The Evolution of Black Neighborhoods Since Kerner

This article studies the evolution of African American neighborhoods since the publication of the ground-breaking Kerner Commission report in 1968. We first examine how black and riot-affected neighborhoods evolved in four representative cities—Detroit, Newark, Los Angeles, and Washington, D.C.—during this period. Among black neighborhoods in these cities, we find that black neighborhoods not directly affected by riots fare better but trend similar to those that were. Notably, a number of disparities the commission identified as policy priorities—such as relatively lower income, higherpoverty, and higher unemployment—persist despite declines in racist attitudes, extreme segregation, and an increased suburbanization ofblacks. Fifty years after its publication, these findings suggest that the concerns of the Kerner Commission report remain relevant.

The National Advisory Commission on Civil Disorders, colloquially known as the Kerner Commission, was tasked by Lyndon B. Johnson (LBJ) to examine the causes and propose solutions to the destructive urban rioting that marked the 1960s. The resulting report (1968) focused primarily on the abject living conditions in many African American communities as a principal cause of the rioting. Its narrative voiced the growing recognition that efforts culminating in the adoption of civil rights legislation and Great Society programs were not enough to quell rising discontent within many black communities. Rampant housing and labor market discrimination driven by white racism, the report argues, contributed to the formation and maintenance of black ghettos, places characterized by extreme segregation, concentrated poverty, poor public good provision, and limited access to mainstream jobs. The commission outlined a program for direct government investments in housing, education, and employment coupled with active antidiscrimination campaigns especially in sub-urban areas. Deemed radical by many then, as well as by many today, the commission's omnibus recommendations were largely ignored (Kerner Report 1968; Russell 2004).

The Kerner Commission understood that neighborhood living conditions are a large component of individual and family-level socioeconomic well-being and the report notes that prior to the 1960s, neighborhood conditions facing blacks in cities outside the South remained quite difficult. The "promised land" in the north—away from the southern Jim Crow regime—had provided little of the prosperity that many black migrants sought (Olzak, Shanahan, and McEneaney 1996). The roughly fifty years since the report's publication, however, have seen a dramatic change in the overall institutional environment. Blacks as a group have made substantive gains both socially and economically; overt efforts to exclude blacks or constrain where blacks can live have largely disappeared. In particular, as Reynolds Farley describes in this volume about Detroit, peripheral city neighborhoods and formerly all-white sub-urbs now boast large or majority-black populations (2018). Yet, despite these apparent advancements in other sectors of society, racial segregation continues to characterize cities; blacks face lower educational attainment levels and higher unemployment; disparities in income and wealth persist at levels near those that commission chairman Otto Kerner and his colleagues described five decades ago.

This article examines how the features of black neighborhoods that concerned the Kerner Commission have evolved. Ultimately, we seek to understand how, in light of the attention brought by the Kerner report and others focused on problems in urban neighborhoods, these black neighborhoods have evolved vis-à-vis their white and other nonblack counterpart neighborhoods. Despite reductions in racial animus and discrimination over time, as recognized by the authors of the Kerner report, blacks still remain largely in segregated neighborhoods. Hence, it is important to determine what relative progress in black neighborhood quality, if any, has occurred since the Kerner report issued its call for black neighborhood improvement fifty years ago.

To conduct our analysis, we combine U.S. census data harmonized across five decades in the Neighborhood Change Database with tract-level information on the location of riots in these cities (Collins and Margo 2004). We match these riot locations to tracts in our set of cities and compare these areas to those that did not directly experience rioting. We focus particularly on the resorting of residents: What were the characteristics of neighborhoods that retained high proportions of black residents? How did neighborhoods elsewhere change? We characterize population counts, racial composition, educational levels, income, poverty, and public assistance use across five decades spanning 1970 through 2010 for black and transitional neighborhoods, and then compare these outcomes across census tracts representative of America's urban core that have disproportionately higher shares of black and other minorities, census tracts outside this core, and urban core census tracts directly affected by rioting of the 1960s.

