Abstract

Our projection of the labour supply until 2060 combines and redesigns two steps of a typical labour force projection: on the one hand the new and also redesigned population projection, on the other hand the future potential labour force, which is identified by predicted participation rates and the projected population.
In contrast to other population forecasts (e.g. the 13th coordinated population projection of the Federal Statistical Office) our model distinguishes Germans and foreigners, because employment behaviour is significantly different, especially for women. Therefore our projection is modelled by age, sex and nationality (German/non-German).
We apply a stochastic approach, an alternative to the wide-spread deterministic modelling. This allows to quantifying the forecast uncertainty by confidence intervals.
Principal component analysis (and different techniques of stochastic time series analyses) is used to forecast group-specific rates, i.e. fertility, mortality and migration rates. As a conclusion from the last IAB population projection, for which migration was partially endogenised, immigration flows and emigration flows are estimated separately. Therefore, the net migration is not given in variants, but results from separate predictions of immigration and emigration. This approach reveals an interesting phenomenon: high immigration subsequently causes higher emigration.
The birth rate can be predicted fairly accurately by the historical data: The confidence band is relatively narrow. The situation is similar for the development of mortality. In contrast, generating a long-term, but nevertheless likely course of migration was a troublesome task.
All in all, the demographic trends are known and, not surprisingly, confirmed by our up-dated forecast: As a result of the low birth rates for some decades the German Population as well as the labour force is shrinking and ageing. The currently high level of net migration will dampen these trends, but will not be able to reverse the long term trend.