Harvard University Tanner Lectures
On Human Values 2004-2005

Stephen Breyer
Associate Justice
Supreme Court of the United States

"Our Democratic Constitution"

Harvard University
Lowell Lecture Hall

November 17, 18 and 19, 2004
Cambridge, Massachusetts

Preliminary Draft
Not Yet Footnoted

The United States is a nation built upon principles of liberty - principles that include a democratic form of government. The French political philosopher Benjamin Constant understood the connection. He distinguished between liberty as understood by the ancient world, the Greeks and the Romans, and liberty as understood by the Nineteenth Century moderns.

Writing thirty years after the French Revolution, not long after the adoption of the American Constitution, Constant pointed out that "the liberty of the ancients" is an active liberty, consisting of a sharing of a Nation's sovereign authority among that Nation's citizens. From the citizen's perspective, it consists of "an active and constant participation in collective power." It means "submitting to all the citizens without exception the care and assessment of their most sacred interests." The sharing of sovereign authority, Constant said, would "ennoble[]" the people's "thoughts and establish[] among them a kind of intellectual equality which forms the glory and power of a people."

Constant distinguished the active, participatory "liberty of the ancients" from a more "modern" negative liberty that consists of "individual independence" from State control. He was no stranger to negative liberty. Having seen the Terror in action, Constant understood the need to protect the individual from the excesses of those who act in the name of a democratic majority. In recognizing the need to protect individuals from the unchecked power of even a democratically constituted State, Constant made clear that both ancient active and modern negative freedoms "are important." He said that we must embrace both, "learn[ing] to combine the two together."

Those who created our Constitution understood the liberty of which Constant spoke. They wrote a document that stresses the importance of modern "negative" liberty, but at its core is a blueprint for "ancient" active liberty, which they might have called "public liberty" and which we might call "democracy." Jefferson, for example, spoke of the rights of the citizen as "a participator in the government of affairs," - while Adams, his rival, added that all citizens have a "positive passion for public good."

My Tanner lectures are about this ancient liberty as embodied in our more modern Constitution. My thesis is that courts should take greater account of the Constitution's democratic nature when they make both constitutional and statutory decisions. That thesis encompasses well-known arguments for judicial modesty: the judge's comparative lack of expertise; the importance of assuring that "the people" themselves develop "the political experience and the moral education that come . . . from correcting their own errors;" an understanding of that doubt, caution, and prudence, that not being "too sure" of oneself, which Learned Hand described as "the spirit of liberty" itself. But my thesis goes beyond these traditional arguments. It finds in the Constitution's democratic objective not simply a restraint on judicial power or a counterweight to its protection of individual liberty but also a source of judicial power and an interpretive aid to making protection of that freedom more effective.

I shall focus on the people's constitutional right to "an active and constant participation in collective power." And I shall explain why I believe that increased judicial emphasis upon this right will help judges better resolve several specific important interpretive issues. In effect, I argue that greater judicial emphasis upon active liberty will help to bring about better law. "Better law," on my view, is law that enables the community to find practical solutions to important social problems through fair, open, and collective deliberation. In addition, my argument will appeal to consequences, to due consideration of whether and how a given interpretation actually promotes this better law. I shall therefore also explain why I believe that an approach to constitutional and statutory interpretation that places great weight upon consequences has considerable virtue.

These three lectures will have several parts of unequal length. First, I shall explain what I mean when I say that the Constitution centrally embodies the principle of active liberty, a principle the framers intended to thrive in practice. And I shall show how that view is consistent with the now-standard historical explanation of the Constitution's creation.

Second, I shall turn to a more lengthy discussion of six examples drawn from the areas of free speech, federalism, privacy, equal protection, statutory interpretation, and judicial review of administrative action. In each instance I shall consider a contemporary problem that seems to call for a governmental response. In each instance I shall focus upon related judicial issues of constitutional or statutory interpretation. In each instance I shall explain why I believe that increased recognition of the Constitution's democratic objectives (along with an appreciation of the role courts have to play in securing those objectives) can help judges deal more effectively with the interpretive issues, thereby helping communities deal better with the problems that have called those issues into being.

Third, I shall link my discussion to a broader interpretive approach, which emphasizes a consideration of consequences. I shall contrast this approach with that of judges who would place primary interpretive weight upon language, history, tradition, and precedent. I shall explain why I believe the latter approach, by undervaluing consequences, undervalues related constitutional objectives, and thereby exacts a constitutional price that is too high.

My discussion reflects throughout three observations related to my professional experience. First, a Supreme Court Justice, faced with a steady diet of constitutional questions, inevitably develops a view of the Constitution as a whole. My own view is that the document creates a coherent framework for a certain kind of government. Described in general terms that government is democratic; it avoids concentration of too much power in too few hands; it protects personal liberties; it insists that the law respect each individual equally; and it acts only on the basis of law itself.

Of course, the document embodies these general objectives in discrete provisions. In respect to democratic government, for example, the Constitution insists that Congress meet at least once each year, that elections take place every two (or four or six) years, that representation be based upon a census that must take place every decade; and it has gradually made suffrage virtually universal, extending it to all adult men and women, irrespective of, for example, race or religion. (It also guarantees the States a "republican form of government.")

But a general constitutional objective, such as that of self-government, plays an interpretive role well beyond the individual provisions that refer to it directly. Individual phrases make up a single document. That document embodies general purposes. And those purposes inform the interpretation of the phrases. By openly and clearly referring to those general purposes, courts can help to create harmonious relationships both among different individual provisions and among the several general constitutional purposes just mentioned. My examples help to show how that is so.

Second, most judges, whether interpreting a statute or the Constitution itself, refer to language, to history, to tradition, to precedent, to purpose, and to consequences. All are important. I do not suggest that purpose itself, or any purposive notion such as "positive liberty," provides a magic key to constitutional interpretation. At most, it will sometimes help a judge in his or her use of this entire set of traditional tools. I do not counsel abandonment of any of them.

Third, individual Justices proceeding case by case normally do not think of themselves as applying an "ideology" or belonging to a "school of thought." But it is nonetheless possible to speak meaningfully about changes in Court "philosophy" over time. That is because the Court, at different periods in its history, has read the Constitution as emphasizing different sets of objectives. Thus, one can characterize the early nineteenth century as a period during which the Court, through its interpretations of the Constitution, helped to establish the authority of the federal government, including the federal judiciary. One can characterize the late nineteenth and early twentieth centuries, known as the Court's Lochner era, as a period during which the Court under-emphasized the constitutional importance of participation by black citizens in our representative democracy and over-emphasized the importance of constitutional protections of property.

Later Courts - the New Deal Court and the Warren Court - emphasized the Constitution's protection of the citizen's freedom to participate in government, i.e., the citizen's "active liberty." The former dismantled various Lochner-era distinctions, thereby expanding the scope of democratic self-government. The latter interpreted the Civil War Amendments in light of their purposes, thereby helping African-Americans become members of the Nation's community of self-governing citizens - a community that the Court expanded further in its "one person, one vote" decisions.

Assuming that this kind of broad characterization is possible, my lectures suggest that the contemporary Court has swung too far back, too often under-emphasizing or overlooking the contemporary importance of active liberty as currently understood.

The Theme

The concept of "active liberty" - as I said at the outset - refers to a sharing of a nation's sovereign authority among its people. Sovereignty involves the legitimacy of a governmental action. And a sharing of sovereign authority suggests several kinds of connection between that legitimacy and the people.

For one thing, it should be possible to trace without much difficulty a line of authority for the making of governmental decisions back to the people themselves - either directly, or indirectly through those whom the people have chosen, perhaps instructed, to make certain kinds of decisions in certain ways. And this authority must be broad. The people must have room to decide and leeway to make mistakes.

For another, the people themselves should participate in government - though their participation may vary in degree. Participation is most forceful when it is direct, involving, for example, voting, town meetings, political party membership, or issue- or interest-related activities. It is weak, but still minimally exists, to the extent that it is vicarious, reflected, say, in the understanding that each individual belongs to the political community with the right to participate should he or she choose to do so.

Finally, the people, and their representatives, must have the capacity to exercise their democratic responsibilities. They should possess the tools, such as information and education, necessary to participate and to govern effectively.

When I speak of active liberty, I mean to suggest connections of this kind between the people and their government - connections that involve responsibility, participation, and capacity. Moreover, active liberty cannot be understood in a vacuum, for it operates in the real world. And in the real world, institutions and methods of interpretation must be designed in a way such that this form of liberty is both sustainable over time and capable of translating the people's will into sound policies.

An Historical Connection

Is it reasonable from an historical perspective to view the Constitution as focusing first and foremost upon active liberty, upon the right of individuals to participate in democratic self-government? I believe so. I have already listed various constitutional provisions that specifically further that objective. And the now-standard historical accounts of the writing of the Constitution - in the works, for example, of Gordon Wood and Bernard Bailyn - make clear that active liberty, the principle of participatory self-government, was a primary force shaping the system of government that the document creates

The primarily democratic nature of the Constitution's governmental structure has not always seemed obvious. John Adams, for example, understood the Constitution as seeking to create an Aristotelian "mixed" form of government. Our government, like the British government, would reflect the structure of Eighteenth Century society. The House of Representatives, like the House of Commons, would constitute the "democratical branch" of the new federal government, embodying the people's basic decency and common sense. The Senate, like the House of Lords, would represent the aristocratic element of society, embodying its wisdom while checking the people's sometimes "barbarous and cruel" passions. The Executive would represent the monarchical element of society, with the President serving as a mediator, a balancer, helping to keep the social forces in equilibrium.

But Adams himself recognized that his notions of constitutionalism were not widely shared. And historians now tell us that the consensus view among the Framers was a different one, namely that the Framers sought to create a governmental structure that would reflect the fact that sovereign authority originated in the people. They believed that the "right to legislate is originally in every Member of the Community," that the Constitution should reflect that fact, and that only certain imperatives - the need for a workable government and the need to protect individual (negative) liberty - might warrant modification of this basic principle.

Democracy, of course, could not mean a Greek city-state. The Nation's geographic size, along with its large and growing population, would prevent replication at the national level of the Athenian agora or a New England town meeting. The people would have to delegate the day-to-day work of governance. But the people could continue to share sovereign authority; they could continue to participate actively in the governing processes. "Delegated democracy" need not represent a significant departure from democratic principle.

Moreover, in the view of modern historians, post-revolutionary American political thought was characterized by suspicion of government, hostility to the Executive Branch, and confidence in democracy as the single effective check upon government's oppressive tendencies. The former colonists, now Americans, saw "radical destruction of magisterial authority" as the way - perhaps the only way - to keep power in check, to prevent its arbitrary exercise. They embraced the concept of "public liberty," believing that "liberty in a State is self-government." A free people, they believed, is a people that government cannot oppress, for the reason that the people have "a constitutional check upon the power to oppress." Thus, during the time between the end of the Revolutionary War and the writing of the Constitution, the American public came to the conclusion that first and foremost democratic principle must underlie the structure of post-revolutionary government.

After the Revolution, the citizens of many former colonies translated their democratic beliefs into highly democratic forms of state government. Pennsylvania, for example, experimented with a constitution that abolished the position of governor, substituting a twelve member elected council; created a unicameral legislature with one-year terms; imposed strict, four-year term limits; insisted that all public decision-making take place in public; and provided for a Board of Censors, a kind of state-wide grand jury with separately elected members who would investigate all actions by the legislature and report to the public. Indeed, in many of the colonies governors were forbidden to participate in the lawmaking function; impeachment was common; and terms of office were short. Most Americans accepted the Whig maxim, "where annual elections end, tyranny begins."

Why then did the Framers not write a Constitution that contained similar democratic structures? Why did they not, like Pennsylvania, approximate a closer-to-Athenian version of democracy? Why did they create so complex a form of government, placing more distance between electors and elected than even the needs of "delegation" of democratic authority might demand?

The reason, in part, is that experience with many of these initial forms of democratic government had proved disappointing. Pennsylvanians found that their government enacted conflicting policies, reflecting the vagaries of shifting public opinion; that through debt repudiation it had produced an insecure climate for business; and that those within government - a continuously changing group - were often at war with one another. Similarly, Massachusetts saw in the Shays Rebellion a public that would fight to avoid not only debt repayment but also taxation of any sort. Other States had faced similar, though perhaps less dramatic, difficulties.

Nonetheless, despite these difficulties, the Framers did not abandon their basically democratic outlook. That is the main point. The Framers continued to believe that power must flow from the people. Madison, emphasizing the Constitution's democratic origins, described it as "a charter of power granted by liberty," not (as in Europe) "a charter of liberty granted by power." The Constitution begins with the words "We the people." And through the use of that phrase, Alexander Meiklejohn tells us, "it is agreed, and with every passing moment it is re-agreed, that the people of the United States shall be self-governed."

The Constitution subsequently implements its Preamble by vesting legislative power in a House of Representatives and a Senate - both bodies made up of individuals who are ultimately responsible to the people. Article One specifies that members of the House will be "chosen every second Year by the People of the several States," i.e., by voters who "shall have the Qualifications requisite for Electors of the most numerous Branch of the State Legislature." That Article also originally specified that Senators would be "chosen by" state "Legislatures." But in so specifying the Framers did not seek to model the "Senate" upon the House of Lords. Rather, eighteenth century supporters of a Senate argued that this second legislative body would increase democracy by providing for "double representation." They pointed out that in every State citizens chose their state legislators through elections. And given the importance of the Senatorial position, it seems likely that the voters would have held their state legislators to account for their national Senatorial choices.

Article II vests Executive power in a President, selected by an electoral college, not the voters. But this mechanism does not create a Presidency free from democratic control. Rather, the Constitution grants state legislators, elected by and accountable to the people, the power to determine how to select the state's electors. In 1789 this meant election by legislators in five states, by the people in four states, and by mixed methods in two states (two States did not participate). By 1836 it meant electors chosen directly by the people in every state but South Carolina (which switched to popular election in 1860). This popular connection now means (and meant at the time) that the President and Senators would consider themselves responsible to, or representing the interests of, not a particular social class, but "We the People."

