In this paper, the author attempts to clarify several implicit hypotheses about local school finance reform, set up tests whereby hypothesis validity can be affirmed or rejected, and outline the policy implications of the results. Two mathematical models of school district behavior are examined, and their implications are tested on a sample of 150 Massachusetts school districts. The first model assumes that the district reacts to the "effective" property base, so that a change in revenue is viewed identically whether the change resulted from a change in property tax rates or in property values. The second model indicates that reactions to changes in tax rates and to property values are different. It is demonstrated that the first model, which has dominated school finance reform litigation of the past five years, is unsupportable; and that its policy recommendations are insufficiently restrictive. (Author/DN)