8. Evaluating Social Theories

1. General observations on the rational
   evaluation of social theories
1. Rational evaluation is comparative
2. Rational evaluation is comprehensive
3. Varieties of government and anarchy
4. Against status quo bias
2. A Simplified conception of human nature
1. Humans are approximately rational
2. Humans are aware of their environment
3. Humans are selfish but not sociopathic
4. n behalf of simplification
5. A historical application
3. Utopianism and realism
1. The principle of realism
2. Prescription for a realistic anarchism
3. Against utopian statism

In the chapters to follow, I ask the reader to consider a broad theory of how society ought to be organized. Before explaining this theory and the arguments in its support, it may be helpful to discuss how theories of this kind ought to be evaluated.

8.1 General observations on the rational evaluation of social theories

8.1.1 Rational evaluation is comparative

We often decide whether to adopt a course of action by asking simply whether the action is good or bad. But the more appropriate question is whether the action is better or worse than the alternatives.[1]

Suppose that while driving my car, I see a dog in the street. I have the option of running over the dog. Should I do it? The answer depends upon what my alternatives are. If I also have the option of stopping the car and waiting for the dog to cross, then I should not hit the dog. But what if my brakes have failed, and I cannot stop the car? What if, furthermore, there is a child playing on the sidewalk to my right and an oncoming car in the lane to my left? My only options are to hit the child, hit the oncoming car, or hit the dog. In that case, I should hit the dog. This option is not good, but it is better than the alternatives.

The same point applies to the evaluation of social theories, where the relevant question is not whether some social structure would be good or bad simply but whether it would be better or worse than the alternatives; that is, the other social structures that we could adopt. This point may be obvious, but it bears stressing, because it is easy to forget in practice; we often criticize or defend political proposals without considering what the alternatives are.

A corollary is that our standard for evaluating social theories ought not to be one of perfection. We ought not, that is, to reject a proposed social structure because under it, some people will suffer from some social problems. Perfection is not one of the available options for human societies. We ought to reject a social system if and only if we can identify a superior alternative.

8.1.2 Rational evaluation is comprehensive

When evaluating a social system, we must consider all of the advantages and disadvantages of that system. We should not allow our evaluation to be unduly influenced by any single social issue.

Imagine a social activist, whom I shall call 'Mom', whose favorite social issue is drunk driving. She attends rallies against drunk driving, quotes statistics about the problem to her friends, and writes letters to her congressman and the local newspapers urging tougher laws to combat drunk driving. Mom knows that there are many other serious social problems; nevertheless, none call forth from her the same passion. This sort of phenomenon, I trust, is familiar enough; every social issue has its Moms. Suppose, furthermore, that Mom becomes convinced that anarchy, while it may adequately address every other problem, cannot address the problem of drunk driving nearly as well as some governmental system. As a result, Mom finds herself unable to accept anarchism.

The story of Mom is meant to illustrate a psychological point: we have emotional attachments to specific social issues which are often disproportionate to the objective significance of those issues, and these emotional attachments can bias our evaluation of social theories. Thinkers must become aware of this problem to guard against it. We must recognize the possibility that the system that is overall best for society may not solve all problems and may not solve the problems to which we feel the greatest psychological commitment.

8.1.3 Varieties of government and anarchy

Suppose that an anarchist, seeking to show the superiority of anarchy over government, takes the Soviet Union under communism as an illustration of the nature of government. Under Soviet communism, tens of millions of innocent people were killed by the government - some for disagreeing with the government, most simply for belonging to the wrong social class.[2] The rest were forced to live for decades under oppression and poverty. Anarchy will be better than that.

Few observers would have difficulty spotting the fallacy in this reasoning: those who defend government do not typically defend any and all forms of government. They need not defend, for example, communist dictatorship; they need defend only some feasible form of government. Thus, in comparing government with anarchy, we should examine the best feasible form of government. It does not matter if we also have available some terrible form of government that we would never want to adopt (unless, having decided to have a government, we lack control of what form of government we get). Hereafter, I shall assume without argument that the best form of government is representative democracy.

