Back in 1968, economist John Kain put forward the "spatial mismatch hypothesis," arguing that African-Americans suffer from a particular form of labor market discrimination—blue-collar jobs were moving to the suburbs while blacks were increasingly concentrated in the inner city—that limits their employment options. More recently, William Julius Wilson discussed the same phenomenon in his 1996 book, When Work Disappears. (Not everyone agrees with the hypothesis, of course—see this link for an overview of the debate. Here's a convincing paper supporting the hypothesis.)

It's not too difficult to see how this would happen, for the most part. Racial segregation in residential areas is a fixture of the American landscape. Douglas Massey and Nancy Denton's 1993 book, American Apartheid, explores this in more detail, showing how white neighborhoods continue to discriminate against minorities; when blacks try to move to white areas, they are often told that the lots are sold, or quoted inflated prices, while banks continue to discriminate in their lending practices. (For a more historical view, see this old post on Richard Nixon's "suburban strategy" to thwart racial integration in the suburbs.)

So black suburbanization has proceeded slowly over the years. What else would we expect? Meanwhile, according to a recent GAO study, over three-fourths of all jobs are now located in suburban areas. That makes transportation an issue for those living in the inner city. It also means that urban workers will have less access to good employment networks. Since firms have classically recruited through local ads or word of mouth, blacks who live in inner cities, away from where the jobs are, are at a disadvantage. (The internet may change this somewhat, but I have no idea how much.)

Anyway, none of that is all that surprising. But there's one other question I've always wondered about: Why do so many firms move to the suburbs—and away from inner cities—in the first place? William Julius Wilson writes about what happens when work disappears, but not why all the work is disappearing. Are businesses making race-neutral decisions about where to locate—and it just so happens that the "best" place for them to locate is where all the white people are (so to speak)?

Probably not. There's an interesting 1998 paper by John Iceland and David R. Harris, economists at the University of Michigan, that I just came across which asks that very question. And what they found was that "firms in neighborhoods with a growing proportion of African-Americans are more likely to express relocation intentions," at least in Boston and Los Angeles, even after controlling for a variety of side factors (the proportion of employees who are black, race of the firm's supervisor, firm sector, changing incomes and population in the neighborhood). So in at least those two cities, businesses really are fleeing away from black people. And the end result of those decisions—whether it's conscious or unconscious—is the structural discrimination described above.

Oddly enough, in Detroit and Atlanta the same effect doesn't hold—firms don't seem to flee the neighborhood when blacks start moving in. Iceland and Harris wondered whether this was because the two cities have both had long histories of racial conflict and "balkanization," such that firms that are sensitive to race have already moved far away from African-American neighborhoods, whereas the process is only just beginning in Boston and Los Angeles. Hard to say. But it seems like an important paper, and I wonder if anyone has followed up on it in the last seven years. Looking around the internet, it doesn't seem so.