III2 JOURNAL OF POLITICAL ECONOMY been accomplished privately, except that the transaction costs, as measured by 2, made this transfer marginally unprofitable. Hence, the government-induced transfer implied by its bond issue can raise net wealth only if the government is more efficient than the private capital market in carrying out this sort of lending and borrowing operation Some additional observations can be made concerning this result. First, if the government is really more efficient than the private market in the lending process (presumably because the benefits of economies of scale in information? and the ability to coerce outweigh the problems government incentive and control), it may be able to exploit this efficiency etter by a direct-loan program, rather than by the sort of bond issue described above. In my simple model, a fraction a of transfers and tax liabilities involved the r group, and this process entailed a dead-weigh loss to the extent that?>0. A program which limited the loan recipients to high-discount-rate individuals would be more efficient in this respect However, the information requirements for this sort of progra much greater than those for a program which does not attempt to dis- criminate--in the transfer and tax liability aspects-among discount rates. The crucial point which can make the bond issue work as a loan program that the purchasers of the bonds automatically discriminate among themselves as to their discount rates Second, the government may be more efficient than the private market only over a certain range of B. In particular, there may be a sufficiently large value of B such that, at the margin, the net-wealth effect of govern- ment debt is zero. If the public choice process leads to this value of B (as it should on efficiency grounds), then, at the margin, the net-wealth effcct of government bonds would be zero, despite the continued existence of"imperfect private capital markets. 25PIl Ill. A Government Monopoly in Liquidity Services Suppose now that government debt provides a form of"liquidity service to the holder, in addition to the direct interest payments. Suppose that, at the margin, these services are valued at the amount L per bond per year. Hence, in the context where all individuals have the same discount rate,r, an additional perpetual government bond would be evaluated as B=(i+l/ The taxes for financing the government debt can be thought of as the interest costs, i, plus any costs involved with the process of creating arse,government debt issue would be"productive in a total sense even lere the marginal net wealth effect was nil. However it is this marginal effect hich enters into analyses of (marginal) fiscal and monetary polici