Restoring Conventions, One Amendment at a Time

Stephen E. Sachs believes we should explore options for a “rolling” convention, one that needn’t meet at a single place at a single time but that nonetheless lets states begin the work of constitutional change. He is the Antonin Scalia Professor of Law at Harvard Law School.
Read Full Biography

Article V, as John Harrison has memorably put it, is the Constitution’s effort to bottle the lightning of popular sovereignty.[1] Through Article V the People can do just about anything that law can do: establish a monarchy, eliminate private property or the Bill of Rights, give war powers to the Supreme Court, or reduce the United States to a province of Canada—anything, that is, but deprive a state of equal suffrage in the Senate without its consent.[2] Yet in doing any of these things, the People never speak for themselves; and who may speak for them, through the device of Article V, is of the utmost importance to our constitutional law.

The Constitution recognizes two ways for the People of the whole United States to speak at once: through regularly meeting legislatures and through single-moment conventions. Our representatives in Congress can propose amendments, or our representatives in state legislatures can apply for a convention to do so; those amendments, in turn, must be ratified in the states, again by legislatures or by conventions.[3] But at both stages the convention method has become rather moribund. Only one amendment (the Twenty-First) was written to be ratified by state conventions;[4] and while states still submit amendment applications, as Gerard Magliocca points out, their function is primarily to spur other kinds of legislative responses.[5] Unofficial efforts like the Peace Conference of 1861 aside,[6] no national convention has been called since the one at Philadelphia.

Conventions aren’t valuable for their own sake. But their desuetude mirrors the decline of the amendment process as a whole. More than half a century has passed since the last constitutional amendment was successfully proposed and ratified, the longest such gap since before the Reconstruction Amendments.[7] Yet American constitutional politics hasn’t slowed down. Instead, Americans increasingly look to closed-door Supreme Court conferences and political-theater confirmation battles to settle constitutional questions that could have been solved by a political process—that is, by open amendments, openly arrived at.

To reopen these channels of constitutional change, we might draw new inspiration from the convention method. Article V used conventions in two ways: to let states be the first movers of constitutional change, and to let the People get around their ordinary lawmakers, whether state or federal. While the convention method remains in place for these purposes, we can also advance them in part, if not in full. The Framers’ design worked best for the blank-slate Framing in which they were engaged; such blank-slate revisions are understandably less popular today. Efforts toward a more “limited” convention, meanwhile, are hard to implement in practice and uncertain in their validity. Instead, we should explore options for a “rolling” convention, one that needn’t meet at a single place at a single time but that nonetheless lets states begin the work of constitutional change. Through a new amendment, we could flip the order of Article V, letting three fourths of states propose an amendment before two thirds of Congress would ratify it.[8] Flipping Article V in this way would preserve the constitutional threshold, the chance for national deliberation, and the relative political power of the states. But it would also help break some of the modern barriers to constitutional amendment—and would help restore constitutional politics to the political process, just as conventions were supposed to do.

1.  Why conventions?

To understand the convention’s role in Article V, we have to understand its far more important role in Article VII. In the former, it was merely a device to draft and ratify amendments; in the latter, it was the key to ratifying the Constitution as a whole. To the Founders, the Article VII conventions let the People of each state go over the heads of their state lawmakers and create new constitutional law for themselves and others. In the same way, Article V conventions let the People of the various states initiate changes to the Constitution, without waiting for Congress to lead the way.

1.1  Article VII

While the several states were surely represented at Philadelphia, the delegates had no power to bind their principals. Their job was to “report[]” on potential “alterations and provisions” in the Articles of Confederation, which alterations would be “agreed to in Congress and confirmed by the States.”[9] The Articles were themselves a sort of treaty, approved by the state legislatures and not to be amended without those legislatures’ unanimous consent.[10] The Convention delegates’ choice to produce something else—a new Constitution, one that would supersede the Articles among as few as nine consenting states—rendered their proposal a mere set of “informal and unauthorised propositions, made by some patriotic and respectable citizen or number of citizens,” and one “to be submitted to the people themselves,”[11] going over the heads of the state legislatures and exceeding their authority.

As Akhil Amar describes, conventions were a “particularly emphatic way” to seek the People’s voice, and to do so “more directly than ordinary legislatures.”[12] A nationwide referendum after Philadelphia would have been unthinkable, not merely because the states jealously safeguarded their independence from each other, but also because of their varying qualifications for the vote. (That’s why the Constitution looked to state eligibility rules to define the electorate for the House.[13]) Even statewide referenda were largely unknown; the Massachusetts and New Hampshire referenda of the 1780s had taken place through town meetings, which not all states used.[14]

