Information Cascade in Politics
Information cascade is a common effect that can occur without realization. For example, we can observe some instances when we are accused of, or accuse others of “joining the bandwagon.” People are said to “join the bandwagon” when they follow a popular viewpoint regardless if that is their true viewpoint. Herding, or information cascade occurs as individuals abandon their own information to make a judgment and instead agree with the inference made from observing the surrounding people. Essentially, any information you may have had is abandoned in order to imitate others’ behaviors. For example, say there is a singer X. You heard his music and articles describing the great things he’s done, so you decide you like singer X and are a fan. Then, one after another, as you ask your friends what they think of this awesome new singer you discovered, everyone around you proceeds to hate on him, claiming he has no talent and they’ve heard he’s a brat in real life. So, you jump in the “bandwagon” and agree with their statements. You essentially abandon your positive judgment of him and all the supporting information that you had gathered and instead accept their judgment and infer that their information that he is a bad singer and person is true despite being contrary. This would be information cascade taking effect. We see this happen with a lot of public figures.
As the election approaches, I thought that there would be many great examples of information cascade. One instance I found in The Atlantic, a highly reputable, and prize winning magazine. This article presents a great example of information cascade happening in current events.
Endorsements are a huge factor in politics. To those who are not or no longer in the running, their endorsement is often as important as their views and policies. The endorsement essentially tie two individuals. This article in The Atlantic accounts and describes an instance where information cascade had occurred in endorsements: in recent Republican endorsements and un-endorsements of Donald Trump. After the release of the infamous 2005 video clip of Trump making lewd comments of women, there was a rush of Republicans “withdrawing their endorsements.” Zollman credits this rush to:
“people who privately want one thing, but feel they can’t declare that. Once a few people start declaring it, the rest say, ‘Oh, I can too.’”
Endorsements, in politics, essentially tie the policies of two individuals which in turn ties their fate. If public backlash is strong and people are all against Trump, by continuing to endorse him, they too will fall with him, losing their own supporters. On the other hand, although they may personally not agree with him, if Trump is doing well, they want to endorse Trump so that his supporters also support them. Thus, it is not in their interest to un-endorse him too quickly. Endorsements are not always whether a politician agrees with another politician but often a power play or predicting the public opinion. In this case, when this video was released, people expected strong Republican figures, such as Paul Ryan, to withdraw their endorsements and in return, everyone else would follow suit, ultimately leading to Trump’s downfall. People expected strong backlash so although their judgment may not have changed, many prepared to withdraw support as public opinion was inferred to be negative of him. As individuals started un-endorsing him, this caused the cascade of individuals to follow suit and also start calling him off. However, this was cut short when Trump called out to his supporters and threatened retaliation. By doing so, instead of being in the many, it gave the impression that these individuals were some of the few leaving him, making them easy and likely targets. As the inference changed, the information cascade stopped. As everyone, or at least the impression of everyone, was not withdrawing support as expected, this effectively stopped the cascade that was happening. Republicans began to infer the opposite, that there was a significant amount of people who still supported and viewed him as a viable candidate; thus, individuals began taking back statements claiming that their comments were half un-endorsements or not un-endorsements at all. Trump’s angry tweet changed the inference so that the few were no longer those endorsing him, but the few were those who were withdrawing. As the inference of the behavior of others changed, due to Trumps quick comments on twitter, Trump was able to reverse the information cascade to also reverse the first effect, causing the “flip-flop” of endorsements we are seeing now.
It is interesting to see information cascade work within the current political campaigns, especially such a clear example. It will be interesting to see what other instances will occur as the election approaches ever so closer.
View article here.