Abstract
Research on how neighborhood racial composition affects where gentrification unfolds yields mixed conclusions, but these studies either capture broad national trends or highly segregated cities. Drawing on the case of Seattle—a majority-white city with low segregation levels and growing ethnoracial diversity, this study uncovers an underexplored mechanism shaping patterns of uneven development and residential selection in the contemporary city: immigrant replenishment. The share of all minorities is negatively associated with gentrification during the 1970s and 1980s, and, in contrast to expectations, shares of blacks positively predicts recent gentrification while shares of Asians negatively predicts it. Increased concentrations of recent immigrants in neighborhoods with greater shares of Asians explain these relationships. These findings suggest that where arriving immigrants move limits residential selection in gentrification and shifts pressures to low-cost black neighborhoods. This study highlights how immigration and points of entry are important factors for understanding uneven development in the contemporary city and has implications for the future of racial stratification as cities transform.
Keywords: gentrification, race and ethnicity, immigration, neighborhoods, Seattle
The ethnoracial composition of neighborhoods is an important factor shaping residential preferences in the US and contributing to the persistence of residential segregation (Massey and Denton, 1993; Charles, 2003). Despite the importance of race in the development of residential patterns, scholarship on how neighborhood ethnoracial composition relates to which low-income neighborhoods gentrify—or experience a socioeconomic upgrading1—yields mixed conclusions. Some scholarship depicts gentrification as synonymous with the transformation of predominantly minority neighborhoods, primarily by higher-class whites—an increasingly common depiction by the media and popular discourse (Brown-Saracino, 2017). Others, however, illustrate its racially selective patterning reflecting the avoidance of such neighborhoods (for a review of this debate, see Brown-Saracino [2017]).
Prior theory and research on the relationship between race and gentrification extend from either national-level analysis or neighborhoods in the advanced stages of gentrification and in highly segregated settings, such as Chicago, New York City, Philadelphia, and Washington, DC (Brown-Saracino 2017). In such places, race and class correlate strongly, and racially mixed neighborhoods are less prevalent, constraining neighborhood change and residential mobility (Charles, 2003; Crowder, Pais, and South, 2012). Further, despite consensus in the literature that gentrification in recent decades exhibits distinct characteristics from gentrification of the past (Hackworth and Smith, 2001), research advancing theory on race and gentrification often do not consider these temporal differences. Understanding this relationship and its mechanisms is important because it has implications for racial and ethnic inequality as gentrification continues to spread to more neighborhoods and cities.
This study draws on the case of Seattle to assess how neighborhood ethnoracial composition affects where gentrification unfolds across several decades. Seattle’s uniqueness relative to the cities typically studied in research on gentrification and residential stratification provides an opportunity to test assumptions based on other settings to advance theory on gentrification (Small, 2009). Seattle, which is experiencing widespread gentrification, is historically a predominantly white city with relatively low segregation levels compared to other large cities2 and thus has few predominantly minority neighborhoods. These conditions may affect residential selection and neighborhood preferences in distinct ways from highly segregated cities, where concentrated poor and minority neighborhoods are prevalent and persistent (Sampson, 2012). In recent decades, Seattle has also become increasingly racially and ethnically diverse because of rising immigration: Its share of non-Hispanic whites declined from 86% to 67% from 1970–2013.3 This feature of Seattle provides an opportunity to examine how the increasingly multiethnic nature of cities and neighborhoods associated with rising immigration may affect contemporary patterns of gentrification.
My findings of where gentrification takes place in Seattle reveal distinct racialized patterns of gentrification over time and uncover an underexplored mechanism—immigrant replenishment—that shapes how gentrification and, more generally, residential selection unfolds in cities today. I find that the share of all minorities is negatively associated with gentrification during the 1970s and 1980s, but the share of blacks has a positive relationship and share of Asians has a negative relationship with gentrification during the 1990s and 2000s. These relationships are partially consistent with existing scholarship on early periods of gentrification and in highly segregated cities, which reflect minority avoidance rather than preferences for diversity (e.g., Spain, 1980; Smith, 1996), but the findings from the recent wave are inconsistent, revealing new processes characterizing neighborhood change in the twenty-first century. What explains these distinct patterns in Seattle? After presenting these results, I examine various explanations in the remainder of the analysis.
The findings suggest that where arriving immigrants, who are predominantly Asian in Seattle, move limits points of entry for gentrification, constraining gentrification’s expected pathway and shifting pressures to low-cost black neighborhoods, which are primarily comprised of African-Americans. This study enhances theory on neighborhood change and residential selection by highlighting the roles of available points of entry and immigration, which studies of gentrification and neighborhood stratification often do not consider. The results suggest that increased immigration to a city with a tight housing market may have unintended consequences on black urban neighborhoods. Because black urban residents may disproportionately face displacement and subsequent disadvantages on the housing market, the findings have implications for the future of racial stratification.
GENTRIFICATION AND RACE
Gentrification is a process by which low-income central city neighborhoods undergo reinvestment and renewal and experience an in-migration of middle- and upper-middle class residents (Smith, 1998, p. 198). Debates between supply- and demand-side explanations of gentrification once dominated scholarship on gentrification, but this debate has become less central as scholars increasingly agree that both the production of capital and preferences are necessary parts of the gentrification process (Zukin, 1987; Zukin, 2016; Brown-Saracino, 2017). While the former emphasizes the production of places that meet the preferences of middle- and upper-class residents, the latter focuses on gentrifiers’ preferences; both sides, nonetheless, point to the importance of residential preferences for understanding patterns of uneven development (Zukin, 1987; Hamnett, 1991). Gentrification thus entails a process of neighborhood selection, in which a variety of actors—including individual households, commercial businesses, state and corporate entities, and other institutions—make decisions to invest in a low-income neighborhood, resulting in the emergent increase in the value of the neighborhood and further attracting middle- and upper-class residents (Hwang and Sampson, 2014).
Although the popular image of gentrification is one in which white gentrifiers establish residence in minority, particularly African-American, neighborhoods (Brown-Saracino, 2017), this pattern is more nuanced. Studies find that gentrification has been historically unlikely to occur in black neighborhoods compared to white or racially mixed neighborhoods (Spain, 1980; Smith, 1996; Wilson and Grammenos, 2005; Freeman, 2009; Owens, 2012; Delmelle, 2016; Hwang, 2016a; Landis, 2016). Smith (1996) attributes this aversion to black neighborhoods to the strength of negative reputations surrounding black poverty and public housing. Studies on residential segregation show that people associate areas with large shares of blacks, and sometimes Latinos, with low neighborhood quality and high levels of crime and disorder, leading them to avoid minority neighborhoods (Ellen, 2000; Quillian and Pager, 2001; Sampson and Raudenbush, 2004; Krysan et al., 2009; Lewis, Emerson, and Klineberg, 2011; Anderson, 2012). Several ethnographic accounts depict white gentrifiers as attracted to racially and ethnically diverse—particularly non-white—neighborhoods instead of homogenous minority neighborhoods, desiring to avoid the homogeneity of white, affluent suburbs and searching for “authenticity” or “grit” (Mele, 2000; Berrey, 2005; Lloyd, 2006; Brown-Saracino, 2009; but see, Bader, 2011). Others depict processes of boundary maintenance in gentrifying neighborhoods adjacent to predominantly African-American neighborhoods (Anderson, 1990; Douglas, 2012; Hwang, 2016b).
While many scholarly accounts depict gentrification occurring in predominantly minority neighborhoods like Harlem or the Lower East Side in New York City and Bronzeville or Pilsen in Chicago, these neighborhoods are either initially predominantly Latino neighborhoods or undergo gentrification by middle-class blacks before the arrival of white gentrifiers (Smith, 1996; Mele, 2000; Taylor, 2002; Bostic and Martin, 2003; Dávila, 2004; Pérez, 2004; Wilson and Grammenos, 2005; Lloyd, 2006; Pattillo, 2007; Boyd, 2008; Hyra, 2008, 2017; Moore, 2009; Freeman, 2011; Wherry, 2011; Anderson and Sternberg, 2013). Anderson and Sternberg’s (2013) comparative analysis of a predominantly African-American neighborhood and a predominantly Latino neighborhood in Chicago illustrates the relatively higher barriers to reinvestment that the predominantly black neighborhood faces due to racialized conceptions. The large socioeconomic gaps between racial groups also explain the relative rarity of gentrification in predominantly black neighborhoods (Brown-Saracino, 2017).
Broader trends on neighborhood trajectories also show that white influx, socioeconomic upgrading, and racial turnover occur in a small share of predominantly black neighborhoods and that most poor black neighborhoods remain poor and predominantly black over time (Logan and Zhang, 2010; Owens, 2012; Sampson, 2012; Freeman and Cai, 2015; Landis, 2016; Timberlake and Johns-Wolfe, 2017; Owens and Candipan, 2018). Further, Hwang and Sampson’s (2014) analysis of neighborhoods in Chicago that were either gentrifying in the 1990s or were adjacent to gentrifying neighborhoods finds that neighborhoods more than 40% black were far less likely to experience continued socioeconomic upgrading. They contend that a racial hierarchy of preferences that operates within general patterns of neighborhood selection in the US—such that people favor white neighbors the most, black neighbors the least, and Asian over Latino neighbors in the middle—also operates in gentrification, reflecting a limit to preferences for ethnoracial diversity among gentrifiers (Hwang and Sampson, 2014). Consistent with this argument, Berrey’s (2005) ethnographic study of a racially and ethnically diverse neighborhood in Chicago finds that gentrifiers state that they value diversity but prefer a limited share of minorities.
Altogether, past research suggests that race-based residential preferences—a well-known factor influencing neighborhood selection and contributing to segregation—is an important process shaping where gentrification takes place within a city. While studies depicting single neighborhoods often focus on gentrification in minority neighborhoods, broader trends show that gentrification is least likely in predominantly black neighborhoods compared with other neighborhood compositions. Higher-class residents, as well as state and corporate entities drawing on such preferences to attract higher-class residents, avoid predominantly black neighborhoods and exhibit limited preferences for diversity. Nonetheless, the geographic scope of these existing studies that inform our understanding of gentrification in minority neighborhoods is limited.
Unlike Seattle, the cities in these studies contain larger numbers of predominantly minority neighborhoods, which are more prevalent in cities with high segregation levels and relatively large African-American populations, and have few racially integrated neighborhoods. On average, these neighborhoods have deteriorated housing, higher crime rates, and lower quality schools, leading residents with greater socioeconomic ability to avoid them (Massey and Denton, 1993; Charles, 2003). Thus, segregation likely constrains the degree to which gentrification takes place in racially mixed or minority neighborhoods. Indeed, Freeman and Cai (2015) find that white influx into majority black neighborhoods is more likely in less segregated cities. Table 1 provides a summary of the studies reviewed above on gentrification occurring in minority neighborhoods, including their location and the ethnoracial compositional changes depicted. The literature presented here is by no means exhaustive but illustrates the geographic limitations of this research.
Table 1.
