Why Rational People End Up In Echo Chambers

By Daniel Tippens

In the wake of Donald Trump’s election, many have been searching for explanations for his victory.  Some anti-Trumpers cite sexism or nation-wide xenophobia, a distaste for the establishment, and so on.  Some Trumpers point to a distaste for politically correct culture, dissatisfaction with the Obama administration’s economic policies, and Clinton’s email scandal.

But one thing that has really drawn my attention in this election is that despite the unprecedented media assault on Trump, he still won. Fact checkers eagerly painted a portrait of Trump’s ignorance, while the press documented numerous instances of “racism or misogyny,” disregard for ethical standards, and a disconcerting lack of political tact. Numerous news outlets explicitly endorsed Hillary and condemned Trump.  Beneath every article about Trump, for example, the Huffington post added a disclaimer:

Note to our readers: Donald Trump is a serial liar, rampant xenophobe, racist, misogynist, birther and bully who has repeatedly pledged to ban all Muslims — 1.6 billion members of an entire religion — from entering the US.

Social media picked up where the press left off, retweeting and sharing just about every anti-Trump article from New York to L.A. Indeed, I have never seen a barrage of this magnitude against a candidate, and yet the Trump train still continued to speed along the tracks.

Some said that this attention to Trump, while negative , actually benefited him, because, as it is said, “any publicity is good publicity.”  I don’t buy this, however, since the very same people who claim that the negative media attention paid to Trump benefited him will, in the same breath, contend that media’s focus on Hillary’s email scandal and the FBI’s investigation into it hamstrung her campaign.

As far as I am concerned, the media’s effect on the presidential candidates’ campaigns was minimal to non-existent — at best indirect — if we are judging their influence by the number of voters whose minds were changed. The most popular explanation for this is confirmation bias; that voters have been cocooning themselves in a bubble of like-minded people and ignoring all the arguments against their views. Trump said that he would bomb the families of ISIS and that he wants to deport millions of people? Even if he did say those things, the media is probably taking it all out of context. Clinton committed a Federal crime? Surely, that’s an exaggeration, given that she hasn’t been arrested.

So the allegation is that we shut out views, simply because we don’t like or agree with them, embed ourselves in communities of those of like mind, because it is comfortable, and are inherently averse to contrary evidence. And of course, it is held that this is all tragically irrational. In deciding who is the best of two candidates, surely one should examine all the evidence for and against them, and not shut out people or information that contradicts our pre-existing views.

While studies have shown that people tend to be somewhat irrational, I am skeptical that this explains our inclination to cocoon ourselves. Instead, I would argue that cocooning and succumbing to confirmation bias is actually the result of a more (though not entirely) rational process than everyone seems to think, and that this may be an even greater cause for alarm. If irrationality was a large part of the problem that led to Trump’s election, education could solve it, at least in theory. But if people are behaving rationally, then only changing the conditions under which rationality is practiced could work, and unfortunately, I don’t know how this would even begin.

Reasonable Skepticism

At the height of the public discussion over the health effects of smoking, tobacco companies put together their own teams of scientific researchers to investigate the issue. What resulted was the widespread dissemination of bunk scientific studies, designed to show that smoking doesn’t cause health problems. Similar industry-related “research” has been done with respect to the health effects of sugar and genetically modified organism (GMO’s).

But academically based scientists, who do not receive industry funding, have also done things that have raised eyebrows. Relatively new scientific fields, like psychology, cancer research, or dietetics have rushed to publicize discoveries and make prescriptions which, in very short time, have been rescinded or entirely reversed.  The scientific consensus on what sort of diet maximizes health has changed drastically over the past several decades.  Flossing, as it turns out, doesn’t prevent cavities, despite the fact that dentists have been telling us that it does.  And until recently, medical researchers were convinced that  radiation therapy is always useful in the treatment of cancer, which no longer appears to be the case.

