It is very nice to see so many nice friends here, many of whom go back over the thirty years I have devoted to thinking about this topic and this book. If this crowd were not so friendly, I would actually take the manuscript out and try to speed-read it to you in the thirty-five or forty minutes that are available to me. But somehow I thought it better to take an alternative course of action, which is to jump from peak to peak, running a serious risk, in an exposition of this sort, of falling into some rather deep valleys, which I will of necessity leave unexplored this evening.

Now, as I was trying to position this talk, I said to myself, "I am now at AEI, and so what I have to do is to assume an air of reasoned moderation out of respect for the strong intellectual corporate culture that has dominated this place." And so what I am going to do is to describe how this book, in fact, makes me Aristotle's main man of the twenty-first century. Moderation is, after all, a virtue to which I have come only relatively late in life.

Substance

My first "moderate" development concerns the substantive positions that I am trying to avoid. Next I shall speak of the methodological traps that lurk along the way. Of course, the appropriate strategy, with respect to both topics is to show that there is somebody to your left and somebody to your right so that you display the virtues of Aristotle's golden mean.

One of the substantive positions that I seek to avoid is one for which I have a great deal of affection: it is the stripped-down version of the libertarian theory of the minimum state, in which the essential role of government is to use its monopoly of force to prevent harms to ordinary individuals, where the relevant harms to others are described in the narrow language of force and fraud. There is no doubt that government must occupy this critical role. It is far less clear that it should be limited to it in light of the many common activities that sensible governments routinely undertake. We must therefore take steps to bolster the constrained vision. Most obviously, it is generally accepted that we have some system of taxation. It is also clear that the government has some role with respect to supplying social infrastructure. Neither of those two functions fit very easily within the narrow libertarian framework.

So the question now is how do we make government a little bit bigger than ardent small-state libertarians would like it to be? It is painful to me to recognize that we must take these steps, in part because I fear that small concessions will lead us to go a great deal further. One of the tendencies on the other side, which I am equally anxious to avoid, says, "Since you cannot give yourself a narrow and limited definition of what government does, essentially what you do, in constitutional and political terms, is recognize that any state can do, more or less, anything that it pleases, so long as it gives at least a passing nod to advancing the common welfare of the population at large. Any intervention, regardless of its content, is going to be sufficient so long as it is passed with benevolent motivation."

This theme, of course, travels in many versions. At one time, the ideal socialist program was to figure out how to attain ideal distribution and ideal production just by running decisions over price and quantity of various goods through some well-informed and well-intentioned central planning agency. I am happy to report that one Friedrich Hayek turned to that issue a little bit before I did, and was able to explain in fairly powerful form why centralized planning does not succeed in its own terms; namely, that that system is not able to produce the level of goods and services to which it aspires. Mr. Khrushchev did not bury us in the 1950s. Rather the Soviet Union disintegrated under its own weight by the late 1980s and the early 1990s.

It turns out, however, advocates for larger government are a very hearty breed who seek to adopt more modest positions in the ashes of their former failures. The fallback position that has the greatest appeal in the United States is, of course, the comprehensive welfare state, which works with a greater success than socialism, precisely because its ambitions are more modest. But that system insists that, in dealing with general matters of economic regulation, no powerful side constraint of constitutional law or legal principle prevents the legislature from doing exactly what it wants and precisely as it pleases, to borrow, and garble, a Rex Harrison line from My Fair Lady.

Now, in responding to these twin perils, I am a moderate after all, because I think that one can escape the dangers of being the hard core libertarian without falling into the trap of thinking comprehensive government power is the order of the day. So the subtitle of this book, which I added when I realized that no one would figure out what Skepticism and Freedom meant, was "A New Defense of Classical Liberalism" so as to highlight the view that the list of legitimate government functions remains relatively limited, such that many tasks fall outside the sway of proper government action.

If one wants a thumbnail sketch without figuring out its philosophical derivation, I would include the functions appropriate to the libertarian state. I would add infrastructure maintenance as a classical public good, and the taxation to support this and its other functions. I would also include, although on delicate tiptoe, the regulation and prohibition of various kinds of monopoly behavior, while recognizing the real dangers that can arise when an administrative state is turned loose on that rather slippery subject. So that is the broad substantive outlook.

Methodology

Now in addressing the second question about methodology and legal rules, I will sound more like a lawyer and less like a political scientist. Basically, we can identify three ways to think about the law. Two give rise to my deep suspicions. Keeping to my Aristotelian theme, I think that the third is much more moderate and sensible than either of the extremes.

When you are young in this world, you believe that the class of deductive truths about social matters is larger than it turns out to be. The great attraction of libertarian thought lay in its deductive power. The hope was that you could axiomatize the system and sort of render social problems amenable to a set of principles that yielded necessary or deductive truths. That vision certainly fired my early academic life.

