When the media reports on climate change, it’s usually built upon the scientific community’s consensus on the issue. Partisans and pundits might reject that consensus, but the media tends to put the evidence first and the politics second. That is precisely how the press should cover what could be the nation’s next gigantic story.

President Donald Trump may not fire Special Counsel Robert Mueller or Deputy Attorney General Rod Rosenstein. But if he does, the critically important story would be the president versus the rule of law and the basic structure of the American republic, not Trump versus the Democrats (or however he’ll frame it).

After all: If journalism is about covering the most newsworthy things, then the inevitable war of parties and personalities that follows are nowhere near as newsworthy as a president breaching a crucial norm. That’s something everyone, regardless of partisan leanings, should be able to agree on.

Good coverage would accurately report on the reactions of the overwhelming bulk of “neutral” experts. Of course, those experts — historians, political scientists, law school professors — are not really neutral. No one is. But for the most part their bias is in favor of the rule of law, not to a political party. If the bulk of them come down on the same side, that’s an important part of the story; it would be a disservice to the public to hunt down a rare case of dissent and portray it as if it was on equal footing with the majority consensus.

Even more important is to accurately convey the weight of partisan dissent if there is any.

Many in the media really botched this during the general election campaign in 2016. The fact that quite a number of high-profile Republicans disowned their nominee should have been a central part of explaining what was going on. The press too often treated the unhappiness of Republican governors and senators as a given, instead focused on what they didn’t do (endorsing the Democrat) rather than what they were doing (refusing to endorse the Republican). Even those who did eventually support Hillary Clinton — including former President George H.W. Bush! — were repurposed as generic Trump enemies. The “neutral” television news, especially CNN, were particularly guilty of this, with a structural format that almost guaranteed that anything would be viewed as balanced. If their panelists weren’t balanced, CNN would hire new Trump-favoring people to keep things that way.

That’s going to be a real temptation for many if Trump acts and is criticized by Republicans such as Senators Jeff Flake, Ben Sasse, Bob Corker, and Susan Collins. In fact, each high-profile Republican defection is really big news, and complicates any “standard partisan fight” frame. But that frame can be inaccurately restored by just treating both Democrats and dissenting Republicans as generic Trump opponents, who naturally would oppose him since that’s what opponents do. In other words, there’s a temptation to heavily discount the importance of what previously dissenting Republicans say; after all, they haven’t been loyal Republicans. But their previous opposition to lesser attacks on democracy is itself a continuing and important story, and if anything makes their reactions more relevant if Trump goes farther.

It’s not about counting votes in a narrowly divided Senate, or even about counting potential votes in an impeachment trial. It’s about conveying the information generated by what members of Congress and other political leaders do, whether by speaking up for the president, staying silent, or speaking against his actions.

Why does all of this matter? Because voters take cues from high-visibility party leaders.

Political participation is not an all-or-nothing affair. Voters aren’t systematically tracking the shifting positions of party politicians on a daily basis. They may not even consciously seek out opinion leaders at all. Instead, an important amount of them get a general sense of what is going on — what’s important right now, what people are saying about it — as they graze through the media landscape. If they determine that there are two more or less equal sides to an issue or an election, and all the people they like are on one side with all the people they don’t like on the other side, they’ll wind up echoing the party position. However, if they hear a muddle, with people they like on both sides, then they will wind up splitting. And if they hear unanimity, they’ll generally go along.

Obviously, it’s not the media’s job to tell voters how to react (whatever that would mean). But it is their job to get the story right. It’s just a fortunate thing that accurately describing whether there’s an elite consensus or not, and whether the parties are unified in opposition to each other or not, will hurt a president who is opposed by both expert consensus and a good deal of his or her own party.

All this is true for Trump coverage even now, since his relentless attacks on investigators has long since crossed the line between normal political back-and-forth and abuse of office. It will be even more important if Trump makes more overt moves to shut down the investigation. The media needs to be ready.

Jonathan Bernstein is a Bloomberg View columnist. He taught political science at the University of Texas at San Antonio and DePauw University and wrote A Plain Blog About Politics.