ArticlePDF Available

Transnational constitutionalism and a limited doctrine of unconstitutional constitutional amendment: A reply to Rosalind Dixon and David Landau

I•CON (2017), Vol. 15 No. 3, 826–832 doi:10.1093/icon/mox058
© The Author 2017. Oxford University Press and New York University School of Law.
All rights reserved. For permissions, please e-mail:
Transnational constitutionalism
and a limited doctrine of
unconstitutional constitutional
amendment: Areply to Rosalind
Dixon and David Landau
As Rosalind Dixon and David Landau observe in their article, “Transnational
Constitutionalism and a Limited Doctrine of Unconstitutional Constitutional
Amendment,”1 the breakdown of constitutional democracy is a central feature of
contemporary politics. Yet it is important to remember that this phenomenon is as old
as democracy itself. For comparative constitutional law, what is important is a shift in
how the breakdown of democracy tends to occur. As I have argued elsewhere, histori-
cally, the paradigmatic example of democratic breakdown was a military coup d’état
that seized power and overthrew a civilian government in blatant contravention of
the existing constitutional order.2 This was later joined by the self-coup or autogolpe,
whereby democratically elected presidents remained in power unconstitutionally and
escaped the connes of term limits and/or electoral losses, for example, by declaring
a state of emergency that suspended many of the constitution’s provisions, and then
amending parts of or rewriting entirely the constitution by decree or convening an
extra-constitutional constituent assembly.
However, as Nancy Bermeo observes, the prevalence of coup d’états and autoglopes
declined steeply with the end of the Cold War, because of the strong international disincen-
tives and the widespread presupposition that democracy is now “the only game in town.”3
What has partially replaced them is democratic backsliding, whereby a democratically
* I. Michael Heyman Professor of Law, University of California, Berkeley; Director, Center for Constitutional
Transitions. Email:
1 Rosalind Dixon & David Landau, Transnational Constitutionalism and a Limited Doctrine of Unconstitutional
Constitutional Amendment, 13 Intl J. Const. law 606 (2015).
2 Sumit Bisarya & Sujit Choudhry, Regional Organizations and Threats to Constitutional Democracy from
Within: Self-Coups and Authoritarian Backsliding, in the Rule of law and ConstItutIon BuIldIng: the Role of
RegIonal oRganIzatIons 183 (Raul Cordenillo & Kristin Sample, eds., 2014), available at http://www.idea.
3 Nancy Bermeo, On Democratic Backsliding, 27 J. demoCRaCy 5 (2016).
Transnational constitutionalism: A reply to Rosalind Dixon and David Landau 827
elected government or president uses legal means to manipulate rules and institutions to
remain in power in future electoral cycles, inter alia with respect to electoral system design,
election administration, political party regulation, presidential term limits, civil and politi-
cal liberties, and independent institutions such as constitutional courts and agencies.
Democratic backsliding is sometimes termed authoritarian backsliding when the democratic
regime under threat replaced a prior authoritarian regime to which it may revert.
To be sure, democratic backsliding is also far from new; the disintegration of
Weimar, and the rise of Nazi Germany, historically stands as the central case of this
kind of capture of democracy from within, and serves as the genesis for the idea of
militant democracy.4 But the fact that democratic backsliding has now emerged as
a major form of democratic breakdown, in response to changes in the international
environment and the diffusion of democracy as a norm of governance, provides the
broader context for Dixon and Landau’s article. Their principal contribution is to link
democratic backsliding to an older debate about judicial review of constitutional
amendments where there is no textual basis for doing so. On their account, “abusive”
constitutional amendments are central techniques for democratic backsliding, and are
often enacted to overrule judgments nding the same actions unconstitutional, and
judicial review of such amendments is justied on the basis that it safeguards democ-
racy from itself, a point that Sam Issacharoff has also made.5 But on their account, the
assertion of such a judicial power carries a potential danger—that courts could over-
reach and second-guess constitutional amendments which do not pose an existential
threat to constitutional democracy. Comparative materials, they argue, could mitigate
this risk, by facilitating a judicial “gut check” in the dialogic tradition of compara-
tive constitutional engagement, to ascertain which constitutional provisions are truly
essential to what the Supreme Court of India has called a basic structure that is a
constitutive element liberal democratic constitutional order.6
In this brief comment, I want to engage Dixon and Landau on two points. First, I
want to suggest that comparative constitutional law should premise its analysis of the
problem of democratic backsliding not on the problem of unconstitutional amend-
ments, but rather on a broader inventory of the various legal tools used by autocrats
that enable democratic backsliding, which encompasses the abuse of the power of con-
stitutional amendment but extends well beyond it. Second, I think the more likely sce-
nario is not that courts will overreach, but rather that they will fail to call democratic
backsliding by its name, which in turn raises important strategic questions for courts
on whether to embrace the power to declare constitutional amendments and on what
basis, or to calibrate the combination of justication and judicial control differently.
