The article reports preliminary observations on data from six colleges on application of a regression model to help explain observed differences in instructional efficiencies of specific curricula. A major finding was that the most efficient distribution of students is in the humanities and social sciences. Tentative explanations for this finding are offered. (DB)