Making a Constitutional Convention Safe for Democracy

For over two decades, conservatives in the United States have been trying to trigger a provision of the Constitution that has never been used—a clause that gives states the power to demand that Congress “call a Convention for proposing amendments” to the Constitution itself. The movement is supported by a wide range of leaders and activists, from a former Comptroller General, David Walker, to a former governor of Wisconsin, Scott Walker. Among currently elected politicians, its supporters include South Carolina senator Lindsey Graham and Florida governor Ron DeSantis. So far, legislatures in twenty-eight states have called for fiscal-integrity amendments—mandating a balanced budget. A powerful and growing faction of the movement, the Convention of States (COS) project, has gone further. Founded in 2013, COS is pushing for a convention to propose amendments that would restrict the scope of federal power and impose term limits on members of Congress and what it calls federal “officials.”

In their new book, The Constitution in Jeopardy, the former Democratic Senator Russ Feingold and the lawyer Peter Prindiville warn that all this organizing aims not to ratify the will of a democratic majority but to push amendments that entrench the beliefs of a right-wing minority.¹ They quote one of the cofounders of Convention of States, Michael Farris, at a simulated convention the group hosted for state legislators in 2016:

If you put enough pressure on state legislatures, you can get stuff done. You don’t need a majority of America, because a majority doesn’t participate.… In state legislative matters, less than 1 percent of the people ever participate. With 1 percent of the American public, that’s three million people, I guarantee you we can get this done.

This movement, Feingold and Prindiville argue, is strong and growing. Recent reports suggest that it has received extensive financial support from conservative backers—including, in the case of COS, the Mercer Family Foundation and the Kochs. It has become a central project of the American Legislative Exchange Council (ALEC), the nonprofit organization of conservative state legislators and other private sector representatives. Their book places the movement in a fascinating and rich historical setting. But their aim in writing it is quite clear: this minoritarian movement, they believe, must be stopped.

Feingold and Prindiville also believe that the Constitution needs reform. Feingold is an outspoken critic of the Supreme Court’s campaign-finance jurisprudence and has pressed for an amendment to address the damage done by Citizens United. Their book is convincing about the need for other amendments as well, including one to change the process by which the Constitution is amended. But because they believe a convention is too dangerous, they insist that it be Congress, not a convention, that proposes the amendments the nation needs.

And yet they both know—Senator Feingold especially—that there is exactly zero chance of Congress mustering the two-thirds majority necessary to propose any amendment of substance. By contrast, the convention movement is already close to garnering the support it needs to obligate Congress to convene a convention. Constitutional reform, Feingold and Prindiville fear, is just over the horizon; it’s just not the sort of reform they want to see.

How should people who are not on the right respond to this prospect? Must a convention be minoritarian? Or is there a way to render a convention safe—even beneficial—for majoritarian democracy?

Every modern written constitution includes a procedure for its own amendment. Ours, drafted in 1787, began that tradition. In Article V, the Constitution describes two ways that amendments might be proposed and two ways proposed amendments might be ratified. Amendments can be proposed either by the vote of “two-thirds of both Houses” of Congress or by “a Convention for Proposing Amendments,” if two thirds of the states make “Application” to Congress to “call” such a convention. Proposed amendments are then ratified by either three fourths of state legislatures, or by conventions in three fourths of the states. Congress gets to decide which of these two modes of ratification is used for any proposed amendment.

In the history of the nation, amendments have only ever been proposed by Congress. And every enacted amendment has been ratified by state legislatures, except one: the repeal of Prohibition, which was ratified by state conventions in 1933. Why then was this essentially unused convention clause included? Its most prominent advocate at the convention that drafted our Constitution was the Virginia delegate George Mason, who feared that if only Congress could propose amendments and “the Government should become oppressive,” then “no amendments of the proper kind would ever be obtained by the people.” The solution was a process whose substance Congress could not control. The states could obligate Congress to call a convention, whether Congress wanted it or not. Congress would then be obligated to send any proposed amendments from that convention to the states. Whether it would actually do so is another question.