It is natural for Dixon and Landau to focus on the judicial review of constitutional
amendments without any textual foundation to do so, because it represents the
4 Sujit Choudhry, Resisting Democratic Backsliding: An Essay on Weimar, Self-Enforcing Constitutions, and the
Frankfurt School (Unpublished).
5 Samuel Issacharoff, fRagIle demoCRaCIes: Contested PoweR In the eRa of ConstItutIonal CouRts (2015).
6 See Sujit Choudhry, Globalization in Search of Justication: Toward a Theory of Comparative Constitutional
Interpretation 74 IndIana l.J. 819 (1999); VICkI JaCkson, ConstItutIonal engagement In a tRansnatIonal eRa (2009).
828 I•CON 15 (2017), 826–832
ultimate assertion of judicial power. Proposals to relax or remove presidential term
limits are the most visible and common example of constitutional amendments in the
service of democratic backsliding, having generated constitutional conict in recent
years across Sub-Saharan Africa (Burkina Faso, Burundi, Cameroon, Chad, Congo
Brazaville, Democratic Republic of Congo, Gabon, Guinea, Malawi, Namibia, Niger,
Nigeria, Rwanda, Senegal, Togo, Uganda, and Zambia) and Latin America (Colombia,
Ecuador, Honduras, Nicaragua, and Venezuela). But an alternative research strategy
would begin with a basic denition of a consolidated democracy, which is marked by
political competition and the alternation of power, and which generates norms of self-
restraint on the part of the executive. Democratic backsliding occurs when political
competition is impaired, and the checks on executive power are undermined. The idea
would be to reason inductively and empirically from comparative practice to develop
a taxonomy of the policy levers of democratic backsliding and the legal sites where it
occurs, and to build a theory of the judicial role in these contexts on that foundation.
The South African constitutional system is a particularly promising case study. The
African National Congress (ANC) has been in power continually since it won South Africa’s
rst post-Apartheid election over two decades ago in 1994, and since then has held con-
trol at the national level and in most provinces. As I have argued elsewhere, the ANC has
sought to perpetuate its dominant status through a variety of techniques that are charac-
teristic of so-called dominant party democracies, e.g., India under the Congress Party and
Mexico under the Institutional Revolutionary Party (PRI).7 These measures, inter alia, are
the manipulation of electoral systems to fragment opposition political parties; the capture
of independent institutions designed to check the partisan abuse of public administration;
the erosion of federalism as a resource for political competition that enables opposition par-
ties to derive the political benets of incumbency; and the subordination of the democrati-
cally elected parliamentary wing of the party to the unelected non-parliamentary wing.
Specic instances of the ANC’s attempt to entrench its dominant status have come
before the Constitutional Court of South Africa in recent years. These cases have
involved the removal of a ban on oor-crossing which permitted the ANC to recruit
legislators from other parties; a parliamentary rule limiting the right to initiate bills
by minority political parties; the extension of the term of ofce of the Chief Justice
of South Africa by the President; the abolition of an independent anticorruption
authority that had investigated the President and his associates; the appointment of
the independent Director of Public Prosecutions, and the ability of the ANC to block
a secret vote on a motion of non-condence in President Zuma brought by opposition
parties.8 What bears mention here is the variety of legal contexts in which these cases
7 Sujit Choudhry, “He had a mandate”: The South African Constitutional Court and the African National Congress
in a Dominant Party Democracy, 2 Const. Ct. ReV. 1 (2009), available at http://www.constitutionalcourtrev-
8 United Democratic Movement v. President of the Republic of South Africa (No. 2) 2003 1 SA 495 (CC)
(oor-crossing); Oriani-Ambrosini, MP v. Sisulu, MP Speaker of the National Assembly, 2012 (6) SA
588 (CC) (minority political party right to introduce bill); Justice Alliance of South Africa v. President of
Republic of South Africa & Others, 2011 (5) SA 388 (CC) (extension of appointment of Chief Justice);
Democratic Alliance v. President of the Republic of South Africa & Others, [2012] ZACC 24, 2013 (1) SA
Transnational constitutionalism: A reply to Rosalind Dixon and David Landau 829
arose. The oor-crossing case (United Democratic Movement I) turned on a challenge
to a constitutional amendment; but the others turned on constitutional challenges
to an internal rule of the Parliament of South Africa (Oriani-Ambrosini), a statu-
tory delegation of a constitutionally vested appointment power (Justice Alliance), the
statutory framework governing an anticorruption authority (Glenister), the exercise
of a statutory appointment power (Democratic Alliance) and the interpretation of the
Constitution (United Democratic Movement II).