Instead, state conventions both allowed for deliberation and had a greater claim to represent the People than the legislatures could. The Virginia Plan had called for approval by assemblies “expressly chosen by the people,”[15] and Madison urged that the Constitution should not, like the Articles of Confederation, rest “on the Legislative sanction only.”[16] If no legislature could bind a future legislature, then the new Constitution couldn’t claim supremacy over contrary state legislation; indeed, if it carried at most the imprimatur of some prior legislature, it couldn’t even override a state constitution.[17] To Madison, then, “the new national constitution ought to have the highest source of authority, at least paramount to the powers of the respective constitutions of the states”;[18] as James Wilson put it, “the people by a convention [were] the only power that can ratify the proposed system of the new government.”[19] Article VII didn’t say who should pick the state convention delegates, but the cover letter sent from Philadelphia to Congress specified “a Convention of Delegates, chosen in each State by the People thereof.”[20] In fact, many states conducted their ratifying conventions under special rules with reduced requirements for voting or officeholding, allowing “a uniquely broad class of citizens to vote” both for and as delegates.[21] In other words, the convention method was the most authentic means of democratic expression the Founders knew.

1.2  Article V

In this context, the use of conventions as a means of Article V proposal and ratification—and as a way to get around either Congress or the state legislatures—was a natural choice. The very first discussion of amendments in the Virginia Plan suggested that “the assent of the National Legislature ought not to be required thereto”;[22] George Mason feared that Congress “may abuse their power, and refuse their consent on that very account.”[23] As it came out of the Committee of Detail, the draft allowed for a “Convention for that Purpose” to be called by Congress “on the Application of the Legislatures of two thirds of the States.”[24] But the absence of a ratification requirement raised eyebrows,[25] and as Alexander Hamilton argued, Congress might also “perceive . . . the necessity of amendments.”[26] Roger Sherman would have allowed congressional proposals, but he also urged a unanimous ratification requirement, with “no amendments . . . binding until consented to by the several States.”[27] James Wilson unsuccessfully tried to change that requirement to two thirds of the states, and then successfully arrived at three fourths instead.[28]

As eventually reframed by Madison, the provision would have let Congress “propose amendments,” either by “two thirds of both Houses” or on “the application of two thirds of the [state] Legislatures,” which would then be ratified by three fourths of the states—through either legislatures or conventions, as Congress might select.[29] Importantly, though, these state applications would have been for amendments, not for a convention for proposing amendments. In other words, Madison wanted to let states agree on the language themselves—on which more below.[30] The change from amendment applications to convention applications was proposed later on, by Gouverneur Morris and Elbridge Gerry.[31] Their rationale isn’t clear from the debates; Madison thought Congress “as much bound to propose amendments” as it would be to “call a Convention.”[32] Yet the point may have been, not the mechanics of proposing amendments, but a desire for greater deliberation and range in penning them. Morris had previously urged “another Convention, that will have the firmness to provide a vigorous Government, which we are afraid to do”;[33] such changes could come only from a convention empowered to debate and reexamine the whole, rather than from piecemeal amendments forwarded by the states one-by-one. And whether or not Gerry agreed with this vision, they might both have wanted more legislative-style deliberation over the content of amendments. (Indeed, Gerry tried to remove conventions from the ratification process, leaving it to state legislatures only.[34]) With added protections for direct taxes, the slave trade, and equal suffrage in the Senate, Article V eventually took its current form.[35]

This history allows at least two conclusions. First, amendments were difficult because they had power to revise what had already been agreed upon. Rather than the three-fourths requirement “stand[ing] out as inexplicable,” as Magliocca argues,[36] it made sense as a compromise between those who emphasized the need to alter the Constitution in the future and those who expressed each state’s reluctance to cast its fate into the others’ hands. Each state had the chance under Article VII to look at the Constitution and say no to the deal; but allowing amendments under Article V with anything less than unanimity meant that other states could alter that deal later on. In that context, it’s hardly surprising that Wilson’s two-thirds standard might have been rejected as too lenient, and that only a broader agreement was allowed to bind an unconsenting state. (Indeed, Roger Sherman, who had favored unanimity, later worried that the standard was too low, and that “three fourths of the States might be brought to do things fatal to particular states.”[37])

Second, conventions were means by which the People of the several states could lead the process of constitutional change. By applying for a convention, they could leapfrog their recalcitrant representatives in Congress; by ratifying a proposal in conventions, they could leapfrog their recalcitrant state legislators. As Magliocca notes, while Congress has always had a choice of ratification by legislatures or by conventions, the only amendment to go the convention route was the Twenty-First—when Congress sought “to prevent malapportioned state legislatures from blocking Prohibition’s repeal,” a policy overwhelmingly favored by their constituents in the delegate elections.[38] Just as the Article VII conventions in 1787 and 1788 had overridden the authority of state legislatures and even state constitutions, Article V conventions were a way of breaking the logjams that had blocked necessary change.