Summary of Literature on Gentrification in Minority Neighborhoods
| Location | Study | Setting or Findings Relevant to Racial/Ethnic Composition and Gentrification |
|---|---|---|
| New York, NY | Smith (1996) | • Setting (Harlem): black neighborhood, black gentrifiers |
| Mele (2000) | • Setting (Lower East Side): Latino neighborhood, white gentrifiers | |
| Taylor (2002) | • Setting (Harlem): black neighborhood, black gentrifiers | |
| Dávila (2004) | • Setting (East Harlem): Latino neighborhood, white gentrifiers | |
| Freeman (2011) | • Setting (Harlem and Clinton Hill): black neighborhood, black gentrifiers | |
| New York, NY & Chicago, IL | Hyra (2008) | • Setting (Harlem and Bronzeville): black neighborhood, black gentrifiers first, white gentrifiers later |
| Timberlake and Johns-Wolfe (2017) | • Finding: % black predicts gentrification by blacks but not by whites in Chicago, % Latino predicts gentrification by whites in New York City | |
| Chicago, IL | Pérez (2004) | • Setting (Near Northwest Side): Latino neighborhood, white gentrifiers |
| Berrey (2005) | • Setting (Andersonville)/Finding: mixed-race neighborhood, residents claim to value diversity but prefer a limited share of minorities | |
| Wilson and Grammenos (2005) | • Setting (Humboldt Park): Latino neighborhood | |
| Lloyd (2006) | • Setting (Wicker Park): Latino neighborhood, white gentrifiers | |
| Pattillo (2007) | • Setting (North Kenwood-Oakland): black neighborhood, black gentrifiers | |
| Boyd (2008) | • Setting (Bronzeville): black neighborhood, black gentrifiers | |
| Bader (2011) | • Finding: white residents do not consider racial mix as important for preferences for redeveloped neighborhoods | |
| Douglas (2012) | • Setting (West Town): Latino neighborhood adjacent to black neighborhood, white gentrifiers | |
| Sampson (2012) | • Finding: neighborhood hierarchy by poverty and % black is durable over time | |
| Anderson and Sternberg (2013) | • Setting (Bronzeville and Pilsen)/Finding: black neighborhood (Bronzeville) and Latino neighborhood (Pilsen), black neighborhood gentrifies slower due to reputation | |
| Hwang and Sampson (2014) | • Finding: % black and % Latino negatively predict gentrification’s advancement in gentrifying neighborhoods and in neighborhoods adjacent to gentrification | |
| Delmelle (2016) | • Finding: most upgrading neighborhoods are racially diverse at baseline | |
| Chicago, IL, Houston, TX, Los Angeles, CA, & New York, NY | Bader and Warkentien (2016) | • Finding: 1970–2010, no evidence of white entry into black neighborhoods |
| Philadelphia, PA | Anderson (1990) | • Setting: mixed-race neighborhood adjacent to black neighborhood, white gentrifiers |
| Moore (2009) | • Setting: black neighborhood, black gentrifiers | |
| Wherry (2011) | • Setting (Centro de Oro): Latino neighborhood | |
| Hwang (2016b) | • Setting (Southwest Center City)/Finding: black neighborhood, white gentrifiers who maintain boundaries associated with areas that are still predominantly black | |
| Washington, DC | Hyra (2017) | • Setting (Shaw/U-Street): black neighborhood, black gentrifiers first, white gentrifiers later |
| U.S. | Spain (1980) | • Finding: 1967–1976, white inmovers replacing black outmovers in central cities is rare |
| Bostic and Martin (2003) | • Finding: 1970–1990, black homeowners contributed to gentrification during the 1970s but not the 1980s | |
| Logan and Zhang (2010) | • Finding: 1980–2000, white entry into predominantly black tracts is rare | |
| Owens (2012) | • Finding: 1970–2010, vast majority of minority urban neighborhoods do not experience socioeconomic ascent, though the proportion increases substantially during the 2000s from prior decades | |
| Freeman and Cai (2015) | • Finding: 1980–2010, white entry into predominantly black tracts is rare but increased substantially in 2000s; more likely in less segregated cities | |
| Landis (2016) | • Finding: 1990–2010, % white positively predicts gentrification in central cities but not % black or % Hispanic | |
| Owens and Candipan (2018) | • Finding: 1990–2010, most majority-minority neighborhoods remain majority-minority; high-SES whites replacing minority neighborhoods is more likely in ascending Hispanic neighborhoods than ascending black neighborhoods |
THE CHANGING NATURE OF GENTRIFICATION
Despite the relative rarity of gentrification occurring in black neighborhoods compared to neighborhoods with greater shares of whites or mixed-race neighborhoods, the increased occurrence of gentrification in historically black neighborhoods in recent decades compared to the past likely contributes to popular images of gentrification disproportionately affecting African-American neighborhoods. Freeman and Cai (2015) find that nearly twice as many predominantly black neighborhoods experienced increases in whites from 2000 to 2010 compared with prior decades, though the relative share of black urban neighborhoods experiencing this influx was only 10.4 percent. Owens (2012) finds a similar uptick of urban minority neighborhoods experiencing socioeconomic ascent from 2000–2009 compared to prior decades.
In general, gentrification has spread to more cities and neighborhoods since the late 1990s compared to its early waves in the 1970s and 1980s (Hackworth and Smith, 2001). In contrast to the relatively slower and sporadic gentrification of the past, Hackworth and Smith (2001) describe the gentrification of the late 1990s and beyond as a new wave of gentrification that is more rapid and widespread, taking place in economically riskier neighborhoods, and increasingly led by developers, investors, and the state. Researchers have attributed this broader trend to a variety of factors, including changes in the geography of high-skilled jobs and amenities; increased cultural tastes for urban living; decreased crime in central cities; and the loosening of mortgage markets and securitization characterizing the housing boom of the early 2000s prior to the Great Recession (for a review, see Hwang and Lin [2016]).
While the increased spread of gentrification may explain its increased occurrence in predominantly black neighborhoods, some scholarship suggests that this is coupled with a distinct shift in gentrifiers’ ethnoracial preferences as gentrifiers increasingly find “authenticity” in historically black neighborhoods (Hyra 2017). While these preferences may hold in historically black neighborhoods, they are nonetheless inconsistent with broader trends that predominantly black neighborhoods are still relatively less likely to gentrify compared to neighborhoods with other compositions. Further, the growth of the middle-class black population may also contribute to increases in the gentrification of black neighborhoods. While middle-class blacks may play a role pioneering gentrification in black neighborhoods, the influx of middle-class blacks can lead to the subsequent influx of high-income white residents (Hyra, 2008, 2017).
Other scholars point specifically to the increased role of the state and institutional investors in facilitating gentrification through pro-development regimes, including transit-oriented development, and shifts in housing policies towards privatization that target predominantly African-American neighborhoods (e.g., public housing demolition, mixed-income developments) (Wyly and Hammel, 1999; Hackworth and Smith, 2001; Lees and Ley, 2008; Goetz, 2011; Hyra, 2012; Zuk et al., 2015). Nonetheless, research on the relationship between these policies and gentrification in minority areas are mixed. While Pollack, Bluestone, and Billingham’s (2010) analysis of 41 metropolitan areas finds that transit-oriented development generally occurs in racially diverse areas with high shares of Latinos and immigrants, the impacts of transit-oriented development on home prices, business growth, and displacement are quite varied (for a review, see Zuk et al. [2015]). Further, while cities across the US, including Seattle, have demolished large-scale public housing developments since the 1990s and redeveloped these sites into mixed-income communities, Tach and Emory’s (2017) analysis of the impact of public housing redevelopment on neighborhood changes across the US find that such changes did not result in increases in nonpoor and white residents.
Although examined less in the scholarship on gentrification, the rising population of recent immigrants to central cities may also explain shifts in gentrification to predominantly minority neighborhoods (Hwang, 2015). First, the arrival of Asian and Latino immigrants to predominantly black or other minority neighborhoods may increase diversity or alter the social and economic conditions that may make them more attractive to gentrifiers, though this may likely differ between Asians and Latinos, as well as within these groups. Studies show that the growth of Asian and Latino immigrants spurred local economies, increased housing demand, repopulated declining areas in central cities, and lowered crime (Winnick, 1990; Sampson, 2012). The growth of black immigrants may also affect the social and economic conditions of these neighborhoods. Hwang (2015) finds that the influx of foreign-born residents, Asians, and Hispanics, especially into black neighborhoods, across several US cities in the early 1970s is positively associated with gentrification in subsequent decades.
The influx of new immigrants to preexisting ethnic enclaves or areas that are forming enclaves, however, may also detract gentrification. Hwang’s (2015) study of early gentrification found that ethnic enclaves were negatively associated with subsequent gentrification. Research on immigration and the diversification of cities and neighborhoods suggests that, in cities with low levels of segregation and growing Asian and/or Hispanic populations, these groups become increasingly segregated (Frey and Farley, 1996; Iceland, 2004; Logan, Stults, and Farley, 2004). Outgroups may increasingly avoid these groups as they grow (Sanchez, 1997; White and Glick, 1999). At the same time, these groups may form communities that favor in-groups by relying on co-ethnic networks and employers for information on resources and housing and preferences by ethnic landlords for in-group tenants (Ball and Yamamura, 1960; Wong, 1998). Winnick (1990) also documents the effective political organization of ethnic groups in neighborhoods that prevent development processes and preserve affordable housing. Altogether, these processes may protect neighborhoods with greater shares of these groups from gentrification and offset gentrification onto other low-cost neighborhoods, such as predominantly black neighborhoods, that may be more vulnerable to the negative consequences of gentrification like displacement (Massey and Denton, 1993; Desmond and Shollenberger, 2015). Such a process creates new dynamics of neighborhood inequality by race in cities today.
SETTING
In this article, I use Seattle as a case study to examine how neighborhood ethnic and racial composition predicts which neighborhoods gentrify and if this relationship differs over time across gentrification waves. Like most major cities, the overall white population in Seattle declined substantially from 1960–1990. Since 1990, however, it has been steadily increasing. Seattle’s share of blacks has wavered between 7–10% of the total population since 1970 and consists of both African-American and foreign-born residents. Notably, Seattle’s foreign-born black population, which comes primarily from East Africa, has more than doubled from 2000 to 2013, while its native-born black population has declined by more than 15%. The Asian population, on the other hand, has increased rapidly in recent decades, surpassing that of blacks by 1990. By 2013, Asians comprised 14% of the total population, doubling in size since 1980. The Hispanic population in Seattle more than tripled since 1980 but only comprised 6% of the overall population by 2013.
Unlike highly segregated cities, Seattle has few majority-minority or ethnic neighborhoods, relatively more racially diverse neighborhoods, and mostly predominantly white areas. Figure 1 presents maps of Seattle’s ethnoracial compositions in 1980 and 20134 for Seattle’s census block groups, which are divisions of census tracts containing 600 to 3,000 residents and the smallest geographic unit for which the US Census provides publicly available demographic estimates. Seattle’s growing Asian and Hispanic populations and black population concentrate in certain areas, but they also reside in most areas throughout the city. Further, other groups also typically reside where each of these groups concentrates. No block groups were over 50% Asian or Hispanic in 1980, and only 5% were in 2013. Less than 3% of block groups were over 50% black in either year. Even Seattle’s International District, a cultural center for Asians and Asian-Americans, was only 50% Asian in 2013.
Figure 1.
Ethnoracial Groups in Seattle by Census Block Groups in 1980 (left) and 2013 (right). + (green) = Asians, ▲ (blue) = blacks, ○ (red) = Hispanics, ☓ (orange) = whites, 1 dot = 25 persons.
