For months now, the big question within the House Democratic Caucus over impeaching President Donald Trump has been framed in one specific way: Would Democrats start an “impeachment inquiry?” Of the 235 House Democrats, 100 now publicly support such a move, per CNN’s count.

Members of a key House committee said on Friday that, well, they’re doing it — kind of. Sort of? Maybe? But also, not really.

House Judiciary Committee Chair Jerry Nadler (D-NY) told reporters that, as part of a lawsuit seeking records from the Mueller investigation, the committee would tell a judge they’re considering whether to use “a constitutional power of the utmost gravity: recommendation of articles of impeachment.”

But asked repeatedly if his probe should now be categorized as that long-awaited “impeachment inquiry,” Nadler demurred. At one point, he said that “too much has been made of” that particular phrase. He characterized it as a continuation of the committee’s existing investigation into President Trump’s “malfeasances” — one that would naturally consider the possibility of impeachment. Eventually.

Rep. Jamie Raskin (D-MD) mused that, well, maybe there already is an inquiry. “A lot of people believe we’ve been in an impeachment inquiry ever since we started looking into high crimes and misdemeanors,” he said. His point was that there was no official rule for how an impeachment inquiry begins. But his own preferred word choice, he said, is that “we are in an impeachment investigation.”

This hair-splitting shows the fundamental dilemma House Democrats are facing on the topic.

On one side, the party’s base is demanding impeachment, and many Democrats are desperate to show they’re doing something to stand up to Trump. Some members of Congress may fervently believe it’s the right thing to do, while others may fear primary challenges from the left.

But party leaders like Speaker Nancy Pelosi remain firmly opposed to an impeachment push. Calculating that there’s zero chance the Senate will remove Trump from office, they fear a controversial impeachment vote would harm Democrats representing districts that voted for Trump — something that could put the House majority at risk.

Why are Democrats split over impeachment?

At the broadest level, the debate over impeachment is quite simple.

One side’s position is, essentially: “We think Trump has done terrible things, so he should be impeached.” (The specific terrible things cited vary, but they include obstruction of justice episodes described in the Mueller report, bigotry, corruption, policies like family separation, and general unfitness of character.)

The other side’s position: “Impeachment will fail to remove Trump, so it’s pointless, and it also could hurt Democrats politically.”

Of course, that’s a generalization. There are some who have made the case that an impeachment push could help Democrats politically (or at least that the potential harms are overstated). There are likely also Democrats who just disagree on the substance, thinking it hasn’t been shown that Trump committed crimes worthy of impeachment. Still, these are the most common arguments on each side.

But another way to understand the divide is as a question of differing interests, among Democrats who serve different constituencies.

Support for impeachment is strongest among Democrats whose constituencies or audiences are very anti-Trump. This includes members of Congress representing left-leaning districts, as well as Democratic presidential candidates trying to woo progressive voters, and activists and pro-Democratic media like the popular Pod Save America podcast. Their very politically engaged constituencies or audiences badly want House Democrats to do more to try and fight back against Trump.

Democratic leaders like Pelosi, though, have a different goal — to protect their House of Representatives majority. To do that, they have to win again with a map that’s biased in Republicans’ favor. So they are looking out for the interests of Democrats in districts Trump won — to shield them from a hugely controversial impeachment vote that they fear could imperil their reelections.

Pelosi herself has offered a confusing, shifting series of arguments for why impeachment would be a mistake (Rolling Stone’s Ryan Bort counted seven). But at the heart of it all is likely her concern for her majority.

So on the whole, the debate resembles previous congressional controversies in which a party’s base makes impassioned demands for action, for a fight, and for doing what they feel is right — to the discomfort of party leadership, who believe it will achieve nothing and backfire politically.

What are the politics of impeachment, exactly?

Nationally, impeachment usually doesn’t poll very well, though there’s some variation depending on how exactly the question is framed and what options are offered.

