[This was first published in The Thought (July-August 2004). I have corrected some typos in this version and added a few passages to make this flow better.]
Some believe that the basic or primal unit of social order is the family. Others use this proposition as an argument for the necessity of government. They see the family as either government in miniature or as the highest level of stable social order possible outside of a society with government. In other words, an anarchist society would never get above the family level. Once government is superadded onto such a society, high levels of order would be possible. (See de Jouvenel 1993) I will consider the latter position here. Even if the family is the basic unit of social order, the argument that this makes governments necessary is weak at best.
Of course, families don't so much interact as individuals do. Thus, while one can look at the family as a real social institution – other examples include the firm, social clubs, and churches – generally people in wider (i.e., extra-familial) society interact as individuals. When I talk to a friend, e.g., I am the one talking – not my family. Yet those who claim that the family is a unit of social order need not mean that families are or act like natural individuals. More likely, they mean that individuals interact inside families. Certainly, this claim makes much more sense.
That people interact on a family level without the need of government should be the first clue that perhaps larger scale social orders or even different types of social order are possible without the need of government. A lot of other social orders are as simple as the family, such as friendships and even some trade relations. If Jeff swaps his stone axe for Jake’s rabbit pelt, it doesn't require an act of parliament for them to interact. If it does, then how did people get along before government?
Let’s not dismiss the idea too quickly. Instead, let’s examine the claim that government creates social order above the family level in more detail. The core argument goes like this:
1. Inside the family unit, people can interact without a government. This is because the family unit is small and simple.
2. Larger social units are too complex for rules and order to arise spontaneously. A weaker form of this claim is that even if larger social unit do arise spontaneously, there's no way to insure that they will be maintained or that they’re as stable as the family.
3. Government is necessary to overcome this problem of complexity by insuring a context in which people can be assured of certain social regularities. Absent government, no larger scale social interaction works in the long run.
Prima facie, the first claim seems obvious. In families, interactions take place face-to-face. Each family member, typically, has a strong stake in keeping the family going as well as in getting along. The members spend much of their development time in the family unit. Families usually represent their first contact with other individuals and often their only contact for many of their early, formative years. Size and complexity do vary among families; suffice it to say at least some families are small and simple.
The second claim is a little less obvious, but it appears to have some merit. Outside of family, individuals often have much less stake in any social interaction and the interactions are usually more limited. Motives tend to be harder to understand and there’s often less incentive to coordinate effectively beyond the limited goals of any organization. Moreover, when people do treat, say, a social club or a business intensively, usually people say “She’s married to her job” or “He's like family.”
There’s a problem with being too quick to dismiss other social units here. For one thing, some of the units last longer than families. Certain corporations have outlasted family units. Families, firms, clubs, churches, and even governments, of course, can all cycle through members. There’s a certain amount of fuzziness in drawing clear lines between these social units. For example, when a group of blood-related individual co-habitat a house, most people will call them a family. If one of them should move away, get married, and have kids, some will say that she or he has started a family. Others will say this is a new branch of the same family. In a way, they’re both right, but in the context of the core argument above, family as a basic social unit depends on a high level of interaction. It’s safe to say that when such interaction diminishes and ceases, the family has ceased to exist.
This does not solve the problem of a family that continues to exist for many generations. Imagine the case where the child is born into a family, grows up maintaining the high level of contact with the family by living in the same house, reproduces with her or his children staying in the family, and eventually dies. After a few cycles of this, none of the initial family members are still alive and the current members have had no contact with those initial members. Is this still a family or is it a progression of separate families?
One can avoid this problem by not going into too much detail, since the above problem applies to all social units above that of an individual. Corporations, clubs, churches, governments, and families all tend to have fuzzy boundaries. Yet just as one might not be sure exactly where the valley ends and the mountain begins doesn't mean all talk about valleys and mountains is nonsense. Let's leave aside the problem of strict delineation as it is not important to the argument.
Some of these social units have come into existence outside or in spite of government. Trade relations that predate government are a case of the former. Churches that cross national boundaries and are actually oppressed by some governments are instances of the latter. A less benign instance is organized crime. Further examples are languages, money, and social conventions such as when to have meals or when to sleep. Some of these are necessary before there can even be a government – a point I will treat below. It seems clear that at least some lasting social order does arise spontaneously above the family level.
