The Convention Method for Proposing Amendments: Essential, Misunderstood, and Broken
Michael B. Rappaport believes there are reforms of the convention method that would avoid the possibility of a runaway convention and provide a workable alternative, but they face considerable obstacles. Rappaport is Hugh and Hazel Darling Foundation Professor of Law, University of San Diego; Director, Center for the Study of Constitutional Originalism; 2025-2026 University Professor.
Read Full Biography
Introduction
One of the potentially most important yet largely overlooked provisions of the original Constitution is the convention method for proposing constitutional amendments. Under this provision, two thirds of the state legislatures can apply to Congress to call a convention for proposing constitutional amendments.[1] The more well-known method for amending the Constitution requires two thirds of each house of Congress to propose amendments. But what if an amendment is needed to reduce or reform Congress’s powers? Congress is unlikely to pass such an amendment, so the convention method is the only route for bypassing this congressional veto.
Unfortunately, the convention method is broken. It does not function. A convention for proposing amendments has never been called. The main reason appears to be that the state legislatures fear the possibility of a “runaway convention” – one that proposes amendments beyond the scope authorized by Congress. As a result, Congress effectively holds a de facto veto over constitutional amendments. That veto may have blocked many popular changes from becoming amendments, such as those requiring a balanced budget, congressional term limits, or a line-item veto.
Significantly, there are reforms of the convention method that would avoid the possibility of a runaway convention and provide a workable alternative to the congressional amendment method. But enacting a constitutional amendment to implement those reforms is difficult because Congress may be reluctant to surrender its veto and the convention method is burdened by the state legislatures’ fear of a runaway convention. Ultimately, reforming the convention method may be possible but only if there is strong popular support that overcomes these obstacles.
This essay explores these themes. It begins by responding to Gerard Magliocca’s essay Article V Constitutional Conventions.[2] It also argues that runaway conventions are properly viewed as unconstitutional, even though that unconstitutionality has not quelled the fear that has crippled the convention process.
Magliocca’s Essay
Gerard Magliocca’s essay Article V Constitutional Conventions provides a good, concise, and accessible introduction to the history of various aspects of constitutional conventions. It traces the origins of conventions in England, discusses the Philadelphia Convention that drafted the United States Constitution, and examines the state conventions that ratified it. The essay also explores instances in which states have applied to Congress, asking it to call a convention for proposing amendments. While state applications have never reached the two thirds threshold that would require Congress to call a convention, Magliocca notes that the threat of such a convention may have prompted Congress to propose amendments on its own initiative. Finally, he discusses the only instance in which state conventions were used to ratify a constitutional amendment – the 21st Amendment, which repealed Prohibition – and considers why that method may have been chosen.
Despite its evident virtues, I do have some disagreements with Magliocca’s essay. First, in the very first line of the essay, he writes “Article V lays out three paths for changing the Constitution.”[3] But Article V actually lays out four paths:[4] (1) Congress can propose an amendment that is then ratified by the state legislatures; (2) Congress can propose an amendment that is then ratified by state conventions; (3) Two thirds of the state legislatures can apply for a proposing convention, which then proposes an amendment, that is then ratified by state legislatures; and (4) Two thirds of the state legislatures can apply for a proposing convention, which then proposes an amendment, that is then ratified by state conventions.
Magliocca does not distinguish between the third and fourth of these methods. But their differences are important. The third allows for a convention to propose and state legislatures to ratify, while the fourth allows for a convention to propose and conventions to ratify, taking legislatures entirely out of the process, especially when such legislatures might block a necessary amendment. Thus, Article V is an elegant provision that permits constitutional amendments to be passed with the participation of the existing legislatures (Congress and/or the state legislatures) or without their participation (the proposing convention and/or the ratification conventions).
