Users can click on RDUs to include or exclude them from a district or simply drag boundary lines. As they update boundaries, the statistical data shown in an embedded table or graph updates dynamically.
Today's redistricting software -- offered by Esri, Caliper and Citygate GIS -- is much easier to use than it was 10 years ago, Greninger says. And while the county's service includes support, he won't have people running to different offices to provide training and address technical problems.
The new system is also more convenient for citizens, who no longer need to travel to use a dedicated computer -- any computer with a browser and Internet access will do. And access is available 24 hours a day. "People can do it in their own homes," Pedersen says.
The unanswered question is whether the tools will produce more public input -- and generate better districts.
"We could get 10 plans or 1,000," Greninger says. Even Esri, which just recently introduced its cloud-based offering, is waiting to see what happens. "Honestly, I'm still convincing myself that the market is really ready for it," said Richard Leadbeater, manager of industry solutions for state government at Redlands, Calif.-based Esri.
Zimmerman is also taking a wait-and-see attitude. "I question whether there are that many people who want to go through the process," he says. While the tools are better, the rules of the game for producing a viable plan are complex, governed by the U.S. Constitution, the Voting Rights Act, court rulings and regulations. And when it comes to crunching the numbers to get optimized outcomes, computers can only take you so far.
A Hollywood portrayal of gerrymandering might include scenes of political operatives in front of high-performance computers in back rooms, crunching terabytes of demographic and voting history data, and optimizing districts to produce the best possible outcome for a given politician or party. Indeed, the political parties do run very exacting analyses of every voter, says Leadbeater. "Do they own cats, eat Ramen Noodles, read Esquire magazine? The parties are doing this in spades." But there are too many variables to automatically optimize district boundaries.
Some variables are straightforward, such as the requirement for each district to contain the same number of people. Others, such the definition of "communities of interest" -- which are demographic groups that must be preserved in the new setup -- are fuzzy. "There's no algorithm for that," Greninger says.
Even determining the racial makeup of a district can be challenging; for example, the 2010 census included 159 race categories. "It's a very complex, massive computation problem, and it's all about trade-offs," Leadbeater says.
Sign up for Computerworld eNewsletters.