In this paper, I examine the differential impact of local personal income levels on infant mortality for white and non-white infants in the United States from 1962 through 2016 using county level mortality data. Non-whites have higher infant mortality rates than whites on average, but also see greater reduction in infant mortality rates associated with residence in a state with higher average non-white per capita income, as well as a state or county with higher average per capita income overall. My analysis shows the increases in average incomes in this period would be expected to be associated with about half of the observed decline in the infant mortality gap. Without the increases in average incomes, the expected gap would be twice as large. Further, without the differential impact of local personal income levels on non-white infant mortality, specifically the greater reduction of non-white infant mortality rates from higher average incomes, the expected gap would be more than three times as large. Over the full period, states with 1000 USD (in 1967 dollars; >8000 2022 USD) more in state average non-white per capita income see about 2 fewer non-white infant deaths per thousand live births. This association is strongest at the start of the study period, beginning in the early 1960s, where states with 1000 USD (in 1967 dollars) more in average non-white per capita income saw 4 fewer non-white infant deaths per thousand live births. Observing this association over rolling time windows, the relationship weakens in the mid to late 1960s. In the later periods the association shrinks to approximately 1 fewer non-white infant death per thousand live births.

This paper considers the hypothesis that hospitals in the U.S. predictably produce wider racial disparities in mortality from infectious respiratory disease when these systems are under greater strain, particularly due to a more severe period of infectious respiratory disease.