I also disagree with Magliocca’s suggestion that the Philadelphia Framers were not very concerned “about the process for future changes” in the Constitution through amendment. Magliocca says that the “Framers’ choice of a three-fourths requirement for state ratification of amendments stands out as inexplicable.” He notes that “[o]nly two thirds of the states (nine out of thirteen) were required to ratify the Constitution. Why, then, did the Framers make it harder to ratify amendments? The simplest explanation is that they just did not give this issue careful consideration.”[5]
But there is a straightforward explanation for why three fourths of the states were required to ratify constitutional amendments but only nine out of thirteen to ratify the Constitution itself.[6] At the time, Rhode Island had been boycotting the Philadelphia Convention, and it seemed virtually certain not to ratify it, at least initially. (This prediction turned out to be accurate, as Rhode Island refused to hold a ratification convention and only ratified the Constitution when threats by the other states forced its hand.) When the Constitution was proposed, the nine states needed to ratify it represented three fourths of the twelve participating states. Thus, the three quarters requirement was the rule that was applied to the states which were actively participating in the decision whether to adopt the Constitution.
But my main disagreements with Magliocca concern his discussion and apparent belief that the Constitution allows runaway conventions – conventions that either exceed the limits of their authority as expressed in Congress’s call for a convention or that propose ratification methods for their recommended amendment that depart from the constitutional requirements. While Magliocca does not explicitly argue that runaway conventions are constitutional, he does claim that the history of the Philadelphia Convention supports their constitutionality, and the overall tenor of his discussion suggests acceptance of such conventions and praise for their role in inducing Congress to propose amendments.
Here, I disagree with Magliocca, arguing that limited conventions are essential and that runaway conventions are best viewed as unconstitutional. I then reject the claim that the runaway Philadelphia Convention supports the constitutionality of runaway conventions generally.
The Central Role of the Convention Method
While normally treated as an afterthought, the convention method for proposing amendments to the Constitution – under which two thirds of the state legislatures can apply to Congress to call a convention to propose constitutional amendments – is actually an essential feature of the constitutional amendment process established by the Constitution.
When the constitutional amendment process was being debated, an earlier provision was rejected because it would have only allowed Congress to propose amendments. As George Mason said at the Convention, “It would be improper to require the consent of the Natl. Legislature [for an amendment], because they may abuse their power, and refuse their consent on that very account.”[7] The Framers recognized that a constitution with a congressional veto on amendments would become distorted over time. Such a veto would block necessary reforms of the national government, particularly of Congress itself.
While the Framers employed the convention method to avoid a congressional veto, unfortunately, that method is broken. It has never been used to call a convention. The main reason, at least in modern times, appears to be the fear of a runaway convention.[8] Suppose that legislatures in two thirds of the states apply to Congress to call a convention on a specific subject, such as proposing a balanced budget amendment. The convention meets but instead of proposing a balanced budget amendment, it proposes an amendment allowing prayer in public schools or permitting governments to practice reverse discrimination. Perhaps the proposed amendment also states that it will become part of the Constitution when ratified by a majority of the states rather than the three fourths that the Constitution requires.
From the perspective of those seeking a balanced budget amendment, the convention might be a disaster. There is now a risk that the new amendment – which is strongly disliked by proponents of the balanced budget amendment – could become part of the Constitution. Even if they are able to defeat the proposed amendment, doing so would require a significant amount of effort. Moreover, if the proposed amendment includes ratification by a majority, rather than three-fourths, of the states, there is an even greater risk it will become part of the Constitution. Further, the constitutional questions raised by this runaway convention and proposed amendment may create a constitutional crisis and would at a minimum be extremely divisive.
The possibility of such a runaway convention leads people who would otherwise support a balanced budget amendment to decline to apply to Congress to call a convention to propose that amendment. Thus, while the convention method is formally available to amend the Constitution, it is effectively nonfunctional.
The ineffectual convention method appears to have had serious effects on our Constitution. In the last few generations, there have been several popular movements to amend the Constitution that would reduce Congress’s power, including proposals for a balanced budget amendment, a congressional term limits amendment, and a line-item veto amendment.[9] Unsurprisingly, Congress has been unwilling to propose these amendments which would restrict its power. But since the convention method does not function, these amendments could not be enacted by bypassing Congress. While there is no way to know whether these amendments would have been enacted under a functioning convention method, their popularity suggests that they might have been.
