Exposing Filter Bubbles and the Value of Diverse Political Ideology
by Carolyne Im
In the wake of the 2016 presidential election, polarization in the U.S. seems to have reached a new high. With John McCain’s death in August 2018, people mourned the loss of the Maverick, the type of politician that seemed to exist beyond party lines. Polarization and the pull towards identity-driven political action currently plagues the country and the globe. It is often manifested in radical right-wing and, at times, post-truth politics, as with the election of Jair Bolsonaro in Brazil, Brexit in Britain, and the prominence of Marine le Pen and the Front National in France.
This polarization is not only limited to politics; researchers have found that people on opposite sides of the political spectrum cannot even agree on science. James Evans, a professor of sociology at the University of Chicago, tracked the Amazon suggested pages for politically biased items in his study, “Millions of Online Book Co-Purchases Reveal Partisan Differences in the Consumption of Science.” Evans and his team analyzed the suggestions under the “Customers who bought this item also bought” heading for Barack Obama’s book Dreams from My Father and Mitt Romney’s No Apology. Looking at the top 100 items in each list, they compiled 1,303,504 unique titles. They found left-leaning customers were directed to science books on basic science topics such as physics, astronomy, and zoology, while right-leaning customers were directed to science books on commercial science, such as criminology, medicine, and geophysics. Evans said conservatives tend to draw on science associated with economic growth because “that’s what they want from science.” In comparison, Evans said for liberals, science is “more like Star Trek… traveling through worlds, searching for new meanings, searching for yourself.” Essentially, he concluded not even science is free from partisan bias, showing a tendency of confirmation bias even in the most non-political subjects.
Evans’ study provided a concrete example of filter bubbles, an increasingly popular term for the algorithmic tendency for internet sites to suggest congenial information. A term coined by Eli Pariser, a left-wing political internet activist, filter bubbles can be seen everywhere online, from YouTube autoplay to the slew of Amazon suggested items to even the top Google results. Pariser revealed that there is no standard internet anymore; it is all custom fit for the specific consumer in an internet culture where the most valuable commodity is sustained attention. Filter bubbles have exacerbated the problem of echo chambers within the media, in which people are isolated to their own perspective and perspectives like theirs. However, they take it to a further, even more dangerous level; while previously isolating oneself to an echo chamber may have been a personal choice, filter bubbles are completely independent of the consumer’s desires. Though one may be exposed to information with balanced political bias, filter bubbles take it completely out of their hands. Participating is completely involuntary and unavoidable. Subsequently, the media has created parallel streams of reality in which different information about the same topic is disseminated, so although two people of opposite political leanings may be discussing the same thing, they are working with a completely different set of facts.
It is this phenomenon that Evans observed in his initial study observing the Amazon algorithm. After this first experiment, he moved on to a second study, entitled, “The Wisdom of Polarized Crowds,” which asked the question, Can polarized individuals cooperate and yield successful results? He and his team found that diversity in political opinion actually leads to better performance on complex tasks. They came to this conclusion by analyzing the effects of conflicting ideology on team performance on Wikipedia edits. They studied 400,000 online teams throughout Wikipedia editors on English articles on politics, social issues, and science. They found teams with high polarization, or a high “variance of political alignments,” yielded higher quality articles than that of homogeneous teams, resulting in “longer, more constructive, competitive, and substantively focused by linguistically diverse debates.” They measured the quality of an article on a six-category scale, ranging from stub, start, C class, B class, good article, and featured article, stub being the lowest quality and featured article being the highest quality. They saw with a 1-unit increase in polarization, the odds of moving from a lower- to higher-quality category multiplied by 18.57 for political articles, 2.06 for social issues articles, and 1.90 for science articles. They concluded this raise in quality was due to the greater diversity of information within the articles. The key, Evans said, was competition. Though many of the debates led to “begrudging” compromise, it yielded a better article because the editors were “motivated to find counter-factual and counter-data arguments that fuel[led] the conversation.”
Evans and his team said this study produced evidence that “ideological polarization can lead to productive, high quality collaboration”--if they are exposed to one another. Filter bubbles, echo chambers, confirmation bias, motivated reasoning, and selective exposure all characterize the modern media culture in which we seek out congenial information and reinforce our preexisting biases. For this reason, the internet is, in fact, not a democratizing source. It is limiting the scope of information people see, not expanding it. By revoking the choice to see uncongenial information, filter bubbles degrade individual decision and undermine democratic institutions. Evans and his team are trying to change that; they’re not just trying to expose disturbing trends in media, they’re trying to create a culture in which people are informed and actively organize a platform against these undemocratic tendencies. At the end of their study, Evans and his colleagues cites a quote from the philosopher John Stuart Mill. He writes, “Not the violent conflict between parts of the truth, but the question suppression of half of it, is the formidable evil; there is always hope when people are forced to listen to both sides.” As technology continues to evolve, Evans, Pariser, and countless more are trying to show us that we must stay vigilant and consciously seek out the other side. If we don’t change the culture of polarization, it is unclear how far identity-driven politics will go.
Carolyne Im is the Editor-in-Chief of the GW Scope and the Managing Editor of the GW Undergraduate Review. She is a junior majoring in Political Communication and minoring in Music. Currently, she is a Luther Rice Undergraduate Research Fellow conducting research on the discrepancy between professed attitudes of racial awareness and performed antiracist behavior.
Comentarios