W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
913
THE STRUCTURE OF CONSTITUTIONAL
AMENDMENT RULES
Richard Albert
INTRODUCTION
Formal constitutional amendment rules are the gatekeepers to
the constitutional text. They detail the procedures for changing the
written constitution,
1
specify what is subject to or immune from
formal amendment,
2
promote deliberation about constitutional
meaning,
3
distinguish the constitutional text from ordinary law,
4
and may also be designed to express constitutional values.
5
Formal
amendment rules are especially useful for channeling popular will
into institutional dialogue
6
and checking informal constitutional
amendments.
7
By their nature, formal amendment rules reflect
both faith and distrust in political actors: they simultaneously
Assistant Professor, Boston College Law School; Yale University (J.D.,
B.A.); Oxford University (B.C.L.); Harvard University (LL.M.). Email:
Brannon Denning, Rosalind Dixon, Oran Doyle, Brian Galle, Frank Garcia,
Tom Ginsburg, Claudia Haupt, Eugene Mazo, William Partlett, Arie Rosen,
Yaniv Roznai, Ozan Varol, John Vile, and workshop participants at Boston
College, the University of San Francisco, and at the 201314 Annual Meetings
of the Law & Society Association and the Association of American Law Schools.
I am also grateful to the editorial staff of the Wake Forest Law Review—
particularly to Christopher Salera, Caroline Massagee, Isabella Shaw, Kim
Sokolich, and Denton Worrell—for their exceptional editorial assistance in
preparing this Article for print.
1
. Rosalind Dixon & Richard Holden, Constitutional Amendment Rules:
The Denominator Problem, in COMPARATIVE CONSTITUTIONAL DESIGN 195, 195
(Tom Ginsburg ed., 2012).
2
. Jon Elster, Constitutionalism in Eastern Europe: An Introduction, 58 U.
CHI. L. REV. 447, 471 (1991).
3
. Raymond Ku, Consensus of the Governed: The Legitimacy of
Constitutional Change, 64 FORDHAM L. REV. 535, 571 (1995).
4
. ANDRÁS SAJÓ, LIMITING GOVERNMENT: AN INTRODUCTION TO
CONSTITUTIONALISM 39–40 (1999).
5
. Richard Albert, The Expressive Function of Constitutional Amendment
Rules, 59 MCGILL L.J. 225, 236 (2013).
6
. Walter Dellinger, The Legitimacy of Constitutional Change: Rethinking
the Amendment Process, 97 HARV. L. REV. 386, 431 (1983).
7
. Rosalind Dixon, Constitutional Amendment Rules: A Comparative
Perspective, in COMPARATIVE CONSTITUTIONAL LAW 96, 97 (Tom Ginsburg &
Rosalind Dixon eds., 2011).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
914 WAKE FOREST LAW REVIEW [Vol. 49
authorize political actors to improve the constitution
8
while limiting
how and when political actors may do so.
9
Given the many essential
functions formal amendment rules serve, we would expect
constitutions to entrench them, and indeed most of them do.
10
Yet the structure of formal amendment rules has received little
scholarly attention. Scholars have devoted considerably more
attention to informal amendment,
11
which we can define as “the
alteration of constitutional meaning in the absence of textual
change.”
12
Though constitutions may today change less frequently
via formal rather than informal amendment, constitutional
designers must nonetheless understand the formal amendment
options available to them in order to structure the formal
amendment rules they entrench in the constitutions they design. It
is therefore surprising that constitutional designers have few
academic resources to explain how to design the rules governing
formal amendment, which John Burgess called in 1890 “the most
important part of a constitution,”
13
and which Akhil Amar has more
recently described as holding “unsurpassed importance, for these
rules define the conditions under which all other constitutional
norms may be legally displaced.”
14
8
. Brannon P. Denning & John R. Vile, The Relevance of Constitutional
Amendments: A Response to David Strauss, 77 TUL. L. REV. 247, 275 (2002).