Though there has never been a convention, the process of states calling for one has been enormously consequential. Between 1893 and 1912, the nation came within one state of the number necessary to require Congress to call a convention. “Fear of a convention,” as Feingold and Prindiville put it, “and the constitutional reforms that fear can inspire,” drove Congress to propose a series of amendments to staunch the movement—including ones that would give the people the power to elect senators, give Congress the power to levy an income tax, give women the right to vote, and give America Prohibition. All four proposals became law. Jill Lepore counts this as one of the periods when the Constitution was “rewritten,” and it would not have happened as it did without the convention movement.

This story has intrigued modern constitutional reformers on both the left and the right. Why shouldn’t they use the same method—the threat of a convention—to get Congress to act today? The answer, according to Feingold and Prindiville, is that the system is untested and that the language in the Constitution—against the background of the founding convention, which acted beyond the expectations of many who initially called for it—creates too much uncertainty. Even if the states are adamant that they intend a convention to have a limited scope, Feingold and Prindiville argue that nothing would stop a convention, once convened, from considering—and recommending—any question its delegates wanted. Courts are not going to enjoin a convention from talking or voting however it likes. “A convention’s amendment proposition authority,” they conclude, is “just like that of Congress”: unlimited. So a convention called for one purpose—say, to consider a balanced budget amendment—could detour to a radically different purpose, like abolishing birthright citizenship.

Even worse, from their perspective, is how the convention would make its proposals. The dominant view among the activists Feingold and Prindiville criticize is that the convention would vote by state, each state getting a single vote. That could mean that if a convention did take up an amendment outside its original mandate, and delegations from just twenty-six states—possibly representing a small fraction of the nation’s population—supported it, then that proposal would be sent to the states for ratification. Scary stuff indeed.

Feingold and Prindiville may overstate their case. Some convention advocates do believe that twenty-six states could pass anything. But the largest actual organizing by pro-convention state legislators is for rules requiring, first, that an amendment needs the support of thirty-six states to be proposed and, second, that no amendment may be proposed whose “subject” was not included in the “Application” of at least thirty-four states. The rules themselves say that they can’t be amended without the support of thirty-four states, and they purport to make it impossible altogether to amend the restriction on which “subjects” the convention can consider. Who knows whether these rules would govern the convention once it was called. If they do, and if they are followed, then the convention simply cannot take up subjects beyond the topics included within the states’ applications. 

Even if you’re not comforted by these rules, you may say: So what? If it takes thirty-eight states to ratify any amendment, then the vote of even a single chamber from thirteen states could stop it. Nothing crazy is ever going to get over such a high threshold. Feingold and Prindiville are not so sure. And even if the worst isn’t enacted, at least immediately, they fear the effect of unratified amendments as well. If a proposal is made without an expiration date, it could haunt the political process within a state and “radically transform state politics in purple states,” they write. “Amendment ratification would become a key wedge issue and attract millions in dark super PAC spending to otherwise backwater races for mainly part-time state legislative positions.” Combine this with the reality that, for strong reasons of judicial restraint, no court would ever stop a convention from considering whatever topic it wished, and you can understand their anxiety: Here is a movement, operating under the national radar in state legislatures around the nation, with generous funding from right-wing billionaires, that could initiate a process by which delegates representing a small minority of the country might have the power to propose amendments to the Constitution. What could possibly go wrong?

Just how close we are to Congress being obligated to call a convention is a matter of significant debate. Thirty-four states must first submit applications by passing resolutions through their state legislatures. So far, according to those keeping track, twenty-eight states have made a call for a fiscal-integrity amendment; five have called for a convention limited to proposing amendments about term limits; three for an amendment to overturn Citizens United; and nineteen for a convention to, as the COS language puts it, “impose fiscal restraints on the federal government” and restrict its “power and jurisdiction.” None on their own are close to thirty-four states—unless you count the much older calls for a “plenary” convention (a convention not limited to particular topics) that have not yet been rescinded. The scholar Robert Natelson, a supporter of the convention movement, counts these plenary calls, which now include four states beyond those supporting a fiscal integrity amendment. If you add those four to the calls for a fiscal-integrity convention, that’s thirty-two—two short of the magic number.