Thus, John Wilson, an influential figure at the Constitutional Convention summed up the Framers' conception of the non-legislative branches as follows:

"The executive and judicial power are now drawn from the same source, are now animated by the same principles, and are now directed to the same ends, with the legislative authority: they who execute, and they who administer the laws, are so much the servants, and therefore as much the friends of the people, as those who make them."

And John Taylor, writing in 1790, described the Constitution's structure in terms that are difficult to reconcile with a retreat from democratic principle. "Power," he said, "is first divided between the government and the people, reserving to the people the control of the dividend allotted to the government." The government's allotment of power is then "distributed in quotas still more minute" to its various branches. But though the power is dispersed, the people themselves continue to control the policy-making activities of these different branches of government.

One might argue that these descriptions vastly over-state the Framers' commitment to democracy. As I have just said, the Constitution seems to create a governmental structure far more complex, and in part far more distant from the people, than principles of delegated democracy demand. Does not that fact reflect a profound retreat from democratic structure, in the direction, as Adams suggested, of aristocratic government? Not necessarily so. That is because we can find in these same constitutional facts not so much a retreat from democratic principle as an effort by the Framers to produce a government committed to democratic principle that also, as a practical matter, would work. That is to say, we can find in the Constitution's structural complexity an effort to create a form of democratic government likely to escape those tendencies to produce the self-destructive public policies that the Pennsylvania and Massachusetts experiments had revealed - a form of democratic government that could produce legislation that would match the needs of the nation.

Consider, for example, what Madison called the problem of "faction." As described by Gordon Wood, the problem grew out of the fact that the new nation encompassed divergent social, economic, and religious interests. There were "rich and poor; creditors and debtors; a landed interest, a moneyed interest, a mercantile interest, a manufacturing interest" and numerous subdivisions within each category. The States' post-Revolutionary experience demonstrated that the natural tendency of these groups was to choose representatives not for their "abilities, integrity, or patriotism" but for their willingness to act solely to advance the group's particular interests. This often meant that "the great objects" of society were "sacrificed constantly to local views." The unicameral state legislatures, with their small electoral districts, large number of seats, and annual terms might have come close to the Athenian vision of true democracy. But these bodies were "bulging and fluctuating" and "filled with such narrow-minded politicians who constantly mistook 'the particular circle' in which they moved for the 'general voice' of society." The framers' goal was to "secure the public good and private rights against the danger of [factionalism], and at the same time to preserve the spirit and form of popular government."

How did they achieve that goal? Madison, in Federalist Number 10, said that the answer was to broaden the electoral base so that more members of government owe their position to the many. "If elected officials were concerned with only the interest of those who elected them, then their outlook was most easily broadened by enlarging their electorate." The base could not be made too broad, to the point where the elected official loses contact with the voter. But it must be broad enough to stifle the propensity "to rash measures and the facility of forming and executing them." It must be broad enough so that "no one common interest or passion will be likely to unite a majority of the whole number in an unjust pursuit." Madison predicted that this broadening would also have the effect of drawing out "representatives whose enlightenment and virtuous sentiments render them superior to local prejudices and schemes of injustice" and more likely to pursue the true interests of the Nation. And a bicameral legislature would prevent usurpation of the people's ultimate power by forcing designing men to control two houses instead of one and by dividing the "trust" of the people between "different bodies of men, who might watch and check each other."

Consistent with Madison's analysis, the Constitution provides that a House Member's electoral district will remain small while a Senator's district and the President's district will encompass the entire state and the entire nation respectively. The larger districts, by including many diverse interests, lessen the likelihood that a particular faction will win influence at the expense of the general well-being of a constituency's citizens. Moreover, House Members can serve only two-year terms and need be only twenty-five years old, but Senators serve for six years and must be thirty. The President, though serving for four years, must be thirty-five. The longer terms of office and the older the minimum required age, the greater the insulation from short-term caprice of public sentiment and the more likely the elected official would be a proven leader rather than an untested political heir. At the same time, Senators and President would remain responsible to the people through election, by state legislators or through the electoral college.

One can also see in the Constitution's structure an effort by the Framers to protect "negative liberty" - the freedom of the individual from oppressive governmental action. The new Americans had initially thought that a thoroughly democratic government based on public liberty would naturally protect the individual rights of it citizens. They believed that securing "the right of the people to participate in the government" was the best way to secure the negative liberty of individuals. But the State government experiments in less disciplined democracy had proved disappointing in this respect as well, bringing about what some called a new form of despotism.

How would the Framers protect the citizen from the practice of this despotism by the very government their Constitution would create? They found one answer in a complex structure with checks and balances among federal branches, along with delegation to the federal government of limited powers, which diffused power and prevented impetuous action by the central government. They thought the courts might provide additional protection, by interpreting the Constitution's delegation of limited powers to the federal government as excluding the authority to take action that deprived individual citizens of their (negative) liberty. State Constitutions added further protections to curb the excesses of State government. But ultimately the Framers, like Benjamin Constant, recognized that neither these protections nor active liberty, the "liberty of the ancients," would prove sufficient. They added a Bill of Rights with explicit protections against government interference with certain fundamental personal liberties. The Civil War amendments later expanded and more fully secured these basic negative liberties.

The important point for present purposes is that history permits me to answer my original question affirmatively. From an historical perspective, one can reasonably view the Constitution as focusing first and foremost upon "active liberty." The Framers wrote a Constitution that is in its "principles, structure, and whole mass, purely and unalterably Democratic." They included elements designed to "control and mitigate" the ill effects of more direct forms of democratic government, but in doing so, "the Framers did not see themselves as repudiating either the Revolution or popular government." Rather, they were "saving both from their excesses." As Madison said of self-government in Federalist 39:

"It is evident that no other form [of government] would be reconcilable with the genius of the American people; with the fundamental principles of the Revolution; or with that honorable determination which animates every votary of freedom to rest all our political experiments on the capacity of mankind for self-government.

Finally, the very act of ratifying the Constitution, by means of special state elections with broad voter eligibility rules, signaled the democratic character of the document itself.

I recognize that this account is abbreviated, leaving out necessary qualifications (including restrictions, for example, of race and sex, upon the Eighteenth Century electorate). And it draws upon the work of only a few modern historians. Yet for present purposes it will suffice. It makes plausible a certain view of the framers' basic intent, a view that sees the framers as seeking to create a form of government in which all citizens share the government's authority, participating in the creation of public policy. It understands the Constitution's structural complexity as responding to certain practical needs, for delegation, for non-destructive (and hopefully sound) public policies, and for protection of basic negative freedoms. And it views the Constitution's democratic imperative as accommodating, even insisting upon, these practical needs. In sum, our constitutional history has been a quest for workable government, workable democratic government, workable democratic government protective of individual (negative) liberty. Our central commitment has been to "government of the people, by the people, for the people." The bulk of these lectures will illustrate how this constitutional understanding helps to resolve important interpretive problems related to modern government.


The principle of active liberty - the need to make room for democratic decision making - argues for judicial modesty in constitutional decision-making, a form of judicial restraint. But there is more to it than that. Increased recognition of the Constitution's democratic objectives - and an appreciation of the role courts can play in securing those objectives - can help guide judges both as actors in the deliberative process and as substantive interpreters of relevant constitutional and statutory provisions. To show how this is so, I shall use examples drawn from the areas of free speech, federalism, privacy, establishment of religion, equal protection, statutory interpretation, and review of administrative action. Each example considers modern government-related problems that call for a democratically-based response. And each raises difficult questions of constitutional or statutory interpretation. In describing how I would approach these questions, I shall show how increased emphasis upon the Constitution's participatory objectives can help bring about better law.


I begin with free speech. I shall consider how the First Amendment applies where the government seeks to regulate certain activities affecting speech, in particular campaign finance, corporate advertising about matters of public concern, and drug-store advertising about the availability of custom-made pharmaceuticals. The discussion of these examples will help show the importance of reading the First Amendment, not in isolation, but as seeking to maintain a system of free expression itself designed to further a basic constitutional purpose: creating and maintaining democratic decision-making institutions.

I begin where courts normally begin in First Amendment cases. They try to classify the speech at issue, distinguishing among different speech-related activities for the purpose of applying a strict, moderately strict, or totally relaxed, presumption of unconstitutionality. Is the speech "political speech," calling for a strong pro-speech presumption, "commercial speech" calling for a mid-range presumption, or simply a form of economic regulation presumed constitutional? Should courts begin in this way?

Some scholars argue that making these kinds of categorical distinctions is a misplaced enterprise. The Constitution's language makes no such distinctions. It simply protects "the freedom of speech" from government restriction. "Speech is speech and that is the end of the matter." But to limit distinctions to the point where First Amendment law embodies the slogan "speech is speech" cannot work. And the fact that the First Amendment seeks to protect active, as well as negative, liberty, helps to explain why.

The democratic government that the Constitution creates now regulates a host of activities that inevitably take place through the medium of speech. Today's workers manipulate information, not wood or metal. And the modern, information-based workplace, no less than its more materially-based predecessors, requires the application of community standards seeking to assure, for example, the absence of anti-competitive restraints, the accuracy of information, the absence of discrimination, the protection of health, safety, the environment, the consumer, and so forth.

Laws that embody these standards obviously affect speech. Warranty laws require private firms to include on labels statements of a specified content. Securities laws and consumer protection laws insist upon the disclosure of information that businesses might prefer to keep private. Health laws forbid tobacco advertising, say, to children. Anti-discrimination laws insist that employers prevent employees from making certain kinds of statements. Communications laws require cable broadcasters to provide network access. Campaign finance laws restrict citizen contributions to candidates.

To treat all these instances alike, to scrutinize them all as if they all represented a similar kind of legislative effort to restrain a citizen's negative liberty to speak would lump together too many different kinds of activities under the aegis of a single standard, thereby creating a dilemma. On the one hand, if strong First Amendment standards were to apply across the board, they would prevent a democratically-elected government from creating necessary regulation. The strong free speech guarantees needed to protect the structural democratic governing process would (if applied without distinction to all governmental efforts to control speech) unreasonably limit the public's substantive economic (or social) regulatory choices. The limits on substantive choice would likely exceed what any liberty-protecting framework for democratic government could require, depriving the people of the democratically necessary room to make decisions, including the leeway to make regulatory mistakes. That, along with a singular lack of modesty, was the failing of Lochner. No one wants to replay that discredited history in modern First Amendment guise.

On the other hand, to apply across the board uniform First Amendment standards weak enough to avoid the shoals of Lochner would weaken the First Amendment to the point where it did not offer sufficient protection for the free exchange of ideas necessary for structural reasons, i.e., to maintain the health of our democracy. Most scholars, including "speech is speech" advocates, consequently see a need for distinctions. The question is, which ones, applied where?

At this point, reference back to the Constitution's more general objectives helps. First, "active liberty" is particularly at risk when law restricts speech directly related to the shaping of public opinion, for example, speech that takes place in areas related to politics and policy-making by elected officials. That special risk justifies especially strong pro-speech judicial presumptions.

Second, where ordinary commercial or economic regulation is at issue, this special risk normally is absent. Moreover, strong pro-speech presumptions risk imposing what is, from the perspective of active liberty, too severe a restriction upon the legislature - a restriction that would dramatically limit the size of the legislative arena that the Constitution opens for public deliberation and action. The presence of this second risk warns against use of special, strong pro-speech judicial presumptions or special regulation-skeptical judicial review.

The upshot is that reference to Constitutional purposes and "active liberty" in particular helps to justify the category of review that the Court applies to a given type of law. But, as we shall now see, those same considerations argue, among other things, against category boundaries that are too rigid or fixed and against too mechanical an application of those categories. Rather, reference to active liberty will help courts define and apply the categories case by case.

Consider campaign finance reform. The campaign finance problem arises out of the explosion of campaign costs, particularly those related to television advertising, together with vast disparity in ability to make a campaign contribution. In the year 2000, for example, election expenditures amounted to $1.4 billion; and the two presidential candidates spent about $310 million. In 2002, an off-year without a presidential contest, campaign expenditures still amounted to more than $1 billion. A typical House election cost $900,000, with an open seat costing $1.2 million; a typical Senate seat cost about $4.8 million, with an open contested seat costing about $7.1 million.

Comparable expenditures in foreign democracies are far lower. A typical British or Canadian parliamentary election involves expenditures for individual seats of about $13,000 and $43,000 respectively. (Television costs explain much of the difference. The cost of television advertising now approximates $10,000 per minute in a major city. In the 2000 election, parties and candidates spent between $770 million and $1 billion on television ads).

A small number of individuals and groups underwrite a very large share of these costs. In 2000, about half the money the parties spent, roughly $500 million, was soft money, i.e., money not subject to regulation under the then-current campaign finance laws. Two-thirds of that money - almost $300 million - came from just 800 donors, each contributing a minimum of $120,000. Of these donors, 435 were corporations or unions. The rest, 365, were individual citizens. At the same time, 99 percent of the 200 million or so citizens eligible to vote gave less than $200. Ninety-six percent gave nothing at all.

The upshot is a concern, reflected in campaign finance laws, that the few who give in large amounts may have special access to, and therefore influence over, their elected representatives or, at least, create the appearance of undue influence. (One study found, for example, that 55% of Americans believe that large contributions have a "great deal" of impact on how decisions are made in Washington; fewer than 1% believed they had no impact). These contributions (particularly if applied to television) may eliminate the need for, and in that sense crowd out, smaller individual contributions. In either case, the public may lose confidence in the political system and become less willing to participate in the political process. That, in important part, is why legislatures have tried to regulate the size of campaign contributions.