2.  the “limited” convention

The first state applications for a convention, from Virginia in 1788 and New York in 1789, sought what are now called “plenary” or “unlimited” conventions—essentially ‘do-overs’ for Philadelphia that could revisit the Constitution in full.[39] At most twenty-one states have sought such conventions (though the number may be lower),[40] and of those ten have since purported to rescind their applications.[41] Although their substance may seem simple, these efforts toward a plenary convention are still fraught with uncertainties—whether the state applications can be rescinded before they vest, how many delegates each state must receive, which procedures the convention may follow, and so on.[42] (In fact, some of these problems were raised by James Madison at Philadelphia, when he urged that states should just submit amendment language instead.[43])

Yet the most important confusions center on the possibility of what are now called “limited” conventions, conventions charged to address—and, often, only to address—amendments on a particular topic or even with particular wording. As Magliocca notes, most state applications are for limited conventions, focusing on particular topics such as direct elections for Senators, one-person-one-vote, a balanced budget, or term limits.[44] (For example, the “Convention of States” project seeks a convention “limited to proposing amendments . . . that impose fiscal restraints on the federal government, limit the power and jurisdiction of the federal government, and limit the terms of office for its officials and for members of Congress.”[45]) These more limited applications also face more serious uncertainties—concerning their interpretation and meaning, their constitutional validity, and their eventual enforcement. Given the depth and seriousness of these uncertainties, those hoping for a revived convention process should look elsewhere.

2.1  Interpretation and meaning

Assume, for the moment, that a state may apply for—and Congress may call—a limited (or less-than-plenary) constitutional convention. How should we understand those limits? If one state applies for a convention “for the purpose of considering a balanced budget amendment,” can that be added to a call for a convention “limited to considering a balanced budget amendment,” or even one “limited to considering an amendment that balances the federal budget on a biennial basis,” for the purpose of reaching two thirds? Can we count plenary applications toward a more limited convention (and, if so, toward which)?[46] Or count certain more limited applications toward a plenary one?[47]

The problem is far from theoretical: as Magliocca notes, many of the convention applications prior to the Seventeenth Amendment were differently worded, forcing Congress to decide how to read them.[48] Article V’s text is silent on the matter: it simply says that Congress, “on the Application of the Legislatures of two thirds of the several States, shall call a Convention for proposing Amendments.”[49] Presumably what matters is whether the applications are equivalent in substance, rather than identical in wording (like bills passed by the House and Senate). But what counts as substance, and how close is close enough?

One way to understand the substance of a state application or a congressional call is to ask about the charge of the convention: what its delegates are obliged to do or not to do. For example, while Virginia’s 1788 application referenced “the defects of this constitution that have been suggested by the State Conventions,” it merely asked that the convention have “full power to take into their consideration” those defects, and that it “report such amendments . . . as they shall find best suited to promote our common interests, and secure to ourselves and our latest prosperity the great and inalienable rights of mankind.”[50] In other words, the delegates would have no obligation, by the terms of their appointment, to consider or avoid any topic whatsoever. New York’s 1789 application was broader in form (asking that the delegates have “full powers to take the said Constitution into their consideration”),[51] but it was identical in substance: the duties of a conscientious delegate would be precisely the same regardless of whether Congress’s resolution ended up using Virginia’s wording or New York’s. On this approach, these two applications are substantively the same, and both could be counted toward the two-thirds requirement for a plenary convention.

The next set of applications a few decades later, however, imposed a different set of obligations. During the tariff fight of the 1830s, South Carolina asked “that a Convention of the States be called as early as practicable, to consider and determine such questions of disputed power as have arisen between the States of this confederacy and the General Government.”[52] This application may seem quite broad. But it meant, for example, that delegates who attended and inquired only as to amendments about congressional procedure or executive privilege would have failed in their charge. If one third of states had submitted South Carolina’s application, and another third had submitted New York’s, Congress in framing its own resolution would have had no idea what to tell the delegates they must do.

On this argument, South Carolina’s application wasn’t really plenary and unlimited (and couldn’t have been aggregated with applications that are), because it added legal obligations that the delegates wouldn’t otherwise have had. These legal obligations might not have been enforceable in court, of course, just like a juror’s obligation to “well and truly try” the case;[53] but they were there nonetheless—and were distinct from the political obligations created, for example, if most of the applying states had added separate language, distinct from the application itself, to “earnestly request” that a particular subject be considered or avoided. The legal obligation was one thing, the political obligation another.

