A Flaw in the Wisdom of the Crowds?
http://www.nytimes.com/2012/07/08/sunday-review/when-the-crowd-isnt-wise.html?_r=0
This article presents us with the interesting idea that the “wisdom of the crowds,” a term coined by James Surowiecki in his 2004 book may not be as reliable as most people take it to be—especially in the context of predicting the outcome of markets. In the case of the Supreme Court’s decision on healthcare, many online sources such as Intrade (a place where people can bet on real-world events), indicated that the mandate would be ruled unconstitutional with a much higher likelihood of 75%. However, contrary to this popular opinion, the mandate was ruled constitutional. What is surprising in this situation is that even up to the day of the decision; the popular opinion had not shifted in the right direction.
There are plenty of other examples of “mistakes” in the market with several potential causes. The example of Washington’s experts being fundamentally incorrect in believing in pro-invasion in the aftermath of the Iraq war shows how a flaw in the wisdom of the crowds may be due to the fact that people are inherently overconfident. Even when given outside information, they might ignore important cues while placing a heavy weight on their own opinion—which is eventually passed on to others. Furthermore, research shows that the reason why there is a lack of wisdom in the markets is oftentimes because the dispersal of relevant information is too slow or because only a small circle of people have the right information. Both of these potential areas of concern suggest that it is usually the case that too much emphasis is placed on opinions that take the early stage.
Drawing on our understanding of the effect of information cascade from class, we can see the legitimacy of the point about how there is extra emphasis placed on early opinions that surface. Based on the simple herding experiment that we explored (which shows that rational participants will factor in others’ opinions in making their own decision), it is clear that opinions that surface earlier are viewed by more people, and therefore have more of an impact on the popular opinion in comparison to individual opinions that surface later. Additionally, this extra emphasis that is placed on these opinions relates directly to the concept of herding/information cascade in that once an opinion is adopted by a large enough group of people, it becomes difficult to discount it later (cascade begins).
At the same time, people who express their opinions first are seen to be more confident in their decisions, which naturally demand more credit. However, interestingly enough, this does not necessarily improve the accuracy of the overall pool of opinions. Instead, it creates a situation where this is much more volatility. As mentioned in the article, the problem of overconfidence can be just as misleading as it is helpful. In this case, information cascade poses the potential to set off a domino effect in the wrong direction.
In class, we also discussed the real world implications behind information cascade and the existence of informational effects. We mainly focused on fixed scenarios through controlled experiments. However, through this article, we can see that informational effects may not be as relevant in real life as they are in the theoretical space of controlled experiments. Because overruling false information as a result of information cascades is not as instantaneous in real life as it is under the assumptions of experimentation (due to the limitations of the dispersal of ideas), we are faced with the interesting question—is the growing amount of information available in the world today helping or hindering the accuracy of the “wisdom of the crowds?”
-blogger23