Our Court in 1976 considered the constitutionality of the congressional legislation that initially regulated campaign contributions; and just last year we considered more recent legislation that tried to close what Congress considered a loophole - the ability to make contributions in the form of unregulated soft money. The basic constitutional question, as you know, does not concern the desirability or wisdom of the legislation but whether, how, and the extent to which, the First Amendment permits the legislature to impose limits on the amounts that individuals or organizations or parties can contribute to a campaign. I shall not go into the details of the Court's decisions, but I shall sketch an approach to decision-making that draws upon the Constitution's democratic objectives.

One cannot (or at least I cannot) find an easy answer to this basic constitutional question in language, in history, or in tradition. The First Amendment's language says that Congress shall not abridge "the freedom of speech." But it does not define "the freedom of speech" in any detail. The nation's founders did not speak directly about campaign contributions. Madison, who decried faction, thought that Members of Congress would fairly represent all their constituents, in part because the "electors" would not be the "rich" any more than the "poor." But this kind of statement, while modestly helpful to the cause of campaign finance reform, is far from determinative.

Neither, in my view, can we find the answer through the use of purely conceptual arguments. Some argue, for example, that "money is speech." Others say, "money is not speech." But neither contention helps. Money is not speech, it is money. But the expenditure of money enables speech; and that expenditure is often necessary to communicate a message, particularly in a political context. A law that forbade the expenditure of money to communicate could effectively suppress the message.

Nor does it resolve the problem simply to point out that campaign contribution limits inhibit the political "speech opportunities" of those who wish to contribute more. Indeed, that is so. But the question is whether, in context, such a limitation is prohibited as an abridgement of "the freedom of speech." To announce that the harm imposed by a contribution limit is under no circumstances justified is simply to state an ultimate constitutional conclusion; it is not to explain the underlying reasons.

The better approach, however, is to consider the First Amendment's relation to the Constitution's basic structural objective, "participatory self-government." To understand the First Amendment as protecting active liberty is to understand it as protecting more than the individual's "negative" freedom. It is to understand the Amendment as seeking to facilitate a conversation among ordinary citizens that will encourage their informed participation in the electoral process. It is to suggest a constitutional purpose that goes beyond protecting the individual from government restriction of information about matters that the Constitution commits to individual, not collective, decision-making. It is to understand the First Amendment as seeking primarily to encourage the exchange of information and ideas necessary for citizens themselves to shape that "public opinion which is the final source of government in a democratic state." In these ways, the Amendment helps to maintain a form of government open to participation (in Constant's words) "by all citizens without exception."

To focus upon that First Amendment's relation to the Constitution's democratic objective is helpful because the campaign laws seek to further a similar objective. They seek to democratize the influence that money can bring to bear upon the electoral process, thereby building public confidence in that process, broadening the base of a candidate's meaningful financial support, and encouraging greater public participation. They seek thereby to maintain the integrity of the political process - a process that itself translates political speech into governmental action. Insofar as they achieve these objectives, those laws, despite the limits they impose, will help to further the kind of open public political discussion that the First Amendment seeks to sustain, both as an end and as a means of achieving a workable democracy.

To emphasize the First Amendment's protection of active liberty is not to find the campaign finance laws automatically constitutional. Rather, it is to recognize that basic democratic objectives, including some of a kind that the First Amendment seeks to further, lie on both sides of the constitutional equation. Seen in terms of "negative liberty," they include protection of the citizen's speech from government interference; seen in terms of "active liberty," they include promotion of a democratic conversation. That, I believe, is why our Court has refused to apply a strong First Amendment presumption that would almost automatically find the laws unconstitutional. Rather the Court has consistently rejected "strict scrutiny" as the proper test, instead examining a campaign finance law "close[ly]" while applying what it calls "heightened" scrutiny. In doing so, the Court has emphasized the power of large campaign contributions to "erode[] public confidence in the electoral process." It has noted that contribution limits are "aimed at protecting the integrity of the process"; pointed out that in doing so they "tangibly benefit public participation in political debate"; and concluded that that is why "there is no place for the strong presumption against constitutionality of the sort often thought to accompany the words 'strict scrutiny.'"

Consequently, the Court will try to look realistically both at a campaign finance law's negative impact upon those primarily wealthier citizens who wish to engage in more electoral communication and its positive impact upon the public's confidence in, and ability to communicate through, the electoral process. And it will apply a constitutional test that I would describe as one of proportionality. Does the statute strike a reasonable balance between electoral speech-restricting and speech-enhancing consequences? Or does it instead impose restrictions on speech that are disproportionate when measured against their electoral and speech-related benefits, taking into account the kind, the importance, and the extent of those benefits, as well as the need for the restrictions in order to secure them?

I doubt that courts, in trying to answer these questions, need totally abandon what I have referred to as judicial modesty. Courts can defer to the legislatures' own judgment insofar as that judgment concerns matters (particularly empirical matters) about which the legislature is comparatively expert, such as the extent of the campaign finance problem, a matter that directly concerns the realities of political life. But courts should not defer when they evaluate the risk that reform legislation will defeat the participatory self-government objective itself. That risk is present, for example, where laws set contribution limits so low that they elevate the reputation-related or media-related advantages of incumbency to the point of insulating incumbent office holders from effective challenge.

A focus upon the Constitution's democratic objective does not offer easy answers to the difficult questions that campaign finance laws pose. But it does clarify the First Amendment's role in promoting active liberty and suggest a legal framework for addressing those and other vexing questions. In turn, that framework can help the Court arrive at tough but faithful answers. Modesty suggests where, and how, courts should defer to the legislature in doing so.

The inquiry is complex. But courts both here and abroad have engaged in similarly complex inquiries where the constitutionality of electoral laws is at issue. That complexity is demanded by a Constitution that grants to Congress the power to secure a fair electoral system while requiring judges to conduct First Amendment review of Congress's decisions.

Reference to the Constitution's "participatory self-government" goal helps resolve other kinds of First Amendment problems, for example those raised by commercial speech such as commercial advertising. To what extent does the First Amendment protect that speech from government regulation?

Several recent cases have focused upon the question. In one of them the Court considered a California law that, as interpreted, allowed any member of the public to bring a "deceptive business practices" claim against a corporation that had advertised and distributed its denials of charges that it was engaging in disreputable business practices abroad. The California court upheld the law, and ultimately our Court refused review.

In another, the Court struck down a law forbidding pharmacists to advertise the availability of individual "compound drugs," drugs that the pharmacist makes up specially for patients with unique requirements due, for example, to drug-related allergies. The Court said that the First Amendment forbade any statutory effort to restrict information in order to help the public make wiser decisions. The Court thought the pharmacy law was precisely such an effort.

In each case I disagreed. I thought that the business practice speech was primarily political and not subject to regulation. I thought that the pharmacist's speech was primarily commercial and subject to regulation. If the Court had seen active liberty as the moving force behind the creation of, and application of, the First Amendment's categorical distinctions, then these cases might have come out differently.

In the corporate advertising case, Nike, a business corporation, tried to defend itself against claims by newspapers, human rights groups, and labor organizations that it had engaged in disreputable business practices abroad. Nike denied these claims in letters sent, for example, to newspapers and to college athletic directors. The California Supreme Court permitted a private citizen to sue Nike for false advertising - on the ground that Nike's denials were themselves false. And that court found that the First Amendment gave Nike no especially strong protection against such a lawsuit. Our Court initially decided to consider whether that decision was correct, but because of certain procedural difficulties, it ultimately dismissed the claim without reaching the merits. I did not agree with the Court's procedural conclusions and I would have reversed the California Supreme Court, which had denied strong First Amendment protection to Nike's responses.

In the advertising case, Congress forbade pharmacists to advertise specific compound drugs because those drugs had not been tested. While doctors would know of their existence and could prescribe them when necessary, advertising would generate strong patient demand, leading doctors to prescribe a given untested compound to patients for whom it was only a convenience, not a necessity. The law, in permitting prescriptions, but forbidding advertising to consumers, struck a compromise between the patient's special need for drug compounds and the special risk of harm due to the fact that compound drugs (being individualized and special) had not met ordinary safety testing requirements.

The Court majority thought that any subsequently created safety risk could not justify a law that, in the Court's view, reflected a "fear" that "people would make bad decisions if given truthful information about compounded drugs." But, for the reasons I have set forth when discussing "speech is speech," that omnipresent function - providing information - is not by itself sufficient to warrant a strong anti-regulatory presumption. And without such a presumption, the existence of widespread prescription drug advertising, the medical belief that "direct-to-consumer advertising pressures physicians into prescribing drugs they would not ordinarily prescribe," and the small but definite safety risk present in untested drugs - together would have justified the "information restricting" law at issue.

From the perspective of a First Amendment that seeks first and foremost to facilitate democratic self-government, the two courts' results in these two cases seem backwards. Nike responded to criticisms of its labor practices. Those criticisms lie at the center of an important public debate in which participants urge or oppose public collective action. Nike's speech sought to shape related public opinion. It sought to do so under the permissive legal standards that govern speech of its opponents. If the "false advertising" lawsuit goes forward, Nike (and other potential speakers), out of reasonable caution or even an excess of caution, might well censor their own expression well beyond what the law could constitutionally demand. What could be more central to basic First Amendment concerns?

The pharmacists' speech, by contrast, did not directly serve any such democratic purpose. The pharmacists did not seek through price advertising to contribute to a public debate about the relative merits of compound drugs. At most, they conveyed information that would help patients make more informed private decisions about what drugs to ask their physicians to prescribe. But this purpose, while important, is not so important that it justifies striking down legislation that regulates speech for sound reasons related to the traditional regulation of public health and safety. A contrary view of the First Amendment standard here fails to further, indeed it impedes, the workings of a democratically determined economic regulatory system. It restricts Congress's regulatory powers, preventing the public from achieving related objectives that the community democratically determines to be important.

I do not mean that the First Amendment leaves Congress free to enact any regulatory law it wishes related to commercial speech or economic regulation. Traditional, "modern," negative liberty - the individual's freedom from government restriction - remains important. Individuals need information to flow freely to make decisions about their own lives. And, irrespective of context, a particular rule affecting speech might, in a particular instance, require individuals to act against conscience, inhibit public debate, threaten artistic expression, censor views in ways unrelated to a program's basic objectives, or create other risks of abuse. These possibilities themselves form the raw material out of which courts will create different presumptions applicable in different speech contexts. And even in the absence of presumptions, courts will examine individual instances with the possibilities of such harms in mind.

My argument is that, in applying First Amendment presumptions, we must distinguish among areas, contexts, and forms of speech. Reference to basic general, constitutional purposes can help generate the relevant distinctions. And reference back to at least one general purpose, "active liberty," helps both to generate proper distinctions and also properly to apply the distinctions generated. The "active liberty" reference helps us to preserve speech that is essential to our democratic form of government, while simultaneously allowing the law to deal effectively with such modern regulatory problems as campaign finance and product or workplace safety.


I turn next to federalism. Recent Court cases on that subject lead me to focus on two general questions of modern government. First, how can we reconcile democratic decision-making with the highly technical nature of many government decisions? Second, what level of government is best suited to the making of which decisions? I cannot provide a general answer to these questions. But I can suggest that, by taking explicit account of the Constitution's liberty-related objectives, judicial decisions will help the three branches of government together arrive at better answers.

In one sense, the Constitution's federal structure helps to protect negative liberty. A division of powers among federal and state governments makes it more difficult for the federal government to tell state and local governments what they must do. And it thereby frees citizens from restraints that a more distant central government might otherwise impose. Yet it leaves citizens subject to similar restraints imposed by the States themselves. Thus it seems more natural to view the structure as helping to secure more effective forms of active liberty, i.e. as facilitating meaningful citizen participation in government by preserving a more local decision-making process.

My colleague Justice O'Connor has explained the connections well. By guaranteeing state and local governments broad decision-making authority, federalist principles secure decisions that rest on knowledge of local circumstances, help to develop a sense of shared purposes and commitments among local citizens, and ultimately facilitate "novel social and economic experiments." Through increased transparency, those principles make it easier for citizens to hold government officials accountable. And by bringing government closer to home, they help maintain a sense of local community. In all these ways they facilitate and encourage the "ancient liberty" that Constant described: Citizen participation in the government's decision-making process.

Today, this participation principle must be implemented against the backdrop of a highly complex set of technology-based social problems that defy decision purely at local or purely at federal levels. Rather, they call for a federal/state cooperation that permits effective action while respecting the liberty I have just described.

A concrete example may show how that is so. Consider the regulation of toxic chemicals. Some toxic chemical regulation must take place at the national level. Chemical substances, traveling through air, water, or soil, often affect the environment in more than one state. Chemical substance regulation demands scientific and technical expertise often more readily available at the federal level. Federal regulation, because it is national in scope, can facilitate the development of a national understanding about chemical dangers, say by promoting a simple, uniform language for talking about safety risks. And only a federal regulator can set minimum substantive standards designed to avoid a race to the bottom among states hoping to attract industry investment.

Yet some aspects of the problem would seem better regulated at a state or local level. Similar amounts of similar chemicals in air or water or soil may have different toxic effects depending upon local conditions. Similar emissions standards may have different economic effects depending upon differing economic circumstances in different communities. And different communities may place different weights upon similar clean-up (risk-reduction) benefits and costs. These differences suggest that citizens in different local communities may come up with different answers to the same basic questions: How clean should our local waste dump be? And at what cost?

The idea of "cooperative federalism" tries to provide regulatory answers that take account of these needs. Federal agencies no longer rely exclusively upon classical command-and-control regulation. They have supplemented that form of regulation with less restrictive, more incentive-based, methods, including taxes and marketable rights. Federal agencies can make expertise available to state and local officials without imposing their will. Those agencies may help separate questions of common interest (particularly those that require expertise) from more locally-oriented questions of fact or value. These approaches mean more decision-making authority for local governments; they place greater power in the hands of individuals; and, in doing so, they further the liberty interests - both active and negative - that underlie federalist principles.