Thus, the failure of the convention method may have been extremely consequential. It may have blocked popular reforms that might have been desirable for the country. And if the failure of the convention method blocked new limits on the national government, that failure may have led to a less balanced polity, artificially making it more nationalist and less federalist.
One objection to this analysis derives from Magliocca’s argument that the threat of a runaway convention may have led Congress to propose certain constitutional amendments, such as the Bill of Rights or the Seventeenth Amendment. If that threat were a significant concern for Congress, the possibility of a runaway convention would not defeat amendments that Congress opposes but instead serve as part of the method for inducing Congress to propose amendments it would otherwise oppose.
The problem is that there is little reason to believe that this threat actually has a strong effect. It is true, as Magliocca mentions, that the threat may have induced Congress to propose the Bill of Rights. But that involved unique circumstances. The Constitution had only been in place for a brief time. The Federalist advocates of the Constitution had promised to propose a Bill in order to secure ratification, but then Congress dragged its heels. Opponents of the Constitution in the state legislatures would have been anxious to hold a new convention. But once the Constitution became popular over time, people have been less cavalier about risking drastic change. Magliocca cites only one other amendment—the Seventeenth—as potentially driven by the convention threat, but he concedes that the evidence for a causal link is inconclusive.
By contrast, the significant threat of a convention has not induced Congress to pass other amendments that had a significant number of applications from state legislatures for a convention, such as the balanced budget amendment or an amendment partially overturning the Reynolds v. Sims apportionment decision.[10] The ineffectiveness of the threat of a runaway convention is not hard to explain. The threat of a convention must be a credible one to induce Congress to act. But it appears that the state legislatures are too scared to actually call a convention and therefore are unwilling to call Congress’s bluff.[11] Thus, the threat of a convention in modern America has not induced Congress to propose amendments it otherwise opposes and is unlikely to do so.
The Constitution Allows Limited Conventions
What is so unfortunate is that the failure of the convention method, due to the fear of a runaway convention, is to a significant extent the result of a failure to discern the Constitution’s original meaning. In my view, the Constitution pretty clearly indicates that limited conventions are constitutional under the original meaning and need to be respected. Therefore, runaway conventions should be viewed as unconstitutional and their products should be treated as illegal nullities.
A limited convention is a convention that is restricted to deciding whether to propose an amendment on a specific subject, such as a balanced budget amendment. The language of the Constitution strongly implies that limited conventions are constitutional. The Constitution says that Congress “on the application of the legislatures of two thirds of the several states, shall call a convention for proposing amendments.” There is nothing in this language that suggests that a state legislature cannot submit an “application” for an amendment on a specific subject. If one asked whether a limited convention was a convention for proposing amendments, the answer would obviously be yes. A limited convention is precisely a convention that decides whether to propose a (specific) amendment.
I would go further and argue that an even more restricted type of limited convention is constitutional. Suppose that two thirds of the state legislatures apply for a convention limited to deciding whether to propose a precisely worded amendment that has been agreed upon by these legislatures ahead of time. Would it be constitutional to call a convention of this type? Again, the answer is yes. Even though its authority is limited to deciding whether to propose a specifically worded amendment, such a convention would still be a “convention for proposing amendments.”
Some commentators have argued that a convention must have broad or even unlimited discretion to decide what amendments to propose. But this argument is weak. There is nothing in the meaning of a convention that implies this type of discretion. Certainly, there were limited conventions prior to the enactment of the Constitution.[12] Moreover, the Georgia Constitution of 1777, one of the constitutional precursors to the United States Constitution, explicitly authorized limited conventions.[13] And once one allows proposing conventions to be subject to limits, it is not clear why they cannot be limited to the decision whether to propose a precisely worded amendment.