Our descriptive analysis yields several key stylized facts. Consistent with earlier research, we find that, on average, the population in tracts directly affected by riots fell dramatically between 1970 and 2010. We document that these areas initially became more non-white and poor in the intervening decades. By contrast, tracts in riot-affected cities that did not directly experience rioting violence had relatively stable populations. However, they have evolved to become much more diverse racially, ethnically, and socioeconomically in part as a consequence of Hispanic population growth. Moreover, we find that neighborhoods directly affected by riots in the cities we study remain among the most economically disadvantaged today. In particular, we find that black movement from the urban core to peripheral city neighborhoods and suburbs accelerated after 1970, that amenity declines consistent with neighborhood divestment coincided with urban riots and ultimately helped foster ongoing gentrification observed in many urban neighborhoods (for example, Hyra 2012), and that socioeconomic gaps persist between black neighborhoods—with and without any history of rioting—and neighborhoods without a large concentration of black residents.

We discuss several policies in the spirit of those advocated by the Kerner Commission that, adopted in recent decades, have been aimed at improving living conditions and life chances for residents of disadvantaged neighborhoods. These policies include those aimed at desegregation, those that have generally focused on promoting economic development and employment, and those geared toward improving the quality of educational choices in these neighborhoods. Echoing the existing literature that seeks to evaluate the success of these policies, it is apparent that desegregation initiatives have had only minimal effect. Likewise, those aimed at increasing economic development and employment in black neighborhoods have largely been ineffective or have engendered complaints of gentrification and displacement. By contrast, policies that have sought to improve educational quality and choice in these neighborhoods have had more, albeit mixed, success.

We expand the scope of our inquiry to provide a descriptive comparison of U.S. neighborhoods where blacks typically live as of 2010. Given that changes in attitudes and other features of American life had led to a racial resorting both within the urban areas affected by the rioting and across the nation, the types of neighborhoods where blacks typically reside in the twenty-first century have broadened. A simple comparison reveals that blacks still live in neighborhoods that lag on a number of key indicators. In concert with the comparisons presented from the riot cities analysis, it only further amplifies the fact that, fifty years on, the concerns of the Kerner Commission report remain relevant.

DATA

Our data are drawn from three sources. The Neighborhood Change Database (NCDB), the linchpin of our analysis, consists of census tracts harmonized to the 2010 boundaries spanning five census decades since 1970. Originally constructed by the Urban Institute and sold commercially by Geolytics, Inc., the NCDB helps overcome changes in the boundaries and definitions of census tracts over time that limit longitudinal analyses in which tracts are the unit of analysis. Hence, the NCDB allow us to study how specific places have changed over time.

To the NCDB data, we match tract-level information denoting the location of riots occurring in 1967 for our representative cities: Detroit, Los Angeles, Newark, New Jersey, and Washington, D.C.¹ These cities experienced particularly deadly and destructive episodes of rioting that were featured in major news outlets, providing some of the impetus for President Johnson to establish the Kerner Commission. Finally, we augment these NCDB files with additional information from the decennial census (1970 through 2010) and the American Community Survey (for 2006 through 2010) to better characterize neighborhood environments.

Table 1 presents a descriptive breakdown of the neighborhoods used in the final analysis dataset. Our data include 2,978 tracts located within our representative cities. We then separately examine tracts nationwide to assess the contemporary status of black neighborhoods.² We define a black neighborhood as a tract whose population was 40 percent black or higher in any census year. Admittedly, this cutoff is somewhat arbitrary, but we believe, given the size of the black population and degree of segregation in most cities, that neighborhoods near this threshold likely would have been characterized as black in past decades. In addition, before 2000 almost all neighborhood transitions are one way and complete (Card, Mas, and Rothstein 2008; Casey 2018). Therefore, it is reasonable to assume that neighborhoods with such high proportions of blacks were likely to continue to trend in that direction. Our organizing frame of characterizing a black neighborhood is motivated by well-established historical and contemporary patterns of racial and socioeconomic sorting and the range of observable inequities that arise from such sorting (Massey and Denton 1993; Cutler and Glaeser 1997).

Table 1.

Breakdown by Neighborhood Type

Our riot cities generally had a high percentage of neighborhoods, on average, classified as black between 1970 and 1990: in 1970, approximately 19 percent and approximately 23 percent in 1980 and 1990. However, in the most recent census, this figure decreased to around 20 percent. This pattern is consistent with reports elsewhere suggesting that many of the former industrial cities affected by riots are losing their black population and becoming more integrated through gentrification and immigration.

