|dc.description.abstract||This thesis comprises three essays which examine how skills gaps form, persist and can be closed in young Latin American children. Skills are the foundation of an individual’s human capital, and since the early 2000s have been broadly deﬁned as cognitive (things that are taught) or non-cognitive (things that are learned through experience). The theoretical underpinning of the development of such skills is the production function for skills introduced by Todd and Wolpin (2003, 2007). The setting of Latin America is a relevant context to undertake this research, as it is one of the most unequal regions in the world. The ﬁrst two studies make use of longitudinal Young Lives data from Peru, while the last uses government data from Honduras.
In the ﬁrst paper, the analysis exploits the longitudinal aspect of Young Lives to examine when nutritional investment is most productive in the ﬁrst ﬁve years of life. The study looks to improve on the existing literature by using two new instrumental variable methods: weak identiﬁcation robust conﬁdence intervals,and instruments generated from assumptions about heteroskedasticity. The second paper uses siblings data to estimate eﬀects of parental and peer relationships on early adolescent children’s self-esteem and pride. It then decomposes socioeconomic and locality gaps to determine how policy can potentially equalise outcomes in each group. The last paper evaluates the impact of a randomised cash transfer in Honduras on early childhood development outcomes. Speciﬁcally, the study examines if a transfer conditional on school attendance promotes intra-household spillovers from the beneﬁciary siblings to younger siblings under the age of 5. The estimates show modest improvements in early childhood development for these younger siblings. The body of evidence suggests that these improvements are real spillover eﬀects and not just a result of an income eﬀect for beneﬁciary households. Overall, the thesis provides evidence which supports the theoretical model on the importance of early intervention, while noting the necessity of using sound methodologies to deal with empirical issues in estimating such eﬀects.||en_US