Following Mare’s (2011) presidential address to the Population Association of America, much research has examined the dynamics of intergenerational mobility and socioeconomic inequality. However, this research is starkly divided on the existence of multigenerational effects and pathways. In this article, we examine educational attainment across three generations using the Panel Study of Income Dynamics. We test three key hypothesized mechanisms of grandparent educational advantage transmission that have been documented in international samples: compensation and augmentation, grandparent lifespan, and genetic transmission. To further interrogate how these mechanisms function across social contexts, we also race-stratify all analyses. As a more stringent test, we also test all hypotheses using cousin fixed effects models. We find no evidence in favor of the compensation hypothesis, but do find descriptive support for the augmentation hypothesis, as well as some limited support among whites for college graduation outcomes in cousin fixed effects models. Findings on the lifespan overlap and genetic transmission hypotheses are highly race-stratified. Overall, our findings support the importance of investigating the mechanisms of three-generation processes of educational attainment across contexts and sub-groups.

继马雷（Mare，2011年）向美国人口协会（APA）致主席讲话之后，许多研究都研究了代际流动和社会经济不平等的动态。但是，这项研究在多代效应和途径的存在上存在明显分歧。在本文中，我们使用收入动态面板研究考察了三代人的教育水平。我们测试了国际样本中已证明的三个主要假设的祖父母教育优势传递机制：补偿和增加，祖父母寿命和遗传传递。为了进一步审视这些机制如何在社会环境中发挥作用，我们还对所有分析进行了种族分层。作为更严格的检验，我们还使用表亲固定效应模型检验所有假设。我们找不到支持补偿假说的证据，但确实找到了对增加假说的描述性支持，以及表亲固定效应模型中白人对大学毕业结果的有限支持。关于寿命重叠和遗传传播假说的发现在种族上高度分层。总体而言，我们的发现支持研究跨情境和亚群体的三代教育素养过程机制的重要性。