For this study, we focus on the areas within four cities affected by the major riots that led to the establishment of the Kerner Commission report and remain among the most prominent examples in the late 1960s. Because the Kerner Commission was also concerned with the conditions facing black Americans nationwide and the potential for civil unrest to spread to a broader set of cities, we close by considering the twenty-first century outlook. Specifically, we compare contemporary conditions within the neighborhoods of America's one hundred largest cities to those within riot-affected cities.

HOW FAR HAVE WE COME? THE EVOLUTION OF BLACK AND RIOT-AFFECTED NEIGHBORHOODS

Figures 1 and 2 present population trends for several comparisons of neighborhoods within riot cities, those that directly experienced riots versus neighborhoods overall within riot cities, and black neighborhoods that experience rioting and those that did not.

As figure 1 shows, neighborhoods in our riot cities that experienced rioting depopulated, on average, after 1970. The most pronounced drop occurred between 1970 and 1980, from roughly 4,500 to less than 4,000, declining more gradually thereafter. This depopulation does not occur when looking over the pooled set of neighborhoods in the riot-city metro areas overall. After a relatively constant population of approximately 3,250 between 1970 and 1980, these neighborhoods grow over time to almost four thousand people per tract, on average, by 2010. Figure 2 compares the same population trends restricted to black neighborhoods that directly experienced riots and those that did not within these cities. The similarities are striking, in that the depopulation of riot-affected neighborhoods mirrors that of unaffected black neighborhoods.

In tables 2 through 5, we summarize changes in the racial and ethnic composition of neighborhoods in our riot-affected cities since 1970. Riot-affected neighborhoods transitioned from being overwhelmingly black in 1970 to later being almost evenly split across black and non-black residents (table 2), including a rising share of foreign-born and Hispanic residents. Neighborhoods also become increasingly racially and ethnically diverse (table 3). This shift holds for the full set of black neighborhoods, including those affected by riots (table 4) and those simply in riot-affected cities (table 5).

Table 6 examines transitions among black neighborhoods since 1970. Specifically, among neighborhoods classified as black in 1970, what

Figure 1.

Population Trends, Overall and Riot-Affected Neighborhoods 4,500

Source: Authors' calculations based on the Neighborhood Change Database, 2010, and the 2010 U.S. census (2010b).

Note: Riot tracts correspond to those identified by Collins and Margo (2004).

Figure 2.

Population Trends, Overall and Riot-Affected Black Neighborhoods

Source: Authors' calculations based on the Neighborhood Change Database, 2010, and the 2010 U.S. census (2010b).

Note: Riot tracts correspond to those identified by Collins and Margo (2004).

Table 2.

Racial-Ethnic Composition, Riot-Affected Neighborhoods

Table 3.

Racial-Ethnic Composition, Neighborhoods Overall

Table 4.

Racial-Ethnic Composition, Black Neighborhoods Affected by Riots

Table 5.

Racial-Ethnic Composition, Black Neighborhoods Overall

Table 6.

Neighborhood Transitions

was their classification in 2010? The first panel presents these statistics for the riot-affected areas we study and the second for neighborhoods not directly affected. By contrast, of the areas classified as white in 1970, 64 percent transitioned to black by 2010. Nearly 60 percent of neighborhoods classified as black in 1970 remained as black neighborhoods in 2010. Among those black neighborhoods that transitioned away, almost none became a white neighborhood: 2 percent of neighborhoods classified as black became white neighborhoods whereas 40 percent became Hispanic. Among neighborhoods not directly affected by riots and those in comparison cities, racial compositions were even more stable: 67 percent of black tracts in 1970 remained black in 2010. Only 1 percent transitioned to white over the period. Compared with riot-affected neighborhoods, only 7 percent of neighborhoods initially classified as white transitioned to black over the period.

