Countering (Massive Numbers of) Lies Doesn’t Work

Lies are dangerous in a number of ways. Putting aside that there are lots of situations where a false belief leads to very bad action (e.g., believing homeopathy is an effective cancer treatment leads to forgoing treatment that would have saved one’s life or mitigated suffering). They are also dangerous because people with bad agendas tend to resort to lying because they can’t win on the merits. And they don’t just resort to a bit of deception, or even clever deception. It turns out that wholesale, massive, shameless, easily rebutted lies are pretty effective, at least for something.

Consider the decades long attack on the EU:

But Britain has a long and well-observed tradition of fabricating facts about Europe—so much so that the European Commission (EC) set up a website to debunk these lies in the early 1990s. Try our interactive quiz below and see if you can spot the myths.

Since then the EC has responded to over 400 myths published by the British media. These range from the absurd (fishing boats will be forced to carry condoms) to the ridiculous (zippers on trousers will be banned). Some are seemingly the result of wilful misunderstandings.

Sadly, for all the commission’s hard work, it is unlikely to be heard. The average rebuttal is read about 1,000 times. The Daily Mail’s website, by contrast, garners 225m visitors each month.

And, of course, the Leave campaign, itself, was almost wholly lie based. Remain made some (economic) predictions that were falsified (and that needs to be understood), but it didn’t traffic in wholesale lies, to my knowledge.

Similarly, we have a decades long campaign, almost entirely easily-debunked-lie based, against Hillary Clinton. Just take claims about her honesty (esp. next to Trump). Robert Mann produced a very interesting graph of Polifact’s fact checking of a selection of politicians:

It isn’t even close! HRC is one of the most honest politicians (in terms of telling falsehoods) and Trump is one of the most dishonest.

Yet, when I was debating folks on Democratic leaning blogs, I had people saying that Clinton was a pathological liar. When presented with this chart, they stuck to their guns. (Note, they didn’t think Obama was a liar.)

You can quibble with the methodology (see Mann’s blog post for a discussion), but Polifact’s fact checker tries to be evenhanded. One should be at least a little struck by this evidence.

But correction often just doesn’t work, backfires, or isn’t effective in changing attitudes and behavior. For example,

Facts don’t necessarily have the power to change our minds. In fact, quite the opposite. In a series of studies in 2005 and 2006, researchers at the University of Michigan found that when misinformed people, particularly political partisans, were exposed to corrected facts in news stories, they rarely changed their minds. In fact, they often became even more strongly set in their beliefs. Facts, they found, were not curing misinformation. Like an underpowered antibiotic, facts could actually make misinformation even stronger.

Or consider Emily Thorson’s concept of belief echoes:

However, through a series of experiments, I find that exposure to a piece of negative political information persists in shaping attitudes even after the information has been successfully discredited. A correction–even when it is fully believed–does not eliminate the effects of misinformation on attitudes. These lingering attitudinal effects,which I call “belief echoes,” are created even when the misinformation is corrected immediately, arguably the gold standard of journalistic fact-checking.

Belief echoes can be affective or cognitive. Affective belief echoes are created through a largely unconscious process in which a piece of negative information has a stronger impact on evaluations than does its correction. Cognitive belief echoes, on the other hand, are created through a conscious cognitive process during which a person recognizes that a particular negative claim about a candidate is false, but reasons that its presence increases the likelihood of other negative information being true. Experimental results suggest that while affective belief echoes are created across party lines, cognitive belief echoes are more likely when a piece of misinformation reinforces a person’s pre-existing political views

We see this in the various formulations of the Clinton Rules.

One major harm of such mechanisms is that it opens up a line of defense for very bad people, e.g., Trump, to wit, that there are “Trump rules” and the bad things pointed out about him are fake. They aren’t, but why trust a gullible media about it?

I’ve had personal experience of this. I used to comment a lot on LGM. One commenter with a propensity for persistently saying very silly things (about, e.g., statistics, causality, politics, and even the law (they are a lawyer)) got to a point where they couldn’t stand my repeated refutations (including pointing out how they’d been refuted before). They embarked on a pretty systematic campaign to lie about me, primarily about my mental health and that I was “stalking” them, on the verge of a breakdown, that they were frightened of me, that I had no sex life or other kind of life, that I spent large period of times looking things up on them (stalking!), etc. These were transparent lies and obvious gaslighting. No one took them directly seriously, but they did have effects. People would see an exchange and assume that there was some fault on my part (however mild). This would pop up elsewhere, in other comments.  Some of these people were more sympathetic to a gaslighting liar than they had any right to be.

So, pretty exemplary behavior and a sterling reputation vs. transparent lies and extremely bizarre slanders and…well, I’m the one not commenting any more. It worked, in a way. (Trump winning had an effect too. It’s not solely due to this bad behavior.)

Given sufficient shamelessness and no structural counter (e.g., moderation) and no big effort on my part (e.g., an active campaign), there’s little penalty for such lying and it advances their noxious cause.

These examples can be multiplied easily (anti-vaccine, pro-tobacco, climate change denial campaigns come to mind).

