The Economist asks, “what happened to Somalia’s pirates,”

But the main reason for the drop in maritime hijackings seems to be that ships are now far better defended against attacks. Armed guards, now carried by more than 60% of vessels, have been essential in discouraging them. Pirates are playing it safe by first scouting for guards, whereas previously they opened fire to intimidate crews; seeing arms on board is a big deterrent. Higher cruising speeds in pirate-infested zones and rerouting also have helped, as have razor wire, high-pressure hoses and citadels—secure spaces on ships from which crews can call for reinforcements.

Also, The Economist links to a short blurb on mercenary naval forces, which shipping companies can hire to protect their convoys.

Historically, how does private policing compare to public policing? Gene Callahan answers this question by looking at a homicide rate time series for England, pointing out that roughly 800 years ago the crime rate was roughly “20 times today’s rate.” He notes that the decline in the homicide rate coincides with a continuous growth of the state. He admits that correlation is not causation, but he thinks it’s fairly clear that the correlation is not mere coincidence. I sympathize with the argument, but it suffers from an omitted variable bias, and in fact the number of omitted variables is probably very large. Further, what did directly cause the decline in crime could be things that would coincide with both public or private security. But, in the end, I say all of this as someone who believes that public security has provided society a net benefit.

I’m of the opinion that private security makes sense in a world where the distribution of force is relatively equitably distributed, such that the costs of violence are, on average, too high. This means that private security hasn’t been viable for most of human history, and it’s not obvious that it’s possible today. This implies that the state, so far, has most likely been the most efficient means of maintaining peace. While its present viability is not obvious, I do think the case for the superiority of private security is pretty high, but only by making one very likely erroneous assumption. I think the kind of justice and the means its enforced by would be very different from present day. An immediate change, I think, would cause a cultural shock. If we do away with the assumption that people will deal with this shock, then private security can’t work until institutions — including culture — develop to that point. This implies a transition over a large number of years.

But, it’s not difficult to see how Gene’s comment is not convincing to the anarchist. Consider the following possible causes of the long-term decline in homicides,

1. Better communication technology, allowing more crime to be reported to the authorities;
2. Faster means of transportation, allowing the authorities to more quickly respond to reported crimes;
3. Better, more expensive weaponry, giving authorities an advantage with economies of scale — individual criminals may not be able to afford them;
4. Growing economic wealth gradually increased the opportunity cost of crime, shifting the supply of crime to the left;
5. Better non-violent home protection systems, including homes that are more difficult to break into, home alarms, et cetera;
6. Criminals typically have non-random backgrounds, meaning that improvements in wealth also reduce the size of environments which are more likely to induce criminality;
7. Changes in weapon technologies make it easier for non-skilled users to fire more lethal weapons, making it easier to stop assailants (raising the cost of crime);
8. Various things that are extremely counter-intuitive that we may not recognize. I have in mind, as an example, Donohue’s and Levitt’s work on the relationship between abortion and falling crime rates. I understand that there are convincing arguments that show a number of problems with their methodology. But, my point is that there are things that are extremely non-obvious, and in fact until someone is creative enough to come up with a novel causal relationship we may even be convinced that there is no causal relationship whatsoever.

I can think of a lot of other plausible explanations for declining homicide rates in England. But, if you assume that private security is possible, none of these things are factors which are preferable to state security, but not to private security. That is, if we strictly model the security industries, the list is entirely composed of variables exogenous to the model. They are all changes with positive externalities, which affected the provision of security. If we assume private security was possible in 1200, then the time series may show exactly the same relationship. In fact, if private security is really more efficient than public security, then we’d see a superior trend.

There are a lot of anarchists that strongly believe private security is possible. If a lot of the reasons for a decline in crime are unrelated to actual institutions of justice, then it’s not difficult to see that the use of the trend line to argue in favor of public security is not convincing. I suspect that the issue that many anarchists have with Steve Pinker’s The Better Angels of our Nature. (In Pinker’s case, though, I think most are underestimating the book.) But, it’s not obvious that all increases in world peace have been caused by the state itself, rather than concurrent cultural, economic, institutional changes which bestowed upon society the benefit of greater peace. Maybe increases in the cost of weaponry, especially when we consider military-grade weaponry, induced economies of scale, by increasing the initial cost of warfare (weaponry, military training, deployment, et cetera). This forced governments to wage larger wars, but at much fewer intervals, and at higher costs (in price and, especially, in human life). These are all externalities that private institutions of justice would also benefit from, if they were possible.

