Estimation and Projection of County Survival Rates

David A. Egan-Robertson, State of Wisconsin
Balkrishna Kale, State of Wisconsin

Age-specific survival rates may vary among counties due to socioeconomic, environmental and health care differences. In small-population counties, deaths in many age-sex groups are small and subject to random variation. Consequently, is actual data or a model more reasonable for projections? This paper compares two approaches to developing county-level survival rates, utilizing Wisconsin data. First, we calculate rates based on vital events from the 1990s. Second, we employ a model that applies state level survival rates to every county's population at the decade's outset to estimate deaths by age and sex over a ten-year period. Then, using the ratio of actual to estimated county deaths, we adjust the estimated distribution of deaths and compute survival rates. We discuss the problems we confronted in these approaches and the solutions we adopted. Furthermore, we compare the two sets of life expectancies generated from the survival rates at various ages and across counties.