You're Viewing the Archives
Return to IVN's Frontpage

The Computer Algorithm Designed to Save Democracy

by Shawn M. Griffiths, published

Imagine if there was a computer algorithm that could redraw legislative districts in a purely nonpartisan manner. Imagine this algorithm could give courts a measurable way to determine what constitutes "too partisan" in the drawing of electoral lines. Imagine it could end partisan gerrymandering for good.

Well, that may now be closer to a reality.

With gerrymandering now at least partly in the mainstream after John Oliver devoted an entire segment to the subject on his HBO show, Last Week Tonight, Vox published a video on an effort by two researchers, Wendy Tam Cho and Yan Liu, to give courts a standard they can use to adequately measure partisan gerrymandering.

The Supreme Court has not touched the partisan aspect of gerrymandering in decades because the high court has yet to "settle on a standard or definition of political fairness," says Cho. Justices currently have no precedent to declare a partisan gerrymander and are reluctant to apply any standard to it themselves.

Enter Cho and Liu's algorithm. It takes the basic criteria required by the court for redrawing legislative districts (population equity, contiguity, preserving political subdivisions, traditional districting principles, etc.) and generates maps based solely on that criteria without using any political information.

"These are by definition nonpartisan maps," says Cho.

The courts are then given a standard to work with. If maps drawn in a state don't look like any of the possibilities produced by Cho and Liu's supercomputer, then there is evidence of partisan motivation behind the redistricting.

Additionally, if political information is added to the algorithm and the maps do match any of the possibilities produced by the supercomputer, then there is further evidence of partisan gerrymandering.

So the key to solving the prevalent issue of lawmakers picking their voters, rather than the other way around, may just be our advancements in technology. If a supercomputer can generate millions if not a billion nonpartisan maps, it may give the court what they need to create a legal precedent for what constitutes "too partisan" in the redistricting process.

What do you think?

About the Author