The MMR scare cascade
A British medical journal (the Lancet) published a paper in 1998, which raised the possibility of a connection between the MMR vaccine and autism. The leader of the group behind the paper was a British surgeon named Andrew Wakefield. The paper’s findings were based on twelve children who had been admitted to a London hospital with behavioural symptoms. The parents of eight of these children had suggested a link between their children’s behaviour and the recently administered MMR vaccine.
The press conference, which followed the paper, caused the story to quickly reach mainstream media as Wakefield expressed his lost support and belief in the vaccine.
The paper itself lacked any sufficient evidence to connect autism with the vaccine, and considering this lack of evidence, one would expect that such a paper would be ultimately disregarded and forgotten.
However a media frenzy ensued and a national health scare resulted. The scare, which mainly affected the parents of young children (who were to receive or had received the vaccine), had a huge effect on attitudes towards vaccines. The most dangerous effect was that the MMR vaccination compliance had dropped from 92% in 1996 to 84% in 2002. These drops have been linked to the large increase in measles and mumps cases following close after.
But how did such an insignificant and inaccurate study manage to become the biggest U.K. story in 2002? An information cascade could help provide an answer to this question.
In this situation, a parent can either support or oppose the vaccine for the child. I’ll call the parents against the vaccine the opponents and the parents for the vaccine the supporters. A single parent cannot physically view the number of parents who support and oppose the vaccine (like they can in the restaurant example*see bottom of page). Instead it is through the media that the parents get this “outside information”. The media works like, and has the same influence as the people in the restaurant who are already sitting down. The more attention the media gives to the vaccine opponents (equivalently; the more people who are sitting in a restaurant), the greater the effect this has on the rest of the population.
Additionally each parent holds some personal information that is a combination of previously held attitudes towards vaccines and information supplied by doctors and studies.
Therefore each parent can decide on his or her stance towards the vaccine by either listening to his or her own information or by listening to the media.
According to this setup, if the press discredited the paper (as it deserved), no change would be made to the proportions of opponents and supporters of the vaccine. However in reality, the press gave too much credibility towards the study, which initiated a cascade. The media’s focus on the paper effectively told the parents that this paper should be given more credit (Similar to the restaurant example, it seemed to the parents, that the media was aware of some extra information about the paper and findings which the parents were unaware of). This resulted in some parents ignoring their private information and instead listening to the media (assuming it knew something they didn’t).
As the number of sceptics grew, the paper got exponentially more presence in the media. As the media presence grew, more previous supporters became opponents (as more people sat in the “opponents” restaurant, people looking to eat were encouraged more to eat at the opponents restaurant).
This cascade reached its peak in 2002. There after the Wakefield paper began to lose significant credit as other large studies showed no link between autism and the vaccine. Further investigation showed that the original paper was fraudulent as large sums of money were backing Wakefield to find a problem with the vaccine. However despite the complete debunking of the paper, the effects of the paper were never fully reversed, leaving some people with a permanent attitude change towards vaccines.
This is an example of a dangerous cascade where following the crowd was not the correct decision.
* In the restaurant example, a person’s decision to eat at a particular restaurant is affected by the person’s personal information on the restaurant as well as the busyness of the restaurant. If a person sees a busy restaurant, he or she might assume that the people in the restaurant know some extra information that he or she does not. Therefore the person may disregard his or her own opinions and follow the crowd in the hope that the crowd knows best.