In the statistics of cities piece, you reduce the enormous space of possible configurations to the well known scaling relationship, evaluated across income scales. How do you think about measuring error relative to this distributional assumption? Relatedly, have you considered the configuration of attributes (including but reaching beyond income) that "best" explain -- or add information the distribution of segregation?
In the statistics of cities piece, you reduce the enormous space of possible configurations to the well known scaling relationship, evaluated across income scales. How do you think about measuring error relative to this distributional assumption? Relatedly, have you considered the configuration of attributes (including but reaching beyond income) that "best" explain -- or add information the distribution of segregation?