If that’s right, then an application like South Carolina’s couldn’t have been aggregated with Georgia’s version, calling for a convention “to amend the constitution aforesaid, in the particulars herein enumerated, and in such others as the people of the other States may deem needful of amendment,”[54] and especially with Alabama’s, which sought a convention “to propose such amendments . . . as may seem necessary and proper to restrain the Congress of the United States from exerting the taxing power for the substantive protection of domestic manufactures.”[55] Clearly Alabama, Georgia, and South Carolina were all addressing the same topic, but Congress’s choice among them would have given the delegates different obligations—and the whole point of the application process is to take out of Congress’s hands the choice of what the convention delegates are to do.[56]

This interpretive distinction makes a real difference. The vast majority of convention applications to date have been limited in one way or another, so whether the two-thirds requirement has already been met depends on which can be aggregated with which. To Michael Stokes Paulsen, for example, an application with words of limitation (like “sole and exclusive purpose”) can’t be aggregated with one without, but an application for a “convention ‘to consider’ or ‘for the purpose of proposing’” a particular amendment may be aggregated with one specifying another amendment, or even with a plenary application; the latter form of words, he argues, doesn’t make its purpose “a condition of the application.”[57] With the category of plenary applications thus broadened, Paulsen concludes that the relevant number was reached several decades ago, and that Congress was obliged to call a plenary convention[58]—though more recent state rescissions again dropped the number below the two-thirds threshold.[59] Yet if Congress’s calling a convention “to consider a Balanced Budget Amendment,” a convention “to consider a Gun Control Amendment,” or a convention simply “for proposing amendments” gives different sets of obligations to conscientious delegates, then Congress in framing the convention has to choose one—and it can’t aggregate across different topics that the delegates might be obliged to consider.

2.2  Constitutional validity

The main dispute over limited conventions—and one of very long standing—is whether they are constitutional at all. Does “a Convention for proposing Amendments” mean “a convention for proposing such amendments as that convention decides to propose”?[60] Or may the state legislatures or Congress “limit what the convention may consider”?[61] A “Convention for proposing Amendments” might mean a convention whose charge is simply that of “proposing Amendments,” and which can’t be altered or limited by state applications; or it could mean any amending convention whose charge includes the proposing of at least one amendment of some kind.

The answer seems at best unclear. Although Michael Rappaport forcefully argues that the limited convention is affirmatively authorized, as Article V’s language is “broad enough” to support a limited convention,[62] the language might also be broad enough to support either reading. As in Chiafalo v. Washington, in which the Supreme Court considered whether Article II’s “Electors” were to be free from outside compulsion or subject to state law,[63] whether Article V’s “Conventions for proposing Amendments” were to have a plenary charge or to be constrained by state applications is fundamentally about contextual understandings rather than text alone. If, for example, Georgia’s 1777 constitution specifically allowed counties and the state assembly to “specify[] the alterations to be made,”[64] does that show that limited conventions were generally thought permissible, or that special language was required to implement them?

Other sources are rendered somewhat ambiguous by the difficulty of distinguishing political expectations and constraints from legal ones. When George Washington wrote that the “constitutional door is open for such amendments as shall be thought necessary by nine States,”[65] was he referring to those states’ mere ability to seek a convention (which would surely include debate on well-known and popular proposals, by delegates specifically chosen for that purpose by the states themselves), or suggesting a power in those states to limit a convention to their desired changes? When Tench Coxe argued that, “[i]f two thirds of those legislatures require it, Congress must call a general convention, even though they dislike the proposed amendments,”[66] was he speaking of amendments which the convention must consider (and consider only), or of amendments which it almost certainly would consider, given the controversies underlying the call? Or when George Nicholas argued in the Virginia ratifying convention that “[t]he conventions which shall be so called will have their deliberations confined to a few points,”[67] was that because they would have been legally so confined by Congress, or because the impetus for amendment would necessarily have focused on some issues to the exclusion of others? And so on.

Part of the problem in reconstructing the context is that a plenary convention, while foreign to us, was highly present to the minds of the Founding generation, some of whom wanted a full re-do of the Philadelphia Convention—including the legislatures in New York and Virginia. George Mason, wishing for fewer federal powers, suggested that if certain points were “improperly settled, his wish would then be to bring the whole subject before another general Convention.”[68] And as noted above, Gouverneur Morris, who wished for more, “had long wished for another Convention, that will have the firmness to provide a vigorous Government, which we are afraid to do.”[69] Once the Constitution was ratified, moreover, Madison worried that any convention would “spread a general alarm, and be but too likely to turn everything into confusion and uncertainty.”[70] So while a limited convention was surely conceivable to the Founders, it’s possible that the plain language of “a Convention for proposing Amendments” more naturally suggested a plenary one.

Finally, there’s the possibility of a dog that didn’t bark. In the lead-up to the Eleventh Amendment, numerous state legislative bodies and executives pressed their congressional delegations to propose an amendment in Congress—even recruiting their fellow state legislatures to the effort. But none of them appear to have applied (or, it seems, even discussed applying) for a limited Article V convention on the topic, which is something they could’ve done by themselves.[71] Perhaps they didn’t think of it, or they didn’t think it necessary, given the likely votes in Congress; perhaps they would have pursued a convention had the (faster) congressional process failed.[72] But perhaps also they thought that any such convention would have to have a plenary charge, and so would be an inappropriate means for pursuing such a limited end.

