Dunbar’s number

We have in the past discussed the issue of the desirable size of an allotted chamber (or more generally, the size of the set of decision makers). The two contrasting constraints are the need for representativity on the one hand which demands the chamber is not too small, and the need to avoid mass political effects which demands that the chamber is not too big.

One important factor which determines the size at which mass political effects become influential is the ability of the group members to have face-to-face social interactions. Once group members are unable to interact with others personally, the system becomes opaque, promoting new ideas becomes increasingly difficult for the average member, and power brokering emerges. In such a situation, power is no longer equally distributed in the chamber.

It turns out that the maximum size of a group at which personal relationships are still possible has been termed “Dunbar’s number“, after British anthropologist Robin Dunbar, who theorized that

this limit is a direct function of relative neocortex size, and that this in turn limits group size … the limit imposed by neocortical processing capacity is simply on the number of individuals with whom a stable inter-personal relationship can be maintained.

According to Wikipedia, the value of Dunbar’s number has been estimated to be in the range of 100 to 290. Importantly, Dunbar’s number depends on factors such as the proportion of time spent on maintaining the social relationship. Dunbar estimated that a group of 150 people would have to spend 40% of its time on maintaining personal relationships.

Presumably, with the high stakes and concentrated resources of a state parliament, the number can be pushed somewhat higher, but for lower powered chambers, the group size would have to be significantly lower for personal relationships to be sustainable.

Advertisements

11 Responses

  1. The relevance of the Dunbar number also depends on the structure and function of the allotted chamber.

    Legislatures typically organize into committees of manageable size, even if the plenary sessions are huge. Of course, small committees would lack the descriptive representativeness desired, and within plenary sessions members would tend to rely on cognitive short-cuts, rather than deep understanding of committee reports.

    If an allotted chamber is going to act like a traditional legislature (initiating, drafting, debating, amending and adopting laws), the Dunbar number is an important consideration. But if it is to function as Keith Sutherland advocates (merely hearing pro/con presentations, and voting without debate among members), the Dunbar number doesn’t matter, as there personal relationships aren’t involved. In this case the relevant number is the size of a body at which members lose the feeling that their individual power is so small, it no longer matters if they pay attention, or put in the effort to make the best possible decision. I don’t know what that point is.

    Like

  2. I do not think personal relationships Dunbar-style is necessary between all the participants in an allotted chamber. Nor would it be possible, since people vary widely in their ability to form personal relationships to 50+ strangers on demand.

    As long as people can reguarly meet and talk in smaller groups that overlap and aren’t all self selected, I think the quality of the communication will be good enough.

    Like

  3. Terry,

    > Legislatures typically organize into committees of manageable size, even if the plenary sessions are huge.

    If important decisions are made in committees, while plenary sessions (and other all-to-all interactions) are too cumbersome to allow efficient transmission of information then influence in committees becomes a power resource that is unequally distributed.

    > But if it is to function as Keith Sutherland advocates (merely hearing pro/con presentations, and voting without debate among members), the Dunbar number doesn’t matter, as there personal relationships aren’t involved.

    Right, but then – as we have discussed many times before – most political power is held by those who write the proposals and prepare the “pro/con” presentations (these presentations could simply be meaningless facades obscuring the real issues – as in mass media political punditry today). In that case, the question of the size of the vetting chamber is of relatively little importance since its powers are very limited.

    Like

  4. Harald,

    > people vary widely in their ability to form personal relationships to 50+ strangers on demand.

    The idea is that the chamber members would not be strangers, but would become familiar with each other well enough to understand their positions, motivations, character, etc. The service period has to be long enough so that this is not an “on demand” situation.

    > As long as people can reguarly meet and talk in smaller groups that overlap and aren’t all self selected, I think the quality of the communication will be good enough.

    I am not sure what you mean by that. The condition that you describe would be fulfilled if we set up, within the existing electoral system, randomly selected discussion groups in which all citizens are invited to participate and then allow those groups to meet regularly and talk about political topics. Wouldn’t you agree that simply convening such groups would not lead to a meaningful democratization of the system? The huge number of the groups would make any discussions and any conclusions made in these groups completely inconsequential.

    Like

  5. Yoram: “meaningless facades obscuring the real issues”

    Hmm, no doubt the former might be labelled superstructure and the latter infrastructure, but Peter has banned me from asking you exactly what you mean by such a distinction.