Although the differing population trends across neighborhood types are interesting, the Kerner Commission was specifically concerned with neighborhood quality experienced by blacks who lived in these neighborhoods. Figure 3 presents educational attainment levels within the neighborhoods of riot-affected cities. At baseline in 1970, levels are higher in neighborhoods overall than in riot-affected neighborhoods. College attainment in riot-affected neighborhoods rises at a rate similar to that of neighborhoods overall but fails to catch up to the higher attainment within the average neighborhood—roughly 25 percent of residents in the average neighborhood of a city that experienced rioting had a college degree by 2010, versus fewer than 15 percent in riot-affected

Figure 3.

Education Trends, Overall and Riot-Affected Neighborhoods

Source: Authors' calculations based on the Neighborhood Change Database, 2010, and the 2010 U.S. census (2010b).

Note: Riot tracts correspond to those identified by Collins and Margo (2004).

Figure 4.

Black Education Trends, Overall and Riot-Affected Black Neighborhoods

Source: Authors' calculations based on the Neighborhood Change Database, 2010, and the 2010 U.S. census (2010b).

Note: Riot tracts correspond to those identified by Collins and Margo (2004).

Figure 5.

Income Trends, Overall and Riot-Affected Neighborhoods

Source: Authors' calculations based on the Neighborhood Change Database, 2010, and the 2010 U.S. census (2010b).

Note: Average tract-level household income is adjusted for inflation using the personal consumption expenditure deflator for 2010. Riot tracts correspond to those identified by Collins and Margo (2004).

neighborhoods. In figure 4, restricting our view to black neighborhoods within our representative cities, we see a similar pattern for college attainment. Black neighborhoods generally look slightly better than riot-affected neighborhoods with respect to the proportion of residents with college degrees, though both follow the same trend over time—mirroring individual-level shifts in educational attainment concurrent with the changing structure of the economy favoring higher level skills (see, for example, Autor 2014).

We next examine household income within neighborhoods (figure 5). Within the four cities, riot and nonriot neighborhoods follow a similar trend in regard to income growth, but show large, persistent gaps. In 1970, the gap is roughly $20,000. By 2000, it is $25,000. By 2010, the average nonriot neighborhood residents had incomes over $70,000, versus approximately $45,000 in riot-affected neighborhoods. The riot-affected neighborhoods experience only $10,000 of real income growth over a forty-year period. When we examine differences between black and riot-affected neighborhoods, black neighborhoods overall fare only slightly better (figure 6). A modest $3,000 to $4,000 advantage persists for black neighborhoods relative to riot-affected neighborhoods, and the trends are almost identical over time.

Unemployment is a primary driver of income statistics; therefore, we next trace unemployment trends in our neighborhoods of interest. In figure 7, we show that riot neighborhoods begin with slightly higher unemployment rates in 1970, though both riot and nonriot neighborhoods have, on average, unemployment rates below 10 percent. However, between 1970 and 1990, a sizable neighborhood unemployment gap emerges—roughly 10 percentage points—before closing somewhat thereafter. Accordingly, this pattern mirrors the income trends shown previously. Black neighborhood unemployment trends in riot cities (figure 8) track similarly to neighborhood trends generally for riot-affected neighborhoods, which again is borne out in the previous series of income-based trends (figures 5 and 6).

We close our discussion of neighborhood economic and amenity trends by examining poverty and public assistance receipt within these neighborhoods. As figure 9 shows, within

Figure 6.

Income Trends, Overall and Riot-Affected Black Neighborhoods

Source: Authors' calculations based on the Neighborhood Change Database, 2010, and the 2010 U.S. census (2010b).

Note: Average tract-level household income is adjusted for inflation using the personal consumption expenditure deflator for 2010. Riot tracts correspond to those identified by Collins and Margo (2004).

Figure 7.

Unemployment Trends, Overall and Riot-Affected Neighborhoods

Source: Authors' calculations based on the Neighborhood Change Database, 2010, and the 2010 U.S. census (2010b).

Note: Average tract-level household income is adjusted for inflation using the personal consumption expenditure deflator for 2010. Riot tracts correspond to those identified by Collins and Margo (2004).

Figure 8.

Unemployment Trends, Overall and Riot-Affected Black Neighborhoods

Source: Authors' calculations based on the Neighborhood Change Database, 2010, and the 2010 U.S. census (2010b).