The same point applies to the other term of the comparison: anarchist theorists need not defend any and all nongovernmental conditions; they need defend only some feasible nongovernmental social structure. It does not matter if, in addition, there are terrible forms of anarchy that we would never want (unless, again, we lack control of which form of anarchy we get). Anarchist thinkers differ over what the best nongovernmental system would be, particularly whether it would have a socialistic or a capitalistic economy.[3] I shall not enter that debate here. I shall simply assume capitalism as the better alternative. This is not because socialist forms of anarchism are not worth considering but simply because the comparison of two social systems, representative democracy and capitalistic anarchy, will prove complex enough to occupy us for the remainder of this book without the addition of further alternatives.

8.1.4 Against status quo bias

Most human beings evince a marked tendency to view the present arrangements in their own society, whatever they may be, as right and good (see Section 6.4). This bias explains how people from extremely different cultures can each regard their own cultures' practices as the best.

One potential form of status quo bias is the practice of assigning a heavy burden of proof to anyone who proposes a new social system very different from the present system. We might hold that the reformer must prove the superiority of a new system and that any doubt about the relative merits of the new system and the old system redounds to the benefit of the old. Such a burden of proof may prove crushing for two reasons. First, the complexity and unpredictability of human societies renders it difficult or impossible to prove nearly anything of interest in social theory.[4] Intractable differences of opinion about the social consequences of policies, institutions, and social events are commonplace. Therefore, adopting a presumption in favor of the status quo might easily prove the decisive move blocking the acceptance of nearly any social change.

Second, the comprehensive comparison of broad social systems is extremely complex. A great many social issues and problems are affected by the overall structure of society - war, poverty, inflation, drunk driving, pollution, racism, drug abuse, school shootings, fossil fuel dependence, health care, abortion, animal rights, capital punishment, human cloning, education, euthanasia, teen pregnancy, gang violence, and so on. No single work could address every important social issue - and if one did, few would have the patience to read it through. Suppose, then, that we adopt the habit of assuming, for each social issue, that the status quo represents the best way of dealing with that issue until proven otherwise. This would, again, give the status quo an essentially insurmountable lead over any radical alternative. Even if a reformer succeeds in arguing painstakingly that the status quo fails on several important issues, there will always remain many more issues on which the status quo triumphs by default in virtue of our not having had time to thoroughly examine them.

What would be wrong with granting the status quo an enormous dialectical advantage over radical alternatives? The problem is not merely one of dialectical 'unfairness'. The problem is that such a methodology is very likely to lock us into an inferior social system. There is no reason to assume that one's own society happens to be arranged in the best way. Therefore, it is desirable to find a methodology in social theory that gives alternative social structures a reasonable chance of being chosen.

In the following chapters, I address what seem to me the most obvious and serious concerns about anarchism. I cannot address how every important social issue would be dealt with in an anarchist society. If, however, I can show that the problems that initially appear most likely to occasion decisive refutations of anarchism can in fact be well handled by an anarchist society, then the burden of proof will be shifted to the critics of anarchism.

8.2 A Simplified conception of human nature

Any attempt to work out the consequences of a broad social system must rely upon some beliefs about human nature. Here, I describe my most important general assumptions about human nature.

8.2.1 Humans are approximately rational

Human beings typically choose actions on the basis of their beliefs and their goals. Aperson is 'instrumentally rational' when he chooses the action that, based on his present beliefs, would appear to do the best job of realizing his goals, whatever those goals may be. In ordinary life, we rightly take it for granted that human beings are approximately, though not perfectly, instrumentally rational.

Consider a simple illustration. You see a small girl climbing a tree while calling out 'Snowball'. You notice that she is moving towards a white cat perched in the tree. How would you interpret the girl's behavior? Barring unusual conditions, the most natural interpretation would be that the girl is attempting to retrieve the cat from the tree. This interpretation takes (approximate) instrumental rationality for granted: it ascribes to the girl a goal, getting the cat out of the tree, such that her behavior makes sense as a way of pursuing that goal. Without an assumption of basic instrumental rationality, there would be no limit to the possible interpretations. If she is irrational, the girl might just as well be climbing the tree to get away from the cat. Or to slake her thirst. Or to depose the president of Egypt. In normal contexts, we would say that such 'explanations' make no sense - and the explanations don't make sense because they fail to make sense of the agent's behavior.

