Abstract
The proportion of a region's population that is elderly may change over time for many different reasons. The proportion may grow because of the net in-migration of elderly individuals, or it may grow because of the net out-migration of nonelderly residents. Furthermore, the proportion may grow if the number of nonmovers in the ''pre-elderly'' cohort is relatively high. This paper discusses in detail the ways in which this proportion may change over time. This is complemented by a state-level empirical study of elderly population growth in the United States during the late 1980s. The spatial pattern of the rate of new entry into the elderly cohort among nonmovers is found to be particularly influential in determining changes in the proportion of a state's population that is elderly.