backfire effect

The backfire effect is a curious response many people have to evidence that conflicts with their beliefs: instead of becoming open to the possibility that the evidence might be correct and one might have to change one's mind, many people become more convinced that they were right in the first place. Yes, that's right. Some people's beliefs get stronger when evidence against their belief is presented to them. You would think that a rational person would base his beliefs on the strength of the evidence and that evidence against his belief should weaken rather than strengthen his belief, but there is a growing body of scientific evidence that has found most of us are not that rational when it comes to dealing with evidence that conflicts with beliefs we already hold.

Journalist David McRaney sums up the backfire effect nicely: "When your deepest convictions are challenged by contradictory evidence, your beliefs get stronger."

Political scientists Brendan Nyhan and Jason Reifler coined the term “backfire effect” to describe this irrational response of coming to hold one's original position even more strongly when confronted with evidence that conflicts with one's belief. For journalists and anyone engaging in debate and argument who hopes to persuade others and change their minds or correct misinformation, the backfire effect is more than annoying; it indicates our goals are pointless and we are guaranteed to fail. Why bother to provide evidence for global warming or point out the errors of the climate change deniers if the better our arguments are the more the deniers will dig in and be encouraged in their continued errors? Why bother explaining evolution to young Earth creationists and rebutting their inane arguments when the likely effect is to strengthen their erroneous beliefs? Why waste time explaining to anti-vaccinationists the benefits of vaccinations and the harm done by not vaccinating children when all we are likely to accomplish is to fuel their hostility toward the truth? It becomes a futile exercise to argue with people who believe Obama is a Muslim or wasn't born in Hawaii. Not only will no amount of evidence change their minds, but the more evidence we provide to show they're wrong, the stronger their conviction becomes that they're right.

Is there any hope, then, of debunking myths such as the birther myth that President Obama was not born in Hawaii, the creationism myth that a magical being created all species at once and there has been no evolution, or the anti-vaxxer myth that vaccinations are chock full of harmful substances that cause everything from mental retardation to autism to death and so should not be given to our children? According to John Cook of the Global Change Institute, University of Queensland, and Stephan Lewandowsky, School of Psychology, University of Western Australia, there is a way to debunk myths that shows some promise.

...an effective debunking requires three major elements. First, the refutation must focus on core facts rather than the myth to avoid the misinformation becoming more familiar. Second, any mention of a myth should be preceded by explicit warnings to notify the reader that the upcoming information is false. Finally, the refutation should include an alternative explanation that accounts for important qualities in the original misinformation. (The Debunking Handbook.)

Sounds simple, eh? Anyway, I think there is little dispute among those of us who frequently engage in public arguments that all we need to do is provide good information and we will change the hearts and minds of those we argue against. On some issues, the propaganda machinery is a Goliath that no army of independent Davids can hope to take down. For example, Donald Prothero [How We Know Global Warming is Real and Human Caused] reports that "the day that the 2007 IPCC report was released (Feb. 2, 2007), the British newspaper The Guardian reported that the conservative American Enterprise Institute (funded largely by oil companies and conservative think tanks) had offered $10,000 plus travel expenses to scientists who would write negatively about the IPCC report." The Intergovernmental Panel on Climate Change provides an abundance of solid information that journalists, government officials, and citizens alike can look to for guidance in what to believe about climate change. The propaganda machine of the self-interested oil and coal companies and the conservative think tanks and media outlets can easily outpace the flow and influence of scientific data, however.

There's an old saying "if you repeat something often enough it becomes true." Of course, repeating something has no effect on the truth-value of any claim, but familiarity with a claim does increase the chances of accepting it as true. If propaganda machines are good at anything it is at getting the same misinformation repeated again and again in various media outlets. Combating the misinformation by providing accurate information may have the undesired effect of strengthening the belief in the misinformation. So, what's a fellow to do? Sounds like one of those damned if you do, damned if you don't situations.

The fact is, though, that those of us who do battle with astrologers, birthers, anti-vaxxers, homeopaths, Holocaust deniers, evolution deniers, climate change deniers, and the like aren't really hoping to change the minds and hearts of those we confront. We're hoping that many of those who read our arguments, attend our debates, hear our presentations, or watch our videos are not fully committed to the beliefs of those we challenge. The hope is that among the bystanders, the audience, the viewers, and the readers there will be many who will be influenced by the information and arguments we provide. The goal of combating the tobacco companies' propaganda about smoking, for example, was not to change the minds of tobacco executives, but to provide good information that the general public, public health officials, and even politicians, might consider in making decisions about smoking and regulations governing the sale of tobacco products. The goal of debunking the myth that the world will end in 2012 as predicted by the Maya centuries ago is not to change the minds of those writing books or posting on websites promoting this ridiculous idea. The goal is to provide some counterpoints to the myth mongers that might alleviate some of the unnecessary fear and anxiety they've created.

