Skip to content
Who's in the Video
James Surowiecki has been a staff writer at The New Yorker since 2000. He writes The Financial Page. Surowiecki came to The New Yorker from Slate, where he wrote the Moneybox[…]

What makes some crowds smarter than others? Hint: you don’t always want people who fit the mold.

Question: What led to the insight that originally caused you to explore how crowds might actually have some wisdom?

James Surowiecki: Well, the story is actually kind of interesting in retrospect because what happened is when I first started writing about business in the mid-1990’s, I really didn’t have much experience thinking about, or writing about business and markets.  And I was struck by the fact that a lot of times, although as we’ve seen recently, by no stretch of the imagination always, markets actually tend to often be smarter than most of the people in them.  So, we know that over time, most investors, even very accomplished investors underperform the stock market and the like.  And the traditional explanation for why this is, is if you looked at classic economics text books, they would talk about how investors are perfectly rational, they have perfect information, long-time horizons and the like.  Well, that’s obviously not true at all.  Investors are often very irrational; they act on very little pieces of information. 

I was interested in how can you take all of these very, very imperfect, and very, very flawed individuals and collectively how can they arrive at reasonably good outcomes; not great outcomes, but reasonably good outcomes?  And when I started looking at that, what I started coming across were a lot of examples that didn’t have anything to do with markets.  Examples of groups, again and again being able to collectively solve problems or come up with answers that were actually better in some cases than the smartest person in the group.  And that oftentimes were very, very complex.  So, over time, what I started to realize, or decide, depending on what you think of my book, is that it wasn’t that there was something about markets that was especially magical, and actually, in some ways I think markets are one of the more imperfect examples of the wisdom of crowds, but that actually they were just a specific example of this bigger phenomenon that I ended up calling “The Wisdom of Crowds.

So, the story that opens the book is the story of the British scientist, Francis Galton, who is sort of famous in terms of – he was one of the first people to really use statistical analysis to try to think about populations and individual behavior and things like that.  But this particular story is; about a century ago, he was at this country fair in the west of England and as he was walking through the fair, he came across this contest.  And the contest was that an ox had been placed on display and people were basically lined up to guess the weight of the ox.  Now, actually what they were being asked to do was to guess the weight of the ox after it had been slaughtered and dressed, which was sort of unfortunate for the ox, but if you think about it, it complicated the task.  It was a little more complicated than just guessing the weight of an ox as it stood there.  And it was actually a big crowd.  It was actually 700 people and there was a reward if you did well, you’d win a prize.  So there was some incentive for people to do a good job of trying to guess the weight of this ox. 

So, after the contest was over, Galton went up to the guys who had organized the contest and he said, “Can I have the slips?”  Sort of testimony to his brilliance as a scientist that he realized this might be useful information.  And then what he did was, he just did a series of statistical tests on them.  So, he graphed them to see if they formed a bell curve, to see if the distribution was normal.  He calculated the standard deviation, and then he did a couple of things.  He calculated the median guess and then he also eventually ended up calculating the mean guess, or the average guess. 

Now, Galton, as a lot of people know, was this notorious elitist.  For all the virtues of his statistical work, he had this great vice, which was basically that he was the founder of eugenics.  So, he basically thought that only a few people in the world really had the capabilities to make decisions or solve problems.  So, he, not surprisingly thought that the group’s average guess was going to be way off the mark.  He basically thought that you were taking the guesses of a few smart people, some mediocre people, and then a lot of morons.  Because he thought these people were dumb, basically.  And so he thought that the group’s estimate was going to be really flawed. 

But as it turned out, the ox weighed – I always get the two numbers confused, but the ox weighed 1,197 pounds, I think, and the group’s average guess was that it would weight 1,198 pounds.  So it was basically, essentially perfect.  And in this case it was actually better than any one person in the group.  Now, that’s an extreme example.  Most of the time the wisdom of crowds is not going to give you that perfect a guess and in fact, if you do experiments like this once, say with a jar of jelly beans, that’s another way you can do it.  The group’s average guess, there’s usually a few people in the group, a couple who do better than the group as a whole, but it’s actually, I think, a kind of great image and a great example of what the wisdom of crowds can actually look like. 

