Good morning, everyone. It's a delight to be here, to see you all flourishing, and to get a sense of this of evaluators continuing to grow in size and in accomplishments despite inhospitable terrain in government today. I see us as being a little bit like suburban dandelions: you know ... people spend their whole weekend trying to get rid of them but when Monday afternoon rolls around, there they are back again in full and glorious bloom: bright, stronger than ever, and not just unabashed, but thriving. This meeting, I know, marks point at which, as a society, we've begun to think about our past, about paths we've travelled and our reasons for choosing them, and about progress we've made, all along line. I've been asked, in particular, to say what I think we've learned about of evaluation. To do that, as you might well imagine, I went back in time-oh, say 10 years-to remember what we used to say or write about issue of in evaluation and then contrasted that with our approaches today. Here, of course, you should note that I'm speaking from my own personal perspective and experience as head of Program Evaluation and Methodology Division (PEMD) at GAO, and I'm also speaking about a process-the political or process-that is so enormously variegated, so full of chiaroscuro, so inclusive and tolerant of exceptions and special cases, that generalizations hold some peril for generalizer and should be understood within context of this experience and this environment. The experience, however, is not insignificant. PEMD has now produced more than 50 reports for different committees of Congress and we have also worked, in a more transient capacity, with many more. What then do we mean when we talk about politics of program evaluation? David Easton wrote 30 years ago that the study of is concerned with understanding how authoritative decisions are made and executed for a society (Easton, 1957, p. 383), and that the output of a political system is a political decision or policy (Easton, p. 395). Public policy, then, is product of politics. Evaluation, with its purpose of providing high-quality information to decisionmakers, thus automatically claims a role for itself in political process, based on idea that best information, made available to decisionmakers, would surely be useful to them in making and executing public policy. Unfortunately, in very early 70's, rumors began to be heard in land (later, alas, confirmed by studies) to effect that evaluations were not, in fact, often used in decisionmaking. Plenary Address to American Evaluation Association, October 30, 1986 in Kansas City, MO. The views and opinions expressed by author are her own and should not be construed to be or position of General Accounting Office. This article previously appeared in Evaluation Practice, Vol. 8, No. 1, February 1987, Sage.