It’s also worth considering that most modern-day anarchists don’t care about what was possible 800 years ago. They care more about what kind of institutions are possible today, or even in the near future. And, so, while there is a high probability that the state was absolutely necessary for most of human history, it doesn’t take away from the likelihood that none of the major factors that reduced the crime rate would cease to exist in a world of private justice. That being said, there are also many anarchists who believe that the state was never necessary. These people are even more likely to dismiss Gene’s argument, and even a weak version of Pinker’s. Although, to these people I’d show the differences in income growth rates between historical stateless, or relatively stateless, societies and societies with centralized governments. The latter, without a doubt, has, to date, shown the most possible results. In any case, what I’m trying to say with respect to Gene, is that it’s not clear that the evidence is so evidently in favor of public security.

On a tangentially related note, the case for the state became clearer to me in a short conversation with a professor on Daron Acemoglu’s and James Robinson’s Why Nations Fail. I mentioned my disagreements with the authors’ assumptions on the benefits of centralized states. He asked me if I knew what they had in mind when they made the claim, and in that exact moment I realized that the benefit they have in mind is that as the number of governments decrease, the probability of inter-state warfare falls. This also means that there are less governments to fight over extractive institutions, which may otherwise suffer from even stronger temporal rigidities. Maybe this already seems obvious to you, but I think the point was poorly articulated in the book. The authors, for example, cite tax increases as evidence of stronger, more inclusive political institutions. And, they fail to acknowledge that Somalia, during its years as an economically growing, virtually stateless society, actually did relatively well (in pace of improvement). They make other assumptions, as well. At one point they make the claim, at least in my interpretation, that relatively centralized governments are necessary prerequisites for vibrant urban trade centers. The claim was made with regards to early cities. But, maybe other factors — such as factor endowments, geographical location, et cetera — invited greater merchant activity, which in turn made extractive institutions like state bureaucracy possible. The way they present the case for centralized governments is not as strong as it could be.

But, look at the case for centralized governments this way. When there are many small geographical jurisdictions, with the assumption that the market is till too uncompetitive to fully discourage war and conquest through private security, the probability of war grows. As governments come to control larger quantities of land, forcing many governments to cease existing, there is a lower probability of border conflict, and war in general. In this sense, the centralized state can be seen as a net benefit. I feel that this is actually what Acemoglu and Robinson actually have in mind. I think it’s a strong argument.

Coming back to the main point on omitted variables, it’s not at all obvious that the decline in England’s homicide rate is directly attributable to the provision of public security. In all actuality, it’s more likely that the market for security benefited from a long list of positive externalities, mostly associated with economic and technologies improvements. But, there are strong arguments for the existence of the state, such as institutional concerns, that would enrich the debate if they were more regularly addressed.

Over the past week and a half I’ve been thinking quite a bit about possible liberal — pluralist — orders, and a thought occurred to me regarding the possibility of taxation in a world where all people have the ability to disassociate themselves from communities or associations. This will be probably just as controversial as my case against anarchy, but I’d like to think of this blog as being based on controversy — like Hernando Cardoso puts it, if all sides dislike you then you must be doing something right.

Imagine a world where all individuals can disassociate themselves from different polities. I don’t mean just the right to migrate in and out, but the ability to disassociate your property as well. So, if you were living in the United States and you wanted to disassociate yourself from Washington D.C., then your house (and whatever other property you have) would become sovereign. In such a world, if some polity were to levy a tax on you, you could respond by leaving that polity.

Would the average person do this, though? There are certain benefits to associating oneself with polities and groups. First, there is the division of labor, and by disassociating yourself you may also cause that polity to restrict your access to certain benefits of being a more direct member of that division of labor. Second, the political processes of the polity might confer advantages to you at certain points, even if at others there are costs (e.g. taxes). Wage workers, for instance, might create a polity — coming close to being a union, I suppose, although it would hold a geographic jurisdiction — and employers might have to associate themselves with this polity to access the pool of labor. The cost of association might be a tax, redistributing income from capitalists to wage workers.

The main concern with this theory is competition. In my above example, for instance, employers have an incentive to seek lower cost labor, and if the supply of labor is relatively abundant then it may be wage workers competing for jobs. More generally, there is always an incentive to reduce costs, and so polities that try to levy taxes might be forced to reduce, or eliminate, these taxes under competitive pressures. At the same time, though, we can’t just claim “competition!” and then assume that these possibilities aren’t likely. Not all costs can be eradicated, and if the benefits are particularly high then the provider will be able to charge more (non-price competition).