In fact, the legal literature contains many claims about the way in which social institutions are organized. These routinely appealed for inspiration to either the rules of logic or to some set of universal physical principles, such as Newtonian gravitational rules. In fact, you could go back to read Locke and Blackstone, only to discover that the use of the word "law" is somehow or other deliberately equivocated so as to cover natural laws on the one hand, that is, the laws of motion, and natural laws of social behavior on the other.

It seems to me that this identification constitutes a very serious mistake. If you look around various societies, you see enough variety in the content of legal rules that all of them cannot be deductively required. In addition, certain levels of evolution take place with respect to legal doctrines over time, so one would be deeply suspicious of how that could happen if immutability were the touchstone of the law. Today it also turns out that whatever consensus we might have had about proper legal principles has been frequently shattered. Appeals to self-evident truths routinely fall on deaf ears by people who may hold to a different set of self-evident truths. So the first danger lies in seeking cheap victories by deduction, introspection, intuition and self-evidence. Normative legal arguments cannot be won today by fiat.

Now it is also possible to lurch toward the other extreme, which is equally deadly in its own way. This second pole is more characteristic of the twentieth century perhaps than the eighteenth century, but it has spurred modern efforts to undermine classical liberalism. Now the exhilarating thought is that, for one reason or another, virtually every proposition that someone may assert about the nature of a proper legal order is necessarily infected with some fatal contradiction. One source I reread in preparing to write Skepticism and Freedom is what I believe to be the best study of Critical Legal Studies, Mark Kelman's A Guide to Critical Legal Studies written in 1987. What is striking about the book is the number of occasions on which he identifies some fatal contradiction to capitalism. No logician invokes the idea of contradiction more freely than Kelman does.

Oddly enough, an insistence that the classical legal order (which is chugging on quite well) is doomed by some fundamental contradiction supplies us with the mirror image of the natural law types. CLS assumes that whenever you articulate various kinds of propositions, you can show not their necessity, but rather their impossibility. It seems to me that this position is wrong for exactly the same reason that the older natural lawyers were wrong. If there are no cheap victories for the defenders of classical liberalism, so too there are no cheap victories for their skeptical welfare-state opponents either.

Skepticism and Human Nature: Of Scarcity and Self-Interest

So then, what on earth do we do to seek out a sensible middle ground? It is here where one starts to ponder about skepticism and its relationship to both freedom and human nature. Essentially, as I have gotten older and maybe a little bit wiser--which is why that thirty years really start to matter--I have discovered, to my infinite regret, that most of the serious debates over the basic principles of any political order have an irreducible empirical content. The analyst must figure out how to obtain a general understanding of human nature that allows him to do two things simultaneously: first, to articulate and defend a set of general principles of social organization; and, two, be sure that this set of principles is sufficiently respectful of the place of individual choice that it does not, to the last jot and tittle, determine what everybody ought to have, when they ought to have it, and how they ought to have it. That is, the analyst has to acquire enough knowledge to allow him to establish a framework, but cautious enough to be sure that the framework does not dictate how all people think and act.

There are, in effect, two forms of skepticism. One form of skepticism says that we know nothing about legal rules, and therefore cannot articulate sound principles of governance. That is a dangerous form of skepticism, which must be contrasted with the sensible form of skepticism that says we know enough about the basic patterns of human behavior to be confident, on average, what set of rules is going to do better by us as a people than some alternatives that might be advanced.

To put the point in some fancy, modern philosophical jargon, my intellectual conversion has had two parts over the past thirty years. The first has been to recognize a slight expansion in the scope of the state over my former libertarian self. The second is a rather radical and, to some people, regrettable--I am looking at Roger Pilon--conversion of my own intellectual beliefs. My classical liberalism derives from a moral consequentialist, rather than a deontological, orientation.

What drives me to this outlook? It is essentially the hope that I can lay firmer foundations for the strong intuitions, not of the modern socialist, but of the natural lawyer. My hope is that one can defend many of the substantive positions that they defended without having to appeal to their claims of immutable self-evidence. That requires an analysis of the real and material effects on human beings that different types of legal rules have on human societies.

This ostensibly novel approach makes me into an ancient in another way--not in terms of my own age, but in terms of how I think about legal problems. One point that has impressed me about all of the great classical writers on government and individual liberty is that, almost without exception, they do not start with the question of political organization or the question of individual rights. Rather, their deeper instincts lead them to start with questions of human psychology: how is it that we think the basic tendencies of human behavior will work themselves out over time, and what we can do in order to understand these tendencies so as to curb their dangerous excesses while promoting their more creative impulses. So it is that the nature of society depends, to some extent, on this somewhat over general and grandiloquent question: What do we know about the nature of man?