Yet not everyone in the conservative convention movement is as fastidious about aggregation as Natelson. Scott Walker believes there are already enough applications to empower Congress to act. Indeed, if you explore the remarkable online database The Article V Library, there are any number of topics (including repealing the Income Tax Amendment) that, including plenary applications, at some point have had more than the number of state applications needed to allow Congress to act. If Congress counted plenary applications and ruled that once the threshold of thirty-four had been crossed, its obligation to call a convention was set, then it could easily conclude it has the power to convene a fiscal integrity convention—today.

Feingold and Prindiville resist this possibility by arguing that Congress should have no power to call a convention unless there are thirty-four contemporaneous applications on precisely the same subject. They would not include the ancient plenary calls that Natelson does. By their measure, we would be at least six states away from constitutional Armageddon.

But Feingold and Prindiville acknowledge a problem with this argument: namely that, for the same reason they believe a convention could not be limited, neither can Congress be limited in its judgment about whether there are enough applications to empower it to call one. Just as the courts would not dictate what topics a convention could consider, so too would they refrain from stopping Congress from aggregating state resolutions as Congress chooses—not because courts are lawless but because they avoid extending their jurisdiction where there is no clear law to be found. Thus, both at the convention and in Congress, if there is political will, then politics will find a way. There might be no chance of convening a convention so long as either chamber of Congress is held by Democrats. But if the Republicans regained control of both chambers, all bets are off.

Feingold and Prindiville believe the convention movement can be stopped. I don’t believe it can: there is too much money and too much anger fueling it just now. But either way, I do believe it is irresponsible—a kind of constitutional malpractice—for progressives to fight to stop an Article V convention altogether rather than trying to engage with the convention process, if only to steer it to majoritarian values. I believe this not because I think that anyone could persuade conservatives not to fight for conservative causes, but because I am convinced that if more than conservatives tried to pass calls for a convention, the convention that those calls yielded could be radically different. Indeed, as I show below, it could be rendered fundamentally majoritarian.

This point is even stronger if you believe, as I do, that our Constitution as currently interpreted is deeply flawed. The conservative convention movement fights primarily for fiscal integrity. That may well be a serious national concern. But even more serious is the gap in representational integrity: the corrupting dependence upon money in politics, weaponized gerrymandering, the lack of meaningful term limits, the lack of an affirmative right to vote, and a system for electing the president that makes irrelevant most voters in America.

These issues too will not be addressed by Congress on its own—even though reforming them is wildly popular on both the left and right. If you believe—and who couldn’t?—that fixing them is the first step to a Congress that could address matters important to America, why wouldn’t we engage with the only process likely to give the country a chance at that reform? More pointedly, if a convention is coming with or without us, why would we work to remove democratic reform from the permitted topics of a convention—as Common Cause, maybe the most important pro-democracy organization in American history, has done by rallying state legislatures to rescind calls for a convention, even those designed to address the corrupting influence of money in politics? Rather than cowering in fear of uncertain risks, we should be asking whether there are ways to minimize these risks, by securing a convention that could only propose reforms that a clear majority of Americans wants.

Because it turns out that this is now possible. In a pair of decisions unrelated to conventions directly, the United States Supreme Court has opened the possibility of taming a convention to majoritarian values. Three years ago, in two cases we could call The Electors Cases, the Court held that the vote of presidential electors may be directed by a legislature and that electors can be removed if they threaten to vote contrary to that direction. Though the Framers plainly imagined electors exercising discretion and playing a much more robust role in determining the chief executive, history, the Court concluded, had rendered them differently. To recognize a state legislature’s power to tell its electors “that they have no ground for reversing the vote of millions of its citizens,” Justice Elena Kagan wrote in her majority opinion in the first of The Electors Cases, Chiafalo v. Washington (2020), “accords with the Constitution—as well as with the trust of a Nation that here, We the People rule.”

