Using data from the British Household Panel Survey and Understanding Society, we examine the saving behaviour of individuals over time. Initially, we explore the determinants of the saving behaviour of children aged 11–15. Our findings suggest that parental allowances/pocket money (earnings from part-time work) lower (increase) the probability that a child saves. There is also evidence that the financial expectations of the head of household have an influence on their offspring’s saving behaviour, where children of optimistic parents have a lower probability of saving by approximately 2 percentage points. However, there is no evidence of an intergenerational correlation in savings behaviour: the saving behaviour of parents appears to have no bearing on the saving decisions of their offspring. We then go on to explore the implications of the saving behaviour of children for their savings decisions in later life, specifically when observed in early adulthood. We find that having saved as a child has a large positive influence both on the probability of saving on a monthly basis and on the amount saved as an adult. This finding is robust to alternative empirical strategies including IV analysis where the most conservative estimates show that having saved as a child increases the probability of saving during adulthood by 12 percentage points.

We investigate the existence and persistence of financial hardship at the household level using data from the British Household Panel Survey. Our modelling strategy makes three important contributions to the existing literature on household finances. Firstly, we model nine different types of household financial problems within a joint framework, allowing for correlation in the random effects across the nine equations. Secondly, we develop a dynamic framework in order to model the persistence of financial problems over time by extending our multi-equation framework to allow the presence or otherwise of different types of financial problems in the previous time period to influence the probability that the household currently experiences such problems. Our third contribution relates to the possibility that experiencing financial problems may be correlated with sample attrition. We model missing observations in the panel in order to allow for such attrition. Our modelling framework allows us to identify any persistence in financial problems over time as well as any interdependence that may exist between different types of financial problems. Our findings reveal interesting variations in the determinants of experiencing different types of financial problems including demographic and regional differences. Our findings also highlight persistence in experiencing financial problems over time as well as the role that saving on a regular basis can play in mitigating current financial problems.