• A late June Washington Post-ABC News poll found that only 37 percent of US adults supported beginning impeachment proceedings while 59 percent opposed it. (Political independents opposed impeachment proceedings by nearly a two-to-one margin.)
• A July NBC News / Wall Street Journal poll found that just 21 percent of registered voters think “there is enough evidence for Congress to begin impeachment hearings now,” while another 27 percent says Congress should look for more evidence before doing that, and 50 percent opposes any impeachment hearings.
• However, there was also a June Fox News poll finding that 43 percent of registered voters said Trump “should be impeached and removed from office,” and 7 percent said he should be impeached but not removed. When added together, that’s 50 percent support for impeachment.

Overall, it’s clear that at the very least that impeachment would start off as highly polarizing on a national level if not outright unpopular.

Yet an impeachment effort would be a months-long process that would go through several different phases with different political implications.

1) House impeachment hearings: The Judiciary Committee would likely hold high-profile hearings about allegations against Trump and draft potential articles of impeachment. And the question of which party would benefit most from this isn’t so clear. Perhaps swing voters might decide Democrats are extreme for pushing impeachment, or Trump voters might be energized in defense of their man.

But numerous high-profile hearings about allegations of Trump’s criminality could also damage Trump for obvious reasons — they’d be the biggest story in the country. Then again, opinion about Trump has been remarkably stable, and it doesn’t seem that House Democrats’ hearings have particularly hurt the president so far this year.

2) A House impeachment vote: The politics of a vote of the whole House on impeachment, however, seem unambiguously bad for Democrats — because of the chamber’s math and map.

• A party needs 218 seats for a House majority.
• Though Trump lost the nationwide popular vote to Clinton by two percentage points in 2016, he won 228 House districts to her 207.
• Currently, there are 235 Democrats in the House, and 31 of those Democrats represent districts Trump won.

An impeachment vote is a major problem for Democrats in those Trump districts. If they vote no, they’ll disappoint or infuriate loyal Democrats who are their core supporters and volunteers. If they vote yes, they’ll be hammered by Republicans (it could well be a major feature of negative ads here in 2020).

For the most part, they’d probably prefer not to take a position one way or the other — particularly when Trump’s acquittal in the Senate seems certain anyway. And a good speaker of the House generally tries to avoid forcing her caucus’s vulnerable members to take a highly controversial vote that won’t actually achieve anything.

House Republicans wouldn’t face the same problem. Not only is the House map slanted in their party’s favor, but the 2018 midterms wiped out almost every Republican representing a district Clinton won. So unless there’s some truly stunning new revelation, expect the whole House GOP to stand by Trump.

3) A Senate trial: If impeachment articles do manage to pass through the House, the next stop is a trial before the Republican-controlled Senate. The details of how that would play out aren’t yet clear. Would Mitch McConnell even hold the trial, or just ignore it like he did Merrick Garland’s Supreme Court nomination? Would he hold the trial but turn it into a farce?

4) The Senate acquits: Should a trial happen, though, the ultimate outcome seems clear. It takes a two-thirds Senate majority to convict an impeached president and remove him from office. Since there are only 47 Senate Democrats, that means 20 Republican senators would have to vote to oust the president who remains overwhelmingly popular among their party’s voters.

Unless some opinion-shaking bombshell unlike anything we’ve yet seen in Trump’s presidency is found, this will not happen. So at the end of the road is defeat and disappointment for Democrats, and Trump still in office.

Whether Trump and his 2020 reelection effort would be helped or hurt by the whole months-long spectacle isn’t clear. (The conventional wisdom is that Bill Clinton’s impeachment and acquittal was bad for Republicans. But George W. Bush ended up winning the next presidential election anyway, and Trump’s alleged offenses are different.)

Still, looking at this whole likely sequence of events, it’s evident why Democratic leaders don’t want to go down this road — it doesn’t end anywhere good, and the politicians put in the toughest spots are House Democrats in districts Trump won.

So where are things right now?

Impeachment supporters argue that we’re in the midst of a major political crisis, that Trump has committed serious crimes and is getting away with them, that responding with politics as usual is the wrong approach, that Pelosi is myopically focusing on reelection rather than what’s right, that depending on how impeachment plays out it could actually end up hurting Trump’s reelection chances — and that it’s simply the right thing to do.

“Dr. King reminds us that there are times that you have to do that which is neither safe nor politic nor popular,” Rep. Al Green (D-TX) told Politico.