And one other thing I’ll say about it that is important is that the group was not just made up of experts, so it wasn’t just made of butchers and farmers who probably were pretty good at doing this, it actually included a lot of other people.  People like clerks, or family members, people who were just there to kind of take a guess as a lark.  And so the diversity of the group actually was something else that makes its intelligence interesting. 

Question: What makes certain crowds smarter than others?

James Surowiecki: You hear the phrase, “the wisdom of crowds.” It’s bandied about a lot, which is nice.  I mean, it’s nice that people have to some extent taken up the idea.  But it isn’t as simple as just kind of casting about out widely, ask as many people as possible and you’re going to end up with good results.  There are a lot of examples, and we just have lived through one of; well, we’ve lived through two if you think about the Internet bubble of the late ‘90’s and now the housing bubble of this decade.  There are a lot of examples of crowds being either really extreme in their behavior.  So, really volatile or taking positions that are way out there.  Or alternatively, a lot of times crowds are often quite mediocre.  They’re just sort of very ordinary in their thinking.  And so I think it is really important to recognize that you need to do certain things to get people to be wise as a group. 

I think the three things that are most important are, first of all you need, and this is kind of obviously, but it’s actually quite important.  You need some way to aggregate lots of different individual judgments to produce a kind of collective judgment.  So, for me, the wisdom of crowds isn’t just about a kind of suggestion box.  A lot of people, for instance, sometimes talk about kind of the Blogosphere, as an example of the wisdom of crowds.  And certainly it’s better to have lots of voices rather than a few, but if you don’t have a mechanism for taking a lot of different opinions and aggregating them in some way, then you just have a lot of different individual opinions and you’re not necessarily coming up with a true kind of wisdom of crowds thing. 

The second thing, which I actually think is probably the most important, is that you really need the group to be diverse.  So, you really want people who are looking at a problem or trying to come up with an answer from lots of different perspectives.  You want people who are using different kinds of tools to think about how to solve a problem.  And that actually can be a big challenge.  It can be a big challenge inside organizations because a lot of organizations push people to be homogeneous in their thinking.  Sometimes organizations just aren’t good at recruiting for diversity.  You hear a lot of lip service paid to diversity in corporate America, but a lot of times organizations tend to hire people who sort of “fit the mold.”  And I’m not really talking her – well, to some extent that’s true just kind of sociologically.  You tend to hire people with similar backgrounds to you, or who look like you, literally, or who come from similar schools.  I think the more important thing is people tend to hire people who have similar perspectives to them.  When companies build teams, they oftentimes tend to hire people, or recruit people who are going to fit in, and I think that’s a big problem.  So diversity is very, very important. 

And then the last, thing which is connected to diversity, but I think is slightly different from it, is that even once you have a diverse team, or a diverse crowd, you also need to make sure that people are able to be independent in their thinking.  You need to make it possible and easy for people to think for themselves rather than them having taking a lot of cues from each other.  And although in principle it seems like, well that should be easy, you know most of us want to be independent thinkers, we don’t think about ourselves as being conformists.  It can actually be pretty challenging.  And I think actually, in some ways, the internet magnifies the problem of what I kind of call “dependent thinking” because it’s very easy for small networks to get set up where people tend to only hear the voices of people they agree with.  And it’s very easy for feedback loops to get set up that make it easy again for the same kind of ideas to get more and more reinforced. 

So I would say the three things are aggregation, and then the two most important are diversity and independence.

Question: What can we learn about the wisdom of crowds?

