Example: in this simulation, agents are willing to live in environments where only 30% of neighbors are like them: https://www.youtube.com/watch?v=dnffIS2EJ30
Another good explanation of it: https://www.youtube.com/watch?v=dFl3Cfw12bo
I wonder if anyone has done a "reverse" model, taking in census data and trying to estimate a rough "Schelling threshold" -- like if all the dots started out randomly placed on the US, what percentage of "tolerance" would lead to the clustering we have in reality.
I spent about a decade working for racial justice think tanks doing research and policy work around racial segregation and its connection to racial disparities, and it was always my go to book to explain to people why the work I was doing was important for racial justice.
I go to school in St. Louis and think that it really shows what the situation is like. Here is a map I just created:
Red box - Forest Park (the big park in the middle of the city)
Green shape - Lambert International Airport (Ferguson is about a mile to the east)
Purple line - City limits -- you can clearly see that St. Louis County and St. Louis city have very different demographics.
Blue line - Imaginary line of segregation. Its a few block north of Forest Park, its not as clear on Google maps, but its very distinct on the dot map.