Add to this the recent and well-publicized reports of significant – even, one might say, massive — replication failures in psychology and medical research, not to mention the widespread perception, backed up by survey data, that academics are overwhelmingly Left-leaning in their political orientation, and academic science has begun to hemorrhage credibility.  Taken all together, then, the phrases “scientific studies show” and “experts agree that” have lost a lot of epistemic authority with the general public.

We find a similar problem with the mass media, which has become increasingly partisan. Pretty much every popular outlet is known to promote one political viewpoint or a narrow cluster of viewpoints at the expense of the others. Things weren’t always like this. Editorial standards for objectivity — or at least the semblance of objectivity, used to be strictly enforced — and so mass media outlets enjoyed a significant influence over public opinion. In the 1970’s, Walter Cronkite was considered “the most trusted man in America,” and Huntley Brinkley won a Peabody award for outstanding, mature, and intelligent treatment of the news. In 1956 an American National Election study found that 66% of Americans believed that newspapers were fair, but in 2013 only 36% thought so.  Like industry-based and academic science, then, media outlets are now increasingly perceived as having goals other than seeking the truth. Phrases like “The New York Times reports that” have also lost what epistemic authority they once had.

Liar Detection and Cocooning

So, let’s return to the presidential election. Suppose that we have a voter named “Joe” who is prepared to do some research on the candidates, so that he can decide for whom to cast his vote come election day. Joe is aware of the fact that many media outlets propagate their political ideologies, and that the scientific community has any number of non-truth-related reasons to promote particular conclusions. As such, in doing his research, he cannot simply trust the judgments of the experts whom he encounters.  Also imagine that  Joe works 40 hours each week and has a wife and two kids, which means that he cannot verify these judgments – he cannot chase down every piece of counter-evidence and every counterargument that is offered.  So, how does he decide whom he should trust?

First, Joe forms some moral, political, and factual beliefs based on what he thinks is his last bastion of reliable evidence: his personal experience. If he lives in a rural, largely evangelical county in the lower Midwest perhaps he hasn’t seen a good reason to believe that he should be concerned about transgendered bathroom access or that climate change is a pressing issue. So, he begins with those beliefs he already has on these subjects, which are  justified by the strongest evidence that he has available to him: his own observations and valuations.

After doing this, Joe checks to see which media outlets promote the same views. He (likely unconsciously) tests which sources are reliable by seeing whether the source’s views line up with the beliefs he has formed on the basis of his experience and prior valuations. In other words, Joe engages in liar detection.

In ordinary life, we call someone a liar (or in the absence of malicious intent, ignorant), when they consistently make statements that we have strong pre-existing reasons to believe are false. When Joe sees The New York Times constantly reporting on the latest in the transgender bathroom wars or the apocalyptic consequences of climate change, Joe spots what he takes to be either ignorance or lies.

When you discover that someone is ignorant or a liar, you don’t trust them or go to them for information. You shut them out, which is exactly what Joe will do. Conversely, when Joe sees that his personal experience and valuations are also endorsed by Townhall and Fox News, he will take these sources to be propagators of truth and return to them for information about the election, throughout the campaign.

As you can see, Joe has now cocooned himself. He will only go to conservative outlets to get his news. Is this irrational? I don’t think so. If you, legitimately, find yourself in a skeptical situation and your most reliable source of knowledge is your own personal experience and valuations, then that is the best foundation upon which to form your beliefs and base your actions.

Now, I don’t want to give American voters too much credit. I think that people have taken some amount of joy in having a reason to cocoon themselves.  Like somebody who is told that he needs to stop dieting for health reasons, this can be a welcome surprise – an excuse to stop putting in the effort that a diet demands of him. Many of us have relished in the opportunity to insulate ourselves in echo-chambers and have likely jumped on that opportunity more hastily than is warranted.

