Analyses of country specific population projections
On the basis of country specific projections (also probabilistic), analyses of consequences of population ageing on socio-economic development were carried out for Austria, Germany, Singapore, Japan, and the USA.
Developing of a new method of probabilistic scenario generation
This method incorporates the decomposition of scenarios by long-term and short-term uncertainties in fertility, mortality and migration dynamics. As a modification of this method, a new algorithm for generating mortality scenarios based on the recent publication in Science by Oeppen and Vaupel was developed (together with Warren Sanderson). This algorithm was presented at the International Symposium on Forecasting (Sydney, July 2004) and applied to several countries. Also in 2004 the research to develop efficient methods that are capable of generating scenarios with a given autocorrelation function began.
Developing of a system for probabilistic population projections
Three methods of probabilistic projections were implemented in the system: random line scenarios (LSS1 approach), autocorrelation in time (LSS2), and an approach that combines long term and short term uncertainties. The projection models were separated from scenarios and baseline data. The system is being now further developed. Probabilistic projections for Germany, Japan and the USA were performed with the help of new system.
Projection of minority populations
In collaboration with David Coleman (Oxford University) projections of ethnic minority populations in some European countries are carried out and social/political consequences of future trends investigated. This work will be presented at the PAA in April 2005.
Forecasting future size of religions
Anne Goujon, Vegard Skirbekk, Katrin Fliegenschnee, Pawel Strzelecki
Following a project by Fliegenshnee et al. (2004) on projections of the Protestant population in Austria, we have extended the methodology to all religious categories in the country with paying particular attention to the Roman Catholics, the Muslim and the population without religion, all experiencing some rapid changes in the last few years. We have looked at how fertility by religious category, marriages, migration and changing religious status within a lifetime may affect the religious composition of the population. This application has resulted into a VID Working Paper (01/2006). This study will be the starting point of a larger comparative project on multi-religion population projections, possibly starting with Canada and Switzerland, applying a similar methodology as the one developed for Austria..