* Thurgood Marshall Professor of Constitutional Law, Harvard Law School. Email: vjackson@law.harvard.edu . With thanks to Richard Albert, Rosalind Dixon, Mark Tushnet, and Marge Baker for very helpful comments, and to Lauren Davis, Kate Epstein, and Noah Marks for helpful research assistance.
Search for other works by this author on:International Journal of Constitutional Law, Volume 13, Issue 3, July 2015, Pages 575–605, https://doi.org/10.1093/icon/mov050
14 November 2015Vicki C. Jackson, The (myth of un)amendability of the US Constitution and the democratic component of constitutionalism, International Journal of Constitutional Law, Volume 13, Issue 3, July 2015, Pages 575–605, https://doi.org/10.1093/icon/mov050
Navbar Search Filter Mobile Enter search term Search Navbar Search Filter Enter search term SearchThe article explores certain claims about the amendment process of the US Constitution. Empirically, it argues that the difficulty posed by formal procedures in obtaining textual amendments of the US Constitution is overstated, if one looks to history and to analogous constitutional requirements for override of presidential vetoes. It suggests that the low amendment rate may result not only from its formal procedures, and exaggerated estimates of their difficulty, but also from ideological or emotional opposition to amendment, as opposed to other methods of legal change. As a normative matter, it argues that, notwithstanding important arguments for caution in seeking constitutional amendment, a constitution that is truly not amendable by its own formal procedures, that relies on indefinitely long tenures for its highest court, and that is committed to judicial supremacy in constitutional interpretation, is in real tension with the democratic component of democratic constitutionalism.
Constitutions are sometimes thought of as embodying the “will” of a particular people. The United States Constitution is widely referred to for its founding declaration, “We the people . . .,” emblematic of a form of popular self-sovereignty. Yet in recent decades some scholars and political actors have come to believe that the US Constitution is no longer capable of being amended through the Article V amendment process. 1 It becomes difficult to see how popular consent can be invoked on behalf of a constitution’s legitimacy if, according to this “impossibility” view, the Constitution can no longer be amended by the organized political action of its constituent people. Other scholars and political actors argue against too ready resort to constitutional amendment, 2 and urge a presumption against formal amendment except in very extraordinary circumstances. We might call this the “textual veneration,” or constitutional reverence, view.
This article suggests that political and constitutional actors in the United States may be too reluctant to resort to popular democratic processes to effect formal constitutional change, and too reliant on seeking constitutional change through other means, especially adjudication and changes in judicial appointments. Reluctance to consider constitutional amendment through Article V procedures raises two different kinds of questions, one empirical, the other normative.
Section 2 discusses empirical claims about the relative difficulty of amending the US Constitution, challenging claims of impossibility by reference to the history of amendment and of congressional overrides of presidential vetoes under a similarly difficult super-majority voting rule. It suggests that sociocultural beliefs in the difficulty of amendment, as well as the impact of normative arguments and cultural attitudes against frequent constitutional change, may contribute to the difficulty of amendment today. Claims of impossibility fail to take into account the self-reinforcing character such claims: 3 it is possible that belief in the near impossibility of amendment contributes to the US constitutional culture’s disposition not to amend. Such a belief is reinforced by normative arguments that the Constitution’s text should not be “tinkered” with and with the development of habits of constitutional “veneration”. Thus, I argue, the claims of impossibility, based on the difficulty of the formal procedures, are overstated: amending the US Constitution is difficult but not impossible.
Section 3 addresses normative arguments about amendment, arguments which may assume special force in response to particular partisan divides. After discussing arguments against resort to amendment, it argues that over-reliance on interpretation and change in the judiciary as the sole methods for constitutional change undervalues the significance of more active forms of public consent for constitutional legitimacy; too strong a convention against formal textual amendment saps the democratic legitimacy of the formal text. To the extent that constitutions depend for their legitimacy on several sources, including democratic consent, unwillingness to use constitutional processes for amendment raises questions of constitutional legitimacy, even if public consent to adjudication provides a distinctive (albeit more indirect) form of legitimacy. Section 4 concludes with some reflections on the implications of this analysis more generally for comparative constitutional study.
The US Constitution is not easily amended. Under Article V, amendments initiated by the Congress—the only method thus far used—require a two-thirds vote of both houses to propose an amendment and ratification by three-fourths of the states, either in their legislatures or by special convention. Not surprisingly, the US Constitution has been less frequently amended than most other national constitutions—and much less frequently amended than many state-level constitutions in the United States. 4 The US Constitution is also one of the most long-lived of constitutions in the world. Its endurance is sometimes attributed to its difficulty of amendment, coupled with an approach to interpretation that has proven flexible over time.
Yet Article V has also been subject to critique for its rigidity, as inconsistent with democratic legitimacy. Indeed, the US Constitution has been sharply criticized for not allowing sufficient flexibility for democratic majorities to amend the Constitution’s text under Article V. 5 And it has been characterized not only as making change too difficult but as making such change almost impossible, at least if the change is significant: Sanford Levinson, for example, refers to Article V as “impermeable,” and as making constitutional amendment “practically impossible.” 6 Below I consider the difficulty posed by the formal rules of the Constitution, by over-estimates of those difficulties, and by ideological or cultural hesitation to employ amendment as compared to other forms of legal change.
The difficulty of amendment can be overstated and can become self-fulfilling, in ways that ignore the important history of Article V amending processes as parts of larger constitutional changes in the United States. As even Sanford Levinson acknowledges, the Constitution, though difficult, is not impossible to amend and, at times, it has been amended with amazing speed. The 26th Amendment, changing the voting age to 18, was proposed by Congress very quickly after the Court’s decision in Oregon v. Mitchell , 7 and was ratified by the requisite number of states within three months. 8 The 24th Amendment, eliminating the poll tax, was also relatively quickly ratified in the 1960s, being proposed by Congress in 1962 and ratified in 1964. At those times, the United States had the same 50 states it currently does.
