Fake News and Junk Science: Stronger Together!

If you voted for President Trump last fall, David A. Graham at The Atlantic says he has “good news and bad news” about your mental capacities. The good news is, contrary to what other liberal journalists have been saying, you “aren’t impervious to reality.”

But don’t get cocky. The bad news is that you’re shallow and primitive, basing your political preferences on “image and tribal identity.”

Without explicitly saying so, Graham leaves the impression that Hillary Clinton enthusiasts like him are, by contrast, models of dispassionate scientific inquiry. Indeed, he claims his bad news about President Trump’s supporters comes straight from “a new political-science paper” whose main authors are “behind some of the most important work on the impact of corrections and fact-checking in recent years.”

Scientifically certifying the mental unfitness of your political opponents is, of course, the opening move in a very nasty game. The final move comes when the newly minted psychological rejects are consigned to maximum-security “psychiatric hospitals” for a long spell of “treatment.” All that’s required is abandoning the reactionary idea that hospitals should be more hospitable than gulags and, voilà, dissent can be hospitalized as well as criminalized. The most advanced progressive regimes realized this. The hospitalization of dissent was right up there with famine-induced genocide as a favorite technique of the USSR and Mao’s China. So, it’s alarming to hear that top tier progressive academics are boldly using their authority as scientists in ways that will marginalize and silence ordinary folk who happen not to share their views.

But, though a casual reading of their paper might lead one to accept this alarming take on its conclusions, a more careful reading shows that, contrary to Graham, the authors don’t explicitly (important word) draw any conclusions particular to the president’s supporters. The official (another important word) conclusion, though only tested on the latter, is supposed to generalize to everyone, whether they stood with the president, Secretary Clinton, or that weird little CIA agent from Utah who just won’t go away.

Here’s their thesis statement:

Are citizens willing to accept fact-checks of false or unsupported claims of candidates they support in the heat of a political campaign? Previous studies have reached conflicting conclusions about people’s willingness to update their factual beliefs in response to counter-attitudinal information. To discriminate between these findings, we conducted two experiments during the 2016 presidential campaign . . . These results suggest that corrective information can reduce misperceptions, but will often have minimal effects on candidate evaluations or vote choice.

So, officially, (ahem) there’s nothing at all here singling out Trump supporters. Indeed, the authors state that one limitation of their work is that they “did not correct a Clinton misstatement and cannot evaluate how her supporters would have reacted.” The point here is that their conclusions are false if it turns out they don’t generalize to Clinton’s supporters. So, not only has Graham mischaracterized the study’s official conclusion, his mischaracterization, if true, would falsify it!

At first glance, it looks like Graham either lied about or completely misinterpreted the paper. But a closer look reveals that the first glance might be too hard on Graham’s intellect and too easy on the researchers’ ethics.

Though the latter avoid framing their conclusions in terms that might recall the horrors of Soviet psychiatry, their paper is drenched in anti-Trump sentiment. They treat it as common knowledge that President Trump “was infamous for extreme exaggerations and misstatements.” His last name occurs 325 times in the paper and the first 17 references all occur in sentences that casually impute dissembling to him as if the subject were the Pope’s religion or a grizzly’s defecatory practices.

And, why didn’t they test their hypothesis on Clinton supporters? Not doing so certainly leaves the impression that there’s something uniquely peculiar about the president’s voters. And, anyone seriously concerned with a general hypothesis about human irrationality, rather than producing a hit piece disguised as research, surely would have done so.

So, a lot would be explained if the scholarly paper and the Atlantic article are a collaborative effort, the paper designed to lead to, without explicitly endorsing, the dangerous scientific marginalizing of the President’s supporters pushed by the article. I can’t say whether that’s right. But I can report that the paper’s principal author, Brendan Nyhan has been featured, and often effusively praised in, at least 119 Atlantic articles: that 13 of these articles are by Graham; that the latter was a Duke undergraduate in the same year the former earned his Ph.D. there; that Nyhan was featured in an article in the Duke Chronicle, (“Forum Talks Politics,” p. 6) for which Graham wrote almost 200 pieces and served as editor; and that Graham has a long a history of producing fake news, including as a major player in perpetuating the infamous Duke lacrosse gang-rape hoax.

So, even if Nyhan is that rare academic who doesn’t notice stories touting his research (which seems most unlikely), it’s very hard to believe that he was unaware of Graham’s gross misinterpretation of his results, and hard to avoid concluding that, in fact, the two colluded to produce it.

Wikileaks exposed the incestuous and underhanded collusion between the Democratic Party and its media allies. And, Graham’s paper looks like the result of collusion between the media and the third leg propping up progressivism’s broken ideas, academia. We are very likely, seeing more than just fake news feeding on junk science. It looks like we’re seeing their creators conspire to advance both their careers and a particularly nasty and dangerous progressive narrative.

And, make no mistake, the science in the paper is at least as junky as the news in the article is fake.

Graham says the authors,

looked at a pair of claims that Trump made while he was the GOP nominee for president in 2016. In one case, during the Republican National Convention in Cleveland, he claimed that crime was up sharply, which was false. In another, during the first presidential debate, he insisted that Ohio and Michigan were hemorrhaging jobs, when in fact both states had unemployment levels below the national average.

Graham’s piece has two functions. One is to reinforce the existing progressive narrative that President Trump dissembles at a rate heretofore unapproached in the annals of American politics. The second function is to water the seeds of the new narrative, implicit in the academic paper, that science deems the President’s supporters cognitively defective.

