The idea that we follow others has some interesting implications. There was a famous campaign in the United States, an anti-drug campaign called the Just Say No campaign. It was started in the early 1980s by Nancy Reagan, and the idea was very simple. Drug rates are increasing, particularly among kids, so public service announcements came out talking about the dangers of drugs and trying to encourage people not to do them. Now, this seems like a pretty simple idea, we want to encourage kids not to do drugs, so the ads had a variety of real-life situations where kids might be asked to do drugs. A neighbor, for example, might say, "Hey, do you want to do drugs?" Or "Hey, the kids at school or doing drugs, do you want to try them?" The idea was by learning to say no, by seeing other kids say no, people would be more likely to say no themselves. More recently, some researchers analyzed some data, not just of the Just Say No campaign, but a more recent campaign against drugs, and they looked at whether this campaign was effective. Did children who saw the antidrug ads, were they less likely to take drugs? You might expect the answer would be yes, that showing kids antidrug ads telling them just to say no would decrease drug use. Unfortunately, that didn't happen, it didn't lead to decreased drug use. In fact, it didn't even lead to no change in drug use, it actually had a perverse effect. Kids that saw more antidrug ads were later more likely to report using drugs. Now, this might come as a big surprise, the ads were particularly designed to get people not to use drugs, so why did it actually have a perverse effect? Well, think about it for a moment. You're a kid, you're sitting at home, you've never thought about trying drugs before, and then an ad comes on television, it says, "Hey little boy or girl, there's something called drugs." Well if you've never heard of drugs before, the ad just told you that they exist, and then the ad proceeds to show other people using drugs. They say, "Hey, the kids at school are using drugs, or the cool kids are using drugs, but you shouldn't." You're sitting there going, "Well, if the cool kids are using them, maybe I should check them out as well." While telling people not to use drugs, the ads were simultaneously saying other people are doing it. Whenever we tell people that other people are doing something, they're going to be more likely to do it themselves. There's another great experiment that shows this really clearly. Some researchers are interested in getting people not to take souvenirs from a petrified forest. There's a famous national forest in the United States and people often take petrified woods as a souvenir on their way out, but if everybody takes a piece of petrified wood, there's no petrified wood left for anybody else. How could they get people to stop taking those souvenirs? Well, one campaign said many park visitors have removed petrified wood from the park, changing the natural state of the forest. Now, you might think that campaign would be effective. It said, hey, lots of people are doing it and it's a bad thing, it's changing the state of the forest. Was it effective? Well, when researchers looked at the data, they found that no message at all, about three percent of people took petrified wood, but the message about removing the wood tripled the number of people that took a piece of wood home. Why? Because the message while simultaneously saying don't take the wood, again, said lots of other people were doing it. A message that was much more effective, said something like, "Please don't remove wood from the forest in order to preserve the natural state of the forest." If we want people not to do something, don't tell them that other people are doing something. The more we tell people that other people are doing something, even if we don't want them to do it, the more likely they'll be to do it. On the other hand, when we want to convince people to do something, we want to point out how many other people are doing it. In Britain, for example, they wanted to raise more tax dollars, not enough Brits were paying their taxes on time, most were, but not all were. They added one line to a letter they sent to people's homes that increased the tax revenues by a huge number by tens of millions of dollars. The only line they added to that letter was, nine out of 10 people of your peers pay their taxes on time, and by merely letting people know that other people were doing it, it increased the chance that they would do it themselves. We've talked about how conformity shapes the individual decisions we make, but they also have some interesting implications for why some products and ideas become popular. Think about something that's recently become popular. Maybe a famous book, or a movie, or music star, take J. K. Rowling, the author of Harry Potter. That book has sold tens of millions, if not hundreds of millions at this point of copies. All around the world, people love the book and it's one of the biggest hits of all time. Well, think about why it succeeded. Was it because that book was just better than others? Indeed, when we see a big hit like a Harry Potter, we think that must just have higher quality. That it won out among all the other books because it's just a better book, but interestingly when we look at the story behind Harry Potter, we see something quite different. J. K. Rowling, the author actually shop that manuscript to a number of publishers who turned it down. Many said it wasn't very good, it had a niche audience, it needs to be rewritten, she had to shop it to dozens of publishers, so someone finally picked it up and made her a multimillionaire in the process. If certain things are better than others, why don't experts realize it? Maybe we wouldn't, maybe I'm not an expert editor, an expert knower of which books are better and which are worse, but someone whose job is to pick good books should be able to tell. Why did lots of publishers turn J. K. Rowling down? Well, it turns out there's something more interesting at work. Success is often unpredictable, not just because it's based on quality, but because success is based a lot on social influence. Imagine you came to a website that had a bunch of different music artists and they had different songs that you could download. There's a long list of artists with different songs, so you can pick whichever one you like. There are songs you've never heard of, bands you've never heard the names of, but you listen to a couple and at the end, you end up downloading a certain song. Most researchers ran a study just like this one, people could listen to whatever music they'd like and download different songs, but in addition to the version I just told you, they also ran the world that included social influence. In addition to the name of the song and the name of the artist, next to it was a number of other people that had downloaded that song previously. Just like the number of views that might appear on a clip on YouTube, or the number of sales that might appear next to a book on a bestseller list. What they looked at is how just that number of what other people had done previously change what people listened to. They found not surprisingly, that people tend to listen to songs that were already popular. People tended to follow what others had done previously. The songs in that social influence world, the hits became much more popular and the failures became much less popular. As we've talked about, people tend to follow others. But there was one more interesting hitch to this story. They didn't just run one social influence world , they ran multiple. Multiple cases where a bunch of individuals were assigned to one situation where they listen to songs. What they found is in those different separate worlds, those different groups of people, completely different songs became popular. If it was just about quality, the same song should've been popular in each world, yet in one world, a certain song might have been the best hit, and in another, it was close to last. Why did that happen? Well, it turned out social influence not only drove success, some things to succeed and others to fail, but it also drove some of the randomness. People tended to follow others. Whoever listened to the song first had a big impact on what others did. Just like that person who might have been the first to speak in the meeting, people tended to follow the others that came before them. They tended to listen to the same songs, to download the same songs, and increase the hit count even further. That's why bestseller lists have such a big impact. We tend to look to those lists to figure out what we should read or watch, but that in turn leads those songs or movies to stay on the list, and then that affects the next people as well. What this study shows is that social influence does two things. First, it increases inequality. It makes popular things even more popular and unpopular things even less popular. But importantly, it also makes success unpredictable. The winner in one version was almost completely different than the winner in others. Just like that first person in the meeting has a big impact on what the rest of the group does, the first person who listened to a song had a big impact on the people that follow them. This idea of heard or bandwagon effects can have a big impact on what catches on. If we're designing a recommender system or thinking about how to provide information to others, we need to be really careful about that social information. In some cases, we want to do what I'll call making the private public. We want to make it more observable what others are doing. The easier it is for us to see what others are doing, the more likely it will be to imitate it. But in other cases, we want to make others' behavior private, if we don't want people to imitate others, we want to make others' behavior harder to see.