But to me, the cocooning phenomenon should be no surprise, and not because we are inherently irrational, but because we have created a situation in which there are no institutions that voters universally trust. The most venerable authorities have compromised their reputations and have forced voters to engage in widespread liar-detection. If we want to prevent cocooning, we need to change the conditions under which rationality is practiced. We need to bring back the Cronkites and Brinkleys, and re-establish trust in some of the news media. Only then can voters evaluate presidential candidates in a more objective way. I would like to think we are moving in this direction, evidenced by things like The New York Times, after having their readers accuse them of biasing their reporting, issuing a rededication to objectivity in the wake of Trump’s election:

As we reflect on the momentous result, and the months of reporting and polling that preceded it, we aim to rededicate ourselves to the fundamental mission of Times journalism. That is to report America and the world honestly, without fear or favor, striving always to understand and reflect all political perspectives and life experiences in the stories that we bring to you. It is also to hold power to account, impartially and unflinchingly. You can rely on The New York Times to bring the same fairness, the same level of scrutiny, the same independence to our coverage of the new president and his team.

Maybe Trump’s election will spur other news outlets to follow suit. But until this happens, we will remain reasonable people within a room of mirrors, listening only to those who will largely agree with us, and being largely rational in doing so.

7 Comments »

  1. The media played Trump’s game all the way up to the debates, soft-balling questions, allowing his worst language audition without comment, trying hard to appear ‘objective’ through false equivalency. He was a TV star and he was good ratings. But we must also admit that Clinton was a terrible campaigner – until September, no rallies, no effort to meet the people, not even photo-ops. He played the Headlines game and she just didn’t. For most media it was Trump all day every day, until it was too late to change direction.

    It’s not a question of bias; but journalistic objectivity requires discernment, and its clear the current crop don’t have it; thus they not only lived in a bubble assuming Americans couldn’t choose Trump, but they sucked many into that bubble as well.

    Then, of course, there are just people who live in their own bubble – or cocoon as DanT notes above. For many of us, the election burst the bubble, which is a good thing. Others, however, are unreachable, like a friend of mine whose convinced that Trump couldn’t possibly have won the majority vote with the Cuban Americans in Florida, there just has to be something wrong with the demographics there.

    With someone that stuck, I don’t know what can be done. Any progressive alliance moving forward has to include my friend , as well others, eg., unyielding feminists and greens, who either see themselves ‘on the right side of history,’ or are are so certain of their cause that they are convinced that others can be ‘led to see the light’ somehow, and if they can’t, then we go hand in hand to hell.

    I am not optimistic, as one can see.

    Like

  2. The technique of information overload, whether practised by the proliferation of almost identical appliance models from the same manufacturer, or by flooding of the media with contradictory stories, is designed to make decision making hard. Swing voters are often characterised as being “low information voters”.

    https://en.wikipedia.org/wiki/Low_information_voter

    One might take the argument from that article cautiously, but “…[n]oting that 95% of incumbents in the highly polarized House of Representatives win re-election despite voters’ preference for centrist representation, [Bawn et al 2012] theorize that voters’ infrequent penalizing of extremist behaviour represents not approval, but a lack of attention and information.” For the polarized party base, extremism is a good, doubly so if it offends the supporter base of the other side.

    The counting of the popular vote has still quite not finished, I believe, but if it was driven more by who stayed away than who actually voted, then maybe it came more down to personalities. If one knows that one doesn’t know much about policy, then using a skill one might be confident in, the judgement of character, it might be a rational way to decide on one’s king, sorry president. Of the two alternatives, I might slightly prefer HRC’s policies (more-of-same), but I don’t really trust her, so I won’t bother.

    As a what-if, I was looking at the plot in this article:

    https://www.statslife.org.uk/politics/3050-does-donald-trump-defy-all-odds

    Down here in Australia we were getting sick and tired of US Presidential electoral coverage by about July. Maybe if the campaign was short and sweet, we might have seen a different outcome.

    Like

  3. Hi Dan, this was well done and I agree for the most part with your analysis regarding how people enact with media, so I don’t have much to add on that point.

