Does higher income inequality adversely influence infant mortality rates? Reconciling descriptive patterns and recent research findings

作者：Siddiqi Arjumand^*; Jones Marcella K; Erwin Paul Campbell

来源：Social Science & Medicine, 2015, 131: 82-88.

DOI：10.1016/j.socscimed.2015.03.010

• Income inequality
• Infant mortality
• Multilevel modeling
• Longitudinal analysis
• State-level
• United states

As the struggle continues to explain the relatively high rates of infant mortality (IMR) exhibited in the United States, a renewed emphasis is being placed on the role of possible 'contextual' determinants. Cross-sectional and short time-series studies have found that higher income inequality is associated with higher IMR at the state level. Yet, descriptively, the longer-term trends in income inequality and in IMR seem to call such results into question. To assess whether, over the period 1990-2007, state-level income inequality is associated with state-level IMR; to examine whether the overall effect of income inequality on IMR over this period varies by state; to test whether the association between income inequality and IMR varies across this time period. IMR data - number of deaths per 1000 live births in a given state and year - were obtained from the U.S. Centers for Disease Control Wonder database. Income inequality was measured using the Gini coefficient, which varies from zero (complete equality) to 100 (complete inequality). Covariates included state-level poverty rate, median income, and proportion of high school graduates. Fixed and random effects regressions were conducted to test hypotheses. Fixed effects models suggested that, overall, during the period 1990-2007, income inequality was inversely associated with IMR(beta = -0.07, SE (0.01)). Random effects models suggested that when the relationship was allowed to vary at the state-level, it remained inverse (beta = -0.05, SE (0.01)). However, an interaction between income inequality and time suggested that, as time increased, the effect of income inequality had an increasingly positive association with total IMR (beta = 0.009, SE (0.002)). The influence of state income inequality on IMR is dependent on time, which may proxy for time-dependent aspects of societal context.

• 出版日期2015-4