The Accuracy of Population Projections for Subcounty Areas

Abstract Population projections for subcounty areas often are made by extrapolating past population trends. Although more sophisticated methods exist and population extrapolations long have been criticized for their weak theoretical foundations, they remain an extremely popular planning tool in part because constraints of data, time, funds, and expertise make the more sophisticated methods infeasible. The objective of this research was to document the level of accuracy of a number of commonly used extrapolative methods and, thereby, to offer suggestions for their use. For example, exponential extrapolation was found to be most accurate for rapidly growing or declining areas, whereas linear extrapolation was most accurate for moderately growing areas. More than thirty-five hundred simulated population projections for 1960 and 1970 were made for most of the methods being tested. Then the simulated projections were compared to the actual populations to measure percentage errors and tendencies to over- or underestimate. The results indicate that extrapolative methods may yield sufficiently accurate projections for many planning purposes.