On numerous occasions over the past few months I’ve heard people refer to Surowiecki’s “The Wisdom of Crowds” when explaining why any group driven project is inherently good.

My favourite has been the explanation regarding how citizen’s assemblies – because they tend to be composed of 100 or more members – are inherently wise and therefor produce a good outcome. To begin with, I find it interesting that those who defend electoral reform rarely talk about the merits of the proposal and instead refer to the soundness of the decision making process used to reach them. Citizen assemblies, it must be said, are not some magical process that produce inherently good outcomes. Indeed, if those who invoke his book had actually read it, they’d realize the Surowiecki’s analysis not only fails to support their contention about the process, it may actually do the opposite.

Pause for a second, and think about the logic that says a solution is good simply because it was arrived at by a large group of people. It is actually quite frightening. Indeed, one of the first things Surowiecki points out is that not all crowds are wise. The statement hardly needs supporting, but Surowiecki nonetheless trots out numerous examples of unwise crowds – angry mobs, investors in a stock bubble, and the various branches of US intelligence services. It’s not simply the size of a crowd that makes it “wise” it is also the rules that guide its behaviour. To be specific Surowiecki cites four key elements (which I’ve cribbed from wikipedia):

Diversity of opinion: Crowds – even those whose members hold ill-informed or eccentric interpretation of the known facts – will be wiser then groups that possess identical data, similar perspectives, or interpret data in a similar fashion.

Independence: Crowd members opinions aren’t determined by the opinions of those around them.

Decentralization: People are able to specialize and draw on local knowledge.

Aggregation: A mechanism for turning private judgments into a collective decisions.

Violate one of these elements and your crowd risks becoming a mob. By my estimation, citizen assemblies run the risk of violating three.

First, the diversity of opinion is at risk. While citizen assemblies’ proponents would have you believe they are composed randomly, this is in fact, not the case. Firstly, there are a number of people who, for reasons of work or family, would not participate. But what really interests me is that people opposed to change (and possibly those who are simply indifferent) are less likely to participate. If you thought a proposition (such as electoral reform) was silly, would you spend a year debating it – or would you simply await your opportunity to vote against it at the end of the process? Participation in these assemblies is almost assuredly tilted toward those predisposed to favour change – e.g. the crowd is more likely to be homogeneous in its desire for change and its perception of the electoral system.

Second, citizen assemblies are less likely to be independent. If you enter into a process to change the voting system the pressure to support change, any change, must be intense. Imagine if you sat around for a year listening, debating and arguing, and came out of the process agreeing that the status quo was ultimately superior to any other option. What a frustrating outcome! Tax payers would question why the money was spent and your friends would ridicule you for “wasting your time.” Worst still, what if the assembly couldn’t agree? What would that say about its constituent members? The internally created pressure on assembly members to put forward something, anything(!) new, and to have a clear majority of assembly members support this proposal was likely intense. I’d wager that once momentum for one solution began to emerge, other members were willing to bandwagon along “for the sake of the process.” In short, assembly members allowed their opinions to be determined by the opinions of those around them. (except, of course for those who held out. The same people who – from first hand accounts – were invariable referred to as stick in the muds and ‘difficult’ people. “Think like the rest of us or we’ll socially ostracize you…” isn’t that a sign of a mob?)

Finally, citizen assemblies have poor mechanisms for aggregation. Although neither the BC nor the Ontario Citizen assemblies required it, both placed strong emphasis on reaching consensus – articulating it as an ideal. If there is one system of decision-making Surowiecki believes makes a crowd dumb, it is a consensus-based approach. In order to reach consensus crowd members sacrifice the previous three elements – diversity, independence and decentralization – in order to gravitate towards the group’s mean. In effect the group’s collective knowledge and diversity of analytical ability is lost. This is the antithesis of a wise crowd. It is a crowd that actually gets dumber with time because it has less data, less analysis and fewer perspectives with which to assess the problem. It isn’t that people agree – they simply censor themselves to prevent disagreement.

This isn’t to say the Citizen Assemblies came to a bad solution for electoral reform (although to confess, I think in both BC and Ontario they did) . Again, all I wish to convey is that the citizen assemblies are not some magical process that produce inherently good outcomes. These process are neither democratic, representative, nor inherently superior – so don’t let supporters of the ballot initiatives bully you with process arguments. Let’s assess these proposals based on their contents – and what they do to democracy in Canada (which, in Ontario’s case, strengthens the parties and the backroom boys).