Stochastic population forecast: an application to the Rome Metropolitan Area

The stochastic method has recently received much attention from researchers and seems to be a useful and efficient tool for implementing multiregional forecasting at a local level. In the paper the results are presented of a project financed by the Province of Rome to make a multiple stochastic population forecast of the Rome Metropolitan Area (RMA) using the so-called Bertino-Sonnino method, based on micro-simulations of birth-death-emigration-immigration point event processes. This forecast is based on a range of assumptions referring to the future demo-graphic dynamics over the period 2009-24 and forming three variants: high, medium and low. The outcome of the stochastic method is compared with deterministic multiregional forecasting to verify the efficiency of both methodologies. This two-step strategy allows control to be maintained over the assumed future demographic variants, at the same time linking in a probability level.