Cliff Huang has created some amazing graphics depicting racial segregation in US cities. What I found fascinating was quite how sharp many of the boundaries are. They're often sharper than you would expect if the causes were simply economic.
That's because there's a powerful social phenomenom at work here, which is simply that people prefer to be with their own 'kind'. If you identify with a particular community, and that community is defined ethnically, then living outside of it can be very uncomfortable.
Which brings me to some recent research by Chris Scheitle, at Pennsylvania State University, and Kevin Dougherty, and Baylor University. They've analyzed data from over 400 people who took part in the US Congregational Life Survey to find out how long people had been attending their congregation, and how that related to whether they were in a minority or majority group.