Abstract
Population estimation methods designed for the state/provincial and county levels generally exhibit high levels of accuracy. However, there are persistent shortcomings that have not been resolved through methodological development. We argue that at least some of these shortcomings would be better understood by linking these methods with the substantive socio-economic and demographic dynamics that clearly must be underlying the changes in population that the methods are designed to measure. To illustrate our main point, we conduct a case study of Indiana over two periods, 1970–1980 and 1980–1990. Indiana is selected because a common population estimation method exhibits a common problem over the two periods: its coefficients change. We link these changes to Indiana's transition to a post-industrial economy and describe how this transition operated through demographic dynamics that ultimately affected the estimation model. The results of the case study not only suggest substantive explanations for changes in the model but also the changes that are likely until the next census – the working life of the model. In turn, these changes provide information about anticipated changes in the model's coefficients which can be used to modify them in order to maximize accuracy over the model's working life. We conclude that while these results are limited in their application, they suggest additional research along these lines would benefit population estimation methods.
Get full access to this article
View all access options for this article.