Constitutional amendment, although certainly present, is not where the legal
action has been. And so not surprisingly, the terrain of legal argument has been very
broad, with parties invoking a variety of traditional public law doctrines, such as
rationality review, the non-delegation doctrine, and the good faith application a statu-
tory appointment power inected by broader concerns of constitutional structure, as
well as the interpretation of constitutional provisions such as the Bill of Rights, those
granting minority political party rights in the legislative process, and governing votes
of non-condence. I have argued that traditional doctrines should be reimagined to
respond to the particular problem of the abuse of power by dominant party democra-
cies, and set out an extended explanation of, justication for them: anti-domination,
anti-capture, anti-centralization, and anti-usurpation.9 The salient point here is that
United Democratic Movement I was the one case that raised the problem of constitu-
tional amendment, squarely and centrally. I suggest this is not a contingent detail par-
ticular to the South African case; rather, it raises deeper questions about how courts
should apprehend, and exercise their role in checking, democratic backsliding. Across
a broad variety of constitutional systems that have experienced and face the risk of
democratic backsliding, we should likewise expect to see a diverse range of legal con-
texts extending beyond constitutional amendment as sites of conict and contesta-
tion over the fate of democracy. Substantive checks on the constitutional amendment
power, albeit important, are accordingly not sufcient as legal responses to the full
array of mechanisms employed by would-be autocrats.
Bermeo has recently made a parallel observation from the standpoint of com-
parative politics to which constitutional lawyers should pay heed. She argues that
democratic backsliding requires that we recognize that “democracy is ‘a collage’ of
institutions” that is “put together piece by piece, and can be taken apart the same
way.” This poses a great challenge for scholars of democratization, who have “focused
on clear cases of democratic collapse” as opposed to “slow slides toward authoritari-
anism.” Political scientists must come to grips with this reality “or risk their own slide
into irrelevance.”10 Comparative constitutional law must too ensure that its view-
nder encompasses high prole, dramatic conict over constitutional amendments
but extends beyond it to ordinary constitutional law, administrative law, and statutory
248 (CC) (appointment of Director of Public Prosecutions); Glenister v. President of the Republic of South
Africa & Others, [2011] (3) SA 347 (CC) (independent anticorruption authority); and United Democratic
Movement v. Speaker of the National Assembly, [2017] ZACC 21 (opposition motion of non-condence).
9 Choudhry, supra note 7.
10 Bermeo, supra note 3, at 14.
830 I•CON 15 (2017), 826–832
interpretation, less it fail in its mission to provide sufcient resources to identify and
control these abuses of public power.
This insight shapes the lessons to be drawn from Hungary, perhaps the most vis-
ible example of democratic backsliding in contemporary times. As Kim Scheppele has
perceptively remarked, the Hungarian reversion to authoritarian rule has occurred
through a very large number of individual constitutional amendments, which if
taken on their own, would not raise red ags on the basis that they fall outside the
scope of a basic structure common to constitutional democracies.11 Nonetheless, in
aggregate, they clearly have undermined the democratic character of the Hungarian
constitutional order. Hungary poses a challenge for Dixon and Landau, because of
their focus on the substantive acceptability of constitutional amendments considered
in isolation. However, I would suggest that on their account, Hungary is a relatively
easy case because of the way in which the separate constitutional amendments were
visibly, and publicly, linked closely with each other over a compressed timeframe as
part of an overall project of constitutional transformation. Far more challenging for
them would be lower visibility, non-constitutional change pursued in a more incremen-
tal way. South Africa, not Hungary, is the case that should stimulate reection on how
constitutional law should recognize and respond to abuses of public power that are
part of a process of democratic backsliding.
Dixon and Landau advocate for the strongest possible form of judicial review—the
power of courts to review on substantive grounds, without any textual basis, the
constitutionality of constitutional amendments that threaten democracy—a super-
majoritarian judicial power. Their concern is that courts asserting this power may go
too far by second-guessing constitutional amendments that do not rise to the level of
posing a threat to the constitutional order, and that comparative materials can serve
as anchors to check judicial over-reach by helping courts to induce what is truly fun-
damental, basic or essential to a constitutional system.