There is even some historical support for a convention limited to deciding whether to propose a specifically worded amendment. While I am not aware of a proposing convention that was limited to a specifically worded amendment – although it is possible the Georgia Constitution contemplated it – there were conventions that were so limited. The ratification conventions under the Constitution serve precisely the function of deciding whether to ratify a precisely worded amendment that has been proposed to them.
Given that a limited convention is constitutional, it would be unconstitutional for either Congress or the states to participate in the amendment produced by the runaway convention. The convention itself derives authority from the Constitution. Thus, if the convention violated the terms of the Constitution, then it would lack authority. Accordingly, the amendment it proposed could neither be sent to nor ratified by the states.
The easiest way to see this is to imagine a runaway convention that proposed an amendment requiring ratification by a majority of the state legislatures rather than the three fourths required by the Constitution. It would be unconstitutional for Congress to send this proposed amendment to the state legislatures for ratification, and it would be unconstitutional for the state legislatures to ratify it on that basis. After all, the Constitution clearly requires that three fourths of the states ratify an amendment.
If these actions would be unconstitutional, so would other attempts to ratify the products of a runaway convention, such as an amendment that exceeded the limits of Congress’s call for a convention. The actions of the convention would be ultra vires. Thus, if Congress called a limited convention and the convention ignored the limits, this would also be unconstitutional since the convention would have exceeded its authority.
If the states did purport to ratify an amendment produced by a runaway convention, then that amendment should not be recognized as part of the Constitution. While it is not entirely clear which entity would have the final say over the constitutionality of the amendment – the Supreme Court or Congress, if it is a political question – that entity should treat it as a nullity.
While I argue that the products of a runaway convention should be treated as a nullity, I am not claiming that whoever decides the question – the Supreme Court or Congress – would actually treat it as a nullity. In fact, if one were confident that the products of the runaway convention would be declared unconstitutional, then there would be much less reason for state legislatures to fear a runaway convention. My argument is only that the proper analysis of the original meaning would treat the products of a runaway convention as unconstitutional.
Unfortunately, there is a significant risk that either the Supreme Court or Congress would not follow the original meaning or would misapply it, especially if the runaway convention’s proposed amendments were ratified in the political process and therefore appeared to enjoy significant public support. Thus, the fear of a runaway convention is entirely rational, even though I believe such conventions are unconstitutional.
Magliocca’s Analysis of The Runaway Philadelphia Convention
While Magliocca does not discuss these issues, he does argue that the runaway Philadelphia Convention supports the idea that runaway conventions are constitutional. The Philadelphia Convention had been sought by the states and by Congress operating under the Articles of Confederation. The call of the Confederation Congress provided that a convention of delegates from the several states be held “for the sole and express purpose of revising the Articles of Confederation and reporting to Congress and the several legislatures such alterations and provisions therein as shall when agreed to in Congress and confirmed by the states render the federal constitution adequate to the exigencies of Government & the preservation of the Union.”[14] The last part of the call was a reference to the provisions of the Articles, which provided that amendments must “be agreed to in a congress of the united states, and be afterwards confirmed by the legislatures of every state.”[15]
The proposed Constitution produced by the Philadelphia Convention, however, did not conform to this call. It replaced the Articles rather than revised them, and it called for ratification by nine of the thirteen states by conventions rather than by all thirteen state legislatures.
Magliocca writes:
“While state petitions under Article V often restrict their convention call to a particular topic, the Philadelphia Convention can be cited as authority for the proposition that a national convention is free to propose anything, including a new document and ratification process. . . . The only limit on a national convention that can be inferred from Philadelphia is that the proposal requires a separate ratification process. But the nature of that process could probably be changed from three-fourths of the state legislatures or state conventions to a simple majority by a national referendum given that Article VII changed the prior ratification process.[16]
Even if one ignores the arguments in the previous section in favor of a limited convention, I believe the suggestion that the runaway Philadelphia Convention supports the case in favor of the constitutionality of runaway conventions under the Constitution is weak.
