The Internal Revenue Code invokes the concept of reasonableness as the major qualification for the stockholder executive compensation deduction for federal income tax purposes. However, neither the Code nor Regulations contain general guidelines for determining reasonable compensation. Consequently, disputes with the IRS are frequent, resulting in substantial litigation. The primary hypothesis of the study was that the IRS guideline variables were incapable of discriminating taxpayers who have won litigated reasonable compensation cases from those who have lost. The secondary hypothesis was that the IRS guideline variable group, the court case variable group, or the two groups combined were equally powerful in discriminating taxpayers who have won litigated reasonable compensation cases from taxpayers who have lost. The study included all unreasonable compensation cases litigated in the Tax Court from 195^ to September, 1980. Only cases related to the reasonableness of officer-shareholder compensation of closely-held corporations were included.

This study deals with three problems—the influence of the study of accounting on the value-level choices of individuals, a comparison of value—level choices among accounting populations, and a comparison of value-level choices between accounting and each of four other disciplines. Two models were developed to test sixteen hypotheses related to these problems. One model isolates any influence of the study of accounting on value-level choice and the other model develops value-level profiles without regard to cause. Multiple Regression Correlation techniques were used to analyze the data.