This monograph summarizes and evaluates "educational production function analyses"--studies of the relation between inputs and outputs in an education system--in order to aid in educational manpower policy making. In addition, data from New York state school districts and from a large national sample of high school students is subjected to multiple regression analyses for information on the cost effectiveness of different inputs. Inputs which policy makers can influence--teacher characteristics (length of service, graduate training, verbal ability), class size, quality and quantity of school administrators, and use of educational technology--are examined. Comparisons are made of the effect on student achievement levels of spending a fixed amount of money on several different types of school inputs. The study finds that all inputs are not equally productive for all grade levels or all subject matter specialties. For instance, neither teacher experience nor degree affect student achievement at grades K-3; the contrary, however, is true in grades 3-6. Incentive pay might, therefore, best be turned to increasing student-teacher ratios at grades K-3 and bolstered in grades 3-6 at the expense of class size. (JH)