Nearly any ordinary intentional action could just as easily be used to illustrate the idea. Driving to the store, putting gas in your car, buying a jar of applesauce, asking for a raise, lying down in bed, going to the dentist, cooking a meal, picking up a telephone - any of these actions can be explained by reference to (i) a goal that can be plausibly ascribed to the agent, (ii) a host of accurate and reasonable beliefs on the agent's part about the world and about the consequences of one's actions, and (iii) the assumption of instrumental rationality on the part of the agent.

There are exceptions to the general rule of instrumental rationality. Psychologists have documented many specific failures of rationality, too numerous to list here.[5] In addition, there are some general conditions that render various failures of rationality more likely: human beings are most likely to make mistakes when facing unfamiliar, complex situations or situations in which abstract reasoning is required to work out the right choice. They are also likely to make errors in decisions that are unimportant to them, where they do not care to devote sufficient thought to identify the best option.

The assumption of instrumental rationality is most likely to hold good when people are facing simple, familiar situations in which the best option is easy to work out. It is also very likely to hold for managers of businesses in competitive industries. Businesses that are managed irrationally will tend to underperform rationally managed businesses. The latter will thus expand while the former shrink, until the market is dominated by relatively rational businesses.

8.2.2 Humans are aware of their environment

Human beings tend to possess a great deal of accurate and practically relevant information about their environments and about the potential consequences of their actions. They usually do not ignore useful information or adopt wildly inaccurate beliefs about matters relevant to decision making. This is a variation of what philosophers call 'the principle of charity'.[6] The same sort of examples that illustrate basic instrumental rationality can also be used to illustrate this principle of charity.

Consider again the girl climbing the tree. To understand her behavior, we must assume that the agent possesses a host of sensible beliefs about herself and her environment - that the cat is in the tree, that the tree is a solid object, that physical objects tend to fall downward when unsupported, that climbing the tree will bring her close to the cat, that her hand will not pass through the cat, and so on. In ordinary life, we would not pause over all these beliefs, but all of them are implicated in the girl's apparent plan to retrieve the cat by climbing the tree - if she were ignorant of any of these facts, she could not make that plan. Again, this is not an unusual example. Any of a great variety of ordinary actions could as easily be used to illustrate the point.

There are a number of exceptions. People tend to be ignorant of complicated, abstract truths about unfamiliar subjects. They tend to ignore information that they find uninteresting and that does not help them to achieve their goals. And if some information is costly to gather, either in monetary terms or in terms of time and effort, then relatively few people will possess that information.

On the other hand, when information is easily and cheaply acquired, easy to understand, and relevant to achieving our goals, then we will generally have that information. It is particularly likely that managers of businesses in competitive industries will possess accurate information relevant to managing their businesses, since the businesses whose managers remain ignorant of such information will tend to underperform the businesses whose managers are well informed, allowing the latter businesses over time to dominate the industry.

8.2.3 Humans are selfish but not sociopathic

I am extremely selfish. I recently bought a winter jacket for myself at a cost of around $200. I already had some sweaters, shirts, and other jackets; I simply wanted a better jacket than any I already had. So I satisfied a fairly trivial desire with this expenditure. Had I donated it to a charity group working to combat world poverty, this same money might have saved another person's life.[7]

This suggests that I value my own welfare perhaps thousands of times more highly than I value the welfare of strangers elsewhere in the world. Yet as disturbing as this may sound, it does not indicate some sort of sociopathic disorder on my part. It does not even indicate an abnormal level of selfishness. The average American gives even less to charity than I do, while a third of American households give nothing at all.[8] Most human beings, when given the chance to give money to others, will choose to keep the money for themselves, particularly if no one is watching.[9]

For another illustration, imagine that you were to learn that you are going to die tomorrow. Though it may be impossible to imagine accurately how you would feel, it is a safe guess that you would be quite upset. Now I will tell you something you probably did not know: based on recent worldwide mortality statistics, there are about 156,000 human beings who will in fact die tomorrow.[10] How do you feel now? You may find this information disturbing. But if you are like most people, you are far less upset at this news than you would be by the news that you yourself were about to die. This suggests, again, that your concern for yourself is perhaps thousands of times stronger than your concern for most other people.

