Modeling income distributions and Lorenz curves
Modeling income distributions and Lorenz curves
复制标题
收入分配和洛伦兹曲线建模
DOI：
10.1007/s10888-010-9132-5
复制
发表时间：
2010
期刊：
影响因子：
--
通讯作者：
D. Slottje
中科院分区：
文献类型：
--
作者：
D. Slottje
她在纪念才华横溢、兼收并蓄的学者卡米洛·达古姆方面做得很好。其次，她在一个地方汇编了我在建模收入分配方面遇到的最好的工作，使用假设的收入分配函数（以下简称“HIDF”）来近似和描述经验收入分配数据。我不打算为每一篇论文写几段，而不是花费读者的时间来叙述显而易见的事情。Chotikapanich教授在她的介绍中做了出色的工作，以实质性的方式描述了每篇论文的内容。此外，摘要是经过深思熟虑的，清楚地传达了每章的内容，所以我也建议感兴趣的读者阅读这些摘要。我要做的就是抓住重点。这是一本优秀的书，我强烈推荐给任何对这一领域感兴趣的人，并敦促你阅读它。这篇评论的其余部分将试图告诉你为什么你应该阅读它。不平等专家总是有很多选择，他们如何量化和报告工资或收入的经验分配中固有的不平等程度。这个问题通常可以归结为使用不平等的汇总度量（如基尼系数）来描述分布中固有的不平等程度，或者使用HIDF（如Dagum模型）来近似和描述整个数据批次;或者两者兼而有之，即使用HIDF并估计其参数，然后估计洛伦兹曲线和随之而来的基尼系数等，HDIF方法的一个显著优点是，您可以根据需要估计尽可能多的参数，然后将这些参数视为“超参数”，并查看其他协变量
she has done a fine job in honoring the memory of the supremely talented and eclectic scholar, Camilo Dagum. Secondly, she has compiled in one place the best collection of work I have encountered on modeling income distributions, using hypothetical income distribution functions (henceforth,“HIDF”) to approximate and describe empirical income distribution data. Rather than expending the reader’s time by recounting the obvious, I am not going to write a few paragraphs about each paper. Professor Chotikapanich has done an excellent job in her introduction to the volume of describing each paper’s content in a substantive way. In addition, the abstracts are well thought out and clearly convey what each chapter is about, so I refer the interested reader to those as well. What I will do is get to the punch line. This is an excellent book and I recommend it strongly to anyone interested in this field and urge you to read it. The rest of this review will attempt to tell you why you should read it. Inequality specialists have always had a lot of choices on how they quantify and report the level of inequality inherent in an empirical distribution of wages or income. The issue frequently boils down to using a summary measure of inequality (like a Gini) to describe the level of inequality inherent in a distribution or using an HIDF (like the Dagum model) to approximate and describe the entire data batch; or doing both, namely using an HIDF and estimating its parameters to then estimate the Lorenz Curve and attendant Gini, etc., from it. One significant advantage of the HDIF approach is that you can estimate up to as many parameters as you wish, and then treat these parameters like “hyper-parameters” and see how other covariates