On Tracking and Disaggregating Center Points of Population

Abstract

In this article we explore methods for tracking and disaggregating five alternately defined mean and median center points of population, measures that can help in interpreting the forces underlying shifting settlement patterns. We argue that the point that minimizes the sum of squared great circle distances is more conceptually appealing than the center point located via the method currently employed by the U.S. Census Bureau. We also suggest that the point of minimum aggregate distance—as deployed in many other geographic applications—provides an interesting alternative to the median center historically used in population analysis, which is the crossing point of the medial lines of latitude and longitude. We then propose methods to disaggregate any of the alternatively defined center points into multiple points useful for tracking and comparing the relative influences of each of the components of population change: births, deaths, domestic (or internal) migration, and immigration. Similarly, we track and examine the shifting locations of the center points of various age, sex, and race or ethnicity groups. In a final section, we suggest that the increasing average and standard distances of individuals from the median and mean centers result from the increasingly bicoastal distribution of the U.S. population. As summary measures of all of the changes in population occurring anywhere across the nation's land area, centers of population provide an interesting conceptual platform for drilling into the variegated geographic patterns and disparate demographic forces that underlie a country's population distribution.