    Like

  6. Did I do that? Man, I must be a real bastard…

    Two sources and then a quick comment. First, Malcolm Gladwell as a rather extensive discussion of Dunbar’s Number in his book BLINK. He uses the firm Gore-Tex as an example; supposedly, the company is internally organized in teams so as to take advantage of the number. Second, Condorcet made a different argument regarding the relationship between size and competence. He argued that the larger the group, the higher would be the probability that it would decide correctly via majority rule for a given average competence level (that’s just his famous Condorcet Jury Theorem in action), but the lower would be the average competence level. I think he rested his case upon elitist considerations (i.e., there just aren’t enough smart/educated people out there), but one could just as easily make the argument on grounds of rational ignorance, etc. And this is an argument that doesn’t depend upon any interaction effects between group members.

    Now for the comment. I think we all agree that in any large legislature, influence will not be distributed equally. Not everyone will talk the same amount, not everyone will be equally persuasive, etc. Even Yoram and Keith agree on this, I think! And so one important question to ask is whether we want to acknowledge this fact institutionally, and therefore control for it. Failure to acknowledge it can lead to the so-called “Tyranny of Structurelessness,” whereby everyone pretends the situation is egalitarian while the real power-holders wield their influence unscrutinized. See http://flag.blackened.net/revolt/hist_texts/structurelessness.html

    Like

  7. Yoram,

    If the entire population was invited to such groups, that would cause the usual problems with rational ignorance.

    Either way, I don’t suggest such groups as a replacement for public speaking to a larger assembly, only an addition to it.

    I’ve spent many years in schools and workplaces with less than 200 people, and I have never managed to even learn all names, much less get any good idea of who all those people were. I don’t think I’m unusual in that. It’s in small groups you get to know people.

    Like

  8. > If the entire population was invited to such groups, that would cause the usual problems with rational ignorance.

    Exactly – so the opportunity to communicate within a small egalitarian group is only useful if that small group is powerful enough to motivate active participation. Having many small subgroups within a large allotted chamber would therefore not serve to overcome the problems of mass politics.

    > I’ve spent many years in schools and workplaces with less than 200 people, and I have never managed to even learn all names, much less get any good idea of who all those people were.

    I agree – familiarity doesn’t occur spontaneously. The members of the group must have incentives to become familiar with each other. In addition, the appropriate organization must exist and necessary resources must be available.

    Like

  9. Peter,

    > one could just as easily make the argument on grounds of rational ignorance, etc. And this is an argument that doesn’t depend upon any interaction effects between group members.

    Not exactly – it seems there is a close connection between the interactions effects and the rational ignorance effects. As long as someone believes that they have the potential to influence the entire group then it is irrational for them to be ignorant. Any understanding they gain by investing time and effort (and other resources) could be transmitted to the other members of the group and thus influence the group’s decision.

    > I think we all agree that in any large legislature, influence will not be distributed equally.

    Yes. The severity of the inequality depends on the size of the group, on institutional design and on cultural norms. It seems to me that if the group is larger than the Dunbar number then inequality would be both large and insurmountable, unless sortition is used. When the group is small, on the other hand, appropriate design and culture could provide a close approximation to equality of power.

    > Failure to acknowledge it can lead to the so-called “Tyranny of Structurelessness,” whereby everyone pretends the situation is egalitarian while the real power-holders wield their influence unscrutinized.

    I agree. This is the situation we have in Western societies where the elections-based system is pretended to provide equality of power, while in fact the electoral elite – colluding with other elites – dominate public policy unchecked.

    Like

  10. > Did I do that? Man, I must be a real bastard…

    Have you been squelching free speech on the Internet again?

    Like

  11. Yoram: “When the group is small, on the other hand, appropriate design and culture could provide a close approximation to equality of power.”

    Since the so-called ’empirical turn’ in deliberative democracy, political scientists have tried to address these inequalities by increasingly sophisticated methods, such as counting the number of words that each deliberator is allowed. I’m not sure if they have got round to constructing a veil of ignorance behind which to conceal each speaker. Perhaps they could borrow Stephen Hawking’s speech simulator (I gather he’s in need of a new one) to equalise the rhetorical impact and employ a parliamentary speech writer to equalise the discursive content of each speech act (with appropriate limits on the number of adjectives allowed to each speaker). What to do with the stupid people and those who have nothing to say might well stymie their efforts, but where there’s a will I’m sure they’ll find a way (given the necessary research funding). Similarly with the police, who are going to find themselves very busy keeping tabs on the army of lobbyists queuing up behind their chosen ventriloquist’s dummy.

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: