Harvard Law Review Harvard Law Review Harvard Law Review

Constitutional Law

Does the Constitution Mean What It Says?

The full text of the Foreword may be found by clicking the PDF link below.

“What do you do when there is a practice that flatly contradicts a clear text of the Constitution? Which of the two prevails?” Justice Scalia asked the Solicitor General that question during the oral argument in NLRB v. Noel Canning, the case about the President’s power to make recess appointments that was decided during the Supreme Court’s 2013 Term. The U.S. Court of Appeals for the District of Columbia Circuit had held that the text of the Recess Appointments Clause did not permit the President to make recess appointments in circumstances in which, the government argued, the President had been making those appointments for decades. If the D.C. Circuit was right about the clear meaning of the text, shouldn’t it — Justice Scalia’s question suggested — prevail over even a well-established practice?

The Solicitor General questioned the premise that the language was clear, but his answer was that “the practice has to prevail,” even over clear language. He explained, though, that “we don’t have that here” because “[t]his provision has been subject to contention as to its meaning since the first days of the Republic.” In fact, the Solicitor General said, “[i]t’s extremely unlikely that [the question] would arise if the text were so free of doubt.”

Actually, it might not be unlikely at all. If we read the text of the Constitution in a straightforward way, American constitutional law “contradicts” the text of the Constitution more often than one might think. Adhering to the text would require us to relinquish many of the most important and well-established principles of constitutional law. Here are some examples of the anomalies — outcomes that are inconsistent with established principles of constitutional law — that following the text of the Constitution would produce:

  • Both the President and the federal courts could abridge the freedom of speech and prohibit the free exercise of religion, because the First Amendment, by its terms, applies only to “Congress.”
  • A state could have an established church.
  • The requirement that “searches” not be “unreasonable” would not apply to the interception of many electronic communications.
  • States could disenfranchise, for example, poor people or gay people, because the Equal Protection Clause of the Fourteenth Amendment does not protect the right to vote.
  • States could discriminate against racial minorities, or any other group, in providing, for example, schools, parks, or transportation, because the Equal Protection Clause, by its terms, refers only to “protection.”
  • The federal government would be free to engage in discrimination on the basis of race or sex, because the Equal Protection Clause applies only to the states.
  • The federal government, under a standard application of the principle of expressio unius, could not enact criminal laws, except in certain narrow areas.
  • The Bill of Rights would not apply, in whole or with the current exceptions, to the states.
  • States would not be immune from suit when federal statutes made them liable, at least in state courts and when sued by their own citizens in federal court.

Some of these examples are well known, like the principle of Bolling v. Sharpe, according to which the law developed under the Equal Protection Clause applies to the federal government. Other instances in which we have departed from the text of the Constitution are obvious but more or less systematically ignored, like the fact that the First Amendment begins with “Congress.” Others require more explanation, which I will provide. But the important point is that these examples cannot be dismissed as curiosities that don’t really have much to do with the nature of American constitutional law. They involve important and well-established principles of constitutional law that are inconsistent with the text. They are anomalies not in the sense that they are odd excrescences but in the sense that they call into question our familiar way of thinking about the relationship between the text of the Constitution and constitutional law.

That familiar way of thinking, I believe, goes something like this: Constitutional law is derived from the text of the Constitution. Sometimes the text is clear; if so, it controls. But many provisions are not clear. They must be “interpreted” according to relatively familiar, if sometimes controversial, principles. Always, though, constitutional law begins with the text and proceeds from there.

That familiar view misdescribes American constitutional law. Clear text does not always govern, as the anomalies show; there are times when established principles are simply inconsistent with the text. Beyond that, constitutional “interpretation” usually has little to do, in practice, with the words of the text. There are times when the text is decisive, and it is never acceptable to announce that you are ignoring the text. But routinely the text, although not flatly inconsistent with the outcome of a case, has very little to do with the way the case is argued or decided. In most litigated cases, constitutional law resembles the common law much more closely than it resembles a text-based system.

So there is a puzzle: How can we make sense of a system that treats the text in this way? If the familiar understanding — start with the text, and proceed from there — is incorrect, just what role does the text play, and how can that role be harmonized with the other sources of constitutional law?

The answer, I think, is that provisions of the text of the Constitution are, to a first approximation, treated in more or less the same way as precedents in a common law system. The effect of constitutional provisions is not fixed at their adoption — or, for that matter, at any other time. Instead, like precedents, provisions are expanded, limited, qualified, reconceived, relegated to the background, or all-but-ignored, depending on what comes afterward — on subsequent decisions and on judgments about the direction in which the law should develop.

As a result, a provision of the Constitution will be most important when there are not a lot of subsequent precedents “interpreting” it. Once a large body of precedent-based law has developed, the text will recede from view. Like a venerated old case, a textual provision might be invoked for rhetorical effect and maybe even for more than that — in the way that, for example, Marbury v. Madison might be invoked when a court is considering the constitutionality of an act of Congress. But just as the reasoning and precise holding of Marbury do not determine the scope of judicial review today, so the text of the Constitution often does not determine the shape of constitutional law today — as the anomalies dramatically show. It is true that the Supreme Court would never “overrule” a provision of the text, in the way it might overrule a precedent. But the anomalies — instances in which the text has been effectively overridden by later developments — suggest that there is less to this difference than meets the eye.

These claims will have to be qualified in some important ways. But I believe the idea that provisions of the Constitution function roughly in the same way as precedents in a common law system is a better description of U.S. constitutional law than the familiar view that constitutional law is a matter of “following” or “interpreting” those provisions. And because the common law is a well-established approach to the law that has plausible (at least) justifications, understanding constitutional law as an evolutionary, common law system avoids the notorious problem of explaining why constitutional provisions adopted decades or centuries ago should continue to bind us today.


* Gerald Ratner Distinguished Service Professor of Law, the University of Chicago. I am grateful to William Baude, Paul Crane, Andrew Crespo, Ryan Doerfler, Vicki Jackson, Richard Lazarus, Brian Leiter, Richard McAdams, Jennifer Nou, Martha Nussbaum, Louis Michael Seidman, Mark Tushnet, Adrian Vermeule, David Weisbach, and participants in workshops at the Harvard Law School and the University of Chicago Law School for discussion and comments; to the Burton and Adrienne Glazov Faculty Fund at the University of Chicago for financial support; and to Lauren Walas and Kendell Coates for excellent research assistance.