Studying time use variations in 18 countries applying a life course perspective

To gain insight in variations in life courses during last decades, and the factors underlying these variations, time use data seem suited. By means of analyzing time use data insight is gained in the (relative) importance of various life spheres as paid work, household work, volunteer aid, care, anc education in and over people's life. The relevance of an integrated insight in the relation between paid work and these other life spheres seems to have grown with the introduction and (policy) application of the idea of "transitional labour markets". Time use variations during (individual) life courses in 18 countries are analysed by means of Hierarchical Age-Period-Cohortmodelling (HAPC). By means of this method the classical APC-riddle, i.e. the fact that the APC model is underidentified due to a linear dependency among age, period, and cohort, can be tackled. This paper compares the fixed versus the random-effects model specifications for APCanalysis. The random-effects HAPC-model appears the most appropriate specification. The analyses find evidence in support of quadratic age effects on time use. Furthermore, the analyses find significant cohort and period effects. Finally, the period effects as well as the welfare state effects indicate a non-negligible sensitivity for economic circumstances and welfare policies.