Chapter 9 Inequality of Opportunity in the United States: Trends and Decomposition
Chapter 9 Inequality of Opportunity in the United States: Trends and Decomposition
复制标题
第九章美国机会不平等：趋势与分解
DOI：
10.1108/s1049-2585(2011)0000019012
复制
发表时间：
2011
期刊：
影响因子：
--
通讯作者：
Juan Gabriel Rodríguez
中科院分区：
文献类型：
--
作者：
G. A. Marrero;Juan Gabriel Rodríguez
目的-我们的最终目标是描述三个方法论问题。首先，比较各种估计方法对长时间序列的相对性能;其次，估计努力与环境的相关程度;最后，根据理想树将总不平等分解为机会不平等和努力不平等。
方法-我们估计参数和nonparametrically在1969年和2007年之间的机会在美国的事前不平等。根据Bjorklund et al.（2011）的提议，计算了努力与环境之间的相关程度。此外，我们还利用变异系数平方的自然分解和Shapley值的嵌套，将完全不等式分解到一个三层分解的理想树上.
结果-我们发现非参数和参数方法之间的显着差异。特别是，我们的研究结果表明，考虑情况之间的交叉影响可能是相关的。此外，在美国，1969年至2007年期间，努力与环境之间的相关程度显著增加，占总IO的5%至20%。另外，在20世纪70年代和80年代，种族是主要的环境因素，占直接信息输出的50%以上，而在最近20年，父母教育则占主导地位。
独创性-我们修改的参数规格，考虑情况之间的交叉影响。我们估计的程度之间的相关性的努力和情况的长时间序列。我们根据一个三层的层次模型分解总不等式。
Purpose – Our ultimate goal is to characterize three methodological issues. First, compare the relative performance of alternative estimation methods for long time series, second, estimate the degree of correlation between effort and circumstances, and, third, decompose total inequality into inequality of opportunity and inequality of effort according to an ideal tree.
Methodology – We estimate parametrically and nonparametrically the ex-ante inequality of opportunity in the United States between 1969 and 2007. The degree of correlation between effort and circumstances is computed following the proposal in Bjorklund et al. (2011). In addition, we decompose total inequality based on an ideal tree with three levels of disaggregation by applying the natural decomposition of the squared coefficient of variation and the Nested Shapley value.
Findings – We find significant differences between the nonparametric and parametric approaches. In particular, our results reveal that considering cross-effects between circumstances may be relevant. Moreover, the degree of correlation between effort and circumstances which has significantly increased over the period 1969 and 2007 in the United States, explains between 5% and 20% of total IO. In addition, race is the main circumstance during the 1970s and 1980s, accounting for more than 50% of the direct IO, while parental education take the lead in the last two decades.
Originality – We modify the parametric specification by considering cross-effects between circumstances. We estimate the degree of correlation between effort and circumstances for long time series. We decompose total inequality according to a three-level hierarchical model.