EDITORIAL

Math on trial

See allHide authors and affiliations

Science  09 Aug 2019:
Vol. 365, Issue 6453, pp. 523
DOI: 10.1126/science.aay9836

Embedded Image
CREDIT: STANFORD LAW SCHOOL

Can science help take the partisanship out of politics? To some extent, that was the question the U.S. Supreme Court confronted in this year's blockbuster partisan gerrymandering case. The Court answered “no” and, by doing so, delivered an object lesson in the limits of math and science to settle the most political of constitutional questions.

The Supreme Court has wrestled for decades with the problem of how to determine when partisanship in the redistricting process becomes so excessive that it violates the Constitution. As with pornography and the First Amendment, it never sufficed for the Court to say, “I know it when I see it.” The Court sought an “administrable standard” to separate tolerable levels of partisan consideration in the redistricting process from intolerable levels of gerrymandering. And in a 2004 Pennsylvania case, Justice Anthony Kennedy all but invited scientists and lawyers to come up with new ideas for clear constitutional tests.


Embedded Image
CREDIT: ADAPTED BY C. AYCOCK/SCIENCE FROM JARED RODRIGUEZ/TRUTHOUT/FLICKR/CC BY-NC-SA

Mathematicians and data scientists rose to the challenge. Specialists in geometry proposed new measures of district compactness, computer scientists offered new ways to randomly generate thousands of plans to assess an enacted plan's partisanship, and “teched-up” political scientists suggested new notions of “partisan symmetry” or “wasted votes” to ensure political fairness. These methods and others were offered in cases such as the ones from North Carolina and Maryland that made their way to the Supreme Court this year.

For the most part, each of these methods tried to prove the obvious: that egregious partisan gerrymanders were as biased and unfair as the subsequent election results proved them to be. But the Court could never rally around a single “administrable standard”—one that could be easily stated and operationalized as a matter of constitutional law. This was particularly important in the gerrymandering context given the high political stakes involved and the risk that judicial decisions would be seen as picking partisan winners and losers.

With the critical replacement of Justice Anthony Kennedy with Justice Brett Kavanaugh, the Court then had a five-member majority willing to put a nail in the coffin of partisan gerrymandering claims. In an opinion for that new majority, Chief Justice John Roberts explained why the problem of partisan gerrymandering was simply too difficult and too political. All the math presented by the plaintiffs, the lower courts, and the dissenters was largely irrelevant for the majority because it did not provide a bright line rule derived from arithmetic akin to “one person, one vote.”

Although scientists have been quick to criticize the Court's decision, Chief Justice Roberts's inhibitions are characteristic of what I see when I work with judges in these cases. (I have been appointed by courts on numerous occasions to draw plans to remedy one or another constitutional violation.) Judges often worry that “you can find statistics to support anything” (as I once heard a lawyer argue in a redistricting case). And in today's highly polarized political environment, judges know that whatever the merits of their decision in a redistricting case, they will be accused of favoring their own preferred party.

All is not lost, however, for scientists wishing to see their favorite method enacted into law. Although the federal courthouse door is shut to them, the state courts remain open. And for the entrepreneurial, legislation at the state level—through popular initiatives and action by less politically riven legislatures—has proven effective recently. Indeed, scientists should feel particularly at home in using the states as “laboratories of democracy” (as Supreme Court Justice Louis Brandeis described them in 1932). Should they find a standard that works, maybe some future U.S. Congress—admittedly, quite far in the future—will provide for national standards that the Court could not find in the Constitution.

View Abstract

Navigate This Article