 |
The 130th Annual Meeting of APHA |
|
The 130th Annual Meeting of APHA |
Andreas Muller, PhD, Department of Health Services Administration, UALR, Ross Hall 207, 2801 S. University Ave., Little Rock, AR 72204-1099, 501-569-8368, axmuller@ualr.edu
Problem: In the U.S. the income gap between rich and poor has widened over the last three decades (1). The extent of income inequality is greatest in the large cities. Income inequality can be considered a public health problem since greater income inequality is related to higher age-adjusted death rates (2,3,4). Reducing income inequality might lead to greater population longevity.
The association between income inequality and mortality is typically assumed to be causal, i.e., that increases in income inequality will lead to increased mortality. Cross-sectional studies cannot test this assumption rigorously; studies of changes are better suited (5).
Data and Methods: The study is based on the State of the Nation’s Cities database (6) that permits limited analysis of change in 74 U.S. census designated metropolitan areas between 1970-90. Due to major definition changes, the multiple linear regression analysis relies on 74 communities and is restricted to the period 1980-1990.
Model Specification and Findings: All variables are expressed in first differences to measure 10-year change. The dependent variable reflects change in crude death rates per 1,000 population. The dependent variable is related to changes in three independent variables: (1) the population over 65 years old, (2) the population under 18 years old and (3) the ratio of percentile family income limits (90th/10th) reported for years 1979 and 1989. Mainly due to the change in the age distribution, the model accounts for 64% of variation in the dependent variable. The income inequality measure is positively and significantly related (b=.094; p<.05) to change in death rates after adjustment for changes in the age distribution. The second model adds two more change variables: (1) median family income and percent of college graduates 18 years and older. The second model fits the data better (R2adj.=.703; p<0001). The partial regression coefficient for the family income limits ratio is substantially reduced (b=.039; p=.42) and not statistically significant. The coefficient for change in the percent of college graduates is statistically significant and negatively related to death rates (b=-.14; p<.001). The findings are similar for large (>1,000,000 people) and small cities.
Summary and Conclusions: The study of 10-year change provides further support that income inequality is positively related to mortality. The income inequality-mortality effect appears to reflect mostly changes in the socioeconomic structure of urban populations.
Learning Objectives: After the presentation the session participant will be able to
Keywords: Social Inequalities,
Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.