The relative difficulty in amendment is a general feature of federal constitutions. Levinson and others have relied on Donald Lutz’s seminal work of the 1990s to show that, among the “selected” countries analyzed, 9 the US has the most difficult amending formula. But it is worth noting that, in Lutz’s study, bicameral approval or initiation requirements—as are likely to be found in federal constitutional states—generally increased the difficulty rating of amending schemes. 10 Lutz’s scale rated countries as varying in amendment difficulty from 0.8 percent (Austria) to 5.60 percent (Yugoslavia). The United States’ formal procedure was rated as 5.10 percent—not that much more difficult than Switzerland (4.75%), Venezuela (4.75%), Australia (4.70%), and Germany, Spain, and Nigeria (each at 4.60%). Requirements for bicameral approval of amendments by as much as a two-thirds vote of each house are not uncommon in federal and unitary systems. 11 Some later studies suggest that it is requirements of bicameral approval—or of multiple veto point and multiple decisions—rather than the federal aspect of the state as such that contributes to constitutional rigidity through formal amending processes. 12
Under existing procedures, it is not at all clear that the US amendment process should be regarded as “impossible” or even “close to impossible.” According to John R. Vile, a leading scholar of the US amendment process, of the thirty-three amendments that have been proposed by a two-thirds vote of both Houses of Congress, only six have failed to be enacted. 13 This record might be read to suggest that the major hurdle for amendments is the Congress, since the great majority of proposals that emerge are passed. How difficult is it to obtain agreement of two-thirds of both houses on proposals? It is not easy, to be sure. But if we consider the other situation in which the Constitution requires action by two-thirds of both Houses of Congress, the two-thirds of both houses requirement does not appear to be an “impermeable” barrier.
To propose an amendment to the states requires two-thirds of both houses; to override a presidential veto also requires two-thirds of both houses. There have been approximately 110 congressional overrides of presidential vetoes since 1789 (including 15 during the highly confrontational presidency of Andrew Johnson). 14 Although this number is under 5 percent of the total number of presidential vetoes, even this data suggests that it is far from impossible to obtain agreement from the two houses by a two-thirds vote on substantive issues. Moreover, more meaningful figures focus on only those “regular” vetoes that were not “pocket vetoes” (which do not go back to Congress for possible override), 15 and on public (rather than private) bills. 16 Focusing on these categories of presidential vetoes of legislation likely to be of consequence and capable of being overridden by Congress, one finds that a significantly higher percentage have been overridden. For example, a 2013 study finds that looking only at “regular vetoes” since 1961, thirty-seven out of 233 vetoes were overridden—for a rate of 15.9 percent. 17 If two-thirds of both Houses of Congress can join together to override Presidential vetoes at these rates, then perhaps the requirement that two-thirds of each House join in proposing amendments may not, by itself, be quite so large an obstacle to proposing amendments as might be thought; and as noted, most amendments proposed by Congress have been ratified by the states.
On the other hand, there are differences between veto overrides and proposals for amendment relating to ease of formulation, 18 and prior congressional investment, 19 that may make it more difficult for Congress to propose amendments than to override vetoes. The most recent proposal for an amendment from Congress—the Equal Rights Amendment proposed in 1972—failed to be ratified; although garnering ratifications in 35 states, it fell three states short of the 38 needed, illustrating that the requirement of ratification by three-fourths of the states poses an additional and real obstacle. Moreover, the first 12 amendments occurred very early in US history—by 1804.
There may, then, be political phenomena that exist now, that did not exist before, that make amendment particularly difficult. 20 Indeed, to the extent that an amendment becomes a highly partisan issue, under conditions of relatively even partisan divide, the amending formula will pose a very substantial obstacle. 21 Whether in the long run that is a good or bad effect is debatable; slowing down amendment in times of intense partisan divides is not necessarily a bad thing. 22 Overriding presidential vetoes also appears to have varied in rate over time; the first override of a presidential veto was not until 1845 (Tyler’s administration), and the pace of overrides has varied with the political makeup of the Congress and the President (with thirty overrides, for example, from the beginning of President Nixon’s term until the end of President Reagan’s). 23
A different way of looking at things is to introduce the possibility of variation in amendment pace over different time periods. 24 The United States has had several long periods without amendments: from 1804 to 1865; from 1870 to 1913; and from 1971 to the present (with the bizarre exception of the 27th Amendment). 25 Between 1865 and 1870, there were three important amendments in the wake of the Civil War; and between 1913 and 1971 there were eleven amendments. We may not, then, know the extent to which the Article V process is “dead,” or simply quiescent, as it has been in the past. There is at least a possibility that enacting amendments makes enacting other amendments somewhat more likely—and conversely, that disuse of the amending procedures may increase the political or cultural resistance to their use. 26
There is no agreed-on ideal pace of amendment—just as there is no agreement on ideal country size, or on ideal duration of constitutions. The widespread acceptance of judicial review in the United States, and of what some call “judicial supremacy” in interpreting the Constitution, 27 suggests that US constitutional culture accommodates a considerable amount of constitutional change through adjudication before the Supreme Court—and might even suggest that the people have consented to this enhanced judicial function.
Nonetheless, comparative experience suggests there may be advantages to a system that provides more flexibility and that also promotes a more representative amendment process. It might well be sensible—were reform of the amendment process on the agenda in the United States—to consider requiring ratification by something less then three-fourths of the states—such as by only two-thirds or three-fifths of the states. Other federal states have similar procedures but require ratification by smaller percentages of their member units—in Canada, by two-thirds of the provinces, 28 and in Australia, by a majority of the states. 29 Some federal states require only supermajorities in the national legislature, without resort to ratification in the subnational units, as in Germany 30 or India; 31 but given greater commonalities in legal origins and constitutional culture, the Australian and Canadian examples are probably closer comparisons. 32
But a comparative perspective also suggests a respect in which the formal rules for the US amending procedure may be too lax. Both Canada and Australia include a requirement lacking from the US Constitution: that, in addition to approval by a majority or two-thirds of the subnational units, the numbers of persons voting in favor of the amendment or the number of states approving the amendment must represent more than 50 percent of those who voted (or the voting population) nationally. 33 Such a requirement would increase the democratic legitimacy of the US amendment procedure, by avoiding the (hypothetical) possibility that an amendment were approved by the thirty-eight smallest states, against the expressed views of voters in the twelve largest population states—which hold a majority of the population. 34 Even if such changes in procedure were made, the process would still have features that would be difficult to meet and could be criticized as departures from simple majority rule. 35 But even as it now stands, is amendment of the US Constitution through article V impossible? History suggests that the answer is probably no. 36
What may increase the difficulty of formal amendment in the United States is a seemingly widespread view that constitutional change cannot , as a practical matter, be sought through amendment, and that as a normative matter amendments should not be the vehicle for legal change (but rather changes in judicial interpretation should be sought). Taking account of these aspects of political culture in assessing the effects of the different amendment procedures is consistent with some recent studies indicating that the formal rigidity of amending formula has, at best, limited value in predicting rates of amendment. 37
Although I cannot offer conclusive proof in this paper that empirical beliefs in impossibility and normative attitudes of constitutional veneration contribute to disuse of the Article V amendment process, the suggestion that they may be causal factors is consistent with emerging work on the significance of culture as a factor influencing amendment rates. 38 On the first of these—overstatements of the difficulty of using the amendment process—as already noted, Levinson entitled a chapter in his book, “The Impermeable Article V,” and refers to Article V as making it “practically impossible” to enact constitutional change. 39 He is by no means alone in doing so. Professor William Eskridge describes the failure of the Equal Rights Amendment as having apparently been the “death” of the Article V amendment process. 40 Joel Colón-Ríos describes Article V as “creat[ing] almost insuperable constraints on constitutional change.” 41 These statements are widely echoed in the scholarly literature.
Predictions of impossibility often combine with asserted reverence for the Constitution, related normative resistance to proposing changes in its text, and, sometimes, fears that any resort to the amendment process will invite populist movements to amend in undesirable ways; these all may cause hesitation to resort to the amendment process even when it would seem a logical response. Consider the reactions from groups that disagreed with the Court’s decision in Citizens United v. Federal Election Comm’n . 42 Public opinion polls after the decision indicated that public sentiment, among both Republican and Democratic voters, was running strongly against the decision. 43 One might have thought this was an opportune moment to go the people, through the amendment process, to obtain constitutional authorization for regulation of campaign expenditures and contributions. Yet some combination of reasons—which may include concerns for overriding the Supreme Court through amendment, reluctance to amend in any way the First Amendment, as well as a belief in the futility of the amendment process, and support for the decision by influential minorities—prevented any groundswell of support for a responsive amendment in the first years after the decision came down. Yet an immediate response through constitutional amendment (as occurred, for example, after Oregon v. Mitchell ) had some compelling reasons behind it, including the benefits of acting before spending and fundraising patterns adjusted to the new regime under Citizens United and expectations by politicians and corporations became entrenched. 44
A leading public interest group that opposed the Citizens United decision was the Brennan Center for Justice at New York University School of Law. The Brennan Center devoted considerable effort to developing strategies for responding to the decision, including hosting conferences and preparation of research/policy papers. The nature of the organization’s early response is illustrated by a paper, Renewing Democracy After Citizens United, which proposed three strategies toward renewing democracy: enacting legislation for small donor public financing; increasing voter registration; and developing new jurisprudence, that is, new doctrinal arguments to persuade future courts of a different view. 45 In 2011, the Brennan Center held a conference, Accountability After Citizens United , also on responses to Citizens United , that focused on statutory approaches through securities and corporate law; activating administrative agencies, like the Federal Election Commission (FEC), Federal Communications Commission (FCC), and Internal Revenue Service (IRS); and, again, developing new jurisprudence around the issue. 46 The Brennan Center’s failure in those responses to devote any attention to the possibility of an amendment as a form of “democratic renewal,” or “accountability,” speaks volumes about its belief in the possibility, and/or desirability, of amendment.
Similarly, an issue analysis prepared for another public interest group, the American Constitution Society (ACS), soon after the Citizens United decision, addressed several issues. 47 They included the need to have better factual records for “strict scrutiny” review to support the claimed need for regulation; increasing disclosure requirements; addressing disclosures and expenditure issues through corporate law and requirements of shareholder approval; moving towards public funding of campaigns; modernizing voter registration; and appointing as Supreme Court justices those with a “voter-centric” view of the First Amendment. 48 Despite its thoughtfulness and seemingly comprehensive overview of possible responses (including the constitutional politics of appointments to the Court), this memo for ACS did not mention the possibility of constitutional amendment. 49
To be sure, some critics of Citizens United have tried, and with some success, to promote a constitutional amendment in response. At least fourteen proposals were introduced in Congress within two years of the decision; these range from amendments sweeping broadly to say that corporations do not hold rights under the Constitution, to those authorizing legislatures to regulate corporate, or all, campaign contributions and expenditures, to those excluding campaign expenditures from First Amendment protections, to those establishing a substantive rule, for example, banning corporate contributions or expenditures. 50 An organization called “Move to Amend” seeks petition signatures for a proposed amendment to “our Constitution to firmly establish that money is not speech, and that human beings, not corporations, are persons entitled to constitutional rights”; they reportedly had about 300,000 signatures in mid-2013. 51 Another organization, “Free Speech for People,” supports amendments to “(1) enable Congress and the states to limit campaign spending and (2) to make it clear that corporations do not have constitutional rights;” in 2013 it claimed to be “one-third” of the way to succeeding with a constitutional amendment. 52 In 2014, the “Democracy for All” amendment, Senate Joint Resolution 19, 113th Congress, did not receive the necessary two-thirds vote, notwithstanding majority support in the Senate. 53 Some well-established public interest groups, including People for the American Way, 54 and Public Citizen, 55 have supported amendment as an approach, as have some legal scholars. 56
But other influential non-governmental organizations (NGOs) and political actors who disagreed with the decision in Citizens United nonetheless ruled out seeking an amendment, either because they believed it is so unlikely to succeed and/or because they believe other approaches have normative advantages. It is possible that those groups that did not include amendment as a possible response to Citizens United did so because they believed a constitutional amendment is a bad idea for reasons other than its unlikelihood of enactment. 57 They might believe that an amendment would be too hard to craft to achieve its goals in the hands of an uncertain Supreme Court; or that entrenching provisions relating to the subject is a bad idea for democratic politics; or even that “progressives” who oppose Citizens United should not seek amendment lest it encourage “conservatives” to seek amendments that progressives disagree with. 58 But a significant part of the opposition to amendment—among those who would like to overturn the effects of Citizens United —has to do with a belief that it would be impossible. For example, Mark Schmitt, Senior Fellow of the Roosevelt Institute, advanced a series of arguments, which appear mostly bottomed on the unlikelihood of an amendment passing. 59 In one debate with proponents of amendment, he stated that proceeding with an amendment would imply to those working on other approaches that “we can’t do anything until we have a constitutional amendment,” and argued that “under the current circumstances, ‘We can’t do anything until we have a constitutional amendment’ is exactly the same as saying, ‘We can’t do anything.’” 60 Later remarks made clear he was not substantively opposed to amendment insofar as it related to campaign finances, and that he agreed that Citizens United was wrongly decided, 61 though he also raised questions about the capacity to draft an adequate amendment. 62 Thus evidence suggests that opposition to amendment is based in part on concerns that it would be impossible and in part on normative worries about amendment as opposed to other mechanisms of change.