Their interesting proposal presupposes that a court recognizes and is willing to assert
its special role in protecting democratic constitutionalism from itself. The Colombian
Constitutional Court is probably the best example of an apex court that conceives of
its function in this way, as the reasoning in its two judgments on presidential term
limits illustrates. The Hungarian Constitutional Court, in the period in between the
coming to power of the Fidesz regime and the adoption of the Fourth Amendment,
also stands as another example as it self-consciously and expressly attempted to stem
the tide of democratic backsliding, as Scheppele has explained.12
However, the orientation of the Colombian and Hungarian constitutional courts
is likely quite exceptional. Thus, the real risk may not be overreach, but under-reach.
Consider the Supreme Court of India, which Dixon and Landau hold out as an
example of an apex court that has asserted its role to check the power of constitutional
11 Kim Lane Scheppele, The Rule of Law and the Frankenstate: Why Governance Checklists do not Work, 26
goVeRnanCe 559 (2013).
12 Kim Lane Scheppele, Constitutional Coups and Judicial Review: How Transnational Institutions can Strengthen
Peak Courts at Times of Crisis, 23 tRansnatl & ContemP. PRoBs. 51 (2014).
Transnational constitutionalism: A reply to Rosalind Dixon and David Landau 831
amendment in the service of constitutional democracy, and which has also over-
reached. The leading case where the Court stood in the way of the antidemocratic
use of the constitutional amendment process is the Indira Gandhi election case.13 The
Supreme Court struck down the 39th Amendment to the Constitution of India, which
had two key features. First, it withdrew the jurisdiction of the ordinary courts over the
conduct of elections of the Prime Minister and the Speaker of the Lok Sabha, autho-
rized Parliament to enact a law to vest authority over electoral disputes with respect
to these two individuals in another body, and immunized that law from constitutional
challenge. Second, the amendment provided that no law made prior to its adoption
applied to the election of the Prime Minister and Speaker, and that any court order
declaring such an election void was itself void and of no effect.
The Court struck down the 39th Amendment on the basis that democracy was an
essential feature of the Indian constitution and that the amendment jeopardized free
and fair elections. But only one Justice (Justice Beg) impugned the amendment on the
basis that it had been enacted to validate Indira Gandhi’s election, and to entrench
the power of the majority Congress Party—that is, it was motivated by personal and
partisan political ends. The other Justices (Chief Justice Raj, Justice Matthew, Justice
Khanna) also held the amendment to be unconstitutional on the basis of democracy
but on different grounds. One line of reasoning held the amendment to be unconsti-
tutional because it declared Gandhi victorious while at the same time repealing the
law pursuant to which she could have been elected; the absence of a law pursuant to
which her election could have occurred raised a formal deciency in the constitutional
amendment. The narrow applicability of the amendment to two political ofcehold-
ers was another formal deciency, because a basic feature of law is its generality. The
other line of reasoning held that free and fair elections in a democracy required the
judicial resolution of electoral disputes, which the amendment contravened.
Thus, most of the justices in Indira Gandhi relied on formal and procedural con-
ceptions of democracy, rather than on a direct attack on the motives underlying the
amendment itself. Iwould suggest that these justices made a strategic calculation that
the formal and procedural notions of democracy offered a narrower and safer basis
for the Court’s ruling, which may have preserved the Court’s institutional capital by
not provoking a direct confrontation with Gandhi. But this may have come at the cost
of establishing a direct link between the Court’s judgment and the way in which the
constitutional harms of the 39th Amendment were understood in the broader con-
stitutional politics. So there was a tradeoff and a choice to be made between these
imperfect alternatives.
South Africa presents another example of a similar set of judicial choices and
tradeoffs. In the cases I set out above, the dangers of a dominant party democracy was
presented to the Constitutional Court by litigants in order to encourage it to be partic-
ularly alert to a broader pattern of the conduct whereby the ANC was entrenching its
13 Indira Nehru Gandhi v. Raj Naraian, 1975 Supp. SCC 1; Sujit Choudhry, How to Do Constitutional Law
and Politics in South Asia, in unstaBle ConstItutIonalIsm: law and PolItICs In south asIa 18 (Mark Tushnet &
Madhav Khosla, eds., 2015).