Nevertheless, there are many exceptions to the rule of human selfishness. Many people voluntarily donate money to charity, albeit far less than they could afford.[11] Most are prepared to make great sacrifices for family, friends, lovers, or others to whom they are close. And a few people, such as Albert Schweitzer or Mother Theresa, virtually devote their lives to helping others.[12]

Perhaps the largest and most pervasive exception to the rule of selfishness concerns the perceived negative rights of others: an ordinary human being may or may not sacrifice $200 to save the life of a stranger in a foreign land, but nearly everyone would be appalled at the thought of killing a stranger for $200. Admittedly, a small number would be happy to kill another person for $200 - sociopaths lack the respect for social norms and the capacities for empathy, guilt, fear, and horror that prevent most human beings from injuring each other under normal circumstances. Fortunately, however, sociopaths comprise only about 2 percent of the population.[13] The remaining 98 percent of human beings are concerned with social norms, are capable of empathy, and possess a rich array of emotions.

Thus, while ordinary human beings are willing to make very little positive effort to aid others with whom they lack close relationships, they tend also to be reluctant to directly attack other human beings or to positively violate generally accepted social norms in other ways, even when they stand to gain by doing so.

8.2.4 On behalf of simplification

The foregoing is a very simplified account of human nature. There are myriad variations among individuals and innumerable human motivations that I have not touched upon. Why is it useful to consider such a simplified picture of humanity?

First, note that the account identifies causal factors in human behavior that are both real and large. This is one requirement for a useful idealization.[14] We human beings really are moved by self-interest, and we really do tend to take actions likely to get us what we want. I have not mentioned every exception to these rules, but I have tried to mention the most important and well-established exceptions.

Second, the principles about human nature listed in the previous subsections are banal and driven by commonsense observations rather than any partisan ideology. In this way, they are unlike the Marxists' contention, for example, that human beings are strongly motivated by class interests.[15] The latter claim is highly controversial and accepted only by those with a particular ideology. This point is very important, because clashing ideologies are the greatest obstacle to progress in social theory.

Third, my simplified account of human nature enables us to derive many qualitative predictions about the effects of social systems relatively straightforwardly, with little need for judgment calls. There is a well-developed and compelling body of theory, comprising economics and game theory, that works out the consequences of the assumption of rational egoism. Those familiar with the theory can generally agree upon a large range of predictions, regardless of their initial moral, religious, or political leanings.[16] This is important, again, for making progress in social theory. The simpler a theory's assumptions are, and the more straightforward its predictions, the less room there is for human bias and irrationality to manipulate the theory. Due to the enormous role that bias plays in political discourse,[17] this is among the most important virtues that a theory in this area can have.

8.2.5 A historical application

The simplified account of human nature I have sketched makes useful predictions about certain social systems. Take the case of a social theory proposing that all citizens should work for the benefit of society, while receiving equal pay. A simple theoretical prediction is that, in such a system, productivity will decline. Individuals have a high degree of control over their own productivity, and greater productivity usually demands greater effort. Since most people are rationally selfish, they will not exert great effort to be productive unless they expect to receive personal benefits from doing so. So if all are paid equally, and if there are no other rewards or punishments attached to quality and quantity of work, then people will not be very productive.

