Grant Description: The researchers examined the relationship between local health department expenditures and county-level disparities in mortality and infant mortality rates for Black and White racial/ethnic groups. They estimated whether changes in expenditures are associated with changes in those rates over time. The objective of this project was to inform the debate about the level of resources that should be allocated to public health systems rather than to medical care or other determinants of population health.
Policy Summary: Overall, we found little evidence that 1990-1997 increases in local health department (LHD) expenditures or percentage share of public revenue allocated to LHDs are related to 1990-1997 decreases in all-cause Black (African American)-White disparities in mortality and infant mortality rates. The implication of this finding is that policy makers cannot simply increase the budgets of local health departments and expect Black-White disparities in mortality to decline. Black infant mortality rates existed only for LHDs in about 110 metropolitan areas with large numbers of Black births. For those areas, we found that increases in 1990-1997 LHD spending and percentage share of public revenue to LHDs were related consistently to declining 1990-1997 Black and White infant mortality rates, but these associations were not statistically significant, perhaps due to the small number of metropolitan areas. Building on the findings for infant mortality, we examined the relationships between LHD spending, LHD services, and Black-White infant mortality rates. We found that greater LHD spending was correlated consistently with providing a broader mix of Maternal/Child Health Services, Clinical Services, and Communicable Disease (CD) Screening and Treatment. In turn, Maternal/Child Health Services were associated strongly with reductions in White infant mortality rates, while Communicable Disease Screening and Treatment was associated with reductions in Black infant mortality rates. These patterns suggest a pathway connecting greater LHD spending with increased services that benefited women and infants, which in turn reduced Black and White infant mortality rates. The lack of an association between changes in LHD spending and changes in all-cause mortality rates for Black and White populations has different policy implications, depending on the underlying reasons for this finding: • The study’s 7-year time series may be too short to observe the consequences of LHD spending over the lifecourse of adults. The policy implication of this explanation is to conduct studies with longer time series. A related point is that LHD spending may be related to disparities in cause-specific Black-White mortality rates, but Black populations in most counties are too small to obtain cause-specific Black mortality rates from the CDS Wonder database. • Racial/ethnic disparities in mortality may not decline if local health departments place more resources into improving population health rather than reducing health disparities. The policy implication is to re-allocate at least some funds toward effective interventions for reducing Black-White disparities in mortality. • LHDs may have increased spending for interventions to reduce racial/ethnic disparities in mortality rates, but the interventions were not effective. A policy implication is to fund only interventions based on evidence that the interventions reduce mortality in Black and White populations. • Individual-level and community-level socioeconomic characteristics shape population health, and Black-White mortality disparities arise from large differences in Black-White social positions in society. The policy implication is to reduce Black-White mortality disparities by increasing public health interventions that target education, employment security, environmental protection, health behavior and other drivers of health disparities. We also found some evidence that increasing LHD resources was associated unexpectedly with increasing Black and White mortality rates. This pattern suggests that LHD spending may have increased in response to increasing mortality rates, but greater LHD spending did not translate into lower mortality rates within the study’s time series. Further analysis with a longer follow-up might clarify these results. Nonetheless, a possible explanation is that greater LHD spending was for activities that were not powerful enough to reverse the strong socioeconomic disadvantages driving Black-White disparities in mortality. More research is needed to identify the forces that produced this pattern of results, which would provide guidance to policy makers.