I said I’d hoped to have some thoughts about Part II of The Problem of Political Authority and I do have some reactions. I have a soft spot for anarchism, and I suppose I was an anarchist for some time in my youth (New Left anarchist, quite individualistic). I think the fact that political philosophers don’t take it seriously is deplorable. At the same time, I find the case for anarchy rather weak, and the second part of MH’s book is no exception.

It’s easy to show that anarchy is possible. We can describe a possible world where there are no states, government, and legal systems (in something like Hart’s sense). And we can also tell a story about how such a world could emerge and maintain itself. Contemporary game theory even provides the elements of the story: in repeated games with large numbers of players there are multiple outcomes that are stable (i.e., Nash equilibria); indeed, just about any outcome can be stable. This is the folk theorem (potted version). So there should be such a story for anarchy. But are we seriously to believe that anarchist communities could emerge and sustain itself in our world? In either the last couple of centuries or the next fifty or hundred years? I am quite sceptical. MH does not claim that anarchy could come into existence in any or even many circumstances; he only wants to show that it is feasible under some realistic conditions, where most people are anarchists. He may be right, but I wonder whether some of the feasible and desirable forms of anarchy might not resemble some of our polities (e.g., small states with little coercion, medieval city-republics).

Concepts. I am a fussbudget about concepts. MH is not. “Semantic questions about the use of ‘government’ and ‘anarchy’ are of no great importance.” (p. 232) I complained earlier about conceptual problems, and Part II really worries me in this regard. Distinctions are needed if we are to assess the merits of anarchy (see the thoughtful comment by ScottA under my first post). We need to know what anarchy is. Voluntariness and competition (p. 232) are not quite enough; some small political societies might satisfy these conditions and still be state-like in some important respects. As I noted in my first post, a distinction between ‘state’ and ‘government’ is needed. The government in many systems can fall with a vote of confidence in parliament, but the state remains intact. The French and Iranian revolutions changed a lot, but they did not put an end to one state and give birth to another. Why are distinctions needed here? For one if a state like France can undergo a serious revolution and be in its fifth republic, then states may be much more durable than governments. They may also be the locus of legitimacy.

In addition, it’d be useful to distinguish between states and empires. The two aren’t the same, and empires are much older forms of political rule, preceding the emergence of states. And the longest-lived empires were much older than any state. Some empires were more tolerant of “diversity” than other polities of their time. In his discussion of war MH notes, with some hope, “The decolonization movement of the twentieth century shows it is particularly difficult for a foreign state to keep control of a territory in modern times.” (p. 294) The “modern times” had better be post-1918 if this statement is to be true. Note that the colonial empires referred to were ones where the “core” did not or could not integrate “foreign” peoples into its polity. Many European states did precisely that with the peoples inside their current state borders. With the exception of the Corsicans and the Basques, the French state stomped out regional languages and cultures; that’s an effective way of making possible rule by foreigners once the age of foreign monarchs was over (e.g., the European monarchies, the House of Hapsburg). How can we judge the effectiveness of states if we do not know whether Rome was a state, or whether China was a couple of centuries ago? (Neither were.) How can we take heart from learning that “Vatican City has never been at war.” (p. 296) Vatican City is just the remnant of one of the greatest empires of Europe – or perhaps two, if we include Rome. That’s a bit like saying that Saddam Hussein was peaceful once in chains.

MH needs distinctions, and they do slip in from time to time. In a discussion of recent events in Egypt he notes that “Mubarak was expelled from office as a result of a predominantly peaceful protest movement.” (p. 293) Mubarak is said to been part of a corrupt regime, his government to have many members, and his administration to collapse suddenly. It is hard to make sense of political events without these concepts. Note of course that the Egyptian state is still in existence. (I add a political comment: the expulsion of Mubarak could hardly be attributed to protest movements; it was a military coup d’état.)

Given what I have said, I don’t think that “state” (or “government”) and “anarchy” exhaust the possibilities. If that is the case, an argument against one does not establish the case for the other. MH does not quite do this, but of course early modern political thinkers often made the case for the state in that manner. The founding father of modern political philosophy, Thomas Hobbes, argued for the state by arguing against anarchy. He tried to define them so that the two logically exhausted the possibilities. MH says that “The standard solution in social philosophy begins by proposing a radical inequality: a single institution with power over all other individuals and organizations. For Hobbes, the solution ends there.” (p. 228) I agree this is a bad way of thinking about possible ways of organizing political life. But I think finding better ways require more distinctions; we need a clearer idea of the alternative forms of political organization. There may be more choices than “government” or “anarchy”.

