Professor Christine Nero Coughlin (JD ’90) blogs on Huffington Post: ‘Election 2016: The Potentially Polarizing Effects Of Search Engines, Social Media And Motivated Reasoning’

Photo of Professor Christine Nero Coughlin (JD '90)

Professor Christine Coughlin (JD ’90) originally published the following post on The Huffington Post Blog here.

 Editor’s Note: The views and opinions of our faculty members that are invited to write in national media outlets are their own, and not reflective of Wake Forest Law as an institution. Our policy is to re-publish all faculty member articles that are published in national media.

According to a new study by the Pew Research Center, 81 percent of supporters of Hillary Clinton and Donald Trump disagree not only on their respective policies and plans but also on the “basic facts” surrounding the candidates and the election. The Pew Research Center stated that its survey was conducted online from Sept. 27 to Oct. 10 among 4,132 adults, including 3,616 registered voters, and was largely completed before the release of the 2005 Hollywood Access videotape showing Trump making lewd comments about women.

According to Pew’s website, this survey shows how “negative attitudes toward the opposition have become a defining feature of the current campaign.” While myriad reasons underlie this perfect storm of negativity, we should be aware of how and why the information we obtain through technology such as search engines and social media contribute to political polarization.
While many search engines exist, Google is far and away the most popular. In fact, its name is now commonly used as a verb. Google adjusts its “top-secret” algorithm – its 200-plus numerical ranking factors used to retrieve and rank information – more than 600 times per year. When an individual enters a search, the results are ranked based on factors such as the terms used, the responsive domain sites, along with their popularity, and how frequently the terms are used on that site, as well factors such as an individual researcher’s location, preferences, and search history. To provide a simple illustration, if I am a sports historian in Columbus, Ohio, conducting research on Heisman Trophy finalists and type “London” into my Google browser, I will likely retrieve highly ranked search results about London, Ohio, located approximately 25 miles southwest of Columbus, and the former home to Warren Amling, an All-American Ohio State football player and 1945 Heisman Trophy finalist. However, if I am a college student located in Raleigh, North Carolina, looking at travel sites for a post-graduation trip and I type in “London,” my highest-ranked search results will likely involve London, England, things to do and budget travel sites.

Scholars who have studied, and even those who publicly criticize Google, all acknowledge that the search engine excels in giving the individual researcher the information sought. In fact, despite the thousands of responsive pages that may exist for any given search, 50 percent of clicks tend to go to the top three items and 91 percent of searchers never go past the first page. In addition, once the researcher clicks on a link, that click illustrates relevance, thus possibly becoming a factor for results in future searches.

Let’s apply this phenomenon to the major party candidates in election 2016. Based on whatever iteration of the algorithm is used, along with the searcher’s location, preferences, and search history, the information retrieved is consistent with what the researcher anticipates finding and what he or she has researched in the past. In the political information realm, this phenomenon builds upon itself so that when supporters of Mr. Trump type in his name, they tend to see positive articles about him more highly ranked and negative articles about Secretary Clinton more highly ranked, and vice versa. Supporters of Mr. Trump are more likely to click on links that support their political views, showing relevance, as are supporters of Secretary Clinton. The end result is that the information obtained and read by the two opposing supporters, even when they use the very same search terms, may be vastly different. This clearly contributes to the polarization we are seeing – an 81 percent disagreement even on the basic facts about the candidates and their positions.

Similarly, the use of social media to obtain information has exploded. There are more than 300 social media sites that enable users to create and share and receive personal or professional information with self-selected friends or followers. In election 2016, both of the major party candidates are using social media – in particular Twitter where Secretary Clinton has 9.89 million followers and Mr. Trump has 12.6 million followers – to generate support. As an aside, similar to the verb “google,” Twitter now has its own verb “tweet.” Also similar to the search engine method of clicking on relevant information, the searcher who “likes” or “reacts” to a post or page, or enters search terms such as “bigly,” or “Wikileaks,” is rewarded with the appearance of more pages, posts or tweets that contain similar information.

When we consider the effect of the Google algorithm and social media in politics today, we shouldn’t adopt a conspiracy theorist attitude that there are nefarious activities or attempts by Google or any social media site to influence the election. Rather, this phenomenon of political information polarization is likely explained as the natural consequence of combining these newer forms of information gathering technology with the more traditional social psychology theory of motivated reasoning – whereby people tend to seek out information that confirms what they already believe.

On his blog site, Yale Law School Professor Dan Kahan illustrates motivating reasoning with the following example: “In the 1950s, psychologists asked experimental subjects, students from two Ivy League colleges, to watch a film that featured a set of controversial officiating calls made during a football game between teams from their respective schools. The students from each school were more likely to see the referees’ calls as correct when it favored their school than when it favored their rival. The researchers concluded that the emotional stake the students had in affirming their loyalty to their respective institutions shaped what they saw on the tape.”

It is natural to trend toward information that supports our beliefs and worldview but if 81 percent of voters cannot even agree on the basic facts involved in the election, this tendency, along with some of the information bias built into obtaining political information in today’s world, could be damaging to our country’s ability to reunite post Nov. 8. We will continue to be influenced by political information we receive through Google, social media and even newer technologies – of this we should be certain. This is not, however, necessarily negative. What we need to do is to understand the potential effect of technology on the information we obtain, view that information through a critical lens, and ensure we are always seeking alternative viewpoints – in other words, engage in critical research and analysis. In addition, we should consider the experience of political canvassers who have made successful strides, particularly in the past two elections, in opening voters up to new ideas and policies by engaging in genuine, and respectful dialogue with each other.