When the history of President Trump’s failing effort to overturn the 2020 election results is written, there will be many bogus claims to chew on. But more than any other, one is emerging as the epitome of bad faith, poor research and our inability to rid our political discourse of a patent falsehood.

For more than a month now, Trump and his allies have been peddling a seemingly devastating anomaly in the election results: that Joe Biden underperformed Hillary Clinton’s 2016 showing in every metro area in the country except four — Atlanta, Detroit, Milwaukee and Philadelphia. The fact that those four cities just happen to be in the decisive states could sure lead one to draw some conclusions.

And it has. Over and over again, the claim has been broadcast by Trump, his aides and his media allies. Some have held it out there as merely suspicious; others, like Trump, have flatly called it evidence of fraud.

The only problem is that it’s demonstrably ridiculous. The claim has been fact-checked repeatedly — as far back as mid-November — by the conservative National Review, USA Today, this blog and the New York Daily News. Fox News political analyst Karl Rove even debunked it on air. Purveyors of it have on multiple occasions issued corrections. But those corrections were often issued long after their claims took root. And over the past week in particular, the claim has metastasized in some right-wing media outlets and Trumpworld.

First, let’s deal with the veracity of the claim. The National Review’s Dan McLaughlin had the most authoritative and extensive fact check. He found that not only did Biden not underperform Clinton in every metro area except the four above, but that he actually also overperformed her in 31 of 36 urban counties. What’s more, one of those five areas in which he actually did worse than Clinton? None other than Philadelphia. And Detroit and Milwaukee were on the low end of his improvements on Clinton’s margins. If anything, the data show Biden did worse in these supposedly decisive metro areas than you’d expect.

That NRO fact check came Nov. 16 — more than three weeks ago. At the time, this claim had been floating around in far corners of the Internet. But despite the claim being roundly disproved, it has survived and increasingly thrived, thanks to Trump’s disinformation campaign and increasingly credulous conservative allies seeking to bolster his allegations of fraud.

It all began Nov. 5, when a data journalist named Richard Baris claimed, “Joe Biden underperformed Hillary Clinton in every major metro area around the country, save for Milwaukee, Detroit, Atlanta and Philadelphia.” The votes were still being counted in many states, but the claim was quickly promoted by the president’s son Eric Trump, who labeled it “Shocking news.”

On Nov. 8, a little-known D.C. think tank called the Democracy Institute promoted Baris’s claim. Director Patrick Basham in Britain’s Sunday Express quoted from the tweet, apparently without checking the data.

(The Sunday Express article no longer appears on the Web, but the text appears on the Democracy Institute’s website.)

By Nov. 14, Townhall’s Matt Vespa cited Basham and, to his credit, appeared to attempt to actually check some of the numbers. He was satisfied enough to write a piece titled, “Democracy Institute Pollster: Yes, the 2020 Election Looks Like it Was Stolen.” But five days later, on Nov. 19, Townhall was forced to correct the piece and change the headline, according to captures of the piece from the Internet Archive. It acknowledged in an editor’s note, “[T]his definite claim turns out to be anything but accurate.”

Before that correction was issued, though, the cat was out of the bag. The same day, Nov. 14, the conspiracy-laden, pro-Trump website Gateway Pundit cited Townhall, citing Basham. By Nov. 15, the conservative, pro-Trump pollster Rasmussen Reports then tweeted out the Gateway Pundit piece, with the tweet earning more than 33,000 likes. By Nov. 16, the conservative Family Research Council’s founder, Tony Perkins, cited Basham’s claim on the FRC website.

Things began to fall apart that day — or so it would seem. McLaughlin’s extensive National Review debunk landed just after 4 p.m. By that evening, Fox News prime-time host Martha MacCallum tossed out Townhall’s claim — only to be shot down by none other than Rove, the former George W. Bush adviser:

MACCALLUM: [Townhall] says … Biden underperformed Hillary Clinton in every major metro around the country, save for Milwaukee, Detroit, Atlanta and Philadelphia. Do you find any of that odd, or what is the explanation for that, in your opinion?
ROVE: Well, I’m not certain it’s necessarily accurate, particularly when it comes to deception in the major metro areas.

