Taking the Long View: Longitudinal Surveys and the Construction of Educational Inequality in America

Abstract

Federally-funded longitudinal datasets have become a primary source of information for anyone attempting to describe, evaluate, or influence the American education system. Every year these datasets provide the raw material for thousands of research articles and assessments of the American school system including claims about educational access, attainment, and achievement gaps. Despite their considerable influence in research and policy conversations, longitudinal datasets have received little scholarly attention from historians of education, leaving the ideas embedded in them and the kinds of thinking and ways of seeing the American school system that they enabled largely unexamined. This paper takes up this issue and examines the way in which early efforts to produce longitudinal information about the operation of America’s schools – through Project Talent and the National Longitudinal Study of 1972 – shaped the research and the claims that could be made about the American education system. I argue that these efforts were intimately tied to the ambition to bring systems analysis to bear on questions of school reform – an ambition that created a demand for larger, more precise statistics on school operations – and helped co-construct new ideas about American educational equity and opportunity. While Project Talent reflected the Progressive Era view of educational equity as providing appropriately differentiated schooling, NLS72 represented a new concern for educational inequity as the difference in the aggregated chances of individual students. Though these early efforts failed to deliver on the promise to identify the precise relationship between identifiable school characteristics and school outcomes, they demonstrated the general viability of these techniques and pointed the way toward this new frontier of research. In attending to this history, I seek to bring attention to what has become a foundational part of the American educational research infrastructure and the influence that infrastructure has had on education research. In providing new capabilities for researchers and advocates, NLS-72 influenced the kinds of claims – for both good and ill – that would define policy conversations for decades to come: bringing into focus a broader picture of national disparities while at the same time obscuring local context and geographic distance. Just as consequentially, the initial survey continues to cast a shadow on American education research through the repetition of initial variables and survey constructs in subsequent decennial surveys (e.g. ELS, HS&B, NELS, etc.) – shaping the way we think about educational success, failure, and inequity in America.