America’s Unamendable Constitution

Virtually all of the world’s codified constitutions are amendable using special procedures that require legislative or popular approval, and sometimes both. But an increasing number of constitutions—over half of the world’s new constitutions since the fall of the Berlin Wall—make at least one constitutional value unamendable, impervious to the people’s amendment power even where large majorities may wish to amend it. In Germany, human dignity is unamendable, as is republicanism in France, federalism in Brazil, socialism in Cuba, secularism in Turkey, and Islamic republicanism in Afghanistan. The U.S. Constitution, for its part, is entirely unamendable, from top to bottom, on all matters both big and small.

But the unamendability of the U.S. Constitution is not an intentional constitutional design, as it is in most countries, for instance in Italy where the authors of the constitution made a deliberate choice to exempt republicanism from the rules of constitutional amendment. Nor does the unamendability of the U.S. Constitution derive from constitutional interpretation, as in India, where the Supreme Court has determined that the constitution’s “basic structure” is unamendable despite there being no mention of unamendability in the constitutional text.

The unamendability of the U.S. Constitution derives instead from constitutional politics. In the present political climate it is practically unimaginable, though it nonetheless remains theoretically possible, to assemble the required supermajorities for a constitutional amendment. This is the gridlock that Sanford Levinson suggests, in the lead essay for this symposium, can lead to a Type 2 constitutional crisis. The deep divisions in Congress and among the states make unlikely even simple majority agreement on nearly anything of importance, let alone the onerous supermajority agreement the Constitution requires for amendment. Under Article V, two-thirds of Congress and three-quarters of the states must agree to a constitutional amendment, no small feat.

The best evidence of the difficulty of constitutional amendment in the United States is the historical rate of amendment failure. Literally thousands of amendments have been proposed since 1789, but only 27 have succeeded. Of those, then ten in the Bill of Rights were passed as a condition of adopting the Constitution to begin with; the Thirteenth, and Fourteenth, and Fifteenth were controversially forced upon the states during Reconstruction; and the Twenty-First repealed the Eighteenth. Others, like the Seventeenth and Twenty-Second, had become law by political practice well before they ultimately became law by formal amendment. It may therefore overestimate the number of actual amendments to simply count the ones that appear in the text of the Constitution.

This suggests to me that the U.S. Constitution is best understood as constructively unamendable. Its unamendability is not a legal fact but rather a political reality that we may infer from the current state of American politics. Reading the text of the Constitution alone gives the mistaken impression that everything is freely amendable, but really nothing today is amendable.

Federalism is also a cause of constructive unamendability. In federal democracies, the rules of constitutional amendment are designed to protect dual interests, and they commonly grant veto powers to both the national and subnational governments for amendments to federalist institutions. In Canada, for example, amending the method for selecting senators requires the consent of both Parliament and a supermajority of provinces. In Australia, amending the powers or boundaries of a state requires the approval of a majority of voters in that affected state in addition to majority approval in both houses of the national legislature and by national referendum. It is no coincidence that federal states like these often rank very high on indices of amendment difficulty.

Or consider a state’s representation in the U.S. Senate, for which the Equal Suffrage Clause requires the consent of the state whose representation is changed, a rule that could well be interpreted to require the unanimous consent of states, not just of the state whose representation is diminished. But amendment would nonetheless remain unlikely if the Clause required the consent only of the state whose representation had been diminished. Why would any state agree to that?

Indeed, the U.S. Constitution is one of the world’s most difficult to amend, if not the most difficult according to scholars of comparative constitutional law, though I have recently shown why the Canadian Constitution may be even more difficult to amend. But wherever the U.S. Constitution ranks on the scale, no one can deny that it is extraordinarily difficult to amend.

It was not always this way. The U.S. Constitution was once thought too easy to amend. In the Progressive Era, there were four constitutional amendments in a short span of fewer than ten years: the Sixteenth Amendment, authorizing a direct income tax in 1913; the Seventeenth Amendment, establishing direct elections to the United States Senate, also in 1913; the Eighteenth Amendment, imposing prohibition in 1919; and the Nineteenth Amendment, constitutionalizing women’s suffrage in 1920.

Constructive unamendability, then, is not a permanent feature of the Constitution. Just as new political fault lines may form to create intractable conditions that prevent the formation of the consensus needed for an amendment, so too may political circumstances evolve to alleviate the pressures that generated the intractable conditions to begin with.

The incompatibilities that today complicate the possibility of political agreement on an amendment may tomorrow dissolve or at the very least become less pointed, or the people may in the future apply sufficient pressure to push political actors past their present impasse, or some emergency may move political actors to rise above their partisan disagreements in the service of the larger good. These are all real possibilities that would be consistent with the lived American experience.

All of which leads me both to agree and to disagree with Levinson that America’s “most fundamental crisis” is the failure to accept that “the Constitution—and our being trapped into its byzantine ‘forms’—is a bug, and not a feature, of our political system.”

I agree that it is worth asking whether the constructive unamendability of the U.S. Constitution may have today descended the country into some kind of constitutional crisis. But it is not clear what kind of crisis it is. Where is the disobedience to law, which characterizes Type 1 crises? Or the blind fidelity to law that triggers Type 2 crises? Where are the extraordinary forms of struggle and opposition that occur in Type 3 crises?

