TAX TREATMENT UNCERTAINTY AND THE IRS INDIVIDUAL RULINGS PROGRAM--A COMPREHENSIVE ANALYSIS

This dissertation sets up two models to explain why the IRS may prefer not to issue rulings in some areas of taxation. The first model is based on the assumption that the IRS is a net revenue maximizer. In this model, the individual rulings program is viewed as an imperfect screening device which brings about savings on auditing costs and alleviates the problem of information asymmetry. Both the taxpayer and the IRS think of the tax liability as a random variable since the court has the final say on how much the taxpayer owes the government. Because using the court to resolve differences of opinion is costly to both parties, pretrial bargaining takes place in a natural fashion and occupies center stage of the first model. The conclusion of the first part is that the IRS may not offer the individual rulings program to discourage the taxpayer from undertaking tax-favored projects and that this incentive not to issue rulings becomes stronger as explicit tax savings and audit costs to the taxpayer get larger.; The IRS is supposed to do more than simply protect the revenue, however, and this leads to my second model. My second model is based on an alternative behavioral assumption about the IRS. In this model, the IRS is assumed to maximize an alternative measure of social welfare. Here the IRS is viewed as caring about the types of investment projects that are undertaken for reasons other than their direct tax revenue consequences. This model explores whether such a social-welfare maximizing IRS would exert an indirect influence upon the size of certain segments of the economy by use of the "no ruling" device. The conclusion is that the IRS' refusal to rule is consistent with social welfare maximization. This suggests that we can view the individual rulings program not only as an information source of the IRS but also as a fiscal policy tool to encourage socially desirable investment and to discourage socially wasteful activities. (Abstract shortened with permission of author.).