The Wisdom of the Mob

Part I of this occasional series on government and political philosophy is here.

Ancient Greece is rightly known as the cradle of democracy. Athens was ruled for centuries through a type of democracy in which regular people rotated through the various democratic institutions, doing their civic duty to help make the right decisions for all Athenians. Athens was far from democratic in a modern sense because many residents were slaves and could not take part in the democratic process. Nor could women.

What is far less well-known is that ancient Athens’ most famous sons, Plato and Aristotle, were staunchly anti-democratic and railed against the excesses of democracy as they witnessed it. Plato, for example, decried the democratic mob — actually a jury of 500 Athenians — that voted to execute Socrates, Plato’s mentor. Socrates was executed for the crimes of corrupting the youth of Athens and of impiety, apparently for advocating recognition of different gods than those recognized by the state (according to Plato’s well-known dialogue, the Apology).

Will Durant, in his excellent The Story of Philosophy, wrote of Plato’s views: “But even democracy ruins itself by excess — of democracy. Its basic principle is the equal right of all to hold office and determine public policy. This is at first glance a delightful arrangement; it becomes disastrous because the people are not properly equipped by education to select the best rulers and the wisest courses.” Plato wrote in the Protagoras: “As to the people, they have no understanding, and only repeat what their rulers are pleased to tell them.”

We hear echoes of Plato’s arguments today on both the right and the left. Conservatives are often of the republican (small r) type who praise representative democracy, as opposed to direct democracy, and the need for professional politicians to focus on the problems of governance that are beyond the ken of the average person’s time, interest, or ability. On the left, we have sometimes similar arguments decrying the lack of wisdom of the working-class white folk who unwisely vote for conservatives who will implement policies that go against the interests of those same voters (see Thomas Frank’s What’s the Matter with Kansas, for example). Even though we often hear conservatives criticizing liberals as “elitists,” the frequent criticisms of direct democracy from both sides of the aisle reveal a strong elitist strain across the spectrum.

Here in California we often hear a newspaper editorial or TV pundit advocating a no vote on a particular initiative, even though the initiative may contain good ideas, because we’ve just gone too far with this experiment in direct democracy. California’s system of direct democracy, we often hear, with its initiatives, referenda, and recalls, has made the state all but ungovernable. We shouldn’t encourage major decisions by initiative because we need a professional class of decision-makers to make those hard decisions for us. That’s why we elect leaders: to make good decisions for us.

Hogwash and poppycock.

Crowd wisdom in the private sector is increasingly being shown to be better at solving serious problems and making tough decisions than the so-called experts. This body of evidence provides strong support for the idea of expanding direct democracy, which I fleshed out in Part I of this series of essays, summarized as follows: We should utilize online voting and an expansion of what voters are allowed to vote on as a way to better utilize the wisdom of the crowd, to reduce corruption, and to reduce the need for all-too-rare enlightened leadership. Power should devolve to the most appropriate level of governmental or nongovernmental organization in order to maximize freedom and to find the best solutions to on-the-ground problems. I labeled these ideas “progressive libertarianism” or “isocracy” — rule by equals.

The difference between the Athenian era and today is that the large majority of the electorate is plenty educated and plenty equipped to make the decisions required to run a modern democracy. Moreover, the speed and spread of information is exponentially higher than it was in the days of ancient Athens. The real problem is that opinion makers, as well as our elected leaders, simply aren’t educated about the wisdom of the crowd, of mobs.

In the last couple of decades, however, we’ve seen an abundance of pretty compelling evidence showing that collective intelligence — the wisdom of the mob, in other words — leads to better decisions than the experts. This is counter-intuitive to most of us, but here and in later columns I’ll explain why this is the case.

Mobs Can Be Smart; Experts Can Be Dumb!

