此三年期的計劃為隨機優越法則的新應用。 第一年計劃為一個實證研究。個人將利用投資人對投資組合之選擇來估計幾乎隨機優越法則中的風險偏好參數。近年來，財務的文獻指出幾乎隨機優越法則可以解釋財務上的一些謎題並可以評估各項風險性資產的績效，然而這些實證發現非常仰賴於Levy et al. (2010)對於幾乎隨機優越法則中風險偏好參數的估計結果。不幸的是，Levy et al. (2010)的結果是根據學生對一些不具真實性的實驗設計選擇而來，同時該篇論文中所使用的幾乎二階隨機優越的條件是錯誤的。本篇論文是文獻上第一篇使用真實的投資選擇資料，並使用正確的幾乎隨機優越的條件來估計該法則中的風險偏好參數。 第二年計劃個人將建構一個全新的法則，以確切衡量社會經濟的健康不均度。近年來有多篇論文嘗試建構不均度的衡量法則，以找到所有趨避建康不均度的決策者均同意健康分配不均度上升的充分必要條件。然而這些條件太過嚴苛，很容易就找到一些例子，是大多數的決策者都同意不均度上升，但這些條件確無法指出不均度上升。本計劃將改善此一缺點，提出新的健康不均度法則，以適用於絕大部分的決策者。本計劃亦將利用開發中國家的健康分配資料來進行實證研究。 第三年計劃則將建構新的所得不均度連續法則，這些法則可以連結所得不均度趨避的法則以及下方所得不均度趨避的法則。所得不均度趨避乃假設政策決策者並不喜歡所得分配向兩邊擴散，是一個與政策決策者對不均度偏好的二階微分相關的法則，當文獻往前走一步要求下方所得不均度趨避，這需要假設政策決策者所使用的不均度指標其斜率須為concave 函數，也是一個與政策決策者對不均度偏好的三階微分相關的法則。然而這個三階微分的假設在實務上太過強烈，有許多政策決策者並非對於所得水準都符合這個三階微分的假設。因此本計劃將建構新的評估所得分配不均度的法則，此法則界於二階與三階之間。計劃也將利用加拿大的所得分配資料來進行實證分析。 ;This is a three-year project. All of them are related to stochastic dominance. For the first year project, I will use empirical data to estimate the preference parameters in almost stochastic dominance (ASD) established by Leshno and Levy (2002). Recently, the finance literature has shown that ASD is helpful in explaining some puzzles in finance. However, their conclusions heavily rely on the estimation of the preference parameters in the ASD rules provided by Levy et al. (2010) which is not only obtained from artificial tasks designed in laboratory but also adopt the incorrect condition in ASD. Our project is the first one in the literature using empirical observations in portfolio choice decisions and adopting the correct conditions to estimate the preference parameters in ASD. Our findings can help understanding investors’ risk preference as well as reexamining the current empirical findings in finance. For the second year project, I will establish the conditions to identify robust ordering of socioeconomic health inequality for most policymakers. The measurement of health inequality is an important issue in economics, public health and epidemiology. Recently, the literature has provided the conditions to identify robust ordering of health inequality, but the conditions are very rigid. It is because that both papers consider all weight functions on health scores for different groups of socioeconomic status which represent the judgments of inequality aversion. These weight functions allow zero weights. However, in reality, most policymakers will not place zero weight on any specific group. Thus, this project will seek for the robust ordering of socioeconomic health inequality by excluding extreme weight functions. Our rule can significantly improve the applicability of the health inequality measurement. For the third year project, I will provide continua of the inequality relations between income inequality aversion and downside inequality aversion. It is almost universally assumed that a mean-preserving spread in income distribution is less preferred to all policymakers, which implies convex inequality indices in the inequality analysis. The recent literature employs additional assumption: policymakers are downside inequality averse, which means that the slope of the inequality indices is concave globally. However, the assumption of downside inequality aversion is very strong. In this project, I relax this strong assumption on the slope of the inequality indices and seek for an unambiguous prediction. The new notion of income inequality measurement derived in this paper can help to distinguish the ranking of income distributions.
