Advancing methodological knowledge within state and local demography: a case study.

Much of the academic literature dealing with state and local demography involves the development and evaluation of methods for estimating population. The focus on estimation methods is not surprising because they are used in many states to allocate resources. The quality control in regard to the validity and reliability afforded these methods by the traditional academic peer review process is important because, among other things, it serves to reduce the high potential for conflict that exists when resources are at stake. There are, however, methods being used by state and local demographers that have not been subject to peer review. While not necessarily un- sound, these "fugitive" methods serve to keep the potential for conflict high because of the uncertainty regarding their validity and reliability. This paper examines just such a situation in the form of a case study. It is a discussion of a regression model developed in Nevada following the 2000 census that led to conflict over its use to estimate the population of Clark County, Nevada in 2002. The discussion reveals statistical and methodological shortcomings in this model that lead to an alternative model not subject to these shortcomings. This example illustrates how this type of analysis and discussion can lead to a wider understanding of methods on the part of practitioners through the corrective process of academic peer review. It also suggests that states in which estimates are used to allocate resources would be well-served by subjecting new methods being considered for use to academic peer review before they are adopted. [ABSTRACT FROM AUTHOR]