Group polarization can be bad enough in real life. As Harvard Law Professor Cass Sunstein shrewdly observed in the late nineties, however, the Internet and social media have the propensity to lead to an exaggeration of this phenomenon because there is a greater propensity for politically and socially like-minded people to form online groups in which discussion and deliberation take place. Diverse groups are less likely to be affected by processes that lead to group polarization presumably because what I called “the initial condition” this past Monday fails to obtain, so no particular outcome is considered in-group desirable. But when a majority of people, or a few powerful group members, have similar agendas, the initial condition is bound to arise and psychological processes then move people to a more extreme point of view.
Blogs, online forums, social media sites and Internet groups tend to be frequented by people with very similar agendas. For example, while lawyers and psychologists might read Leiter Reports, the dominant readership are philosophers; while an older white woman may join a Facebook group for young, black philosophers, the majority of members will tend to be young black philosophers; and while people with no interest feminist debates may end up reading the blog Feministing, the majority of the readership will be people with a strong interest in, and inclination toward, feminism.
Another factor that can make group polarization more likely to occur on the Internet is the existence of rival “outgroups” (Sunstein, 1999). Because information flows so quickly on the internet, it very quickly reaches a lot of people some of which will disagree with asserted viewpoints. So, rival outgroups are more likely to form on the Internet than in real life, further increasing the likelihood of group polarization.
A third factor that can increase the likelihood of group polarization on the Internet is what is known as “exit.” This is a scenario in which group members leave a group because they do not like the direction in which people’s opinions are moving. This makes the remaining group members more like-minded, and this is a factor that can motivate people in groups to move to a more extreme viewpoint. This effect is hugely exaggerated on the Internet because it is so easy to leave or join online groups. Exiting a real-life group has multifaceted personal effects. It could sacrifice newfangled friendships and liaisons, mess with people’s daily schedules, and leave people without a real-life support net. Leaving an online group is often considerably less complicated and has fewer real-life consequences. In-group online friendships, if formed at all, are less likely to be as close and deep as real-life friendships. Exiting an online group may free up some time here and there but is not going to mess with people’s daily activity schedules, and while an online group can serve as a support net, the impersonal nature of the support net makes the group readily replaceable by another online community.
A fourth factor relevant to the exaggeration of group polarization on the Internet is the occurrence of emotionally charged events. A classical example is the assassination of Martin Luther King that immediately led to more extremist opinions within opposing groups. The 9/11 terrorist attack is a more recent emotionally charged event that shifted people’s opinion about Muslims to the extreme and the opinions of supporters of Muslims to the other extreme. After 9/11 the FBI reported that there was a 1700 percent increase in hate crimes committed against Muslims in the United States. There was also a significant increase in negative stereotypes associated with Muslim Americans. The Internet plays a central role in quickly circulating news reports of emotionally charged events, making people within groups feel more strongly about their group identity. The induced strong emotions can then move groups to more extreme positions regarding the groups responsible for the events in question.
There are other factors that can contribute to a higher likelihood of group polarization on the Internet. The feasibility of commenting anonymously and posing as a different person with a pseudonym is greatly increased on the Internet. When people are relatively or fully anonymous and group membership is made salient, then the chance of group polarization increases significantly. Anonymity is erasing individual differences, making it seem to group members as if their fellow group members are very similar to them. This, in turn, strengthens the feelings of belonging and group identity. When no or little attention is drawn to individual in-group differences, the majority opinions or the opinions of people of perceived authority come to seem desirable for the group as a whole, making it more likely for people to move toward that outcome or engage in one-upmanship that can lead to a more extreme perspective.
Another reason the Internet escalates polarization is that search engines allow people to selectively seek out information that confirm their existing opinions rather than arbitrary information that could challenge their viewpoints. A large array of information is generally positively valued, as access to multifaceted information appears to be the bread and butter of democracy and unbiased decision-making. But when a huge load of information becomes available, it becomes possible for people to do selective searches for information that confirm pre-existing biases or for allies who can assist in activism. Selective searches for confirming information allows people to present persuasive argument that can move individual people’s more moderate viewpoints to the more extreme. Selective searches for allies can help establish groups susceptible to the force of the arguments. Additional in-group discussion can then help people conjure up further arguments and move opinions to further extremes. Search engines readily available to everyone on the Internet thus contributes strongly to group polarization.
As Sunstein has argued, the Internet has created the need for a distinction between planned and spontaneous polarization. The spontaneous version of the phenomenon is what we have been discussing up until now. The planned version takes place when people purposely create groups of like-minded people with the goal of taking the positions of other group members to the extreme. It’s relatively hard to generate or decide to join groups in real life. Being a member of a real-life group is time-consuming and costly. In comparison, it’s extremely easy to generate or join groups on the Internet. Anyone can generate a Facebook group and invite selected individuals to participate or find an already-existing group to join. Although being active in groups online can be time-consuming, it is far less time-consuming than having to spend time on transportation, irrelevant small talk and useless food or cigarette breaks. So planned polarization is strongly encouraged by the structure of the Internet. As an example, consider a group of girls in high school who decide that they don’t like another girl from their class and plan to bully her. Prior to the age of the Internet, bullying would mostly take place on school property or at after-school activities, and it would be relatively easy for teachers to discover the abuse and intervene. With access to the Internet the scheming girls can meet any time they choose online to plan their attack and reinforce each others’ opinions, and they can conduct the offense completely online without the knowledge of parents and teachers.