This conclusion was surprising, at least to me—I had argued the other side before the Court. But it has enormous potential for making an Article V convention safe. The Court has expressly held that in ratifying an amendment, no legislature can be controlled by any outside force, including a referendum. But The Electors Cases should mean that delegates to a convention could be legally controlled—in just the sense that if they vote contrary to how they are pledged, they can be removed and replaced by delegates who will vote as directed. If presidential electors, a position created by the Constitution itself, can be directed by the peoples’ vote, then “we the People” certainly should have the power to direct delegates to an Article V convention.

Controlling delegates is not a new idea. Feingold and Prindiville discuss the laws that exist in many states that purport to forbid delegates to a convention from voting contrary to how the legislature directs. They insist, however, that these laws are “toothless” because they don’t stop the legislature from “chang[ing] course once a convention is called”—by, for example, directing delegates to vote differently from how the legislature originally determined.

Yet this misses a critical possibility that The Electors Cases throw into relief: if the delegates are directed to reflect the will of the people, rather than the legislature, then the legislature should have no power to change how they must vote. That should be clear with presidential electors; it should be even clearer with delegates to an Article V convention. Imagine, for example, that a state legislature directs its presidential electors to vote as the people in that state vote. But then imagine that after one candidate wins in that state, the legislature changes its mind and passes a law directing its electors to vote for the candidate who lost. Would that change be allowed?

This is the critical ambiguity that the Supreme Court left unresolved in The Electors Cases, but only because it didn’t have to resolve it. Yet it would make no sense of the principle behind those cases—“here, We the People rule”—to allow a legislature to override the people’s will because it didn’t like how the people voted. “Here, We the People rule” must mean that if the legislature gives the people the vote, the people’s choice becomes binding.

That same principle has enormous potential for an Article V convention. There is already a rich history of binding delegates to state ratifying conventions: when state conventions voted to repeal Prohibition, more than a dozen states obligated delegates to vote as they had been pledged when elected. But the more interesting potential raised by The Electors Cases is binding the delegates at the national convention itself.

The idea would be this: across the world—perhaps most prominently in Ireland—there has been a rapid growth of “citizen assemblies,” gatherings convened to consider matters thought too controversial or too difficult for ordinary politics. These assemblies bring together ordinary citizens—in Ireland, along with legislators—in large, randomly chosen, but representative bodies. The assemblies are provided with the information they need to determine some important public issue. They then deliberate, sometimes over extended periods. At the end of their deliberation, they release their views. Think of it as a public opinion poll, but with an informed public given a chance to engage with each other. Depending on the jurisdiction, the assembly’s views might be binding. Ordinarily, they are just persuasive. In Ireland, assemblies have addressed questions about abortion and same-sex marriage. Their conclusions were wildly popular, ratified by national referenda, and impossible to imagine the ordinary Irish legislature ever producing.

Imagine, then, citizen assemblies convened within the states to constrain the discretion of delegates to an Article V convention. Such assemblies could be constituted in a far more robust way than in other countries, modeling the process on Stanford professor James Fishkin’s deliberative polls—large, representative, randomly selected samples of a state—but, like jurors on a jury, required to show up. The citizens would gather in some convenient place; they would be compensated for the true cost of their participation—not the $15 per day you get for jury duty, but travel expenses and the per diem of a government employee, plus lost wages, job protection, and maybe a bonus. They would be given briefing material in advance, and then meet and deliberate in small and large groups about which amendments or topics they would support. Delegates to an Article V convention would then be constrained by the conclusions of these citizen assemblies. For example, the state legislature could require that “no delegate from the State of X may vote to support a proposed amendment whose topic at least 60 percent of the state’s Citizen Assembly did not support.” And with that one rule, an Article V convention would be constrained by majoritarian democracy.

Some will resist the idea of a citizen assembly for fear that it would be dominated by extremists. This is a mistake. No doubt there will be extremists. But if the assembly is randomly selected and representative, the extremists would be a tiny minority because they are only a tiny minority of America. Article V could thus embrace the most ambitious experiment in representative democracy in generations—and, because of The Electors Cases, its results could be made legally enforceable within the convention.