Rove then quickly produced data from many metro areas undermining the claim, adding that there was “not a big differential between them and if you look at across the country.”

But not everyone was so interested in actually checking the numbers as Rove was.

After yet another fact check from USA Today, on Nov. 20, the Federalist on Nov. 23 breathed a whole new degree of life into the claim. In a piece titled “5 More Ways Joe Biden Magically Outperformed Election Norms,” J.B. Shurk again cited Basham’s quote as one of his five examples, without delving into the data.

Like Townhall, the Federalist was later forced to correct its piece. But it did so a full nine days after publication. And in the meantime, the bogus claim spread even further.

On Nov. 27, Fox Business Network host Lou Dobbs quoted that portion of the Federalist’s piece and urged Trump to “Never Concede.” It got 15,000 likes.

It soon spread to other conservative hosts. Fox News’s Sean Hannity cited the Federalist’s piece on Nov. 30, saying, “Get this: Joe Biden actually underperformed Clinton in major liberal cities across America, including New York, Chicago and L.A. But he outperformed Clinton in swing-state cities Milwaukee, Atlanta, Detroit, Philly. How convenient!”

The same day, former Fox host Bill O’Reilly also cited the piece on his show and offered similar sarcasm. “That’s pretty — wow! Is that coincidence or what!?

Not to be outdone, the pro-Trump outlet One America News did a full piece on the Federalist article that included the debunked claim. On Dec. 1, Trump promoted the clip, featuring White House correspondent Jenn Pellegrino.

That night, White House press secretary Kayleigh McEnany appeared on Hannity’s show and, rather than citing someone else, stated the claim as fact.

“All across the country, we hear stories like this,” she said, “but most specifically in those four cities where Joe Biden somehow overperformed Hillary [Clinton], though he underperformed her in every other metro.” She specifically listed the exceptions to this supposed trend as the four cities noted above.

I fact-checked McEnany on Dec. 2, which is the same day the Federalist withdrew the claim from its piece. Despite the headline still reading “5 More Ways,” the piece is now without a No. 3 — just 1, 2, and then 4 and 5.

The New York Daily News added its own fact check on Dec. 5 — the fifth of its kind, if you include Rove — but that wasn’t stopping the Trump team.

That evening, pro-Trump Fox host Jeanine Pirro ranted, “How is it that Biden underperformed Clinton in New York, Chicago and L.A. but won in the swing-state cities of Milwaukee, Atlanta, Detroit and Philadelphia, each known historically for voter corruption?”

Trump attorney Rudolph W. Giuliani alluded to the claim on Maria Bartiromo’s Sunday Fox show the next morning. The same day, fellow Fox weekend host Mark Levin featured Basham, who claimed Biden had “done very, very poorly in most of the country, except where [it] absolutely mattered.”

Not to be outdone, the increasingly prominent and credulous airwaves of Newsmax featured the claim the same day, courtesy of host Benny Johnson.

“Biden underperformed Hillary Clinton everywhere in the country, except in a few key counties,” Johnson claimed. “Pollster Richard Baris and analyst Robert Barnes have noted that Joe Biden underperformed Hillary Clinton in every major metro area around the country, except four: Milwaukee, Detroit, Atlanta and Philadelphia. How magical!”

It would be magical if it were anywhere approaching being true.

Trump, again, promoted the clip. And on Monday, Trump campaign senior adviser Steve Cortes attempted to rehabilitate the claim yet again.

“Are we really to believe that Joe Biden threaded the needle to the degree needed to only outperform in precisely the places where he needed to outperform?” Cortes asked rhetorically in a video posted to Twitter. “Or does that improbability speak to fraud and malfeasance. I firmly believe it’s the latter.”

The data is clear, though, and there is no improbability. If anything, the data shows Trump’s performances in the key metro areas lagged behind most others.

To this point, the claim has been debunked not just by mainstream outlets, but also by Rove and two of those who played a major role in initially spreading it. None of that has deterred those who want to believe it — or at least would very much like you to believe it — from pressing forward. But that they have been so willing to propagate this bogus and roundly rejected claim is certainly a commentary on their inability or lack of desire to actually parse the various voter fraud claims that are out there.