Note: Riot tracts correspond to those identified by Collins and Margo (2004).

our sample, poverty rates in riot-affected neighborhoods are roughly 15 percentage points higher than in neighborhoods overall. Riot neighborhoods see a growth trend in poverty from 1970 through 1990 and a leveling off thereafter, whereas average neighborhoods (neighborhoods overall) within these cities experience very gradual growth in poverty over time, and from a much lower baseline. A sharp decrease in public assistance use occurs between 2000 and 2010 for both riot neighborhoods and neighborhoods overall, with public assistance use between both types of neighborhoods converging to less than 10 percent.

In figure 10, among the black neighborhoods, we observe that riot-affected neighborhoods have even higher poverty rates than the already-elevated levels generally, by approximately 3 to 5 percentage points. These rates, above 30 percent after 1970, are higher than national averages. For black neighborhoods, these levels roughly mirror trends at the individual and family level for black Americans, for whom poverty has held steady at between approximately one in four for individuals and one in three for families (Semega, Fontenot, and Kolla 2017). Meanwhile, public assistance use again falls and converges between 2000 and 2010 to under 10 percent. As shown in figure 9, public assistance falls dramatically, a divergence from poverty trends between 2000 and 2010. Given that no comparable, large reduction in poverty, the divergence in contemporary poverty-welfare assistance trends shown in figures 9 and 10 is broadly consistent with major policy changes to the nation's cash welfare program for the poor in 1996 that, by many accounts, have led to poverty without welfare benefits (Blank 2009; Ziliak 2016; Shaefer, Edin, and Talbert 2015).

What Worked and Did Not Work: Neighborhood Policy Since Kerner

Aside from its ostensibly controversial assignment of blame, the Kerner report is notable for proposing a broad set investments—"enrichment" in its language—in the areas of housing, education, employment, and general welfare, coupled with "integration." The commission members believed investments in these areas would satisfy the dual aims of reducing the likelihood of additional violence and improving the life conditions and future prospects of black families and their children.

Figure 9.

Poverty and Welfare Trends, Overall and Riot-Affected Neighborhoods

Source: Authors' calculations based on the Neighborhood Change Database, 2010, and the 2010 U.S. census (2010b).

Note: Percent poor is calculated as the proportion of families under sixty-five living below the poverty line. Riot tracts correspond to those identified by Collins and Margo (2004).

Figure 10.

Poverty and Welfare Trends, Overall and Riot-Affected Black Neighborhoods

Source: Authors' calculations based on the Neighborhood Change Database, 2010, and the 2010 U.S. census (2010b).

Note: Percent poor is calculated as the proportion of families under sixty-five living below the poverty line.

The recommendations and observations of Kerner, many of which Rick Loessberg and John Koskinen summarize in this volume (2018), turned out to be especially prescient, given the link between neighborhood-level conditions and long-term individual-level socioeconomic outcomes. Almost none of these proposals, however, were directly implemented in the immediate aftermath of publication. Although a comprehensive review is beyond the scope of this article, we briefly discuss several of these programs that have had impacts on black (and other) neighborhoods and comment on their progress relative to the goals outlined in the Kerner Commission report.

A range of federally funded place-based policies, whether focused directly on people or in-directly by facilitating business growth to address poverty, high unemployment, and urban blight, have been implemented since the late 1960s. Though these policies were not necessarily as ambitious and targeted as those described in the report, they were similar in spirit, particularly in terms of promoting integration of neighborhoods, encouraging neighborhood economic development, and improving access to quality education in desegregated schools. In addition, though not place-based, research suggests that several key transfer programs such as SNAP (food stamps) had substantial positive impacts on economically disadvantaged communities including improved long-term outcomes for recipient children (see, for example, Hoynes, Schanzenbach, and Almond 2016).

The passage of legislation such as the Fair Housing Act (FHA) of 1968 and the Equal Credit Opportunity Act (ECOA) of 1974 were key steps toward the comprehensive federal framework for combating discrimination in the housing market and promoting desegregation that the commission considered of paramount importance. The FHA prohibited discrimination in both rental and owner-occupied housing and, at the time, many hoped it would help promote integration of blacks into higher quality, white neighborhoods (Massey 2015). However, it included no provisions prohibiting discrimination in mortgage lending markets, an omission corrected by the ECOA. Despite the symbolic power of this and related legislation, segregation between blacks and whites has declined nationally only slightly, though with some variation across cities (Rugh and Massey 2014). Moreover, audit studies suggest that discrimination in housing markets as well as in mortgage lending continue to affect black house-holds.