Compared to other major cities, the Asian population in Seattle is relatively large and comprised of a diverse group of ethnic origins. Prior to major legislative reforms surrounding immigration in 1965, Seattle’s Asians were primarily Japanese, Chinese, and Filipino (Taylor, 1994).5 Following 1965, the Chinese and Filipino populations grew rapidly, and Koreans and Vietnamese began arriving in large numbers. Most of Seattle’s Asian growth, however, occurred after 1980 and is attributable to these groups’ continued growth and new arrivals from Southeast Asia. In 2013, 65% of Seattle’s Asians were foreign-born, and Asians comprised more than half of the foreign-born population.6 Nonetheless, Asians of different ethnic origins and both foreign- and native-born Asians tend to live in the same block groups.
Blacks have also comprised a substantial proportion of Seattle’s minority population since World War II, which brought large influxes of African-Americans in search of labor opportunities (Taylor, 1994). Despite early claims of Seattle’s racial tolerance, blacks, along with Asians, experienced intense housing discrimination through the widespread use of restrictive covenants for much of the twentieth century (Taylor, 1994). While this likely deterred many blacks and Asians from living in the city in the first place, such discrimination relegated blacks and Asians to specific areas of the city. Most blacks lived in the Central District, while most Asians lived in the International District (Taylor, 1994). Few blocks in the Central District, however, were predominantly black; many whites and Asians also resided there (Taylor, 1994). Following the passage of the 1968 Fair Housing Act, which banned restrictive covenants, blacks moved to other sections of Seattle, particularly the southeast, and the suburbs and became far less concentrated: While 80% of Seattle’s black population lived in the Central District in 1960, only 38% did so by 1980 (Taylor, 1994). Figure 1 illustrates that this share was far lower by 2013.
Seattle’s Hispanic population has grown substantially but is relatively small. Only one-third of the Hispanic population was foreign-born in 2013, and only two-thirds have origins in Latin American countries. Approximately half of Hispanics have origins in Mexico, and most migrants after 1980 came from Central and South America. In 1990, they had a per capita income of $27,580 (in 2013 constant dollars)—higher than both Asians ($22,574) and blacks ($19,956) in Seattle but lower than whites ($39,226).
HYPOTHESES AND STRATEGY
Many of Seattle’s neighborhoods have gentrified, especially in recent decades in step with aggressive growth policies and a rapidly growing tech industry (Morrill, 2003). Despite Seattle’s present-day reputation of widespread gentrification and unaffordability, the city still has its share of working class neighborhoods: over a quarter of the city’s neighborhoods had median household incomes below national levels as of 2013. Prior research suggests two possibilities predicting the relationship between gentrification and ethnoracial composition in a city with low segregation levels like Seattle. On the one hand, other factors, such as housing and socioeconomic characteristics or proximity to downtown, may predict gentrification instead of racial composition. Alternatively, if ethnoracial composition is indeed associated with gentrification in Seattle, gentrification may be more likely to occur in ethnically or racially diverse neighborhoods that have some shares of minorities, which are far more prevalent than in highly segregated cities, rather than highly concentrated minority or white neighborhoods. Nonetheless, Seattle may also exhibit similar processes of minority avoidance as in highly segregated cities. If there is a relationship between ethnoracial composition and gentrification, I would expect that it would reflect a racial hierarchy, either consistent with general trends of racial stratification, favoring whites over Asians, Asians over Hispanics, and Hispanics over blacks (Charles, 2003) or reflecting the socioeconomic order of ethnoracial groups in Seattle, reversing the ordering of Hispanics and Asians.
The literature on gentrification also describes marked distinctions in the character of gentrification over time, with minority avoidance in the early wave of gentrification but mixed claims of its increased occurrence in minority neighborhoods, particularly black ones, in the recent wave of gentrification. Therefore, I separately examine each period of gentrification by relying on distinct datasets and match these data to the 1970–2000 US Censuses and additional data sources. As I will demonstrate, the relationships between ethnoracial composition and gentrification in Seattle do not conform to the literature and hypotheses outlined above. After presenting these results, I then draw from multiple supplementary data sources to explain these results. Table 2 presents a summary of these hypotheses and explanations based on the discussion and literature summarized above.
Table 2.
Hypotheses and Explanations
| Hypotheses | Context | Explanation |
|---|---|---|
| • Nonracial Amenities: Racial and ethnic composition of neighborhoods does not affect which neighborhoods gentrify. | Low-segregation | • racial and ethnic composition of neighborhoods does not constrain housing choices in low-segregation places |
| • Diversity: Gentrification favors racially diverse neighborhoods over homogeneous white or minority neighborhoods | Low-segregation | • distinct cultural preferences of gentrifiers for “diversity” can be fulfilled in low-segregation places where mixed-race neighborhoods are more prevalent |
| • Minority Avoidance/Racial Hierarchy: gentrification favors neighborhoods with fewer minorities and exhibits a racial hierarchy in which neighborhoods with more blacks are at the bottom | High or low-segregation | • neighborhood preferences/bias |
| • Minority Attraction/Reverse Racial Hierarchy: gentrification favors neighborhoods with more minorities and exhibits a reversed racial hierarchy in which neighborhoods with more blacks are at the top | Recent wave gentrification (1990s/2000s) | • changing preferences/biases or distinct hierarchy in Seattle |
| • state-driven policies (transit, public housing transformation) | ||
| • rise of middle-class minorities | ||
| • immigration to black neighborhoods attracting gentrification | ||
| • immigration to other neighborhoods deflecting gentrification to black neighborhoods |
DATA AND METHODS
Early Wave Gentrification, 1970s-1980s
To examine early wave gentrification, I draw on data from an influential field survey conducted by geographers Elvin Wyly and Daniel Hammel (hereafter, WH) during the 1990s across 23 US cities, including Seattle (for details, see Wyly and Hammel [1999]). WH’s approach reliably captures direct and distinctly visible indicators of neighborhood upgrading that are inherent to gentrification, particularly aesthetic changes to the built environment. These measures are consistent with demographic changes associated with gentrification, such as increased shares of college-educated residents and high increases in home values, but more reliably capture the visible transformation of neighborhoods (Wyly and Hammel, 1999; Hwang and Sampson, 2014).7 WH conducted block-by-block field surveys of only gentrifiable census tracts—those with a median household income in 1970 below the citywide median in 1970, when cities in the West experienced large population declines after steady growth in preceding decades. Among gentrifiable census tracts, they considered tracts to be gentrifying if most blocks had at least one improved structure or at least one block in the tract had at least one-third of its structures improved, and they triangulated their findings with archival resources, such as city planning documents and local press.
In their survey of Seattle, which they conducted in 1998, 41 of the city’s 124 tracts were gentrifiable, of which they categorized 22 as gentrifying. Although Seattle’s census tracts span larger geographic areas than in other, denser major cities, familiar neighborhood identities span tract boundaries. The map of Seattle in Figure 2 shows the gentrification categories of Seattle’s census tracts. Many of the tracts that were gentrifying by 1998 were in or near the downtown area and the University District by the University of Washington. Notably, the southeastern area of the city, where many African-Americans moved over the last several decades, has since become gentrifiable—having median incomes below the citywide median in more recent years, and I assess these areas in the analysis of the recent wave of gentrification.
Figure 2.
Map of Early Gentrification from 1998 Gentrification Field Surveys
The New Wave of Gentrification, 1990s-2000s
To identify gentrification in recent decades, I use census-based variables from the 1990 and 2000 US Census harmonized to 2010 Census block group boundaries and the 2009–2013 ACS 5-year estimates.8 While visible indicators may identify gentrification more reliably than census-based measures, observational data over this entire time period do not exist, and census data offer a way to compare similar spatial units over time.9 I conduct this analysis using census block groups instead of census tracts because, unlike the data for the early wave of gentrification, this analysis is not restricted to data collected by others.
Building on prior approaches to measure gentrification (Wyly and Hammel 1999; Freeman, 2009), I consider block groups to be gentrifiable if their median household income is below the citywide median household income in either 1990 or 2000.10 To identify if block groups were gentrifying over these periods, I selected variables that best matched the changes comparing WH’s survey results and 1970–1990 Census data. I consider a block group to be gentrifying if it had an increase in either its median rent or median home value above the median citywide increase and an increase in either its share of college-educated residents or median household income above the median citywide increase from either 1990–2013 or from 2000–2013, allowing for both slower and more rapid gentrification.
Appendix Table A1 presents comparisons of these measures with correlates of gentrification, including both demographic and housing census-based variables, as well as coffee shops and building permits. These correlations lend support for these indicators in Seattle. Figure 3 maps the recent wave of gentrification using this measure. In addition to the areas that WH had examined, gentrifiable areas in this period also include parts in the far north of the city and include more adjacent areas in the southern part of the city. There is substantial overlap with the areas that WH identified as gentrifying in 1998, but there is also considerable expansion well beyond adjacent areas.
Figure 3.
Map of Recent Gentrification in Seattle for 1990–2013 Census-Based Gentrification Measures
Independent Variables
Data for ethnoracial, socioeconomic, and housing characteristics are from the decennial US Censuses from the Geolytics Neighborhood Change Database. To examine predictors of early gentrification, the 1970 data are harmonized to 2000 Census tract boundaries because the field surveys are based on these boundaries. I use 1970 as the baseline year because the census data shows that, on average, tracts that WH observed as gentrifying were gentrifying as early as the 1970s, but findings using 1980 data are similar.11 The analysis of the new wave of gentrification uses 1990 data harmonized to 2010 Census block group boundaries. The main ethnoracial compositional variables I included in the models are the shares of Asians, blacks, and Hispanics. Ethnic groups or nativity within these broader ethnoracial categories are not available with data from these years. I also present results for models replacing the compositional variables with the overall share of non-white residents.
I control for several additional factors based on prior research that may predict variation in where gentrification occurs using principal components—an ideal approach for small sample sizes because it transforms a set of related variables into linearly uncorrelated variables and, therefore, minimizes multicollinearity. To capture the presence of an available, affordable, and older housing supply (Zukin, 1987; Ley, 1996; Smith, 1996), I include median rent and home value (logged), the share of residents who have lived in the same residence for over 10 years, homeownership rate, vacancy rate, the share of multiunit housing, and the share of buildings over 30 years old. To capture proximity to downtown and institutions, where jobs are primarily located, I calculate the nearest distance to either Seattle’s Downtown or the University of Washington.12 Because neighborhoods that have relatively higher socioeconomic status among gentrifiable neighborhoods may be more likely to gentrify, I also include variables for income per capita (logged), median household income (logged), poverty rate, the share of college-educated residents, and the share of residents in professional or managerial occupations.
The first two components from the principal component analysis explain over 90% of the variance for these characteristics across census tracts and block groups in both periods. The first component reflects housing opportunities for development—low housing costs and homeownership; high vacancies, multiunit housing, and older buildings; and proximity to downtown. The second component reflects high socioeconomic status—high shares of college-educated residents and professionals.13 Appendix Table A2 presents factor loadings and correlations for each variable.