Constitutions. I am a fan of the 9^th and 10^th amendments to the (second) American Constitution, and I am quite sad they have been ignored. But if we are to evaluate how successful the American constitution has been, we need to know what it is? For MH a constitution seems to be something like the American document that bears its name. The question then arises, “Who will enforce the constitution?” This question answers itself: “No other organization has the power to coerce the government. Therefore, we will have to rely upon the government to enforce constitutional restrictions against itself.” (p. 221; see also p. 335). Some questions: does a “constitution” need to be enforced? Is there no agent other than “the government” who could enforce it? And, of course, what is a constitution to begin with? Take the US Constitution: does it consists of the text approved or the text + the first ten amendments? The tradition says the latter. This document needs to be interpreted; are the interpretations in some way part of the constitution? Many think so. Can the ignored 9^th be said to be part of the Constitution any more? I know that MH’s main point is that the constraints on government and federal power in the US Constitution have proved ineffective in a number of ways. I think he is largely right. But the idea of a constitution is not examined, and this has consequences for our understanding of state and anarchy.

The older idea of a constitution points to the structure of a polity. On Hart’s theory this might be something like the set of secondary rules, in particular the rule of recognition. I mention his account, as the important rule of recognition does not need to be enforced; it is essentially a convention among officials. Other accounts would have constitutions be self-enforcing or, more technically, coordinative equilibria (e.g., R. Hardin. B. Weingast). If a constitution, in the important sense of the structure of the polity, is self-enforcing, then that should have some implications for the project of evaluating anarchy. It might even be a fact that lends some support to the anarchist. And it might show that many states depend less on coercion than is thought.

The case for anarchy. The libertarian tradition has distinguished itself in pointing to and devising all kinds of ways of doing without states. And it is to be regretted that that this literature is not better known outside of libertarian circles. The first job of states is to provide security, and the biggest problem facing anarchy seems to be security (see David Friedman’s little book on anarchy). Robert Nozick’s book introduced many political philosophers to protective agencies, and we are now all familiar with competition between such agencies. MH thinks these agencies will be a more promising way of providing security than states. His arguments are not bad, but it is hard to evaluate the case for or against. On one hand the casualties of states are not negligible; in the twentieth century they are breathtaking. On the other it is hard to be optimistic about the prospects for anarchist communities. MH thinks that competition between protective agencies will have a moderating effect; he speculates that “it is unlikely that a protection agency would wish to staff itself with” murderers (p. 235). I note that many do in fact do just that. It’s not just that big law firms hire bullies with legal training; the mob employs murderers. If one’s business is predation – call it “protection” – then competition may not help us very much. In many part of the world, clans and their “protective agencies” battle away, none of much inspiration to lovers of liberty and the result horrific.

As I argued in my book on the state, the history of the world has been a running experiment regarding anarchy, with results hardly encouraging. (The short section is entitled “History of the World”, and I have just reread it and recommend it!) Thousands of years ago all humans lived in anarchist communities. Today the only thing close to them are semi-anarchist communities like kibbutzim, most of them sheltered by states. Forms of centralized governance replaced them, culminating in the modern state. The latter’s history is quite remarkable. In medieval Europe multiple forms of political rule co-existed, with overlapping authority. In early modern times the state started replacing these. Older “feudal’ structures, independent cities or city republic, leagues of cities, federations, empires were conquered or assimilated – and the Church, a power over kingdoms and other polities, was weaked. Today virtually the entire landmass of the globe is the territory of a state (or what is misleadingly called a “nation-state”). That’s as close as we can get to a controlled experiment. In fact, it’s better than a controlled experiment. We want to know how anarchy would fare in the world of humans, and we do know: anarchist communities get gobbled up. Americans should know this better than many, as we are well aware of the fate of the multiple communities that preceded us on this continent.

This is all to say that I did not find Part II entirely convincing. MH makes a good attempt to show the feasibility of anarchy under favorable conditions. Appealing as the picture may be, it’s not clear it has much importance for our world, where anarchist communities can’t last long. More importantly, we need to know what sorts of forms of political organization are both reasonably favorable to liberty and possible for us. And this task requires more analysis than that provided.