2.3  Enforcement

Assuming a limited convention is constitutional, how is it to be enforced? Whatever the nature of the charge or the obligation of a delegate on the front end, if a “runaway” convention goes beyond that charge and its delegates disobey their instructions, what status do their proposed amendments have on the back end? Article V speaks of Congress “call[ing] a Convention for proposing Amendments, which . . . shall be valid to all Intents and Purposes, as Part of this Constitution, when ratified” by three fourths of states.[73] Suppose that the states apply for, and Congress calls, a convention “for proposing amendments to limit the budget deficit”—but that the convention’s proposal, in addition to a cap on annual spending-above-revenue, also includes a line-item veto. Assuming that this veto provision is too distantly related to fall within the charge (itself a potential subject of controversy), would the resulting amendment, or perhaps just that section of the amendment, therefore not be valid, even when ratified by three fourths of the states?

Again, the Founding-era evidence is less than clear. As James Iredell put it in the North Carolina ratification debates,

Any amendments which either Congress shall propose, or which shall be proposed by such general convention, are afterwards to be submitted to the legislatures of the different states, or conventions called for that purpose, . . . and, upon the ratification of three fourths of the states, will become a part of the Constitution.[74]

The word “general” in “general convention” might refer to geography rather than topic;[75] but we still have to ask whether “[a]ny” means any, or whether some convention-proposed amendments will nonetheless fail to become valid “upon the[ir] ratification.”

And, again, the reason for the uncertainty is in part the fine distinction between political and legal expectations and constraints. For example, Magliocca invokes the Philadelphia Convention as an example of a runaway convention. The “Framers famously went beyond their commissions” by proposing a document with a nine-out-of-thirteen ratification rule, which no one could pretend was simply amending the Articles.[76] Perhaps Philadelphia serves as precedent “that a national convention is free to propose anything, including a new document and ratification process”—say, by “national referendum.”[77]

Yet this claim of precedent elides the distinction between a convention’s changing the law, within an existing legal system, and its changing the legal system itself. The Constitution’s adoption changed our legal system, just as the American Revolution surely had; declaring the thirteen colonies to be “Free and Independent States,”[78] British law notwithstanding, was a departure from prior law rather than an application of it. Yet the Articles of Confederation had changed the law within an existing legal system, employing those thirteen independent states’ lawful “Power to . . . contract Alliances”[79]; and the same is true of all of the Article V amendments (even significant ones, like the Reconstruction Amendments), which employed preexisting legal powers toward new ends.[80] Article VII’s departure from the Articles of Confederation, on the other hand, would have been invalid under preexisting law, and it was recognized as such at the time: Madison insisted that “such institutions must be sacrificed,” in cases of “absolute necessity,” to “self-preservation” and “the safety and happiness of society”—only then adding that “Perhaps” the Articles’ violation by some states rendered it voidable by others.[81] Like the Revolution, Article VII can be cited as a political precedent for departing from a prior system when necessary, for a political “Right of the People to alter or abolish” their “Form of Government” and “to institute new Government” in its place.[82] But what it doesn’t establish is a legal precedent that such an alteration or abolition would be legally valid within our current legal system: indeed, Article V says outright what the appropriate ratification method must be, namely legislatures or conventions in three fourths of the states.

Of course, none of this means that a departure from prior law wouldn’t work.[83] The Revolution sure did. And if a runaway amendment were genuinely popular and actually ratified by three fourths of state legislatures, or if its new ratification-by-referendum procedure were actually carried out with success, then it’s entirely possible that the new constitutional order would attract more loyalty from more people (or from people with more guns) than the old. A national convention, even one called under the clearest limits and with delegates bound by the most fearsome oaths, can only be called “limited” or “plenary” in retrospect: in such cases “all that succeeds is success.”[84] And this may be the strongest argument against the convention method: that bottled lightning, like a bottle of aerated water,[85] sometimes explodes.

3.  The Rolling Convention

So what is to be done? The congressional-proposal channel is running shallow; the convention channel is even drier and full of shoals. In such a case another channel might be dug. Madison’s idea at Philadelphia was to have the states call, not for conventions, but for specific amendment text. A version of this proposal might achieve part of the Founders’ goals: it would let the states be the first movers in proposing amendments, even if it didn’t wholly take ordinary lawmakers out of the process of extraordinary lawmaking. A new amendment could create a third channel of constitutional change, a “rolling” convention that flips the order of Article V: it would let three fourths of states propose amendments, whether by legislature or convention, and then have two thirds of each House of Congress ratify them.

