In this book, originally presented as the third series of H. Rowan Gaither Lectures in Systems Science at the University of California (Berkeley), are examined the contributions that systematic analysis has made to decision making in the government's "social action" programs--education, health, manpower training, and income maintenance. Drawing on her own experience in government, the author indicates where the analysts have been helpful in finding solutions and where they have been no help at all. It is urged that widespread implementation of social experimentation and accountability by the Federal government be effected. Underlying both kinds of programs should be the requirement for comprehensive, reliable performance measures. (Author/JW)