<![CDATA[Fragile and conflict affected states (FCAS) are those in which the government lacks the political will and/or capacity to provide the basic functions necessary for poverty reduction, economic development, and the security of human rights of their populations.Until recent history, unfortunately, the majority of research conducted and universal health care debates have been centered around middle income and emerging economies. As a result, FCAS have been neglected from many global discussions and decisions. Due to this neglect, many FCAS do not have proper vaccinations and antibiotics. Seemingly, well estimated health care costs are a necessary stepping stone in improving the health of citizens among FCAS. Fortunately, developments in statistical learning theory combined with data obtained by the WBG and Transparency International make it possible to accurately model health care cost among FCAS. The data used in this paper consisted of 35 countries and 89 variables. Of these 89 variables, health care expenditure (HCE) was the only response variable. With 88 predictor variables, there was expected to be multicollinearity, which occurs when multiple variables share relatively large absolute correlation. Since multicollinearity is expected and the number of variables is far greater than the number of observations, this paper adopts Zou and Hastie’s method of regularization via elastic net (ENET). In order to accurately estimate the maximum and expected maximum HCE among FCAS, well-known risk measures, such as Value at Risk and Conditional Value at Risk, and related quantities were obtained via Monte Carlo simulations. This paper obtained risk measures at 95 security level.]]>
Copyrights © 2018