The goal of this amendment wouldn’t be to change the threshold for amendment. The two-thirds and three-fourths requirements would be left precisely intact. Instead, it would change the order in which they apply, allowing state legislatures to develop and debate amendment proposals with real legal effect. Requiring that two thirds of each House of Congress agree on anything seems fanciful in this polarized age; but requiring that they reach this level of agreement before an amendment can even get started might explain why such amendments are so few and far between. On the other hand, there may well be proposals with broad popular support that could find their way through a substantial number of state legislatures, putting them on the legislative (and, indeed, congressional) agenda with demonstrated bipartisan appeal.[86]

It’s conceivable that something like this method could be achieved without a new amendment. Rappaport argues, for example, that “if the great majority of the states agreed on a single proposal when applying for a convention,” they could increase the pressure on the convention delegates to return that and only that proposal—and thus increase the likelihood that other states would be willing to join the convention call.[87] But the use of a national convention always opens the door for something odd to happen between the application, proposal, and ratification stages, which might be enough to scare off all but the most committed states.

Moreover, while part of Madison’s vision had been to cut Congress out of the amendment loop, that project might be less attractive to many today. First, population changes mean that having two thirds or even three fourths of the states on your side means less than it used to. In the 2020 Census, the twelve largest states accounted for roughly 59% of the population of all fifty.[88] If two thirds of states may call for a convention and then three fourths may ratify its products, it’s theoretically possible—though politically quite infeasible—for the fewer than half of Americans who live in the smallest thirty-eight states to propose and ratify an amendment opposed by the majority of Americans who live in the largest twelve. Any states-only amendment process that lets a minority of Americans change the Constitution against the majority’s wishes will encounter serious resistance. Requiring ratification by two thirds of the House and Senate, by contrast, means that a minority-rule amendment has no chance of passing.

Second, a states-only amendment process might face a real deliberation deficit. In the same way that ballot initiatives sometimes produce worse governance than ongoing legislatures (producing decisions that might make sense one-by-one but are inconsistent when combined), constitutional amendments that sound good in isolation to individual state legislatures might pose serious problems when assembled together. In a world of state-led and state-ratified amendments, no one legislature voting on a proposal would face all of its burdens and benefits at once. That’s one argument for an actual convention, in which state-selected delegates would at least meet and discuss issues as a body; in the absence of such a convention, it’s an argument for keeping some national institution involved that bears responsibility for the whole. Even though Representatives and Senators all have their own districts or constituencies, some process of nationwide debate and revision might be necessary before the constitutional text is changed. And because Congress can propose amendments too, requiring congressional ratification gives Congress the chance to edit out any flaws in a state-drafted amendment and to send out a new version as its own proposal—something that a state-led and state-ratified process might never do. If bad amendments can do more damage than good ones can easily repair, we should want to keep the threshold for fundamental change rather high, and we should want every opportunity for correcting errors before they make it into the Constitution’s text.

Preserving the two-thirds-of-each-House requirement would, it’s true, make it harder to pass amendments that might reduce congressional power (whether by balancing the budget, imposing term limits, or constricting current interpretations of Article I, Section 8). To some, that’s the main benefit of the convention process. But that channel would still exist; the rolling convention needn’t displace the actual convention, which can still go over Congress’s head when the issues are serious and the risk of a runaway convention worth running. But a “rolling” convention would achieve many of the benefits of that process without running the risks: by going to the state legislatures first, it could produce clear indicators of popular support even for an amendment that Congress wouldn’t have come up with on its own. If the absence of a real constitutional amendment process is doing real damage to the Republic—and the current state of constitutional debate suggests that it may be—a new channel of constitutional change may be worth digging.

Conclusion

Magliocca is right that “the lack of a second national convention after 1787 does not mean that the state summoning power is irrelevant.”[89] The very possibility of a convention may give Congress a reason to act: once the state petitions are pouring in, as was the case during the run-up to the Seventeenth Amendment,[90] Congress may well choose to preempt any risk that we’ll reach the two-thirds threshold. But the same uncertainties of the convention process that make it a useful cudgel against Congress also make it a dangerous tool if used on purpose. When convention applications are difficult to read on their own, and when serious constitutional questions dog their validity and eventual enforcement, only the strongest reasons could support calling a true national convention. The goal of reviving the amendment process should be reform, not revolution. We should restore the convention process, but we should do it one amendment at a time.

 _______________________________

[1]     John Harrison, The Lawfulness of the Reconstruction Amendments, 68 U. Chi. L. Rev. 375, 375 (2001).

[2]     U.S. Const. art. V.

[3]     Id.

[4]     S.J. Res. 211, 47 Stat. 1625 (1933).

[5]     See Gerard N. Magliocca, Article V Constitutional Conventions 9–19.

[6]     See David M. Potter, The Impending Crisis: America Before the Civil War: 1848–1861, at 546–47 (Don E. Fehrenbacher ed., 1976).

[7]     Compare U.S. Const. amend. XXVI (proposed in 1971, see S.J. Res. 7, 85 Stat. 825 (1971), and ratified later that year), with U.S. Const. amend. XXVII (proposed in 1789, see 1 Stat. 97, 97 (1789), and ratified in 1992).