Another concern is that tax is “immoral” and “theft.” Since I’m a moral subjectivist, I don’t recognize this as being necessarily true. Morality is in the eye of the beholder, even if we all tend to hold similar basic moral principles. Nevertheless, I feel this sentiment is fueled by the fact that most libertarians probably believe that the costs of living under the state outweigh the benefits — which is true, since the state has to employ coercion to collect tax revenue (otherwise few would pay them). In future possible versions of a state, though, the costs and benefits of membership may be internalized to a greater extent, and those costs that are externalized may still be accepted as long as the benefits are great enough. A wealthy person may accept a relatively high (for that society) tax as long as she receives certain benefits, otherwise she’ll disassociate herself.

How is this differentiated from any other business? The benefits of these types of polities aren’t just security and/or justice, if these states would provide these services at all. Rather, they might be political processes by which people can publicly debate policies such as progressive (or even regressive) tax systems, basic income guarantees, regulations, et cetera. Even in the realm of justice, taxes in these societies could be used to pay for courts which specialize in dealing with cases that other institutions of justice can’t, such as where there exists transaction costs that make strictly voluntary means of solving externalities too expensive. These are all hypothetical examples, which may turn out to be wrong or right. The point is that it’s not farfetched to think of cases where something like the state, which levies taxes, might be beneficial — even in a world of free association. A political process might be useful to solve issues that the market may not be able to at any given point in time.

I don’t think we’d see this is on the scale of modern states. The scope of the state would probably be much reduced, as it recedes from the sphere of the market process and other areas where other institutions are better adapted for the kind of outcomes participants are looking for. My main purposes here is to provide an illustration of how something similar may still exist, even given extensive evolution in our institutions. The way most libertarians think of the state is as something deeply flawed that either has to be overthrown or very restricted based on rules which are relatively outdated (e.g. constitutional minarchists). I think this is the wrong way to think about it. States can simply evolve, retaining similar functions and forms, even if dramatically refined.

While I don’t know the history of the intellectual movement well, I think it’s safe to say that anarchism, broadly defined, is held in more serious regard in the present than it has been in the past. In recent decades, we’ve seen a lot of anthropological research of early anarchist societies and, at least it seems that, academically anarchist political philosophy has gained a strong base. Contemporary libertarianism, although maybe only in the circles I run in, also has stronger anarchist overtones than it did in the past, probably thanks in large part to Murray Rothbard. But, there are strong reasons not to be an anarchist. Instead, libertarians ought to be more open-ended and instead advocate the, admittedly vague, concept of continual institutional development.

As many faults as present governments have, including democratic institutions, they still provide an order. The sudden dismantling of these institutions would likely cause disorder. Instead, it may make more sense to assume a gradual transition to stateless societies. In this scenario, institutional change is more piecemeal, therefore the shock to order wouldn’t be as disruptive. We can even think about this gradual transition as a perpetual process of institutional competition, where orders arise simultaneously and then are chosen against each other. There might be disagreement about how difficult it would be for society to create the necessary anarchic institutions, but we nevertheless should all more-or-less agree on the necessity of transition.

Most libertarians probably believe in the theory of spontaneous order. But, in spontaneous order ends are chosen by all individuals, and therefore it represents a plurality of opinions, preferences, and beliefs. And, in fact, we don’t actually know what the aggregate outcome will be. An important tenet of the theory is the limitation of human knowledge, implying the inability of one mind to develop complex institutions. In other words, not only is the range of potential outcomes broad, but we don’t know even know what some, or most , of these are.

Doesn’t this present libertarian anarchism with a dilemma? How can we argue that anarchy is preferable to all other possible outcomes, while admitting that we don’t even know what these alternatives are? It could be that a system of political governance is developed, vastly superior to modern day democracy, leading to a different, unique path of institutional evolution that may completely bypass the “anarchy phase.” Alternatively, maybe society will progress into anarchy, but thereafter transition to something else entirely. Advocating anarchy, in other words, is arguing that there is one end that we all ought to work towards, and this ideology, in my opinion, is opposed to liberalism (which is a pluralist philosophy).

