This is a hierarchical cluster analysis of 2017 foreclosures in Connecticut with race and income done at the town level, one of four geographic units I used along with Census block groups and tracts and sub-regional drainage basins.

For this year and geographic unit, I found a strong association between towns with a lower median income and a higher than average number of foreclosures per person. I found a weaker association between lower median income, higher than average percentage of Black residents, and higher than average number of foreclosures.