Norway and other industrialized countries have experienced massive changes in family behaviour over the last few decades: the age at marriage has increased, a larger proportion has remained unmarried, informal cohabitation has become more common, and rates of union disruption have escalated. At the same time, women got fewer children, and those who became mothers did so later in life. These changes are, of course, closely linked to each other. For example, people who do not live in a stable relationship usually do not want children, and conversely, the childless may see little need to formalize a consensual union and they may find it easier to dissolve a relationship.

The intention of the project was to learn more about the reasons for and consequences of these changes in family changes, with special emphasis on the Norwegian setting. The drift away from the formal marriage has been particularly pronounced in Norway and other Nordic countries. Yet, fertility is relatively high. This situation makes Norway very interesting from an international perspective. In addition, the country has quite unique register data that allow detailed exploration of demographic behaviour. However, it would obviously also be important to contrast the development in Norway with that seen in other industrialized countries.

Another goal of the project was to develop better tools for analysing and forecasting future family patterns.

More work was done on fertility, and less on other family changes, than originally intended. This was because the register data on cohabitation have more limitations than was assumed when the proposal to CAS was written, and because an attempt to get funding from the Norwegian Research Council for a special survey among cohabitants failed (in spite of extremely positive reviewer reports). In addition, the interests among some of the project participants had drifted more towards fertility.

Most of the work was based on Norwegian register data. Participants from other countries were very enthusiastic about the potentials of those data and wanted to use them rather than data from their own country (or other countries). Thus, less comparative work was done than originally planned, but we have learned more about Norway, and when the publications from the project appear, several researchers in this field may become aware of the potentials of the data, which may trigger further initiatives to analyse Norwegian family behaviour.

The methodological part of the project proceeded as planned. A method for stochastic family forecasts was developed, and as part of that, methodological papers on marriage markets and the two-sex problem were also written.

The researchers at CAS have concrete plans about further co-operation. For example, Alho and Keilman will continue their work together on stochastic forecasting, and Kravdal will do more work with Rindfuss (and others in North Carolina) on fertility and with Emily Grundy on the link between mortality and fertility. In addition, co-operation with other British participants is very likely (Aassve, Sigle-Rushton, Steele), and Kravdal and H.-P. Kohler plan to use information (from the Norwegian Institute of Public Health) about di-zygotic twins to identify various effects in models for fertility and family behaviour.

To facilitate such co-operation and increase the chance of a successful completion of the projects started at CAS, all participants – as well as the discussants in the 2007 workshop - will be invited to a workshop in the spring 2008.