THE STRUCTURE STILL STANDS: A REPLY TO NARVESON, WINSTON, & TUSHNET
Randy E. Barnett
8 Good Society 56
One should always be flattered by comments on one's works, whether favorable or not, from the likes of Professors
Narveson, Tushnet, and Winston. In this regard, I am no exception. That they should have agreed to contribute to this issue, I appreciate very much. It now falls to me to reply to their thoughtful comments without
seeming ungrateful for their willingness to comment at all.
I begin with Jan Narveson because we obviously share so much in common. In The Structure of Liberty, I attempt to explain the conception of
justice and the rule of law that, given the nature of human beings and the world in which we live, is required to enable persons to pursue happiness, peace and prosperity. And the pursuit of happiness, peace and prosperity
requires that we handle somehow the pervasive social problems of knowledge, interest, and power. This analysis depends, quite explicitly, on the assumption that the reader shares with me the objective of enabling persons to
pursue happiness, peace, and prosperity and on the fact that legislators implicitly claim their statutes are binding in conscience because they contribute to these ends. In this regard, the analysis is, as I note in the
introduction, "hypothetical" (as distinct from categorical) and, for this reason, in Narveson's view, it is incomplete and insufficient to justify fully the principles of justice and the rule of law that I identify.
"[Ilt will ... not do simply to specify that the purpose of law is to enable us all to live together happily. We must further show you, the rational individual, why you should buy into that purpose, and thus to pursue
your happiness in a way that commits you to allowing others to do so as well." (Narveson, 46) I do not disagree that my argument is incomplete and
therefore insufficient to justify fully the liberal conception of justice and the rule of law. I quite consciously refrained from providing the additional step that he urges upon me for at least three reasons. The first
and most obvious is one of competence. Dealing with philosophical foundations of this sort-if I may call them this-would be likely to push me beyond my ken even more than does the analysis I do present. This is not
false modesty on my part, but simple realism. In the words of Harry Callahan (as played by Clint Eastwood), "a man's got to know his limitations."
But more than a well-justified modesty prevented me from attempting to supply the foundation that Narveson understandably seeks. For, to the extent that we
are speaking of what would motivate the reader to accept with me the goal of a world in which people are able to pursue happiness, peace, and prosperity while living in close proximity to others, I suspect-and it has been my
experience that different arguments will appeal to different people. Some like Jan will find a social contractarian account persuasive. Some desire a teleological approach. Others will desire a more efficiency-based
analysis. Still others seek a deontological argument. I am not convinced that these different modes of justificatory analysis are mutually inconsistent. Instead, I think they can be mutually reinforcing where they
converge on the same results and mutually corrective where they diverge and must somehow be reconciled. But I would not like to defend this suspicion here so much as suggest that the debate about which justification is
sufficient to motivate people to find a way to enable people to pursue happiness, peace, and prosperity is a quite different and separable one than the issues I took 350 pages to examine.
My final reason for adopting the limited focus is that I believe that, whatever motivates one to accept the goal of enabling persons to pursue happiness, peace and
prosperity, it is necessary to be far more comprehensive and systematic about the means necessitated by that goal than purely "philosophical" accounts have typically been. Most philosophical treatments of rights are
not nearly instrumental enough. For this reason, it becomes necessary to argue that certain means-in particular certain principles of justice and the rule of law-are in fact essential means by which persons may pursue
happiness, peace, and prosperity while living in close proximity to other. Showing why this is so against those who would deny this "hypothetical" claim, is the burden I have undertaken in The Structure of Liberty
, and it is more than enough burden for me to bear, In any event, I trust that most readers do share the goal-even if they do not realize that the means that are necessary to its achievement may conflict with other goals they
may have. And I further contend that, for several reasons I provide in the introduction, legislation is not binding in conscience on the individual if it undermines this goal.
