Data for social justice and public health are akin to the proverbial two edged sword. To the extent we base any of our claims about social injustice in evidence, we must use data—whether of the quantitative or qualitative sort. But data do not simply exist. By contrast with the literal definition of “data” as “that which is given,” data instead are duly conceived and collected, via the ideas and labour of those who would obtain the requisite evidence.¹ In the case of epidemiology, moreover, we must often use population data appearing in categories that are far from ideal—precisely because the assumptions of those with the power to shape and accrue the data often differ from those who seek to use these data to illuminate and oppose social inequalities in health.

Instructively highlighting these tensions are issues that …