korver-Glenn, 2020; Taylor, 2019). As such, the home values illustrated here are subject to the levels of segregation each group is subjected to within their respective neighborhoods. Nevertheless, in all cases except Central Americans, Black–white inequalities remain significant.
Conclusion
In this study, we examined how race and ethnicity shape homeownership and home values among Latinxs in the United States. Building on our insights from the literature of Latinx racial ideologies, we expected that self-identified measures of race would obscure the extent to which racism shapes Latinx housing tenure. More specifically, we hypothesized that anti-Blackness would provide a more precise theoretical lens to interpret homeownership inequality among Latinxs. We uncover three significant findings in our study and relate them to their theoretical and practical implications.
First, measures of both race and ethnicity shape Latinx homeownership. These findings build on previous research (Aja et al., 2019; Rosenbaum, 1996; Martinez, 2020), showing similar findings. However, our results are novel in that they show how racial housing inequality persists after accounting for detailed racial and ethnic identification. As such, we show that race shapes housing outcomes within ethnicity and that ethnicity shapes housing outcomes within race. Because Latinidad can be racialized (Celia Olivia Lacayo, 2017), and thus acted upon as a distinctly non-Black and non-white category, our findings point to the importance of centering anti-Black racism in research on Latinx inequality, as well as expanding survey “measures” of race/racism.
Second, the role that race and racism play for Latinxs is highly contextual and is often obscured by patterns of white identity. The process of identifying as white among Latinxs is likely shaped by various material and immaterial factors. For example, ideological stances rooted in anti-Blackness and anti-Indigenousness may lead Latinxs to identify as white, even if they are racialized in other ways (Darity et al., 2005; Golash-Boza and Darity, 2008). Previous research indicates that identification can diverge from racialization in ways that artificially inflate (white) racial distributions (Vargas, 2015; Vargas and Stainback, 2016). In other words, survey data may reflect how Latinxs feel about their racial status and not how they are racialized by others. These findings have practical implications for the ways that race is “measured” in survey data, but also in the ways that researchers interpret racialized findings.
Finally, we show that the homeownership status of Black Latinxs is relatively consistent across ethnic groups. This pattern contrasts with all other race groups, who exhibit significantly more heterogeneity across ethnic groups. Given these findings, we argue that anti-Black racism and racism as a fundamental cause may be better approaches for understanding homeownership inequality among Latinxs. Nevertheless, the overall racial structure we illustrate reflects the theorized future of inequality (Bonilla-Silva, 2002). Because self-reported identification among Latinxs is subject to their relations with colonial pasts/presents, racist ideologies, and material statuses, relying on secondary data analysis can obscure the lived experiences of racism in the housing market for Latinx ethnic groups. Thus, understanding racialized inequalities among Latinxs through the lens of anti-Blackness and fundamental causes would help identify the mechanisms that create and sustain these inequalities.
The overall implications of our findings are twofold. First, our results show that racism and racial inequality are defining characteristics of homeownership and home values among Latinxs. Thus, any policy-level approach that attempts to close housing inequality gaps or equitability distribute housing must consider existing patterns of ethnic and racial inequality. Second, quantitative surveys should continue to collect data on race and ethnicity so that changes in racial-ethnic inequality can be better understood across time and space. However, because self-identified race can obscure racialized experiences in the housing market, we argue that future research and surveys seriously consider using additional measures of race and racialization. The importance of multiple racialization measures is most clear in the predicted likelihoods of white Latinx housing outcomes, which significantly vary despite identifying as one group. While some of these differences may be related to group characteristics of ethnic groups, they are likely shaped by a preference for whiteness among Latinx groups (Darity et al., 2005). A more accurate approach to understanding how race and racism shape housing outcomes would be to ask respondents how they believe others identify them at the everyday “street” level (López, 2014). Such an approach better contextualizes how race “operates” in everyday “market” life in shaping housing status and can inform interventions crucial to deter anti-Black discrimination.
We began this project with the intent of exploring how both race and ethnicity shape Latinx homeownership. We found clarity in understanding race and racism’s centrality in shaping housing inequality for Latinxs. However, as we detail, there is substantial ambiguity about the racialized experience of Latinxs. This is partly because of the lack of data on “street race”—a limitation of our study and the use of the particular data—and because of the varying ideologies that underpin racial identity among Latinx groups. Nevertheless, we conclude our study hoping that future research centers on race as a fundamental cause of housing inequality and aims to employ improved “measures” of race, racism, and racialized experiences.
Footnotes
Acknowledgements
The authors would like to thank Tamara Nopper, the anonymous reviewers, and editor David Fasenfest for their support, comments, and suggestions throughout our writing and reviewing process.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
