First patterns were described by (Kiernan, 1986) comparing the age of moving out and the form
of residence of young people in six European countries. She finds that young people in Denmark
are the first ones to leave, followed by West Germany, France, the Netherlands, Ireland and the
UK. Woman leaving earlier than men appears across all of these six. Countries with early-leaver
the share of young people living without a partner or with a partner and unmarried is significant
higher.
Several papers proof a systematic difference between geographical regions in Europe. (Holdsworth, 2000)
compares Britain and Spain and indicates the patterns (Billari et al., 2001) finds looking at cohorts
born around 1960 in whole Europe. He states an increase of the leaving age north to south.
While the eastern European countries match the tendency of the southern European countries. The
median for leaving parental home in Nordic countries is around 20 for southern and eastern European
countries around 27. Germany ranks in this data just behind the Nordic countries and on the same
stage as France with a median around 22. Across all countries he finds that women leave earlier than men. The reason young people leave home also differs systematically. While in southern countries children leave their parent’s home in a high share for forming a union with their partner and not before finishing education. In the Nordic countries and central European it is the other way around.
Assave et al. show the same systematic difference by comparing EHPC data. They try to lead this pattern back to the difference in culture but also on the difference in welfare policies. Assave  et al 2002 studie US data and find the same reasons for leaving as in southern european countries. Therefore they design a model, similar to the one of job search, to determine the decision of moving out as the successful match on the marriage market.
Laferre and blanc 2004 look at the rental assistance reform in France in 1992. The reform included an extension of the assistance from only familys with children to all low income housholds, i.e. students. They find that there is a modest rise in the share of students forming new household compared to the total number of new formed households with rised by 3% through the reform.

Theoretical Framework

This chapter deals with the theoretical basis of our analysis,  a simple Roy model setup. Introduced by Roy in 1951 the conceptual framework of the generalized Roy model provides the opportunity to explain self selection behavior by heterogeneity in the characteristics of the agents . We will use this approach to illustrate the decision process of individuals to leave their parent's household during tertiary education.  Note that we will follow the notation in Heckman 2001. \cite{Heckman_2007}