On this recent discussion, I brought up the notion of an alternative form of government which would have advantages over democracy. The general concept was:
Government without territory. Consider government as something we would subscribe to in the same manner as long distance phone service (or insurance). Real consent of the governed would imply that you choose your own government -- i.e. that courts, police, etc. would be services that you purchase not institutions that one must automatically accept.
I was, of course, rather quickly accused of being an anarchist. It is true that protection (a general term I will use for the services provided by government) is typically a service that we assume will be provided by a single entity for any given geographic area. We typically assume that there should be the police, the court system or the army.
Oddly enough, this is rarely actually the case -- and historically few human beings have ever lived in a society where such services are provided by one entity and even fewer where all three of those protective services were provided in totality by one entity. For example, we already live in a world where some law enforcement functions are held by local government, some by the state government and some by the federal government. There is also significant overlap, where the same offense can be potentially punished by multiple entities. Arbitration of civil cases can be conducted via a private arbitrator or through several layers of state courts. Even military functions are distributed (to a degree) amongst national guard, formal military and many defensive functions are shared with customs, coast guard, border patrol and the TSA. Clearly, it is possible and even commonplace for defensive services to be provided by multiple entities in the same geographic area and it is even possible for their respective jurisdicitions to overlap.
(As an aside, the suggestion that all such services be provided by a single entity as would occur under martial law is not only unlikely to be an improvement -- it is almost the definition of bad government.)
So, the idea that we can have competing agencies of protection is not all that radical. It already occurs to some degree, and could potentially occur to a greater degree. I wish to propose the theory that increasing the degree of jurisdictional overlap will confer the same sorts of benefits that competition brings in other industries.
The primary objection to this thesis, no doubt, is that protection is "special" -- that it is a service of a different nature from other services. However, it is impossible to maintain that protection cannot be provided in a competitive market. The existence of private security, arbitration and militias both historically and in the modern world make it clear that such services can be provided by entities that compete for market share rather than have their jurisdiction defined by force or statute. The objection cannot be that protection can't be provided in that manner -- just that it is "bad" to do so or at least better to provide them within rigidly determined jurisdictions.
The primary argument against jurisdictional overlap is an efficiency argument. If you have multiple agencies providing protection that it is possible that there will be more resources devoted to protection than are actually needed to keep the peace. For example, if a neighborhood has 20% of its households purchase security provided by Agency A and 80% by Agency B -- both A & B must patrol the neighborhood when it would be better if one or the other did. This may be true, but there are several simple solutions if this truly proves to be inefficient. Both agencies have an incentive to increase their market share to 100% and will thus compete in price and quality to achieve that end. Second, the homeowners themselves may decide to form a covenant whereby the whole neighborhood purchases a bulk security package from a single agency. Third, the homeowners and/or the agencies can coordinate to achieve geographic contiguity between the two portions of the neighborhood (i.e. so that all Agency A customers are on the west side of the neighborhood). It is, of course, possible that the homeowners don't care about this inefficiency and neither do the agencies. In that circumstance, one has to wonder why we would want to impose a more "efficient" system upon them if they clearly don't want it.
An interesting thing to note about the competitve police services above is that the system overall has tendency to err on the side of extra security. Contrast this with the enforced monopoly of a municipal police department where there is a tendency toward insufficient security (because there is no way to fire a city police department). To use terminology from economics, the externalities of competitive protection are more likely to be positive, while the externalities of a forced monopoly on protection are more likely to be negative.
While I would advocate the fully competitive model described above, it is quite possible to propose a moderate system that is halfway between what we have now and the fully competitive model. Let's say that a given city has 40 police precincts covering 40 distinct regions of the city. We could have a system where each region has a local board which collects funds (let's assume property taxes) to fund the police precinct in that region. Each year (or every few years) the board will hire a single agency to provide police services for that precinct. If an agency underperforms it may be replaced by a different competing agency. Agencies would thus be encouraged to compete by demonstrating more effective provision of police services. The fully competitive model that I described above differs from this system primarily in that the precincts themselves would be determined by market forces.
Similar arguments could be made for courts or civil defense, but I don't think it is necessary here to make an argument for competitive provision of every kind of protection. Rather, it is merely sufficient to demonstrate that such an arrangement is possible and preferable to current systems. Just because we have always* provided such services in a particular manner doesn't mean that it is the best way and it certainly doesn't mean it is the only way.
It is entirely possible that the "anarchist" accusation stems from concern as to what would happen if all protection was provided in a competitive manner. It is entirely possible that current technology and/or social development only allows for a certain degree of protection to be provided in a market-driven manner. But that question is rather academic since it is unlikely that we would achieve a fully competitive protection market overnight. Rather, such competitive elements would gradually be implemented as they were seen to be more effective. If we reached a point where this evolution needed to stop (temporarily or otherwise) it would be fairly easy to recognize those problems as they arose and act on them. In other words, if the system got too loose, a more traditional protection system would likely be implemented in response.
I welcome any questions or comments (so long as they are civil).
* Actually, many of the protection arrangements that we consider "normal" are recent inventions. For example, city-wide police forces are, for the US, a late 19th century invention.