“If Pelosi treats Trump as an aberration and continues to be passive in the hopes that we can all power through until next November, there’s no accountability mechanism built into our system of democracy that has any real credibility,” Elizabeth Spiers writes in the New Republic.

Impeachment’s Democratic critics, meanwhile, think a focus on reelection isn’t myopic, but rather quite important — if Republicans retake the House, a new Democratic president would be seriously hemmed in, and a second-term Trump would be able to try to pass sweeping new conservative legislation. They view impeachment as a performative stunt that would achieve nothing beyond the (temporary) emotional validation of Democrats’ anti-Trump base.

“He’s just not worth it,” Pelosi has said.

With this divide persisting, the position many in the party have gravitated toward is supporting an “impeachment inquiry.”

That does not commit to a vote in favor of impeachment. It sounds like an open-minded effort to get more facts. And also, when Democrats’ supporters ask them why they aren’t supporting impeachment, it’s something they can say in response: “I support an impeachment inquiry.” 100 out of 235 House Democrats have now taken this position, per CNN.

Still, once an inquiry gets started, it will likely be difficult to bottle up. Pressure will then build for the next step: the actual vote that Democratic leaders so dread.

That’s why, despite Pelosi’s insistence on Friday that she’s “not trying to run out the clock” on impeachment, many believe she’s doing just that — avoiding it and putting it off until it’s too late to make it happen before the 2020 elections.

But the pressure may eventually prove too much for Pelosi to withstand. Take, for instance, Nadler’s shift. When the year began, he frequently insisted he would only proceed to impeachment if he thought he could win over some Trump supporters. But he’s been hammered by activists and constituents for months — as well as a primary challenger.

Nadler still isn’t publicly calling for an impeachment inquiry, but he’s reportedly pressing for one in private. It is, after all, what his voters want.

With former special counsel Robert Mueller’s testimony having come and gone, the White House is trying to turn the tables on President Donald Trump’s political opponents by portraying the years-long Russia investigation as a total waste of time. Toward that end, administration spokesperson Hogan Gidley went on Fox News on Friday and proclaimed it now has been “proven” that Russian interference had no impact on the outcome of the 2016 election.

“It’s now been proven by every metric that Russian meddling had no effect on the outcome of the election,” Gidley said.

This talking point is meant to help Americans rest easy, secure in their knowledge that Trump didn’t wind up in office in part because of crimes committed by a foreign adversary on his behalf. But there’s just one problem: It’s not true.

Hogan’s position would be defensible if you interpret what he’s saying very narrowly: Neither Mueller’s 22-month investigation nor the Senate Intelligence Committee’s own two-year inquiry found any evidence that any votes Americans cast on Election Day were changed by Russian hackers.

But the reality of Russia’s sophisticated, multi-pronged effort to meddle in the US election went much deeper than probing states’ voting infrastructure, and the circumstantial case it had an impact is overwhelming. What remains in question is just how decisive that impact was.

Russia helped Trump, even though it didn’t change votes for him

First, it’s worth remember what Russia was up to: Contrary to what the conspiracy theory Sean Hannity and Trump pushed during their latest Fox News gabfest on Thursday, Moscow was not attempting to aid Hillary Clinton’s campaign.

Both US intelligence agencies and Mueller’s investigation affirmed that Russian hackers wanted to help Trump win office and committed crimes toward that end. Trump even acknowledged it at one point.

A Senate Intelligence Committee report released on Thursday indicated that beyond hacking and propaganda campaigns, the Kremlin’s efforts included attempts to penetrate elections systems in all 50 states. It also affirms there is no evidence Russian hackers messed with vote totals or were able to change votes. So in that very narrow sense, the claim that Russia didn’t affect the outcome of the election is defensible.

Gidley, however, went much further.

Votes not being changed doesn’t mean votes weren’t affected

To determine whether Russia had an effect on the 2016 presidential election (as Gidley ruled out), you wouldn’t just need to look at Moscow’s attempted penetration of voting systems. You’d also need to consider how the Kremlin’s $1.25-million-a-month online propaganda campaign and WikiLeaks’ publication of emails hacked from the Democratic National Committee and Clinton campaign played in American voters’ minds on Election Day. While there’s no incontrovertible proof that those efforts swayed voters in Trump’s favor, there is a convincing circumstantial case.

