GOVERNMENT BONDS generation(generation 2). Generation 2 expects to receive a transfer of mount S while old, financed by a lump-sum tax levy on(young) genera tion 3, etc. It is assumed here that an individual's payment received while ld is independent of his own contribution to the scheme while young, and hat neither the old receipt nor the young payment depends on the amount of work, income, etc. Assuming interior bequests(which would be guaranteed by a sufficiently high value of S), a change in S would induce the current old generation(generation 1)to maintain its choice of ci and correspondingly, to raise Ai by 1/(1-r)times the change in S. This increased inheritance would just offset the increased tax liability imposed on(young) generation 2. With its consumption unchanged, generation 2 would use its own higher social security receipt to raise its bequest to generation 3, A2, by 1/(1 -r) times the change in S. As in the ernment debt. if the solutions for be impact of a marginal change in S would be solely on the size of bequests and not at all on the pattern of consumption. 16 The same results would follow in the case of operative intergenerational transfers from young to old, with a marginal increase in S implying a corresponding reduction in of gifts fre to old In particular, p oublic support of education involves a forced transfer resources from old to young. In the main, this sort of imposed transfer would be offset by adjustments in the opposite direction of discretionary transfer D. Inheritance Taxes Suppose now that inheritances(or gifts )are taxed at a proportionate rate In particular, the bequest from a member of generation i, Ai, yields a 16 As in the case of government debt issue, the formal proof depends on the assur rould also not be affected by a social security scheme that involved the accumulation of a government"trust fund, ' Assuming that the fund were held in the form of earning asse an increase in the fund would be equivalent to a negative government debt issue. Real ffects of a social security system would arise if the payme behavior of the old generation. In that case there would be allocative effects produced by the disincentive to work in later years. of educa at that(l)there vel, government education programs will involve real effects to n,(2) public expenditure on education is pressed sufficiently far so that a re- transfers cannot occur on a one-for-one basis, and 3) there ar distributional effects involving relative educational expenditures and tax liabilities across families, As an empirical matter, Peltzman(1973) has shown that public subsidies for higher education are offset to an extent of about 75 percent by reductions in private expenditures for higher education. However, Peltzman's 75 percent figure does ne coincide with the desired estimate of the effect on discretionary transfers, since othe omponents of discretionary transfers may also be affected and (on the other side) since not all private expenditures for education constitute intergenerational transfers