Population forecasting is the work to obtain the basic preconditions for socio-economic planning or city planning. However, there are always errors in the prediction results. In addition, although elaborate methods have been developed, it is inevitable that there will be some error as long as it is a prediction. Therefore, it is necessary to use the forecast result after understanding how much error can occur. It is expected that the error will increase if there occur a calamity. Since the target period of this study is the Great East Japan Earthquake, we will try to measure such impact.
Based on the above, in this study, we will compare and analyze the errors caused by three forecasting methods from the results of population forecast of cities, wards, towns, and villages in Japan.
Three forecasting methods are discussed, that are trend forecast, cohort change rate method and cohort factor method. Given the data at two time points in 2005 and 2010, the former two prediction methods are applied. In addition, we will use the results of population estimation by the cohort factor method by the National Institute of Population and Social Security Research, which is used as the standard for population forecast in Japan. It will be the basic forecast verification of how much error will occur in the forecast results of the 2015 population by comparing with the realized values.
人口予測における誤差から考察する予測方法・地域の影響
.png)
tuad-bulletin28-6-koto
(1.6MB)
