On the basis of a microeconometric examination of Philadelphia School District data, it is concluded that (1) many school inputs do matter, (2) disadvantaged students can be helped by particular types of inputs, and (3) these optimistic results are accounted for by the use of pupil-specific data, and appropriate statistical data. The school inputs examined include teacher quality and school quality (e.g., class size). Peer group effects are discussed. Genetic and socioeconomic inputs of the students are examined. The importance of examining school inputs differentially is established by a comparison of the results with and without interaction analysis. The importance of using disaggregated data is emphasized by a comparison of the results when pupil-specific data for school inputs are used with the results when school averages are used. (Author/MC)