THE toughest part of being a Supreme Court justice is not deciding who wins. It is writing the opinion explaining why the winning party should prevail. And since Supreme Court decisions serve as guides to courts in future cases, a majority opinion needs to do more than provide a justification for why A wins over B. It must anticipate similar cases that could arise and draw lines showing how those ought to be resolved.

On March 28th, in a case contesting a Maryland map from 2011 that painted a solidly red congressional district blue, the justices seemed to feel a particularly heavy burden. This is not surprising. Benisek v Lamone marks the second time the court has stepped into an electoral politics minefield this term, and in three months’ time, the justices must decide whether to try to put the brakes on partisan gerrymandering—a scourge of American democracy whereby politicians rig elections in their own favour. The problem is, defusing some mines risks tripping others.

Benisek concerns a rather brazen move by Maryland Democrats to buy themselves an extra representative in Congress. In 2012, after the new map was in place, an upstart Democrat unseated a Republican who had spent 20 years in the House of Representatives. It wasn’t close: the Democrat won by 21 points, two years after his rival had won re-election by 28 points. Justice Elena Kagan explains the reversal of fortunes: the “Maryland legislature...shuffle[d] 360,000 people out and [brought] in 350,000 people. The result of that is that the district went from 47% Republican and 36% Democratic to exactly the opposite, 45% Democratic and 34% Republican”. This was not happenstance. “How much more evidence of partisan intent could we need?” Justice Kagan asked Michael Kimberly, the lawyer for Maryland.

There is more, though. In a deposition last year, Maryland’s former governor, Martin O’Malley, was candid about his support for refashioning the sixth congressional district. Defeating the 10-term Republican was “my hope," Mr O’Malley said. "It was also my intent to create … a district where the people would be more likely to elect a Democrat”. All this adds up to rather solid evidence that Maryland legislators meant to, as the plaintiffs say, “retaliate” against Republican voters in violation of their right to freely associate under the First Amendment. But Justice Stephen Breyer started thinking down the road to cases where the culprits were less careless in broadcasting their intent. The Maryland example is “extreme” he said, “100% partisan”. “Now could we say that? Yes, I think we could.” But chastising Maryland Democrats for this sin is “not going to solve other cases and we’ll never have such a record again”. How, Justice Breyer wondered, should the court deal with a national “problem of serious dimensions”? Gerrymandering has already been weaponised by sophisticated computer modelling, but we haven’t seen anything yet. “If you think what’s happened now is something, wait until you see those computers really working”, Justice Breyer warned Steven Sullivan, the lawyer for Maryland.

In October, when the justices heard Gill v Whitford, a 14th Amendment equal-protection challenge to a gerrymander engineered by Wisconsin Republicans to hurt Democrats running for the state legislature, even Justice Samuel Alito admitted that highly skewed districts are “distasteful”. But then and now, Justice Alito doubted whether any line could be drawn between acceptable and excessive levels of partisanship in creating district maps. In Benisek, he painted Mr Kimberly into a corner with a hypothetical involving a legislature’s choice between two otherwise unobjectionable voting districts—one slightly benefiting the majority party and the other slightly benefiting the minority party. Let’s assume the majority party picks the district that serves its interests. Justice Alito: “Now, in your view, that's unconstitutional, I gather?”

The problem, as Justice Alito showed, is that the Supreme Court has always held that states have a wide berth in deciding how to draw their district lines. “Hasn't this court said time and again you can't take all consideration of partisan advantage out of districting?” Under the First Amendment retaliation theory, “I really don't see how any legislature will ever be able to redistrict”, he said.

So there is the dilemma. If the Supreme Court concludes that the First or 14th Amendment bars excessive partisanship in redistricting, it will have to lay out what would qualify as an acceptable level of partisanship that won’t get states into constitutional trouble. Yet the plaintiff’s proffered standard seems to swallow partisanship whole. What then? Will the courts be faced with a sheaf of lawsuits that, in Justice Breyer’s words, “get judges involved in dozens and dozens and dozens of very important political decisions?” No federal judge wants that, least of all Supreme Court justices.

It seems that some may be looking for a way out of the morass. Chief Justice John Roberts, as well as Justices Alito, Ruth Bader Ginsburg, Sonia Sotomayor and Anthony Kennedy, filled 14 transcript pages at the beginning of the hearing noting that nothing the court says this spring could possibly make a difference in the 2018 elections coming up in the autumn. This has the makings of a procedural punt. Combined with Justice Breyer’s musing over the possibility of saving the matter for the autumn (when the parties in Gill, Benisek and a similar case pending in North Carolina could assemble with “blackboard” in hand), it suggests that kick-the-can may be one strategy the justices are considering. But even this would have its costs—and constitute a tough decision in itself. If the justices delay things much longer, they may have to do without swing Justice Kennedy who, at 81, is rumoured to be thinking about retirement. Without his vote, any curbs on America's most anti-democratic manoeuver may prove elusive.