Just as The Structure of Liberty
is not a book about these philosophical issues, it is also not a book about democracy-notwithstanding the title of Winston's provocative essay. I am not sure I agree with him that libertarians are more or less "afraid" of democracy than socialists or welfare-statists or social conservatives or academics generally (or elected politicians for that matter), but the merits and demerits of democracy is an important issue that I only touch upon in passing in
The Structure of Liberty. This occurs during my discussion of advocates of monocentric legal orders who assume there must be a single ultimate authority above all others, and who seek to constrain that authority by a
system of voting. I praise the objective of establishing some sort of "reciprocity" between the ruled and their rulers, which is the principal virtue of voting. But I then go on to identify some of the
well-known weaknesses of electoral constraints on power: e.g. that voters lack sufficient information to make knowledgeable choices among policies (where given a choice) or candidates, that individual candidates and parties
represent such an amalgam of policy views that it is difficult to chose the lessor of the evils, and that majorities of voters are likely to vote out of their interests where those interests conflict with the principles of justice
that are needed to handle the problems of knowledge, interest, and power. My limited objective in noting these weaknesses is to show how other methods of
establishing reciprocity are likely to be more effective than voting. The weaknesses of depending upon voting to constrain both enforcement errors and abuse are so well-known-by libertarians and nonlibertarians alike-that I
was surprised that Winston would call them into question. Indeed, I would be shocked if they were not shared to some degree by Winston himself. Does he believe in the necessity of judicial review-that is, a judicially
enforceable "bill of rights" (whether written, like the right of free speech, or unwritten, like the right of privacy) that would override majority rule on occasion? (Did he support the nomination of
democratic-majoritarian Robert Bork to the Supreme Court?) If he shrinks from unrestrained majoritarianism, what is he "afraid" of? Probably the same things he says "libertarians" fear. If his
"fears" lead him to qualify his commitment to "democracy," then his approach is not so categorically different than mine-or than "libertarians" generally-as the title of his essay supposes.
Winston's repeated appeal to "democracy," however, misses the point of the thesis I present in The Structure of Liberty. This is revealed in
the very last sentence of his essay: "The challenge is to find the fight balance between individual flourishing and democratic solidarity." (Winston, 52) But how is one to go about finding such a balance? Appealing
to the virtue of "the idea of the community engaged in regulating itself," (Winston, 51) cannot answer this question for it is one of the two ideals he wants to balance. There must be some independent analytic or
institutional framework to help us decide the "balance" between "individual flourishing" and "democratic solidarity." Moreover, his formulation assumes what must be shown: that "democratic
solidarity" is the appropriate approach to civic order" (Winston, 51) and "associational values." (Winston, 51). Despite Winston's
claim to the contrary, I nowhere disparage either civic order or associational values and, in the final chapter, I specifically discuss their importance and their consistency with my conclusions. And, in focusing on my use of
illustrations based on Ann, Ben, and Cynthia, Winston overlooks the extent to which I insist that the problems of knowledge, interest, and power are social problems that would not exist if people were the atomistic individuals that
the left loves to accuse classical liberals of assuming. Moreover, given that I view them as solutions to pervasive social problems, the rights I defend in The Structure of Liberty,
could hardly be characterized as "pre-social rights" (Winston, 50). But assuming that democratic solidarity is the appropriate method of achieving
civic order and associational values, Winston's reply still misses the necessity of some such analysis as I present. For even one who strongly supports democratic self-rule, must still decide the sort of rights, if any, he or
she would vote for in a democratic regime. And if the voter (or legislator) shares the goal of a society in which people are able to pursue happiness, peace, and prosperity then- given the nature of human beings and the world
in which we live- the voter or legislator should respect the rights I defend and adhere to the requirements of the rule of law that I identify. In other
words, appeal to democratic solidity, community self-rule, or majoritarianism does not help much in telling the participant in such a community what rights and procedures they should vote to recognize. Something like the
analysis I present (or some alternative to it) needs to be presented. And it is this analysis with which Winston fails to come adequately to grips.
For example, I stress that the first-order problem of knowledge is enabling persons to put into practice the dispersed knowledge of time, place, and circumstances while somehow taking into account the vast ignorance that each of us
has of the circumstances of others. Winston suggests that there are other forms of knowledge-like theoretical knowledge or the knowledge of experts-besides personal and local knowledge and he is quite correct. But I
never denied the existence or importance of these other forms of knowledge. I only insisted on the need to enable people to make use of the knowledge of particular circumstances of time, place, etc. they have and that experts
necessarily lack. About this problem, Winston has nothing to say. He thinks it is "odd" (Winston, 51) for me to be concerned with the
problems of ignorant or advantage-taking voters, "given an absence of worry about consumer intelligence" (Winston, 51). But it always amazes me to hear about the virtues of democratic self-rule from those who would
not trust consumers to buy an automobile or meal without the coercively imposed assistance of experts. When it comes to making a decision concerning their personal situations, consumers are said to be incompetent and in need
of substantial assistance and protection; but these very same incompetent people can knowledgeably make laws to govern 250 million other people, or knowledgeably select from among their number the competent legislators who can do
this while keeping a constant eye on them too. Which is the more difficult or complex type of decisions? About which decision are consumers more likely to have or obtain reliable information? In sum, which is the
more realistic view of personal competence? Such limited confidence in individual competence to make personal decisions must also undermine a commitment to true
collective self-rule as well-and it always does. So under the guise of "democratic solidarity" invariably lies rule by benign legislators guided by benign experts.