To what extent have the Court's recent federalism decisions taken these considerations into account? Each Court holding helps to some degree to protect negative liberty - in the narrow sense described. That is to say, the decisions limit the federal government's ability to control the activities of individuals and businesses. But in respect to the furtherance of active liberty these decisions are often retrograde. They discourage use of the cooperative, incentive-based regulatory methods that I have just mentioned. Thus, in unanticipated ways, they paradoxically threaten to shift regulatory activity from the State and local, to the federal, level - the likely opposite of their objective.

Consider first the Court's holding that federalism means that Congress may not write laws that "commandeer" a State's legislative or executive officials. Thus, Congress cannot require a state legislature to write a particular kind of law, for example, a nuclear waste storage law. Nor can Congress enact a law that requires a local official to spend time enforcing a federal policy, for example, a law that requires a local sheriff to see whether a potential gun buyer has a criminal record. These interpretations of the Constitution's federalism requirements stop Congress from enlisting local officials to check compliance with federal minimum standards. They thereby force Congress to either forgo the program in question altogether or, perhaps more likely, to expand the size of the program-related federal enforcement bureaucracy. Other things being equal and given ordinary bureaucratic tendencies, this fact will make it harder, not easier, to shift regulatory power from the federal government to state and local governments. And it will make it harder, not easier, to experiment with incentive-based regulatory methods.

Justice Stevens illustrated the problem. In a dissent, presciently written before the terrorist attacks of September 11, 2001, he wrote that the "threat of an international terrorist, may require a national response before federal personnel can be made available to respond. . . . [What in the Constitution] forbids the enlistment of state officials to make that response effective?" Would freedom to enlist state officials not help to advance both the cause of national security and the cause of co-operative federalism?

Consider next, the Court's decisions that have significantly limited Congress's power (under the Commerce Clause or the Fourteenth Amendment) to require a State to waive its Eleventh Amendment immunity from suit by private citizens. Judged in terms of their consequences, it is difficult to see how these decisions advance cooperative federalism. To the contrary, they make it more difficult for Congress to create uniform individual remedies under legislation dealing with nationwide problems, for example, private civil damages actions for citizens injured by a state's unlawful use of their intellectual property. Less obviously, but as importantly, they will prevent Congress from adopting certain forms of "less restrictive" regulation - forms that mean less federal government interference in state, local, or personal matters.

Suppose, for example, that Congress, reluctant to expand a federal regulatory bureaucracy, sees citizen suits as a way to ensure that State entities (as well as private entities) comply with toxic waste dump legislation. Suppose that Congress, in an effort to achieve a particular environmental goal, directs State governments to impose environmental taxes and permits private citizens to sue a State to protest a particular tax assessment or to obtain a tax refund. Or suppose that Congress, anxious to shrink the size of, say, its federal maritime law enforcement staff, permits individual citizens to bring federal administrative proceedings against state port authorities, providing for federal enforcement only where administrative proceedings determine them to be justified. The Court's Eleventh Amendment decisions rule out these less restrictive, less bureaucratic methods of enforcing federal law - a consequence inconsistent with both negative freedom and the Constitution's vision of active liberty.

Finally, consider cases in which the Court has limited the scope of Congress's Commerce Clause powers. The Court has found that gun possession near local schools and violence against women in local communities do not sufficiently "affect" interstate commerce to permit Congress to legislate. Decisions in this third category do mean less federal regulation. They do not directly discourage citizen participation in "incentive-based," or "cooperative" state/federal, regulatory programs. But in these instances the public has participated in the legislative process at the national level. Indeed, Congress held elaborate public hearings only to find its legislative work nullified.

Moreover, these cases indirectly may discourage the development of complex cooperative programs. That is because the Court's own close scrutiny of, and re-weighing of, the evidence that Congress found sufficient to show "inter-state effects" create uncertainty about how much evidence is needed to find the constitutionally-requisite effect. Certain portions of the Court's reasoning, such as its refusal to consider the "aggregate" effect on interstate commerce of individually small "non-economic" events, aggravate that uncertainty. Congress then may not know whether it can, or it cannot, legislate the details of a particular "cooperative" - federal, state, local - regulatory framework. And there is no speedy way for it to find out. Those circumstances, other things being equal, make it less likely that Congress will enact complex laws that might well embody cooperative federalist principles.

The consequences that I have just described do not show the Court's federalism decisions are wrong. (Though I believe they are wrong and that the Court was right in yet more recent cases to slow down, perhaps to halt, this development.) The examples simply raise such questions as: Why should courts try to answer difficult federalism questions on the basis of logical deduction from text or precedent, alone? Why not ask about the consequences of decision-making on the active liberty that federalism seeks to further? Why not at least consider the practical effects on local democratic self-government of decisions interpreting the Constitution's principles of federalism - principles that themselves seek to further that very kind of government?

The examples also suggest a need for a better "dialogue" between Congress and the Court in this area. Judge-made law interpreting the Dormant Commerce Clause foresees such a dialogue. That doctrine asks courts to decide whether a state regulatory law unreasonably inhibits trade with other states. The doctrine weighs basic principles of federalism against local economic protectionism, a matter highly relevant in an increasingly global economy. Does a state law that, for example, prohibits importing peaches grown with certain pesticides, insists on the use of special steel for elevator cables, or prevents interstate trucks from transporting dynamite during daylight hours, reasonably protect the State's citizens from dangerous pesticides, faulty elevators, and risks of explosion? Or does it unreasonably protect the State's peach growers, steel makers, and contractors from out-of-state competition?

In this area of constitutional "federalism" law, Congress is free to overturn by statute a judicial decision with which it disagrees. Congress can even delegate the power to decide such matters to an expert agency. The federal Department of Transportation, for example, after providing opportunity for public comment, can decide many "dormant commerce clause" questions subject only to judicial review of the reasonableness of its decision. The doctrine, while facilitating the use of expert opinion, thereby permits the public acting through its elected representatives to have the last word. It encourages judicial modesty in enforcing Commerce Clause objectives, leading courts to defer to the conclusions of the democratic process.

The Court might also consider specific legal doctrines that would promote a similar dialogue in respect to federalism more generally. Through a hard look requirement, for example, the Court would communicate to Congress the precise constitutional difficulty the Court has with the statute at issue without resorting to permanent invalidation. Congress, in reenacting the statute, would revisit the matter and respond to the Court's concerns. A clear statement rule would have the Court call upon Congress to provide an unambiguous articulation of the precise contours and reach of a given policy solution. Those doctrines would lead the Court to focus upon the thoroughness of the legislature's consideration of a matter, thereby encouraging public participation, and the explicit nature of its conclusion, thereby promoting clarity and consequent accountability. As their names suggest, they would require Congress to look hard at and speak clearly on a matter, but they rarely would create an absolute "federalism-based" bar to legislation. Such an approach treads carefully and with restraint when courts consider the validity of a legislative enactment, and it is consequently consistent with a Constitution that emphasizes active liberty.

I am not arguing here, however, whether the Constitution permits development of these kinds of legal doctrine. Rather, I point to the Constitution's democratic objectives, explain the complexity involved in attaining those objectives where modern technical decision-making is at issue, note the related tension between those objectives and the recent cases, and suggest that proper resolution of many such federalism issues cannot be left to the judiciary alone. There are likely better ways.


My third example focuses upon privacy. By privacy, I mean a person's power to control what others can come to know about him. It illustrates constitutional decision-making under conditions of uncertainty - uncertainty brought about by rapid changes in technology. Where technological advance means significant change in the regulatory environment, Americans normally search pragmatically for new legal answers, and they often participate in a democratic conversation along the way. Judicial respect for this process often counsels a special degree of judicial caution.

Let me illustrate what I have in mind by describing the privacy-related legal problem. That problem arises out of three factors: the variety of values implicated by our concern for privacy; the need for already complicated legal regimes to accommodate new technologies; and the difficulty of balancing competing (sometimes conflicting) concerns in this complex area of law.

First, an array of different values underlie the need to protect personal privacy. Some scholars emphasize the values related to an individual's need to be left alone, not bothered by others, perhaps adding that privacy prevents us from being judged on the basis of a single preserved private fact taken out of context. Others emphasize the way in which important personal relationships, of love and friendship, depend upon trust, which, in turn implies a sharing of information not available to all. Others find connections between personal privacy and individualism in that privacy may encourage non-conformity and more free expression. Still others, for similar reasons, find connections between privacy and equality; for example, an inability to obtain highly individualized information about customers can lead businesses to treat all customers alike. One might weight these different considerations differently; but still almost everyone finds in them important relationships to an individual's dignity; and almost all Americans accept the need for legal rules to protect that dignity.

Second, most of our privacy-related legal challenges lie at the intersection of a legal circumstance and a technological circumstance. The legal circumstance consists of the fact that several different types of laws are involved in regulating privacy. Some laws, such as trespass, wiretapping, eaves-dropping, and search-and-seizure statutes, protect particular places or sites, such as homes or telephones, from searches and monitoring. Other laws protect, not places, but kinds of information, for example, certain personal data about another person. These different laws protect privacy to different degrees depending upon place, an individual's status, the type of information, and the kind of intrusion at issue.

The technological circumstance consists of the fact that advancing technology has made the protective effects of present law uncertain, unpredictable, and incomplete. Video cameras can now monitor shopping malls, schools, parks, office buildings, city streets, and other places that current law had left unprotected. Scanners and interceptors can overhear virtually any electronic conversation. Thermal imaging devices can detect from outside the home activities taking place within it. Technology now provides us with the ability to observe, collate, and permanently preserve a vast amount of information about individuals - information that the law did not prohibit people from collecting but which, in practice, was not readily collectible or easily preserved. These technological changes have altered the practical, privacy-related effect of the set of previously-existing laws.

The legal circumstance and the technological circumstance taken together mean 1) a complex set of pre-existing laws 2) applied in rapidly changing circumstances. That application means changed, perhaps diminished, privacy protection with the extent to which protection diminishes varying depending upon individual circumstances. To maintain pre-existing protection, we must look for new legal bottles to hold our old wine.

Third, revision of our laws affecting privacy requires balancing (not always agreed-upon) interests in a host of different areas of human activity in light of uncertain predictions about the technological future. The answer to the balancing question - how to balance the interests - is often far from clear.

Suppose, for example, that businesses using computers, obtain detailed consumer purchasing information and create individualized customer profiles. Some believe that their possession of those profiles significantly diminishes the customer's privacy. But the profiles may also help firms provide products better tailored to fit customers' desires, and at lower costs.

Suppose, for example, that hospitals place an individual's medical records on line. Doing so may compromise the patient's privacy. But the ready on-line availability of those records may also lower insurance costs or help a patient who arrives unconscious at an emergency room.

Suppose, for example, that the law insists that information about an individual's genetic make-up must remain confidential. That law will protect that individual's privacy. But what happens if a close relative, a nephew or cousin, needs that information to assess her own cancer risks?

It is tempting to think we can resolve these dilemmas simply by requiring that an individual whose privacy is threatened be informed and grant consent. But an "informed consent" requirement does not necessarily work. Consent forms can be signed without understanding. And, in any event, a decision by one individual to release information or to keep it confidential often affects the lives of others.

All of this is bound by what is technologically possible. Should the law require programming video cameras on public streets to turn off at certain times? When? Should the law require software that instructs computers to delete certain kinds of information? Which? Should the law require encrypted cell-phones? Should the law impose upon certain web sites a requirement that they permit users with certain privacy preferences to negotiate access-related privacy conditions? How? Is such software available?

Think back over what I have been saying and you will begin to understand the legal, technological, and value-balancing complexity involved in trying to resolve the legal aspects of the personal privacy problem. I cannot offer solutions. But I can suggest how twenty-first century Americans go about finding solutions. The way they do so is best described as a form of participatory democracy.

Ideally, in America, the law making process does not involve legislators, administrators, or judges imposing law from above. Rather, it involves changes that bubble up from below. Serious complex legal change often is made in the context of a national conversation involving, among others, scientists, engineers, businessmen and women, the media, along with legislators, judges, and many ordinary citizens whose lives the new technology will affect. That conversation takes place through meetings, symposia, and discussions, through journal articles and media reports, through administrative and legislative hearings and through court cases. Lawyers participate in this discussion, translating specialized knowledge into ordinary English, defining issues, often creating consensus. Typically administrators and legislators make decisions only after the conversation is well under way. Courts participate later in the process, determining whether, say, the legal result reached through this "bubbling up" is consistent with basic constitutional norms. This "conversation" is the democratic process in action.

The nature of the law-revision problem together with the process of democratic resolution counsels a special degree of judicial modesty and caution. That is because a premature judicial decision risks short-circuiting, or pre-empting, the "conversational" law-making process - a process that embodies our modern understanding of constitutional democracy.

A recent case will illustrate the point. The Court considered a private cell-phone conversation that an unknown private individual had intercepted with a scanner and delivered to a radio station. A statute forbade the broadcast of that conversation, even though the radio station itself had not planned or participated in the intercept. The case required the Court to determine the scope of the station's First Amendment right to broadcast, given the privacy interests that the statute sought to protect.

Justice Stevens, speaking for four members of the Court, wrote that the key constitutional value at issue was a First Amendment interest in furthering public discussion of matters of public concern. The First Amendment trumped the statute, permitting the station to broadcast the information. But the opinion nonetheless favored a narrow holding. It focused upon the particular circumstances present in the case - the fact, for example, that the station had had nothing at all to do with obtaining the intercept. Justice O'Connor and I concurred emphasizing the potential importance of, and the current uncertainty about, the privacy interests at issue. We explicitly left open the possibility that a broadcaster would be liable in less innocent circumstances or where less pressing public concerns favored disclosure.

The narrowness of the holding itself serves a constitutional purpose. The democratic "conversation" about privacy is ongoing. In those circumstances, a Court decision that mentions its concerns without creating a binding rule could lead Congress to rewrite eaves-dropping statutes, tailoring them to take account of current technological facts, such as the widespread availability of scanners and the possibility of protecting conversations through encryption. A broader constitutional rule might itself limit legislative options in ways now unforeseeable. And a broad decision is particularly dangerous where statutory protection of an important personal liberty is at issue. By way of contrast, the Court also recently held unconstitutional police efforts to use, without a warrant, a thermal imaging device placed on a public sidewalk in order to identify activities within a private home. The case is different because it required the Court simply to ask whether the residents had a reasonable expectation that their activities within the house would not be disclosed to the public in this way, i.e., the privacy harm at issue is relatively clear and the applicable Fourth Amendment principle, comparatively speaking, is well-established. The case required the Court, not to look for new legal categories, but rather to fit new technology into old categories. It was less likely that doing so would interfere with any ongoing democratic policy debate.

The privacy example suggests more, in respect to judicial caution. It warns against adopting an overly rigid method of interpreting the Constitution - placing weight upon eighteenth century details to the point where it becomes difficult for a twenty-first century court to apply the document's underlying values. At a minimum it suggests that courts, in determining the breadth of a constitutional holding, should look to the effect of a holding of a certain breadth on the ongoing policy-creating process. They should distinguish for those purposes between, say, the "eavesdropping" and the "thermal imaging" kinds of cases.

The example also makes clear that it is misleading to contrast "practical" and "legal" judicial concerns. In exercising caution, a judge is not deserting the "judicial" role of law interpreter in order to be "practical." Rather, the judge is following the law, interpreting the Constitution in light of the Constitution's own concern for active liberty. Because, as I have previously noted, the constitutional concern itself is a practical concern, the Constitution authorizes, perhaps requires, courts to proceed "practically" when they examine new laws in light of the Constitution's enduring values.

Affirmative Action

My first three examples have focused upon problems of participatory government and problems of free speech - areas that one might describe as "active liberty's" natural home. My fourth example looks further afield. It focuses upon judicial efforts to determine whether a law school's affirmative action program was consistent with the Equal Protection Clause. It illustrates how reference to democratic self-government can help a court decide a very different kind of constitutional question.

In the recent affirmative action case Grutter v. Bollinger, the Court considered the University of Michigan's use of race as a law school admissions criterion. The Law School, an elite institution, receives about 3,500 applications each year for admission to a class of about 350. The School said that it seeks "students who individually and collectively . . . [are] the most capable," who have "substantial promise for success" in, and after, law school, and who will likely contribute "to the well-being of others." To obtain those students, the school considered an applicant's grade point average, Law School Admissions Test score, and recommendations. After ruling out any applicant whom it believed would not "do well enough to graduate" without "serious academic problems," it factored into its decision certain "soft variables," including the quality of the applicant's essay, the difficulty of undergraduate courses taken, unusual life experiences, and - most important for present purposes - minority race.

The School sought to enroll a "critical mass" of minority students, i.e., a number sufficient to encourage "underrepresented minority students to participate in the classroom and not feel isolated." Why? The School said that it considered race as an admissions factor in order to achieve racial "diversity." It wanted diversity in order "to enrich everyone's education." And to achieve diversity it needed affirmative action, favoring groups "historically discriminated against, like African-Americans, Hispanics, and Native Americans." Without affirmative action, it added, those groups "would not be represented" in the law school student body "in meaningful numbers."

The question before the Court was whether this use of race as an admissions factor by a State school was consistent with the Equal Protection Clause - a Clause that forbids any State to "deny to any person . . . the equal protection of the laws." The answer depended in significant part upon which of two possible interpretations of the Clause that the Court would accept.

On the first view the Clause insists that State activity must be "color blind." Justice Thomas, writing in dissent, explained that view as follows:

The Constitution abhors classifications based on race, not only because those classifications can harm favored races or are based on illegitimate motives, but also because every time the government places citizens on racial registers and makes race relevant to the provisions of burdens or benefits, it demeans us all. 'Purchased at the price of immeasurable human suffering, the equal protection principle reflects our nation's understanding that such classifications ultimately have a destructive impact on the individual and our society.

On the second view courts must understand the Clause as more narrowly purposive. It grows out of a history that includes this Nation's efforts to end slavery and the segregated society that followed. It reflects that history. It consequently demands laws that equally respect each individual; it forbids laws based on race when those laws reflect a lack of equivalent respect for members of the disfavored race; but it does not similarly disfavor race-based laws in other circumstances. Justice Ginsburg, writing in a companion case, explained that view as follows:

In implementing [the Constitution's] equality instruction . . . government decision makers may properly distinguish between policies of exclusion and inclusion. . . . Actions designed to burden groups long denied full citizenship stature are not sensibly ranked with measures taken to hasten the day when entrenched discrimination and its after effects have been extirpated.

The Civil War Amendments sought to permit and to encourage those "long denied full citizenship stature" to participate fully and with equal rights in the democratic political community. Experience suggested that a "color blind" interpretation of those Amendments, while producing a form of equal opportunity, was insufficient to bring about that result. Hence, in purposive terms, invidious discrimination and positive discrimination were not equivalent.

These two views, one color-blind, one purposive, reflect not polar opposites but rather different interpretive tendencies. Those who favor the "color blind" nonetheless concede that sometimes, on rare occasions, the Clause permits distinctions based on race. Those who favor the "narrowly purposive" view concede that courts nonetheless must carefully scrutinize any legal classification based on race, for without careful examination, courts may fail to "ferret out classifications [that] in reality [are] malign, but [which are] masquerading as benign." The Court must also take a hard look to ensure that benign racial preferences are "not so large as to trammel unduly upon the opportunities of others or interfere too harshly with the legitimate expectations of persons in once-preferred groups."

The Court majority in Grutter ultimately adopted a form of the second view. It scrutinized the law school's use of race carefully, indeed it said "strictly." But it nonetheless found the Law School's diversity rationale "compelling." Because the School considered each application individually, it believed that the School's affirmative action program was "narrowly tailored" to achieve that objective. It added that it "expect[ed] that twenty-five years from now" this "use of racial preferences would no longer be necessary."

For present purposes, I want to focus on only one part of the Court's argument, namely why the Court accepted an interpretation of the Equal Protection Clause that was closer to the second than to the first view that I have advanced. The grounds for accepting that interpretation might have involved the claim that past discrimination against minorities can justify special efforts to help members of minority groups today. This claim rests upon considerations of equality. And equality is the underlying objective of the Equal Protection Clause. Justice Brennan explained the claim many years ago when he said that the "constitution is color conscious to prevent discrimination being perpetuated and to undo the effects of past discrimination." The Law School's admissions policy similarly referred to the university's commitment to "diversity with special reference to the inclusion of students from groups that have been historically discriminated against." But the Law School did not press this kind of equality-based remedial claim in our Court. Its hesitancy may have reflected the fact that the Court in earlier cases has cast doubt on the constitutional validity of affirmative action that seeks simply to remedy earlier "general societal discrimination."

The grounds for accepting the "narrowly purposive" view might have included a liberty-based claim - a claim that the Constitution grants universities especially broad authority to determine for themselves the composition of their student bodies. Justice Powell, for example, wrote in University of California Regents v. Bakke that the "freedom of a university to make its own judgments as to education includes the selection of its student body." The Law School argued for a more fully participatory form of education, stating that "in the context of higher education, a compelling state interest includes a "diversity" that promises "educational benefits," "help[ing to] break down racial stereotypes," enabling "students to better understand persons of different races," "promot[ing] cross-racial understanding," and producing "livelier, more spirited, and simply more enlightening and interesting" discussions. The Court, accepting a form of this argument, pointed out that "given the important purpose of public education and the expansive freedoms of speech and thought associated with the university environment, universities occupy a special niche in our constitutional tradition." But the Court's opinion does not suggest that these considerations, related to free expression, are determinative.

Instead, the Court placed important weight upon certain practical considerations, which Justice O'Connor, writing for the Court, described as follows:

Major American businesses have made clear that the skills needed in today's increasingly global marketplace can only be developed through exposure to widely diverse people, cultures, ideas, and viewpoints.

She added:

High-ranking retired officers and civilian leaders of the United States military assert that 'based on their decades of experience,' a 'highly qualified, racially diverse officer corps . . . is essential to the military's ability to fulfill its principal mission to provide national security.

She then said:

Student body diversity . . . better prepares students for an increasingly diverse work force and society and better prepares them as professionals," adding that "education is pivotal to 'sustaining our political and cultural heritage [and plays] a fundamental role in maintaining the fabric of our society.

She drew these themes together with these words:

Nowhere is the importance of openness more acute than in the context of higher education. Effective participation by members of all racial and ethnic groups in the civic life of our Nation is essential if the dream of one Nation, indivisible is to be realized. . . . [Indeed,] the path to leadership must be visibly open to talented and qualified individuals of every race and ethnicity. All members of our heterogeneous society must have confidence in the openness and integrity of the educational institutions that provide this training. . . . [And] all must participate.

What are these arguments but an appeal to principles of solidarity, to principles of fraternity if you like, or as I would claim for present purposes, to principles of active liberty? They find some form of affirmative action necessary to maintain a well-functioning participatory democracy. They say that an interpretation of the Equal Protection Clause that would outlaw the Law School's affirmative action program is an interpretation that, from the perspective of the Constitution's basic democratic objectives, would not work. Too many individuals of all races would lack experience with a racially diverse educational environment helpful for their later effective participation in today's diverse civil society. Too many individuals of minority race would find the doors of higher education closed; those closed doors would shut them out of positions of leadership in the armed forces, in business, and in government as well; and too many would conclude that the Nation and its governmental processes are not ours, but theirs. If these are the likely consequences - as many knowledgeable groups told the Court they were - could our democratic form of government then function as the Framers intended?

When faced with one interpretation of the Equal Protection Clause that, through efforts to include, would facilitate the functioning of democracy and a different interpretation of the Equal Protection Clause that, through perceived exclusion, might impede the functioning of that democracy, is it surprising that the Court majority chose the former? Is not that interpretation more compatible with a Constitution that seeks to create a democratic government able, as a practical matter, to function? Given that constitutional objective, it is not surprising that the Court interpreted the Equal Protection Clause in a way that diminishes the risk of serious racial division - a division that exclusion from elite educational institutions would aggravate. Nor should it be completely surprising that, in light of similar risks created through excessive racial self-identification and resulting strife, the majority wrote of its "expectation" that in twenty-five years policies like the Law School's would no longer be necessary.

I explore affirmative action no further because I have made my point. Sometimes reference to active liberty can help a court choose between competing interpretations of constitutional provisions that, on their face, seem based upon other values. Grutter's interpretation of the Equal Protection Clause provides one such example.

Statutory Interpretation

My fifth example focuses on statutory interpretation. It contrasts a more literal text-based approach with an approach that places more emphasis on statutory purpose and congressional intent. It illustrates why judges should pay primary attention to a statute's purpose in difficult cases of interpretation where language is not clear. It shows how over-emphasis on text can lead courts astray, divorcing law from life - indeed, creating law that harms those whom Congress meant to help. And it explains why a purposive approach is more consistent with the framework for a "delegated democracy" that the Constitution creates.

The interpretive problem arises when statutory language does not clearly answer the question of what the statute means or how it applies. Why does a statute contain such language? Perhaps Congress used inappropriate language. Perhaps it failed to use its own drafting expertise or failed to have committee hearings, writing legislation on the floor instead. Perhaps it chose politically symbolic language or ambiguous language over more precise language - possibilities that modern, highly-partisan, interest group based politics (responding to overly-simplified media accounts) often make likely. Perhaps no one in Congress thought about how the statute would apply in certain circumstances. Perhaps it is impossible to use language that foresees how a statute should apply in all relevant circumstances.

The founding generation of Americans understood these or similar possibilities. They realized that judges, though mere "fallible men," would have to exercise judgment and discretion in applying newly codified law. But they expected that judges, when doing so, would remain faithful to the legislators' will. The problem of statutory interpretation is how to meet that expectation.

Most judges start in the same way. They look first to the statute's language, its structure, and its history in an effort to determine the statute's purpose. They then use that purpose (along with the language, structure and history) to determine the proper interpretation. Thus far, there is agreement. But when the problem is truly difficult, these factors without more may simply limit the universe of possible answers without clearly identifying a final choice. What then?

At this point judges tend to divide in their approach. Some look primarily to language and text-related circumstances for further enlightenment. They may try to tease further meaning from the language and structure of the statute itself. They may look to language-based canons of interpretation in the search for an "objective" key to the statute's proper interpretation, say a canon like noscitur sociis, which tells a judge to interpret a word so that it has "the same kind of meaning as its neighbors." Textualism, it has been argued, searches for "meaning in structures." It means "preferring the language and structure of the law whenever possible over its legislative history and imputed values." It asks judges to avoid invocation of vague or broad statutory purposes and instead to consider such purposes at "lower levels of generality." It hopes thereby to reduce the risk that judges will interpret statutes subjectively, substituting their own ideas of what is good for those of Congress.

Other judges look primarily to the statute's purposes for enlightenment. They avoid the use of interpretive canons. They allow context to determine the level of generality at which they will describe a statute's purpose - in the way that context tells us not to answer the lost driver's request for directions, "where am I?" with the words "in a car." They speak in terms of Congressional "intent," while understanding that legal conventions govern the use of that term to describe, not the intent of any, or every, individual legislator, but the intent of the group - in the way that linguistic conventions allow us to speak of the intentions of an Army or a team, even where they differ from those of any, or every, soldier or member. And they examine legislative history, often closely, in the hope that the history will help them better understand the context, the enacting legislators' objectives, and ultimately the statute's purposes. At the heart of a purpose-based approach stands the "reasonable Member of Congress" - a legal fiction that applies, for example, even when Congress did not in fact consider a particular problem. The judge will ask how this person (real or fictional), aware of the statute's language, structure, and general objectives (actually or hypothetically), would have wanted a court to interpret the statute in light of present circumstances in the particular case.

Three recent cases illustrate the difference between the two approaches. In each the majority followed a more textual approach; the dissent, a more purposive approach.

Case One. The Foreign Sovereign Immunities Act says that a foreign nation, when it is a defendant in commercial litigation, can sometimes successfully assert a defense of "sovereign immunity" and thereby avoid liability. Where the applicability of the Act (including its many exceptions to the sovereign immunity rule) is in doubt - say because the plaintiff claims that a defendant corporation, even if part of a foreign government, is an ordinary commercial enterprise - the defendant can remove the case to federal court to permit a federal judge to decide whether or how the Act should apply. But a defendant can do so only if "a majority of" its "shares or other ownership interest is owned by" a foreign nation.

I can illustrate the issue in the case before us as follows: Suppose that Ruritania's government owns 100 percent of Ruritania Furniture Company's stock. If a plaintiff sues Ruritania Furniture Company in a state court, Ruritania Furniture can remove the case to federal court to permit the federal judge to determine how the Act applies. But suppose the state court plaintiff sues, not Ruritania Furniture Company, but Ruritania Chair Company, a subsidiary of Ruritania Furniture. Suppose further that Ruritania Furniture owns 100 percent of the stock of Ruritania Chair. If Ruritania owns Ruritania Furniture, which in turn owns Ruritania Chair, can Ruritania Chair remove the case to federal court? Does it qualify under the statute? Does Ruritania's ownership of a parent that in turn owns a subsidiary mean, in the words of the statute, that Ruritania possesses an "other ownership interest" in the subsidiary? If so, Ruritania Chair can remove the case; otherwise it cannot.

Judges using a more literal text-based approach are likely to find that the wholly owned subsidiary of the wholly owned parent cannot remove the case. The majority of my Court reasoned: American corporate law ordinarily considers a corporation and its shareholders as distinct entities. Thus, ordinarily the law deems that the corporation, not the government that owns the corporation, is the owner of the corporate subsidiary. Other linguistic clues reinforce this point. For example, had Congress wished to depart from this ordinary rule, it might have used the phrase "direct and indirect ownership." That phrase, found in some other statutes, would have signaled a Congressional intent to disregard the ordinary corporate "structural ownership rule[]." But Congress did not use it. Instead, Congress wrote language that refers both to 1) "other ownership interest" and also 2) "ownership of a majority of . . . shares." To give strong meaning to the word "other," we should interpret it to mean "other than ownership by way of share ownership." And the need to read the word "other" in this way is supported by the interpretive canon that says that "a statute must, if possible, be construed in such fashion that every word has some operative effect." These linguistic and structural facts, taken together, indicate that the words "other ownership interest" do not include ownership of shares in a parent that, in turn, owns a subsidiary.

A more purpose-oriented judge will likely come to the opposite conclusion. The dissenters in the same case reasoned: The purpose of the Act's jurisdictional provision is to bring into federal court cases in which a foreign government owns a commercial defendant. The Act will thereby allow the foreign government to take advantage of federal procedural protections that state court systems sometimes lack. Given this purpose, why would Congress, or any reasonable member of Congress, want to grant "this protection to a Foreign Nation acting through a Corporate Parent but deny the same protection to a Foreign Nation acting through a Corporate Subsidiary?" There is no good answer to this question. At the same time, the dissenters said, a more literal textual interpretation would unnecessarily complicate the business of corporate-structuring. It would tell those engaged in that work that they must consider a new non-business-related factor - a legal jurisdictional factor - when they decide whether to structure a government business with one, or with two or more, tiers. There is no reason to do so. The dissenters recognize that the statute's language must permit their purpose-based interpretation. They pointed to Justice Holmes comment in a similar case: The purpose-based interpretation does not "ignore the distinction between a corporation and its members." It simply interpret[s] an untechnical word "owner" in the liberal way that Congress intended.

Case Two. The Federal Arbitration Act, written to overcome judicial hostility to arbitration, says that all courts, including state courts, must enforce arbitration clauses written into contracts - indeed, written into any contract that the Commerce Clause gives Congress the power to control. The Act makes an exception for arbitration clauses contained in "contracts of employment of seamen, railroad employees, or any other class of workers engaged in foreign or interstate commerce." A retail store and one of its employees enter into an ordinary employment contract that contains an arbitration clause. The Arbitration Act applies - and state courts must enforce the clause - unless the employment contract falls within the exception. And it does so only if retail store employees fall within the term "any other class of workers."

A Court majority, following a more literal, text-based approach, concluded that the words "any other class of workers" did not include those who work in retail stores. The relevant words, it pointed out, follow an explicit reference to "seamen" and to "railroad workers." A canon of statutory interpretation, ejusdem generis, says that, if "general words follow specific words in a statutory enumeration," courts should construe the "general words" as "embrac[ing] only objects similar in nature to those objects enumerated by the preceding specific words." And retail store workers are not similar in nature to seamen and railway workers.

The majority drew support from other words in the relevant phrase, the words "in commerce." Those words have become a term of art, signaling that Congress intends a limited, rather than a full, exercise of its Commerce Clause power. The majority added that its interpretation, which limits "other class of workers" to, say, transportation workers, is consistent with a limited exercise of Congress's commerce power, but an interpretation broad enough to include workers in retail stores is too broad to count as limited. The majority recognized that the words "in commerce" had not become words of art in 1923 when Congress enacted the statute. But it thought that reading those words more broadly in older statutes would "bring instability" to the interpretive task.

The dissent believed that the words "any other class of workers" referred to all workers, retail store employees included. Taking a more directly purposive approach, it began by asking why Congress included the exception at all. The Act's legislative history - embodied in testimony before the relevant Senate committee - made clear that the seamen's union had opposed the Arbitration Act's enactment because it feared that arbitration in respect to employment would disfavor ordinary workers. Seeking to overcome this opposition, the American Bar Association, the statute's chief proponent (and drafter of its language), testified that it wanted to help bring about arbitration of commercial disputes, not of employment disputes. It said that it had no interest in affecting "labor disputes." The testifying ABA official added, "if your honorable committee should feel that there is any danger of that, it should add to the bill the following language, 'but nothing herein contained shall apply to seamen or any class of workers in interstate or foreign commerce.'" Herbert Hoover, then Secretary of Commerce, seconded the thought in testimony that referred to the "railroad man" as well as to "the seaman" and "sundry other workers."

If this testimony adequately describes the exception's purpose - and nothing in the history suggests any contrary purpose - Congress would have wanted the exception to cover all workers. To the extent that it did not, Congress's objective would be compromised. The words "in commerce" do not show the contrary, for that phrase was not a term of art in 1923 when Congress wrote the Act. Nor does the fact that the exception lists "seamen" and "railroad workers." After all, in 1923 Court decisions had led Congress to believe that its commerce powers were limited - perhaps in this area primarily to workers like those in transportation who helped to move goods from state to state. Congress might have wanted its language to emphasize that it had exempted the groups of workers then most directly affected, particularly since, by doing so, it could assure those opposed to the law, the seamen, that they had nothing to fear. Over time courts have read the Commerce Clause more broadly, and, since the Act's language ties its scope to the Clause, they have thereby extended the reach of the Act itself. Why would Congress not have wanted an expanding exception of similar scope?

Does it matter that the "textualist" view of the statute prevailed? As a result, States cannot disfavor arbitration clauses in most employment contracts. Some would argue that this result is good. Labor arbitration has worked well in areas subject to Labor Board regulation - where labor disputes are typically subject to arbitration for other legal reasons. Others might argue the contrary. But if one sees the interpretive process as an effort to locate, and remain faithful to, the human purposes embodied in a statute, how can one admire this result? The only direct evidence available (I would say the only evidence available) indicates that, at the time of the statute's enactment, members of Congress saw a problem - a problem involving arbitration of commercial contracts. They tried to attack that problem with a statute tailored to the problem's scope - deliberately eliminating labor contracts from the statute's coverage. Yet the Court has responded by extending the statute's scope so that it now regulates an area of life that members of Congress would originally have thought excluded, which exclusion they desired and sought.

Case Three. The federal habeas corpus statute is ambiguous in respect to the time limits that apply when a state prisoner seeks access to federal habeas corpus. It says that a state prisoner (ordinarily) must file a federal petition within one year after his state court conviction becomes final. But the statute tolls that one-year period during the time that "a properly filed application for State post-conviction or other collateral review" is pending. Do the words "other collateral review" include an earlier application for a federal habeas corpus petition? Should the one-year period be tolled, for example, when a state prisoner mistakenly files a habeas petition in federal court before he exhausts all his state collateral remedies?

It is unlikely that anyone in Congress thought about this question, for it is highly technical. Yet it is important. More than half of all federal habeas corpus petitions fall into the relevant category - i.e., state prisoners file them prematurely before the prisoner has tried to take advantage of available state remedies. In those cases, the federal court often dismisses the petition and the state prisoner must return to state court to exhaust available state remedies before he can once again file his federal habeas petition in federal court. If the one-year statute of limitations is not tolled while the first federal habeas petition was pending, that state prisoner will likely find that the one year has run - and his federal petition is time-barred - before he can return to federal court.

A literal reading of the statute suggests that this is just what Congress had in mind. It suggests that the one-year time limit is tolled only during the time that state collateral-review (or similar) proceedings are in process. And that reading is supported by various linguistic canons of construction.

Nonetheless, the language does not foreclose an alternative interpretation - an interpretation under which such petitions would fall within the scope of the phrase "other collateral review." The word "State" could be read to modify the phrase "post-conviction . . . review," permitting "other collateral review" to refer to federal proceedings. The phrase "properly filed" could be interpreted to refer to purely formal filing requirements rather than calling into play more important remedial questions such as the presence or absence of "exhaustion." A purposive approach favors this latter linguistic interpretation.

Why? To answer this question, let us refer back to our hypothetical legislator, the "reasonable Member of Congress." Which interpretation would that Member favor (if he had thought of the problem, which he likely did not)? Consider the consequences of the more literal interpretation. That interpretation would close the doors of federal habeas courts to many or most state prisoners who mistakenly filed a federal habeas petition too soon, but not all. Whether the one-year window was still open would depend in large part on how long the federal court considering the premature federal petition took to dismiss it. In cases where the court ruled quickly, the short time the federal petition was (wrongly) present in the federal court might not matter. But if a premature federal petition languishes on the federal court's docket while the one-year runs, the petitioner would likely lose his one meaningful chance to seek federal habeas relief. By way of contrast, state court delay in considering a prisoner petition in state court would not matter. Where state proceedings are at issue, the statute tolls the one-year limitations period.

Now let us ask why our reasonable legislator would want to bring about these consequences? He might believe that state prisoners have too often abused the federal writ by filing too many petitions. But the distinction that a literal interpretation would make between those allowed to file and those not allowed to file - a distinction that in essence rests upon federal court processing delay - is a random distinction, bearing no logical relation to any abuse-related purpose. Would our reasonable legislator, even if concerned about abuse of the writ, choose to deny access to the Great Writ on a random basis? Given our traditions, including those the Constitution grants through its habeas corpus guarantees, the answer to this question should be "no." Would those using a more literal text-based approach answer this question differently? I do not think so. But my real objection to the text-based approach is that it would prevent them from posing the question at all.

I mean my three examples to suggest the danger that lurks where judges rely too heavily upon just text and textual aids when interpreting a statute. I mean them to indicate why, where difficult statutory questions are at issue, courts do better to focus foremost upon statutory purpose, ruling out neither legislative history nor any other form of help in order to locate the role that Congress intended the statutory words in question to play.

For one thing, near-exclusive reliance upon canons and other linguistic interpretive aids in close cases can undermine the Constitution's democratic objectives. Legislation in a "delegated democracy" is meant to embody the people's will, either directly (insofar as legislators see themselves as translating how their constituents feel about each proposed law) or indirectly (insofar as legislators see themselves as exercising delegated authority to vote in accordance with what they see as the public interest). Either way, an interpretation of a statute that tends to implement the legislator's will helps to implement the public's will and is therefore consistent with the Constitution's democratic purpose. For similar reasons an interpretation that undercuts the statute's objectives tends to undercut that Constitutional objective.

I concede that: Were the courts fully consistent in their use of the canons, were Congressional drafters fully aware of those canons; were Congress to rely consistently upon the work of those drafters; in a word, were the same linguistic conventions known and used similarly by all; then reliance upon those conventions alone could provide interpretations likely to reflect Congressional purposes. But in the world as it is, we shall do better to use whatever tools best identify congressional purpose in the circumstances.

Use of a "reasonable legislator" fiction also facilitates legislative accountability. Ordinary citizens think in terms of general purposes. They readily understand their elected legislators' thinking similarly. It is not impossible to ask an ordinary citizen to determine whether a particular law is consistent with a general purpose the ordinary citizen might support. It is not impossible to ask an ordinary citizen to determine what general purpose a legislator sought to achieve in enacting a particular statute. And it is not impossible for the ordinary citizen to judge the legislator accordingly. But it is impossible to ask an ordinary citizen (or an ordinary legislator) to understand the operation of linguistic canons of interpretation. And it is impossible to ask any ordinary citizen to draw any relevant electoral conclusion from consequences that might flow when courts reach a purpose-thwarting interpretation of the statute based upon their near-exclusive use of interpretive canons. Were a segment of the public unhappy about application of the Arbitration Act to ordinary employment contracts, whom should it blame?

For another thing, that approach means that laws will work better for the people they are presently meant to affect. Law is tied to life; and a failure to understand how a statute is so tied can undermine the very human activity that the law seeks to benefit. The more literal text-based, canon-based interpretation of the Foreign Sovereign Immunities jurisdictional statute, for example, means that foreign nations, those using tiered corporate ownership, will find their access to federal courts cutoff, undermining the statute's basic jurisdictional objectives. The textual approach to the habeas corpus statute randomly closes courthouse doors in a way that runs contrary to our commitment to basic human liberty. And it does so because it tends to stop judges from asking a relevant purpose-based question: Why would Congress have wanted a statute that produces those consequences?

In sum, a "reasonable legislator" approach is a workable method of implementing the constitution's democratic objectives. It readily translates the general desire of the public for certain ends, through the legislator's efforts to embody those ends in legislation, into a set of statutory words that will carry out those general objectives. I have argued that the Framers created the Constitution's complex governmental mechanism in order better to translate public will, determined through collective deliberation, into sound public policy. The courts constitute part of that mechanism. And judicial use of "will of the reasonable legislator" - even if at times it is a fiction - helps statutes match their means to their overall public policy objectives, a match that helps translate the popular will into sound public policy. An overly literal reading of a text can too often stand in the way.

Administrative Law

My final example returns to a question common to all modern democracies, and one that I discussed earlier in the context of federalism. How can we reconcile democratic control of government with the technical nature of modern life? The former calls for decision-making by citizens or their elected representative; the latter for decision-making by administrators or experts. If we delegate too much decision-making authority to experts, administration and democracy conflict. We lose control. Yet if we delegate too little authority, we also find democracy weakened. To achieve our democratically-chosen ends in a modern, populous society requires some amount of administration, involving administrative, not democratic, decision-making. To achieve those same ends in a technologically advanced society requires expertise. The average citizen normally lacks the time, knowledge, and experience necessary to understand certain technical matters related, for example, to the environment, energy, communications, or modern weaponry. Without delegation to experts an inexpert public, possessing the will, would lack the way. The public understands this fact. Who would want to vote about how an Army battalion should take the next hill? The Framers foresaw this possibility. They sought to create a workable democracy - a democratic process capable of acting for the public good.

To reconcile democratically chosen ends with administrative expertise requires striking a balance - some delegation, but not too much. The right balance avoids conflict between democracy and administration. The latter then complements the former by implementing legislatively determined general policy objectives. How to strike that balance? That is the mystery. The Constitution, not surprisingly, leaves the matter primarily in the hands of the legislature. Its legislative handiwork, the statute, is subject to court interpretation on this point as on others. And that is where administrative law plays a role. As classically conceived, administrative law helps to implement the legislature's choice of when and how to delegate decision-making to administrators and experts.

My last example focuses upon a related principle of administrative law, a principle of judicial "deference" to agency interpretation of statutes. The principle applies when judges and administrative agencies seek to interpret the same statutory provision. Should a judge give weight to the agency's interpretation of its governing statute, perhaps deferring to the agency, substituting its view of the statute for the judge's own? If so, when? And why? In Chevron v. EPA the Court held that a judge should defer to a reasonable agency interpretation of an ambiguous statute. It said that the "power of an administrative agency to administer a congressionally created . . . program necessarily requires the formulation of policy and the making of rules to fill any gap left implicitly or explicitly by Congress." And "a court may not substitute its own construction of a statutory provision for a reasonable interpretation made by the administrator of an agency." As one of my colleagues has written Chevron seems to require deference to "an authoritative agency position," unless "(1) the statute is unambiguous so there is no room for administrative interpretation; (2) no interpretation has been made by personnel of the agency responsible for administering the statute; or (3) the [reasonable] interpretation made by such personnel was not authoritative, in the sense that it does not represent the official position of the expert agency."

Suppose, for example, that a labor relations statute requires a federal agency to "meet" with its employees' collective bargaining representative and "negotiate in good faith for the purposes of arriving at a collective bargaining agreement." These words do not say when that good-faith negotiation must take place. They do not say whether the agency must re-negotiate if an important new matter comes up, say, in the third year of a five year contract. The statutory language is ambiguous. And the answers yes, no, or it depends (say, on what the parties agreed to about re-opening in mid-stream) all seem reasonable. Hence, given Chevron, a judge should defer to the agency's own statutory answer. The agency, not the court, will thereby determine the meaning of the statute.

For present purposes, I note that judges do not agree about how absolute Chevron's approach is meant to be. Is it a judicially-created absolute rule? Is it just a rule of thumb? How shall we interpret that rule of interpretation? To refer back to the democratic origins and purposes of delegation itself will help answer this question.

What lies behind Chevron? What is its rationale? The answers in part seem practical. No one can foresee all possible applications of a statute. Legislation inevitably contains ambiguities and gaps. The agency that administers the statute is likely better able than a court to determine how best to fill those gaps. The agency, experienced in administering the statute, will likely better understand the practical implications of competing alternative interpretations, consistency with Congressional objectives, administrative difficulties, the consequences for the public.

But the answer is not entirely practical. Principles of active liberty also matter. Looked at from a democratic perspective, Chevron helps a judge answer an important question about Congress: How, in this statute, did Congress strike the democratic/administrative balance? Did Congress intend the courts to defer to the agency's own interpretive views or did Congress intend not to delegate the interpretive decision to the agency (thereby asking the courts to treat the statute like any other)? It is quite possible that no Member of Congress actually thought about the matter. But a judge still can ask how a reasonable Member of Congress would have answered it had the question come to mind. The judge can ask whether, given the statutory aims and circumstances, a hypothetical Member would likely have wanted judicial deference in this situation.

Does framing the question as referring to a Member of Congress matter? Will it not normally lead to the same answer that practical considerations alone would suggest? After all, a reasonable Member of Congress would want the statute to work well; that Member would know that interpretation plays an important role in assuring that result; and that Member would realize that in most instances judges possess less relevant expertise than does the administering agency. Hence that Member would likely conclude, as did Chevron, that ordinarily judges should listen carefully to the agency's views about proper interpretation. If the statute is ambiguous, courts should defer to a reasonable agency interpretation of a statute. If the deference-related result is the same, why bother imagining a hypothetical Member of Congress?

Active liberty, however, suggests that it does matter. Indeed, it suggests that Chevron's rule is not absolute, but simply a rule of thumb. Deference to a reasonable agency interpretation of an ambiguous statutory provision often makes sense, but not always. Suppose, for example, a statutory ambiguity arises in the context of a question of national importance. Does the National Labor Relations Act term "employee" cover corporate supervisors thereby protecting a company's supervisors' decision to organize a supervisors' union? Does the statute forbidding discrimination in employment based upon "age" forbid discrimination against younger workers, as well as against older workers? Would our hypothetical reasonable Member of Congress have wanted a regulatory agency to decide such questions of major importance? Where a statutory term, though ambiguous, concerns a matter that Congress is likely to have wanted to decide for itself, our construct - the "reasonable Member" - leads us to conclude that courts should not defer. And that is how our Court has treated most such questions.

To take another example, suppose that the agency has expressed its views in an informal manner - a manual of internal procedures, a press release, or a letter of written by low level officials. Should that agency viewpoint nonetheless carry weight in the interpretive enterprise? The "reasonable Member of Congress" approach will lead the courts to ask whether it makes sense to defer to a particular kind of agency interpretation given the particular kind of statute and the particular circumstances at issue. The answer will be, sometimes yes, sometimes no - more or less what our cases have held.

Treating Chevron in this way - not as an absolute rule, but as a rule of thumb - may seem to complicate life for agencies, lawyers, and judges. But particular circumstances can generate clear (if narrow and specific) legal answers - and answers that may make more sense than those that would flow from a more absolute, overarching interpretive rule. As importantly, those answers make democratic sense. In all likelihood a hypothetical reasonable Member of Congress would have decided the delegation/deference question so as to help the statute work better to achieve its ends. And those ends usually reflect the general desires of the public. Use of the fiction thereby helps the statute work better, in both the functional and the democratic sense of the term. And, in doing so, it makes it easier for the public to hold accountable for the results both Congress and those whom Congress has charged with the task of implementing the statute.

The practical need and the theoretical democratic reasons for using the "reasonable legislator" fiction merge. The fiction helps to make the law reflect the public's desire for a law that implements its general instructions. Active liberty provides a democratic rationale for better functioning administrative law.


My six examples focused on different areas of the law - free speech, federalism, privacy, equal protection, statutory interpretation, and judicial review of administrative action. The discussion of each involved contemporary problems of modern government - campaign finance, environmental regulation, technology-based privacy risks, affirmative action, the legislative and administrative processes. I have tried to show how, in varying contexts, reference to the Constitution's basic democratic objectives can help courts shape constitutional doctrine, reconcile competing constitutional objectives, time judicial intervention, interpret statutory ambiguities, and create room for agency interpretations. The discussion has suggested how the courts, by paying somewhat greater attention to those democratic objectives, can bring about better law.

By now it should be clear that when I argue for greater attention, I am not arguing for a new theory of constitutional law. In my experience most judges approach and decide most cases, including constitutional cases, quite similarly. They are professionals. And their professional training and experience leads them to examine language, history, tradition, precedent, purpose, and consequences. Given roughly similar forms of legal education and professional experience, it is not surprising that judges often agree about how these factors, taken together, point to the proper result in a particular case. Even where they differ, the degree of difference is often small. Our Court, which normally steps in where other judges disagree, decides close to 40 percent of its cases unanimously. Most of the rest involve only one or two dissenting votes. In only about 20 percent of our case load do we divide five-four. And the same Justices are not always on the same side of the split. Only a handful of constitutional and statutory issues are as open in respect to language, history, and basic purpose, as those I have here described.

I have taken this professional framework as a given. Within that framework, I have argued for greater awareness of, and emphasis upon, the Constitution's democratic imperative. My argument has not rested upon logical or scientifically convincing empirical demonstration. Rather it has used examples to suggest a pattern. And that pattern suggests that supplementing ordinary professional judicial approaches with increased emphasis on the Constitution's democratic objectives will help Americans better resolve contemporary problems through law.

A Serious Objection

I now want to broaden my argument's appeal - and to tie the argument to broader questions of interpretation. Throughout, I have urged attention to purpose and consequences. My discussion sees individual constitutional provisions as embodying certain basic purposes, often expressed in highly general terms. It sees the Constitution itself as a single document designed to further certain basic general purposes as a whole. It argues that an understanding of, and a focus upon, those general purposes will help a judge better to understand and to apply specific provisions. And it identifies consequences as an important yardstick to measure a given interpretation's faithfulness to these democratic purposes. In short, a focus on purpose seeks to promote active liberty by insisting on interpretations, statutory as well as constitutional, that are consistent with the people's will. A focus on consequences, in turn, allows us to gauge whether and to what extent we have succeeded in facilitating workable outcomes which reflect that will.

Some lawyers, judges, and scholars, however, would caution strongly against the reliance upon purposes (particularly abstractly-stated purposes) and assessment of consequences. They ask judges to focus primarily upon text, upon the Framers' original expectations narrowly conceived and upon historical tradition. They do not deny the occasional relevance of consequences or purposes (including such general purposes as democracy), but they believe that judges should use them sparingly in the interpretive endeavor. They ask judges who tend to find interpretive answers in those decision-making elements to rethink the problem to see whether language, history, tradition, and precedent, by themselves will not yield an answer. They fear that, once judges become used to justifying legal conclusions through appeal to real-world consequences, they will too often act subjectively and undemocratically, substituting an elite's views of good policy for sound law. They hope that language, history, tradition, and precedent will provide important safeguards against a judge confusing his or her personal, undemocratic, notion of what is good for that which the Constitution or statute demands. They tend also to emphasize the need for judicial opinions that set forth their legal conclusions in terms of rules that will guide other institutions, including lower courts.

This view, which I shall call "textualist" (in respect to statutes) or "originalist" (in respect to the Constitution) or "literalist" (shorthand for both), while logically consistent with emphasizing the Constitution's democratic objectives, is not hospitable to the kinds of arguments I have advanced. Nor is it easily reconciled with my illustrations. Why, then, does it not undercut my entire argument?

The answer, in my view, lies in the unsatisfactory nature of that interpretive approach. First, the more "originalist" judges cannot appeal to the Framers themselves in support of their interpretive views. The Framers did not say specifically what factors judges should take into account when they interpret statutes or the Constitution. This is obvious in the case of statutes. Why would the Framers have preferred 1) a system of interpretation that relies heavily on linguistic canons to 2) a system that seeks more directly to find the intent of the legislators who enacted the statute? It is close-to-obvious in respect to the Constitution. Why would Framers, who disagreed even about the necessity of including a Bill of Rights in the Constitution, who disagreed about the content of that Bill of Rights, nonetheless have agreed about what school of interpretive thought should prove dominant in interpreting that Bill of Rights in the centuries to come?

In respect to content, the Constitution itself says that the "enumeration" in the Constitution of some rights "shall not be construed to deny or disparage others retained by the people." Professor Bailyn concludes that the Framers added this language to make clear that "rights, like law itself, should never be fixed, frozen, that new dangers and needs will emerge, and that to respond to these dangers and needs, rights must be newly specified to protect the individual's integrity and inherent dignity." Given the open-ended nature of content, why should one expect to find fixed views about the nature of interpretive practices?

If, however, justification for the literalist's interpretive practices cannot be found in the Framers' intentions, where can it be found - other than in an appeal to consequences, i.e. in an appeal to the presumed beneficial consequences for the law or for the Nation that will flow from adopting those practices? And that is just what we find argued. That is to say, literalist arguments often try to show that that approach will have favorable results, for example, that it will deter judges from substituting their own views about what is good for the public for those of Congress or for those embodied in the Constitution. They argue, in other words, that a more literal approach to interpretation will better control judicial subjectivity. Thus, while literalists eschew consideration of consequences case by case, their interpretive rationale is consequentialist in this important sense.

Second, I would ask whether it is true that judges who reject literalism necessarily open the door to subjectivity. They do not endorse subjectivity. And under their approach important safeguards of objectivity remain. For one thing, a judge who emphasizes consequences, no less than any other, is aware of the legal precedents, rules, standards, practices, and institutional understanding that a decision will affect. He or she also takes account of the way in which this system of legally related rules, institutions, and practices affects the world.

To be sure, a court focused on consequences may decide a case in a way that radically changes the law. But this is not always a bad thing. For example, after the late-nineteenth-century Court decided Plessy v. Ferguson, the case which permitted racial segregation that was, in principle, "separate but equal," it became apparent that segregation did not mean equality but meant disrespect for members of a minority race and led to a segregated society that was totally unequal, a consequence directly contrary to the purpose and demands of the Fourteenth Amendment. The Court, in Brown v. Board of Education, overruled Plessy, and the law changed in a way that profoundly affected the lives of many.

In any event, to focus upon consequences does not automatically invite frequent dramatic legal change. Judges, including those who look to consequences, understand the human need to plan in reliance upon law, the need for predictability, the need for stability. And they understand that too radical, too frequent, legal change has, as a consequence, a tendency to undercut those important law-related human needs. Similarly, each judge's individual need to be consistent over time constrains subjectivity. As Justice O'Connor has explained, a constitutional judge's initial decisions leave "footprints" that the judge, in later decisions, will almost inevitably follow.

For another thing, to consider consequences is not to consider simply whether the consequences of a proposed decision are good or bad, in a particular judge's opinion. Rather, to emphasize consequences is to emphasize consequences related to the particular textual provision at issue. The judge must examine the consequences through the lens of the relevant constitutional value or purpose. The relevant values limit interpretive possibilities. If they are democratic values, they may well counsel modesty or restraint as well. And I believe that when a judge candidly acknowledges that in addition to text, history, and precedent, consequences also guide his decision-making, he is more likely to be disciplined in emphasizing, for example, constitutionally relevant consequences rather than allowing his own subjectively held values to be outcome determinative. In all these ways, I think a focus on consequences will itself constrain subjectivity.

Let me provide an example of how some of these principles apply. The First Amendment says that "Congress shall make no law respecting an establishment of religion." I recently wrote (in dissent) that this Clause prohibits government from providing vouchers to parents to help pay for the education of their children in parochial schools. The basic reason, in my view, is that the Clause seeks to avoid the "social conflict potentially created when government becomes involved in religious education." Nineteenth and Twentieth Century immigration had produced a Nation with fifty or more different religions. And that fact made the risk of "social conflict" far more serious after the Civil War and in Twentieth Century America than the Framers, with their Eighteenth Century experience might have anticipated. The Twentieth Century Supreme Court had held in applicable precedent that, given the changing nature of our society, in order to implement the basic value that the Framers wrote the Clause to protect, it was necessary to interpret the Clause more broadly than the Framers' might have thought likely.

My opinion then turned to consequences. It says that voucher programs, if widely adopted, could provide billions of dollars to religious schools. At first blush, that may seem a fine idea. But will different religious groups become concerned about which groups are getting the money and how? What are the criteria? How are programs being implemented? Is a particular program biased against particular sects, say, because it forbids certain kinds of teaching? Are rival sects failing to live up to the relevant criteria, say, by teaching "civil disobedience" to "unjust laws"? How will claims for money, say, of one religious group against another, be adjudicated? In a society as religiously diverse as ours, I saw in the administration of huge grant programs for religious education, the potential for religious strife. And that it seemed to me was the kind of problem the First Amendment's religion clauses seek to avoid.

I am not arguing here that I was right. I am arguing that the opinion sought to identify a critical value underlying the Religion Clauses; it considered how that value applied in modern-day America; it looked for consequences relevant to that value; and it sought to evaluate likely consequences in terms of that value. That to me is what is meant by an interpretive approach that emphasizes consequences. Under that approach language, history, precedent, and factual consequence all constrain judicial subjectivity.

Third, "subjectivity" is a two-edged criticism, which the literalist himself cannot escape. The literalist's tools - language and structure, history, and tradition - often fail to provide objective guidance in those truly difficult cases about which I have spoken. Will canons of interpretation provide objective answers? One canon tells the court to choose an interpretation that gives every statutory word a meaning. Another permits the court to ignore a word, treating it as surplus, if otherwise the construction is repugnant to the statute's purpose. Shall the court read the statute narrowly as in keeping with the common law or broadly as remedial in purpose? Canons to the left of them, canons to the right of them, which canons shall the judges choose to follow?

Recall the case about the Foreign Sovereign Immunities Act, in which the question was whether the statutory word "control" has a broad or a narrow meaning. Should the Court, choosing a narrow meaning, have emphasized Congress's use of a different phrase "direct or indirect control" in other statutes where it intended a broad meaning? Or should it instead have emphasized Congress's use of the same phrase - the single word "control" - in a yet different set of statutes where Congress also intended a broad-meaning? Or should it have counted how many times Congress did each? Why place greater weight on the canon suggesting a need to give every statutory word a separate meaning than upon the statute's overall purpose?

Recall Chevron. Reference to the hypothetical "reasonable Member of Congress" introduces no more subjectivity that the Court's own efforts to construct workable interpretive rules without any such reference. On both views Chevron requires Courts to decide when an agency interpretation is reasonable - thereby opening the door to subjectivity.

Consider a recent Ex Post Facto Clause case from our court. We had to decide whether the Clause barred California from enacting a new statute of limitations, applying it retroactively, and then prosecuting an individual for child molesting twenty years after the old limitations period had expired. The Court used one of the literalist's tools, history, to find the answer. Two hundred years ago, Justice Chase, borrowing Blackstone's three century old language, had said the Ex post facto Clause was applicable when, for example, a law made a crime "greater than" it previously was. Did California's retroactive statute of limitations make the crime "greater than" it previously was?

The answer, historically speaking, depended upon the nature of certain events in England that inspired Blackstone's formulation. All Members of the Court carefully examined two abusive Parliamentary prosecutions of 350 years ago -- those of the Earl of Clarendon and the Bishop of Atterbury. Everyone agreed that the Ex Post Facto Clause barred prosecutions of this kind. But there the agreement ended. The dissent thought that these historic parliamentary prosecutions had nothing to do with the case at hand. The majority thought that they offered it considerable support. Who was right? The truthful answer - and one I believe a reading of the opinions will support - is that no one but an expert historian could possibly know. And even the experts might disagree. Judges are not expert historians. How does reliance upon history bring upon certainty or objectivity in such a case?

Why do I remind you of the uncertainties, in close cases, of linguistic structure, of canons of interpretation, and of history? Because those difficulties mean that the "textualist," "traditionalist," and "originalist" approaches themselves possess inherently subjective elements. Which linguistic characteristics are determinative? Which canons shall we choose? Which historical account shall we use? Which tradition shall we apply? And how does that history, or that tradition, apply now?

Significantly, an effort to answer these questions can produce a decision that is not only subjective, but which is also unclear, i.e., one that lacks transparency about the factors that the judge considers truly significant. A decision that directly addresses consequences, purposes and values, is no more subjective and has the added value of exposing underlying judicial motivations, specifying the points of doubt for all to read. This is particularly important because transparency of rationale permits informed public criticism of opinions; and that criticism, in a democracy, plays an important role in checking abuse of judicial power.

Fourth, I do not believe that textualist or originalist methods of interpretation are more likely to produce clear, workable legal rules. But even were they to do so, the advantages of legal rules can be overstated. Rules must be interpreted and applied. Every student whose class grade is borderline knows that the benefits that rules produce for cases that fall within the heartland are often lost in cases that arise at the boundaries.

Nor is clarity exclusively promoted through use of rules. Metaphors and examples also can illuminate. Section 3 of the Clayton Act, for example, prohibits a seller from making a sale on the condition that the buyer will "not use or deal in the goods . . . of [the seller's] competitor." This language most obviously applies to a seller who actually exacts a promise from his buyers. But it also captures cases in which the seller conditions the sale of one product on the buyer also purchasing a different product as well. The courts have introduced the metaphor of "tying" to describe this practice. In doing so, they have made the scope of the statute's proscription more vivid and concrete to the benefit of lawyers, businesses, and courts alike. The common law, too, has long set forth law by way of example - embodied in the facts of the individual case. That precedent has guided ordinary buyers and sellers, promoting prosperity through commercial certainty, for close to three centuries.

In any event, insistence upon clear rules can exact a high constitutional price. California enacted a high-penalty mandatory sentencing law, "three strikes and you're out." California, applying this law, imposed a fifty-year sentence on an individual whose third crime was stealing $153 worth of video tapes; it imposed a twenty-five year sentence on another individual whose third crime was theft of a golf club. Our Court had to decide whether these sentences violated the Constitution's Cruel and Unusual Punishment Clause, a question that turned on whether the sentences were "grossly disproportionate." The majority thought not, with some Justices expressing concern about the workability of any contrary holding.

I was in the minority. I conceded that striking down the law would leave the Court without a clear rule that would readily distinguish those sentences that are "grossly disproportionate" from those that are not. Courts might have to exercise judgment in each case. But that judgment would not lack guidance. A Supreme Court opinion, based on examples, discussing how, and why, the standard applied to the cases before us would help to provide that guidance. Regardless, the administrative gains from rule-based clarity were not worth the constitutional price. Insisting on a clear rule has made the Clause virtually inapplicable to any specific length prison sentence.

Fifth, textualist and originalist doctrines may themselves produce seriously harmful consequences - outweighing whatever risks of subjectivity or uncertainty inherent in other approaches. I have deliberately chosen examples to illustrate that harm. In respect to statutory interpretation, a canon-based approach meant more complex jurisdictional law that closed the federal courthouse doors to certain foreign state-owned enterprises, thereby increasing foreign-relations' friction, just what Congress enacted the statute (the Federal Sovereign Immunities Act) to avoid. Emphasizing a statute's literal language meant a habeas corpus law that randomly closes the doors of federal courts to a set of state prisoners. In respect to administrative law, a canon-like interpretation of Chevron risks judicial deference to an agency interpretation of a statute in circumstances where no reasonable legislator could have so intended, thereby substituting an agency view of the statute for an interpretation likely more consistent with the views of the public's elected representatives.

In respect to the First Amendment, a more "textualist" (if not more "originalist") approach would treat all speech alike - "speech is speech and that is the end of the matter." The prescription drug example showed that application of such a view could unreasonably impede reasonable health-based regulatory objectives. In respect to federalism, a more "originalist" approach could impede efforts to draw citizens into local administration of federal programs - thereby inhibiting the development of a cooperative federalism that can mean more effective regulatory programs. In Eleventh Amendment cases that approach (if, indeed, the Court's approach can be called "originalist") would prevent Congress from creating uniform national remedies in such areas as protection of intellectual property - an area where modern global commercial circumstances make uniform enforcement methods important, if not a necessity.

In respect to privacy, a more literal, less consequence-oriented, approach would not necessarily value the reasons for judicial hesitation; thereby taking inadequate account of ongoing legislative processes and consequently leading to premature judicial interference with legislative development. In respect to equal protection, a more literal approach could have read the Equal Protection Clause divisively, impeding rather than furthering the democratic unity needed to make the Constitution's institutions work as intended.

Of course, my examples are no more than that, examples. They do not prove the general superiority of the interpretations they illustrate. But if you agree that examination of consequences can help us determine whether our interpretations promote specific democratic purposes and general constitutional objectives, I will have made my point. That point is that a more literal approach has serious drawbacks. Whatever "subjectivity-limiting" benefits a more literal, "textual," or "originalist" approach may bring, and I believe those benefits are small, it will also bring with it serious accompanying consequential harm.

Much of the harm at stake is a constitutional harm. Literalism has a tendency to undermine the Constitution's efforts to create a framework for democratic government - a government that, while protecting basic individual liberties, permits citizens to govern themselves, and to govern themselves effectively. Insofar as a more literal interpretive approach undermines this basic objective, it is inconsistent with the most fundamental original intention of the Framers themselves.

For any or all of these reasons, I hope that the audience of those strongly committed to "textualist" or "literalist" views - those whom I am almost bound not to convince - is fairly small. I hope to have convinced some of the rest that active liberty has an important role to play in constitutional (and statutory) interpretation.

That role, I repeat, does not involve radical change in current professional interpretive methods nor does it involve ignoring the protection the Constitution grants fundamental (negative) liberties. It takes Thomas Jefferson's statement as a statement of goals that the Constitution now seeks to fulfill: "All men are created equal." They are endowed by their Creator with certain "inalienable rights." "To secure these rights governments are instituted among men deriving their just powers from the consent of the governed." It underscores, emphasizes, or re-emphasizes the final democratic part of the famous phrase. That re-emphasis, I believe, has practical value when judges seek to assure fidelity, in our modern society, to these ancient and unchanging ideals.


To re-emphasize the constitutional importance of democratic self-government carries with it an additional practical benefit. We are all aware of figures that show that public knows ever less about, and is ever less interested in, the processes of government. Foundation reports criticize the lack of high school civics education. A credible national survey reveals that more students know the names of the three stooges than the three branches of government. Law school graduates are ever less inclined to work for government - with the percentage of those entering government work (or non-government public interest work) declining at one major law school from 12% to 3% over the course of a generation. Polling figures suggest that during that same time, the percentage of the public trusting the government has declined at a similar rate.

This trend, however, is not irreversible. Indeed, trust in government, and interest in public service, has shown a remarkable rebound in response to the terrorist attacks of September 11, 2001. Courts alone cannot maintain this trend; nor obviously is it their job to do so. But courts, as highly trusted government institutions, can help. They can explain in terms the public can understand just what the Constitution is about. They can make clear, above all, that the Constitution is not a document designed to solve the problems of a community at any level - local, state, or national. Rather it is a document that trusts people to solve those problems for themselves. And it creates a framework for a government that will help them do so. That framework foresees democratically-determined solutions, protective of the individual's basic liberties. It assures each individual that the law will treat him or her with equal respect. It seeks a form of democratic government that will prove workable over time.

That is the democratic ideal. It is as relevant today as it was 200 or 2000 years ago. Two thousand years ago, Thucydides quoted Pericles as telling his fellow Athenians: "We do not say that the man who fails to participate in politics is a man who minds his own business. We say that he is a man who has no business here." Related ideals, the sharing of political authority, a free people delegating that authority to a democratically elected government, participation by those people in that democratic process, moved the Framers. And they wrote a Constitution that embodied these ideals. We judges cannot insist that Americans participate in that government. But we can make clear that our Constitution depends upon it. Their participation is necessary. It is a critical part of that "positive passion for the public good" that John Adams, like so many others, believed was a necessary condition for any "real liberty" and for the "Republican Government" that the Constitution creates.