Skip to main content

Network Effects in Bullying – Robertson – Hirshon

The spread of the bullying has strong network effects stemming from both direct and informational benefits. The articles I found describe some causes of bullying, from following the crowd to maintaining popularity. In both situations, network effects are present without being explicitly named. 

The article “Why Do Decent People Bully” (Robertson) presents cognitive dissonance and the brain’s need for consistency as causes for why a person with no initial intention to bully might begin. Robertson uses the story of a group of teenagers bullying their bus monitor to illustrate this point. The bullying begins small. Then with encouragement, the initial small tease escalates to harassment. As each incremental insult is justified by the previous one. The brain thinks: “If I am insulting person A, I must have a reason à Person A must deserve to be insulted à I should continue to do so”.

The podcast “Popularity and Bullying” (Hirshon) explores how aggression increases with popularity. The second-most popular tier tends to be most aggressive. The most popular tier however, is significantly less so.


A “network effect” is the effect that someone else’s action has on your own. Network effects result from “direct benefits” and “informational benefits”. A direct benefit is the advantage a person gains by doing the same action as the crowd. An Informational benefit is the use of the information revealed by other’s choices. For example, if an unfamiliar restaurant is crowded it might be because the crowd knows something you don’t – that the food is good.

There are two main direct benefits of bullying presented here: First, increased aggression is linked to increased popularity. There is a subconscious feeling that by bullying those less popular, an individual can increase his or her own popularity. Second, joining the crowd can make you safe from bullying. If everyone is bullying person A, then they are not bullying you.

In this case, the informational benefit is similar to the cognitive “consistency” discussed in Robertson: If everyone is bullying person A and/or I am bullying person A, there must be something bad about person A. Thus the chain of “Person A must deserve to be insulted à I should insult Person A” begins.


The spread of bullying with networks effects presents a scary, though not surprising reality. Network effects lead to information cascades where everyone does the same thing because that is was the previous people did. In information cascades, people are behaving rationally based on the available information but their actions are not what is actually best for themselves or others. The example in class uses he red and blue marbles in the urns. After two red marbles are pulled, everyone thereafter guessed red – even if people 3 through 100 pick blue marbles. If after guess everyone looked at all the marbles, they would regret all guessing red.

Bullying spreads like an information cascade. The story in Robertson is a clear example of this. The bullying began relatively suddenly and spread rapidly. After being confronted for their actions, many of the teenagers “expressed apparently genuine remorse”. This parallel suggests that in some cases, bullying is the rational choice – just like guess red in the marbles and urn example. Bullying is an expected part of almost every child’s life – whether they bully, are bullied, or know someone who is bullied; network effects suggest this will not change. On the bright side, network effects also suggest ways to prevent bullying: Information cascades can be easily disrupted by revealing what people actually know instead of just what their choice or action was. By giving kids information to recognize when they are becoming a bully, future “bullying cascades” can potentially be prevented.


Leave a Reply

Blogging Calendar

November 2015
« Oct   Dec »