This paper examines the origins and influences of the introduction of longitudinal student data-sets as a way of gaining insight into the operation of American schools and as a tool for policy-makers. The paper argues that the creation of this new form of data in the 1960s and 1970s represented a relatively new way of thinking about American schools that allowed policy-makers to view the American education system as relatively uniform and the goal of policy to optimize its function. The use of data in this way produced, somewhat paradoxically, a higher precision and more distorted view of the American schooling. The detailed picture provided by these data-sets offered policy-makers new and important insights into the pathways through schooling of individual students while at the same time presented scholars and policy-makers with an increasingly abstract and decontextualized view of the American education system as a whole. That is to say, to view the system uniform and national terms. The upshot of the unprecedented scale and detail of the data collection effort was a more stylized portrait of the American education "system"--one that appeared more amenable to federal intervention but no more likely to secure educational equity for its students.