Sunstein’s hypothesis from the late 1990s that polarization will be worse on the Internet has been confirmed for online communities (Daniels, 2009) and for Twitter (Yardi and Boyd 2010). Yardi and Boyd (2010) analyzed 30,000 tweets on Twitter about the shooting of George Tiller, a late term abortion doctor, on May 31, 2009 in Wichita, Kansas and found that conversations between like-minded individuals quickly led to strengthened group identity among both pro-life and pro-choice people. Conversations between people from different groups, on the other hand, widened the gap between the opposing communities.
Because the group polarization is hugely exaggerated on the Internet, a majority of the beliefs formed as a result of deliberation or the airing of opinions on blogs, social media sites and group forums are unreliably formed and hence do not have a great propensity to be true overall. Some beliefs end up being misleading or exaggerated versions of the truths. Other beliefs will be outright false.
As Sunstein argues in On Rumors, group polarization is also a trigger of a greater within-group commitment to the truths of rumors. Group polarization can explain rumors, such as the widely circulating falsehoods that Obama was not born in the United States, that the CIA was responsible for the assassination of John F. Kennedy and that Governor Sarah Palin thought that Africa was a country rather than a continent. The claim about Sarah Palin, which appeared on MSNBC, was supposedly traced back to Martin Eisenstadt, an alleged McCain policy adviser. But Martin Eisenstadt is entirely fake. The think tank where he was claimed to work, the Harding Institute for Freedom and Democracy, is merely a website and his blog and the online TV clips of him are all fake. The hoax was put on by filmmakers Eitan Gorlin and Dan Mirvish. They had previously fooled Jonathan Stein, a reporter for Mother Jones, and a Los Angeles Times political blogger. They subsequently claimed that they weren’t the source of the Palin rumor. Whoever was the source of it is not the only person to blame. The journalists and bloggers who didn’t vet the incoming information are to blame as well. But so is the phenomenon of group polarization. As Sunstein argues, people believed the Sarah Palin rumor, because they already held the view that she was an airhead. It was a simple step to move to a more extreme viewpoint.
When people’s conviction in the truth of a rumor is strengthened due to group polarization, this clearly can have devastating effects for individuals, as everyone who googles a person’s name will immediately see the false information but will not necessarily find out that the information was false. Lots of folks still believe the Palin story, for example. As Sunstein puts it:
In the era of the Internet, it has become easy to spread false or misleading rumors about almost anyone. A high school student, a salesperson, a professor, a banker, an employer, an insurance broker, a real estate agent-each of these is vulnerable to an allegation that can have a painful, damaging, or even devastating effect. If an allegation of misconduct appears on the Internet, those who Google the relevant name will immediately learn about it. The allegation will help to define the person. (It might even end up on Wikipedia, at least for a time.) (On Rumors, Chapter 1)
There are situations in which group polarization is a good thing, for example, within social support systems. Alcoholic Anonymous is an example of a social support system where group polarization is very likely to occur. Alcoholics tend to develop extreme viewpoints about alcohol and abstinence. Some of these beliefs may be exaggerations of true beliefs. Qua exaggerations, they are false. But in a setting like this it doesn’t matter whether people’s extremist beliefs about alcohol are true or false. It is far more important that alcoholics receive the group support they need to quit their addiction. So, in support groups where the aim is psychological support, group polarization can be a good thing. The problem it presents in other cases is that it is an unreliable belief-forming method, as unreliable belief-forming methods aren’t likely to yield true beliefs.
Daniels, J. (2009). Cyber Racism: White Supremacy Online and the New Attack on Civil Rights: Rowman & Littlefield.
Yardi, S. & Boyd, D. (2010) “Dynamic Debates: An Analysis of Group Polarization Over Time on Twitter,” Bulletin of Science Technology Society October, 30, 5, 316-327
Sunstein CR (1999) “The Law of Group Polarization”, December 1999, University of Chicago Law School, John M. Olin Law & Economics Working Paper No. 91.
Sunstein CR (2002) “The Law of Group Polarization” Journal of Political Philosophy,10, Issue 2, pages 175–195.
Sunstein, CR. (2009). Going To Extremes: How Like Minds Unite and Divide, Oxford University Press.
Sunstein, CR. (2009). On Rumors: How Falsehoods Spread, Why We Believe Them, What Can Be Done, Farrar, Straus and Giroux.