Because crime is an additional factor that may influence which neighborhoods gentrify (Kirk and Laub, 2010), I also control for overall crime rates reported in 1970 in the Testing Theories of Criminality and Victimization Study in Seattle14—a survey of a random sample of 5,302 residents conducted in 1990—for the analysis of early gentrification and logged crime rates15 per 100,000 residents reported by the Seattle Police Department in 1996—the next closest year to 1990 for which crime data are publicly available.16
Methods
To examine the relationship between ethnoracial composition and the location of early gentrification, I use logistic regression models predicting the binary measure of whether a tract was gentrifying on ethnoracial composition in 1970 among gentrifiable tracts. Given that WH only observed 41 tracts, I use Firth’s (1993) penalized likelihood approach to adjust for bias in the estimates that can result from having a small sample size and when predictors with values above a certain point have the same outcome.17 Because of the small sample size, readers should interpret the coefficient estimates with caution. To examine contemporary gentrification, I use a logistic regression model predicting whether a block group was gentrifying by 2013 on ethnoracial composition in 1990 among gentrifiable block groups.
Explaining Recent Wave Gentrification
The findings from the recent wave of gentrification, which I present below, are inconsistent with expectations: there is a positive association of subsequent gentrification with shares of blacks and a negative association with shares of Asians. After demonstrating the peculiar results of recent gentrification, I consider various explanations drawing from the literature reviewed earlier. I separately include variables described below to test if each of the various explanations mediate the observed relationships. Specifically, I test if Seattle’s unique racial dynamics, transit and public housing policies, middle-class minorities, and immigrant settlement patterns explain the findings.
To examine Seattle’s unique racial dynamics, I draw on data from the Testing Theories of Criminality and Victimization Study in Seattle and the Seattle Neighborhood and Crime Survey18—a similar survey of 2,200 Seattle residents conducted in 2003. I examine responses on preferences for various race groups as neighbors using the 2003 survey to capture explicit biases, as well as measures of neighborhood perceptions of disorder and danger to capture implicit biases against neighborhoods based on their racial compositions using data for both surveys. Following Sampson and Raudenbush (2004), I use scaled measures of neighborhood perceptions of danger (based on fear of walking alone at night and a safety rating of the neighborhood [1990] and concerns with safety, worries of attack, and a safety rating [2003]) and neighborhood disorder (based on the presence of teens hanging out, vandalism, and abandoned and run-down housing [1990] and teens hanging out, litter, graffiti, abandoned and run-down housing, and neighbors causing trouble [2003]).
To assess state-driven policies, I draw from public geocoded data on the locations of Seattle’s public housing developments and light rail stations. I identify the block groups containing now former public housing developments and calculate distances from the centroid of tracts to the nearest light rail station. To assess if middle-class minorities play a distinct role in gentrifying predominantly black neighborhoods, I include census-based measures on poverty rates and median incomes by race groups. Lastly, I examine the role of the Asian, Hispanic, and foreign-born population changes using Census data.
RESULTS
Early Gentrification and Minority Avoidance
Table 3 presents average characteristics of the variables used in the analysis based on 1970 and 2000 US Census data for the sample of tracts by their gentrification categories. Although all gentrifiable tracts were over 50% white on average in 1970, gentrifying tracts had higher shares of whites and lower shares of blacks, Asians, and Hispanics in 1970 compared to nongentrifying tracts. Gentrifying tracts were also more advantaged on some indicators in 1970. They had more professionals and higher home values, but they also had more residential turnover and multiunit structures. The averages were similar for other indicators. Over time, gentrifying tracts experienced larger increases in income, share of college-educated residents, and median home values compared to nongentrifying tracts.
Table 3.
Average Tract Characteristics and Analysis Measures by Wyly and Hammel’s 1998 Gentrification Field Survey Categories
| Gentrifiable | Nongentrifiable | |||||
|---|---|---|---|---|---|---|
| Nongentrifying | Gentrifying | |||||
| 1970 | 2000 | 1970 | 2000 | 1970 | 2000 | |
| N | 19 | 22 | 83 | |||
| % white | 58.3** | 48.9** | 88.5 | 77.7 | 90.9 | 72.9 |
| % black | 25.8* | 19.9** | 6.1 | 7.2 | 3.5 | 7.2 |
| % Hispanic | 2.5† | 9.0* | 1.5 | 4.7 | 2.1* | 4.5 |
| % Asian | 11.2** | 19.7* | 2.2 | 9.1 | 2.9 | 14.5** |
| % foreign-born | 12.7 | 22.9* | 10.3 | 12.7 | 8.3** | 15.8* |
| % below poverty | 19.3 | 22.6 | 17.4 | 17.0 | 7.5** | 8.9** |
| Median household income | $50,562 | $47,818 | $46,517 | $50,439 | $74,538** | $76,259** |
| Income per capita | $25,263 | $29,222** | $25,884 | $47,708 | $31,898** | $43,957 |
| % college-educated | 10.3 | 31.5** | 16.6 | 55.7 | 17.2 | 49.3* |
| % professional/managerial | 17.6** | 37.0** | 25.8 | 51.6 | 28.8 | 49.9 |
| Median home value | $113,055* | $301,474** | $154,224 | $451,805 | $134,612 | $402,765 |
| Median gross rent | $534† | $846† | $612 | $960 | $764** | $1,134** |
| % new resident in last 10 years | 71.9** | 76.8** | 81.5 | 86.1 | 65.3** | 65.2** |
| % homeownership | 47.6 | 33.5† | 33.6 | 22.9 | 67.9** | 61.0** |
| % vacant units | 11.9 | 6.2 | 9.9 | 6.8 | 5.1** | 3.5** |
| % multiunit structures | 53.2** | 59.1** | 77.8 | 83.9 | 23.7** | 31.3** |
| % units built over 30 years ago | 60.2 | 70.7 | 65.8 | 64.5 | 39.6** | 74.9** |
| Measures for regression analysis (1970) | Mean | Std. dev. | Mean | Std. dev. | Mean | Std. dev. |
| Distance (in feet) (sq. rt.) | 53.0* | 54.8 | 20.2 | 31.0 | 99.3** | 46.6 |
| First PC (residential opportunity) | 33.9* | 61.7 | 76.9 | 33.5 | −28.1** | 48.6 |
| Second PC (socioeconomic status) | −7.70 | 21.8 | −14.1 | 24.4 | 5.50** | 25.3 |
| Crime rate (logged) | 8.90 | 0.70 | 8.64 | 0.66 | 7.82** | 0.51 |
Note:
p<0.01,
p<0.05,
p<0.10 (two-tailed t-test).
T-tests compare gentrifying tracts to nongentrifying and nongentrifiable tracts. Dollars are in 2013 constant dollars.
Controlling for additional factors related to the likelihood of gentrification, logistic regression results, which are presented in the first two columns of Table 4, confirm patterns that early gentrification was less likely to take place in gentrifiable census tracts with more minorities. Neighborhoods with greater shares of Asians and Hispanics in 1970 are negatively associated with early gentrification in Seattle census tracts after controlling for residential and socioeconomic characteristics (Model 1a). The coefficient for percent black is also negative but is smaller and significant only at the p<.10-level. Grouping all minorities together (Model 1b) shows that gentrification was less likely to occur in neighborhoods with fewer shares of whites. Comparing Figures 1 and 2 shed light on this finding: The areas where Asians primarily resided throughout the southeastern section of the city did not gentrify and the few areas where Hispanics represented higher shares of the population—primarily the areas with low population density near South Lake Union and Georgetown—did not gentrify. Further, while some parts of the Central District, where a relatively larger share of blacks resided, were gentrifying, many of the remaining areas that had higher shares of blacks did not gentrify. Areas with higher levels of the first principal component (residential opportunities) were also more likely to gentrify.
Table 4.
Regression Coefficients and Standard Errors Predicting Gentrification on Ethnoracial Composition
| Early Gentrification (Gentrification Field Surveys, 1998) | Recent Gentrification (Census-Based Gentrification, 1990–2013) | |||
|---|---|---|---|---|
| (1a) | (1b) | (2a) | (2b) | |
| % composition | % minority | % composition | % minority | |
| % black | −0.076† | 0.029** | ||
| (0.038) | (0.010) | |||
| % Asian | −0.231** | −0.040** | ||
| (0.101) | (0.012) | |||
| % Hispanic | −2.336* | −0.029 | ||
| (0.979) | (0.039) | |||
| % minority | −0.047* | −0.002 | ||
| (0.023) | (0.005) | |||
| First PC (residential opportunity) | 0.064** | 0.024** | 0.000 | 0.001 |
| (0.028) | (0.009) | (0.003) | (0.003) | |
| Second PC (socioeconomic status) | −0.042 | −0.008 | 0.017** | 0.018** |
| (0.032) | (0.018) | (0.006) | (0.005) | |
| Crime rate (logged) | −0.348 | −0.341 | 0.060 | 0.082 |
| (1.091) | (0.741) | (0.242) | (0.233) | |
| Prior gentrification | 0.158 | 0.333 | ||
| (0.362) | (0.345) | |||
| AIC | −24.5 | −13.1 | 311.9 | 327.2 |
| N | 40 | 40 | 241 | 241 |
| Model | Penalized logistic regression | Logistic regression | ||
Note:
p<.01;
p<.05;
p<.10 (two-tailed test).
Overall, in contrast to the nonracial amenities and diversity hypotheses expected in a city with low segregation, early gentrification in Seattle was much more likely to take place in low-income neighborhoods that were predominantly white and had few minorities. The patterns of early gentrification in Seattle reflected broader trends of minority avoidance found in other highly segregated cities. A clear hierarchy among ethnoracial minority groups was not evident, though this is likely due to the small sample size.
Contemporary Gentrification and Hierarchy Reversal
Table 5 shows the neighborhood characteristics in 1990 and 2013 and the variables used in the analysis for block groups by their gentrification categories for the recent wave of gentrification. Gentrifying and nongentrifying block groups were similar on many dimensions in 1990, including the share of whites and Hispanics, poverty and income levels, college-educated residents, housing and rental values, homeownership rates, and multiunit structures. Block groups that were gentrifying, however, began the period with higher shares of blacks, lower shares of Asians and foreign-born residents, and an older housing stock in 1990 compared to nongentrifying block groups. By 2013, gentrifying block groups had higher shares of whites, college-educated residents, income levels, and ownership rates and lower shares of blacks and Hispanics compared to nongentrifying block groups—consistent with changes commonly associated with gentrification. Not surprisingly, block groups that were gentrifying also were closer to Downtown or the University of Washington compared to nongentrifying ones. The gentrifiable block groups differed from nongentrifiable block groups on nearly every characteristic, and, although nearly two-thirds of the population in gentrifiable block groups was white on average, these block groups also had greater shares of minorities compared with nongentrifiable ones.
Table 5.