The core of Levinson’s critique of the Constitution is not the gridlock itself but rather what allows the gridlock to persist. For instance, the design of the Senate—which gives population-poor Wyoming the same power as population-rich California—is both a symptom and a cause of the constitutional crisis he identifies. But as he has written elsewhere, perhaps the biggest problem is America’s veneration of its Constitution. Americans resist the idea of too much tinkering with the majestic creation of the Philadelphia Convention. What lies beneath this resistance is a culture of framer-worship, a deep respect for tradition, a fear of the unknown, and in some ways a belief in manifest destiny that only the Constitution as originally designed can secure. On their own and collectively, these are serious barriers to constitutional amendment, whether through the Congress-centric amendment process that has been used since after the founding or the convention-centric amendment process that Article V allows but that the people have never once used since 1789.

Although the Constitution has been amended relatively few times, its meaning has changed many times over the years, both in ways that do not appear in the constitutional text and in others that in fact belie it. The source of most of these unwritten changes—what some call informal amendments—has been judicial interpretation. The Supreme Court has become the modern Philadelphia Convention, a continuing constituent assembly that has merged into itself the two historically separated functions of interpreting and amending the Constitution.

But those who oppose what is derisively called “judicial activism” should remember that Congress, the states, and the people have the authority to reverse a Supreme Court opinion using the amendment power in Article V. Article V has been used in this way a handful of times thus far in American history. It may not be possible to use Article V in this way or indeed at all today, but the constitutional text is not fated to its present stasis.

True, getting two-thirds of Congress to agree to propose an amendment is unlikely. In the near term, then, the congressional pre-veto will stand in the way of an Article V amendment. And it is unlikely, though they have come close in the past, that two-thirds of the states could one day properly petition Congress to call a convention to propose amendments for state ratification.

One strategy to counteract the constructive unamendability of the Constitution is to find new ways to send amendment proposals to the states for their possible ratification. The modern transformation of American politics into a strong form of presidential government suggests that the President could offer a solution, if even a temporary one, where congressional inaction threatens to generate or exacerbate a constitutional crisis of any of three types Levinson identifies.

The President could take over the function the Constitution assigns to Congress and issue an amendment proposal directly to the states, with instructions to vote on the proposal in state conventions. States could then deliberate on the amendment in extraordinary assemblies, the amendment ultimately being adopted if three-quarters of the states vote in favor.

On one reading, this would itself be a Type 1 crisis: It amounts to disobedience of the rules of Article V, which make no mention of the President and authorize only Congress or a constitutional convention to propose amendments for the states to consider in either state legislatures or conventions.

But on another reading, Article V is not the only way to amend the Constitution. It specifies only certain but not all ways to alter the text. This president-initiated amendment procedure would not be a violation of Article V but rather a creative innovation within its prescribed structure of proposal by some national institution of government and ratification by the state institutions.

The reason why only state conventions should have the power to ratify a president-initiated amendment is the same as why the Philadelphia Convention instructed states to ratify the proposed constitution in state conventions: only extraordinary assemblies of the people can legitimate a change that may on its face appear to violate the formal rules of the existing constitution. Just as the Philadelphia Convention exceeded its mandate to revise the Articles of Confederation by proposing an altogether new constitution, here the President could be accused of exceeding the boundaries of Article V. But successful ratification by state conventions would transform what some would assuredly challenge as an illegal amendment into a legitimate one.

All of this of course presumes that the United States Constitution desperately needs to be amended or replaced. On that front, I take a different view from Levinson. I do not believe the Constitution is broken, nor that it is mired in crisis, nor that it has caused one today—nor however do I agree with Alexis de Tocqueville, who once wrote that it is “the most perfect federal constitution that ever existed.” It might still be improved. And just because Article V is unusable today does not mean that it will remain that way tomorrow.

Also from this issue

Lead Essay

  • Sanford Levinson proposes a typology of the constitutional crisis. Some crises, he argues, stem from political actors disregarding the written constitution. Others, he says, are the product of too much adhesion to the written constitution, even in circumstances in which the written text proves self-destructive. He reviews several incidents of each from American history, and he closes by suggesting, provocatively, that the U.S. Constitution may be undergoing a crisis even today.

Response Essays

  • Presidentialist systems are empirically unstable, and parliamentary systems seem superior in this regard, writes Jacob T. Levy. Not only that, but presidentialism encourages the false belief that the president at any given time represents a unified national will. These faults, though, are not indicative of a constitutional crisis particular to United States. All liberal democracies have been experiencing a crisis of confidence in recent years, at least since the end of the post–Cold War euphoria. Confidence in representative institutions has recently been low in general, not just in presidentialist systems. The role of American constitutional design in all of this would appear to be incidental.

  • Richard Albert notes that the U.S. Constitution has become practically unamendable: Congress is simply too divided to agree with the supermajority needed to report out an amendment, and the public appears permanently disinclined to make use of Article V’s provision for an amending convention. As a result, changes to the U.S. constitution have been taking place de facto, but almost never de jure; interpretations today do the work that amendments should probably do, and the potential for constitutional crisis grows. Albert closes with a bold proposal: Perhaps Article V does not provide an exhaustive list of methods by which the Constitution might be amended.

  • The slowness of the U.S. Constitution is a feature, not a bug, says Tom Ginsburg. It shields us from politically motivated, short-term thinking in our constitutional design. The alternative, a constitution that is too easy to change, is if anything more likely to provoke crises, as have recently happened in Hungary and Poland. Madisonian constitutional design lowers the stakes of politics, meaning that a president Trump or a president Sanders will be less able to do anything truly damaging. Even in a world of rapidly changing technology, Ginsburg says, we still would have to calibrate the degree of difficulty in constitutional change to match the degree of change needed to suit the situation, and this is likely a hopeless task. Perhaps some constant low-level degree of constitutional tension is the only alternative, and that may not be so bad after all.