When this program was first introduced, hard knowledge about human behavior was hard to come by. Take a quick peek at Aristotle's de Animalia, and it is clear that at one time we did not know all that much about the nature of man. But today that project is on firmer foundations. My approach in Skepticism and Freedom is to figure out how to translate some of the basic insights of evolutionary psychology (or social biology or general evolutionary theory) in order to see whether they give us enough information to structure the inquiry into the basis of various social institutions. This approach means that we have to start with some stylized facts to determine the main features of our social institutions.

What are the key features of this system? It is best to start with three. The first has to do with individual self-interest; the second has to do with scarcity; and the third, which is a refinement on the first, explains why individual self-interest is, in fact, not quite as individual as is sometimes supposed, but has a rather more complicated, cooperative quality that is captured under the term "inclusive fitness." So the biological material that I learned in the 1970s and 1980s comes back to first haunt and then help my legal work at the beginning of the twenty-first century.

Now why does one want to start with scarcity? Well, if there is one truth which we can truly regard as self-evident, this one is going to be it. If we look around the world, we observe that there is a finite space, finite amount of land, and finite amount of water. It also turns out that our level of desire for these resources always exceeds the resources available to satisfy them, no matter what the state of humankind.

Even before you get to the question of how legal rules intervene in the operation of a political system, you have to ask what kind of personality types can survive in an unstructured environment where, in fact, scarcity operates as the dominant constraint. The answer is there is going to be one kind of characteristic, which is going to dominate over all others, and that turns out to be self-interest, subject to the qualification of family and inclusive fitness that I mentioned. So let us just kind of look at this expanded version of self-interest.

Let us just assume, in a thought experiment, that we have two kinds of folks, some of whom are benevolent and indifferent to their own welfare, relative to the sentiments of their fellow man, and others who turn out to be egotistical. If you run the calculations, what is going to happen when one guy is quite happy with a fifty-fifty split, while the other guy seeks seventy-five-twenty-five in his favor? There is only one direction in which the overall balance of power can move. Over time, the self-interested individuals will manage to co-opt more and more of the resource base, which would otherwise have been made available to others. Selection pressures, therefore, will exhibit an enormous bias in favor of the selfish relative to those who have a disinterested altruism. The repeated application of a single mechanism helps shape so much of the behavior of all living beings, humans included.

The question then arises, what is the break on individual self-interest? The answer turns out to be genetic inheritance. Basically, if any organism is going to succeed in this process, it has to be in it for the long run. If you simply want to gorge yourself on sweets and fancies, you (or those like you) can die out in a single generation, at which point people who are altruistic toward their own children will fare far better than those who are selfish in their dealings with them. So the dominant feature of human behavior is captured by inclusive fitness, which takes into account not only ourselves, but those people who are related to us genetically, discounted by the degree of that relationship.

Just this one simple modification of the basic principle of human behavior vastly complicates the strategies that individuals, as social actors, take toward each other. It is no longer plausible to take a naive Hobbesian view that individual self-interest translates into the importance of individual autonomy. It also means that many of the classical critics of laissez faire have a point which should be, in fact, incorporated into our basic understandings, which is the idea that it is dangerous to proceed as if the world were composed of atomistic individuals who care nothing about their connections with other human beings. That sort of stripped-down model of individual egoism is, in fact, not an outgrowth of our biological theory, but stands in direct contradiction to it. The level of cooperation necessarily required for any person to implement his or her genetic program, therefore, requires a complex set of social behaviors. Once you understand this impulse, it helps explain how to escape the Hobbesian dilemma without getting yourself into the terrible pitfalls of the absolute state. So this one assumption takes on a lot of heft.

From Human Nature to Human Laws

The next question, which follows from the first two, asks how do the resource constraint and the self-interest constraint work in tandem in setting up rules of property to deal with resources in the external world? Here, the basic biological proposition that I want to rely on is a very simple one. It has to do with the fact that variations in all traits appear in every population and why these variations are necessary for the survival of the species as a whole. If every individual were exactly like every other individual, we would lose two enormous benefits: One is it would no longer be possible to have cooperation between individuals with slightly different skills because, in fact, the gains that you would get from trade would be reduced substantially by the identity of the individuals in question. Two is that your vulnerability to attacks by external perils, for example, disease, would be markedly increased. Once some parasite has created Dutch elm disease, it wipes out the entire species because the lack of variation reduces the level of potential resistance.

But variation over what? It is here I think that we must be aware of some of the dangers of moral relativism. One of the words that we use in everyday language to describe the products that we can purchase in a marketplace is "goods." The term "goods" is not just simply fortuitously related to the idea of individual good--quite the opposite. The reason we tend to call these things "goods" is that we know, on average, that they are going to improve the fitness of those individuals who consume them, unlike "waste" and "bads," which are going to have exactly the opposite effect.

