This table shows the death and population data we were working with.

Institutional residents living in long term health care facilities were excluded by design because the income level of the census tract in which the institution was located might be unrelated to the income of the institutional residents.

In addition, 1 to 2% of deaths could not be coded to census tract, or CT-level income data were not available.

Altogether, after all exclusions, we had nearly 360,000 deaths in 61 million person-years of observation to work with.