A few examples of how smart mobs can be:

• Crowdsourcing has helped find solution to tough problems that have resisted solution in the private sector. Innocentive has pioneered this model, under which companies pay fees for the ability to present unsolved problems to anonymous outsiders, an army of volunteers that Innocentive coordinates and pays for successful solutions. The company provides solutions anonymously in order not to bias the customer against the solutions, so solutions can be provided by literally anyone. Here’s a paper from Harvard Business Review examining how the Innocentive platform provides solutions to problems that stymie the private sectors best scientists. Jeff Howe’s 2008 book, Crowdsourcing, contains a fascinating account of Innocentive’s history.

• Innocentive’s “mob” is actually a collection of experts in various fields from all over the world, so isn’t this more of an example of bringing in outside expertise, rather than going to a true “mob”? Well, yes. But there are many examples of more traditional “mobs” being really smart in solving problems. An older example goes back to the 19th century: Francis Galton, a cousin to Darwin, came across a competition at a local fair in which people were asked to guess the weight of a live ox after it was slaughtered and dressed. The 800 or so guesses, when averaged, were only one pound off (1,197 vs. 1,198 pounds).

• Was this just lucky? Maybe, but there are many other examples. The “crowd” on the TV quiz show Who Wants To Be a Millionaire picked the right answer to questions posed to them 91 percent of the time — compared to only 65 percent of the time for the experts called upon. James Surowiecki, an expert on crowdsourcing and author of the seminal 2004 book The Wisdom of the Crowd, wrote: “With most things, the average is mediocrity. With decision making, it’s often excellence. You could say it’s as if we’ve been programmed to be collectively smart.”

• The Good Judgment Project, now in its third year, has found that volunteers, collected into small groups and asked to make difficult predictions about world events, are far better than the so-called experts.

Here’s a final fun example. A curious reader of Surowiecki’s book The Wisdom of the Crowd ran his own experiment on Google+ asking his contacts to guess how many Cheerios were in a strange-shaped vase. The median score from 436 guessers was off by only 17 pieces (the crowd’s median guess was 450 and the actual answer was 467).

On the flip side, who knew that experts could be so, well, dumb? A few examples:

• More than 95 percent of all managed bond funds underperform (yes, under) the market.

• Wharton professor J. Scott Armstrong found in a study of expert forecasts: “I could find no studies that showed an important advantage for expertise.”

• James Shanteau, an expert on expertise, found in his work that “experts’ decisions are seriously flawed.”

• Surowiecki writes in his book: “Heretical or not, it’s the truth: the value of expertise is, in many contexts, overrated.”

• Philip Tetlock, an expert on political judgment and self-described “expert on experts of experts,” concludes in his 2006 book, Expert Political Judgment: “Beyond a stark minimum, subject matter expertise in world politics translates less into forecasting accuracy than it does into overconfidence.” Tetlock also reviewed a number of books on how wrong experts can be in an excellent 2010 article in The National Interest.

There are, of course, certain individuals who can indeed be smarter than the mob. Warren Buffet, for example, has consistently beat the market in his investments in various companies. Another example: Nate Silver, a writer for the New York Times who has created his own forecasting methods for election predictions, has been incredibly accurate in his predictions. In the 2012 election he got almost every prediction correct, including the presidential, House, and Senate elections.

However, smart mobs were also very good at predicting the 2012 election outcomes. For example, the prediction market Intrade got most of the election calls (for the presidency and key swing states) correct.

With respect to stock-picking, some companies, including Piqqem, have pioneered crowd-sourced stock picking. It remains to be seen whether this approach can rival the likes of Warren Buffet, but time will tell.

What Makes a Mob Smart?

Mobs aren’t always smart. Mobs can be quite dumb, and historically often have been. What makes a mob smart, according to Surowiecki, are three important features: diversity, independence, and decentralization. Surowiecki writes: “Groups do not need to be dominated by exceptionally intelligent people in order to be smart. Even if most of the people within a group are not especially well-informed or rational, it can still reach a collectively wise decision.”

Diversity is important for collective intelligence because people who are considered smart are generally smart in the same ways and have similar backgrounds. This means they might not be as innovative in suggesting solutions to problems posed to the crowd. Intelligence is a famously slippery concept, but we can define it for present purposes as simply the ability to solve problems. Colloquially we think of intelligence as the ability to solve conceptual problems and people who are “smart” are those who are good with words or mathematical symbols and good at communicating. However, the array of possible problems extends far beyond this traditional conception of intelligence, and it is for this reason that diversity in crowds is so important.