The resistance to amendment as a response to judicial interpretation of the Constitution is also illustrated in the earlier controversy surrounding the Court’s flag-burning cases. In 1989, the Supreme Court held unconstitutional, under the First Amendment, a Texas state law that made it a criminal offense to “desecrate a venerated object,” specifically defined to include a “state or national flag,” as by burning an American flag at a public political demonstration. 63 The decision was controversial, and unpopular, generating calls for response including a constitutional amendment supported by President Bush and many Republicans. 64 Leading Democrats opposed an amendment, but supported a federal law prohibiting flag burning, as did a number of leading constitutional scholars, in part to avoid “tinkering” with the Constitution. 65 Both Professors Tribe and Dellinger testified that the Court would probably uphold a statute prohibiting flag destruction if written in more neutral terms (focused on “destruction” rather than “desecration”), while Robert Bork, a leading constitutional scholar and unsuccessful Republican nominee for the Supreme Court, testified that the only way to make lawful a statute directed at flag burning was by constitutional amendment. 66 Congress enacted a federal statute banning destruction of the flag, whether out of then-genuine belief that such a statute was a good policy that would be upheld by the Supreme Court, or out of a hypocritical desire to display patriotism in a likely unconstitutional act; 67 the federal statute was promptly found unconstitutional by the same five-justice majority that struck down the earlier statute. 68 By the time of the second decision, however, passions had cooled, 69 and although efforts to enact an amendment continued at least through 2006, no amendment was proposed by the requisite votes of both houses of Congress. 70
In both of these instances, some apparent opponents of a Supreme Court decision invoked normative arguments in favor of responses that did not involve a constitutional amendment. It is to such normative arguments that I now turn more directly.
Reluctance to resort to amendment for normative reasons may contribute to the relative absence of successful amendment movements, in ways that reinforce the sense of futility in trying; the perceived difficulty and the normative arguments against constitutional amendment (even when there is strong disagreement with the Court) may reinforce each other in an overall cultural message: whatever you do, don’t amend the Constitution. This section explores the normative case for a presumption against amendment as a response to decisions one disagrees with, as it has been developed by its leading proponents in the United States, and then considers affirmative arguments in favor of amendment as a response that may at times be preferable.
In the 1990s, as numerous proposals for constitutional amendments received consideration in Congress, a public interest group, the Constitution Project, formed and issued a report, “Great and Extraordinary Occasions: Developing Guidelines for Constitutional Change,” 71 which cautioned against resort to amendment in response to judicial rulings. Adrian Vermeule has referred critically to what he calls a “generic . . . presumption” in the United States that constitutional amendments are a mistake. 72 One cluster of normative arguments include opposition to “clutter[ing]” or “trivializing” the Constitution, 73 or undermining its “unifying force” and “symbolic significance” by causing it to be viewed as a special interest vehicle. 74 Efforts to amend “often put on the table fundamental issues about our character as a nation, thereby bringing to the fore the most divisive questions on the political agenda,” which could harm national cohesion, because “no country can . . . [argue] continuously” about fundamentals. 75 (Similar arguments have been advanced recently in the UK for not engaging in a process to create a written constitution. 76 ) Indeed, some opponents of amendment in the United States might be understood to seek an unwritten constitutional convention against formal textual amendment, to promote that “veneration” of the Constitution seen as an important basis for rights protection and the rule of law. 77
While very frequent amendment may detract from the rule of law and diminish attachment to the constitution, stability and the rule of law are surely not threatened by anything approaching the very low amendment rate in the United States; the last time the Constitution was formally amended was more than twenty years ago, in 1992, and then by an amendment originally proposed in 1791 in a bizarre and sui generis ratification process. 78 The difficult hurdles for amendment are themselves likely to stand as obstacles to the kind of frequent amendment, or “cluttering” that might pose real threats to either stability or the rule of law.
Apart from arguments already noted, Kathleen Sullivan has argued that a presumption against amendment is warranted by interests in constitutional coherence, in constitutional generality, and in preserving the role of the Supreme Court. 79 Arguments from constitutional coherence may be predicated on a greater presumption of existing coherence than is warranted. 80 And the related worry over drafting challenges, including the unintended consequences of a constitutional text later interpreted by courts, 81 arguably support care in drafting more than to hesitation to amend. 82
The interest in generality is important for some aspects of constitutions; indeed, the harder a constitution is to amend, the more significant it is that it contain sufficiently general principles; on the other hand, some parts of a constitution need to be specific (e.g., in identifying when elections are held). This concern does relate to what are appropriate subjects for amendments; the caution against entrenching specific policies, of a kind on which democratic polities might legitimately change their minds, in a very difficult to amend constitution, does seem like a relevant factor. 83 But a “general principle” might include giving an authority to a legislature to act in an area denied them under judicial decisions, as in the proposed Child Labor Amendment and, in part, the 21st Amendment repealing Prohibition (which gave authority to state legislatures to regulate intoxicating liquor). A leading proposal to respond to Citizens United takes the form of authorizing legislative regulation of campaign expenditures. 84 Sullivan’s concern for not “impos[ing] a controversial social policy” or not entrenching “short-term policy goals” 85 (which she illustrates by reference to Prohibition), can be quite difficult to apply ex ante : surely, the 13th Amendment abolishing slavery was, even if controversial, normatively appropriate; and if Prohibition (of alcoholic beverages) was a “short-term” goal, perhaps it could only be known as such over the passage of time (which resulted in its repeal). 86
Finally, on the role of the Supreme Court: too frequent use of amendment to overrule Supreme Court decisions with which a majority disagrees would have the potential to diminish the role of the Court. But too frequent enactment of statutes that are in effect efforts to overrule might have the same effect, as the Court may have believed was the case when Congress, in response to the Court’s decision in Employment Division v. Smith , 87 enacted the Religious Freedom Restoration Act; 88 the Court struck down major portions of this Act, implicitly rebuking Congress for giving insufficient respect to Smith . 89 Acceptance of the Court’s results in adjudicated decisions is important to the rule of law and to legal stability; having the Court perform a “settlement” function on constitutional contests has real value (especially when its decisions protect rights of vulnerable and relatively powerless minorities). But the requirements for enacting amendments are themselves so rigorous, that those procedures—and the history of their use—suggests that the risks of too frequent amendment are low. 90 Whether particular decisions should be overturned is a question not only of the Court’s settlement role, but also of the merits of the decision.
A related concern about frequent amendment threatening minority rights was raised by Walter Dellinger’s testimony on the proposed flag-destruction statute after Johnson v. Texas . 91 For Dellinger, a statutory response was preferable to an amendment, because the latter might whet “the appetites of many for quick ways to leap over constitutional barriers” protecting the rights of “unpopular group[s].” 92 Professor Michelman dubbed these and other concerns the fear of “addictive” amendment. 93 Michelman argued that a flag burning statute, if upheld, might have broader adverse effects on the Constitution’s protection of freedom of speech than would a constitutional amendment: on conventional understandings, amendments can draw arbitrary lines to limit change to particular subject areas, in ways that the principled consistency requirements of judicial review would not permit. 94
Even under criteria discussed by some supporters of a presumption against amendment, the subject matter of a proposed amendment to authorize state and federal legislatures to engage in reasonable regulation of campaign expenditures and contributions in response to Citizens United is arguably justified. Sullivan, for example, wrote that amendment is appropriate “when changes consistent with [the Constitution’s] broad purposes are unlikely to be implemented by ordinary legislative means.” 95 Sullivan’s example is franchise-expanding amendments which, she says, were appropriate as constitutional changes because existing electors would be hesitant to “dilute” the value of their own suffrage by making change at the state level without being sure other states would do the same. Whatever the merits of this argument, an a fortiori argument might well apply when an amendment is—short of persuading the Court to overrule itself—the only way to authorize regulation that so many believe is important to preserving the value of individual votes in elections. 96
Nor has the Constitution Project, which published guidelines concerning when amendment is or is not appropriate, argued that the Constitution should never be amended. Although its report opposes any constitutional amendment to overcome the effects of Buckley v. Valeo , 97 it draws a distinction between amendments that take a definitive position on a contested issue of policy as opposed to those that authorize governments to regulate. 98 A provision with such authorization for legislative action could overrule Citizens United without raising concerns about unduly entrenching particular substantive policies.
I move now to arguments in favor of considering amendment as a response to judicial decisions that seem importantly wrong. While there are reasons to be concerned about too frequent amendment, so, too, are there reasons to be concerned about too infrequent use of the Article V amending power. There are some significant benefits of amendment, as compared to reliance on adjudication (or ordinary legislation), to respond to a perceived need for legal change.
First, constitutional amendment focuses more political attention on the object of the amendment once it gains momentum. In its early phases, an amending process focuses proponents’ attention on identifying the constitutional principle at stake. As it moves forward, and positions are articulated, areas of consensus on possible legislative solutions may emerge; greater identification of sources of opposition will also emerge; and—at least in theory—a better deliberative process can result.
This process is likely to be more costly than the process of adjudication, or of ordinary legislation. But precisely because of its more diffuse and iterative processes, constitutional change by virtue of the amendment process of Article V may be more likely to be treated as final by those opposed and more likely to endure than contested judicial decisions. 99 (For these reasons, some scholars have suggested that even though the initial decision costs of adjudication are lower than of amendment, over the longer run, amendment may end up being a more efficient way to assure enduring constitutional change. 100 ) To be sure, the more incremental processes of constitutional adjudication—what Heather Gerken refers to as “informal constitutional amendments”—have some countervailing advantages, including their open-ness to continued contestation about meaning. 101 But just as not everything needs to be resolved by an authoritative text, so too not everything needs to be as open-textured as adjudicated constitutionalism may tend to be.
