Understanding Changing Inequality within Families in the United States, 1960-2015

stephanie MollerJoseph whitmeyer北卡罗来纳大学夏洛特分校美国家庭收入不平等自20世纪下半叶以来有所增加，反映了长期收入不平等下降的逆转。在解释这种上升趋势时，许多研究人员把注意力集中在中产阶级的萎缩和收入两极分化上。然而，这一解释与双亲家庭最为相关。收入不平等的变化在其他家庭类型中表现得不同，这一模式在现有文献中经常被忽视。因此，研究人员还不知道形成不平等总体模式的因素是否适用于按种族和家庭结构划分的家庭子部分的收入不平等变化。这是文学上的一个重要空白，因为美国是一个多元化的社会。拟议中的研究将评估自20世纪60年代以来经济、人口和政治变化对不同家庭家庭收入不平等的影响程度。这项研究将通过开发两个全面的历史数据集，为数据基础设施做出贡献：美国县比较数据集，其中包括1960-2010年的十年数据；美国州比较数据集，其中包括1960-2015年的年度数据。这些数据将与人口普查局（Census Bureau）和劳工统计局（Bureau of Labor Statistics）的家庭数据相结合，以检验不平等总体变化的预测因子在多大程度上解释了群体内部的不平等。这些数据集将分发给其他研究人员，研究结果将通过研究文章和政策简报分发。该项目为研究生研究助理提供了宝贵的培训，并为其他研究生提供了利用这些数据进行论文和论文研究的机会。调查人员将根据种族和家庭结构对不同家庭群体的收入不平等进行调查。拟议的研究将包括两套分析。首先，利用十年一次的人口普查数据（1960-2010），调查人员将有孩子的家庭分配到家庭收入分配的五分之一。然后，他们将创建县级指标，衡量每种家庭类型在收入分配的每五分之一中所占的百分比。调查人员还将创建一个对每个群体内部不平等的总结衡量标准。分层重复测量模型将允许研究者评估有助于解释总收入不平等随时间变化的理论在多大程度上适用于家庭的子部分。在第二组分析中，研究者将分析当前人口调查（1967年至2015年）的家庭收入数据。家庭将按种族和家庭结构分类。分层分位数回归将评估经济和劳动力市场、社会人口和政治变量在收入分配中不同位置解释家庭收入的程度。这些分析将澄清在前一步中确定的宏观层面的关联是否仅仅反映了不同的家庭？暴露于宏观变化（例如与不稳定工作增加相关的半年制就业）还是总体模式预测家庭？独立于个人层面人口统计、就业和教育的收入分配位置。

SES-1627479Stephanie MollerJoseph WhitmeyerUniversity of North Carolina at CharlotteFamily income inequality in the United States has increased since the latter half of the twentieth century, reflecting a reversal of a long period of declining income inequality. In explaining this upswing, many researchers have focused on a shrinking middle class and polarization of income. Yet, this explanation is most relevant for White, two parent families. Changes in income inequality have unfolded differently for other family types, a pattern that is often overlooked in the extant literature. As a result, researchers do not yet know whether factors that shape aggregate patterns of inequality apply to changes in income inequality for sub-segments of families by race and family structure. This is an important gap in the literature as the United States is a diverse society. The proposed research will assess the extent that economic, demographic, and political shifts since the 1960s affect household income inequality among diverse families. This research will contribute to data infrastructure by developing two comprehensive, historical datasets: The Comparative U.S. Counties Dataset which will include decennial data 1960-2010 and the Comparative U.S. States Dataset which will include annual data 1960-2015. These data will be combined with household data from the Census Bureau and the Bureau of Labor Statistics to examine the extent that predictors of aggregate changes in inequality explain within-group inequality. The datasets will be distributed to other researchers, and the findings will be distributed through research articles and policy briefs. This project provides invaluable training for a graduate research assistant and opportunities for other graduate students to utilize these data for thesis and dissertation research. The investigators will examine income inequality among subgroups of families, dividing them by race and family structure. The proposed study will include two sets of analyses. First, utilizing decennial census data (1960-2010), the investigators will assign families with children to quintiles of the family income distribution. They will then create county-level measures of the percentage of each family type in each quintile of the income distribution. The investigators will also create a summary measure of inequality within each group. A hierarchical repeated measures model will allow the investigators to assess the extent that theories that help explain changes in aggregate income inequality over time apply to sub-segments of families. In the second set of analyses, the investigators will analyze family income data from the Current Population Survey (1967 to 2015). Families will be disaggregated by race and family structure. Hierarchical quantile regression will assess the extent that economic and labor market, sociodemographic, and political variables explain family income at different locations in the income distribution. These analyses will clarify whether macro level associations identified in the previous step simply reflect different families? exposure to macro shifts (such as part-year employment associated with the rise in precarious work) or if the aggregate patterns predict families? locations in income distributions independent of individual-level demographics, employment and education.