The other day, someone reminded me of the health craze a couple of years ago involving trans-fat. Laws were passed attempting to ban this horrible substance from foods in restaurants, every other food package in the grocery store had a “trans-fat free” label, and people worried about the effects of this chemical on their health. Now, however, it seems that the trend has faded, being mentioned only occasionally, much like various other short-lived dietary fads (remember Atkins?) I started wondering about why such trends seem to suddenly spring into existence, and wondered if this might be the effect of an information cascade. That is, once a critical number of people begin to accept that a certain health regimen or diet is beneficial, everyone else assumes that these initial few must be correct.

I managed to find a somewhat interesting article from a few years ago that was written expressing this same belief. The author, John Tierney, believes that the trans-fat craze was an example of an information cascade. Although he admits that information cascades can converge upon the correct decision, he seems skeptical of whether this occurred in the trans-fat case. He seems to believe that the trend was motivated by a few individuals who subsequently influenced others to follow their decision.

Although the author does not go into much detail regarding the possible network effect, he does make a few interesting points regarding how various people can influence the decision of many others. He specifically mentions how a few people deciding to act in favor of their own opinions (that is, not necessarily following the opinion of the cascade) can end the cascade. As discussed in class, the cascade begins when the number of votes in favor of one option exceeds the number of votes for the alternative by a certain number. However, if enough people were to break the pattern and vote against the cascade, then it is possible to bring the votes back inside this margin and end the cascade.

However, there does seem to be one odd flaw in the author’s argument. He initially mentions that the arguments fit best to binary choices, or situations in which someone has only two possible options to choose from. He then however, starts discussing how one could recommend a mixture of low-fat and regular diets, which is not a clear binary example. However, the argument would still work if it weren’t for another error. He gives the example of being the second person to choose and disagreeing with the opinion of the first. In this case, it seems likely that one would either continue to advocate one’s original opinion or favor up to an even division of opinion between the two choices. However, he states that this second person would then advocate slightly more in favor of the first person’s opinion. Regardless of how reliable or unreliable one believes the signals to be in any similar situation, one would not switch away from one’s own information on the basis of second-hand information from one other person. Regardless of these odd choices in argumentation, however, the article remains interesting to read.