In this proposed pilot project, we aim to demonstrate that in the presence of intergenerationally-correlated endowments, the many studies that purport to estimate the effects of parental and grandparental characteristics on child outcomes are misinterpreted in ways that are likely to overstate substantially direct parental effects and indirect grandparental effects, and maybe even have the sign wrong for direct grandparental effects. By “intergenerationally-correlated endowments” we mean important factors that affect the outcomes of interest that are usually not observed in social science data, and that are transmitted from one generation to the next, such as family culture or genes. The basic purposes of this pilot project are to develop this point, summarize and reinterpret three-generational studies in the literature in light of this point, develop a model that will permit identification of grandparental effects with data on four generations, locate data across four generations, and then prepare a National Institute of Health (NIH) grant application to explore and reinterpret intergenerational relations in depth to understand better their causal nature.
In recent years, social scientists have increasingly become interested in the effects of grandparents on grandchildren’s demographic, educational and occupational outcomes, leading to the burgeoning literature on multigenerational effects. The growing interest in multigenerational effects reflects the aging trend and growing economic inequality under which grandparents’ roles for grandchildren’s demographic, educational, and economic outcomes may be increasingly relevant. In the proposed pilot project, we will summarize and critically reinterpret three-generational studies in the literature to demonstrate that in the presence of intergenerationally-correlated endowments, the many studies that purport to estimate the effects of parental and grandparental characteristics on child outcomes are misinterpreted in ways that are likely to overstate substantially direct parental effects and indirect grandparental effects, and maybe even have the sign wrong for direct grandparental effects. By “intergenerationally-correlated endowments” we mean important factors that affect the outcomes of interest that are usually not observed in social science data, and that are transmitted from one generation to the next, such as family culture or genes. To evaluate existing studies, we will apply the model of intergenerational relations with consideration of endowments, extending earlier work of Behrman. Based on our evaluation of the literature, we will develop a model that will permit identification of grandparental effects with data on four generations, locate data across four generations, explore and reinterpret intergenerational relations in depth to understand better their causal nature, and prepare an NIH application to undertake this research.