Exploring the rural paradox: A spatial investigation of rurality and mortality

Open Access
Author:
Yang, Tse-Chuan
Graduate Program:
Rural Sociology
Degree:
Doctor of Philosophy
Document Type:
Dissertation
Date of Defense:
April 14, 2008
Committee Members:
  • Leif Jensen, Committee Chair
  • C Shannon Stokes, Committee Member
  • Diane K Mc Laughlin, Committee Member
  • Murali Haran, Committee Member
Keywords:
  • inequality
  • environmental hazards
  • social capital
  • health disparity
  • mortality
  • rurality
Abstract:
While life expectancy in the U.S. has lengthened over the 20th century, demographic sub-populations have not shared this improvement equally. Health disparities exist along several social dimensions such as gender, race/ethnicity, and socioeconomic status. Among them, somewhat less attention has been paid to geographic health disparities. In contrast to urban places, rural areas are characterized by lower insurance coverage rates, fewer health service facilities, and unfavorable socioeconomic conditions. However, these disadvantages do not translate into higher death rates. Rather, the mortality rates are lower in rural than urban areas – constituting a puzzle known as the “rural paradox.” There are three shortcomings in the literature on residential mortality differentials: 1) underdeveloped conceptualization and measurement of rurality, 2) a lack of a spatial perspective and attention to spatial dependency, and 3) incomplete attention to the array of possible determinants of the rural paradox. By addressing these issues, a more nuanced measure of rurality and four mediators between rurality and mortality are employed – population composition, social capital, internal inequality, and environmental hazards. Rurality is further divided into three dimensions: “denseness,” “exogenous economic integration” (EEI), and “farming, fishing, and forestry industries” (FFF). Methodologically, not only is the traditional analytic approach (ordinary least squares) explored, but also four other spatial modeling techniques are used: spatial error, spatial lag, mixed, and the conditional autoregressive method. An overall comparison among these methods is also provided. By analyzing county level data in the contiguous U.S., the following conclusions are reached. First, the effect of rurality on mortality differs by dimensions. EEI and FFF confirm the rural paradox, but denseness has a countervailing effect on mortality. Second, the exploratory spatial data analyses show a strong spatial clustering of mortality in the U.S.The Black Belt, Appalachia, Mississippi Valley, and Delta Region comprise the high mortality zone. The Midwest, Great Plains, and the U.S.-Mexico border region have relatively low mortality rates. Third, controlling for environmental hazards, denseness has a beneficial impact on mortality, which follows the intuitive expectation that urban residents benefit from better social conditions and other health-related services. Fourth, social capital and environmental hazards could explain the influence of FFF on mortality, though FFF still has a statistically significant effect in the full model. Fifth, population composition could partially account for EEI’s adverse impact on human health. Sixth, inequality shows a non-linear and unfavorable effect on mortality but could not help to explain the rural paradox. Seventh, spatial modeling techniques are necessary for ecological studies of mortality because they outperform the traditional method with respect to model fit and predictive ability. Finally, residence still matters because the rurality effects on mortality could not be fully explained by the four mediators. However, a clearer picture of the rural paradox is depicted.