Second, amendment in response to judicial error is an important check on the judiciary, particularly given the unlimited tenure for the Article III judiciary and the absence of a mandatory retirement age. Retired Justice John Paul Stevens recently called for six different amendments, each to respond to six lines of judicial decisions he argues are in need of correction. 102 The role of amendment in maintaining the democratic roots of a constitution may be more important in a country like the United States, where great stability in judicial attitude can persist across multiple presidential elections, depending on contingencies of aging, than in countries with more regular turnover in their constitutional courts. And correction by amendment is, arguably, more consistent with the rule of law and with the legitimacy of courts than judicial overruling. 103 Opponents of amendment seem to assume that judicial overruling poses no threat to the rule of law; but I am not sure why this would be so: any authorized changes from existing rules—whether by amendment, statute, or new judicial decisions—holds some threat to stability and rule of law values. 104
Third, as Adrian Vermeule has argued, it is sometimes a benefit that broader ranges of people participate in determining fundamental legal constraints. 105 Legislators, themselves, are likely to be representative of a broader range of views than the nine justices who sit on the US Supreme Court, thereby providing more “public” involvement even in the proposal for amendment. 106 Moreover, the US amending process draws on a much wider range of the public than those who can participate in litigation, and offers different kinds of opportunities for participation than does participation in social movements.
Moreover, the proposal of an amendment by Congress and subsequent responses in the states—even if not in the end sufficient to ratify—may influence jurisprudence in directions similar to those of the amendment itself. 107 Adjudicated results may be unstable unless supported by political mobilizations; thus, adjudicatory decisions are likely to work best when they harmonize with more broadly held views. Even unsuccessful amendment proposals can sometimes shift constitutional understandings: the unsuccessful fight for the Equal Rights Amendment (ERA) (first introduced in 1923, but not sent out to the states by Congress until 1972), may have contributed to—or could be seen as of a piece with—social movements leading to changed understandings of gender that underlay the Court’s change in equal protection jurisprudence to provide a substantial part—though probably not the whole—of what the adoption of the ERA would have accomplished. 108 Reva Siegel has drawn attention to the way in which ERA proponents envisaged that the political mobilization for the amendment could have positive effects on adjudicated understandings of gender equality. 109 Thus, focusing only on the difficulty of formal amendment may understate the capacity of those amending procedures to influence constitutional development even when the formal requisites are not met.
Finally, some use of the amending process would be consistent with recognizing the idea of constitutional imperfection and the roots of the Constitution in more active forms of democratic consent. 110 The amendment provisions of the Constitution represent an important reminder that instruments made by human beings are imperfect (as may be judicial decisions interpreting those provisions), and of the need for humility about quality of existing law. Thus, “absent flagrant disregard for constitutional language, some amendments will be required as defects become apparent, or changes are desired.” 111 Some use of the amending process should be possible without threatening public attachment or affection towards the Constitution as a whole. 112 Refreshing the Constitution, through resort to amendments to clarify and put into the basic law a fundamental public norm—possibly, for example, towards reducing the role of highly concentrated forms of wealth in elections—might even increase public attachment, especially as those amendments are announced and solemnized through appropriate formalities. 113
This is not to suggest that judicial development and change of constitutional law over time is illegitimate or anti-democratic. The legitimacy of judicial review is beyond question at this point in the United States—legally, morally, and sociologically. But judicial review is connected through much more indirect processes to developing public views about constitutional values, and long time lags may exist depending on accidents of the appointment process and medical histories and life spans of those appointed. And while the standard interpretive approaches of constitutional law have over time looked not only at text, original meaning, and precedent but also at developing values, historic practice, and pragmatic consequences, 114 the application of constitutional law is a place where tradition meets present needs; by virtue of its embeddedness in a common law system of adjudication judicial review is likely to proceed incrementally; the past can thus prevail over present understandings for long periods. It is not an indictment of judicial review to suggest that some direct use of the tools of political democracy—i.e., voting on proposed constitutional changes in Congress and in the states—is valuable in sustaining a healthy balance among institutions and processes in the authoritative articulation of constitutional law. 115
This conclusion is reinforced by considering more generally the elements of constitutional legitimacy.
Within US constitutional theory, constitutional legitimacy is sometimes described as deriving entirely from the degree to which the Constitution can be understood to embody the will of the particular people of the United States. 116 If this were correct, the relatively infrequent resort to formal amendment would be highly anomalous and in need of even greater justification. However much one can argue that public acquiescence in constitutional interpretations is equivalent to consent, it seems likely that a more active form of democratic engagement on constitutional fundamentals would yield some differences to those produced through the incremental process of adjudication. But there is sound reason to think that what gives constitutions—and their amendments—legitimacy is far more complex, and can be understood to depend on several components.
David Beetham, a theorist of political power and democracy, has argued that the legitimacy of power can be understood through three perspectives: (i) “legal validity”—referring to the acquisition and use of power through existing rules; (ii) the “moral justifiability” of power relations, that is, the degree to which the rules for acquiring and exercising power are justified in light of the beliefs of both dominant and subordinate groups in the society; and,(iii) “evidence of consent,” particularly by the “subordinate to the particular power relation.” 117 Writing of the concept of legitimacy in constitutional law, Richard Fallon has likewise argued that there are “three distinct kinds of criteria that in turn support three distinct but partly overlapping concepts of legitimacy—legal, sociological, and moral.” 118 These two accounts, though similar, differ in their treatment of consent. 119 Beetham explicitly links consent, as a legitimating factor, to certain formalities: “evidence of consent expressed through actions which are understood as demonstrating consent within the conventions of a particular society.” 120
These frameworks of analysis suggest that constitutional legitimacy is more complex than can be captured by invocation of the “will of the people,” or the “constituent power,” at any given moment. Rather, constitutional legitimacy rests on three elements: the roots of the constitution (or its amendment) in positive consent or active acceptance by the people who it governs; the constitution’s (or amendment’s) embodiment of good and/or just principles; 121 and the constitution’s (or amendment’s) compliance with or promotion of the rule of law. 122 These considerations overlap: for example, consent contributes to sociological legitimacy and may also be understood as required by the rule of law, to the extent that a constitution provides rules for its own amendment; consent also sounds in the kind of good principles on which many political theories ground the authority to govern. (Thus, popular consent may be understood not only as a necessary form of sociological acceptance but also as a morally legitimating factor, 123 and as a separate criterion for legitimacy in its own right.) But these considerations do not fully overlap: for example, a constitutional text may embody an immoral rule of law (e.g., the now-defunct Fugitive Slave Clause of the US Constitution) or one inconsistent with good government (as Levinson has argued is true of the requirement that each state, regardless of population, have two Senators); yet the rule of law and its emphasis on obedience to controlling law until it is lawfully changed would tend to support their legitimacy.
