This essay examines the evolution of data-collection practices in American expenditure surveys over much of the twentieth century. Economists conducting expenditure surveys faced one fundamental concern: their success hinged on the cooperation of interviewees (typically housewives) and the reliability of their testimony. Investigators recognized that both dependencies posed serious problems, and they struggled to devise effective solutions. I argue that over the course of the twentieth century, the methodology of US expenditure surveys evolved in a nonteleological way, as new approaches led the BLS to adopt previously rejected techniques and to abandon other strategies that were formerly of central importance. In charting this history, the essay reveals the many challenges and complexity of gathering a seemingly straightforward form of economic data.

The text of this article is only available as a PDF.
You do not currently have access to this content.