The goal of this paper is to carefully document the characteristics of within-family monetary transfers in the United States, using all nine waves of the Health and Retirement Study (HRS). Using the HRS, we construct a novel child-level longitudinal dataset and augment it with detailed information at the parent level. Consistent with previous studies, we document that intra-family transfers are significant in their incidence and magnitude and that, on average, they flow downward, from parents to children. Since we observe families for as long as 18 years, we are able to document new facts and establish a link between the early parental transfers and children's characteristics later in life. Among many facts we document, we find a strong incidence of parental transfers during child episodes of negative financial shocks, such as a job loss or divorce. We also find that children receiving larger transfers early in life are more likely to belong to a higher income class later. Parental giving is positively related to parental wealth, age, and liquidity, and children's permanent income, while it is related negatively to children's age and children's income category at the time of giving. We find that providing parents with help and attention is not a significant determinant of parental transfers overall. However, after conditioning on the parental decision to give to at least one child, levels of attention and help provided by a given child are important determinants of the relative (to one's siblings) parental transfers. Overall, we find that, independent of the motive for giving, parental transfers help improve children's welfare
