In light of the evidence on the origins and consequences of early inequalities in the UK gathered throughout the chapter, we conclude in Section 8 by highlighting key trends in the early years policy landscape since the turn of the century and sharing our reflections on where policy should go next. [...] 4 The estimates and standard errors of the coefficients of the multivariate regression models for cognitive development and socio-emotional and behavioural difficulties are reported in the first column of Appendix Tables 1 and 2, respectively. [...] We also see the model controls explain far more of the variation in socio- emotional development than in cognition: the R-squared, which measures the fraction of variation in the measure of development that is explained by variation in the correlates, is 0.45 for socio-emotional development and 0.22 for cognitive development. [...] In order to assess the extent to which the raw gaps in early development depicted in Figures 1 and 2 are explained by systematic differences in children’s environments, we report in Appendix Table 3 the raw association between each of the variables included in the model in the estimation sample, next to the adjusted association (i.e. [...] By contrast, looking at socio-emotional development in the lower part of the figure, the most dominant factor to emerge is the child’s emotional environment, which explains 35% of the variation in child socio-emotional and behavioural difficulties, or over three-quarters of the overall variation explained by the model.