This prediction is in fact correct. The twentieth century's experiments with social systems in this vicinity are well-known, so I shall not dwell on them.[18] An interesting but little-known illustration is provided by America's first experiment with communism, which took place at Jamestown, the first permanent English settlement in America.[19] When the colony was established in 1607, its founding charter stipulated that each colonist would be entitled to an equal share of the colony's product, regardless of how much that individual personally produced. The result: the colonists did little work, and little food was produced. Of the 104 founding colonists, two-thirds died in the first year - partly due to unclean water but mostly due to starvation. More colonists arrived from England, so that in 1609 there were 500 colonists. Of those, only 60 survived the winter of 1609-10.[20] In 1611, England sent a new governor, Sir Thomas Dale, who found the skeletal colonists bowling in the streets instead of working. Their main source of food was wild plants and animals, which they gathered secretly at night so as to evade the obligation to share with their neighbors. Dale later converted the colony to a system based on private property, granting every colonist a three-acre plot to tend for his own individual benefit. The result was a dramatic increase in production. According to Captain John Smith's contemporaneous history,

When our people were fed out of the common store and labored jointly together, glad was he [who] could slip from his labor or slumber over his task, he care not how; nay, the most honest among them would hardly take so much true pains in a week as now for themselves they will do in a day ... so that we reaped not so much corn from the labors of thirty, as now three or four do provide for themselves.[21]

One lesson from this episode is that, simple as the account of human nature I have advanced is, it can yield very useful predictions. If the company that created the Jamestown charter had known a little economics, hundreds of lives might have been spared. Another lesson is that the impact of human selfishness depends greatly on the social system in which people are embedded: in one kind of system, selfishness may have disastrous consequences, while in another, it promotes prosperity.

8.3 Utopianism and realism

8.3.1 The principle of realism

When proposed political and social arrangements are criticized, it is usually for being either harmful or unjust. But sometimes a vision of society is both perfectly just and highly beneficial, and yet we reject it on the grounds that the idea is too unrealistic, or 'utopian'.[22]

Consider, for example, a position that we might call 'utopian socialism'. The utopian socialist holds that everyone should be paid equally, regardless of productivity. In the last section, we saw that arrangements of this kind can lead to underproduction, putting things mildly. The utopian socialist addresses this problem by simply proposing that we all henceforth agree to behave selflessly. If we all voluntarily work our hardest for the good of society, then there will be no problem. Granted, this has not happened in the past; nevertheless, says the utopian socialist, it is what ought to happen.

Intuitively, this idea is defective as a political and social philosophy, regardless of whether the society it depicts is just or desirable, because the idea is too unrealistic. It seems to call for an alteration in a robust aspect of human nature, without proposing a plausible mechanism for bringing this about.

Many believe that anarchism, too, falls to the charge of excessive utopianism. These critics accuse anarchists of relying upon unrealistically optimistic assumptions about human nature.[23] How can we avoid this objection?

8.3.2 Prescription for a realistic anarchism

To avoid excessive utopianism, we should respect the following constraints:

1. We may not assume unrealistic levels of altruistic motivation. In examining the advantages and disadvantages of the system, we must assume that the anarchist society will be populated by people with relatively normal levels of selfishness, and we must accept the consequences of that selfishness within the particular social structure.
2. We may not assume perfect rationality or knowledge. Our defense of anarchism must be consistent with the fact that people periodically make mistakes.
3. We may not assume psychological uniformity. We must recognize the fact that human beings have a variety of motivations and character traits; for example, that some individuals are unusually aggressive or reckless.
4. We may not simply assume persistence of the system over time. Rather, we must be able to argue that, once adopted, the system would be able to resist forces that might be thought to undermine it.
5. We may not assume simultaneous, worldwide adoption of the system, since there is no plausible way in which that could come about. We must imagine that (perhaps as a transitional stage) some limited region or group becomes anarchist in a world otherwise dominated by states.

On the other hand, it is worth noting two things that do not render a social theory overly utopian:

1. If a theory 'cannot' be implemented simply in the sense that people cannot be convinced to implement it, this does not render the theory too utopian.