The reason we know of these distributions on both the positive and negative domains stems from our knowledge of the caloric minimum necessary to support life. This daily necessity creates a set of imperatives that it is easy to overlook in times of plenty, but which all of us share at the most basic level. So these basic needs explain why people need territory, which then becomes property; why people need bodily integrity, which then becomes autonomy; why people must protect their children, which becomes family; and why people need cooperation to expand their possibilities, which then becomes contract.

The question then is how do we integrate our knowledge of these impulses into some general and objective theory. What purpose are we trying to achieve when we impose laws? Simply to talk about this project as though the sole goal is individual liberty misses the very important question of whether actions by one individual will have systematic negative consequences on other individuals. And if those negative consequences do occur, what, if anything, should be done in order to counteract them?

We can now start to put the pieces together into a larger whole. The autonomy principle will make a strong debut in any legal system because the moment the law allows somebody to break the bone or to penetrate the skin of another person, then death may ensue by infection or exhaustion or blood-letting or whatever. Yet the gains to the victor from this incursion are (as an empirical matter to be sure) relatively small. The judgment therefore of all social contract theorists is to require the mutual renunciation of force. We are confident, given the commonality of our human origins, that all of us will gain more from the imposition of this prohibition than they will suffer from being bound by it.

This conclusion does not state a necessary truth. It is a proposition which you could deny without self-contradiction. But it is not something, I dare say, that you could deny with a straight face. The great reason why it is that social contract theory has always had unacknowledged links with utilitarian theory is their convergence on a common object. The first of these praises the virtues of hypothetical exchange while the second starts talking about the utility gains that real exchanges and, by definition, hypothetical exchanges ought to achieve. They are somewhat different manifestations of the same basic impulse. Both the utilitarian tradition and the social contract tradition at their best find rather different ways to express this same fundamental truth.

We can now take the analysis of these duties of nonintervention a second step. The only way, it turns out, that they will be effective is to make them universal. It does me no good to say that old Smith cannot knife me because he disagrees with my political theory. Even if he toes the line, I can be killed only one time, and thus face the fear that any one person out of the mass of humanity may end my life even if all others wish for it to continue. What is essentially critical on the issue of self-preservation is absolute unanimity in adherence to the principle. So long as your one worst enemy can do you in, the gains that you get from interacting with everybody else will go for naught.

We also know, however, that actual contracting across a huge population is not possible. The individuals who are alive today may die tomorrow and others will take their place. Voluntary agreements could never keep up with these shifts in population. Accordingly, this absolute universal social imperative, nonintervention, must be enforced by some method. Whatever the variations in temperament and taste amongst ourselves as individuals, if we do not follow this particular kind of rule, we are not going to succeed with any lesser venture.

Now the next question is how do we take best into account the variations among human beings? Here we can take advantage of a powerful and instructive legal parallel. Generalized rules in rem, that is, rights that are good against the rest of the world, tend to be negative in form, universal in application, and easily complied with by all members of the community. They are rules, in effect, that work no matter how large or small the population is. So, whether you are talking about Rome of 1 million or the United States of 300 million, the same negative rules can apply perfectly well to these somewhat anonymous large person interactions.

But as I have also indicated, there is a parallel level of individual variation. How should we best exploit that? We have, basically, a one-word answer: exchange. The basic theory runs as follows: As a structural matter, in dealing voluntary exchanges, you can be confident that the Eleventh Commandment--the one they forgot to write on Sinai, but should have written there--will apply. It is one that has been beaten into my head for twenty-odd years, so I am going to say it again: Voluntary exchanges result in mutual gains for the parties. So the critical presumption is--and I use the word "presumption" here very advisedly--that the mutual gains from trade drive the institution of contract.

Now our confidence in this general proposition about the role of contract depends only on our general knowledge of the bare features of human behavior: self-interest (suitably modified) and individual variation. Our confidence does not depend upon our particular knowledge of whether you like carrots or whether I happen to be allergic to broccoli. What the analysis says is this: We know about these natural variations, and choose private ordering through voluntary contracts to harness the private knowledge that people have about their own preferences structure, because we believe that transacting this knowledge base is manifestly superior to our collective knowledge about preference structures writ large. The whole point in privatizing exchange is to exploit that localized private information to maximize the gains from trade.

The genius of the legal system is it does not have to know what the private information is in order to be able to validate this basic claim. All it has to know is that the preference maps of different individuals at different times are not identical to one another, so that gain from exchange can produce improvements in overall welfare. This regime has one other hidden virtue. One good idea applied a million times is one idea that is capable of generating a million gains.