Average Block Group Characteristics and Analysis Measures by 1990–2013 Census-Based Gentrification Measures
| Gentrifiable | Nongentrifiable | |||||
|---|---|---|---|---|---|---|
| Nongentrifying | Gentrifying | |||||
| 1990 | 2013 | 1990 | 2013 | 1990 | 2013 | |
| N | 133 | 111 | 231 | |||
| % white | 64.5 | 53.3** | 65.0 | 64.9 | 82.2** | 76.5** |
| % black | 11.5* | 11.9* | 17.7 | 8.4 | 5.7** | 3.7** |
| % Hispanic | 4.4 | 9.4** | 3.9 | 6.4 | 2.6** | 4.5** |
| % Asian | 17.5** | 19.6** | 11.6 | 14.5 | 8.4* | 10.6* |
| % foreign-born | 18.4** | -- | 13.2 | -- | 10.4** | -- |
| % families below poverty | 4.1 | 7.0** | 3.4 | 3.9 | 0.8** | 2.2** |
| Median household income | $44,549† | $47,060** | $41,867 | $64,341 | $73,403** | $90,352** |
| Income per capita | $27,640 | $31,035** | $28,150 | $42,823 | $40,432** | $51,961** |
| % college-educated | 30.8 | 44.8** | 30.6 | 59.9 | 42.4** | 63.4* |
| % professional/managerial | -- | 44.9** | -- | 54.5 | -- | 60.5** |
| Median home value | $234,505 | $345,731** | $221,929 | $416,426 | $301,518** | $515,206** |
| Median gross rent | $846 | $892** | $815 | $1,039 | $1,083** | $1,206** |
| % new resident in last 10 years | 77.2 | 80.3 | 76.6 | 80.3 | 59.1** | 62.1** |
| % homeownership | 33.5 | 33.9† | 32.5 | 38.1 | 69.5** | 68.7** |
| % vacant units | 5.7* | 6.7 | 6.8 | 7.3 | 3.2** | 5.3** |
| % multiunit structures | 61.7 | 62.6 | 62.7 | 60.8 | 20.1** | 21.7** |
| % units built over 30 years ago | 53.2** | 64.2 | 64.7 | 64.5 | 76.4** | 82.1** |
| Measures for regression analysis (1990) | Mean | Std. dev. | Mean | Std. dev. | Mean | Std. dev. |
| Distance (in feet) (sq. rt.) | 91.3** | 62.1 | 68.6 | 53.2 | 103** | 43.1 |
| First PC (residential opportunity) | 15.8* | 66.2 | 34.8 | 60.3 | −25.9** | 44.0 |
| Second PC (socioeconomic status) | −27.5** | 29.9 | −12.6 | 25.2 | 21.8** | 27.4 |
| Crime rate (logged) (1996) | 9.25 | 0.66 | 9.36 | 0.68 | 8.64** | 0.40 |
Note:
p<0.01,
p<0.05,
p<0.10 (two-tailed t-test).
T-tests compare nongentrifying tracts to gentrifying tracts and nongentrifiable tracts to gentrifying tracts. Dollars are in 2013 constant dollars. % poverty for individuals is not available for block groups; missing values are not available in normalized block group data.
The last two columns in Table 4 present logistic regression results predicting the log odds of gentrification on ethnoracial composition and overall minority composition, respectively, among gentrifiable block groups and controlling for other factors associated with the likelihood of gentrification. Gentrification in recent decades is positively associated with neighborhoods with greater shares of blacks but negatively associated with shares of Asians (Model 2a). There is no association between the share of minorities and gentrification (Model 2b), which is expected given the opposite direction of these relationships. This pattern is also evident in Figure 3: Although virtually all the Central District and International District—historically and symbolically black and Asian areas, respectively—were gentrifying, the areas in the southeastern section of the city that were not gentrifying are areas where there are more Asians, such as the area between Beacon Hill and the International District. Moreover, more areas on the eastern side of Rainier Valley in the southeastern section of the city, which include the Columbia City neighborhood and have relatively higher shares of blacks, were gentrifying.
In sum, the findings for recent gentrification are distinct from those for earlier gentrification and are inconsistent with the nonracial amenities, diversity, and minority avoidance hypotheses. Figure 4 displays the predicted probabilities of gentrification in early and recent gentrification by the shares of blacks and Asians in a neighborhood with all other control variables held at their means, illustrating the distinct shifts between race groups and their relationship to gentrification in Seattle over time. Further, the relationships are inconsistent with a racial hierarchy of residential preferences or the socioeconomic order of groups in Seattle, such that neighborhoods with greater shares of Asians should be more likely to gentrify than those with greater shares of blacks (Charles, 2003). They do not fully support the minority attraction hypothesis either because the relationships with gentrification for shares of Asians and blacks are opposite. Nonetheless, they reflect recent observations of gentrification’s increased prevalence in predominantly black neighborhoods, or a reversed racial hierarchy. Next, I examine potential explanations to understand this shift in Seattle.
Figure 4.
Predicted Probability of (a) Early Gentrification and (b) Recent Gentrification by Percent Asian and Percent Black
ASSESSING EXPLANATIONS
Racial and Ethnic Stratification in Seattle
Given Asians’ relatively long history and lower socioeconomic status in Seattle relative to other major cities, one possibility is that a racial hierarchy reflecting residential preferences may be distinct in Seattle. Both blacks and Asians faced high levels of discrimination historically in Seattle (Taylor, 1994; Chin, 2001). However, data from the 2003 Seattle Neighborhood and Crime Survey show that a significantly (p<.001) greater share of residents indicated that they favored living in a neighborhood that was one-fourth Asian (44%) rather than one that was one-fourth black (38%). These differences in preferences for all residents are also similar for white, college-educated, and high-income respondents.
While these results suggest that residents hold stronger explicit biases against blacks relative to Asians as neighbors, consistent with past findings illustrating a racial hierarchy of preferences among neighbors (Charles 2003), implicit biases could be driving the observed relationships. Although past research generally demonstrates this relationship for black neighborhoods (e.g., Ellen 2000; Quillian and Pager, 2001), implicit biases against Asian neighborhoods in the context of Seattle may drive avoidance of Asian neighborhoods. Drawing from survey responses on neighborhood perceptions of danger and disorder from the 1990 Testing Theories of Criminality and Victimization Study in Seattle and the 2003 Seattle Neighborhood and Crime Survey described earlier, I find that among gentrifiable block groups, the share of Asians and blacks have similar positive correlations with perceived neighborhood crime and disorder in 2003 (rblack/disorder=.29, p<.001; rAsian/disorder=.34, p<.001; rblack/danger=.27, p<.001; rAsian/danger=.31, p<.001). However, the 1990 measures are positively correlated with the share of blacks (rdisorder=.53, p<.001; rdanger=.42, p<.001) but not the share of Asians (rdisorder=.10, p>.05; rdanger=.13; p>.05). Nonetheless, these perceptions do not change the results when either perceived disorder or danger is included in the model. Models 1 and 2 in Table 6 display these results.
Table 6.
Assessing Explanations Predicting Recent Gentrification
| (1) | (2) | (3) | (4) | (5) | (6) | (7) | |
|---|---|---|---|---|---|---|---|
| Perceived Disorder | Perceived Danger | Public Housing Redevelop-ment | Transit Develop-ment | Middle-Class Minorities | Asian/Hispanic Growth | Recent Immigrants | |
| % black, 1990 | 0.029** | 0.028** | 0.028** | 0.035** | 0.033** | 0.026* | 0.031** |
| (0.010) | (0.010) | (0.010) | (0.011) | (0.011) | (0.010) | (0.010) | |
| % Asian, 1990 | −0.036** | −0.040** | −0.040** | −0.030* | −0.034** | −0.017 | −0.020 |
| (0.012) | (0.012) | (0.012) | (0.013) | (0.012) | (0.018) | (0.015) | |
| % Hispanic, 1990 | −0.021 | −0.030 | −0.031 | −0.025 | −0.034 | 0.218† | −0.012 |
| (0.040) | (0.040) | (0.039) | (0.039) | (0.046) | (0.116) | (0.040) | |
| First PC (residential opportunity), 1990 | 0.001 | 0.001 | 0.000 | 0.003 | 0.001 | 0.001 | 0.002 |
| (0.003) | (0.003) | (0.003) | (0.003) | (0.003) | (0.003) | (0.003) | |
| Second PC (socioeconomic status), 1990 | 0.017** | 0.017** | 0.017** | 0.020** | 0.020** | 0.011† | 0.014* |
| (0.006) | (0.006) | (0.006) | (0.006) | (0.006) | (0.006) | (0.006) | |
| Crime rate (logged), 1996 | 0.087 | −0.017 | 0.061 | 0.165 | 0.063 | −0.149 | −0.088 |
| (0.268) | (0.271) | (0.243) | (0.248) | (0.253) | (0.254) | (0.254) | |
| Prior gentrification, 1980–1990 | 0.153 | 0.149 | 0.160 | 0.128 | 0.422 | 0.117 | 0.140 |
| (0.363) | (0.363) | (0.363) | (0.365) | (0.387) | (0.368) | (0.365) | |
| Perceived disorder (2003) | −2.407 | ||||||
| (3.146) | |||||||
| Peceived danger (2003) | 0.343 | ||||||
| (0.996) | |||||||
| Public housing redevelopment (1995) | 0.213 | ||||||
| (0.825) | |||||||
| Distance to transit redevelopment (2003) | 0.011* | ||||||
| (0.005) | |||||||
| Asian poverty rate, 1990 | 0.003 | ||||||
| (0.007) | |||||||
| Black poverty rate, 1990 | −0.009 | ||||||
| (0.006) | |||||||
| Asian population change, 1980–1990 | −0.005† | ||||||
| (0.003) | |||||||
| Hispanic population change, 1980–1990 | −0.034** | ||||||
| (0.013) | |||||||
| Recent immigrant population (logged), 1990 | −0.358* | ||||||
| (0.178) | |||||||
| AIC | 312.7 | 313.2 | 313.8 | 309.1 | 284.3 | 304.8 | 309.7 |
| N | 240 | 240 | 241 | 241 | 223 | 241 | 241 |
Note:
p<.01;
p<.05;
p<.10 (two-tailed test).
These general survey results do not support an explicit preference for black neighborhoods, but the positive relationships with perceptions of danger and disorder may also reflect the features that attract some residents to these neighborhoods in their search for “authentic” black neighborhoods (Hyra, 2017). However, I also test if there are distinct processes for neighborhoods with high concentrations of blacks that may hold more symbolic significance as historically black neighborhoods. Excluding the 17 block groups that are over 50% black, the share of blacks is no longer statistically significant but remains positive and does not explain the negative effect for Asians (model not shown). Further supplementary analysis does not indicate statistically significant nonlinear effects.
State-Driven Policies
Other research has argued that state-driven policies explain the shift over time from gentrification avoiding black neighborhoods to targeting them (Goetz, 2011; Hyra, 2012). In Seattle, unlike other major US cities, public housing is intentionally racially integrated, and Asians comprise a substantial share of the public housing population (Taylor, 1994). Yesler Terrace, which was Seattle’s largest and only remaining public housing development until its recent conversion to mixed-income housing in 2014, originally imposed racial and ethnic group size restrictions but primarily houses blacks and Asians (Taylor, 1994). The remaining smaller housing projects, which were all converted to mixed-income housing beginning in 1995, were intentionally built in predominantly white areas (Taylor, 1994). Thus, shifts in public housing policy that promote development should affect both groups and thus cannot explain the divergent results for blacks and Asians. Model 3 in Table 6 includes an indicator for whether public housing redevelopment was present in a block group, and the results do not change. Indeed, none of the block groups that were over 50% black that were gentrifying previously contained public housing.
Transit development—another state-driven policy—may also be responsible for the observed shift. Seattle’s public transportation light rail system, which was implemented in 2003, may contribute to the trends observed in recent decades. I test in Model 4 of Table 6 if the distance from a tract or block group to the nearest stop along the light rail line that is either completed, under construction, or planned explains the findings. Although the distance to transit is positively associated with gentrification, the main results for Asians and blacks remain.