Anarchy does have its appeal. There is a substantial literature on anarchic organizations and institutions, from free banking to broader work on anarchic institutions of justice. Much of it could very well be more-or-less right. But, the actual value of this work isn’t in proving the absolute superiority of anarchy, rather its use is in illustrating that there is good reason to believe that we can do better than the present. It helps reduce some of the uncertainty attached to unplanned order. However, it should not be interpreted as this “spontaneous” dialectic, if you will, that somehow must lead to statelessness.

I suppose that one could argue that anarchists aren’t actually advocating for “The” anarchic society, but rather imply a broader concept inclusive of all possible stateless societies. Yet, then we should consider that the term is then an empty concept (apart from the above-mentioned fact that it’s also possible that there will always be a state — even if it looks radically different from present day ones). The institutional possibilities becomes so broad that the term “anarchy” is hardly useful for knowing anything at all about these institutions, apart from the very general and vague concept of “statelessness” (property is territorial monopoly, as such there is an entire spectrum of possible configurations between completely individual sovereignty and the modern state).

Nevertheless, a legitimate argument may be that there is a strong possibility that anarchy is where we’re headed. In this sense, it’s okay to argue a strong feeling in favor of anarchy. In the back of our heads, however, we should always be aware of our fallibility. As Hayek argued, liberals — libertarians or otherwise — shouldn’t be “rationalistic individualists.” We shouldn’t accept theories which necessarily assume a common goal that we’re aware of and understand. Instead, we should accept the limits of our knowledge and instead restrict the range of choice to the area that each, unique, individual — or organization (e.g. the firm) — has enough knowledge on. This process of restriction doesn’t arbitrarily and/or suddenly come into existence, but rather that it has to be developed over time and is therefore necessarily imperfect.

Admittedly, this kind of open-ended individualism can be relatively unappealing. First, it’s hard to come up with a romantic name for it. Second, it invites criticism of nihilism, but even if it were true I’m not sure being skeptical (even very skeptical) is a shortcoming. Third, it forces us to admit that we have no idea where society is headed, undermining our ability to dogmatically oppose concepts we dislike or are uncomfortable with. In some cases there are legitimate arguments against political institutions we know and understand and our opposition to these is sensible, but these represent a minority of potential institutional outcomes. Fourth, it also leads us to recognize the probability of regressive outcomes. The process of spontaneous order, like all complex outcomes of human action, is imperfect too. In this case, though, I think the unappealing position is the most coherent one to take. Otherwise, I’m not sure how you can reconcile the plurality of liberalism with the homogeneity of “The” objective — it ceases to be an individualistic political philosophy.

Addendum: Thank you to whoever linked to this post on Reddit! The first comment on that thread argues that “spontaneous order is anarchy.” Anarchy in what sense? The state is the outcome of a spontaneous order. Is the state, therefore, an anarchic institution? Again, I’m concerned that “anarchy” is becoming an empty concept.

Just a quick thought that springs out of yesterday’s discussion of democracy. In a comment responding to Roberto, I wrote that I see anarchy is a significant increase in the degree of pluralism over democracy and that such a system of governance would achieve a more equitable distribution of power. What does this imply with regards to power relations? Two points,

1. I don’t think the existence of hierarchies of power are inconsistent with an equitable distribution of power;
2. That power can be equally distributed doesn’t mean that an equal distribution necessarily prevails.

In his article “The Problem of Social Cost,” Ronald Coase suggests treating property rights as goods that can be bought and sold. This was his way of conceptualizing the idea of bargaining to resolve property disputes, when the costs of bargaining were low enough to make it possible. I think we can talk about power in the same way. Power influences the ability constrain another person’s range of choices. An equitable distribution of power implies that the playing field is level, or that everyone has equal ability to skew others’ range of choices — this might mean only that you can protect yourself from others’ attempts to limit your choices. But, an equal share of power isn’t always desirable.

Someone might want to sell power in return for something else. Wage workers do this when they sign contracts with their employers, giving the latter the power to use the former in ways stipulated in the contract. These are voluntary power relations. It doesn’t make sense to see these as “unfair,” since they come about because they’re desirable given the conditions of that time and place. The same is often true within families. When I lived with my dad I didn’t enjoy him bossing me around, but I essentially sold him that right to exercise power in return for his welfare. The distribution of power between me and him was unequal, but I found this situation “ideal” nonetheless.

I can’t think of a short, accurate way of summing up my point. Maybe this will do: broader distributions of power imply more equitable access to power, but this doesn’t necessarily mean that society will distribute it equally between themselves — oftentimes, people sell power in return for goods they attach greater value to.