This paper reconsiders long-term, national or socially contextualized trends in a number of important education areas commonly addressed by policy interventions. These include widening participation, teacher training, school performance, the use of targets, and student assessment. In each case, it is very difficult to sustain an argument that intervening policy changes do anything at all to disrupt or ameliorate their intended target. This raises the intriguing possibility that policy is often an epiphenomena, formalizing and recognizing the ongoing changes that are its ostensible and stated purpose. Given the proclaimed success of many of the same policies by their advocates, this paper also raises questions about the extent to which research and evaluation genuinely tries to find out what works. (Contains 4 tables and 4 figures.)