To analyze this argument, it is useful to distinguish between two aspects of the Philadelphia Convention. First, instead of proposing a revision to the Articles, the convention proposed a replacement of the Articles through a different ratification method. Thus, the convention refused to follow the call from Congress.
While at first this might seem like a precedent for allowing departures from the call, that impression evaporates upon examination. The Articles did not provide for a convention. Instead, they provided that Congress should propose amendments. The Philadelphia Convention was an extraconstitutional meeting. It is best understood as an advisory convention – one that
merely offered advice to Congress. While the Philadelphia Convention did exceed its authority, it is not clear that its actions were illegal, since the convention was merely advisory. It had no binding or recognized role. And even if the convention’s proposal had in some sense been illegal, Congress was presumably free to ignore that illegality since it was acting on its own authority in adopting the proposal of the Convention.
Thus, the fact that the Philadelphia Convention ignored the call does not have any implications for an Article V proposing convention under the Constitution. The latter convention has a binding role and authority under the Constitution. When it violates that role and exceeds its authority, consequences follow.
Now, consider the second aspect involving the Philadelphia Convention –that the Confederation Congress sent the Philadelphia Convention’s proposal to the states. This aspect of the runaway convention is problematic. While Congress could lawfully choose to send a proposal from an advisory convention that exceeded its call, Congress could not lawfully send a proposal that violated the Articles’ requirements as to ratification. Thus, Congress acted unlawfully. This illegal action does not count as evidence that an Article V proposing convention can ignore the call. Instead, it counts as evidence that conventions and legislatures sometimes behave in illegal and unconstitutional ways.[17]
In sum, the runaway Philadelphia Convention is not evidence that it would be constitutional for an Article V proposing convention to exceed its call or to propose a ratification method that departs from the one specified in the Constitution. Instead, the Philadelphia Convention shows that conventions pose a risk of illegal action, and one must be very careful to address that risk.
What Is to Be Done?
Given the infirmities of the convention method, the Constitution has a significant defect that requires remedying. This raises two questions. What should replace the currently broken convention method and how can that replacement be enacted?
I recommend replacing the convention method with a process that authorizes two thirds of the state legislatures to propose an amendment by agreeing on the exact language that would become part of the Constitution. To allow the state legislatures to debate and deliberate on the substance of the proposed amendment, the Constitution would authorize states to convene an advisory convention. This convention would have no formal powers – it could not formally propose an amendment – but it would allow the states to determine what specific amendments might be desirable and would have support among the different states. Under this process of proposed amendments authored by state legislatures, as informed by an advisory convention, the problem of a runaway convention would be eliminated, because the state legislatures would be proposing a specifically worded amendment (and the advisory convention would have no power to formally propose an amendment).
While this reform would avoid a runaway convention, it would upset the balance of the Constitution because it would give the states a dominant role at both the proposing and ratification stages. Under the existing Article V, the nation is represented at the proposing stage either through a congressional proposal or a convention proposal. To ensure that national perspectives are represented, the reform could require that ratification of the amendment proposed by the state legislatures should occur through a national referendum requiring a supermajority of votes.
This reform leads to the second question: How can the reform be enacted? Congress is unlikely to propose it as a constitutional amendment, since the reform would eliminate its veto over amendments. And the reform cannot easily be passed through the existing convention method due to the fear of a runaway convention.
Yet, the cause is not hopeless. If strong popular and political support for a functional alternative to the convention method developed – especially bipartisan support, which is quite possible because people across the political spectrum favor significant changes to the Constitution – then major reform efforts might be undertaken.
First, state legislatures could apply for a limited convention – perhaps even one with a specifically worded amendment – that would consider whether to propose the reform. The state legislatures might also state in their applications that they regard a runaway convention as unconstitutional and that they would refuse to ratify the products of any such convention. The state legislatures might also state that delegates to the convention should be required to take an oath that they would not vote for any proposal that exceeds the limits on the convention.