   As an analogy, suppose it is proposed to me that I should donate $200 to charity this month. I reply: 'No, that is not realistic, because I refuse to do it.' Intuitively, I have not articulated a valid objection to my donating $200, however intransigent my refusal may be. Similarly, the refusal of most members of society to take anarchism seriously, let alone to attempt to implement it, does not create a valid objection to anarchism. Of course, if most people have some reason for rejecting the theory, that reason might be cited as an objection to the theory. Anarchism is to be evaluated by supposing, perhaps improbably, that the system is adopted and considering from there whether a desirable and just state of affairs would result.

2. To be sufficiently realistic, a model for society need not be feasible or desirable under all social conditions. It need only be argued that there are some conditions, likely to be realized now or in the future, under which the model would succeed.

In subsequent chapters, then, I shall attempt to characterize a kind of anarchist society that would be stable and livable under some realistic conditions, despite the existence of criminals, dissidents, ordinary human selfishness, and foreign governments.

8.3.3 Against utopian statism

It is widely recognized that anarchists face a significant challenge of avoiding utopianism. It is widely recognized, as well, that some nonanarchist theories, such as certain forms of socialism, face charges of utopianism. What is less well recognized is that even very conventional, moderate political theories can be utopian.

For example, the theory of liberal democracy might be too utopian. It would not be excessively utopian simply to advocate that we maintain current institutions, functioning exactly as they now do. But few thinkers adopt this position. Most believe that some degree of reform is called for - for example, that the democratic process should be less influenced by special- interest groups. The more one knows about how government actually functions, the less likely it is that one can sincerely assert that it operates as it should. This opens up the possibility that the changes we would like to see in democratic states are overly utopian. Advocates of liberal democracy face the same strictures against utopianism as advocates of more radical positions, such as anarchism or socialism.

It may seem strange to suggest that a mainstream, liberal democratic position might turn out to be too utopian, while some radical anarchist alternative might be sufficiently realistic. But the distinction between utopianism and realism is not a matter either of how far a proposal is from the status quo or of how far it is from the mainstream of political thought. The distinction between utopianism and realism chiefly concerns, roughly, whether a political or social idea requires violations of human nature. A mainstream political view might turn out to require such violations, while some radical alternative does not. It is perfectly possible for a small change to be unfeasible, while some much larger change is feasible.

One common form of utopianism consists of confusing the way individuals and organizations are 'supposed to' behave with the way they will behave. When social systems are evaluated, it does not matter how a system is supposed to work; what matters is how it can be expected to work under realistic assumptions about human nature. For example, we can say that the function of government is to protect the rights of its citizens, but nothing follows from this about what government will actually do. In the absence of an effective mechanism for inducing government agents to efficiently dispatch their stated functions, we cannot assume that citizens will receive appropriate protection. The point here is not that government will not protect us; the point is that whether government is an effective mechanism for protecting individual rights, promoting social welfare, or promoting any other aim needs to be established by argument and evidence, not taken for granted simply because of the stated purpose of government.

A related form of utopianism consists of suspending general assumptions about human nature when considering agents of the state. Defenders of government are often keen to point out the harms that might result from the widespread greed and selfishness of mankind in the absence of a government able to restrain our worst excesses. Yet they seldom pause to consider what might result from the very same greed and selfishness in the presence of government, on the assumption that government agents are equally prone to those very failings. It is not that statists have some account of why government employees are more virtuous than average people. Nor do they have some plan for making that be the case. Rather, it seems simply to have never occurred to most statists to apply realistic assumptions about human nature to the government itself. The state is treated as if it stood above the empirical human world, transcending not only the moral constraints but also the psychological forces that apply to individual human beings.

Any social system, whether it be anarchist or statist, must be judged by how it would perform when inhabited by real people, such as we find in the actual world. It does not matter if anarchy would work well in a world in which all individuals were selfless and always respected each other's rights. By the same token, it does not matter if government would work well in a world in which all government agents were selfless and always respected the rights of individuals.

---

Previous Chapter       Next Chapter

The Problem of Political Authority

An Examination of the Right to Coerce and the Duty to Obey

Michael Huemer

Index

Previous
Chapter

Next
Chapter

---

Hogeye's Anarchism Page

Library of Liberty