You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.

Why the Pundits Got Impeachment Wrong

Live by the polls, die by the polls.

Nate Silver (Slaven Vlasic/Getty Images for AWXII)

For much of this year, House Speaker Nancy Pelosi used Abraham Lincoln as a human shield. Pressed by the media on calls to impeach President Trump, she regularly quoted Lincoln, who said three years before becoming president: “Public sentiment is everything. With public sentiment, nothing can fail. Without it, nothing can succeed.”

It was a convenient mantra, precisely what you’d expect a politician to say if she agreed with the prevailing public sentiment. For months, that sentiment was clear: Opinion polls consistently showed that a majority of Americans did not think Trump should be impeached, versus about 35 percent who felt the opposite. So Pelosi urged patience and “process”—that is, when she wasn’t arguing that impeachment was “so divisive,” Trump is “just not worth it,” or, oxymoronically, that he was “becoming self-impeachable.”

Pelosi was often portrayed as the only person standing in the way of impeachment, but she was hardly alone. A majority of her Democratic caucus stood by her, refusing to publicly endorse impeachment, and prominent members of the political media—across the ideological spectrum—wrote articles endorsing her rationale. Look at the polls, they argued.

But now, in the wake of the Ukraine scandal, the polls have swung suddenly and drastically: A majority of Americans consistently support Trump’s impeachment. It turns out that public opinion is not immutable, but subject to political persuasion and unforeseen events. It’s vital to remember this episode, because the next time these pundits resolve to dismiss the next big idea on the left, rest assured they will cite polling to make their case.


Nate Silver, the statistician who founded FiveThirtyEight, was especially insistent that impeachment was a terrible idea. When special counsel Robert Mueller’s testimony to Congress in July failed to move the needle on support for impeachment, he tweeted, “Not a great poll for the ‘public opinion on impeachment will change once the public hears more about the evidence’ crowd.” In late August, he dismissed commentators on the left who argued that Pelosi and other reluctant House Democrats could shift public opinion by supporting impeachment, writing that “despite the insistence that public support for impeachment would go up if more Democrats came out in favor of it, a whole bunch of Democrats have come out in favor of impeachment and public support has ... decreased.”

As recently as September 12, just days before news broke of the whistleblower’s complaint about Trump’s call with Ukrainian President Volodymyr Zelenskiy, he was getting in flame wars like this:

Many others echoed these takes, using polls to build a narrative about what voters did and did not want. Writing in Slate in May, William Saletan argued that Democrats should do everything to hold the president accountable except impeach him. “Something about the word impeachment—maybe the impression that it reflects an agenda rather than an open-minded assessment of facts—turns people off,” Saletan concluded, citing polls showing that support for impeachment never broke 40 percent. Instead, he argued, Democrats should focus on investigations and the 2020 election.

Writing the same month in RealClearPolitics, Bill Scher made a similar case, arguing that an impeachment investigation would be a political disaster for Democrats. “Impeachment doesn’t just poll poorly; it polls poorly even though most Americans already believe Trump has committed crimes,” he wrote. “Clearly, there is a bloc of voters who believe Trump committed crimes yet blanch at impeachment-and-conviction. It’s an assumption without evidence that an impeachment inquiry will produce evidence of crimes not yet already presumed by most voters, and yet another assumption that any such evidence would move the public opinion needle.”

The New York Times’s David Brooks echoed Saletan and Scher in late September, after it became clear that Trump had asked Zelenskiy to aid his re-election effort. Brooks acknowledged in his opening sentence that “Trump committed an impeachable offense on that call with the Ukrainian president.” But impeachment was “elitist,” he claimed. “This is not what the country wants to talk about,” wrote the flyover-man expert. “Pelosi said she would not proceed with impeachment unless there was a bipartisan groundswell of support. There is no bipartisan groundswell, and yet she’s proceeding. According to a Quinnipiac University poll, only 37 percent of Americans support impeachment.”

Some of these pundits have since seen the light. “Four months ago in this space, I was a skeptic of the politics of impeachment, as polls showed a clear majority in opposition, even though a majority believed Trump was a criminal,” Scher wrote late last month. “And it may be that the short-term politics are still bad. But a documented attempt by the president to subvert democracy not only cannot be ignored, it must be challenged.” Saletan argued that Trump should be impeached for an “overwhelming pattern of treachery,” citing seven countries in which Trump “pursued personal advantage at the expense of the United States.” And Silver has acknowledged the shift in public opinion. “There is already polling showing impeachment becoming more popular… and while things are fluid, I buy that the shift is real,” he said in a video posted on September 29.

There were good reasons, it must be noted, to be skeptical. Impeachment polling had barely budged all year; major new developments, like the completion of Mueller’s report and his congressional testimony, caused only minor, temporary shifts. Even many proponents of impeachment acknowledged that the issue polled badly and might be politically damaging, and instead argued that it was a moral and legal necessity. But those poll numbers were treated by many as gospel. Voters could not be swayed. Impeachment was unpopular now and would likely be unpopular forever with half of the country, a consequence of our polarized political moment. The few available precedents, moreover, were inconclusive at best. Bill Clinton grew more popular during the impeachment investigation. While Watergate showed that public opinion can shift on impeachment, it took major developments like the Saturday Night Massacre and the revelations in the White House tapes; plus, partisan loyalties were less hardened in the early ’70s than they are today.

Staking a claim against impeachment also seemed like a safe bet, and confident prognostication is highly valued, if not outright required, in today’s political media. As my colleague Osita Nwanevu wrote recently, “Speculation about what voters might think or might decide about the candidates and how they might fare in the elections in question still seems to occupy much more space in our papers and take up much more time in television news than straightforward coverage of what we firmly know about them—their personal biographies, their political backgrounds, and their platforms. Campaign reporters spend more time telling voters what voters think than they do relaying facts … voters might not know about.” This is yet more true of pundits, who tend to trust their political gut more than the entire sweep of contemporary and historical evidence.

It’s too early to confidently state why polling has moved in favor of impeachment so dramatically, but one possibility is that Pelosi’s decision to endorse an inquiry has made a substantial difference. As Brian Beutler wrote before the inquiry was announced, Pelosi had previously purposefully kept her caucus divided on the issue. While impossible to definitively prove, Pelosi’s decision to launch an “official impeachment inquiry” appears to have been significant in terms of moving polling opinion. Before Pelosi opened the inquiry, 37 to 41 percent of voters consistently supported impeachment. Last week, polling showed that nearly 60 percent approved of Pelosi’s decision to open an inquiry. Polling released on Wednesday showed that 52 percent supported impeachment and removal. It turns out that persuasion in politics is still possible. Leaders are not shackled to public opinion and have the power to change people’s minds.

This realization has broader implications for political discourse in the coming months and years. The polling is mixed at best on bold policy objectives like the Green New Deal and Medicare for All. On the Green New Deal, as with much early impeachment polling, Republicans are united and Democrats are divided. The polling on Medicare for All, meanwhile, suggests that it has a small plurality of support, but is declining in popularity. That said, the data itself is murky. But this should not necessarily dissuade either Democrats or pundits: Voters are more than capable of changing their minds, and that seems to be especially true when party leadership endorses ambitious or seemingly risky initiatives. Public sentiment may be everything, but it is not some esoteric force that’s immune to outside influence and understood by a select few. If the impeachment question has shown us anything, the more someone claims to know what Americans think, the less likely they really do.