Second, Congress might also take actions to guard against a runaway convention. It might pass legislation (or legislative rules) stating it would treat an application for a limited convention to be valid. And Congress might prohibit itself from sending to the states (or elsewhere) for ratification any convention proposal that either departs from the limits on the convention or the constitutionally prescribed ratification method. While a supermajority of each house of Congress might not support amendments that reformed the convention method, a mere majority in each house may be willing to pass legislation of this type.
Of course, even if Congress and the states adopted these safeguards, they could not completely eliminate the possibility of a runaway convention or the ratification of an amendment it might propose. But these safeguards might sufficiently reduce the risk so that state legislatures would be willing to apply for a limited convention. And given the great benefits that a reform of the convention method would produce, it would be a risk worth taking.
____________________________
[1] U.S. Const. Art. V.
[2] Gerard Magliocca, Article V Constitutional Conventions (2025).
[3] Magliocca at 1.
[4] Article V provides in pertinent part:
The Congress, whenever two thirds of both Houses shall deem it necessary, shall propose Amendments to this Constitution, or, on the Application of the Legislatures of two thirds of the several States, shall call a Convention for proposing Amendments, which, in either Case, shall be valid to all Intents and Purposes, as Part of this Constitution, when ratified by the Legislatures of three fourths of the several States, or by Conventions in three fourths thereof, as the one or the other Mode of Ratification may be proposed by the Congress.
[5] Magliocca at 8 (emphasis added).
[6] John O. McGinnis & Michael B. Rappaport, Our Supermajoritarian Constitution, 800 Tex. L. Rev. 703, 798-799 (2002).
[7] James Madison, The Debates in the Federal Convention of 1787, at 89-90 (Gaillard Hunt & James Brown Scott eds, (1920).
[8] There are additional reasons as well. One is that applications from different states may vary slightly in their language or scope, creating uncertainty over whether they pertain to the same convention. Another derives from the fact that Congress is likely the body responsible for making that determination. Since Congress will probably oppose a convention since it bypasses Congress’s veto on amendments under the congressional proposal method, Congress may apply a strict standard and use those differences as a basis to deny a convention call. 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, 1526-1528 (2010).
[9] Rappaport, Reforming Article V, at 1548-1551.
[10] See Magliocca at 14-15, 20-21.
[11] Congress might also have leverage in this game of chicken with state legislatures. Since Congress is likely the judge of whether two thirds of the states have applied for the same convention, it can risk the states appearing to reach the threshold, knowing that it can determine that all of the applications do not pertain to the same convention.
[12] See Michael B. Rappaport, The Constitutionality of a Limited Convention: An Originalist Analysis, 28 Const. Comm. 53, 68-69 (2012) (discussing evidence of limited conventions prior to the Philadelphia Convention, including the Annapolis Convention on commerce).
[13] See GA Const of 1777, art. LXIII (“No alteration shall be made in this constitution without petitions from a majority of the counties, and the petitions from each county to be signed by a majority of voters in each county within this State; at which time the assembly shall order a convention to be called for that purpose, specifying the alternations to be made, according to the petitions preferred to the assembly by the majority of the counties as aforesaid”).
[14] Report of Proceedings in Congress (Feb. 21, 1787) https://avalon.law.yale.edu/18th_century/const04.asp?utm_source=chatgpt.com.
[15] Articles of Confederation, Article XIII, at https://www.archives.gov/milestone-documents/articles-of-confederation.
[16] Magliocca at 6-7.
[17] Akhil Amar has argued that the Constitution’s ratification method did not actually violate the Articles of Confederation. He claims that the Articles were a treaty and that treaties were voidable if they were violated, as Amar argues the Articles were. Akhil Reed Amar, The Consent of the Governed: Constitutional Amendment Outside Article V, 94 Colum. L. Rev. 457 (1994). But the existing Constitution is not a treaty and thus is not voidable in the way that this argument asserts the Articles were. Therefore, this way of viewing the runaway Philadelphia Convention does not provide support for the constitutionality of a runaway convention under the Constitution.