Likewise, the Housing and Urban Development Acts of 1968 that provided funding for integrated developments outside central cities, the Community Reinvestment Act of 1974 that prohibited redlining of black neighborhoods, and the Housing and Community Development Act of 1974 that established a block grant program to support community development in the decade subsequent to the Kerner Commission report's publication represent important attempts to improve the quality and stability of black neighborhoods (see, for example, Massey 2015). Since this early legislation, much neighborhood development policy has come in the form of place-based economic development initiatives, some of which Jamein Cunningham and Rob Gillezeau also discuss in this volume (2018).

Among the most ambitious of these policies that are in the spirit, if not in the implementation, of the Kerner Commission report have been the establishment of enterprise and empowerment zones to promote economic development and private investment, primarily beginning by federal initiative in the 1980s and 1990s and later by states (Neumark and Simpson 2014). These strategies, which vary in intensity and size across jurisdictions, are often organized as tax incentives to promote hiring local workers, firm location, and investment within distressed areas (see, for example, Ladd 1994).

Evaluations of the net benefits of establishing such zones, however, are somewhat controversial (see Neumark and Simpson 2014). David Neumark and Jed Kolko, for example, argue in an evaluation of California's enterprise zones that the program was largely ineffective (2010). Like a number of studies published earlier, they find little evidence of an impact from designating a place as an enterprise zone on economic activity. In contrast, recent evidence on empowerment zones suggests that, in areas they were instituted, place-based policies substantially increase jobs and wages relative to places considered for status but rejected (Busso, Gregory, and Kline 2013). Even if the establishment of such zones were beneficial, additional concerns are that such policies help spur gentrification forces that raise rents and other costs in recipient neighborhoods, leading to fears of displacement of typically poorer, nonwhite incumbent residents. As a consequence, potential benefits accruing to existing black residents may be limited. This can occur when, for example, tax incentives to promote development are financed by diminished educational investments—which themselves promote economic development, and when incentives are targeted neither toward firms with a strong propensity to hire nor toward adults who wish to reenter the workforce (Bartik 2018).

Education policies implemented since Kerner include expansion of the federally funded Head Start prekindergarten program, enacted in 1965 to provide educational programming, meals, and other developmental activities to three- and four-year-old children living in poor families. The program was part of President Johnson's war on poverty and expanded in generosity throughout the 1960s and early 1970s with the goal of improving child development and subsequent outcomes during school-age years and into adulthood. Today, several states supplement Head Start with their own pre-K and early pre-K programs (Garces, Thomas, and Currie 2002; Currie 2006).

Within the domain of secondary education, policy responses to the racial stratification inherent in many cities sought to both integrate schools and improve school quality experienced by blacks. Policies adopted in this context include school district-level efforts to equalize educational opportunities across neighborhoods via the busing and transfer of students into higher quality schools and neighborhoods, often based on race and economic status. Frequently the result of court-mandated intervention to desegregate schools, such efforts have borne positive results for students yet at times met with resistance from families who fall within the school boundary (see, for example, Billings, Deming, and Rockoff 2014); many of these programs were abandoned throughout the 1990s and 2000s.

As a result of the pushback stemming, in part, from the altered racial composition of schools and removal of students from their neighborhoods, decentralized school choice mechanisms to improve educational opportunities among black students became increasingly attractive. School choice approaches to improving educational quality among low-income children in disadvantaged neighborhoods introduce competitive market principles into the K-12 educational space by providing vouchers for low- and moderate-income families to attend private or parochial schools or via the introduction of charter schools, privately run public schools. Although proponents laud competition and accountability by giving low-income students and families increased market power, the evidence on school choice and academic achievement is mixed at best and in some instances negative (Angrist, Pathak, and Walters 2013; Baude et al. 2018; Dynarski and Nichols 2017).

