This thesis presents five standalone essays that demonstrate the feasibility and utility of employing advanced analytic techniques to cross-sectional data from Indonesia in order to deal with some technical challenges typically encountered either in the estimation of social gradient in health or in the monitoring and evaluation of well-being as a multidimensional construct. The first essay estimates the causal effect of poverty on mental health by exploiting a natural experiment induced by weather variability across 440 districts in the Indonesian archipelago. The second essay applies parametric anchoring vignette methodology to investigate the extent to which the estimates of demographic and socio-economic inequalities in self-rated health are biased by survey respondents' differential reporting behaviour. The third essay formally assesses the existence and identifies the social determinants of the double burden of malnutrition in Indonesia using a variant of a generalised linear mixed model. The fourth essay maps the social and spatial distributions of malaria in 27 districts in Indonesian Papua using a probabilistic disease mapping technique that is capable of accounting for the complex dependency structure of spatially-correlated multilevel data. The fifth essay examines the extent and patterns of multidimensional poverty in Indonesia over the last decade using a novel poverty measurement method that is sensitive to both the incidence and intensity of multiple deprivations in income, health and education domains. Together, these essays show that although health and social researchers in the developing world have little choice but to conduct cross-sectional studies, new insights can sometimes be gained if one is willing to look at existing data through a new lens. In all five cases presented here, this approach is proved to be useful in shaping practical policy-making.