Thomas Schelling was a great economist, who won the Nobel. One book had this: In 1969 and 1971, Schelling published widely cited articles dealing with racial dynamics and what he termed "a general theory of tipping."[21] In these papers he showed that a preference that one's neighbors be of the same color, or even a preference for a mixture "up to some limit," could lead to total segregation,
I think generally it was taken as explaining why "block-busting" in the 60's led to a change from lily-white ownership to all-black ownership--whites had a preference for living among whites. But when you look at what he said, at least as summarized in Wikipedia, it's race-blind. In other words, blacks could have a preference for living among blacks, which could also affect housing patterns. That point was, I think, missed because we didn't have a real world example.
One factor in the discussion is the starting point. In the 1960's you had areas which were inhabited by one race. Another factor is whether you think it's a "preference", or whether it's an emergent property from network effects: you buy where a friend has bought first, "Networks" has become a new buzzword for analysis. And these days we're more familiar with "chain" immigration, where a neighborhood in the US is peopled mostly by immigrants from one town in Mexico, or wherever. .
I bought in Reston in the mid-70's, partly because it was founded from the beginning as equal opportunity housing. I was part of an outflow of people from DC moving out to the suburbs. As it turned out, despite Reston's open appeal, it didn't attract a lot of blacks--I don't think it's ever gotten much over 10 percent blacks--; most blacks moving from the District went to Prince George County, which is now majority black. These patterns fit Schelling's analysis, but how much of the underlying cause is preference, and how much is network effects is still unknown.