• Subscribe

Blue Waters restores fairness to elections

Speed read
  • Electoral districts are hand drawn by incumbent political parties.
  • Research tool promises to identify gerrymandering and restore fairness to elections. 
  • Blue Waters supercomputer made possible the intractable computations.

About six weeks from now, US citizens will head to the polling booths for the 2016 election.  Americans are wary of their choices, and part of the suspicion has to do with the voting districts, hand drawn to benefit political interests.

The US constitution mandates these districts be redrawn every decade, in an attempt to respond to ever-shifting demographics. However, in most states the majority party maintains control of drawing district lines. When vested politicians purposefully draw districts to unfairly favor their interests, the result is skewed districts where election results are largely a foregone conclusion.

Restoring democracy. A social science research project at University of Illinois at Urbana-Champaign brought Blue Waters to bear on the problem of gerrymandering. By allowing vested political interests to draw the voting districts, election outcomes are virtually predetermined. Courtesy National Center for Supercomputing Applications.

Take a look at the re-election rate in the House of Representatives, as an illustration. Over the last four decades (1972– 2012) incumbents have been reelected 93.6 percent of the time – yet voter approval of these representatives has perennially hovered around 15 percent.

“Placing self-interested actors at the helm of such a crucially important process seems ill-advised, and the incongruous discrepancy between the reelection rate and the approval ratings suggests that something is badly awry,” says Wendy K. Tam Cho, professor at the University of Illinois at Urbana-Champaign. “It is not, as we wish to assume in a democracy, the voters choosing their representatives.”

In response to President Obama’s call to change this undemocratic dynamic, Cho decided to bring high-performance computing (HPC) power to the fight. 

Cho and Yan Liu recently were awarded first prize in the Common Cause 2016 First Amendment Gerrymander Standard Writing Competition for their proposal for a partisan standard.  They used the Blue Waters supercomputer, an NSF-funded HPC resource to create a digital means to draw hypothetical district maps – about 800 million of them, in fact.

This large number of maps helps to identify political gerrymandering by providing a comprehensive context for officials to consult. Their approach will be published later this year in the Election Law Journal.

"The discrepancy between the reelection rate and the approval ratings suggests that something is badly awry."  ~Wendy K. Tam Cho

“Our computational tool has the ability to make the process eminently more fair and transparent and has the potential to engage a much broader array of interested citizens,” Cho observes. “Legislators and others who participate in the redistricting process would no longer need to rely on their own limited computational capacities to shape districts. Judges could tap into this resource to help them adjudicate proposed redistricting plans, assess their fairness, and propose districts on the basis of selected criteria.”

Though maps can be drawn by hand, in order to accurately characterize the redistricting problem, an enormous number of maps are required. Drawing electoral maps is equivalent to the set-partitioning problem where the number of possibilities quickly exceeds the number of seconds in the history of the universe – in other words, a computationally challenging task.

Blue Waters provided supercomputing power through two resource allocation awards for the researchers to develop scalable HPC solutions.  Liu, Cho, and Wang approached redistricting as a combinatorial optimization problem, tailored to satisfy legal requirements by developing a parallel evolutionary algorithm for redistricting, scalable to 131,072 processor cores on Blue Waters. 

The technical work for this approach was recently published in the journal Swarm and Evolutionary Computation

<strong>Feeling Minnesota.</strong> Images of counties and a superimposed adjacency matrix are from a research project seeking to provide election officials with computational tools to identify gerrymandering in election districts. Courtesy Wendy K Tam Cho.

“Even with a modest number of units, the scale of the unconstrained map-making problem is awesome,” says Cho. “The use of HPC is essential to synthesize the information for this astronomically large combinatorial problem.”

America’s democracy problem wasn’t created overnight, and Cho knows it won’t be undone quickly either. But by bringing HPC into the mix, an objective measure can level the electoral map and restore fairness to the political process. Maybe then accurate representation of the public can be implemented.

“The political process is sufficiently insulated that few citizens possess more than a cursory knowledge of how election districts are formed.  But if successful, our redistricting tool has the potential to encourage broad, intelligent participation by watchdog groups, members of the media, and ordinary citizens. Without it, democratic goals would otherwise be inconceivable.”

Join the conversation

Do you have story ideas or something to contribute? Let us know!

Copyright © 2022 Science Node ™  |  Privacy Notice  |  Sitemap

Disclaimer: While Science Node ™ does its best to provide complete and up-to-date information, it does not warrant that the information is error-free and disclaims all liability with respect to results from the use of the information.


We encourage you to republish this article online and in print, it’s free under our creative commons attribution license, but please follow some simple guidelines:
  1. You have to credit our authors.
  2. You have to credit ScienceNode.org — where possible include our logo with a link back to the original article.
  3. You can simply run the first few lines of the article and then add: “Read the full article on ScienceNode.org” containing a link back to the original article.
  4. The easiest way to get the article on your site is to embed the code below.