Robert Nozick--our late friend, whose recent death we all deplore--emphasized just this point brilliantly in Anarchy, State, and Utopia when he talked about the recursive use of the basic rules of transfer. Whatever you receive in goods and services from one set of transfers you can use in your next bargain, so that the exchange principle allows you to develop incredibly complicated forms by the successive application of a single bright idea which, in turn, informs what becomes the major task in the law of contracts--how to reduce the transaction costs associated with these exchanges so as to maximize their velocity. In some cases (land sales), it may require formalities; in other cases (stock sales), it may require their elimination, but the basic technique is there, given the objective which comes out of what it is that we have. And we should thank my colleague Ronald Coase for placing this task of evaluating transaction costs at the center of our intellectual agenda.

Now the libertarian prohibition on force and fraud, of course, makes sense for exactly the same reason, going in reverse. Given our confidence that voluntary exchanges produce mutual gain, we start to accept profound restrictions against use of your force or aggression. What is a victory to you is a loss to somebody else. We know therefore that this transaction is not a Pareto improvement (in which somebody is better off and no one worse off). Most of the time we do not believe that the new state of affairs constitutes any kind of improvement at all. We tend to think that the losses are enormous and the gains are very small, which explains the familiar prohibition on the use of force and the use of deception. Deception, it should be clear, is a little bit more complicated than force. But the basic argument is that when people lie, they so misrepresent the relative value of two things in exchange that the victim of the fraud will turn out to be a net loser from the exchange. We therefore stop the fraud to be sure that exchange generates mutual gains.

The Role of Presumptions

But there are limitations. Recall that I described these basic rules as presumptions, and it is extremely important to explain why it is that the language of presumptions is so critical to the overall inquiry. At root, phrasing legal rules as presumptions represents another form of sensible skepticism that recognizes the imperfect general of our general knowledge. If you had perfect confidence in your general propositions, then you would not bother to have recourse to presumptions. You would tend to believe in absolutes. And whenever you believe in absolutes, you come up against cases that seem to fall within the principle but, by God, also seem to cry out for the opposite treatment from what the absolute rule directs.

One of the great weaknesses of standard libertarian theory is that it tends to push too hard by elevating presumptions into absolutes. It needs some coherent way of figuring out what counterexamples give rise to concern, and how these ought to be resolved. So what I am going to do is to just talk for a minute or two about how to use presumptions to enrich our set of social institutions without falling into an endless sea of ad hocery that comes when we despair of having any rules at all.

Let us first start with the presumption in favor of voluntary exchange. Is there any kind of circumstance in which a voluntary exchange ought to be distrusted? Well, in the usual case, the answer is no because most of the externalities to the transaction turn out to be positive rather than negative. If it is an ordinary contract for the sale of goods and both the buyer and seller benefit from it, each of them has a larger resource base which, in turn, gives greater opportunities to trade for third persons. The only externalities you have are positive. Even if those are not routinely or formally acknowledged in the legal system, their presence offers another additional and strong reason as to why private contracts should typically be enforced, not limited.

But suppose a particular contract is to kill the third person and now the externalities are negative. Here gains from trades between the parties spell greater peril to the third party. And so once you have the third party. In response the social response toward this subclass of contracts has to flip over. We now invoke a negative language. Contract becomes combination, or worse, conspiracy for aiding and abetting. So long as we are duty bound to prevent the underlying activity from taking place, we have to attack any contractual arrangement that seeks to advance that activity.

The same principle can be extended elsewhere. One obvious example is cases of bribery. Sometimes economists refer to bribes as the function of a price system, when what is desired is to increase the supply of goods by "bribing" people with higher offers. But that observation is only a misleading metaphor. In most cases we think of efforts by people to subvert public officials in the performance of their lawful duties, so once again the sign flips over such that activities that are normally legal become criminal offenses. In other cases the same logic applies with a bit more caution. I am thinking here of antitrust-type situations and monopolies, where the neoclassical theory is strong enough to suggest that the gains to the insiders who agree to raise prices or restrict output are smaller than the welfare loss inflicted upon third parties. So we do not call those ordinary contracts any more. Now we start to call them combinations, cartels and collusive arrangements because of our intuitive appreciation of what is going on. The great challenge to antitrust law, the great challenge in the law of regulated industries is to summon the institutional wherewithal to distinguish between those business arrangements that are positive-sum when the effect on consumers as well as producers are taken into account. If judges always got things wrong, then it might be better to stop the costly intervention in these matters, but the more judges know--and Judge Stephen Williams, one of our very finest, is in the audience--then the more plausible it is to undertake this ambitious form of regulation. Here too in principle the global analysis could reverse the result that is obtained solely by looking at the effects of the contract on the parties to it. The great antitrust revolution of the last forty years has been trying to get the sign right on those arrangements, whether horizontal or vertical, and I think, for the most part, we have done a tolerably good job.

What about the other situation that sets up the presumption against the use of force against other individuals? One problem for libertarian theories has been in finding the proper treatment for cases of private necessity. The classic illustration allows me to break into your cabin without your consent in order to prevent myself from being killed in a storm? Can I do so even when you try to resist my entry on the ground that it is your property, so that you have an "absolute" right to exclude?