Middle-Class Minorities
Another possible explanation is that the black middle-class is substantially larger in Seattle relative to Asians, thus driving gentrification in neighborhoods with greater shares of blacks while neighborhoods with greater shares of Asians remain isolated from co-ethnic middle-class residents. Past research has highlighted the prominent role of middle-class blacks in the gentrification of historically black neighborhoods (e.g., Pattillo 2007; Boyd, 2008). In Seattle, although the cost of living is certainly higher than most US cities, blacks are more socioeconomically advantaged on average compared to blacks in other highly segregated US cities and thus may be more likely to contribute to the spread of gentrification to black neighborhoods. The per capita income and poverty rate for blacks in Seattle in 1990 were $19,956 (in 2013 constant dollars) and 25%, respectively, while these median figures were $16,390 and 30% among the 10 most segregated large US cities.19 Further, the socioeconomic gap between blacks and Asians, whose per capita income was $22,574, is smaller compared to these other cities.
Unfortunately, data for household income by race are not available in 1990 and are missing for many block groups, though they show that median incomes in gentrifiable block groups among blacks are nearly double that of Asians in 2000. The substantive results, however, remain after including group poverty rates in Model 5. In addition, more advantaged blacks and more advantaged Asians appear to be concentrated in higher-income, nongentrifiable neighborhoods as both the black and Asian poverty rates are much higher, and their incomes are much lower in gentrifiable block groups than in nongentrifiable ones. Thus, the findings instead suggest that middle-class minorities are less constrained to low-income minority neighborhoods in Seattle compared to other highly segregated cities.
Immigrant Growth
Lastly, I examine if the growth of immigrants, who are primarily from Asia, explains the results. While some research suggests that the rise of immigrants, especially to black neighborhoods, promotes gentrification (Hwang 2015), rising immigration may also prohibit gentrification by deflecting gentrification pressures to neighborhoods with greater shares of blacks in Seattle’s tight housing market. Results from Model 6 show that the Asian and Hispanic population changes from 1980–1990 are negatively associated with gentrification, though the coefficient for Asians is only significant at the p<.10-level. Notably, the negative effect of the share of Asians is no longer statistically significant while the positive relationship of the share of blacks remains.20
In Model 7, the recent immigrant population (logged number of immigrants in 1990 who entered the US in the last 10 years) is also negatively associated with gentrification, and the variable mediates the negative relationship between the share of Asians and gentrification. While Asians, Hispanics, and immigrants have become increasingly concentrated as their populations have grown, the neighborhoods where gentrification does not occur are not necessarily enclaves. The findings hold when I exclude the 11 block groups that are over 50% Asian, 50% Hispanic, or 50% foreign-born, as well as those that are part of the International District. Block groups that did not gentrify experienced significantly (p<.05) greater increases in their foreign-born populations relative to gentrifying neighborhoods from 1990–2000, suggesting that these nongentrifying neighborhoods, even those that were not immigrant enclaves but had relatively larger shares of these groups compared to other block groups with higher shares of blacks, attracted immigrant residents, while gentrifying neighborhoods attracted college-educated, higher-income gentrifiers.
An additional consideration is that black immigrants, who tend to live in tracts with higher shares of native-born blacks relative to the remainder of the city, contribute to gentrification in neighborhoods with high shares of blacks. While data on race by nativity is not available until after 2000 and only at the census tract level, I examined the census tracts where these groups lived in 2000 and 2013. Foreign-born blacks comprise only a small share of the black population in gentrifying areas that began the period with high shares of blacks. Instead, they comprise a substantial share of the black population in areas with low shares of blacks overall and in areas that are either nongentrifiable or are not gentrifying.21 Thus, the influx of foreign-born blacks to other neighborhoods may also be offsetting gentrification pressures to areas with higher shares of native-born blacks.
Altogether, these findings are consistent with Morrill’s (2003) analysis of demographic shifts in Seattle neighborhoods from 1990–2000, which documents an expansion of Asian population growth due to immigration throughout many parts of the city that were not gentrifying. Although areas with higher shares of Asians tend to have residential characteristics that favor the likelihood of gentrification, such as older and low-cost housing, low ownership rates, and high shares of multiunit housing, distinct housing market mechanisms appear to occur in areas with higher shares of Asians relative to areas with higher shares of blacks.
DISCUSSION AND CONCLUSION
Over the last two decades, cities have transformed. Gentrification has become increasingly widespread, and the growing scale of immigration has made cities increasingly diverse. Such changes not only alter the ethnoracial composition and the social and economic conditions of neighborhoods but increase demand for low-cost central city neighborhoods. This article demonstrates that these dynamics of urban change are interdependent and sheds light on an important factor shaping how gentrification and, more generally, residential selection unfold today. Past theory on gentrification and race has relied on findings primarily from cities with high levels of segregation or on neighborhoods in the advanced stages of gentrification (Brown-Saracino, 2017). By examining gentrification in the city of Seattle, where high concentrations of minority neighborhoods are less prevalent and racially mixed neighborhoods are more prevalent compared to other cities with high segregation levels, this study extends theory on both gentrification and neighborhood selection by uncovering a mechanism shaping uneven patterns of development today—immigrant replenishment.
This study demonstrates that, even in Seattle, a city with low levels of segregation, ethnoracial composition is still an important factor predicting uneven patterns of development. This relationship, however, has changed over time as immigration has grown substantially. First, counter to expectations that gentrification in less segregated cities would follow patterns that align with nonracial amenities or that value ethnoracial diversity, early gentrification in Seattle exhibits patterns similar to highly segregated cities—minority avoidance. I show that gentrification reflects an aversion to areas with disproportionately higher shares of minorities even after controlling for neighborhood, housing, and proximity characteristics that predict gentrification. Second, in recent decades, the findings support a process of minority attraction to some groups but not others, reflecting a reversed racial hierarchy. Neighborhoods with higher shares of blacks were more likely to gentrify, while shares of Asians had a negative relationship with subsequent gentrification.
In Seattle, rising immigration plays an important role in shaping within-city dynamics of gentrification. While the overall supply and demand for low-cost housing in a city or metropolitan area shapes the broader context in which gentrification unfolds, it does not explain how gentrification unfolds within cities. Within cities, expected preferences of neighborhood selection are insufficient for explaining recent trends in Seattle. The looser city-wide housing demand during the 1970s and 1980s in Seattle, similar to most US cities, likely explains why gentrification primarily took place in neighborhoods with fewer minorities; the tightened housing market combined with the rapid rise of immigrants explains why gentrification in the 1990s and 2000s was more likely to take place in neighborhoods with greater shares of blacks and less likely to take place in nieghborhoods with lower shares of Asians. Gentrifiers—conceptualized broadly as a set of actors that include not only individual households but also commercial businesses, state and corporate entities, and institutions choosing to invest in a neighborhood to attract middle- and upper-class residents—may have racially-ordered neighborhood preferences with limited preferences for diversity, as prior studies show, but other contextual factors can affect gentrifiers’ ability to realize these preferences. Nonetheless, while gentrifiers may be more likely to move to predominantly black neighborhoods, their preferences and forms of avoidance may be taking place at smaller geographies than the block group (Fowler, 2015) or in other contexts rather than their place of residence, such as schools and public space.
An alternative interpretation of the results is that they reflect shifting cultural preferences by gentrifiers. Hyra (2017) documents white gentrifiers’ valorization of predominantly black neighborhoods as gentrifiction expands beyond downtown in Washington, DC, and others document rising nativism in the general population (Sanchez, 1997), potentially deterring gentrifiers from areas with growing immigrant populations. Nonetheless, gentrification in black neighborhoods, though more common than in the past, is still relatively rare compared with neighborhoods with other compositions based on boarder trends (Freeman and Cai, 2015; Owens and Candipan, 2018), and others document the “authenticity” valued by the presence of other ethnoracial groups (Anderson and Sternberg 2013).
The findings also extend to research on residential selection, which emphasizes resources and preferences by movers themselves and the barriers to entry faced by disadvantaged residents (Krysan and Crowder, 2017). This literature rarely considers barriers to neighborhood entry faced by more advantaged residents, but the findings from this study suggest that such barriers can result in emergent patterns of neighborhood selection. For example, landlords in neighborhoods with greater shares of Asians may rely on the continued demand for low-cost housing by new immigrants without having to invest heavily in properties, consequently shifting gentrification pressures onto other neighborhoods that would have otherwise been less desirable. Further research should examine the role of specific actors, such as property owners, and the housing preferences of recent immigrants, especially as immigration continues to play an increasingly important role in shaping contemporary housing market dynamics.
While the spatial configuration of Seattle’s ethnoracial groups may be distinct from highly segregated cities often studied in gentrification, the findings may extend to other diversifying cities in the US with low levels of segregation and experiencing gentrification, such as Portland, OR or Portland, ME. Similar analysis should also be extended to other cities, like those in the Sunbelt and new immigrant destinations, which have distinct underlying racial structures and immigrant histories for which other patterns may emerge. Moreover, multicity, multilevel analyses of the effects of metropolitan-level segregation, minority group size, and immigration on neighborhood-level changes would help provide fruitful bases. A new framework that can account for the increasingly dynamic and complex processes of residential selection and gentrification that come with the growth of multiethnic cities and neighborhoods and that also considers the historically dependent context in which such changes take place is needed to sharpen our understanding of the contemporary city.
Altogether, this study contributes to theoretical and empirical understandings surrounding neighborhood change, residential selection, and racial and ethnic inequality. By considering a city with low segregation levels, this case study of Seattle uncovered how uneven development has similar racialized processes as highly segregated cities in early waves of gentrification and how settlement patterns among growing immigrant populations shape patterns of uneven development in the more recent wave of gentrification. The findings suggest the need for scholars to consider how the spread of gentrification and growing immigration shapes residential selection and stratification in the contemporary city. Growing demand for low-cost urban neighborhoods by both middle- and upper-class residents and recently arriving immigrants create new housing market dynamics that can facilitate or prohibit gentrification within cities and can influence broader patterns of residential stratification and segregation.
The mechanisms shaping patterns of gentrification also have important implications for neighborhood inequality, particularly for blacks. In Seattle, neighborhoods with higher shares of blacks were more likely to gentrify in recent decades, and these neighborhoods experienced large declines in their black populations over the last two decades. If distinct housing markets exist that preserve affordable housing for incoming immigrants and co-ethnics, then affordable housing will decline disproportionately for other minority populations, placing greater displacement pressures on these groups, who may face more disadvantages on the housing market (Massey and Denton 1993; Charles 2003). Furthering our understanding of the mechanisms of the relationship between gentrification and race and ethnicity is important for fostering solutions for equitable and inclusive development and breaking persistent patterns of neighborhood inequality.
