Over the 20th century, US cities became increasingly segregated by race. Existing scholarship on the role of ‘white flight’ has focused on post-war suburbanisation, for instance by documenting the importance of highways in enabling whites to leave central cities (Baum-Snow 2007). Furthermore, economists and sociologists have argued that whites left neighbourhoods in response to black arrivals in the post-war decades, generating the distinctive American urban pattern of blacks residing in the urban core with whites populating the suburban periphery (Massey and Denton 1993, Boustan 2010).
In new research, we argue that whites began resorting themselves away from black arrivals in the first decades of the 20th century, decades before the opening of the suburbs (Shertzer and Walsh 2016). Our analysis isolates the channel of white flight from institutional barriers that constrained where blacks could live in cities. We argue that accelerating white population departures in response to black arrivals at the neighbourhood level can explain up to 34% of the increase in segregation over the 1910s and 50% over the 1920s. Importantly, our analysis suggests that, while discriminatory institutions faced by blacks were clearly important, segregation may have emerged in US cities even in their absence simply as a consequence of market choices made by white families.
The early 20th century was crucially important for the establishment of racial residential segregation in the US. Figure 1 shows the trend in isolation and dissimilarity, two measures of segregation used in the social sciences, over the entire 20th century, for ten large cities. By either measure, segregation increased markedly between 1900 and 1970. However, fully 97% of the overall increase in dissimilarity had occurred by 1930, and 63% of the overall isolation had already taken place by the same year. We thus focus on the 1900 to 1930 period in our analysis.
Figure 1 Segregation trends in the largest ten US cities, 1890-2000
Notes: Data are taken from the dataset used in Cutler et al (1999) and show the average segregation indices across Baltimore, Boston, Brooklyn, Chicago, Cincinnati, Cleveland, Detroit, Manhattan, Philadelphia, Pittsburgh, and St. Louis. We employ their adjustment factor to make the ward-level indices from 1930 and before comparable to the 1940 and onward tract-level indices.
To undertake this study, we constructed finely detailed neighbourhood demographic data for the cities of New York, Philadelphia, Pittsburgh, Boston, and Baltimore in the Northwest, and Cincinnati, Detroit, Chicago, Cleveland, and St. Louis in the Midwest, all of which saw significant influxes of black migrants from the South. While data at the level of the census tract is generally available for cities from 1940 onward, this dataset is the first to provide neighbourhood comparability at a fine unit of urban geography for earlier decades (see Shertzer et al. 2016).
The prevailing wisdom on the cause of the particularly rapid growth of black ghettos in the early 20th century is that white residents used collective action to prevent African Americans from freely settling across more northern cities. An initial wave of policies, most notably the Fair Housing Act, focused on dismantling these structural barriers that prevented blacks from locating in white neighbourhoods. The motivation behind these laws and rulings is consistent with the scholarly consensus that collective action by whites, including violence and intimidation, produced the American ghetto in the first half of the twentieth century (Massey and Denton 1993, Cutler et al. 1999).
More recent policies, such as state-level school finance equalisation schemes and federal community development block grants, have instead focused on funding disparities across jurisdictions arising from the departure of wealthier white residents from central cities. These policies address the fact that segregation – and inequality – can arise as a consequence of uncoordinated choices in the housing market. Our paper demonstrates the importance of these decentralised location decisions in generating urban segregation.
Our empirical approach aims to isolate the residential response of whites to black arrivals. In particular, we leverage exogenous changes in neighbourhood-level black populations generated by variation in out-migration rates of blacks from the South and the cross-neighbourhood location of early black arrivals in the spirit of the immigration shock literature (Altonji and Card 1991, Saiz and Wachter 2011). This strategy aims to isolate the impact of white flight from other neighbourhood-level factors that may have prompted whites to leave.
Our causal analysis confirms the existence and acceleration of white flight over the early 20th century. Our results indicate that one exogenous black arrival was associated with 1.9 white departures in the 1910s and 3.4 white departures during the 1920s. In contrast to studies on the post-war era, when whites were leaving central cities for better housing in the suburbs, we find that blacks and whites were both attracted to growing central city neighbourhoods during this time.
In the final portion of our analysis, we construct a series of counterfactual exercises aimed at understanding how much of the observed increase in segregation over the 1900-1930 period can be attributed to white flight from black arrivals. The most striking finding is the sharp increase in the contribution of flight in each subsequent decade. We estimate that flight was responsible for 34% of the increase in segregation (as measured by dissimilarity) over the 1910s, and 50% of the increase over the 1920s. The impact of flight in the latter decade is particularly important given that the 1920s saw the largest increase in segregation of any decade in the twentieth century.
Our finding, that sorting by whites out of neighbourhoods with growing black populations was a quantitatively important phenomenon, decades before the post-war opening of the suburbs, is novel. Furthermore, these results suggest that segregation may have arisen in US cities even in the absence of discrimination in the housing market, simply as a consequence of the widespread and decentralised relocation decisions of white individuals within an urban area.
These results do not mean that discriminatory institutions played no role in the construction of the US ghetto. For instance, barriers in housing markets no doubt increased the costs of settling in particular neighbourhoods, and were instrumental in determining the location of majority black areas. However, an important implication of our research is that policies that reduce barriers faced by blacks in the housing market may not prevent or reverse segregation as long as white households continue to resort themselves away from potential black neighbours.
Altonji, J, and D Card (1991), “The Effects of Immigration on the Labor Market Outcomes of Less-skilled Natives Immigration”, in J Abowd and R Freeman (eds.) Trade and the Labor Market, pp. 201-234, University of Chicago Press
Baum-Snow, N (2007), "Did highways cause suburbanization?", The Quarterly Journal of Economics 122(2) 775-805
Boustan, L P (2010), "Was Postwar Suburbanization ‘White Flight’? Evidence from the Black Migration", The Quarterly Journal of Economics 125(1) 417-443
Cutler, D M, E L Glaeser and J L Vigdor (1997), “The Rise and Decline of the American Ghetto”, Journal of Political Economy, 107 455-506
Massey, D S. and N A Denton (1993), American Apartheid: Segregation and the Making of the Underclass, Harvard University Press
Saiz, A, and S Wachter (2011), “Immigration and the Neighborhood.” American Economic Journal: Economic Policy, pp. 169-188.
Shertzer, A, R P Walsh and J R. Logan (2016), “Segregation and Neighborhood Change in Northern Cities: New Historical GIS Data from 1900 to 1930”, Historical Methods, forthcoming
Shertzer, A, and R P Walsh (2016), “Racial Sorting and the Emergence of Segregation in American Cities.” NBER Working Paper 22077