9
. Donald J. Boudreaux & A.C. Pritchard, Rewriting the Constitution: An
Economic Analysis of the Constitutional Amendment Process, 62 FORDHAM L.
REV. 111, 123–24 (1993).
10
. Bjørn Erik Rasch & Roger D. Congleton, Amendment Procedures and
Constitutional Stability, in DEMOCRATIC CONSTITUTIONAL DESIGN AND PUBLIC
POLICY 319, 325 (Roger D. Congleton & Birgitta Swedenborg eds., 2006).
11
. See, e.g., BRUCE ACKERMAN, WE THE PEOPLE: FOUNDATIONS 266–94
(1991) (developing the theory of “constitutional moments”); BRUCE ACKERMAN,
WE THE PEOPLE: TRANSFORMATIONS 383–420 (1998) (developing the same
theory); JEREMY WEBBER, REIMAGINING CANADA: LANGUAGE, CULTURE,
COMMUNITY, AND THE CANADIAN CONSTITUTION 260–305 (1994) (discussing
informal amendment in Canada); Brannon P. Denning, Means to Amend:
Theories of Constitutional Change, 65 TENN. L. REV. 155, 180–209 (1997)
(surveying theories of informal amendment); Rosalind Dixon, Partial
Constitutional Amendments, 13 U. PA. J. CONST. L. 643, 664–80 (2011)
(developing the concept of “partial constitutional amendment”); Peter Oliver,
Canada, Quebec, and Constitutional Amendment, 49 U. TORONTO L.J. 519, 526–
31 (1999); David A. Strauss, The Irrelevance of Constitutional Amendments, 114
HARV. L. REV. 1457, 1469–86 (2001). See generally RESPONDING TO
IMPERFECTION: THE THEORY AND PRACTICE OF CONSTITUTIONAL AMENDMENT
(Sanford Levinson ed., 1995) (compiling essays on constitutional change).
12
. Heather K. Gerken, The Hydraulics of Constitutional Reform: A
Skeptical Response to Our Undemocratic Constitution, 55 DRAKE L. REV. 925,
929 (2007).
13
. JOHN BURGESS, I POLITICAL SCIENCE AND COMPARATIVE CONSTITUTIONAL
LAW 137 (1890).
14
. Akhil Reed Amar, The Consent of the Governed: Constitutional
Amendment Outside Article V, 94 COLUM. L. REV. 457, 461 (1994).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 915
Scholars have developed criteria to classify formal amendment
rules. Their classifications help constitutional designers compare
formal amendment rules across nations. Yet these classifications
are either over-inclusive, under-inclusive, or both. As I will
demonstrate, some classify formal amendment rules according only
to voting thresholds, thereby failing to account for nonvoting criteria
that often govern formal constitutional change. Others classify
formal amendment rules according to both voting thresholds and
nonvoting criteria, but they fail to appreciate the larger structures
around which formal amendment rules are built. Still others
classify constitutions whose textual democratic commitments serve
only to cloak nondemocratic practices. These classifications may be
interesting, but they are largely unhelpful to democratic
constitutional designers. Although sham constitutions entrench
formal amendment rules, their entrenched rules do not actually bind
political actors, nor do citizens accept them as accurate and
legitimate reflections of how power is actually exercised and
constrained.
15
Insofar as democratic constitutional design should
reflect democratic practices in fact, formal amendment
classifications are useful only where they exclude sham
constitutions.
In this Article, I construct a new classification of formal
amendment rules based on my analysis of formal amendment rules
in the world’s highest-performing democratic countries. I identify
the world’s leading democracies from the annual Democracy Index
published by the Economist Intelligence Unit.
16
The 2012
Democracy Index “provides a snapshot of the state of democracy
worldwide for 165 independent states and two territories—this
15
. Giovanni Sartori, Constitutionalism: A Preliminary Discussion, 56 AM.
POL. SCI. REV. 853, 861–62 (1962).
