A Dartmouth researcher says the mapping technology of GIS, or Geographic Information Systems, is a powerful political tool, but it does not resolve the basic conflict of how to create voting districts that are both representative and competitive.
Benjamin Forest, an associate professor of geography, has studied the recent history of redistricting, and says that three things currently affect political representation. First, the U.S. population has become more racially and ethnically diverse in the last 40 years. Second, there has been a growing legal commitment to fair political representation. And third, GIS allows states to create districts with very precise political and demographic characteristics.
“The growing power of GIS technology has increased the potential for abusive gerrymandering,” he says. “On one hand, the technology has been very useful for voting rights enforcement, and particularly for creating districts with African American or Hispanic majorities,” he says. “But GIS also increases the potential for sophisticated gerrymandering. In most states, legislatures control redistricting, and they typically use it for partisan advantage and for incumbent protection. The ability to evaluate and predict voting behavior and to then create districts based on these analyses can give political parties more control over election results.”
Forest’s study appears in the Oct. 17 issue of the Proceedings of the National Academy of Sciences. In his paper, he explains that GIS offers both benefits and risks because it can be used in the redistricting process to manipulate the outcome of elections, and courts have had a difficult time articulating clear and precise rules governing the use of the technology.
“There is a real irony with GIS. The technology theoretically permits states to consider a vast amount of information during the redistricting process, including things like ‘communities of interest.’ Unfortunately, a concept like ‘community of interest’ is difficult to quantify. Texas, for example, made a genuine attempt to represent ‘communities of interest’ using GIS, but this information ultimately had little impact on the state’s redistricting plans. Even if GIS provides somewhat better data on ‘communities of interest’, it also contains much, much more precise partisan and demographic information.”
Forest explains that state legislatures have tremendous incentives to use partisan information inherent in GIS, but few incentives to respect ‘communities of interest’ or to create competitive election districts. Nonpartisan redistricting commissions are one response to such concerns, and he cites Arizona as an example (California and Ohio are also considering redistricting commissions). In principle, he says, such commissions can use GIS to create districts that are relatively representative and that provide for relatively vigorous political competition.
“Just having a commission is not enough, however. States can design redistricting commissions well, to insulate them from partisan pressures, or they can design commissions poorly, to simply create proxies for partisan interests. In any case, even a well-designed commission needs a set of clear rules for redistricting.”
Forest thinks that GIS technology is a great tool for redistricting, but it will not improve political representation by itself.
“I think the real lesson is that the technology is less important than institutional arrangements. We need to reconsider how we want to create electoral districts in the first place.”
Dartmouth’s Rockefeller Center for Public Policy and the Social Sciences provided funding for this research.