Earlier this month, for instance, a University of Tennessee Knoxville study funded by the Defense Department found that Trump’s polling upticks during the 2016 campaign correlated with social media activity by Russian trolls and bots. According to the study, every 25,000 retweets from troll and bot accounts connected with Russia’s Internet Research Agency predicted a 1 percent bump in Trump’s polling.

Damian Ruck, the study’s lead researcher, told NBC’s Ken Dilanian that his findings indicate Russia played a very key role in Trump’s victory:

In an interview with NBC News, Ruck said the research suggests that Russian trolls helped shift U.S public opinion in Trump’s favor. As to whether it affected the outcome of the election: “The answer is that we still don’t know, but we can’t rule it out.”

Given that the election turned on 75,000 votes in Michigan, Wisconsin and Pennsylvania, “it is a prospect that should be taken seriously,” Ruck wrote, adding that more study was needed in those swing states.

He points out that 13 percent of voters didn’t make their final choice until the last week before the election.

There is also a strong argument to be made that WikiLeaks, which published the first tranche of emails purloined from Clinton campaign chair John Podesta by Russian hackers just hours after the Washington Post published the Access Hollywood tape of Trump on October 7, swayed voters during the final month of the campaign. In this period, Trump overcame a string of sexual misconduct allegations and a 7-point deficit in the polls to win the election.

As Harry Enten noted for FiveThirtyEight in an analysis of WikiLeaks’ impact during the campaign’s closing stretch, the case remains circumstantial, but Americans were definitely paying attention to WikiLeaks. Enten found that for much of October, there was almost twice as much search interest in WikiLeaks than there was in the FBI, which was also in the news that month because of a letter then-Director James Comey sent to Congress publicizing the Clinton email investigation. Here are a couple additional important data points from Enten’s piece:

Trump, for instance, won among voters who decided who to vote for in October 51 percent to 37 percent, according to national exit polls. That’s Trump’s best time period. He carried voters who decided in the final week, when you might expect Comey’s letter to have had the largest impact, 45 percent to 42 percent.

It’s worth remembering that Trump’s closing message centered largely around WikiLeaks. He mentioned Julian Assange’s operation about five days a day during the campaign’s final month, but now pretends that never happened. (“Problematic is an understatement,” Mueller said on Wednesday about Trump’s promotion of WikiLeaks.) Is it possible the Clinton campaign email dumps and Trump’s relentless hyping of them on the campaign trail had no impact on the outcome of the election? It seems exceedingly unlikely.

So while it’s impossible to prove that Russian propaganda and/or WikiLeaks played a decisive role in swaying anyone’s vote, that’s not what Gidley said. He claimed it’s “been proven by every metric that Russian meddling had no affect on the outcome of the election.” Not only is that not true, but it’s a particularly brazen piece of gaslighting.

The Trump administration has self-interested reasons for downplaying Russian meddling

The president and his Republican enablers in Congress have a vested interest in the narrative that Russia didn’t play a role in Trump’s election, as disconnected from reality as it may be. But the impact of this narrative goes beyond confusing Fox News viewers — it also provides Senate Majority Leader Mitch McConnell with a pretext to block a number of bills meant to prevent foreign meddling from happening again during the very same week Mueller warned the country that it’s an ongoing concern.

It’s worth noting that Trump’s denialism about Russian interference goes even further than Gidley’s. During a press availability shortly after Mueller’s testimony, Trump not only indicated he’s not taking the Russia issue seriously, but went further and dismissed the entire investigation into election interference as based on “a fake set of facts that the Democrats used, and others, to try and do really an illegal overthrow, but we’re going find out about that.”

Then again, perhaps expecting Trump to take Russian interference seriously is too much to ask, considering that a full accounting of what happened in 2016 could potentially undermine Americans’ sense of security that their president beat Hillary Clinton fair and square.

The news moves fast. To stay updated, follow Aaron Rupar on Twitter, and read more of Vox’s policy and politics coverage.