The relative weight of these considerations will vary in different contexts and institutional settings. 124 To the extent constitutions rest on power, consent may be a necessary condition for legitimacy, though it can be manifested in different ways. 125 But if consent by the people is necessary, it may not be sufficient, for reasons in part relating to why consent is valued at all and in part relating to the transnational setting in which constitutions must function.
Walter Murphy has argued that consent is valued because of human capacities for autonomous moral reasoning and, therefore, constitutional provisions inconsistent with recognition of these capacities in others can be found unconstitutional. 126 Quoting Joseph Raz, Murphy argues, “[c]onsent can function as a legitimator only to the extent that the people agree as reasoning, morally autonomous, and responsible human beings. . . . ‘[T]o the extent that the validity of consent rest[s] on the intrinsic value of autonomy it cannot extend to acts of consent that authorize another person to deprive people of their autonomy.’” 127 A similar argument can be developed grounded on equality as the basis for giving legitimacy to democratic consent: if consent to laws enacted by majorities legitimates those laws because we presume human beings equal by virtue of their being human, and majority voting is a decision rule that recognizes this equality, then use of majority voting to deprive some of their equality is inconsistent with the justification for the decision procedure: 128 if the reason justifying voting and the majoritarian principle is an underlying moral commitment to equality of persons, then even a supermajority deliberative decision to entrench slavery would not be legitimated by its basis in democratic consent, because the premise for the legitimating force of consent are legitimating would be absent. 129
Moreover, increasingly, constitutional legitimacy may depend on meeting minimal transnational norms (e.g., against wars of aggression, or piracy, or torture), as well as being consented to by the people a constitution most directly controls. The reasons for the increased importance of what we might call external legitimacy are many and are discussed elsewhere. 130 Transnational discourses may at times converge on the identification of principles of justice or public morality in ways that both reflect and in turn influence constitution-making efforts. For example, there is now widespread support—at least rhetorically—for judicial independence as a fundamental feature of governance. 131 Whether the violation of norms of justice or good government is sufficient, by itself, to make constitutional amendments or revisions subject to judicial review is, of course, a separate and distinct question, 132 though it has been argued that transnational constitutional practices can serve as a useful guide to those courts that do engage in substantive review of procedurally proper constitutional amendment. 133
But notwithstanding important other sources of legitimacy, having roots in popular consent remains a central feature of any constitution that provides the basic law for a constitutional democracy or democratic republic. Indeed, consent may—over time—be the most central feature of sustainable constitutionalism, a necessary if not sufficient feature. Consent, to be sure, can be given in different ways; important change may legitimately occur in constitutional law through processes other than formal amendments. 134 But almost all written constitutions make provision for their own amendment, 135 reflecting the normative demands that constitutions rooted in basic propositions of democracy be subject to revision through processes in which the people can formally participate, either directly or though their elected representatives, in order to remain instruments of self-government. 136 The Article V process has advantages in focusing public attention on a particular text, and using specified procedures known in advance to constitute lawful formal acts of constitutional amendment through publicly participatory processes.
These understandings provide further reasons to modulate the intense normative resistances to using the amendment process in the United States. For a constitution to remain the legitimate product of “we the people,” amendment—through actions in which the public participates and which are formalized through specific events of proposal and ratification—has an important role to play. 137
This paper has argued, first, that the difficulty posed by formal procedures in obtaining textual amendments of the US Constitution is at times somewhat overstated; second, that overstatement of the difficulty may become “self-fulfilling”—that is, may contribute to increasing the difficulty of formal amendment; that together with normative arguments against the use of amendment to effectuate legal change, the perception of difficulty may become even more self-fulfilling; and finally, that notwithstanding important normative arguments for caution in seeking constitutional amendment, a constitution that was truly not amendable by the formal methods that it provides for popular input (either directly or through representatives), would be inconsistent with democratic constitutionalism. I elaborate here on this last point, from an internal perspective on US constitutionalism and from a comparative perspective.
There may be as much reason to be concerned about “amendophobia”—that is, an irrational fear about ever resorting to amendment, even when the occasion is appropriate—as to worry about what Sullivan termed “amendmentitis,” a too ready resort to amendment in response to political problems or judicial decisions. 138 To the extent it exists, the myth of impossibility of amendment may do harm to the democratic roots of US constitutionalism, contributing to a diminution in the Constitution’s democratic legitimacy.
There is no question that the US amendment procedures are difficult. Some rigidity expressed in supermajority, multi-vote procedures for amendment is not itself necessarily an anti -democratic feature. Multiple supermajority voting requirements—and even, arguably, some malapportionment in voting requirements—for constitutional amendment can be understood as features of, or efforts to assure, a more deliberative form of democracy than simple one-off majority voting provides. 139 To be sure, the US voting rules allow very small minorities of the population to block amendment, if the defenders of the status quo are concentrated in small population states. 140 But in some respects, it is the degree to which the amendment process has been foresworn in preference to seeking change through other means (especially adjudication)—the belief that the constitution is and should be unamendable—that is a key “anti-democratic” feature of US constitutionalism. This belief is sustained in part by a reverence for the Constitution, in part by important normative arguments against ready amendment, and in part, by mistaken beliefs as to its impossibility. To the extent formal amendment is actively foresworn, it deprives the US Constitution of important roots in democratic consent on which its long-term legitimacy may depend.
