The increase in the number of young adults living with their parents in the last 20 years has attracted considerable attention in the press and literature at a global scale. In the UK, for example, the number of young adults between the ages of 20 and 34 living with their parents is said to have risen by 37.5% over two decades according to one recent source, but the extent of the rise and the underlying mechanisms are not yet well understood.

This dissertation aims to better understand the decisions to remain/leave parental homes and the subsequent choices, in the case of leaving, to move to new residential destinations by linking them together. The focus is motivated by a long-lasting separation of demographic and sociological investigations on leaving parental homes on the one hand, and the spatial economic studies on locational choices and movements to new residential destinations on the other. This separation has resulted in a gap in knowledge and quantification that we aim to fill. We hypothesise that these two choice decisions are interdependent, and the quality of quantification will benefit from them being modelled jointly.

The research has produced a new statistical modelling approach that makes sense and quantifies the trade-offs by jointly considering the young adults’ own circumstances, their parents’ circumstances, the locational attributes where their parents live, and the attributes of the destinations within the choice sets of their moves. To the best of our knowledge, this is the first theoretical and empirical model of its kind in the literature.

To develop and test this model, we build on the UK Household Longitudinal Study (UKHLS) by supplementing with related data to enable robust estimation that relates the decision outcomes with young adults’ own personal attributes, their families’ attributes, and locational attributes of their parents’ residential areas. For those who decided to move, we also consider the locational attributes of the new places they choose to live. Our data analysis has discovered serious panel attrition biases in the UKHLS which we address through panel gap filling and inverse probability weighting. We then develop separately single-tier models for decisions to remain/leave home and for choices of new residential locations in line with the existing literature. Subsequently, we build a joint model for those choices to allow for a clear comparison. Non-graduates and graduates are modelled separately.

The model results have uncovered three new findings. Firstly, the jointly estimated model provides unambiguous evidence that single-tier models without this link between them lead to far poorer model fit and less robust quantifications of the effects. Secondly, the models quantify, for the first time, the extent to which affordable housing and accessibility to jobs affect young adults’ propensity to leave the parental home and choose their own residential locations. Contrary to conventional wisdom, both non-graduates and graduates are sensitive to both housing costs and job access, and their differences are relatively small. However, there are strong indications of restricted social mobility over the period: graduates whose at least one parent is educated to a degree level are more likely to move to districts with high accessibility to jobs. Thirdly, it is important to separately estimate models for non-graduates and graduates due to the significant differences uncovered by our analyses and modelling regarding the young adults’ personal attributes and the attributes of their parents. These findings are set to revolutionise how future spatial equilibrium models represent young adults’ decisions and choices.

The research conducted in this study has not only contributed to the current understanding of the topic but has also set a new research agenda for the future. One of the most significant areas for further exploration is the utilization of the quantifications presented in this dissertation within the framework of integrated spatial equilibrium modelling. As we move forward, it would be beneficial to integrate the newly developed model into such frameworks, especially in research schemes aimed at testing policies that support young adults in socioeconomically disadvantaged regions. This integration will provide valuable insights into the effectiveness of policy interventions and their impact on the decision-making processes of young adults, ultimately contributing to the development of evidence-based strategies to address their unique challenges and needs.