The third claim also seems obvious. Governments, after all, produce some even if not all social regularity. For example, governments can mandate traffic rules. Noise ordinances can insure quiet times when most people are at rest. A host of other laws guide people in their lives.
However, the claim goes beyond stating the obvious to arguing about necessity. It’s not a casual observation about what governments do, but a claim about causal necessity. In other words, without government, there would be no social order above the family level. That languages, money, and other extra-familial social institutions arise is another clue that this argument has problems.
Let’s turn the argument around. It’s making the empirical claim that absent government, all social bonds and social order itself will breakdown – except for those at the family level – then there are two questions must be dealt with. One is how could government arise in the first place? If social order can never get above the family level without government, then there's no way for government to arise. How would such an extra-familial institution evolve in the first place?
Two, once government did arise, how would it stay together? Why wouldn't people or families blast it apart when it no longer served their immediate wants and desires? Why wouldn't government be a marriage of convenience, rather than a necessary social institution, ready to be destroyed or altered any more or less than any other social institution? In other words, why is social order above the family level ephemeral when it’s under a polycentric or anarchic legal system, but somehow permanent when they are in the context of a monocentric or government legal system? (See Leoni 1991. A polycentric legal system is one where this is no single legal authority. Anarchies are, by definition, polycentric legal systems – literally, systems with many centers of legal authority. A monocentric legal system is one where these is a single legal authority. In a monocentric system, the single authority is the government. These terms are not in Leoni’s book, but they’re implied by it.)
If the argument were true, then society would need a mechanism for social order other than government. Such does exist and it is basically customs and institutions that arise spontaneously through social interaction – as outlined by Hayek (1991) and others. However, if this is so, then government is not necessary for social order, since if it can arise and be sustained without government – and, in fact, government relies on social order for its very existence – then the raison d'etre for government must be found elsewhere if at all.
Some might still argue that constitutions are the extra mechanism that brings social order in via government. Constitutional government no more guarantees anything than custom. In fact, constitutions grow out of and are sustained by customs. Yes, they might change customs – the interaction is dialectical, but one side grounds the other, so it's asymmetric. Yet if one were to graft, say, the ideal libertarian constitution onto a society filled with people who did not respect private property, what would make them change? (If you want examples of that, look at almost any of the European colonies in sub-Saharan Africa. Almost none of them became pluralistic liberal democracies despite the fact that many Europeans thought that was their mission in Africa. No doubt, much the same is happening with the social engineering in the Balkans, Afghanistan, and Iraq.)
Constitutionalism looks like a failed gambit – or at best not the whole story to social order. Constitutions are only respected as long as they are convenient for those in power and do not clash too much with the general culture in a society. When they become inconvenient, constitutional limitations are ignored, changed, or reinterpreted. Look at, e.g., the First, Second, and Fourth Amendments to the U.S. Constitution. Take the First in particular. The U.S. Supreme Court has watered it down to the point that only certain forms of speech are now protected. Commercial and obscene expressions are not among these. In certain contexts, such as product labeling, people are not allowed to print what they wish. The same is now being applied to political campaigns.
This is not to say that in order to be useful constitutions must be perfect and eternal. That would be asking too much, but they are based on a lot of prior social ordering. If such social ordering can and must take place before a constitution works, then the argument that constitutional governments create social order as such is flawed. At best, constitutions made add to an already existing, elaborate social order. Typically, they codify existing ordering.
With the argument refuted, social order need not be married to government. This weakens the conflation of “government” with “order” and “anarchy” with “chaos.” Naturally, this does not mean that government cannot be an ordering force or that removing a government won't lead to chaos. There are better governments and worse anarchies. From a libertarian perspective, e.g., it would be better to live under a current welfare state in the West than, say, Somalia. But it is a false choice to equate the worst form of government-less societies with all forms of government-less societies. This is no different from equating the worst form of government with all forms of government.
* This brief essay is based on a public email discussion I had with Peter Taylor.
Works Cited:
Freidrich A. Hayek. 1991 [1988] The Fatal Conceit. University of Chicago Press: Chicago.
Bertrand de Jouvenel. 1993 [1945] On Power: The Natural History of Its Growth. Liberty Fund: Indianapolis, IN.
Bruno Leoni. 1991 [1961] Freedom and the Law. Liberty Fund: Indianapolis, IN.