Choose Your Friends Wisely

But don’t despair.  There is nothing you can do about it.

Matthew Jackson explains why in The Human Network: How Your Social Position Determines Your Power, Beliefs, and Behaviors.  The book is an explanation of network theory—how things relate to one another.  Curiously, the same sort of models can explain not just the relationships you have with your friends, but how disease spreads and how financial panics occur.  It’s all about the network.

Start with why your friends are more popular than you.  Draw a picture with a circle representing each person.  Draw a line connecting your circle to every one of your friends. Then imagine doing this for everyone else—a line for every friendship.  (It’s a messy picture.) 

Now the obvious statement: popular people have more connecting lines.  That is the definition of “popular” after all. 

But, it has an additional implication that is not so obvious.  Suppose you have a dozen friends, and your popular friend has four dozen friends. That means there are 11 other people who have you as a friend, but 47 other people have your popular friend as a friend.  It is, in other words, way more likely that someone else is friends with your popular friend than they are with you.  In fact, it is also way more likely that your 11 other friends are people who have lots of friends, so they are also more popular than you. It is highly unlikely, for example, that one of your 11 other friends only has one friend who happens to be you. So, on average, your friends really are more popular than you.

The ripple effect of this observation is incredible.  Consider a college student.  How much time do the friends of the college student spend socializing?  How much time do the friends spend studying?  It turns out students overestimate the amount of time the average student spends socializing and underestimate the time spent studying. 

Why?  Socializing is a public act which generates more friend connections.  So, people who socialize a lot have more friends and are seen socializing. Studying on the other hand is a private activity.  You don’t see people studying as much.  And people who are studying a lot make fewer friends.

So, our hypothetical student will have more friends who socialize a lot and will see them socializing a lot, so will think college students must socialize a lot.  Our hypothetical student will have fewer friends who study a lot and will not see them studying, so will think college students don’t study that much.

Now add in the phenomenon of homophily—liking people who are the same as you.  People like people who are like them.  Even if this is only a slight preference, it has dramatic effects. You want friends who are like you means that you want to be like your friends. You see your friends socializing more than you see them studying for the reasons above.  So, you change your behavior to spend more time socializing than you otherwise would or even more than you would if you knew the actual average amount of time the average person spends socializing.

Fads come about for the same reason.  Suppose two relatively popular people for some reason or another decide to wear purple tennis shoes.  Then, since these first two purple tennis shoe wearers are popular, many people are friends with both and thus will think that wearing purple tennis shoes is a good idea and adopt the style.  Now there are more people doing this, so even more people conform.  Next thing you know, everyone is wearing purple tennis shoes solely because of an unrelated decision to do so by a very small number of popular people. 

(This is the same model as health epidemics.  A few sick people with lots of connections can infect a disproportionate number of other people, who in turn have a relatively disproportionate number of other people who will also be infected and so on.)

Now replace “wear purple tennis shoes” with “have a particular political belief.”  And now add in a technology that allows people to announce their political beliefs to their friends.  (Yeah, stretch your mind to imagine what such a technological innovation could possibly be.) 

A relatively small number of relatively popular people announcing a particular political view will ripple through the friend network until everyone in the friend network is announcing the same political view.  It doesn’t take any nefarious outside influencers. It doesn’t even take particularly influential popularizers.  It just takes the fact that some people have more friend connections than others and everyone has at least a small degree of desire to be like their friends. The simple power of network connections along with homophily is all you need.

Now really stretch your mind and imagine that two separate groups emerge.  Suddenly nobody knows very many people, and maybe even anybody at all, with different opinions, despite the fact that half the population has a different opinion.  No brainwashing is necessary.  News organizations devoted to one or the other set of views may even emerge as an effect, not a cause, of the spread of views in a network.

Of course this is all hypothetical.