Given the long-standing inequities that blacks face when it comes to residential inequality and the housing market, the findings of this study call for targeted interventions to stem processes of neighborhood change exacerbating disadvantage for black urban residents. Whereas black neighborhoods were once neglected and in decline, the recent investment and socioeconomic upgrading in these areas should not benefit only its landlords and newcomers. Strong targeted policies that protect long-term and low-income renters and homeowners from displacement, such as subsidies and legal protections from price increases and evictions, and ensure an affordable housing supply, such as inclusionary zoning requirements, are necessary. In addition, inclusive and equitable development processes can help ensure that the benefits of neighborhood change provide opportunities and resources for all the neighborhood’s residents. Finally, policies and programs that address racial wealth disparities can contribute to stabilizing and strengthening these neighborhoods for their existing residents. The transformation of cities provides an opportunity to end the persistence of segregation within cities and mitigate the detrimental consequences that come with it, but, without strong policy interventions, racial inequities may only worsen.
Acknowledgments
This research was supported in part by the Eunice Kennedy Shriver National Institute of Child Health & Human Development of the National Institutes of Health (Grant No. T32HD007163) and the Joint Center for Housing Studies at Harvard University. I wish to thank the Editor and anonymous reviewers, Zawadi Rucks Ahidiana, Asad, Jacob Faber, Ryan Gabriel, Alexandra Killewald, Elizabeth Roberto, Robert Sampson, Ariela Schachter, Mary Waters, William Julius Wilson, and audiences of various workshops, seminars, and conferences for helpful comments on drafts of this paper. I would also like to thank Daniel Hammel, Ross Matsueda, and Elvin Wyly for sharing their data.
APPENDIX
Table A1.
Correlations between 1990–2013 Block Group Census-Based Gentrification Measure and Various Indicators (N=241)
| 1990 | 2013 | |
|---|---|---|
| % white | 0.01 | 0.24** |
| % black | 0.17** | −0.13* |
| % Hispanic | −0.08 | −0.15* |
| % Asian | −0.20** | −0.18** |
| % college-educated | −0.01 | 0.40** |
| % below poverty | −0.05 | −0.17** |
| % professional/managerial | -- | 0.32** |
| Median household income (logged) | −0.08 | 0.42** |
| Income per capita (logged) | 0.06 | 0.38** |
| Median home value (logged) | −0.08 | 0.31** |
| Median gross rent (logged) | −0.08 | 0.33** |
| Coffee shops (per 100 residents) | -- | 0.11† |
| Building permits (per sq. km.) | -- | 0.13* |
Note:
p<.01;
p<.05;
p<.10 (two-tailed test).
N=241. Correlations with coffee shops exclude one outlier.
Table A2.
Factor Loadings and Correlations for Principal Components
| Early Gentrification (Gentrification Field Surveys, 1998) |
Recent Gentrification (Census-Based Gentrification, 1990–2013) |
|||||||
|---|---|---|---|---|---|---|---|---|
| First PC (1970 tracts) |
Second PC (1970 tracts) |
First PC (1990 block groups) |
Second PC (1990 block groups) |
|||||
| Factor Loadings | Correlation | Factor Loadings | Correlation | Factor Loadings | Correlation | Factor Loadings | Correlation | |
| N | 40 | 241 | ||||||
| % vacant units | −0.04 | 0.49 | −0.08 | −0.45 | 0.02 | 0.39 | −0.03 | −0.29 |
| % homeownership | 0.24 | −0.62 | 0.65 | 0.66 | −0.32 | −0.75 | 0.46 | 0.62 |
| Median home value (logged) | 0.00 | 0.12 | 0.00 | 0.18 | 0.00 | 0.20 | 0.00 | 0.43 |
| Median gross rent (logged) | 0.00 | −0.50 | 0.00 | 0.20 | 0.00 | −0.41 | 0.00 | 0.39 |
| % multiunit structures | −0.39 | 0.84 | −0.51 | −0.45 | 0.42 | 0.80 | −0.52 | −0.57 |
| % units built over 30 years ago | −0.25 | 0.67 | 0.12 | 0.13 | −0.04 | −0.11 | 0.41 | 0.74 |
| % new resident in last 10 years | −0.09 | 0.61 | −0.22 | −0.57 | 0.17 | 0.69 | −0.24 | −0.56 |
| Distance (in feet) (sq. rt.) | 0.84 | −0.98 | −0.41 | −0.19 | −0.83 | −0.94 | −0.48 | −0.32 |
| % college-educated | −0.04 | 0.21 | 0.18 | 0.41 | 0.08 | 0.30 | 0.25 | 0.52 |
| % below poverty | −0.08 | 0.58 | −0.09 | −0.27 | 0.00 | −0.01 | −0.04 | −0.30 |
| Median household income (logged) | 0.00 | −0.56 | 0.01 | 0.44 | 0.00 | −0.53 | 0.01 | 0.58 |
| Income per capita (logged) | 0.00 | −0.22 | 0.01 | 0.35 | 0.00 | −0.01 | 0.00 | 0.43 |
| % professional/managerial | 0.00 | 0.01 | 0.20 | 0.49 | -- | -- | -- | -- |
Note: Data for percent professional and managerial are not available for 1990 block groups harmonized to 2010
census boundaries.
Footnotes
This definition distinguishes socioeconomic changes in neighborhoods from ethnoracial changes and residential displacement, which do not necessarily occur (for a recent review of definitional debates, see Brown-Saracino [2017]).
Across metropolitan areas with over 500,000 residents, Seattle ranked in the bottom third for black-white, Hispanic-white, and Asian-white dissimilarity indices, which are standard measures of the uneven distribution of residents across census tracts in a metropolitan area. Source: http://www.psc.isr.umich.edu/dis/census/segregation2010.html.
Population and migration data are author’s calculations using US Census and ACS data unless otherwise noted.
Based on 2009–2013 ACS 5-year estimates, referred to as 2013 hereafter.
All demographic calculations using data prior to 1970 are from Taylor (1994).
The remainder came from Europe (15%), East Africa (12%), Latin America (12%), and Central America (9%).
Census-based measures of gentrification can also reflect incumbent mobility, changes in poverty policies, or price spillovers from adjacent neighborhoods (Waldorf, 1991; Wyly and Hammel, 1999; Owens, 2012; Hwang and Sampson, 2014).
A few variables were not available in the Neighborhood Change Database, and I use variables from Brown University’s Longitudinal Tract Database instead.
Supplementary analysis of early gentrification using census-based measures yield coefficients with distinct magnitudes but similar conclusions and are available upon request along with all supplementary analyses described.
Measures that consider block groups to be gentrifiable based on the metropolitan area median household income produce similar findings.
They had greater increases in their share of college-educated residents compared to nongentrifying tracts in the 1970s and experienced declines in poverty rates during the 1980s, while nongentrifying tracts experienced large poverty increases.
I use the square root of this distance assuming the importance of proximity in predicting gentrification decreases nonlinearly as the distance increases. Using log distance alters the principal components but does not change the main results.
Using components constructed separately for the residential and proximity variables and socioeconomic status variables yield similar main results.
For details, see: http://www.icpsr.umich.edu/icpsrweb/NACJD/studies/9741.
Results are similar in supplementary models that include only violent crime.
Crime rates in 1970 are not reported for the area occupying the University of Washington, and 1990 crime rates are not available for three block groups located in two census tracts that are partially outside of city boundaries. I also exclude these from analysis, but the main results are similar in models excluding crime rates and including tracts or block groups with missing crime data.
The method uses an alternative function in the maximum likelihood estimation that guarantees finite estimates and reduces both the bias and variance that occurs in logistic regression that is particularly problematic for small sample sizes.
For details, see: http://www.icpsr.umich.edu/icpsrweb/ICPSR/studies/28701.
Based on the 1990 black-white dissimilarity index. Source: http://www.s4.brown.edu/us2010/segregation2010/Default.aspx.
The change in the black population during the period is also negative and statistically significant when included in the model, but it does not mediate the effects of the share of Asians or blacks and does not significantly explain additional variation in the outcome.
Hwang’s (2015) analysis of the stage of gentrification among WH’s sample of gentrifying census tracts in Seattle found that neighborhoods in the more advanced stages of gentrification were associated with declines in their foreign-born black population and increases in the foreign-born Asian population from 2000–2009. This sample of tracts, however, is distinct from the analysis sample of contemporary gentrification and represents tracts in their advanced stages of gentrification.