    About the only thing I would question is if the media had such a small role in helping Trump’s campaign. People running for office spend a lot of money on advertising. Trump had tons of free advertising from news outlets. You address this with the following…

    “Some said that this attention to Trump, while negative , actually benefited him, because, as it is said, “any publicity is good publicity.” I don’t buy this, however, since the very same people who claim that the negative media attention paid to Trump benefited him will, in the same breath, contend that media’s focus on Hillary’s email scandal and the FBI’s investigation into it hamstrung her campaign.”

    I guess I am a counter-factual to that argument. Media attention, negative or otherwise kept both their names going. Both were helped (largely along the lines you argue) by all the media attention. And Trump got a lot more, and since he took the time to address accusations against him publicly that gave him still more time, while Hillary ended up waving off chances to get her face and voice out there. She got tired of doing what he was willing to do for publicity.

    Of course, the nature of “bad publicity” does count. Saying something that offends a bunch of people is one thing, Doing something which raises the question if you will actually do what your supporters want is another. I don’t think Hillary was helped by the differential nature of their “bad publicity”.

    Like

  4. “Flossing, as it turns out, doesn’t prevent cavities, despite the fact that dentists have been telling us that it does. ”

    Incidentally, I don’t think that flossing was ever supposed to prevent cavities. It was supposed to prevent gum disease or stop it getting worse once it happens. I am told that there is no serious doubt that this is still the case, but then how could I tell who was telling the truth?

    Like

  5. Without getting all Bahfest about it, I think it is reasonable that people believe socially, that by and large people believe what a group we belong to believes. I have often felt the tug to start believing something held by a group which is accepting of me, even though I have evidence that what they hold is not true.

    Like

  6. “If we want to prevent cocooning, we need to change the conditions under which rationality is practiced. We need to bring back the Cronkites and Brinkleys, and re-establish trust in some of the news media. Only then can voters evaluate presidential candidates in a more objective way. I would like to think we are moving in this direction, evidenced by things like The New York Times, after having their readers accuse them of biasing their reporting, issuing a rededication to objectivity in the wake of Trump’s election.”

    I don’t think the NYT’s statement should be taken at face value (as you appear to do). To me it looks like an attempt to salvage something from the wreckage, to save face with respect to their core liberal audience.

    More broadly, I don’t see any reason to think that what you say should happen (in terms of the media becoming more objective) will happen. Maybe that media ‘golden age’ you allude to was possible only because the *society* was more homogeneous (or at least harmonious) than it now is.

    I agree with the bulk of your analysis however.

    Like

  7. DB,

    I think I can agree with you that the media had some sway, but my suspicion is its influence was restricted to the point in time prior to Hiliary and Trump receiving the party nominations.

    Hi Robin,

    “Incidentally, I don’t think that flossing was ever supposed to prevent cavities. It was supposed to prevent gum disease or stop it getting worse once it happens.”

    Well, I took that from this recent NY times article http://www.nytimes.com/2016/08/03/health/flossing-teeth-cavities.html

    Though I’m not a dentist, so I’m happy to stand corrected.

    “I think it is reasonable that people believe socially, that by and large people believe what a group we belong to believes. I have often felt the tug to start believing something held by a group which is accepting of me, even though I have evidence that what they hold is not true.”

    Agreed, but I still think this fails to explain the growing cocooning phenomenon, especially in the face of data about growing distrust toward mass media.

    Hi Mark,

    “I don’t think the NYT’s statement should be taken at face value (as you appear to do). To me it looks like an attempt to salvage something from the wreckage, to save face with respect to their core liberal audience.”

    Yeah I can certainly agree with you there. Definitely possible (and probable). 😦

    “More broadly, I don’t see any reason to think that what you say should happen (in terms of the media becoming more objective) will happen. Maybe that media ‘golden age’ you allude to was possible only because the *society* was more homogeneous (or at least harmonious) than it now is.”

    Good points. On the issue of having reason to think that what I say should happen will happen, yeah, I alluded to this myself at the beginning of the essay when I said, “unfortunately, I don’t even know how this would begin.”

    Like