Independence refers to the need for independent decision-making by members of the crowd. Interestingly, the more crowd members confer with each other, generally, the less wise the crowd becomes.

Decentralization refers to the lack of any hierarchy in coming to decisions or solutions for problems posed to the crowd. Hierarchical information flows and decision-making are antithetical to smart mobs because the opinions of all members must be considered equal for the averaging approach to work.

I found it very interesting in reading Surowiecki’s book, which doesn’t touch very much on the wisdom of crowds in democracies, that his three criteria for smart crowds (diversity, independence, and decentralization) apply well in the context of elections. A sufficiently large electorate will almost always naturally be highly diverse, independent, and decentralized. It is only when electoral districts are highly gerrymandered that the natural diversity of most of our communities is reduced, sometimes dramatically.

With respect to decision-making — which, of course, translates directly to the political context that is the topic of this column — Surowiecki writes: “There’s no real evidence that one can become expert in something as broad as ‘decision making’ or ‘policy’ or ‘strategy.’” While I resent this statement, since I am an alleged expert on energy policy, I must bow to the evidence that Surowiecki provides. I also recognize that elected officials, like lawyers, are asked to be experts in some manner on all things that come before them. This is, of course, impossible, as it is impossible for any person to become expert in all fields that concern a modern society.

The very point of enhanced direct democracy, or “wiki democracy,” however, is that it allows society to crowd-source expertise and decision-making. And this can be done, importantly, in a way that avoids the corruption, implicit or explicit, that comes with relying on industry expertise or entrenched politicians and staff.

The alleged experts in political judgment that we elect to make decisions for us in representative democracy here in the U.S., even if they were better than the crowd at political judgments, don’t have time to exercise that judgment well in many cases. Congressmembers, for example, spend so much time on fundraising that they have little time left for their real job — deliberating and voting on bills.

This is a big reason why most elected officials don’t even read the bills they vote on — they don’t have time. So why not crowd-source those decisions to those who do have time and passion to dig into the issues? Why not, as a strong first step in the direction of wiki democracy, allow regular people to comment and mark up bills that come up in Congress and state legislatures? Finland has gone further and allowed its electorate to propose entire bills in an online forum. If a bill gets at least 50,000 votes Parliament must vote up or down on the bill.

Don’t get me wrong: I highly value expertise. I’m an expert in my day job — energy policy. I often have to suppress the urge to shake those I’m trying to convince and tell them “Why don’t you get it? I’m making good arguments here, backed up by good data, and you still don’t agree with me?” At the same time, I also recognize the strong evidence showing the limitations of expertise, particularly in making value judgments and forecasting the future. Combine these limitations with my belief that autonomy and democracy are basic principles for good governance and you get my strong support for more, and better, democracy as the cure for what ails our current version of democracy.

Plato didn’t reject his era’s democratic system without offering alternatives. He wrote a whole book describing his ideal system, which involved a 30-year education for the best and the brightest that would eventually produce men who could offer enlightened rule to the masses. Ironically, the “philosopher king” model of ideal leadership that Plato described in The Republic can now best be found in collective intelligence, in the mob.

I’ll end with my own expert (ahem) prediction: We will see a blossoming of direct democracy in the coming decade around the world as the lessons of crowd-sourced wisdom and decision-making spread far and wide. It is far less certain whether we here in the United States will join this revolution in that same timeframe, a revolution that we arguably started in the modern era but have since abandoned due to misplaced fears about the limits of crowd wisdom.

Tam Hunt is trained as a lawyer and biologist and has studied philosophy for decades. He is a renewable energy consultant and lawyer by day, and avid reader by night. He also teaches part-time at UCSB’s Bren School of Environmental Science & Management. He lives in Santa Barbara, plays tennis, and strums a guitar occasionally.