Three voting booths can handle the load for voters in the Russell and Grant Townships Tuesday, Nov. 5, 1996, at the 4-H building in Russell, Kan. Many locals will be voting for their native son Bob Dole. (AP Photo/Cliff Schiappa)

John Barrow, a Democrat, represented Georgia in the U.S. House of Representatives from 2005 to 2015.

The U.S. Supreme Court is slated to tackle one of the most divisive issues in election law: partisan gerrymandering. The June announcement that the court will review the problem in a case from Wisconsin has caused some to wonder: Do we really want the Supreme Court to decide how partisan is too partisan?

This misstates the problem. The question is not whether the courts should decide how partisan is too partisan. It is how to make it possible for voters to decide that for themselves.

Right now, due to a combination of polarized voting and partisan gerrymandering, primary voters rule the roost. As a result, a relatively small number of primary voters are vastly overrepresented in our legislative bodies while voters who show up only in general elections are vastly underrepresented. The most partisan voters end up deciding how partisan is too partisan. That's the problem.

As a former member of Congress who faced one of the most egregious cases of gerrymandering in our country, I have a modest suggestion: We should draw representative districts so that moderate voters have as much say as partisan voters. That means doing away with as many reliably partisan districts as possible.

This approach would solve two problems at once: It would prevent any partisan majority in control of the redistricting process from abusing that power to enhance their power, and it would also be a relatively easy standard for the courts to administer.

Consider what would happen if the courts were to adopt the following rule: Any map that maximizes partisan competitiveness in as many districts as possible — and maximizes partisan competitiveness as much as possible in all the remaining districts — would be presumed to be constitutional. Any map falling short of that standard would be presumed unconstitutional.

Such a map would produce the largest number of districts where there is no partisan advantage whatsoever. In such districts, the nominee of either party would have to compete for and win the support of independent voters. The partisan tilt of the overall population would be reflected in the remainder of districts where the partisan advantage of the dominant party simply cannot be neutralized.

In most states, neutral districts would become a majority; in others, a minority. But in any case, the goal should be to get as many neutral districts as possible because only these districts can temper partisan impulses. And only a neutral district gives independent voters as much say so as partisan voters.

There are plenty good government reasons for such a rule. But as far as courts are concerned, only a rule that requires a maximum number of neutral districts can effectively prevent a partisan majority from running up the score for their side. That's the constitutional reason for such a rule.

This would not require judges to wade into the political thicket and decide how partisan is too partisan. Rather, any litigation would take on the character of judicially supervised baseball arbitration. Instead of engaging in positional bargaining, forcing the judiciary to attempt the line-drawing by splitting the difference between relatively partisan maps, each side would have the incentive to come closest to hitting the goal of maximum competitiveness, on the understanding that if their map fell short, the other side's map would be adopted.

Representatives from so-called safe seats may not like this change, especially if they had to run in more representative districts for a change. But having 200 to 300 genuinely competitive seats in the House is just what we need. We have a few remaining moderates in such seats, but there aren't enough of them, and they're too easily pushed around (or pushed out) by the combination of outside money and the polarized votes of the primary voters.

Under our current system, all votes get counted, but some votes count a whole lot more than others. Primary votes count too much, and general election votes hardly count at all. That's the problem.

The solution is to make everyone's vote count as much as possible and in as many districts as possible. Lord knows we've exhausted the alternative, and it ain't working.