REFERENCES
- Anderson Elijah. 2012. “The Iconic Ghetto.” The ANNALS of the American Academy of Political and Social Science 642(1):8–24. [Google Scholar]
- -----. 1990. Streetwise: Race, Class, and Change in an Urban Community. Chicago, IL: University of Chicago Press. [Google Scholar]
- Anderson Matthew B., and Sternberg Carolina. 2013. “‘Non-White’ Gentrification in Chicago’s Bronzeville and Pilsen: Racial Economy and the Intraurban Contingency of Urban Redevelopment.” Urban Affairs Review 49(3):435–67. [Google Scholar]
- Bader Michael D.M. 2011. “Reassessing Residential Preferences for Redevelopment.” City and Community 10(3):311–337. [Google Scholar]
- Ball Harry V., and Yamamura Douglas S.. 1960. “Ethnic Discrimination and the Marketplace: A Study of Landlords’ Preferences in a Polyethnic Community.” American Sociological Review 25(5):687–694. [Google Scholar]
- Berrey Ellen C. 2005. “Divided over Diversity: Political Discourse in a Chicago Neighborhood.” City & Community 4(2):143–170. [Google Scholar]
- Bostic Raphael W., and Martin Richard W.. 2003. “Black Home-owners as a Gentrifying Force? Neighbourhood Dynamics in the Context of Minority Home-ownership.” Urban Studies 40(12):2427–2449. [Google Scholar]
- Boyd Michelle. 2008. “Defensive Development: The Role of Racial Conflict in Gentrification.” Urban Affairs Review 43(6):751–76. [Google Scholar]
- Brown-Saracino Japonica. 2017. “Explicating Divided Approaches to Gentrification and Growing Income Inequality.” Annual Review of Sociology 43:515–539. [Google Scholar]
- -----. 2009. A Neighborhood That Never Changes: Gentrification, Social Preservation, and the Search for Authenticity. Chicago, IL: The University of Chicago Press. [Google Scholar]
- Charles Camille Zubrinsky. 2003. “The Dynamics of Racial Residential Segregation.” Annual Review of Sociology 29:167–207. [Google Scholar]
- Chin Doug. 2001. Seattle’s International District: The Making of a Pan-Asian American Community. Seattle, WA: International Examiner Press. [Google Scholar]
- Crowder Kyle, Pais Jeremy, and South Scott J.. 2012. “Neighborhood Diversity, Metropolitan Constraints, and Household Migration.” American Sociological Review 77(3):325–353. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dávila Arlene. 2004. Barrio Dreams: Puerto Ricans, Latinos, and the Neoliberal City. Berkeley, CA: University of California Press. [Google Scholar]
- Desmond Matthew, and Shollenberger Tracy. 2015. “Forced Displacement from Rental Housing: Prevalence and Neighborhood Consequences.” Demography 52:1751–72. [DOI] [PubMed] [Google Scholar]
- Douglas Gordon C. 2012. The Edge of the Island: Cultural Ideology and Neighbourhood Identity at the Gentrification Frontier. Urban Studies 49(16):3579–94. [Google Scholar]
- Ellen Ingrid Gould. 2000. Sharing America’s Neighborhoods: The Prospects for Stable Racial Integration. Cambridge, MA: Harvard University Press. [Google Scholar]
- Firth David. 1993. “Bias Reduction of Maximum Likelihood Estimates.” Biometrika 80(1):27–38. [Google Scholar]
- Freeman Lance. 2011. There Goes the Hood: Views of Gentrification from the Ground Up. Philadelphia, PA: Temple University Press. [Google Scholar]
- -----. 2009. “Neighbourhood Diversity, Metropolitan Segregation, and Gentrification: What Are The Links in the US?” Urban Studies 46(10):2079–2101. [Google Scholar]
- -----, and Cai Tiancheng. 2015. “White Entry into Black Neighborhoods: Advent of a New Era?” The ANNALS of the American Academy of Political and Social Science 660(1):302–318. [Google Scholar]
- Frey William H., and Farley Reynolds. 1996. “Latino, Asian, and Black Segregation in US Metropolitan Areas: Are Muliethnic Metros Different?” Demography 33(1):35–50. [PubMed] [Google Scholar]
- Goetz Edward. 2011. “Gentrification in Black and White the Racial Impact of Public Housing Demolition in American Cities.” Urban Studies 48(8):1581–1604. [DOI] [PubMed] [Google Scholar]
- Hackworth Jason, and Smith Neil. 2001. “The Changing State of Gentrification.” Tijdschrift Voor Economische En Sociale Geografie 92(4):464–477. [Google Scholar]
- Hamnett Chris. 1991. “The blind men and the elephant: the explanation of gentrification.” Transactions of the Institute of British Geographers 16(2):173–189. [Google Scholar]
- Hwang Jackelyn. 2016a. “Pioneers of Gentrification: Transformation in Global Neighborhoods in Urban America in the Late Twentieth Century.” Demography 53(1):189–213. [DOI] [PMC free article] [PubMed] [Google Scholar]
- -----. 2016b. “The Social Construction of a Gentrifying Neighborhood: Reifying and Redefining Identity and Boundaries in Inequality.” Urban Affairs Review 52(1):98–128. [Google Scholar]
- -----. 2015. “Gentrification in Changing Cities: Immigration, New Diversity, and Racial Inequality in Neighborhood Renewal.” The Annals of the American Academy of Political and Social Science 660(1):319–40. [Google Scholar]
- -----, and Lin Jeffrey. 2016. “What Have We Learned About the Causes of Recent Gentrification?.” Cityscape 18(3):9–26. [Google Scholar]
- -----, and Sampson Robert J.. 2014. “Divergent Pathways of Gentrification: Racial Inequality and the Social Order of Renewal in Chicago Neighborhoods.” American Sociological Review 79(4):726–51. [Google Scholar]
- Hyra Derek S. 2017. Race, Class, and Politics in the Cappuccino City. Chicago, IL: University of Chicago Press. [Google Scholar]
- -----. 2012. “Conceptualizing the New Urban Renewal: Comparing the Past to the Present.” Urban Affairs Review 48(4):498–527. [Google Scholar]
- -----. 2008. The New Urban Renewal: The Economic Transformation of Harlem and Bronzeville. Chicago, IL: University of Chicago Press. [Google Scholar]
- Iceland John. 2004. “Beyond Black and White: Metropolitan Residential Segregation in Multi-Ethnic America.” Social Science Research 33(2):248–271. [Google Scholar]
- Kirk David S., and Laub John H.. 2010. “Neighborhood Change and Crime in the Modern Metropolis.” Crime and Justice 39(1):441–502. [Google Scholar]
- Krysan Maria, Couper Mick P., Farley Reynolds, and Forman Tyrone A.. 2009. “Does Race Matter in Neighborhood Preferences? Results from a Video Experiment.” American Journal of Sociology 115(2):527–559. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Krysan Maria, and Crowder Kyle. 2017. Cycle of Segregation: Social Processes and Residential Stratification. New York: Russell Sage Foundation. [Google Scholar]
- Landis John D. 2016. “Tracking and Explaining Neighborhood Socioeconomic Change in US Metropolitan Areas between 1990 and 2010.” Housing Policy Debate 26(1):2–52. [Google Scholar]
- Lees Loretta, and Ley David. 2008. “Introduction to Special Issue on Gentrification and Public Policy.” Urban Studies 45(12):2379–84. [Google Scholar]
- Lewis Valerie A., Emerson Michael O., and Klineberg Stephen L.. 2011. “Who We’ll Live With: Neighborhood Racial Composition Preferences of Whites, Blacks and Latinos.” Social Forces 89(4):1385–1407. [Google Scholar]
- Ley David. 1996. The New Middle Class and the Remaking of the Central City. New York: Oxford University Press. [Google Scholar]
- Lloyd Richard. 2006. Neo-Bohemia: Art and Commerce in the Postindustrial City. New York: Routledge. [Google Scholar]
- Logan John R., Stults Brian J., and Farley Reynolds. 2004. “Segregation of Minorities in the Metropolis: Two Decades of Change.” Demography 41(1):1–22. [DOI] [PubMed] [Google Scholar]
- Logan John R., and Zhang Charles. 2010. “Global Neighborhoods: New Pathways to Diversity and Separation.” American Journal of Sociology 115(4):1069–1109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Massey Douglas S., and Denton Nancy A.. 1993. American Apartheid: Segregation and the Making of the Underclass. Cambridge, MA: Harvard University Press. [Google Scholar]
- Mele Christopher. 2000. Selling the Lower East Side: Culture, Real Estate, and Resistance in New York City. Minneapolis, MN: University of Minnesota Press. [Google Scholar]
- Moore Kesha S. 2009. “Gentrification in Black Face?: The Return of the Black Middle Class to Urban Neighborhoods.” Urban Geography 30(2):118–142. [Google Scholar]
- Morrill Richard. 2003. “Gentrification, Class, and Growth Management in Seattle, 1990–2000” Pp. 43–92 in Global Perspectives on Urbanization, edited by Pomeroy G and Webster G. Lanham, MD: University Press of America. [Google Scholar]
- Owens Ann. 2012. “Neighborhoods on the Rise: A Typology of Neighborhoods Experiencing Socioeconomic Ascent.” City & Community 11(4):345–369. [Google Scholar]
- -----, and Candipan Jennifer. 2018. “Racial/Ethnic Transition and Hierarchy Among Ascending Neighborhoods.” Urban Affairs Review. Advanced online publication. Doi: 10.1177/1078087418770810. [DOI] [Google Scholar]
- Pattillo Mary E. 2007. Black on the Block: The Politics of Race and Class in the City. Chicago, IL: University of Chicago Press. [Google Scholar]
- Pérez Gina. 2004. The Near Northwest Side Story: Migration, Displacement, and Puerto Rican Families. Berkeley, CA: University of California Press. [Google Scholar]
- Pollack Stephanie, Bluestone Barry, and Billingham Chase. 2010. “Maintaining Diversity in America’s Transit-rich Neighborhoods: Tools for Equitable Neighborhood Change.” Boston, MA: Dukakis Center for Urban and Regional Policy. [Google Scholar]
- Quillian Lincoln, and Pager Devah. 2001. “Black Neighbors, Higher Crime? The Role of Racial Stereotypes in Evaluations of Neighborhood Crime.” American Journal of Sociology 107(3):717–767. [Google Scholar]
- Sampson Robert J. 2012. Great American City: Chicago and the Enduring Neighborhood Effect. Chicago and London: The University of Chicago Press. [Google Scholar]
- -----, and Raudenbush Stephen W.. 2004. “Seeing Disorder: Neighborhood Stigma and the Social Construction of ‘Broken Windows.’” Social Psychology Quarterly 67(4):319–342. [Google Scholar]
- Sanchez George J. 1997. “Face the Nation: Race, Immigration, and the Rise of Nativism in Late Twentieth Century America.” International Migration Review 31(4):1009–1030. [Google Scholar]
- Small Mario. 2009. “How Many Cases Do I Need? On Science and the Logic of Case Selection in Field-Based Research.” Ethnography 10(1):5–38. [Google Scholar]
- Smith Neil. 1998. “Gentrification” Pp. 198 in The Encyclopedia of Housing, edited by Vliet WV. London: Taylor and Francis. [Google Scholar]
- -----. 1996. The New Urban Frontier: Gentrification and the Revanchist City. London: Routledge. [Google Scholar]
- Spain Daphne. 1980. “Indicators of Urban Revitalization: Racial and Socioeconomic Changes in Central-City Housing” Pp. 27–41 in Back to the City: Issues in Neighborhood Renovation, edited by Laska SB and Spain D. Elmsford, NY: Pergamon Press. [Google Scholar]
- Tach Laura, and Emory Allison Dwyer. 2017. “Public Housing Redevelopment, Neighborhood Change, and the Restructuring of Urban Inequality.” American Journal of Sociology 123(3):686–739. [Google Scholar]
- Taylor Monique M. 2002. Harlem: Between Heaven and Hell. Minneapolis, MN: University of Minnesota Press. [Google Scholar]
- Taylor Quintard. 1994. The Forging of a Black Community: Seattle’s Central District, from 1870 through the Civil Rights Era. Seattle, WA: University of Washington Press. [Google Scholar]
- Timberlake Jeffrey M., and Johns-Wolfe Elaina. 2017. “Neighborhood Ethnoracial Composition and Gentrification in Chicago and New York, 1980 to 2010.” Urban Affairs Review 53(2):236–72. [Google Scholar]
- Waldorf Brigitte S. 1991. “A Spatial Analysis of Rent Shifts in the Chicago Rental Housing Market, 1970–1980.” Urban Geography 12(5):450–468. [Google Scholar]
- Wherry Frederick F. 2011. The Philadelphia Barrio: The Arts, Branding, and Neighborhood Transformation. Chicago, IL: University of Chicago Press. [Google Scholar]
- White Michael J., and Glick Jennifer E.. 1999. “The Impact of Immigration on Residential Segregation” Pp. 345–72 in Immigration and Opportunity: Race, Ethnicity, and Employment in the United States, edited by Bean FD and Bell-Rose S. New York: Russell Sage Foundation. [Google Scholar]
- Wilson David, and Grammenos Dennis. 2005. “Gentrification, Discourse, and the Body: Chicago’s Humboldt Park.” Environment and Planning D: Society and Space 23(2):295–312. [Google Scholar]
- Winnick Louis. 1990. New People in Old Neighborhoods: The Role of Immigrants in Rejuvenating New York’s Communities. New York: Russell Sage Foundation. [Google Scholar]
- Wong Bernard. 1998. Ethnicity and Entrepreneurship: The New Chinese Immigrants in the San Francisco Bay Area. Needham Heights, MA: Allyn & Bacon. [Google Scholar]
- Wyly Elvin K., and Hammel Daniel J.. 1999. “Islands of Decay in Seas of Renewal: Housing Policy and the Resurgence of Gentrification.” Housing Policy Debate 10(4):711–771. [Google Scholar]
- Zuk Miriam, Bierbaum Ariel H., Chapple Karen, Gorska Karolina, and Loukaitou-Sideris Anastasia. 2018. “Gentrification, Displacement, and the Role of Public Investment.” Journal of Planning Literature 33(1):31–44. [Google Scholar]
- Zukin Sharon. 2016. “Gentrification in Three Paradoxes.” City & Community 15(3):202–7. [Google Scholar]
- -----. 1987. “Gentrification: Culture and Capital in the Urban Core.” Annual Review of Sociology 13:129–147. [Google Scholar]