The Stupidity of Crowds
Posted By Rob Millard - 0 Comments -
The September 2006 Harvard Business Review has a short article by Cass R. Sunstein titled When Crowds Aren't Wise. The article raises an extremely important point concerning how strategic and other decisions are reached, whether by consensus or majority decision, in profession service firms.
Conventional wisdom holds that crowds are "wise" and that all things being equal, the more inclusive a decision making process and the larger the number of people in a group making a decision, the more likely the "correct" decision will emerge.
In his landmark book on this topic, The Wisdom of Crowds, James Surowieki asserts that the conclusions of a large group can often be better than those of a small group of experts, simply because large groups can aggregate a large amount of dispersed wisdom.
Whether this is true or not depends at least in part on a phenomenon first described by the Marquis de Condorcet, one of the architects of the French Revolution and possibly the first person to apply mathematics with any systematic rigour to the social sciences.
Condorcet's Jury Theorum, as it is now known, asserts that where the average chance of a group voting will make the correct decision is greater than 50%, the chances of the group itself making the correct decision increases as the size of the group increases.
Consider, for instance, a coin that is very slightly weighted so that when flipped, the chances of "heads" is 51% and "tails" is 49%. If one flipped the coin five or ten times, the chances of getting heads more often than tails would be less than if one flipped the coin 1000 times. The larger the number of iterations, in this case, or the larger the number of people in the group, the greater the probability that the "correct" decision will be revealed.
All well and good, but in real life there are a few other factors that come into play:
1) The people voting must have sufficient information about an issue, to make an informed decision. Otherwise, one person's "truth" will differ even purely objectively from another's. Where the information is complex and an inappropriate level of investment would be required to educate "the crowd" adequately, then it may be better to dispense with consensual decision making altogether and go with the advice of an expert.
2) Not only must they have the information, but people must have the freedom and inclination to voice the correct decision. If self-interest suggests a different judgement, then human nature will skew the likelihood of a purely objective vote. Consider a (correct) decision to devote more resources to business unit "A" rather than to business unit "B." Even if they know that the decision is inherently correct, partners in business unit "B" are likely to be influenced by self-interest, against the correct decision. When many conflicting positions exist, as is common in many professional service firms, reliance on the 'wisdom of crowds' can cause the entire process to seize up.
3) The question needs to be unambiguous, or at least understood similarly by the whole group. This is so obvious as not to require expansion.
In formulating professional service firm strategy, it is very common to rely on the dispersed wisdom of the team, rather than detailed empirical analysis. A skilled strategist can quickly and accurately mine a group for a very credible range of strategic alternatives, then move on to developing the pro's and con's of each and finally deciding on which make the most sense. This method is far quicker and cheaper than commissioning detailed analysis.
It is obviously critical, though, to be mindful of the pitfalls that can make a crowd "stupid." To summarize, these are:
1) Relying on dispersed wisdom when that wisdom (not to be confused with strongly held opinions) simply isn't there.
2) Ignoring pressures that prevent the "wise majority" from prevailing.
3) Trying to resolve complex issues with simple "yes/no" decisions that do not adequately address the depth of complexity involved.
Finally, when it is appropriate to rely on the 'wisdom of crowds,' Condorcet's Jury Theorum dictates that the wider one throws the net in soliciting input, the better.
Comments, as always, are most welcome and may be posted below.
http://www.robmillard.com/admin/trackback/16442