James Surowiecki: Crowd sourcing which is a term that Jeff Howe came up with a few years ago.  I think that crowd sourcing describes a wide range of activities.  Some of which include wisdom of crowds sort of problem solving, but also include other things.  And I think the way I would think about it is that some crowd sourcing actually involves not tapping the collective knowledge of the group, it actually involves sort of casting your net very widely in order to find the one or two people in the group who can solve the problem you’re looking for, or design the product you want, or whatever it is. 

And so, I think that can actually be quite useful and I think as a lesson in the fact that not all talent is located where we think to look, that traditional experts don’t have all the answers, that actually there are a lot of individuals out there that know quite a lot and you can get their input and their information.  That’s a very valuable lesson.  But I think that, from my perspective, the best techniques are those that try to tap the collective knowledge of the group.  And I think actually there are some that actually combine the two in a useful way.  So, I’m not going to remember what it was.  But I think Sony did this animated film basically, via a kind of crowd sourcing technique.  And what it basically entailed was, obviously individuals came up with the various components of it, but the crowd was constantly weighing in on the editing, basically; which should be included, which shouldn’t.  And I think that is a really useful model where you actually have the crowd serve as the filter.  The crowd serves as the judge.  And that, I think, has enormous promise going forward. 

You know, if you think about the wisdom of crowds in every day life.  I mean, one of the great examples for me that I talk about in the book and I think is obvious is Google Search Engine.  I mean, that’s a search engine that’s build basically on the collective judgment of people on the internet because the page rank algorithms relies on hyper links from one website to another as a judgment of the value of that link and to the extent that you think – and now most search engines now use some form of this.  To the extent that you think they do a good job of finding information, which I think they do an exceptionally good job.  That’s a great example of the wisdom of crowds in action. 

Question: How dependent, if at all, is the wisdom of crowds on leadership figures within the crowd?

James Surowiecki: Yeah.  So, that’s a really important question because one of the lessons of the wisdom of crowds obviously is that to some extent, we overrate leaders, at least to the degree that we think of leaders as synonymous with decision makers.  I actually think one of the fundamental lessons of “The Wisdom of Crowds,” particularly for organizations is that to the degree that you can push decision and at the very least kind of information input out into the organization as a whole the better the decisions you make are going to be.  And that we need to move away from the kind of corporate savior model of the CEO. 

Now, there are counter examples.  Whenever I talk about this people always say, what about Steve Jobs?  And my answer is, okay, if you know your CEO is Steve Jobs, and that he has that kind of exceptional judgment, which he seems to have, okay, well then in that case you can really rely on him.  But, generally you want to move away from that.  Having said that, I think that it is very true that inside organizations, leaders are important, not so much again in terms of making the decisions, they are important in terms of implementing the idea of the wisdom of crowds, convincing people that this can actually work and then just as important, taking the judgments, or the idea, that crowds generate and putting them into practice.  When I think about the role of leadership in the kind of new model organization, if I had my druthers, I really think that the role of the leader shifts to put less of an emphasis of straight decision making and more of an emphasis on aggregating knowledge, finding ways to get people to work together in useful ways, and then taking the results of that process and really putting it into practice.  Those are not easy things to do, but to the extent that you can do it I think it’s incredibly valuable. 

One CEO who comes to mind in terms of this is actually Eric Schmidt at Google, who is someone who has talked about the book, “The Wisdom of Crowds,” quite a bit, but he’s someone who I think is trying to instill that ethos inside the organization.  And it’s a tough switch for most CEO’s.  Most CEO’s didn’t get to where they were by saying, “Okay, let’s here what all of you think,” basically. 

I remember talking to a management consultant a long time ago who said for some bizarre reason, these parents he knew asked him, “How will we know if our child is going to be a good…” is CEO material or something.  One of the strangest questions every asked I think.  And he said, “If his Kindergarten teacher says, does not play well with others, then he’s a good CEO,” basically.  And that’s the old model CEO and I think that’s not the model CEO we really need going forward.

Recorded on January 15, 2010


Related