Apart from using such terms, Winston offers no alternative ways to those I specify to handle the first-order problem of knowledge; nor does he deny the existence of
the problem. Nor for that matter does he acknowledge the second-and third-order problems of knowledge I discuss, or the problems of interest (partiality, incentives or compliance problems), or the problems of power
(enforcement error and abuse). My principal claim in The Structure of Liberty
is that these very real, very pervasive-indeed universal-problems can best be solved by adhering to the liberal conception of justice and the rule of law. If this analysis is correct, then voters or legislatures in a democracy should not violate these strictures. One can love democracy to the fullest extent and still accept the necessity for adhering to these rights and procedures.
Of course, a better understanding of the rights and procedures that ought to be respected if people are to pursue happiness, peace and prosperity
might also call into question the efficacy of "democratic solidarity" or "community self-rule" as a means of protecting or generating such rights-assuming some particular system of "democracy" is
specified. This would also require that some alternative to democracy needs to be identified. In the context of the legal system, I take some stabs at such an alternative in discussing a polycentric legal order
consisting of multiple legal systems. And, as Lon Fuller famously observed, "such multiple legal systems do exist and have in history been more common than unitary systems." (Lon L. Fuller, The Morality of Law
, p. 123) I doubt that Winston would refer to Fuller's confidence in such a legal order as "breezy" (Winston, 51), though Fuller well knew that for the better part of history such orders have lacked "regulatory
oversight" (Winston, 51) of the multiple legal systems they contain. Indeed, the phrase "regulatory oversight" reveals much about
the political commitment of Winston. He believes in "regulatory oversight" above all else. Oversight of the individual, oversight of the association, oversight of businesses ("the market"),
presumably oversight of the overseers-but by whom? There is the rub, though its not the only rub. The failure of those who believe in "regulatory oversight" to come to grips with the myriad, pervasive, and
universal problems of knowledge, interest, and power is the main thesis of The Structure of Liberty- along with the necessity for the sort of liberty structured by justice and the rule of law that believers in pervasive
"regulatory oversight" would easily override and even reject. Another believer in regulatory oversight is Mark Tushnet, though he takes
greater pains to show how such oversight is consistent with a natural law approach. Tushnet's argument is that a natural rights analysis of the sort I present in The Structure of Liberty is too abstract, uninformed by
particular facts, and indeterminate to exclude something as apparently illiberal as health and safety regulations. In his final sentence he says: "Information about the real world might not change the conclusions one
would draw, but in its absence Barnett provides a weaker case than people of his persuasion should like" (Tushnet, 53). In an important respect, I can agree with him and I discuss this in the book in a discussion of the
merits of "public policy" analysis. There I say:
Principles alone do not make us expert about any social problem; Take drug prohibition, for example. The abstract right of freedom to contract that I defend using a natural rights analysis would suggest
strongly that coercion should not be used to interfere with transactions between willing buyers and willing sellers of intoxicating substances. If, however, one is to address all the practical objections likely to be made
against this application of the general principle of freedom of contract, one must know a good deal about intoxicating substances and the myriad effects of legal prohibition on users, the legal system, and the public at large.
Nevertheless, even in the absence of this information, the principle of freedom of contract does indeed, at a minimum, argue for upholding a risk assessment term
when "a person knows she is signing a contract with a risk-wage term." (Tushnet, 52) Although I can imagine circumstances in which such a term (or many others) might be unenforceable given the existence of a valid
contract defense, the presumption created by the right of freedom of contract would require a great deal more to overcome than Tushnet is willing to offer. Further, the rule of law would require the formulation of a more
general exception to the normal enforceability of such a contract term that could be applied equally and predictably in like cases. Ironically, in the end, Tushnet offers only hypothetical empirical objections to my analysis
which he says is inadequately grounded in the real world. And he suggests no rule of law to apply in this circumstance in place of the norm of contractual freedom.
But it is wrong to conclude from an appropriate modesty about the persuasiveness of general principles that they can be dispensed with as easily as both Tushnet and
Winston would like. General principles are like a "cheat sheet" to a multiple choice exam. They go a great distance in telling you what the right answers are (or more accurately excluding a host of wrong
answers), without telling you all the "real world" considerations that make right answers right and wrong answers wrong. Nevertheless, the existence of
the cheat sheet is of crucial value in at least two ways. First, it helps one to focus one's "real world" inquiries on the relevant information one would otherwise be likely to overlook: Say the intense corruption
of law enforcement officers, the erosion of constitutional rights, and the development of increasingly potent and dangerous drugs, all caused, I submit, by drug prohibition.