Are we justified in believing that we can influence some people by providing good arguments for accepting evolution, anthropogenic climate change, science-based medicine, etc.? If we're not, then we may as well abandon education altogether. People do change their minds about many things and all of us have learned new things from time to time. Still, it is worthwhile to know that our brains may not be the unbiased truth-seeking missiles we imagine them to be. Knowing our own weaknesses can help us in our attempts at persuading others. For example, studies have shown that people are more receptive to ideas that conflict with their worldviews when they're in a good mood. We might try priming a hostile audience with some affect bias by getting them to think about how wonderful they are or about some time in the past when they felt really great because they acted on a value that was important to them (self-affirmation). Once they're feeling good about themselves maybe they won't feel so threatened by information that conflicts with their beliefs. This might work, but I wouldn't count on it.


I'll finish this post with some comments on a recent exchange I had with a practitioner of reiki. She asserted that I was wrong to attribute the positive changes in health or well-being after receiving reiki to the placebo effect. In defense of her claim, she provided two anecdotes about horses recovering from disabilities after she treated them with reiki. She claimed that "There is no placebo effect with animals." She also cited a blinded study in support of the claim that reiki done by uncertified people isn't as effective as reiki done by those with proper training and certification.

Now, it is unlikely her arguments will change my mind about reiki or any other form of energy medicine, but it is also not the case that her counterpoints have made me dig in deeper and that my belief that energy medicine is an illusion is now stronger. It is also unlikely that I am going to get her to change her mind about reiki or strengthen her belief in its efficacy. A few things are worth noting, however. I would not have responded to her had she not shown me the respect and civility she did by simply telling me I was wrong and then giving me her reasons for saying so. And she would not read my response if I did not respond to her civilly. While I probably won't change her mind about reiki, I may provide her with enough information about the placebo effect to get her to accept that when animals respond to conditioning they are demonstrating one behavior that is often put under the heading of the placebo effect. I may be able to expand her understanding of the placebo effect and, despite her current beliefs, get her to change her mind about animals being susceptible to what is sometimes referred to as a placebo effect. In addition, some readers of our exchange may learn something about the placebo effect that they weren't aware of. Anyway, this is her belief:

Animals have no preconceived ideas about reiki; no expectations about a treatment. There is no placebo effect with animals.

I replied that there is a lot of confusion over the use of the expression 'the placebo effect' and that the expression has become a catchall term for many things, including effects due to patient expectations. I review the many different kinds of things that have been attributed to placebos in my Skeptic's Dictionary entry on the placebo effect, so I won't go over the whole thing here. Anyway, one thing that has been attributed to a placebo effect is the effect of what is called 'conditioning.' You remember Pavlov's dogs: he rang the bell when he fed them and he conditioned them to salivate by just ringing the bell without presenting them with food. Other studies have found that a ritual like getting stuck with a needle in a clinical setting can condition both humans and animals to have physical responses, especially of the opioid system. Humans and other animals can be conditioned to release such chemical substances as endorphins, catecholamines, cortisol, and adrenaline. It is not strictly true that animals have no expectations in treatment. None of us do if it is the first time we are being treated, but dogs, horses, and humans do expect effects from treatments. Shots of morphine will make a dog salivate. Any shots, even of saline solution, after that will make the dog salivate. Of course, the animals may not be aware of their expectation, but conditioning studies show conclusively that the bodies of animals do respond in expectation of results. Not all cases that are attributed to the placebo effect are attributed to expectation effects, however, and of those that are, the expectation does not have to be something the patient--human or animal--is consciously aware of. Anyway, I wouldn't argue that the attribution of benefits to reiki treatments are all due to expectation, in either horse or human.

I admitted to her that the horses who recovered after reiki may have recovered due to reiki, but I also offered other possible explanations and explained why I found them more plausible than the reiki one. The first horse had been treated by a veterinarian with antibiotics and anti-inflammatories. I suggested that it was possible that this treatment kicked in after the vet had given up and that the reiki was given credit because the horse showed improvement soon after the reiki. The conventional treatment had been going on for several weeks, and it seems more plausible that medicine we know works much of the time kicked in here than that some subtle energy was manipulated and somehow magically healed the horse. I also suggested that there might be some unknown factor that could account for, at least in part, the horse's recovery at the time it recovered. Since there is no plausible mechanism by which reiki would work, it would be imperative to look to blinded well-designed studies that strongly suggest reiki works better than sham reiki. The site she cited, however, doesn't really provide much more than optimistic claims and positive assertions. It provides little scientific evidence. I referred her to the work of Edzard Ernst (who is quoted on the page she cites), R. Barker Baussell, and the Cochrane Collaboration to support my point that there haven't been the large-sample, high quality, blinded studies on reiki that are needed to support the idea that reiki might be a plausible healing method. There are studies, to be sure, that have found positive things to say about reiki, but there has not been a large, well-designed, controlled study, much less any replication of such studies. Until such studies have been published in reputable journals, I think I am right to remain skeptical.

I offered an alternative explanation for the second horse she thinks she healed with reiki by wondering whether the horse had been tested before the reiki treatments. The horse hadn't been ridden in several years because of fear of injuring it. She may have ridden the horse after the reiki, but the surgery done several years earlier may have healed the horse. Perhaps nobody had actually tested the horse's leg for fear of harming it. (She did not respond to my suggestion. Make of that what you will.) Anyway, until there is some plausibility that there is a subtle energy that affects health and can be manipulated by trained reiki professionals, I will remain skeptical. To those who say I'm closed-minded, I say: sometimes we don't change our minds when confronted with evidence contrary to what we believe because the evidence isn't very good.