It’s very difficult to deal with. We need to.


How severe is the problem? I just saw a report on a survey using Trump’s and Obama’s inauguration crowd photos:

For the question about which image went with which inauguration, 41 percent of Trump supporters gave the wrong answer; that’s significantly more than the wrong answers given by 8 percent of Clinton voters and 21 percent of those who did not vote.

But what’s even more noteworthy is that 15 percent of people who voted for Trump told us that more people were in the image on the left — the photo from Trump’s inauguration — than the picture on the right. We got that answer from only 2 percent of Clinton voters and 3 percent of nonvoters.

The article discusses the idea of “expressive responding”:

Why would anyone give the wrong answer to a pretty simple question?

To many political psychologists, this exercise will be familiar. A growing body of research documents how fully Americans appear to hold biased positions about basic political facts. But scholars also debate whether partisans actually believe the misinformation and how many are knowingly giving the wrong answer to support their partisan team (a process called expressive responding).

Expressive responding is yet another form of lying with potentially far reaching consequences.


3 thoughts on “Countering (Massive Numbers of) Lies Doesn’t Work

  1. Hi Bijan!

    I have quite a lot to say about your post here, but let me start by the Brexit economics. Krugman was always pretty skeptical of the dire predictions of an immediate recession in the UK after Brexit. He always maintained that the bad economic effects would happen in the medium to long term, while UK’s productivity slowly erodes due to less trade opportunities. See, e.g this. I tend to trust Krugman on economics, his analysis has been spot on during the recession. I don’t really know who the economists predicting a huge Brexit fallout were, so I don’t know their track record either, and if they did seriously engage with the evidence or just went with their gut.

    The second thing I wanted to say : whenever there’s a study showing evidence of some sort of widespread human cognitive bias, I always try to have a moment or two of introspection, and try to identify where this specific bias is affecting me. (I mean, political junkies aren’t immune to those biases, I think a big reason I never fell for the Clinton rules and the Leave campaign’s lies is because I was already predisposed to disbelieve those lies.) And I did find a few examples where I believed a specific lie for a long time : I used to think the ETA was responsible for the Madrid metro bombs, I used to believe that centre-right governments in various countries were more fiscally responsible than centre-left governments. I eventually got rid of those after a few years, but I wonder what kind of vague untrue thing is tainting my worldview right now.

    So you have a very interesting discussion of why it’s hopeless to convince the general public that a lie is a lie, but what about the more modest challenges of trying to mitigate this bias in ourselves*? Get more information? But there are a lot of fake news out there. Vet the information channel before believing it? There’s always the risk of enclosing ourselves into an echo chamber…

    *ourselves = people who are aware of this bias.

    Wow, that’s a huge comment. I should probably get my own blog 😛

    • “I don’t really know who the economists predicting a huge Brexit fallout were, so I don’t know their track record either, and if they did seriously engage with the evidence or just went with their gut.”

      It would be nice to have a comprehensive retrospective. One thing I think could have been confounding is that Cameron didn’t trigger Art 50 right away (or on a short time frame).

      “So you have a very interesting discussion of why it’s hopeless to convince the general public that a lie is a lie, but what about the more modest challenges of trying to mitigate this bias in ourselves*?”

      Yes, this is a big challenge, and a bigger challenge when we thing about mitigating communal biases, and a yet bigger challenge when we still want our messaging to work and our appropriate qualifiers to not be used against us.

      But taking the simplest difficult problem, mitigating our own biases, I think it all comes down, in the end, to establishing verification and validation procedures and using them often.

      Now you are correct that this has become very difficult because of fake news, echo chambers, etc. which is why I try to ground out in observables or the scientific literature. But it’s possible to go too far in the other direction: I’m not going to question the equality of women etc.

      And you should get your own blog! Joe from Lowell did so and I think it was good for him.

  2. First of all, I hope your hand surgery went well! It was today, right?

    “One thing I think could have been confounding is that Cameron didn’t trigger Art 50 right away (or on a short time frame).”

    Good point. A lot of the catastrophic predictions supposed an immediate start to the Brexit negotiations. I can imagine the market panicking as the negotiations proceeded and that it would be clear that the UK would get a bad deal.

    Re: the biases. I was talking about that mostly because I’ve been involved in a leftist political organisation in my campus since I was an undergrad, and I’ve been slowly distancing myself from them. They work very well when it’s time to fight for student housing, against tuition rises, etc. But they have somewhat of a racism and sexism problem, and are refusing to face it. I was also unimpressed when they tried to paint their NIMBY opposition to development next to campus as a social justice issue. This made me question some of the mindset I had unconsciously acquired when hanging out with them. And it’s an awful feeling, what I imagine must be like when someone moves away from a religion they were raised in.

    (Note: I’ve not abandoned any of my social justice commitments, and I still think the organisation does good work. I was just taking this opportunity to think about the biases I had acquired because it was the common wisdom there, and how it’s emotionally hard to get rid of them.)

    And, uh, I’ve kind of started one blog. It’s in French, and there’s just the one post there. Maybe I’ll vent there when I feel like it.

Comments are closed.