Simple solutions to hard problems in the estimation and prediction of welfare distributions

The aim of my PhD projects is to investigate various problems in applied parametric and nonparametric estimation, and eventually in prediction. The main focus is on welfare analysis. For instance, in both nonparametric density estimation and regression, the so-called boundary effects, i.e. the bias and variance increase due to one-sided data information, can be quite serious. Therefore, a new, straightforward and very simple boundary correction is proposed in the first essay, applying local bandwidth variation at the boundaries. In a comprehensive simulation study this method is shown to perform very well. Furthermore, it is an excellent method for estimating the world income distribution, and Engel curves in economics. In the second and third essays, an integration-based procedure is introduced for estimating and predicting of population distributions. This is done by data matching with applications to the economics of wealth and health. In conclusion, both essays 2 and 3, it is evident that the proposed method can be applied to estimate both (unconditional) discrete and continuous distributions. It is applicable irrespective of the mean regression or model, and can be easily extended to other contexts, such as small area statistics, nonparametric ...