FISCAL POLICY AND ENDOGENOUS GROWTH IN A BEQUEST-CONSTRAINED ECONOMY

Abstract

Evidence suggests that the vast majority of planned altruistic transfers between generations are human capital investments in children (households are generally bequest-constrained). This paper demonstrates that empirically-based calibrations of simple overlapping generations models with altruism generate binding bequest constraints with realistic endogenous growth rates and returns to capital. We also show how intergenerational and intragenerational redistributions of wealth affect long-run growth in bequest-constrained economies. Redistribution is a feature of fiscal policy abstracted away from by the standard infinitelylived representative agent models used to analyze endogenous growth. THE INFINITELY-LIVED representative agent model (oo-model) has been the predominate tool used to examine long-run endogenous growth (Romer 1986; Lucas 1988; Rebelo 1991). The implicit justification for assuming an infinite planning horizon for the agent is that the current generation is linked to future generations by altruistic transfers (Barro 1974). This paper makes these transfers explicit by using an overlapping generations model with altruistic parents that invest in the human and financial wealth of their children (OLGA-model). In environments with and without uncertainty, we establish conditions under which the model (i) exhibits endogenous growth and (ii) behaves differently than the oo-model. In addition we examine how growth rates respond to fiscal policies that involve intergenerational and intragenerational redistributions of wealth, features ignored by the literature using the oo-model. The essential insight behind the difference between the OLGA-model and the ci&-model is given by Drazen (1978) and Becker (1981). Their model assumes overlapping generations, where each household is headed by altruistic parents that may affect their children's well being through transfers of both human capital and financial assets (referred to as 'bequests'). Legal restrictions constrain the bequests to be non-negative; parents cannot leave legally binding debt to their children. If it is optimal for the household to make human capital investments and leave non-negative bequests, then the model behaves as if the household is infinitely-lived. However, if the desired bequest is negative and the legal restriction binds, then parents only make transfers in the form of human capital investments. When households are bequest-constrained the model posseses many of the characteristics of the standard overlapping generations model without altruism. In particular, lump sum intergenerational transfers carried out by the government can have real effects (despite the presence of altruistic human capital transfers (Drazen 1978)).