Perhaps even more importantly, adherence to abstract and incompletely informative principles helps us to avoid potentially catastrophic social
experimentation. In their absence we are far more likely to enact "schemes of coercion that rarely if ever can be repealed, regardless of how useless and damaging they may turn out to be. Such repeal then only
comes after costing countless persons their happiness and even their lives." (SOL 324) Politicians and intellectuals who present themselves as "pragmatic" invariably fail to appreciate the pragmatic value of
relatively abstract principles of justice grounded on the sort of "weak premises" (Tushnet, 52) 1 employ. Tushnet thinks he has shown how a
regime of workplace regulations is generally consistent with freedom of contract and even with the underlying analysis that I offered in support of this natural fight. But anyone who reads The Structure of Liberty
would see that he could accomplish this transmutation only by changing substantially both the principles I defend there and the rationales for them. For example, I spend considerable time discussing personal knowledge which I define as "the knowledge unique to any particular persons of their personal perception, of their personal preferences, needs, and desires of their personal abilities, and of their personal opportunities." (31) Rather than quote my one sentence definition (which I then go on to elaborate), he refers instead to "Barnett's definition of personal knowledge, with appears to be a person's inner thoughts and feelings." (1) These definitions are far from the same thing and he never mentions the "local knowledge" about which I also am concerned.
Tushnet complains that "[Barnett] does not tell us whether one can have erroneous personal knowledge" (Tushnet, 1), whereas I say that
"Hayek's account does not assume that everything that people believe is true. Rather, it maintains that (a) there are many things each of us believes that are true and (b) access to these truths by others is severely
limited," (30). He says I do "not quite define liberty" (Tushnet, 52), yet, as I explain in the introduction, the whole project of the book is to define "liberty" (as distinct from license) as just
those freedoms permitted and protected by the structure of rights that comprise the liberal conception of justice. Given that the entire book is a "definition" of liberty, one sentence definitions of liberty of the
sort offered by Professor Tushnet cannot fail to seriously distort the concept I develop. Finally, and most importantly, by limiting himself to
considering only (his conception of personal knowledge but not the specific problem of knowledge I identify, he neglects the other pervasive social problems that are addressed by the right of freedom of contract: the
problems of interest-consisting of the partiality, incentive and compliance problems-and the problems of power consisting of the problems of enforcement error and abuse.
The thesis of The Structure Liberty is that the multiplicity
of pervasive social problems handled by the liberal conception of justice and the rule of law provides a very strong reason to adhere to its prescriptions. This thesis unavoidably required me to treat each of the problems less than completely comprehensively in
The Structure of Liberty, thus leaving me vulnerable to the sorts of complaints on this score made by Tushnet and Winston. Singling out one tive schemes simply does not confront the thesis I provide.
And even with all this jiggering and redefinition, what conclusions does Tushnet reach? After imagining "non-government mechanisms" to
"certify the honesty of the employer's offers and workers' demands" (Tushnet, 53),-mechanisms that would be entirely consistent with the liberal conception of justice I defend, he asserts: "But it is not obvious that
these mechanism are clearly better than government certification systems. Workers might
think that disclosures required by the government and enforced by public sanctions are more credible than disclosures monitored by private agencies enforced by contract remedies for fraud" (Tushnet, 52 [emphasis added]). But then again they might not. Or they might under a Democratic administration, but not under a Republican one, or ... This is just the sort of not-so-subtle burden-shifting based on "empirical" speculations that a commitment to principles of justice which handle the social problems of knowledge, interest and power is needed to avoid.
To seriously challenge the affirmative arguments I present in The Structure of Liberty, advocates of "regulatory oversight"
would need to present more than challenging questions and possibilities, or niggling disagreements with analogies or examples I employ, however interesting and helpful these may be. They would need to show (1) how restricting
or overriding the liberal conception of justice and the rule of law in favor of "regulatory oversight" will not undermine the ability of institutions governed by these rights and procedures to handle the problems of
knowledge, interest, and power and then (2) they must identify a specific scheme of "regulatory oversight" and provide good reasons to think it will effectively handle the problems of knowledge, interest, and power that
such schemes inevitably confront. In the alternative, they must either persuasively (3) deny these are indeed pervasive social problems, or (4) deny that the liberal conception of justice and the rule of law are needed to
handle them. Though I certainly cannot rule out the possibility of someone making any of these types of arguments, because neither Winston nor Tushnet
has directly addressed my central thesis in any of these ways, neither has presented what must be shown to call it into serious question. For this reason, if no other, and despite the intensity and ingenuity of their
critiques, I think the structure of liberty I identify and defend remains standing. Of course, readers of The Structure of Liberty will have to judge this for themselves.
Randy Barnett is the Austin B. Fletcher Professor Boston University School of Law.
that requires the sort of information that good public policy and
some experience on the ground can provide. In this way, public
policy analysis reinforces and sometimes helps reshape our principles (325).