IMPLICATIONS FOR THE TWENTY-FIRST CENTURY: BLACK NEIGHBORHOODS IN THE CHANGING CITY

As we approach the third decade of the twenty-first century, it is important to recognize that the neighborhoods where blacks live are much more variable than they were when the Kerner report was issued. In addition to suburbanization in many of the larger metropolitan areas, the last two to three decades have seen many blacks return to the South as well as move to other regions of the country—the Southwest, for example (Frazier, Tettey-Fio, and Henry 2016). Thus, in this final section we highlight 2010 socioeconomic statistics for black neighborhoods from a set of comparison cities not directly affected by the riots but generally areas where blacks live. Specifically, these are defined as the one hundred largest metropolitan areas in the 2010 census.³ A number of these cities were affected by riots—in some cases much earlier in the decade or in the aftermath of the assassination of Martin Luther King Jr. in 1968.

Table 7.

National Neighborhood Characteristics, 2010

We begin by comparing black neighborhoods in these municipalities to the average neighborhood nationwide (see table 7). Similar to the riot-affected areas we focused on initially, these black neighborhoods generally lag the average U.S. neighborhood in 2010 across several socioeconomic indicators. Specifically, poverty and unemployment are generally higher, roughly 10 and 6 percentage points, respectively. Moreover, we observe a nearly $15,000 household income gap. Perhaps unsurprising as well, college attainment of those living in these neighborhoods lags that of the average national neighborhood by 7 percentage points. Overall, this fits a pattern that emerges across the indicators presented: riot-affected neighborhoods—and black neighborhoods more generally—lag the nation as a whole on a broad range of amenity and socioeconomic indicators.

RACIAL SORTING AND ACCESS TO HIGH AMENITY NEIGHBORHOODS

Central to the Kerner Commission's concerns was desegregation and improving the quality of living conditions for black citizens. Given that black neighborhoods in riot-affected cities remain largely stable and persistently segregated, it is of interest to explore the degree to which blacks have access to higher amenity neighborhoods. As noted, most neighborhood diversity outside the core black neighborhoods in 1970 has been driven by growth in the Hispanic population. Neighborhood changes involving black migration into former white neighborhoods have largely been one way and complete (Card, Mas, and Rothstein 2008). In an era of gentrification, however, there remains the possibility that desegregation will occur through the mechanism of white entry into black neighborhoods.

Table 8.

Evolution of Neighborhoods

Table 8 explores how such access to amenities has evolved over time by exploring the joint distribution of average educational attainment and percentage black. While not a perfect proxy, the educational level of residents is usually correlated with a whole host of related components of neighborhood quality. In particular, highly educated neighborhoods typically have higher quality schools, more diversity in food and grocery options, and lower crime. Scholars have speculated that the prevalence of socioeconomically stratified neighborhoods is contributing to social inequality (see, for example, Putnam 2015). Studying how this joint distribution changes over time should provide some insight into how the neighborhoods where blacks live have changed.

We begin by exploring black neighborhoods overall, focusing on neighborhoods that are at least 20 percent or more (plus) black and those 50 percent or more black, respectively. The first panel of table 8, 20 percent or more black, reflects the rise in the number of people holding a college degree nationwide. In 1970, only 5 percent of the 683 neighborhoods that were at least 20 percent black had a population in which at least 20 percent held a college degree; roughly 1 percent of these neighborhoods featured a population in which 60 percent of the population held a college degree. By 2010, the number of neighborhoods where the population was at least 20 percent black had grown to 839. The fraction of these neighborhoods where at least 20 percent of the residents held a college degree had grown to 29 percent, the fraction in which 40 percent held a college degree was around 11 percent, while the fraction of these neighborhoods where the majority of the population held a college degree had grown to roughly 5 percent.

The lower panel of table 8 focuses on neighborhoods where blacks made up at least 60 percent of the population. Reflecting the broader trends in the United States, we observe steady growth in the share of these neighborhoods with relatively well-educated populations. In 1970, only 2 percent of neighborhoods had a population that was at least 60 percent black in which 20 percent of the population held a college degree; only 1 percent had a population where 60 percent of the population held a college degree. By contrast, in 2010 the fraction of neighborhoods with these characteristics had grown to 28 percent and 5 percent respectively. It's clear that recent concerns surrounding the societal implications of class-based segregation, driven by differences in income and education, very clearly have a racial component as well (see, for example, Putnam 2015). Tracts with large shares of highly educated residents rarely contain large shares of black residents.

