Small Area Population Projections Using Stochastic Simulation

Stuart H. Sweeney, University of California, Santa Barbara

Small area estimates and projections play a central role in some of the most important and contentious planning exercises at the metropolitan level. For example, regional transportation planning, metropolitan water planning, and the spatial allocation of future housing growth all emerge from city-level population and household projections. Despite their prominent role in planning, the methodology used to produce small area estimates and projections has seen relatively little innovation since the introduction of the housing unit method in the 1950s. That methodological stagnancy is due in part to real data limitations and the inherent complexity of the projection exercise. This paper proposes a stochastic simulation version of the housing unit method that was used to produce the 2030 population and household projections for the Southern California Association of Governments.