The provisions for formal amendment do not exist or operate in a vacuum but are part of an inter-related set of constitutional arrangements by which both constitutional continuities and constitutional transformations are made possible. Judicial review in the United States is distinctive from that in many other constitutional democracies. It extends to many important questions of individual rights, as well as to the enforcement of structural limits; in this respect it is broader than the scope of judicial review in, for example, Australia. It is conducted by both state and federal judges, unlike in many continental legal systems; but it operates under the ultimate supervision of the US Supreme Court, whose nine justices are appointed for indefinite terms of office, with no fixed retirement age or clear conventions relating to the timing of resignation; in popular parlance, they are appointed “for life.” 141 The regular input provided by fresh members to high constitutional courts that occurs in many other systems is lacking. In most countries that rely on specialized constitutional courts, the justices serve for a fixed and nonrenewable term of years (in Germany, for example, for twelve years), and are then replaced by political actors with new members. Australia and Canada, with more “generalist” supreme courts, as in the United States, originally provided for indefinite tenure in office; their justices now face mandatory retirement at age 70 or 75, as a result of constitutional amendments in the mid-twentieth century. 142 As of this writing, four Justices of the US Supreme Court are over the age of 75. 143
Moreover, the decisions of these very long-serving justices, many of whom at any moment in time may have been appointed by political actors whose views have since been rejected by the polity, exercise a strong power of judicial review, given widespread acceptance of the “supersupremacy” of the Court’s decisions—the presumption that they are binding in their holdings and reasoning on other branches and levels of government. To be sure, there are some more “departmentalist” moments in US constitutional history, where presidents, for example, have acted on a matter within their constitutional competence on a view of the Constitution different from that of the Court, as when President Andrew Jackson vetoed the national bank bill on constitutional grounds even though the Court had previously upheld national power to provide for such a bank. 144 But traditions of compliance with Supreme Court decisions (and reasoning), even when legislators and executive officers disagree with them are strong, as evidenced by Vice-President Gore’s acceptance of the Supreme Court’s decision on the 2000 election.
To foreswear amendment as a mechanism for legal change, in light of all of these features of US constitutional law, may place simply too much weight on judicial appointments and the adjudicative process as mechanisms to provide appropriate amounts of constitutional flexibility and change. This is not to say that every Supreme Court decision with which large majorities disagree should be subject to amendment: these decisions may be correct, in terms of principle, and should be defended and argued in public fora; some may have effects that are capable of being contained through subsequent legislative and adjudicatory development. But it is, rather, to argue that amendment should not be presumptively foresworn, taken off the table of possible responses; that doing so removes a potentially significant legal tool for maintaining a peaceful and civilized form of constitutional contestation in a constitutional democracy.
A brief, more comparative perspective: A certain degree of “US exceptionalism” appears to be a feature of contemporary comparative constitutionalism; it is increasingly common to assert the outlier status of the United States on a range of contemporary constitutional issues. 145 The US Constitution might appear to be exceptional in its difficulty of amendment, no less than in its present commitment to a version of freedom of expression at odds with that of many other constitutional democracies (and implicit in leading human rights covenants as applied to “hate speech”), or in its lack of commitment to social welfare rights. But in Australia constitutional amendments face considerable hurdles, 146 and some Canadians assert that their constitution has become impossible to amend. 147 It is important in this area as in others not to overstate the degree of US exceptionalism—important for accurate comparative understandings and important, as well, to avoid contributing to the unhealthy, self-sustaining effects of prophecies of impossibility of amendment within the United States.
It is also important to bear in mind the contexts within which particular functions, doctrines, or institutions operate, to bring to bear what I have elsewhere described as a contextualized form of functionalism in making comparisons. 148 Amendment rules must be analyzed not only in how they operate in combination with such features as the appointment mechanisms for judges, the structures for judicial review, and the degree of finality or supremacy accorded judicial interpretations, but also in light of national self-understandings about the degree to which a constitution should be subject to amendment. Such self-understandings, or conceptions of national identity as bound up in particular historical documents, may themselves be or become parts of the socio-legal culture in which constitutions operate. 149 Appreciating the importance of such self-understandings about constitutions and their amendment affords fascinating opportunities for comparison, 150 while at the same time raising epistemological cautions about efforts to derive generally applicable, functional “rules of thumb” for constitutional design.
See, e.g ., Sanford Levinson , Our Undemocratic Constitution: Where the Constitution Goes Wrong (And How We the People Can Correct It) (2006); Rosalind Dixon, Updating Constitutional Rules , 2009 Sup. Ct. Rev . 319, 319 (2010); Richard H. Fallon, Jr., American Constitutionalism, Almost (But Not Quite) Version 2.0, 65 Me. L. Rev . 77, 92 (2012).
For a related point, that constitutional powers not used over a long period of time will “atrophy,” and will come to seem illegitimate to use, or “unexercisable,” see Adrian Vermeule, The Atrophy of Constitutional Powers , 32 Oxford J. Legal Stud . 421, 422–423 (2012) (arguing that the failure to use a constitutional power over a long period can become a “political precedent heuristic, according to which an attempt to revive the power at a later time amounts to an illegitimate attempt to change the rules of the political game”); cf. Richard Albert, Constitutional Amendment by Constitutional Desuetude , 62 Am. J. Comp. L. 641 (2014) (arguing that constitutions can be amended by desuetude). I do not claim that amendment is currently regarded as illegitimate in the United States, but do raise concerns about the effects of atrophy in the use of the amendment power to address major constitutional concerns.
See Donald S. Lutz, Toward a Theory of Constitutional Amendment , in Responding to Imperfection 237, 247, 261 (Sanford Levinson ed., 1995) (describing the US Constitution as the second most difficult to amend in the world, after Yugoslavia’s); Zachary Elkins, Tom Ginsburg, & James Melton, The Endurance of National Constitutions 101 (2009) (describing the United States as “one of the most inflexible” constitutions in the world).
See, e.g ., Joel Colón-Ríos, Weak Constitutionalism 66 (2012); Levinson, supra note 1.
Levinson , supra note 1, at 159, 160.