But if each state gets one vote, couldn’t the convention still be minoritarian? This is where the rules governing the convention become critical. Under the rules proposed by the state legislators, the convention needs the vote of thirty-six states to propose an amendment. If you arrayed the states from most conservative to least, the vote of thirty-six states would represent almost 60 percent of America. If you arrayed the states from most liberal to most conservative, the vote of thirty-six states would represent more than 80 percent of America. The democratic constraint could assure that a supermajority of this supermajority would be required before any amendment could even be proposed—and thus that any proposed amendment would have substantial, even majority, popular support.

For this plan to work, delegates would have to be bound at the state level to vote as the state’s citizen assembly directs. What could assure that binding statutes—as well as the assemblies themselves—would be adopted in enough states? Here is the final part of the strategy for transforming what Feingold and Prindiville fear is dangerously antidemocratic into something potentially pro-majoritarian.

No amendment, however proposed, can be ratified unless thirty-eight states support it. If ratified in the ordinary way, that means that no amendment can be ratified if one house in each of thirteen states opposes it. Ordinarily, such opposition happens after an amendment is proposed. But why must that be so? Why couldn’t state legislatures reject an amendment preemptively in order to influence the convention process itself?

Imagine then that a state passes a law to convene a citizen assembly, requires that its delegates to a federal convention follow the results from that assembly, and requires delegates at a state ratifying convention to be elected and vote as they have been pledged. Long before any federal convention is convened, the state’s citizen assembly could begin its work, surveying proposals for constitutional amendments and deliberating about them. Then imagine the legislature passes a resolution something like this: that the state “hereby preemptively rejects any amendment proposed by a convention that does not entrench the voice of the People in both the convention and ratification process.” That resolution would not insist that each state create a citizen assembly; assemblies are ideal, but it would be enough if convention delegates were constrained by, for example, popular referenda and if delegates to a ratifying convention were constrained to vote as they were pledged when elected.

But here’s the key: If thirteen state legislatures passed such a resolution, in a single stroke the politics of a convention would change fundamentally. No one knows whether such preemptive rejections have any ultimate legal effect, even assuming they could be challenged in court. But their potential legal force would have an enormous political effect. By passing such resolutions, these thirteen states would be warning adherents of a minoritarian convention movement to abide by democratic checks or risk that their amendments would be stillborn. Why waste the effort to convene a convention in the face of state resolutions that could veto any amendment emerging from that convention in advance? Much more sensible would be to convene a convention that met these pro-majoritarian standards—in which case we could see a convention movement develop that embedded majoritarian constraints throughout the nation.

In the summer before the 2016 election, Cenk Uygur—the founder of WolfPAC, the most prominent group behind the anti–Citizens United convention movement—and I met with a prominent progressive to try to win them over to the idea of a convention. This progressive listened with understanding and care, but in the end, they were not convinced. “Why risk it?” they told us. “We’ll win in the fall, and Hillary will give us two, maybe three justices on the Supreme Court. Those justices will then give us everything we need from a reformed Constitution.”

That judgment in the summer of 2016 was not implausible. Yet I remember feeling a certain sadness in their response. When had we liberals become so afraid of the people? When had we become so captured by a legal elite? When, in a word, had we given up on the idea of a constitutional democracy?

I am not a partisan for progressivism. I am a partisan for the people. I believe that ordinary people, properly constituted, given a chance to understand and deliberate, would produce better political judgments on a wide range of critical issues than any group of elected representatives—especially the issues that worried Mason, where Congress itself is the problem.

Progressives in particular should understand this. Rather than retreating in the face of organizing on the right, they should be advancing ideals of democratic empowerment. The convention movement is on the cusp of giving Congress the power to call an Article V convention. The question we should be asking urgently now is how to make sure such a convention represents “the People.” The Supreme Court has shown us how. It has given us a mechanism to require that the delegates at a convention remain faithful to their charge. Let us use that principle and that mechanism to craft a majoritarian convention—and thereby give constitutional reform a fighting chance.