Leah Boustan, an economist at UCLA and the National Bureau of Economic Research, published an Op-Ed in the New York Times last weekend that distilled the essence of her new book, Competition in the Promised Land: Black Migrants in Northern Cities and Labor Markets, described thusly by Princeton University Press:
Traditionally, the Great Black Migration has been lauded as a path to general black economic progress. Leah Boustan challenges this view, arguing instead that the migration produced winners and losers within the black community. Boustan shows that migrants themselves gained tremendously, more than doubling their earnings by moving North. But these new arrivals competed with existing black workers, limiting black–white wage convergence in Northern labor markets and slowing black economic growth. Furthermore, many white households responded to the black migration by relocating to the suburbs. White flight was motivated not only by neighborhood racial change but also by the desire on the part of white residents to avoid participating in the local public services and fiscal obligations of increasingly diverse cities.
I’m going to offer two caveats for my analysis right off the bat: First, the whole book is ambitious in scope, and proposes a provocative that migration was less clearly a Good Thing for the interests of Black advancement (reading the blurb, one might approach the book asking “compared to what?” but….). And second, it’s quite difficult to express complex research findings in short form. The book seems highly worth reading, among other reasons because of the kind of media traction it’s getting through, well, NYT Op-Eds.
That said, I found the article to be flawed in its basic assumptions about the definition and nature of racism in American urban areas, particularly as related to metropolitan real estate markets in the 20th century, and not very informed about the historiography of that phenomenon.
As an American Studies scholar and historian, I’ve worked to develop an understanding of racism as a social system of categorical inequality, a sociological understanding that I’ve found best summarized by Douglas Massey in terms of the unequal distribution of material, symbolic, and emotional resources, and by Barbara Reskin as a social web of inequalities across multiple domains, whose interactions produce racial inequity as an emergent product of their interactions. There are two takeaways here that are tremendously important. One, following Massey, is that any historian who wants to understand why people in the past did what they did had better be able to grasp the idea that the rewards and gains people pursue and the penalties and losses they try to avoid are not always strictly financial. The other, following Reskin, is that inequalities in one domain of life aren’t just shrugged off when entering another domain.
This understanding is not universally embraced, and I’ve dealt with peer reviewers whose principal grievances with my work amounted to asking whether I was sure that the actions I was describing were “caused by racism” (i.e. racial bigotry) and not some other factor (when the argument I made was that they were manifestations of systemic racism). It’s a hunch rooted in my own experience with interdisciplinary scholarship that researchers trained in more quantitative methods and those, like economists, who assume rational individualism and pursuit of utility are most likely to embrace this way of thinking, in which actions with perceptible financial rewards are presumptively judged innocent of racial intent absent decisive evidence.
I’m not altogether sure that Leah Boustan thinks this way; because it often seems like she’s writing the strictures of her discipline’s approach to racism when discussing the relationship between the growth of Black populations in northern cities and the behavior of whites in the housing market.
Here’s a representative quote qualifying an earlier claim that racial motives didn’t drive white suburbanization:
That doesn’t mean racism wasn’t a motivating factor. For the third of white households near a black enclave in 1940, concerns about new black neighbors was indeed a primary motivation. And those households moved out of the city at a higher rate than others, contributing more than a third to the white exodus. But for the remainder of urban whites, most of whom never interacted with a black family, leaving for the resource-rich suburbs was an economic calculus, one that was accelerated by the steady stream of poor migrants, both white and black, into central cities.
I’m not exactly sure what Boustan is driving at here. The literature on urban desegregation and suburban growth both before and after World War II emphasizes that close residential contact was only one way in which urban whites experienced a perceived racial threat. The prospects of school integration, for which busing across broad geographical areas, and the incorporation of Black or other minority voters into municipal politics as a strong interest group (and particularly as one that might demand school desegregation!) were also important. And city governments do a thing called planning in which future needs are provided for by present action. Spoiler alert, planning often involved anticipating demographic change and publicizing those changes in ways that could influence contemporary politics. In fact, these points seem virtually self-evident in light of recent urban histories.
So what gives here? At risk of a very uncharitable reading of the article, I’d say that Boustan generally follows a series of “yeah, but” arguments to maintain the narrowest possible definition of racism (I’m the parent of a six year-old. Let’s just say I’m familiar with yeah, but). Lots of whites left cities as Black migrants settled there? Seems like racism. Yeah, but they weren’t experiencing integration in their own neighborhoods:
Yet only a portion of white flight can be traced back to the now-classic dynamic of racial turnover. Cities were simply too segregated by race for many urban whites to encounter black neighbors. Newly available Census Bureau maps show that in 1940, the average white urban household lived three miles away from a black enclave.
