The Durability of Rurality: Connecting Deaths of Despair and Livelihood Characteristics in U.S. Counties

Abstract

Despite the decrease in overall mortality rate, drug related overdose deaths in the U.S. continue to increase. Factors such as unemployment and the decline of traditional livelihoods have been suggested as potential causes of deaths of despair among rural communities. However, varying definition of rurality challenge the authenticity of such relationships. The study draws on Durkheim and Marx to anticipate conditions that link livelihoods and deaths of despair through anomie and alienation. Thus, the objective of this empirical study is to examine the durability of the relationship between drug related overdose deaths and shifting sources of livelihood as moderated by different definitions of rurality. Rurality is treated as a moderator variable, one that affects the strength of relationship between independent and dependent variables. Data on the dependent variable, age adjusted drug poisoning mortality rate, an indicator of Deaths of Despair (DoD), was obtained from National Center for Health Statistics (NCHS) - Drug Poisoning Mortality by County: United States. Measures of livelihood dependence, unemployment rate, poverty rate, education, and other selected control variables were obtained from United States Department of Agriculture-Economic Research Service (USDA-ERS) Atlas of Rural and Small-Town America and American Community Survey (ACS) 5-Year Data. Five definition of rurality were purposefully selected from Office of Management and Budget (OMB), United States Department of Agriculture Economic Research Service (USDA-ERS) and the U.S. Census Bureau. Pearson correlation and multiple linear regression analysis were used to test hypotheses. Result showed that average drug poisoning mortality rate increased from 8.4 percent in 2007 to 16.7 percent in 2016. At the same time change is unemployment percent was only 0.36 from 2007 to 2016. The observed correlations between dependent and independent variables were weak or moderate (< 0.5). Results showed that despite slight variations, the relationship between sources of livelihood and drug related overdose death rates were relatively stable across different definition of rurality. Further, keeping other variables constant, unemployment rate and source of livelihood (construction and mining) had a significant positive association with drug related overdose death. For other variables, higher percentage of white non-Hispanic, older age, and poverty rate in a county were significantly associated with drug related overdose deaths. On the other hand, education level, marital rate, fertility rate, having insurance, and higher level of income for a county were significantly associated with lower drug related overdose death rates. Gender ratio was not a significant predictor of drug related overdose death. In conclusion, despite slight variations the relationship between the unemployment and DoD was relative stable across different definition of rurality. In some instances, the relationship between livelihood and deaths of despair was sharper in the most rural category of counties, but this pattern was neither marked nor consistent. This finding was in contrary to other studies. Future study should examine the durability of the rural definition using other more granular units of analysis such as tracts or zip code areas.