With great inequality comes great responsibility: the role of government spending on population health in the presence of changing income distributions.

Objectives: To determine the association between provincial government health and social spending and population health outcomes in Canada, separately for men and women, and account for the potential role of income inequality in modifying the association.

Methods: We used data for nine Canadian provinces, 1981 to 2017. Health outcomes and demographic data are from Statistics Canada; provincial spending data are from provincial public accounts. We model the ratio of social-to-health spending ("the ratio") on potentially avoidable mortality (PAM), life expectancy (LE), potential years of life lost (PYLL), infant mortality, and low birth weight baby incidence. We interact the ratio with the Gini coefficient to allow for income inequality modification.

Results: When the Gini coefficient is equal to its average (0.294), the ratio is associated with desirable health outcomes for adult men and women. For example, among women, a 1% increase in the ratio is associated with a 0.04% decrease in PAM, a 0.05% decrease in PYLL, and a 0.002% increase in LE. When the Gini coefficient is 0.02 higher than average, the relationship between the ratio and outcomes is twice as strong as when the Gini is at its average, other than for PAM for women. Infant-related outcomes do not have a statistically significant association with the ratio.

Conclusion: Overall, outcomes for men and women have similar associations with the ratio. Inequality increases the return to social spending, implying that those who benefit the most from social spending reap higher benefits during periods of higher inequality.