16
. THE ECONOMIST INTELLIGENCE UNIT, DEMOCRACY INDEX 2012:
DEMOCRACY AT A STANDSTILL (2013), available at http://www.eiu.com/public
/topical_report.aspx?campaignid=DemocracyIndex12 (last visited Apr. 25,
2014). I have found the Economist Intelligence Unit’s Democracy Index more
informative and comprehensive than Freedom House’s own index, which,
though perhaps more commonly used in studies of democracy, contains fewer
democratic indicators. Specifically, the Freedom House index reports a
country’s freedom status (free, partly free, or not free) and its score on a scale of
one to seven on political rights and civil liberties. See FREEDOM HOUSE,
FREEDOM IN THE WORLD 2013: DEMOCRATIC BREAKTHROUGHS IN THE BALANCE
(2013), available at http://www.freedomhouse.org/sites/default/files/FIW
%202013%20Booklet.pdf (last visited Apr. 25, 2014). In contrast, the Economist
Intelligence Unit’s Democracy Index classifies countries in one of four categories
(full democracies, flawed democracies, hybrid regimes, and authoritarian
regimes), ranks countries with an overall score on a descending scale, and also
quantifies a country’s performance relative to other countries on five democratic
indicators (electoral process and pluralism, functioning of government, political
participation, political culture, and civil liberties). See THE ECONOMIST
INTELLIGENCE UNIT, supra.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
916 WAKE FOREST LAW REVIEW [Vol. 49
covers almost the entire population of the world and the vast
majority of the world’s states (micro states are excluded).”
17
The
2012 Democracy Index allocates to each country a score on five
measures of democracy—electoral process and pluralism, civil
liberties, the functioning of government, political participation, and
political culture—and assigns a total score and overall rank for
every country included in the study. Each country is also classified
under one of four types: full democracies, flawed democracies, hybrid
regimes, or authoritarian regimes. I examine the constitutions of
countries classified as either full or flawed democracies with overall
scores above 7.5 out of ten—thirty-six countries in total—to
construct this new classification of formal amendment rules.
18
My
17
. THE ECONOMIST INTELLIGENCE UNIT, supra note 16, at 1.
18
. The Democracy Index defines full democracies as
[c]ountries in which not only basic political freedoms and civil liberties
are respected, but these will also tend to be underpinned by a political
culture conducive to the flourishing of democracy. The functioning of
government is satisfactory. Media are independent and diverse.
There is an effective system of checks and balances. The judiciary is
independent and judicial decisions are enforced. There are only
limited problems in the functioning of democracies.
Id. at 28. Flawed democracies “also have free and fair elections and even if
there are problems (such as infringements on media freedom), basic civil
liberties will be respected. However, there are significant weaknesses in other
aspects of democracy, including problems in governance, an underdeveloped
political culture and low levels of political participation.” Id. I exclude hybrid
regimes where
[e]lections have substantial irregularities that often prevent them
from being both free and fair. Government pressure on opposition
parties and candidates may be common. Serious weaknesses are more
prevalent than in flawed democracies—in political culture, functioning
of government and political participation. Corruption tends to be
widespread and the rule of law is weak. Civil society is weak.
Typically there is harassment of and pressure on journalists, and the
judiciary is not independent.
Id. I also exclude authoritarian regimes:
In these states state political pluralism is absent or heavily
circumscribed. Many countries in this category are outright
dictatorships. Some formal institutions of democracy may exist, but
these have little substance. Elections, if they do occur, are not free
and fair. There is disregard for abuses and infringements of civil
liberties. Media are typically state-owned or controlled by groups
connected to the ruling regime. There is repression of criticism of the
government and pervasive censorship. There is no independent
judiciary.
Id. Given that my study examines formal amendment rules, I exclude full and
flawed democracies with unwritten constitutions or constitutions without one or
more master texts (like Israel and the United Kingdom), but I include
democracies whose constitutions include one or more comprehensive
constitutional acts (for instance, Canada and New Zealand) and where formal
amendment occurs via constitutional act or fundamental law (for example,
Austria, the Czech Republic, and Denmark).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 917
analysis considers only current, not historical or superseded,
democratic constitutions.
