The Centre for the Economics of Education was asked to bring together a wide range of academic evidence (primarily England-based) to investigate the extent to which academic and non-academic childhood outcomes are complementary to each other, or are in some way traded-off against each other. The report also investigates the drivers of both academic and non-academic outcomes and the extent to which child outcomes persist throughout a child's life and across generations. There is also a brief discussion of the implications of this evidence to education policy. The report finds that the relationships between academic and non-academic outcomes are complex in nature. For example, pupils who are bullied or who take unauthorised absence at age 14 have significantly lower educational achievement at GCSE. Pupils who experienced bullying at age 14 were also much more likely to experience bullying at age 16. Conversely pupils who participate in positive extra-curricular activities, such as clubs, were also found to have better academic achievement later in their schooling. These childhood outcomes are themselves determined by a wide variety of influences (such as the quality of parenting they receive) and environmental factors (for example whether they are exposed to passive smoke). It has been well established that children from disadvantaged backgrounds have relatively poor academic outcomes and tend to have weaker social skills than children from more advantaged households. However the evidence also suggests that these children also go on to experience more negative outcomes in adulthood, such as lower probability of employment and lower wages. Furthermore key social and academic outcomes of parents--cognitive skills, attitudes to education, smoking and drinking - are related to similar behaviours in their children. The report concludes that the complex nature of the drivers of child development, the interdependence of child outcomes, and the way that outcomes persist through an individual's life and across generations needs to be recognised in order to develop truly effective policy.While very little of the evidence highlighted in this report identifies true causal relationships (i.e. that a factor X actually directly causes a change in outcome Y), the report draws on some of the highest quality research and analysis currently available, using detailed longitudinal datasets, including the Department's own Longitudinal Study of Young People in England. This enables us to identify at the very least "robust associations" as well as the data allows. However it does suggest that further research is required to better understand the associations outlined in this report to move to a position where we can identify credible causal relationships. This is important to foster more justified and increasingly effective policymaking. (Contains 85 footnotes.)