[8]     See, e.g., Stephen E. Sachs, Let States Get the Ball Rolling on Amendments, Boston Globe (Dec. 15, 2021), https://apps.bostonglobe.com/ideas/graphics/2021/12/editing-the-constitution/let-states-get-the-ball-rolling-on-amendments; Stephen E. Sachs, Amendments Should Start with States, Harv. Gazette (Dec. 5, 2022), https://news.harvard.edu/gazette/story/2022/12/amendments-should-start-with-states-stephen-sachs/.

[9]     32 J. Continental Cong. 74 (1787).

[10]    See Articles of Confederation of 1781 art. XIII, cl. 1–2.

[11]    The Federalist No. 40, at 265 (James Madison) (Jacob E. Cooke ed., 1961).

[12]    Akhil Reed Amar, America’s Constitution: A Biography 7 (2005).

[13]    See U.S. Const. art. I, § 2, cl. 1.

[14]    Amar, supra note 12, at 309.

[15]    1 The Records of the Federal Convention of 1787, at 22 (Max Farrand ed., 1911) [hereinafter Farrand].

[16]    Id. at 122.

[17]    Id. at 122, 126.

[18]    Id. at 126.

[19]    Id. at 127; see also 2 id. at 92 (describing Rufus King’s call for “a reference to the authority of the people expressly delegated to Conventions, as the most certain means of obviating all disputes & doubts concerning the legitimacy of the new Constitution”).

[20]    2 id. at 665; see Amar, supra note 12, at 308 n.*.

[21]    Amar, supra note 12, at 7.

[22]    1 Farrand, supra note 15, at 22.

[23]    Id. at 203.

[24]    2 id. at 159.

[25]    Id. at 557–58.

[26]    Id. at 558.

[27]    Id.

[28]    Id. at 558–59.

[29]    Id. at 559.

[30]    See id. (“The Legislature of the U— S— whenever two thirds of both Houses shall deem necessary, or on the application of two thirds of the Legislatures of the several States, shall propose amendments to this Constitution, which shall be valid to all intents and purposes as part thereof, when the same shall have been ratified by three fourths at least of the Legislatures of the several States, or by Conventions in three fourths thereof, as one or the other mode of ratification may be proposed by the Legislature of the U.S[.]”).

[31]    Id. at 629.

[32]    Id. at 629–30.

[33]    Id. at 479.

[34]    Id. at 630.

[35]    Id. at 559, 630.

[36]    Magliocca, supra note 5, at 7 & n.10 (comparing the ratification threshold to the two-thirds threshold for expelling members of Congress, convicting after impeachment, ratifying treaties, overriding vetos, and so on).

[37]    2 Farrand, supra note 15, at 629.

[38]    Magliocca, supra note 5, at 19, 21.

[39]    See 1 Annals of Cong. 258–59 (1789) (Joseph Gales ed., 1834) [hereinafter Virginia Application]; H.R. J., 1st Cong., 1st Sess. 29–30 (1789) [hereinafter New York Application].

[40]    See The Article V Library, Article V Convention Application Analysis, http://article5library.org/analyze.php?topic=Plenary&res=1&gen=1&ylimit=0 (last visited Aug. 3, 2025) [hereinafter Plenary Applications]; infra text accompanying notes 52–57 (reclassifying some applications).

[41]    See Plenary Applications, supra note 40.

[42]    See Magliocca, supra note 5, at 13–14.

[43]    See 2 Farrand, supra note 15, at 630 (noting Madison’s concern “that difficulties might arise as to the form, the quorum &c. which in Constitutional regulations ought to be as much as possible avoided”).

[44]    See Magliocca, supra note 5, at 10, 12, 16, 18.

[45]    Convention of States Action, https://conventionofstates.com/files/model-convention-of-states-application/download (last visited Aug. 3, 2025).

[46]    See Robert G. Natelson, Counting to Two Thirds: How Close Are We to a Convention for Proposing Amendments to the Constitution?, 19 Fed. Soc. Rev. 50 (2018).

[47]    See Michael Stokes Paulsen, A General Theory of Article V: The Constitutional Lessons of the Twenty-Seventh Amendment, 103 Yale L.J. 677, 746 (1993).

[48]    Magliocca, supra note 5, at 10 n.17.

[49]    U.S. Const. art. V.

[50]    Virginia Application, supra note 39, at 259–60.

[51]    New York Application, supra note 39, at 30.

[52]    H.R. J., 22d Cong., 2d Sess. 220 (1833).

[53]    See, e.g., United States v. Green, 556 F.2d 71, 71 n.1 (D.C. Cir. 1977) (describing the then-applicable oath) (citation and internal quotation marks omitted).

