Statecraft by Algorithms

Abstract

Throughout the 1920s Congress failed to pass an apportionment bill. Among the various reasons for the failure was Congress’s inability to decide which method, or algorithm, should govern apportionment. Two competing methods, major fractions and equal proportions, rose to the top and the supporter of each claimed that his method offered the only “fair” and “unbiased” solution to the problem. Following this early debate, this chapter argues that contemporary concerns about algorithm fairness and equality do not emerge out of the nature of their complexity. Rather, they inhere in the incommensurability between mathematical and social rationales and the wide room for interpretation yawning between the two. Not only are definitions of “fairness” multiple but how algorithms are described, either through method or through principle, can lead to completely different results.