Information sovereignty and GIS: the evolution of “communities of interest” in political redistricting

Geographic information systems (GIS) first gained a prominent role in US political redistricting following the 1990 Census. Yet, this new technology—combined with a series of political and legal developments—raised a host of difficult questions about the nature of representation, identity, and community. In litigation, courts often referred to “communities of interest” (rather than racial groups) as appropriate political units to use in redistricting, but seldom defined these entities or described their characteristics explicitly. During the late 1990s, GIS practitioners responded to these qualitative judicial guidelines by attempting to identify, quantify and incorporate different measures of “communities of interest” in GIS redistricting databases. The evolution of this technology reflects such legal requirements and, more generally, the social and political systems in which GIS are embedded. Drawing on a set of in-depth interviews with legislative staff members in Texas, I argue that the state practices associated with the development of redistricting GIS created a system that reproduced a limited, conventional, geographically bounded conception of communities of interest. The case illustrates how “political representation” is both enabled and constrained by the kinds of graphical and conceptual representations that are possible with modern GIS. More broadly, the article uses Foucault’s principle of “governmentality” and James Scott’s idea of “legibility” to develop the concept of information sovereignty. The state filters out certain kinds of knowledge in order to maintain its legitimacy as a rational, unbiased observer, and thus helps create a GIS that offers only “thin” representations of community.