I will ask Glenn to jump in for sure.
What we did was look at the communities that had 100 housing units or less, and when there were 75% or more homes in the community that they had self-assessed as in need of major repair, we put them in the poorest bucket. In the others, less than 75% of their houses needed to be replaced or needed major repair. When we look at the average overall in those communities, it was clear to us that those with the poorest conditions—those where more than 75% of their homes needed repair—received, on average, less funding than communities of similar size.
I don't know if there's more detail you would like that maybe Glenn could provide about our determination of that status.