Three essays on heterogeneous capabilities, poverty trap thresholds, and the persistence of inequality

In the presentation of three essays, this study investigates how individuals and households that are endowed with heterogeneous capabilities convert productive assets into economic well-being through the lenses of ex-post empirical analysis and an ex-ante macro-micro simulation.;The opening essay traces the common origins, divergent evolution, and reconcilable differences across asset-based and Human Development/Capabilities perspectives of poverty. It is argued that asset-based studies have embedded in them a strong temptation to focus solely on asset accumulation policies without giving the conversion process of assets into livelihood its due study. Although the asset-based literature has made advances on the theoretical front in explaining how poverty trap thresholds are unique and dependent on intrinsic ability, the empirical analysis of what intrinsic ability may encompass remains understudied.;The essay proposes that empirical asset-based studies of poverty trap thresholds stand to benefit from insights of the Human Development/Capabilities literature by viewing intrinsic ability as capability constraints which leads to differing opportunity costs. To illustrate the bridging role of opportunity costs, a simple, two-household model with heterogeneous opportunity costs is presented and applied to South Africa's most populated province.;The second essay extends the first by asking how particular asset holdings are associated with capabilities to take on new economic opportunities. Knowledge of the patterns and linkages between capabilities and particular asset holdings has been relatively under- realized, particularly in the empirical poverty traps literature. Using the same household survey data of essay one, this essay seeks to empirically decompose how early period asset endowments impact future levels of well-being into direct and indirect mechanisms. A direct impact of asset endowments on future well-being would include consumption of the asset or the direct use of it to produce incomes (i.e. grain stock consumption or the sale of livestock offspring, respectively). The indirect impacts of endowments are of greater interest to this study and are of two forms: asset-to-asset complementarities and how household capabilities (or deprivations thereof) interact with particular asset holdings. To achieve this decomposition, this paper employs a method of path analysis akin to early heritability of traits studies which were aimed at distinguishing between the effects of nature versus nurture.;This second essay contributes to the prior literature in three primary ways. First, it adds empirical robustness to prior theoretical work linking intrinsic ability with household-specific poverty trap thresholds. Second, it bridges the quantitative work on poverty traps with qualitative insights from the human development/capabilities literature by identifying which particular asset holdings are associated with different household capability constraints. Lastly, it serves as a reminder to policy that measuring poverty as asset stocks requires additional knowledge about the process of converting those assets into well-being.;After the first two essays tackle the issue of poverty measurement and its conversion into economic well-being, the final essay reviews a host of macro, micro, and macro-micro modeling strategies in order to draw out central features of a framework that can address the micro impacts of macro changes in the presence of heterogeneous behavioral responses. Additionally, this essay presents a preliminary framework of that model and explores how capabilities, that heterogeneously impact the occupational choice of individuals, might be incorporated.;When there are heterogeneous responses to changes in the macroeconomic employment situation, the identification of winners and losers of potential macro-policy changes in an ex-ante fashion is more complex. Standard computable general equilibrium (CGE) models are only able to identify between group changes in income distributions and not within group changes. Additionally, there is a lack of capacity to include unique behavioral responses. One alternative is to link a behavioral microsimulation model (MSM) to a CGE. The benefit of this approach is that the outcomes of behavior are aggregated rather than behavior itself being homogenized and aggregated as is implicitly done in models with representative agents or household groups. (Abstract shortened by UMI.)...