The narratives aren’t unrelated; the more widely taken for granted it is that the president is a pathological prevaricator, the easier it is to sell the idea that his followers, who don’t believe he is, have a screw loose. And, since the same lies and misrepresentations about the President fuel both narratives, we will have to discuss both. Fortunately, we can dispense with the old and comparatively anodyne one concerning the President’s veracity quickly.

Most of the President’s claims about crime rates in his convention speech were very specific and undisputed. (For example, “that killings have risen by 50 percent” in Washington, D.C.) The only part of the speech that resembles Graham’s attribution that crime is “up sharply” is: “Decades of progress made in bringing down crime are now being reversed.” Graham disputes this by linking to an earlier Atlantic article, in which he claimed crime hasn’t gone up since it’s at a much lower rate than 2006 and the early ’90s. But, the president explicitly spoke of reversing decades of progress. So, if anything, Graham’s comment would tend to support him. Graham goes on to admit that “the violent crime rate did tick up in 2015,” but gives us no figures. For good reason; it turns out that violent crime increased by 3.9 percent, including a 10.8 percent increase in murder and a 6.3 percent increase in rape. So, any propensity for prevaricating here is Graham’s, rather than President Trump’s.

Graham’s rebuttal of the president concerning job losses in Ohio and Michigan is an even more transparent failure. Graham’s claim that they have lower unemployment rates than the national average is a completely obvious piece of misdirection since it’s plainly irrelevant to whether the states themselves have lost jobs. Trump was speaking about a well-discussed problem—namely, manufacturers fleeing the United States for Mexico. Since NAFTA was implemented, Ohio and Michigan have lost one-third of their manufacturing jobs. So, again, it’s Graham who’s dissembling here, not Trump.

A Study Both Mistaken and Misleading
Aside from, (perhaps intentionally) misreporting the paper’s conclusion, Graham also misreports its take on the president’s claims. Graham has them baldly calling the president’s claims “untrue,” while the “political scientists” are more careful, making the much weaker charge that President Trump’s remarks are “misleading.” But raising the alarm about a 33 percent decline in manufacturing jobs wouldn’t be called “misleading” by any competent nonpartisan English speaker. And characterizing a 10.8 percent increase in murder plus a 6.3 percent increase in rape as an increase in crime is as far from being misleading as possible—it’s a tautology.

So, what we have is a study purporting to show people will accept corrections to misleading information, in which the corrections turn out to be mistaken and the initial information perfectly accurate! This alone makes it hard to know what to make of its conclusions, let alone how to understand its authors as competent scientists.

But their study has another limitation that makes it even more worthless and that the authors themselves acknowledge. They say:

[E]ach study evaluated a correction of a single misstatement. Correcting a series of inaccurate claims might have stronger effects on candidate evaluations.

Indeed. But why on earth would anyone expect the correction of a single inaccurate claim to have any effect? President Trump had been campaigning for over a year; he was a well-known celebrity for decades; he authored best-selling books; he produced and starred in a successful television show; and, of course, he had a much used and viewed Twitter feed. Is it even remotely interesting that getting his supporters to accept (falsely) that he’s made one misleading claim won’t cause them to withdraw their support?

Besides acknowledging this limitation, the study’s authors also acknowledge its seriousness. For they respond by drastically weakening their thesis and assert:

[O]ur results provide compelling evidence that citizens can accept factual corrections of misstatements even when they are made by one’s preferred candidate during a presidential election.

In the end, that is what they tout as their great contribution to human knowledge: people are sometimes able to accept that the person they back in a presidential race has misstated something. Has anyone ever been silly enough to deny this?

Well, as it turns out, yes: the two principal authors of the very paper we are discussing, Brendan Nyhan and Jason Reifler, denied it in an earlier unpublished paper! That paper is what they reference to give a false air of significance to their new one.

Spend a moment taking it in: these two guys wrote a paper whose only conceivable merit is that it rebuts some of their own unpublished and almost certainly worthless work. And because the new paper conveniently has two additional authors, the reader isn’t likely to notice that its relevance is all academic smoke and mirrors.

The ideas here are so silly and ill-defined that it’s hard to imagine the authors believe or have any real interest in them. At best, their “results” are junk science, awash in partisan lies, and designed to pad their publication records to attract promotions, raises, and more money to go on vacations, (which, in academic jargon, oddly enough, go by the name “conferences”). At worst, their work is all the above plus an example of collusion between academics and journalists designed to forward a repulsive and dangerous narrative that progressive slanders against President Trump’s supporters are backed by science.

Wikileaks confirmed our worst fears about collusion between the Democratic Party and its mainstream media boosters. If all President Trump’s accomplishes is to help expose this malfeasance and destroy the corporate media’s credibility, he’ll deserve our lasting gratitude. But, perhaps, when he’s finished swatting all the media flies buzzing around his head, he can have someone attend to the academic pests that lurk in the shadows and bite at his feet.


Get the news corporate media won't tell you.

Get caught up on today's must read stores!

By submitting your information, you agree to receive exclusive AG+ content, including special promotions, and agree to our Privacy Policy and Terms. By providing your phone number and checking the box to opt in, you are consenting to receive recurring SMS/MMS messages, including automated texts, to that number from my short code. Msg & data rates may apply. Reply HELP for help, STOP to end. SMS opt-in will not be sold, rented, or shared.