832 I•CON 15 (2017), 826–832
dominant status through the abuse of public power. The case for framing the argument
this way is that it would allow the Court to call into a question an implicit assump-
tion it applied in many cases—that South Africa is a consolidated liberal democracy in
which political competition and the alternation of power serve as an additional check,
above and beyond judicial review, on the abuse of political power, and that courts
could therefore leave some kinds of issues to the political process. However, while the
Constitutional Court has increasingly shown a willingness to check the abuse of pub-
lic power by the ANC government, it did so without resort to a substantive theory of
its role in preventing democratic backsliding. Theunis Roux has argued this enabled
the Court to preserve its institutional independence against the risk of backlash by a
powerful government.14
Let me conclude by linking the pervasiveness of means short of constitutional
amendment in democratic backsliding, and an unwillingness of some—perhaps
most—apex courts to call these abuses of public power by their name. It could very
well be that inviting courts to assert the ultimate power of judicial review over consti-
tutional amendments, and to use the most politically charged justication for doing
so, is strategically unwise in fraught political contexts where the very political forces
that create the risk of backsliding also pose a risk to the court itself. But principle need
not yield to pragmatism in every circumstance. An easier context for the court to use
this kind of justication could be cases where the power of judicial review is rela-
tively normal and not exceptional—that is, the traditional terrain of constitutional
and administrative law and statutory interpretation and not the substantive review
of constitutional amendments. Indeed, perhaps courts should forestall the ultimate
clash over constituent power by focusing their energies on these kinds of cases. The
dialogic engagement with comparative materials could assist courts in grappling with
these dilemmas.
14 Theunis Roux, The South African Constitutional Court’s Democratic Rights Jurisprudence, 5 Const. Ct. ReV.
33 (2014), available at
... Although deadlines for holding elections are often entrenched in a country's legal or constitutional framework, international law allows for the derogation of some rights -including the right to vote and stand for election -in emergency situations with stringent rules to prevent abuse . Some countries have either constitutional provisions or/and legal rules to entertain election postponements due to emergency situations (Choudhry, 2017). During state of emergency, the Indian 7 Later TPLF, a party that administers one of the ten regions of Ethiopia, waged full scale war with the federal government over the disagreement of the election postponement and held its own regional elections. ...
Full-text available
Election postponement in Ethiopia, due to the COVID-19 pandemic, has raised critical constitutional questions that have never been really thought before in the country’s constitutional law jurisprudence. This is because the state of emergency measure in Ethiopia, due to the spread of COVID-19, is in conflict with constitutional deadlines for elections. The constitutional lacuna is complicated by the absence of explicit constitutional provisions that indisputably govern election postponement. Although any legal measures to postpone election schedule and pass constitutional deadlock is far from simple, the Ethiopian government has suggested four possible ‘options’ to the constitutional dilemma: dissolving the parliament, declaring state of emergency, amendment of the Constitution and constitutional interpretation. Finally, the House of Federation (HoF), the Ethiopian upper House entrusted to interpret the constitution decided and postponed the election indefinitely until the pandemic no longer poses a risk to public health confirmed by the parliament which has direct vested interest in the outcome. This research investigates whether the constitutional interpretation option adheres to the premises of the Ethiopian Constitution or is it extra constitutional. Accordingly, the HoF provided superficial analysis and fallacious reasoning and failed to meaningfully grapple with the serious constitutional issues. The constitutional interpretation is not constitutionally bound and is defective. The manner the HoF managed the constitutional vacuum concerning election postponement, indisputably, was ‘constitutional interpretation’ by name but a political decision in practice.
The unconstitutional constitutional amendment doctrine has emerged as a highly successful, albeit still controversial, export in comparative constitutional law. The doctrine has often been defended as protecting a delegation from the people to the political institutions that they created. Other work has noted the doctrine’s potential utility in guarding against abusive constitutionalism. In this article, we consider how these justifications fare when expanded to encompass claims against the original constitution itself, rather than a later amendment to the text. That is, beyond the unconstitutional constitutional amendment doctrine, can or should there be a doctrine of an unconstitutional constitution? Our question is spurred by a puzzling 2015 case from Honduras where the Supreme Court held an unamendable one-term limit on presidential terms, as well as protective provisions punishing attempts to alter that limit, to be unconstitutional. What is particularly striking about the case is that these provisions were not later amendments to the constitution, but rather parts of the original 1982 constitution itself. Thus, this article examines the possibility of ‘an unconstitutional constitution’, what we predict to be the next trend in global constitutionalism.
ResearchGate has not been able to resolve any references for this publication.