During the 1960's, program innovation and social analysis bloomed concurrently. Especially during the mid-years of that decade, many new social enterprises were spurred by the confident analyses of social scientists. The 1970's, however, have ushered in a period of program evaluation and public retrospection. The emphasis no longer is on building a record of program initiatives or on projecting the cost-effectiveness of prospective federal policies, but on looking backward to measure what has been accomplished by means of the activities already undertaken. The difference between analysis and evaluation is tied to the policy-making context within which social science operates. The uses of analysis are most in demand when underutilized slack resources are available, when there is a great deal of confidence in the efficacy of public action, and when policy-makers want to forge new program initiatives. When these conditions change, the dominant tone of both social science and public policy shifts to evaluation. The drainage of resources, confidence, and the will to innovate thus account for the recent rise of evaluation.