Interpolation and forecasting of population census data

• Published in: Journal of population research (Canberra, A.C.T.) A.C.T.), 2010, Vol.27 (1), p.1-13
• Main Author: Fukuda, Kosei
• Format: Article
• Language: English
• Subjects: Age; Age groups; Bayesian analysis; Bayesian cohort model; Census; Census data; Census of Population; Censuses; Cohort effect; Decomposition; Demography; Forecasting; Forecasting models; Interpolation; Japan; Modeling; Population estimates; Population forecasting; Social Sciences; Sociology; Statistical forecasts; Statistics; Studies; Time series forecasting
• ISSN: 1443-2447; 1835-9469
• DOI: 10.1007/s12546-010-9028-z

Since population censuses are not annually implemented, population estimates are needed for the intercensal period. This paper describes simultaneous implementations of the temporal interpolation and forecasting of the population census data, aggregated by age and period. Since age equals period minus cohort, age-period-cohort decomposition suffers from the identification problem. In order to overcome this problem, the Bayesian cohort model is applied. The efficacy of the model for temporal interpolation is examined in comparison with official Japanese population estimates. Empirical results suggest that the model is expected to work well in temporal interpolation. Regarding the age-period-cohort decomposition of the Japanese census data, it is shown that the cohort effect is the largest while the other two effects are very small but not negligible. With regard to the forecasting of the Japanese population, the official population forecast considerably outperforms the forecast in most forecast horizons. However, the pace of increase in root mean square error for longer-term forecasting is larger in the official population forecast than in the forecasts. As a result, a variant of the forecast is best for 10-year forecast.