Building: Main Venue Building
Room: room 9
Date: 2017-02-10 09:00 AM – 10:30 AM
Last modified: 2017-01-23
Abstract
Official population projections are commonly based on deterministic cohort-component-models. In comparison to deterministic approaches, stochastic forecasts are favorable, since they are able to forecast not only the most probable future scenario, but infinitely many possible scenarios, quantifying all of them with respective probabilities. Stochastic models too can be funded on the components fertility, migration and mortality. This paper features an approach, which allows for age-specific forecasting of the components fertility and mortality in different European countries simultaneously.
The focus of this contribution is the promotion of the statistical concepts for forecasting purposes to a broader audience, especially to statistical offices and scientific organizations. After a short description of the modeling approach, the practical implementation is illustrated by forecasting the age-specific fertility rates (ASFR) for 30-year-old women as well as the age-specific survival rates (ASSR) for 70-year-old men in Italy, Germany and Austria until the year 2040, including 75%-Predictive Intervals. Furthermore the model indeed includes a wide range of other countries in Europe as well. The model is very well designed, since it makes use of the historical data and accounts for cross- as well as auto-correlation as well as high dimensionality problems.