Accession Number : AD0761661

Title :   Modeling and Forecasting Populations by Time Series: The Swedish Case,

Corporate Author : CALIFORNIA UNIV BERKELEY OPERATIONS RESEARCH CENTER

Personal Author(s) : Saboia,Joao

Report Date : MAY 1973

Pagination or Media Count : 32

Abstract : Time series analysis technique is used to model and forecast populations. An autoregressive (AR) and a moving average (MA) model are found which seems to fit the population of Sweden very well. Forecasts are calculated using both models, and it turns out that the AR and MA forecasts for 1965 and 1970 are better than the usual ones, which assume constant age-specific birth and death rates and no migration. Forecasts for 1975 and 1980 are also calculated. The technique is very general and, although this study is confined to Sweden, it can be used for other populations. (Author)

Descriptors :   (*POPULATION, *SWEDEN), (*TIME SERIES ANALYSIS, SWEDEN), MATHEMATICAL MODELS, RANDOM VARIABLES, LEAST SQUARES METHOD, MATHEMATICAL PREDICTION

Subject Categories : Statistics and Probability

Distribution Statement : APPROVED FOR PUBLIC RELEASE