It turns out, again, much to my happy amazement, that the common lawyers were more sophisticated on this question than many of the abstract theorists. Blackstone said it, Pufendorf said it, Grotius said it: they all recognized that necessity works a suspension of ordinary property rights. The cash value of that proposition is that the outsider becomes for the moment the insider and may treat the property as his. But once the danger had passed, the interloper had to pay compensation for the loss of use or for any permanent damages created. The necessity became the occasion for a taking with just compensation. The sentiment was so strong that the outsider could use force to overcome the landowner who sought to keep him off his property in times of imminent peril, but only in those times.

The role of forced exchanges is not confined to the odd cases of necessity. In practice, most of these disappear for most owners are only too anxious to help people in peril no matter what the law says. That has certainly been the tradition in maritime cases, where elaborate, reciprocal practices have been developed to improve the odds of successful rescue. The just compensation paradigm, however, comes really into its own in connection with the eminent domain power, which allows the state to take property for public use upon the payment of just compensation.

In this setting the state necessity no longer relates to immediate survival. Rather it involves the coordination of social resources in such a fashion that allows, for example, the construction of a highway when, without it, some single landowner might choose to block the construction of a road across his property. As the holdout power becomes too severe, the condemnation power (which has its own abuses) is used to overcome it. This, in fact, seems like a small emendation of the basic libertarian theory, but, in fact, it represents a major change in worldview, and it is one, I might add, that has an impeccable constitutional pedigree. Stating the problem as a choice between holdout questions and aggression shows that we cannot solve one problem unless we invite to some degree the other. The mere suggestion that just compensation has to be paid requires someone to determine when it is owed and what that level is. It raises question of whether we can identify occasions in which any exercise of the power to take and pay is inappropriate, so that the state cannot go forward at all. It also requires an examination of those cases in which regulation (under the police power) is allowed without any compensation at all.

To put the shift in orientation in slightly more technical terms, the standard libertarian acts as though what law should achieve is to maximize social welfare by minimizing the use of force and fraud. Once the takings issue is introduced, the social welfare function changes: Now the task is to minimize two sorts of institutional defects that cut in opposite directions. One of them is risk of coercion and aggression; the other is the risk of holdout. It turns out that the moment you try to curb the one, you necessarily increase the other. The upshot is that the maximization strategy is more difficult to identify and implement than would otherwise be the case. Figuring out the optimal system of taxation and the optimal systems of regulation requires government to solve that incredibly ticklish problem. Can it be done? I think the answer is that is can. I wrote Takings (1985) and Bargaining with the State (1993) in order to take one stab at the problem. I am not going to try to rehearse those arguments here. I am just trying to point out the long-term intellectual agenda.

Now the second problem facing any normative theory is brutally practical. Is it possible for people to create a government that can enforce the rights and duties that the theory endorses? The baseline for this inquiry is whether we can recognize a set of improvements, shared by all, relative to their position in the state of nature in which no rights (save the use of power) are observed. The rights of property, tort, and contract are fine places to start, but we must develop a comprehensive system of public institutions to finance the protection of the rights just created.

Here, the optimal theory of taxation does not point to a zero-taxation world, any more than it endorses a theory of rights that allows everyone to do exactly what he wants. The question here is not how taxes, writ large, preserve liberty, writ large. Of course taxation is necessary to protect liberty, but the key question is to find that optimal theory of taxation that does it with as little disruption to the basic fabric of social institutions as is possible in this rough and tumble world. Interestingly, on this particular question, all careful libertarian-type thinkers have come to the same conclusion, which is to strongly favor proportionate taxation. By the same token, they are deeply suspicious of head taxes, and they are almost as deeply suspicious of progressive taxes. I think the reason for this basic approach can be summarized in one or two sentences. Their elaboration will have to wait for another day.

The great advantage of proportionate taxation is that it places no dollar limitation on the amount of money that the state can raise to discharge its public functions. It is not as though in time of war some abstract political theory requires low taxation when the survival of a nation is at stake. But this theory does place strong limitations on the distribution of the tax burden, so as to be able to avoid or to limit the political rent-seeking that takes place when everybody tries to garner a large share of a public benefit while seeking to push off most of its cost onto somebody else.

In some circumstances, broad-based taxes are not appropriate because of the skewed nature of the benefits. One of the fine points that the nineteenth century lawyers understood but which the twentieth century lawyers have forgotten is that the close matching cost and benefits across persons lies at the heart of any system of special taxation, such as those for constructing local streets. Allowing a proliferation of special assessments without a matching principle invites major factional intrigue.

