The complementarities of child health achievements in developing countries’. Georg- August-Universität Wirtschaftswissenschaftliche Fakultät, Goettingen. eDiss 3/2015

This dissertation, which is grounded on a human-centred and multidimensional vision of development, aims at analyzing empirically the correlates and determinants of child health in low and middle income countries. The first essay, co-authored with Stephan Klasen, applies bivariate cluster analysis to study the relationship of improvement in different health and education indicators of the Millennium Development Goals (MDGs). The MDGs represent key aspects of human well-being, and as such, they should be closely interlinked. Nevertheless, despite the substantial theoretical case for large synergies between MDG goals, empirically the heterogeneity of the linkage of progress across different MDGs can be quite substantial. The central research question in the first essay of this thesis is therefore whether one can identify clear clusters of countries where such synergies exist (and those where they do not exist), and what drives membership in these clusters. Our particular contribution is to define MDG progress in terms of relative performance indicators which describe the rate of change that exceeds the empirically ‘expected’ rate of change, given initial conditions. Such a measure allows us to control for the different level of ambition implied by the MDGs for countries with different initial conditions and essentially ask whether countries that succeeded in achieving extraordinary progress in MDG achievements were benefiting from synergies or not By applying cluster analysis to these indicators of relative performance on different health and education MDG indicators we are able to distinguish countries where the pair of indicators move in the same direction (‘good’ or ‘bad’ performers) from countries where the pair of indicators moved in opposite directions (‘partial’ performers). Our results suggest that there are synergies, but these only apply to a restricted group of countries (the ‘good’ performers where progress goes hand in hand, and the ‘bad performers’ where regress goes hand-in-hand). On the other hand, we find that in a sizable number of countries synergies are weak or absent, leading to contrary movements in MDG progress at least in one or two pairs of non-income MDG indicators. As our regression results show this partial performance is largely explained by rising inequality and low institutional quality. On the other hand, economic growth, which seems to be the most robust determinant of cluster membership, is particularly able to distinguish the good from bad performers. The second essay, in single authorship, analyzes the short and long term determinants of child health achievements, using longitudinal survey data from Indonesia. In particular, based on the analytical framework proposed by Mosley and Chen (1984), this essay investigates the effects of a set of child-specific, household and community characteristics on child nutritional status, measured by her height-for-age z-scores. The particular contribution of this study is the use of micro-level panel data and of a methodology (i.e. the Mundlak approach of household-average fixed effects) which not only delivers robust and unbiased estimates but also allows me to contribute to prior literature by identifying the short and long term effects of proximate and socio-economic determinants of child health. Results suggest that maternal education has a positive and long term effect on child health which is partly reflected in reproductive behaviour and partly conveyed through child caring practices (i.e. breastfeeding). On the other hand, while temporal increases in income allow for improved expenditure capacity and therefore to better nutritional status, there isn’t any long term effect of income, apart from the one that is conveyed through the acquisition and use of the intermediate inputs such as improved sanitation infrastructure which strongly and significantly affect child health status. Overall, the findings imply that income poverty-alleviating policies complemented with investment in basic health infrastructure might strongly contribute to improving children health conditions, Also, as the linkages between maternal education, child nurturing practices and child health are found to be particularly strong, policies which are aimed at enhancing the quality of education can remarkably improve the nutritional status of children. The third essay, on single authorship, examines the effect of nutritional status on subsequent educational achievements, using micro level panel data from Indonesia and applying a maternal fixed effect specification augmented by an instrumental variables estimator in order to control for possible correlation between some of the components of the error term and the main independent variable. Differences in nutritional status between siblings are identified by using exposure in the earliest months of life to the drought associated with the Indonesian wildfires of late 1997. Estimation results show that health capital (measured by height-for-age z-scores at childhood) significantly and positively affects the number of completed grades of schooling and the score on cognitive test. This imply that from a policy perspective, school and nutrition objectives should not be seen as competing goals but are closely interlinked. Therefore, financial resources devoted to child nutrition policies do not necessarily compete with those for the education agenda; instead, as implied in this essay, they can be regarded as a more cost effective way to raise present and future socio-economic development.