Table 9 presents these statistics concentrating only on the riot-affected neighborhoods in the data. These neighborhoods generally lag behind on the share of the population that is college educated. For example, in 1970, just over I percent of these riot-affected neighborhoods in our data had a population that was at least 20 percent black and had a population where at least 20 percent held college degrees; almost none had populations where at least 50 percent held college degrees. By contrast, in 1970 majority black riot-affected neighborhoods where at least 20 percent of the population held a college degree were nonexistent. By 2010, however, roughly 20 percent of neighborhoods that were at least 20 percent black had a population with 20 percent of their population holding degrees; II percent had populations where 40 percent of the population held college degrees; 5 percent of these neighborhoods had populations where 60 percent of the population held degrees.

Table 9.

Evolution of Riot-Affected Neighborhoods

Turning to the riot-affected neighborhoods that were at least 60 percent black, by 2010, 24 percent had a population that was 20 percent college educated. However, very few of these heavily black riot-affected neighborhoods featured high fractions of college-educated residents: only 7 percent of these neighborhoods had 40 percent or more of their populations that held a college degree, whereas roughly 1 percent featured more than 60 percent of their populations with college degrees. In sum, despite the growth in educational attainment, residents of these neighborhoods typically continue to lag that of black neighborhoods pooled nationally.

DISCUSSION AND CONCLUSION

This article studies the evolution of black neighborhoods in the decades after the publication of the Kerner Commission report. Focusing on Detroit, Los Angeles, Newark (New Jersey), and Washington, D.C., we match riot locations in these cities with tract-level census data harmonized over five censuses. We use these data to assess how neighborhoods in these areas changed over the period, focusing on a descriptive comparison of the evolution of a set of neighborhoods directly affected by rioting and others not directly affected.

We document a number of interesting stylized facts. First, in the years after the riots, in all riot neighborhoods, especially those that were heavily black, declines on a number of quality-of-life indicators were substantial. Second, the riot-affected neighborhoods in our sample remain among the most economically challenged. In our twenty-first-century look ahead, which is perhaps most pressing, black neighborhoods across cities have tended to appear comparable to riot-affected neighborhoods in these riot-affected cities despite improvements in education among people living in these neighborhoods.

A number of caveats, however, apply to the analysis. The study is descriptive and thus limited in its prescription. In addition, many of the 2010 economic measures potentially reflect the residual effects of the Great Recession and slow recovery thereafter. Because black communities are historically among the most vulnerable, they may have been disproportionately affected.

Nevertheless, the differences in socioeconomic evolution across neighborhood types are present prior to 2010 and illustrate an important conundrum. In the fifty years since the Kerner Commission issued its report, the United States has seen tremendous advancements in educational achievement, access to elite employment, income, and wealth for some blacks—it has even elected a black president. Yet, black neighborhoods, especially those directly affected by the period of rioting studied here, persistently lag the nation as a whole on a number of key dimensions. These neighborhoods typically face greater poverty and unemployment, have lower household income, and have relatively few college graduates. Perhaps more important, convergence across neighborhoods on these measures over the five decades since the Kerner report has been minimal. Within the context of current research, the importance of these findings is clear. A growing body of work shows that place matters for individual economic well-being in large part due to the amenities available in more socioeconomically elite neighborhoods (Chetty and Hendron 2018a, 2018b; Chetty et al. 2014, 2018; Andrews et al. 2017; Islam, Minier, and Ziliak 2015). Improvement of black neighborhoods therefore remains an important policy problem for the nation into the twenty-first century.

In sum, the Kerner Commission report issued a clarion call concerning the disparity in life quality experiences and life chances facing black people as a consequence of institutionalized features of American life. Although many of its fundamental suggestions were never adopted, its lasting legacy remains that it succeeded in making the forceful case that the government indeed had a responsibility and a role to play in mitigating the social and economic harm imposed on blacks by racism and discrimination. The persistent challenges that black neighborhoods face suggest that this argument remains relevant.