True enough. But presenting this picture of enduring segregation as a reflection of primordial settlement patterns scarcely disturbed is both ahistorical and blind to the broader spatial dynamics of segregation. Indeed, prior waves of black urban migration, after Reconstruction and during the first World War, created tremendous pressure for local governments, real estate boards, and neighborhood activists to invent and codify racial segregation, displacing prior patterns of chaotic and irregular separation with the more bounded and defended racial borders. Richard Rothstein gives an accessible overview of this process here (and much more here), and it’s covered in much greater depth by Carl Nightingale in his global history of segregation. The point is not simply that public policies and private actors defended racial patterns that already existed; they remade metropolitan space in the early twentieth century, establishing in the process a set of material, symbolic, and emotional resources embodied in homes and neighborhoods. Whites held privileged access to these resources, and inevitably Black and other minority Americans who sought housing challenged parochial segregation along some racial borders, often in areas where residents were, for reasons of class or ethnicity, marginal whites. But focusing on borders as the only places where housing was racialized ignores that monoracial neighborhoods were just as much parts of the system of segregation as the border areas, and that their secure and carefully managed whiteness, reflected and codified in FHA residential security maps, related to the tense and frequently violent border areas much like levees along the banks of a rising river.
So, whites whose neighborhoods weren’t in immediate danger of integration moved out of cities as Black migrants moved in because that migration was widely understood in the political culture of urban America to signify a danger to social order and property value, something expressed in the documentary record of local governments? Seems like racism. Yeah, but white out-migrants could get a better range of services for their tax dollars in the suburbs. Here, Boustan essentially restates the famous 1956 Tiebout hypothesis (link to .pdf) underlying public choice theory–that people buy or rent homes in places where they get the best balance of taxes and services. Charles Tiebout’s hypothesis rather infamously assumed racial discrimination out of its model of multiple metropolitan jurisdictions, and Boustan does likewise.
City voters were more racially diverse and poorer than the suburban electorate, and thus less able to offer low property taxes or high-quality public services. If border residents also fled the city as black migrants arrived, even though black enclaves were miles away, these departures signaled a concern about broader city finances rather than a dislike of immediate black neighbors.
There is a very telling conflation in the first sentence there. While racial diversity and poverty characterized new urban migrants, only the latter had any logical direct effect on the fiscal health of a city. But of course no one would seriously argue that a white city resident in 1942 noticed the poverty or the social service needs of new migrants without also noting their blackness. Boustan (and Tiebout) simply strap blackness onto poverty, which preserves the illusion that the former is epiphenomenal to the real pursuit of rational advantage; Yeah, the new poor residents are black, but since they’re also poor, it doesn’t matter.
Needless to say, I find this explanation inadequate. And I think that Boustan’s rather uncomfortable with it, too. The next paragraphs hedge considerably, identifying the racial motives of whites who resided in border neighborhoods, but otherwise defending the economic motives of the other good white people. Again, at risk of being a very uncharitable reader, it seems that the main intention of this article is to preserve a myth of racial innocence among good white people that hinges on two points: if whites observed any characteristic of their new fellow urbanites as undesirable other than their race, or if they gained any monetary advantage by shunning them, their actions were not racist. Finally, Boustan argues, since the decision to move was an individual one, we can’t possibly know why people made it. Yeah, your grandfather said they left the neighborhood when those people moved in, but until we interview everybody’s grandpa, we’re going to have to assume it was purely about money.
We can see what an inadequate conclusion this is by looking at it in terms of how it conceptualizes race and racism, and how it relies on an epistemological dodge: yeah, it seems racist, but nobody really knows. First the concepts. Recall I referred to Massey’s idea of categorical (i.e. racial) inequality of material (e.g. housing, jobs, income), emotional (e.g. education, perceived mobility, empowerment), and symbolic (e.g. public esteem, regard as a good neighbor and citizen) resources. And recall I referred to Reskin’s concept of a race discrimination system of inequalities in multiple overlapping domains. Put them together and you can imagine a social context in which whites have privileged access to income and first dibs in the housing market. In this context, they also control first dibs on the best schools, either through segregation, manipulated district boundaries, or the fact that white and nonwhite residential neighborhoods are so separated. They also are accustomed to receive preferable treatment from local officials, and to command greater deference from them as citizens. Further, broad social stigmas rooted in American popular culture from its earliest development stress the moral, industrial, civic, and intellectual superiority of whites.
From those circumstances, we can imagine consequences. Black neighborhoods lack power to demand better services, creating an appearance of blight that reinforces racial stigma. This racial stigma would affect the work and educational opportunities of black residents, marginalize them as a constituency worthy of respect, and, for reasons linked to these and other causes, diminish the market appeal of property in areas where black residents lived, certainly, but also in areas where they exercised political influence, or where they might do so in the future.
