This article first presents the conceptual framework of the "improved" school finance. This approach clarifies that effective school resources include compound resources, complex resources, and abstract resources in addition to the simple resources usually included in production functions. The implications of this approach are then explored with the National Educational Longitudinal Survey of the Class of 1988 (NELS88), data rich enough to measure many school resources and many outcomes. The results indicate that simple resources are much less powerful than compound, complex, and abstract resources. Many effective resources are unaffected by spending levels and must be constructed within schools, explaining why money often does not make a difference to outcomes. The results also indicate that, while a few powerful resources affect all outcomes, some affect test scores but not progress through high school, while others affect progress but not learning. (Contains 2 tables and 21 notes.)