Full Citation

Title: The Use of Stochastic Methods in Local Area Population Forecasts

Citation Type: Conference Paper

Publication Year: 2001

ISBN:

ISSN:

DOI:

NSFID:

PMCID:

PMID:

Abstract: This paper extends the method of stochastic forecasting to small area populations where migration becomes a stochastic component in the forecast. This method is applied to Minnesota in order to project trauma volume for a large primary care hospital in Minneapolis. Fertility and mortality data come from the Minnesota Department of Health. Migration data come from the census IPUMS 5% sample and the Internal Revenue Service. Fertility, mortality, and trauma are modeled using a random walk with drift model based upon the Lee Carter method. Migration is modeled as a random draw from the observed yearly values. The results provide a measure of future trauma volume that can incorporate statistical confidence intervals. Trauma is expected to increase by 38% with a upper value of 55% and a lower value of 22% on a 95% confidence interval. The method used is promising, but both more highly developed models of age-specific migration and data on migration are needed.

User Submitted?: No

Authors: Moen, Jesse; Gullickson, Aaron

Conference Name: Population Association of America

Publisher Location: Washington DC

Data Collections: IPUMS USA

Topics: Fertility and Mortality, Migration and Immigration

Countries: