Almost every recorded political statement John Roberts has made throughout his life, from adolescence to his nomination as chief justice, suggests that he has strong conservative political convictions and instincts, and many people naturally fear that he will use his great power on the Supreme Court in the service of his politics. He promised that he would not, but the Senate Judiciary Committee should have been more effective than it was in testing that promise. In fact it failed dramatically in its responsibility to do so.

In his public career Roberts has opposed improving protection for the voting rights of minorities; held that it would be constitutional for Congress to strip the federal courts of their powers to supervise racial integration; denigrated efforts by a group of women legislators to reduce gender inequality in the workplace; referred to the right of privacy as “so-called”; signed a brief advising the Supreme Court to overrule Roe v. Wade; and described a Supreme Court decision outlawing a moment of silence that might be used for prayer in schools as “indefensible.” (See William L. Taylor’s recent article on Roberts’s career in these pages.1 )

In the Senate hearings now concluded, Roberts repeatedly said that the more outrageous of these opinions were only those of a lawyer doing his job, in the Justice Department and the White House, for a very conservative client: the Reagan administration. But of course he did not have to join that administration and the tone of many of his comments suggests not just an obedient staff member but an enthusiastic proponent. The Bush administration has refused to allow senators to see Roberts’s more recent memoranda, written when, as deputy solicitor general in the first Bush administration, he had more responsibility for making policy. We can only speculate about what political opinions these would reveal.

Nor is there much in Roberts’s record as disclosed to the Senate that suggests any personal impulse to moderate the right-wing policies he has defended or any concern for those at the bottom of our society whom such policies would injure. He did advise his superiors that stripping the courts of their jurisdiction over the use of busing as a remedy for racial inequality, though in his opinion constitutional, would be “bad policy.” But he said it would be bad for technical reasons: because, as he told Senator Herbert Kohl, it might “lead to a situation where there’s arguable inconsistency and disuniformity in federal law.” He did advise the administration to sign a treaty condemning genocide, but only because a failure to sign it would be bad for the nation’s image. As a lawyer in private practice he acted pro bono in advising a gay rights group preparing for a Supreme Court challenge to Colorado’s discriminatory constitutional amendment. But he did so, he told Senator Richard Durbin, because the partners of his firm asked him to, and he seemed happy to add, in response to Durbin’s further question, that he would “of course” have been willing to advise Colorado how to resist the challenge if its officials had walked through his door first.

We find no greater reassurance that Roberts will not use his power on the Court to advance a right-wing political agenda when we study his decisions during his brief term as a judge on the District of Columbia Circuit Court. In one of the two most politically sensitive of his cases he declared that the federal government has no power under the Constitution’s interstate commerce clause to force a California developer to protect an endangered species of toad that has so far been found only in that state. In the other case he joined an opinion declaring that the courts must show great deference to the President’s opinion that international treaties, including the Geneva Conventions, do not protect the Guantánamo prisoners. Whether or not these positions are correct in law, they offer no reason to think that Roberts would often hold, in difficult cases, that the law is contrary to what a conservative would wish it to be.

Roberts plainly saw the Senate confirmation hearings as his best opportunity to dispel any suspicion that he would be a political judge, so we must look with some care at the record of those hearings. It is important to be clear what we are looking for. Roberts declared often enough throughout the hearings—indeed, he took every possible opportunity to declare—that he would be guided in all his decisions by “the law” and not by his personal politics or his faith or by anything but the law. In his introductory remarks he said that he would decide every case “according to the rule of law” and that his job was that of an umpire calling balls and strikes but not pitching or batting. He repeated his vow to decide “according to the rule of law” throughout the hearings. We must ask whether, for Roberts, these repeated declarations can have any substance. We can only answer that question by discovering how he decides what the law is when the pertinent constitutional or statutory provisions are vague or abstract. Does his method of legal reasoning in such cases provide a filter, or protective screen, between his politics and his judgment about what the law requires?


As Senator Joseph Biden reminded Roberts, an umpire cannot set the strike zone: that is determined by the explicit rules of baseball. But a Supreme Court justice faces no such constraint. The great constitutional clauses and much crucial legislation are drafted in highly abstract terms that demand interpretation. Much of this is abstract moral language: the Constitution demands “due” process, forbids punishments that are “cruel” as well as “unusual,” commands “equal” protection of the laws, and outlaws “unreasonable” searches and seizures. Important Supreme Court precedents rule out “undue” burdens on women seeking abortions, for example, and make the constitutionality of laws depend on their “rationality.” Statutes often make the liability of a person or corporation turn on whether they have behaved in ways that are “unreasonable” or even “unconscionable.” The bare statement that a judge should enforce “the law” when dealing with clauses that are so abstract tells us nothing: the crucial question is how the judge should decide what the law is.

Roberts himself demonstrated the difficulty and importance of that question in a crucially important statement he had carefully prepared. He pointedly disagreed with the constitutional philosophy often called “originalism,” which is the thesis that judges ought to interpret the abstract moral clauses of the Constitution by asking how the framers of each clause would have expected that clause to be applied—that judges should decide what punishments are “cruel,” for example, by referring to what punishments those who wrote the Eighth Amendment in the eighteenth century thought cruel. On that test, capital punishment, which was of course practiced very widely then, would not be cruel. Some very conservative jurists, including Justices Antonin Scalia and Clarence Thomas and former judge Robert Bork, have explicitly adopted that view of constitutional interpretation.

When Scalia tried to defend this view in a discussion of his judicial methods at Princeton some years ago, the objection was made that originalism, so understood, ignores a crucial distinction between what the framers intended to say and what they expected would be the effect of their saying what they intended to say.2 The framers might have set out their own particular views about what counts as cruelty in punishment, what counts as a denial of equal protection in legislation, and so forth in the constitutional clauses they wrote. But they did not. Instead they chose to lay down general moral principles. So true fidelity to their intentions requires judges to ignore the framers’ concrete opinions and do their best to apply these principles as moral principles: to decide, for themselves, that is, what punishments are in fact cruel and what treatment is in fact equal.

Roberts explicitly accepted this objection to Scalia’s version of originalism. In reply to a question from Senator Charles Grassley, he said, about the equal protection clause that was adopted after the Civil War,

There are some who may think they’re being originalists who will tell you, Well, the problem they [the framers] were getting at were the rights of the newly freed slaves. And so that’s all that the equal protection clause applies to. But, in fact, they didn’t write the equal protection clause in such narrow terms. They wrote more generally…. We should take them at their word, so that is perfectly appropriate to apply the equal protection clause to issues of gender and other types of discrimination beyond the racial discrimination that was obviously the driving force behind it.

Liberal commentators were pleased to learn that Roberts rejects a method of interpretation that they associate with Scalia and other right-wing justices. But he said nothing about how he would decide the great moral questions about due process and equal treatment that, on his understanding, the Constitution requires justices to decide. Rejecting Scalia’s “originalism” increases the responsibility of contemporary judges, but it does not tell them how to exercise that responsibility. Suppose judges must decide, for instance, whether it is consistent with the Constitution’s command of “equal protection” of the laws that a state university grant some limited preference to minority applicants, or whether a state government respects “due process” when it outlaws early abortion. If they may not rely on historical evidence about whether those who wrote those phrases would have expected them to condemn affirmative action or permit making abortion a crime—if judges must ask themselves directly what answer follows from the abstract moral language the framers laid down—then how can they help putting into effect their own convictions, conservative or liberal, about whether affirmative action is fair or whether abortion is immoral?


The cynical view of many lawyers is that judges simply follow their own political preferences and disguise their doing so by announcing that they are following “the law.” These lawyers think that constitutional adjudication is only politics by a grander name, that if a president appoints someone with the policy preferences of John Roberts to the Supreme Court, the nation will then be governed by a very conservative interpretation of abstract clauses of the Constitution. The Supreme Court’s outrageous decision in Bush v. Gore reinforced that cynical view because in order to ensure that Bush became president, the justices who preferred him ignored the doctrines of constitutional interpretations they had previously embraced.3 But most judges insist that there is a difference between what the law requires them to decide, even when they are applying the very abstract moral language of the Constitution, and how they would vote if they were legislators free to vote in accordance with their own policy preferences. How is that possible?

A judge can draw that distinction in such cases only if he has, in addition to his partisan commitments and policy preferences, political convictions of a different and independent kind: convictions about the proper role of a judicially enforceable constitution in a democracy. A constitution shapes democracy by assigning powers to different institutions—by specifying the composition and responsibilities of the legislative, executive, and judicial branches—and it regulates democracy by creating individual rights that act as constraints on what those different branches of government may do. Our constitution, for example, limits the powers of Congress to matters of national concern, leaving purely local issues to state government, and it also limits the power of all branches of government by specifying that no branch may use censorship as a tool of government. But as I have said, it imposes these structures and constraints in very abstract language: it declares that Congress has power only over “interstate commerce,” and that government must not invade “the freedom of speech.” Judges can interpret that abstract language only by appealing to a vision of a desirable, workable form of democracy that they believe both fits and justifies the overall structure of the Constitution. They can then justify choosing one reading of the abstract clauses rather than another by explaining how that reading makes a better contribution to democracy so conceived.

They can interpret their own proper role in a democracy in the same way: by asking what view of the powers of unelected judges to check what other branches of government have done follows from the view of democracy they have identified as best fitting and justifying our Constitution and our practices. That judgment must include a view of how far it serves the appropriate understanding of democracy to require judges to be governed by their own past decisions as precedents and when and why they may depart from precedent in search of what they take to be a more effective democracy. Judges must ask, for example, whether it better serves what they take to be the right understanding of democracy for Supreme Court justices who think Roe v. Wade was wrongly decided to correct what they believe to have been a serious mistake or to respect the nation’s reliance on that decision for three decades.

We must try to discover from his statements at the Judiciary Committee hearings whether Roberts has such a constitutional philosophy—a vision of democracy that can filter out his partisan commitments and policy preferences when he decides what the Constitution requires. It might be helpful first to briefly notice examples of constitutional philosophies that other judges have embraced and how these philosophies can act as filters in this way. Scalia’s announced form of originalism, if in fact he held to it with any important degree of consistency,4 would constitute such a philosophy. He might suppose that a constitution serves democracy best when it limits the power of democratic majorities only in the ways that were widely accepted when the nation began. A judge who held this conception of good democracy would sometimes be required to refuse constitutional challenges that his own policy preferences would encourage him to sustain: it would require him to uphold capital punishment against constitutional challenge, for example, even if he himself detested state-mandated killings. Roberts was right to reject this unattractive vision of democracy, particularly since it has no basis in our history. As he said, the framers themselves rejected it by using the abstract moral language they chose.

Justice Stephen Breyer, in his recent book setting out his own constitutional philosophy, offers a more attractive example. He argues that the liberty protected by an appropriate conception of democracy embraces not only a citizen’s freedom from undue government interference but a more active freedom to participate in self-government as an equal; and he undertakes to show that an understanding of the Constitution as aiming to promote that form of liberty can guide constitutional adjudication in several matters, including free speech, federalism, and the constitutionality of affirmative action.5

I have myself defended a similar view of the Constitution: that it aims to create what I called a “partnership” rather than a majoritarian form of democracy by insisting that all citizens are entitled to an equal role and voice in their self-government, that government at all levels must treat citizens with equal concern, and that government must leave individual citizens free to make the personal decisions for themselves that they cannot yield to others without compromising their self-respect.6

That partnership conception is, of course, only one possible vision of democracy. Still, it illustrates the crucial distinction between the kind of political convictions a constitutional philosophy requires a justice to employ in reaching a constitutional decision and the kind of political positions it excludes. No judge could be guided by such a partnership conception without convictions about, for example, whether minority preferences deny equal concern to white applicants or whether questions about the use of one’s own body in reproduction are among those that dignity requires be left to oneself as an individual. But these general convictions about the Constitution are independent of a judge’s personal opinions about whether affirmative action is wise policy or whether abortion is immoral, and they may therefore act as a screen insulating a judge’s constitutional opinions from his personal or partisan preferences.

It would be nonsense to say that a judge who has a constitutional philosophy of the kind these examples illustrate has set aside all his own moral and political convictions to decide just in accordance with what the Constitution requires. He relies on his own convictions about the best conception of democracy and the role of the Constitution in creating that form of democracy in order to decide what the Constitution, properly understood, does require. But the crucial point is that these convictions about the character of democracy are independent of the more immediate and partisan policy preferences that lead a judge to vote for one party or the other or to favor one legislative scheme about some particular regulatory issue over another. It is the possibility of a constitutional philosophy of this sort—and only that possibility—that can give meaning to a Supreme Court justice’s claim that he sets his own political preferences aside in deciding constitutional cases.


So we must look at the record of the Senate hearings carefully to discover whether Roberts revealed a constitutional philosophy, even indirectly, and if so what his constitutional philosophy is. Senator Orrin Hatch put that question to Roberts early in the hearings. He asked him whether he is “an originalist, a strict constructionist, a fundamentalist, perfectionist, a majoritarian or minimalist….” Roberts replied that he resists labels and wants to be known only as a modest judge, that is, a judge who does not legislate or execute the laws but simply enforces the law correctly. Hatch was not satisfied and continued: “You are probably eclectic [in] that you would take whatever is the correct way of judging out of each one of those provisions? There may be truths in each one of those positions, and none of them absolutely creates an absolute way of judging.” Roberts replied: “Well, I have said that I do not have an overarching judicial philosophy that I bring to every case.”

Senator Grassley returned to the issue of Roberts’s constitutional philosophy later. He summarized an exchange Roberts had had during the hearings considering his successful nomination to the circuit court. Grassley recalled Roberts as saying then, “I do not have an all-encompassing approach to constitutional interpretation…. I would not hew to a particular school of interpretation, but would rather follow the approach or approaches that seem most suited in the particular case to correctly discerning the meaning of the provision at issue.” Roberts accepted Grassley’s recollection of his earlier statement and then added, falling back once again on his hollow reference to law, “And I don’t have an overarching view…. [I] take a more practical and pragmatic approach to trying to reach the best decision consistent with the rule of law.”

This more “pragmatic” approach, which lets a judge’s sense of how the case should be decided govern the method of interpretation he uses to decide it, rather than the other way around, is the most open invitation possible to result-driven adjudication, that is, to a judge putting his own policy preferences into his decision rather than trying to filter them out. Of course Roberts is right that a constitutional philosophy should not be an academic straitjacket that makes the facts of particular cases and the practical consequences of a decision one way or the other irrelevant. But that means only that a decent philosophy must show how facts and consequences are relevant, not that no such philosophy is needed.

The effect of these discussions was that Roberts declined to describe and embrace any general theory of constitutional adjudication.7 But he did discuss what must be an important part of any such theory—the doctrine of precedent—extensively. Most of the Judiciary Committee was interested in his opinions about precedent for a single reason: they wanted him to say either that he would or would not vote to overrule the central holding in the famous 1973 Supreme Court decision in Roe v. Wade that states may not prohibit early abortion. The Court had itself reaffirmed that central holding in 1992, in its decision in Planned Parenthood v. Casey. The hearings therefore offered the unusual spectacle of liberal senators insisting that established precedents should not be disturbed and conservative senators insisting with equal fervor that they sometimes should be.

Roberts declared that since the question whether Roe should be overturned would very likely come before the Court, he should avoid expressing any opinion about whether that case was correctly decided or whether, if not, it should now be overturned. But he did say, several times, that he had great respect for the doctrine of precedent, that any earlier decision of the Court is entitled to “deference,” that a justice should therefore not vote to overrule an earlier Supreme Court decision just because he thinks it was wrongly decided, and that several other factors bear on that issue, including whether people had changed their positions in reliance on the past decision, whether that decision had proved “unworkable,” and whether its basis in law had been eroded by other Supreme Court decisions taken since it was decided. He noted that the three-judge plurality in the case of Planned Parenthood v. Casey had themselves proposed those tests and had concluded that they should not overrule Roe; but he refused to say whether he thinks they were right in that conclusion. He also noted that the Court sometimes has overruled its own past decisions of long standing, and cited, as the obvious example, the Court’s 1954 decision in the Brown case overruling its much earlier decision, in Plessy v. Ferguson, that public school segregation by race is constitutional.

Roberts succeeded in leaving it entirely unclear whether he would vote to overrule Roe when the Court is next presented with the opportunity to do so, though it is worth noting that conservative commentators think he may do so.8 But his careful equivocation also left it unclear whether he has any view of precedent that would prevent his own opinions about the morality of abortion from dictating that decision, because he left open what he means when he says that a past decision has proved “unworkable.”

Is a decision unworkable when after three decades it continues to provoke passionate and even violent opposition and has therefore failed to resolve a divisive social conflict? That is a crucial question in considering Roberts’s likely votes not only about abortion but about many of the Court’s other precedents. Would his theory of precedent permit him to repeal long-established precedents barring prayer in public schools on the ground that these too had proved “unworkable” in resolving deeply divisive issues?

In an important though little-noticed exchange, Senator Arlen Specter asked Roberts whether overruling Roe would mean bowing to public pressure, which the Court ought not to do. Roberts replied only that public condemnation of a precedent is “a factor that is played different ways in different precedents of the Court.” He said that in one case, in which the Court had reversed its earlier opinion that victims may not testify in the sentencing phase of a criminal trial, the fact of wide public dissatisfaction with the precedent was taken as a ground for overruling it. He did not, however, indicate whether he thinks that ground is appropriate.

So Roberts’s discussion of precedent, though extensive, offers no reassurance that he will not follow his own political preferences in deciding which precedents to sustain and which to overrule. We must reach the same conclusion about his discussion of another jurisprudential issue: “judicial activism.” Conservative politicians once gave that name to the practice of liberal justices who overruled state and national statutes—including statutes making abortion or homosexual sodomy a crime—to create what the liberals considered a more just society. In recent years conservative judges have been much more willing than liberals to strike down congressional statutes in order, according to liberal critics, to create the more limited federal government that conservatives favor. In the two most notorious such decisions the Court ruled that Congress has no power to make carrying a gun and bullets into school a crime or to give women who are the victims of “gender-motivated” violence a civil remedy. In both cases, the conservative justices declared, the Constitution limits congressional power to matters of “interstate commerce,” and gun possession and violence against women are matters of local, not interstate, transaction. So liberals now accuse conservative justices of “activism.”

Several senators, both liberal and conservative, asked Roberts to comment on whether and when the Supreme Court should be active in that way. Roberts properly replied that judges have a constitutional duty to test legislation against the Constitution and that it is not any sin of activism but simply a mistake when judges strike down a statute they ought to sustain, just as it is not activism but simply a mistake when they sustain a statute they ought to invalidate. He therefore rejected—in my view correctly—another possible constitutional philosophy that might insulate a judge’s decisions from his own policy preferences: that judges should respect any elected legislature’s decision about the scope of its own powers except when its violation of the Constitution is undeniable. But once again he put nothing in the place of that theory except yet another anodyne appeal to the illusory constraint that judges should follow “the law.”


I do not expect Roberts to vote to overturn Roe v. Wade, or even to overturn the Court’s repeated decisions sustaining limited affirmative action programs in state universities and professional schools. It seems likely, moreover, that neither decision will be seen as in the best interests of political conservatives. Overruling Roe would suddenly make abortion again an urgent national political issue for many millions of women who have come to take for granted the right that women have enjoyed for two generations and who now vote to express their views on other issues, often for Republicans. Overruling Roe would not be helpful to the national Republican Party. Nor does the Bush administration actually oppose using the law to achieve greater diversity in universities and professional schools, a goal endorsed by the military and major corporations in briefs submitted to the Supreme Court in the Michigan cases of 2003 testing the constitutionality of affirmative action programs. The administration itself submitted a brief urging the Court to declare the Michigan programs unconstitutional, but as I said in my discussion of those cases in these pages, it did not in fact call for the elimination of all programs whose explicit goal is to improve racial diversity.9 It is argued only that different means should be found for pursuing that goal.

The danger is greater, in my view, that Roberts will join with the other conservative justices in extending the President’s power to conduct his war against terrorism without regard for either international law or the traditional rights of prisoners. During the hearings he insisted that the Bill of Rights remains in full force during a war, and he seemed to reject Chief Justice Rehnquist’s famous comment that though the laws are not silent in times of war they speak in a quieter voice. Yet he emphasized the president’s powers as commander in chief and suggested, when Senator Patrick Leahy asked him whether the president had the power to order the torture of prisoners, that that depended on whether Congress was “supportive” of the president’s action.

That view hardly seems consistent with the assumption that the most fundamental rights hold with full force even in wartime. When Senator Russell Feingold asked him whether he had any “concerns about the practice of extraordinary rendition, of our government secretly sending people to countries that we know use torture,” he replied that he could not comment because the issue might come before the Court “in one form or another.” But it hardly seems likely that the Supreme Court will be called upon to judge the odious practice of sending prisoners abroad to be tortured because it is unlikely that any government would admit to the practice.

In one of Roberts’s decisions as a circuit court judge that I mentioned earlier, Hamdan v. Rumsfeld, he joined an opinion that took an extraordinarily broad view of the president’s war powers, a view that was unnecessary to the decision.10 The Geneva Conventions, to which the United States is a party, provide that military prisoners who do not qualify for the full protection of prisoner-of-war status are nevertheless entitled to “humane” treatment and to “the judicial guarantees which are recognized as indispensable by civilized peoples.” President Bush has determined, by executive fiat, that the prisoners at Guantánamo whom he accuses of aiding al-Qaeda are not entitled even to that very modest protection.

A prisoner who is accused of being Osama bin Laden’s driver and bodyguard brought a habeas corpus challenge to his forthcoming trial by a military commission citing that provision of the Geneva Conventions among other reasons why he should not be tried before such a commission. Though Circuit Court Judge A. Raymond Randolph held that in any event treaties cannot be enforced in federal courts, he also declared that Bush’s opinion that the treaty should be interpreted as not applying to particular prisoners is entitled to judicial deference, and Roberts, without writing a separate opinion, joined in that declaration. A third judge, Senior Circuit Court Judge Stephen Fain Williams, joined in the decision on the first ground but protested that though a president’s “construction” is entitled to “great weight,” Bush’s interpretation of the Geneva Conventions’ language was so clearly wrong that it should not be accepted. Roberts could have joined with Williams rather than Randolph without affecting the overall decision in the case, but he took the opportunity to declare an amazingly broad view of the president’s powers.


As this is written, there seems no doubt that the committee and then the Senate will confirm Judge Roberts’s nomination, probably, in the latter case, by a large margin. He is a stunningly intelligent lawyer who may well prove to be an excellent chief justice. The country will have to wait and see. But Senator Biden was right when he said that in approving his nomination the Senate is “rolling dice.” The Judiciary Committee allowed him to keep his jurisprudential convictions, if he has any, almost entirely hidden. The senators asked him to comment on very specific cases and issues, an invitation he steadily—though with at least one notable exception—refused.11 I believe he was wrong to refuse to answer these specific questions. His argument that it is unfair to litigants to reveal his present opinion of issues he might later confront is very weak. His honest statement of his present views would in no sense be a promise or commitment. He will have to consider arguments in specific cases before making a decision, and he will join a Court most of whose other members have publicly stated their opinions on many of the issues that will come before them without raising any question of fairness to future litigants, who must often argue knowing that certain justices are disposed to vote against them. His argument, moreover, wholly neglects a very powerful contrary consideration: that according to any plausible view of democracy the public has a right to know his views on matters affecting their fundamental rights in some detail before their representatives award him lifetime power over those rights.

Whether or not Roberts answered the committee’s detailed questions about particular issues, however, it should have pressed him on the character of the more general constitutional philosophy he would employ finally to decide those issues when they arise. It should not have accepted his reiterated banalities about being guided by the law, or deferring to the rule of law, or taking due account of precedent, or deciding legal issues in a practical, pragmatic way, or allowing the facts their “proper role.” It should have asked him what all those words and phrases actually mean. The Senate has no right to gamble with the nation’s constitution and its future.

The committee will shortly have another, equally important, opportunity to protect both when Bush nominates a successor to Sandra Day O’Connor. We must hope that it has learned from its failures in the Roberts nomination. It should demand to know the new nominee’s constitutional philosophy. If he or she refuses to disclose it, or claims that it is only to respect the rule of law and adds nothing more helpful about what that means, then its constitutional duty is to advise the Senate to reject that nominee as either disingenuous or incompetent.

—September 21, 2005

This Issue

October 20, 2005