In one of the most closely watched gerrymandering controversies in the country, the Pennsylvania Supreme Court just redrew the state’s congressional maps, in a way that has Democrats pleased and Republicans livid. The result makes a Democratic takeover of the U.S. House of Representatives this fall significantly more likely.

But more than that, it forces us to ask whether we are going to fundamentally change a system that has been in place for most of our country’s history.

The question we have to answer is this: What should determine how district lines are drawn? On one side we have fairness — not always possible to achieve with perfect certainty, but at least a goal one could seek — and on the other side we have the exercise of raw political power for partisan advantage.

You’ll never guess which side the Republican Party comes down on.

As helpful as the court’s new map may be to Democrats right at this moment, that’s only because the existing map is so absurdly skewed in favor of Republicans. After the 2010 Census, Republicans in control of Pennsylvania’s legislature drew a map that gave them control of 13 of the state’s 18 House seats, despite the fact that according to the state government, there are currently 4 million registered Democrats in Pennsylvania and only 3.2 million registered Republicans. The state Supreme Court ruled that the map violated the state’s constitution, then ordered the legislature and the governor to come up with one that treated all voters fairly. When they couldn’t agree on one, the court did it itself, with the help of Nathaniel Persily, a Stanford Law School professor and redistricting expert who has provided the same service to other courts.

The map they came up with is awfully hard for Republicans to criticize on a substantive basis. Of the 18 new districts, eight were won by Hillary Clinton in 2016 and 10 were won by Donald Trump. The districts are much more compact than in the old map, without any of the stretched and winding lines that characterized them before and with only 13 of the state’s 67 counties split between multiple districts (the number was 28 in the old map).

But when your starting point is a map as skewed as Pennsylvania’s, any attempt to reset to something fair looks like a pro-Democratic effort. As the New York Times described the consequences, “Over all, a half-dozen competitive Republican-held congressional districts move to the left, endangering several incumbent Republicans, one of whom may now be all but doomed to defeat, and improving Democratic standing in two open races.”

Naturally, Republicans are portraying this as an unconscionably unfair decision. The National Republican Congressional Committee announced that “State and federal GOP officials will sue in federal court as soon as tomorrow to prevent the new partisan map from taking effect.” And President Trump is outraged:

What’s so remarkable about Republican anger at the Pennsylvania Supreme Court’s decision is that they’re acting as though the court did the reverse of what Republicans had done, which isn’t the case at all. Had the court drawn a map that produced a 13-5 Democratic advantage, Republicans could then reasonably say, “The court is being as unfairly partisan as we were!” That might be hypocritical, but it wouldn’t be inaccurate.

However, that’s not what the court did at all. It drew a map that will likely produce a congressional delegation evenly divided between Republicans and Democrats, give or take a seat or two in either direction. In other words, they seemingly sought to produce a situation of partisan parity, despite the fact that there are about 800,000 more Democrats in Pennsylvania than Republicans.

To the GOP, this result — a congressional delegation that accurately represents the state’s voters — is apparently a horror.

So this is the question: When district lines are drawn, should the two parties should wind up with about as many seats as their support among the voters would predict? There will of course be some variation based on geography, the skills of particular candidates, and what’s happening in the broader environment in a given year. And no map should be expected to perfectly reflect the voters’ will. But we have to decide whether partisan fairness is one of our goals.

The U.S. Supreme Court hasn’t held partisan fairness as a requirement, and this question is currently before the high court, which is deciding on a challenge to a state legislative map that Republicans rigged in their favor in Wisconsin. Which means the court may soon declare in some form that some level of partisan fairness is a requirement.

That isn’t something Republicans want. They’re quite happy with the situation as it is, one in which whoever’s in power gets to essentially pick their voters, with tools that have become so sophisticated and easy to use that a party can all but guarantee the election outcome it wants before voters ever step into the voting booth. That’s because Republicans happen to have had a great election in 2010, just as redistricting software was really coming of age, and used it ruthlessly in the round of redistricting that happened after that year’s census. Now they want to lock in their advantage.

I have to make clear that when they’ve had the opportunity, Democrats have used their power to construct gerrymanders to favor themselves, too. There may be no better current example than Maryland, where Democrats used some extremely creative cartography to enable them to win seven of the state’s eight congressional seats. This case is also going to be heard by the U.S. Supreme Court.

But despite the advantage they’ve gained in places such as Maryland, as a whole Democrats would be more than happy to give up that advantage and turn every state’s redistricting over to a nonpartisan commission. Depending on how the Supreme Court rules, that may be where we’re headed. But we’ll have to decide that fairness is actually something we value. It won’t guarantee any outcome, because there’s no such thing as a perfect map in any state. But it’s a place to start.