Now imagine that nonwhite residents seek to challenge aspects of this oppression in multiple but simultaneous ways. Would not a challenge to political marginalization, even as simple as organizing a growing minority population as a voting bloc, portend challenges to white advantage in other domains? Would not suburban migration represent a response to this broad and diffuse perception of threat? Would not the higher value of property across suburban boundaries reflect not just what the new housing was, but what it wasn’t?
So, which hypothetical do you think resembles 1950s metropolitan America the most? I’ve got my answer, Boustan has hers. Who’s right? Well, here we get epistemological. For Boustan, the analysis of property values and census data is hard evidence, and assumptions based on the statements of people in the era of white flight are weak and theoretical:
To complicate the picture, few of them left personal accounts, and they may not have been able to articulate exactly why they moved. We are left reconstructing the pieces through careful detective work.
There have been many detectives working this case. And, while it’s impossible to conduct a complete retroactive canvass of every white urban American’s thoughts and feelings about integration and the housing market, good historical scholarship digs deep to find the voices of those people. It looks for what they wrote to public officials, what they said in public forums, the organizations they formed, and what those organizations did. Boustan gives a rather blithe dismissal of what is a large and important corpus of scholarship. And to the voices of a lot of those people who did manage to make their views known, one way or another, for good or ill.
I must say, too, that based on my own scholarship, Boustan’s presumption of municipal boundaries as given features of the landscape is a head-scratcher. Even among urban historians, I’ve been much more concerned than many with the politics of boundaries and city incorporation. So I’m acutely aware to a very geeky extent that the borders of cities are not permanent, and while they change relatively rarely compared to the shifting geographies of housing, political districting, or school attendance areas, they do change. And they change frequently for deeply racial reasons.
Consider Atlanta. In my own research (informed significantly by Kevin Kruse’s great book White Flight) on Atlanta’s territorial growth by annexation and suburban resistance to annexation, a common thread emerges: the manipulation of political boundaries in order to limit the influence of Black residents, through politicians accountable to their votes, over decisions affecting property. Atlanta’s long-serving mayor William Hartsfield played multiple parts in this play. Early in his mayoralty, he courted an organized vote of black Atlantans to remain in office, earning the ire and suspicion of his local opponents and state politicians like the segregationist Eugene Talmadge, who openly condemned the “block vote” in Atlanta as a danger to good government. Later, as Atlanta’s black population grew, Hartsfield viewed the “bloc vote” with more suspicion, backing the Plan of Improvement that annexed the affluent Buckhead area to Atlanta and significantly diluted the power of the “bloc vote.” Hartsfield made no bones about this intent, and he also made no pretense of trying to build the city’s tax base. In a 1943 letter to Buckhead businessmen seeking to win support for the annexation, Hartsfield declared, “we want voters, not money.”
This theme was a constant for decades in Atlanta’s regional politics, underlying the resistance of the unincorporated Sandy Springs area to annexation in the 1960s. Sandy Springs in the 1960s was just the sort of lily-white suburban area that Boustan imagined was not under pressure of residential integration. Indeed, I argued the exact same thing; the area was shielded by Atlanta’s policies of managed racial transition, which expanded existing areas of black settlement at the margins. The vast white and affluent spaces of Buckhead and northside Atlanta stood between Sandy Springs and existing black neighborhoods. But residents still resisted annexation out of the fear that Atlanta’s “bloc vote,” which they saw as a key swing constituency, would give black voters substantial influence over taxing their property and spending the proceeds. If they annexed, Sandy Springs residents were warned they’d be entering a government where “the lowest, least educated, and most irresponsible biped has the same vote as does the educated, responsible and conscientious taxpayer.” I mean, I guess that’s not explicitly racial, but it’ll do until explicitly racial gets here.
This political stance, that local and county governments where black voters exercised influence were profligate wasters of tax money gathered by soaking homeowners underlay major tax revolts in the 1990s, and the formation of new suburban cities in the 2000s and beyond. And it continues today as a more ambiguous, but still present factor in the politics of forming new cities in the region, as Sam Rosen has written in The Atlantic.
Again, at risk of being an uncharitable critic, Leah Boustan’s argument would benefit greatly from reading more history and, frankly, from dropping the yeah, but approach to identifying racism.
Oh, there’s this too:
Just like Trump voters in 2016, different people in the same group — white urban households — took the same action, around the same time, but for different reasons.
It’s perhaps timely to point out that a major reason a lot of those voters shared was aggrieved suburban entitlement and resentment of “political correctness,” i.e. nonwhites arguing for the existence of racism, not economic anxiety.