Gregor Schroeder, Nivedya Madankara Kottayi, Bryce Burdge
This study is the first in the San Diego region to evaluate population forecast error across multiple forecast vintages and increment years spanning more than five decades. Forecast accuracy declines with increasing forecast horizon, while forecast errors are also influenced by population growth rate, population size, and jump-off discrepancy. Findings confirm the generalizability of established relationships in the forecast evaluation literature while extending prior research through analysis of errors in population change rather than solely through population level. Simple naive extrapolation methods frequently matched or exceeded the accuracy of more complex forecasting approaches, demonstrating the importance of benchmarking forecast performance.