[54]    S.J., 22d Cong., 2d Sess. 66 (1833).

[55]    H.R. J., 22d Cong., 2d Sess. 362 (1833).

[56]    The Federalist No. 85, at 593 (Alexander Hamilton) (Jacob E. Cooke ed., 1961) (“[T]he national rulers . . . will have no option upon the subject. . . . The words of this article are peremptory. The congress ‘shall call a convention.’ Nothing in this particular is left to the discretion of that body.”).

[57]    Paulsen, supra note 47, at 744–45 (emphasis omitted).

[58]    Id. at 756.

[59]    Michael Stokes Paulsen, How to Count to Thirty-Four: The Constitutional Case for a Constitutional Convention, 34 Harv. J.L. & Pub. Pol’y 837, 856–58 (2011).

[60]    Paulsen, supra note 47, at 738 (internal quotation marks omitted) (quoting Charles L. Black, Jr., Amending the Constitution: A Letter to a Congressman, 82 Yale L.J. 189, 199 (1972)).

[61]    Michael B. Rappaport, The Constitutionality of a Limited Convention: An Originalist Analysis, 81 Const. Comment. 53, 63 (2012).

[62]    Id. at 69.

[63]    591 U.S. 578 (2020).

[64]    Ga. Const. of 1777, art. LXIII (cited in Rappaport, supra note 61, at 68).

[65]    Letter from George Washington to John Armstrong (Apr. 25, 1788), in 6 The Papers of George Washington: Confederation Series 224, 226 (W.W. Abbot ed., 1997).

[66]    [Tench Coxe], A Pennsylvanian to the New York Convention, Pa. Gazette, June 11, 1788, reprinted in 20 Documentary History of the Ratification of the Constitution 1139, 1142–43 (John P. Kaminski et al. eds., 2004).

[67]    3 The Debates in the Several State Conventions on the Adoption of the Federal Constitution 102 (Jonathan Elliot ed., Phila., J.B. Lippincott Co. 2d. ed. 1836) [hereinafter Elliot’s Debates].

[68]    2 Farrand, supra note 15, at 479.

[69]    Id.

[70]    Paulsen, supra note 47, at 741 n.215 (quoting Letter from James Madison to George Eve (Jan. 2, 1789), in 5 The Writings of James Madison 321 (Gaillard Hunt ed., 1904)).

[71]    See 5 Documentary History of the Supreme Court of the United States, 1789–1800, at 134–35, 162, 237 (Maeva Marcus ed., 1994) (Georgia); id. at 338–39 (Virginia); id. at 440 (Massachusetts); id. at 609 (Connecticut); id. at 611 (South Carolina); id. at 611–12 (Maryland); id. at 612–13 (Pennsylvania); id. at 615 (North Carolina); id. at 618 (New Hampshire).

[72]    See Michael B. Rappaport, Reforming Article V: The Problems Created by the National Convention Amendment Method and How To Fix Them, 96 Va. L. Rev. 1509, 1543 (2010) (suggesting that they might).

[73]    U.S. Const. art. V.

[74]    4 Elliot’s Debates, supra note 66, at 177.

[75]    See Rappaport, supra note 61, at 87 n.78.

[76]    Magliocca, supra note 5, at 4.

[77]    Id. at 5–6.

[78]    The Declaration of Independence para. 32 (U.S. 1776).

[79]    Id.

[80]    See generally Harrison, supra note 1.

[81]    The Federalist No. 43, at 297 (James Madison) (Jacob E. Cooke ed., 1961).

[82]    The Declaration of Independence para. 2 (U.S. 1776).

[83]    See Stephen E. Sachs, Originalism as a Theory of Legal Change, 38 Harv. J.L. & Pub. Pol’y 817, 843 (2015).

[84]    H.L.A. Hart, The Concept of Law 153 (3d ed. 2012).

[85]    See MacPherson v. Buick Motor Co., 217 N.Y. 382, 387 (1916) (citing Torgesen v. Schultz, 192 N.Y. 156 (1908)).

[86]    For further discussion, see sources cited supra note 8; for potential amendment language, see Stephen E. Sachs (@stephenesachs), X (Aug. 4, 2025, 9:29 a.m.), https://x.com/StephenESachs/status/1952361368197190001.

[87]    See Rappaport, supra note 72, at 1515.

[88]    See U.S. Census Bureau, Change in Resident Population of the 50 States, the District of Columbia, and Puerto Rico: 1910 to 2020, https://www2.census.gov/programs-surveys/decennial/2020/data/apportionment/population-change-data-table.pdf (last visited Aug. 4, 2025); List of U.S. States and Territories by Population, Wikipedia, https://en.wikipedia.org/wiki/List_of_U.S._states_and_territories_by_population (last visited August 3, 2025) (organizing the data from largest to smallest).

[89]    See Magliocca, supra note 5, at 9.

[90]    See id. at 10.