How Do People Decide to Allocate Transfers Among Family Members?

Despite recent advances in data collection and the growing number of empirical studies that examine private intergenerational transfers, there still exist significant gaps in our knowledge. Who transfers what to whom, and why do they it? I argue that some of these gaps could be filled by departing from the standard parentchild framework and concentrating instead on fathers, mothers, sons and daughters in a way that accounts for fundamental—and sometimes obvious—male-female differences in concerns and objectives in family life. Elementary sex differences in reproductive biology constitute the basic building blocks of studies of family behavior in many disciplines, but despite recent progress they get far less attention than they deserve in economic studies of the family. I explore, separately, the implications of three basic biological facts for intergenerational transfer behavior. The first is paternity uncertainty: how does it affect the incentives of fathers, mothers and of various grandparents to invest in children? The second is differing reproductive prospects of sons versus daughters: when are sons a better investment than daughters and vice versa? The third is conflict: How much acrimony might we expect to occur in families, and why? In examining these issues I also explore household survey data from the United States. This preliminary evidence is consistent with non-biological as well as biological explanations of behavior. Nonetheless, the biological focus confers two advantages, by generating falsifiable predictions and by illuminating new avenues for empirical work. There is enormous potential for further micro-data-based empirical work in this area. ________________________________________________________________________ This paper was prepared for “The Role and Impact of Gifts and Estates,” a conference sponsored by the Center for Retirement Research at Boston College, Woodstock, VT, October 21-23, 2001. I wish to thank Tracy Vietze for comments on an earlier draft. Introduction How does a parent decide how much to support a child? Does it matter whether the parent in question is a father or mother? Or whether the child is a son or daughter? And how do we know it is the parent who is really doing the deciding? Might it be that the child has already figured out how much he or she wants, and is now prepared for a lengthy campaign to get the parent to hand it over? Perhaps surprisingly, the existing empirical literature on intergenerational transfer behavior contains few answers to these basic questions. Most analyses, for example, are gender-blind, with generic parents and generic children, rather than mothers, fathers, sisters or brothers. Models that contain husbands and wives usually do not feature anything special about being male or female—they might as well be persons 1 and 2. Not that these limitations have necessarily impeded this fast-growing literature too much to date. Advances in data collection and ever expanding empirical interest in the economics of the family have generated significant new knowledge about intergenerational transfer behavior. Recent progress notwithstanding, I argue that expanding the domain of analysis to recognize separate behavior of mothers and fathers, and sons and daughters, could generate substantial new insight into how families function. Such knowledge is of considerable policy interest. The existence, responsiveness and nature of familial transfers each matter for public policies that redistribute income. For example: if my grandmother qualifies for Medicaid benefits that pay for her stay in a nursing home, is she the true Medicaid beneficiary? Or is it my mother, who in absence of Medicaid would have had to care for her? And if such “crowding out” occurs—the substitution of government financed care for familial care—can it really be characterized as an inconsequential “neutralization”? Or does it matter what goes on inside the family? Is it better for elderly women to be cared for by their daughters or by non-relatives? Would it matter whether my mother and her mother got along well or not? What if my grandmother had only sons to rely upon rather than daughters?