My favorite account of the problem comes from a report from now Judge Guido Calabresi, who described years ago a case that, if I recall correctly, he encountered while clerking for the Supreme Court. I have not checked the case out, but the pattern (whether true or false) illustrates the problem of skewed taxation. In a sense it is too good not to be true. The usual rule on special assessments allowed for local improvements to be taxed on a front footage basis. The more land you had adjacent to the road, the higher your taxes. In this case it turned out for a long stretch the proposed road was bordered on one side by a railroad in direct competition with the highway. Here return benefit from taxation is not well proxied by the front footage tax, but it appears that the United States Supreme Court did not think that the railroad had a case to be rid of this special tax so long as the legislature was content to impose it. If so, the case represents a complete breakdown in the way in which the burdens of our collective life should be allocated.

The Constitutional Payoff: Abrams v. Lochner

Now my last question for this discussion is how does this problem translate itself back into constitutional law; that is, what does it tell us about the proper principles that should guide the basic role and function of government. The concern here is with seeking to stake out the easy cases, not with deciding all of the hard ones.

Well, essentially, this tale becomes one of two sorts of approaches for judicial review of legislative action. One approach looks at constitutional guarantees, either under a strict scrutiny or an intermediate scrutiny standard, which are fairly close together, in many cases, in a practical sense. The second approach looks at things under a rational basis test, which allows just about any legislation to pass constitutional muster so long as someone benefits from its invocation.

How then should we forge the link between political theory and constitutional law? Once a court or an analyst starts down the first path and takes seriously the constitutional guarantees, in almost every case they will come up with either the right answer or at least a reasonably good approximation to it. That is why the law on the First Amendment or the dormant commerce clause is far better than the law of takings.

But on the other hand, if the court or analyst starts down that second road and assumes that rational basis is okay, it is tantamount to saying that if some witness or legislator is prepared to swear that two plus two equal five, then, in effect, everybody with a calculator is forced to accept that determination in assessing the legality of the statute in question. That is, to some extent, the position that Robert Bork has taken in his defense of legislative supremacy, which is why the American Enterprise Institute has to supply a broad umbrella to take into account the different brands of conservatism or libertarianism that lie underneath the tent.

I happen to be a consistent believer in the first view and deeply suspicious of the second. If you look at First Amendment law, I think the wonderful opinion which illustrates how to get the skepticism right is, in fact, our friend, Oliver Wendell Holmes writing in dissent in the Abrams case, when he starts to talk about it, "When men have realized that time has upset many fighting faiths, they may come to realize even more than they believe the very foundations of their own conduct that the ultimate good desired is better reached by free trade in ideas --that the best test of truth is the power of the thought to get itself accepted in the competition of the market."

What Holmes basically meant by this assertion is before we, as the state, start to coerce those individuals who we think will prove dangerous to us, we have to be damn sure that they pose an imminent peril to our own safety, by which he meant, in precise analogy to the necessity cases at common law, that there has to be the immediate threat for the use of force. Essentially, he developed, or at least retraced, in conjunction with the First Amendment law, the general judicial attitude on injunctions, which says that courts hold back against injunction unless there is an imminent peril of a serious danger. Once speech is stifled it cannot be re-created. But speech that does cause real damage remains subject to an after-the-fact remedy that should usually be adequate to the circumstances at hand.

Of course, Holmes struck the right mood because he did not over claim. He did not think that his approach had to be foolproof, which it is not, or achieve the status of a self-evident truth, which it cannot. Rather, he made a judgment that market competition was the best test of the truth of an idea. His implicit theory is that voluntary exchange in ideas--which, by the way, is a more complicated market than one in goods--is, in fact, the only ground on which we can organize our political institutions unless we are willing to precommit ourselves to the belief that some set of ideas are bad and some ideas are good. He just did not believe in that precommitment.

But what makes the Holmesian approach so persuasive is that we do have enough knowledge about the dynamics of political debate to see that freedom of speech does work on average to our social advantage, and that this regularity is powerful enough to justify the erection of the First Amendment guarantees at the outset of our republic. At this point, Holmes ties into the sensible form of skepticism I referred to above. Knowledge of political processes, like knowledge of voluntary markets, warns us of the abiding danger in the suppression of political dissent, precisely because we do not know enough to know which set of substantive ideas turn out to be right or wrong.

We should all draw real comfort from knowing that a review of the First Amendment tends by and large to follow a fairly close classical liberal handbook. It turns out, therefore, that the law allows for guarded state intervention to respond to threats of force and fraud. It allows (a bit too little) protection against defamation. On matters of infrastructure, it allows proportionate taxes of speech activities to support the limited functions of maintaining social order, but is deeply suspicious of progressive taxation against various speech activities. Likewise its constant emphasis is on nondiscriminatory regulation in order to create a check on political intrigue. Much of this enduring synthesis has been created by judges who have little knowledge of or sympathy for the classical liberal synthesis that I have defended here. But they manage to gravitate to the right solutions by case-by-case adjudication nonetheless. A system that may look forbidding in the abstract resonates well in connection with particular cases.

