[FoRK] Misunderestimating "The" Tea Party

Jeff Bone jbone at place.org
Sat Mar 20 14:06:04 PDT 2010

Re:  greater social dissatisfaction (on any significantly divisive issue) with greater-scale units of aggregation, Russell asks:

> Is that necessarily the case?

Yes;  that's the whole point.  It's a structural implication.  To see this, let's consider a simplified formal example.

Assume a group of people, a nation, called "Big."  Assume a divisive issue "A".  50.01% of the people in Big are pro-A, and 49.99% of the people in Big are pro-B.  Ignore geographic distribution for a moment as well as any other issues that might be in question.  If Big splits into "Little One" and "Little Two" and the former citizens of Big are allowed to choose which successor nation to belong to, and assuming that nations somehow have a social choice function that is intended to maximize the satisfaction of its constituents with respect to some set of issues, then clearly the factored state(s) result in greater social satisfaction with respect to A (~100%) than the monolithic state, which instead guarantees that about half the people are dissatisfied about A.

Now, if we introduce geographic distribution, it is only in the case of purely isomorphic distribution that dividing the state will result in no net increase in satisfaction.  But that's not realistic;  we know (cf. at least a couple of papers referenced on-list "recently," i.e. in the last several months) that opinions on a given issue tend to cluster geographically.  So dividing things up would realistically tend to increase satisfaction of any given with respect to the policy of their governmental unit with respect to A.

Furthermore, let's introduce multiple issues.  Again, we know (cf. referenced research) that opinions about multiple issues tend to cluster in both preference-space and geographic space.  So realistically, dividing into smaller geographically-clustered aggregation units and pushing authority for those issues down to the smaller units will tend to increase the net social satisfaction over a set of issues (at least, satisfaction with the unit in which one finds ones self or chooses to associate ones self with.)

The only remaining questionable assumption, then, is the reflection of preference at the national level through some social choice function.  We know the limits of democratic processes from Arrow;  but the above demonstrates that, given the ability to form small enough communities of interest, you maximize the effectiveness of the democratic process in ensuring social satisfaction by minimizing the unit of aggregation.  It's also worth pointing out that smaller units of aggregation, empowered to choose their own social choice functions and decision processes, would also allow greater concurrent exploration of social choice function fitness space.

This isn't a question of history, it's an algorithm problem.  Specifically, it's a set covering / tiling type of problem.  More, smaller, and potentially dissimilar tiles enabling greater coverage of the space.  (That's a simplification of course;  the space is n-dimensional.  But the point remains valid.)

Re:  the borders / free trade / immigration issues, your points are valid but those are independent concerns, easily (and more appropriately) dealt with independently.  Simply dividing into more, smaller units of social preference aggregation does not by itself necessarily imply greater friction in associating with the optimal social unit.  Furthermore, in the specific case of the US and delegation of authority to smaller units therein, it's reasonable to assume that this would not by itself increase any such friction.  Separation of concerns, please.  


More information about the FoRK mailing list