But this result does not hold across the board. Holmes was a very eloquent writer, who some years earlier penned Lochner v. New York (1905). In this context, his generalized skepticism led him to write a blank check to legislative power. If the state wants to pass a statute that limits certain kinds of bakers to working only ten hours a day, Holmes said it was just a matter of political will, one way or another. He says, and I quote, "I think that the word liberty in the Fourteenth Amendment is perverted when it is held to prevent the natural outcome of a dominant opinion, unless it can be said that a rational and fair man necessarily would admit that the statute proposed would infringe fundamental principles as they have been understood by the traditions of our people and our law." In his view the idea of liberty did not apply to the right to carry out a trade, which is how it was read in the majority opinion. Rather, that more capacious reading of the Constitution was regarded as "perverted."

It is useful to recall the conception that Holmes opposed: Thus Justice Rufus Peckham had written before Lochner that the "liberty mentioned in the [Constitution] means not only the right of the citizen to be free from mere physical restraint of his person, as by incarceration, but the term is deemed to embrace the right of the citizen to be free in the enjoyment of all his faculties; to be free to use them in all lawful ways; to live and work where he will; to earn his livelihood by any lawful calling; to pursue any livelihood or avocation, and for that purpose to enter into all contracts which may be proper, necessary and essential to his carrying out to a successful conclusion the purposes mentioned above." We should also add freedom of conscience and religion to the list, as later decisions did. If that approach counts as perversion, then we are all the richer for the decision of some judges on at least some occasions to protect such perverted activities

It is ironic that the judges who understood the indispensable role for a broad conception of liberty are the same judges who are widely reviled today because of their narrow understanding of the mainstreams of human conduct. I refer to Justice Rufus Peckham, who gave the broader definition against which Holmes was regaling in the Lochner case, and (oddly enough) Justice McReynolds, who extended the definition of liberty under the due process clause to protect parents who wanted to send their children to private religious schools, were not swept up by a government craze into public institutions. So let us just understand that classical liberalism suffers in the courts today because its foundations have been too frequently rejected.

Yet there are serious dangers with the Holmes position, which basically holds that any justification the state puts forward is going to suffice. So in the Lochner case, where the statute in question looked like a piece of class legislation, the very ground on which it was attacked, his attitude was, "I can't figure any of this stuff out. Since I don't have any general vision of the way in which people act in the legislature, I shall let it pass. If the dominant opinion wants it, then it's justified on the score of health. It is only the 'shibboleth' [his dismissive word] of laissez faire which stands in the way of political reform."

It is worthy of note that Holmes's then iconoclastic views did not attract the support of a single other justice of the United States Supreme Court. Holmes was, and should be, understood not as the arbiter of our law, but as the outlier from our constitutional tradition.

Why so? Because he said our Constitution "is not intended to embody a particular economic theory, whether of paternalism and the organic relation of the citizen to the State, or of laissez-faire." But this has it exactly backwards. Our Constitution could not have succeeded if it had been so indifferent to the choice of government theory. Indeed any reference to the "organic relation" of the individual to the state should give rise to real anxieties about the weakness of our constitutional safeguards against the risk of fascism. Contrast Holmes's resignation in Lochner with his more confident words in Abrams, where we have staked our all on this experiment that free trade in ideas becomes a fundamental substantive commitment of our democracy. Well, why accept free trade in Abrams, unless there is some reason to support it? And if there is, why do we want to fractionate the inquiry into individual freedom when the classical view of the subject covers not only speech, but also contract, and covers the two of them indifferently, that is, in exactly the same theoretical fashion?

The great mistake of modern constitutional law with respect to individual rights, whether it be with occupational and personal liberties or with takings, rests in the implicit assumption that a worldview which is good enough to explain our commitment to freedom of speech under First Amendment areas is worthless and completely contrary to sensible facts when we are trying to understand individual activities associated with economic behavior, property rights, or the sort. The great battle, as far as I can see, in constitutional discourse is whether or not we can persuade ourselves that the principles of classical liberalism, with their deep and powerful roots, are strong enough to energize our understanding of the Constitution. Recall that our Constitution is not a pure libertarian doctrine. It does recognize taxation; it does recognize the power to condemn property; it does allow the state to act when necessary to overcome collective action problems. It thus meets all the standard needs of government while preserving large and important areas for individual liberty. It embodies in a word the kind of moderation that I defended at the outset of this talk. The mission in the next generation is to revive all aspects of that heritage so that the forces of freedom and innovation can burn more brightly in this land, for the benefit of us all.

Richard A. Epstein is the James Parker Hall Distinguished Service Professor of Law at the University of Chicago Law School and the Peter and Kirstin Senior Fellow at the Hoover Institution.