The purpose of this Article is twofold. My main purpose is to
probe the structure and uses of formal amendment rules in
constitutional democracies. I deconstruct the architecture of formal
amendment rules in constitutional democracies, as well as the ways
political actors deploy those rules. This Article endeavors to offer
constitutional designers in both democratic and democratizing
states a comprehensive roadmap to build or refine their own formal
amendment rules.
19
My second purpose is to generate an agenda for
further research into the structure of formal amendment. I identify
patterns, similarities, and distinctions in order to generate renewed
interest in the study of formal amendment from perspectives in
constitutional design, law, history, and theory. This Article is
therefore both an inquiry into formal amendment and an invitation
to further research.
In Part I, I begin by explaining why existing formal amendment
classifications are useful but lacking. In Part II, I create a new
classification to respond to the weaknesses of existing ones. I posit
that formal amendment rules are conceptually structured in three
tiers: foundations, frameworks, and specifications. I suggest that
formal amendment rules are anchored either explicitly or implicitly
in the foundational distinction between constitutional amendment
and revision. I also demonstrate that they operate pursuant to one
of six amendment frameworks: comprehensive single-track,
comprehensive multi-track, restricted single-track, restricted multi-
track, exceptional single-track, or exceptional multi-track. I then
show that formal amendment rules moreover consist of
distinguishable but combinable specifications that supplement their
foundations and frameworks: voting thresholds and quorum
requirements, content restrictions, temporal requirements, electoral
preconditions, and defense mechanisms.
Part III then illustrates how constitutional designers may use
this three-tiered classification of formal amendment rules to achieve
such objectives as managing federalism, expressing constitutional
values, enhancing or diminishing the judicial role, and pursuing
democratic outcomes related to governance, constitutional
endurance, and amendment difficulty. I conclude with suggestions
for future research into the design of formal amendment rules.
19
. More cynically, constitutional autocrats could exploit the structure of
formal amendment rules in constitutional democracies to clothe their regime in
the legitimacy of entrenched amendment rules. See KARL LOEWENSTEIN,
POLITICAL POWER AND THE GOVERNMENTAL PROCESS 136 (2d ed. 1965) (“So deeply
implanted is the conviction that a sovereign state must possess a written
constitution that even modern autocracies feel compelled to pay tribute to the
democratic legitimacy inherent in the written constitution.”).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
918 WAKE FOREST LAW REVIEW [Vol. 49
I. AMENDMENT CLASSIFICATIONS
Existing formal amendment classifications have proven to be
useful for limited purposes. Some focus principally on the voting
thresholds that distinguish one constitution’s formal amendment
rules from another. These classifications are helpful in
understanding how constitutions incorporate majoritarian or
supermajoritarian procedures in formal amendment.
20
Other
classifications are attentive to both voting thresholds and nonvoting
criteria in formal amendment rules, and accordingly illustrate how
both may be used in combination to design formal amendment
rules.
21
Still others classify formal amendment rules in tandem
with methods of informal amendment in an effort to demonstrate
the relationship between the emergence and frequency of informal
amendment and the flexibility or rigidity of formal amendment
rules.
22
Yet all three types of existing formal amendment
classifications are incomplete. None serves as a comprehensive
guide for constitutional designers to understand how formal
amendment rules are structured or to operationalize that structure
in designing their own rules of formal amendment.
A. Voting Thresholds
Formal amendment rules may be classified according to the
quantum of agreement needed to alter the constitutional text. All
formal amendment rules in some way incorporate ordinary or
extraordinary voting thresholds by representative assemblies or
citizens. Arend Lijphart has classified formal amendment rules
according to voting thresholds in thirty-six countries identified as
democracies by Freedom House.
23
Lijphart identifies four categories
of voting thresholds in his classification: approval by ordinary
majority, two-thirds majority, less than two-thirds majority but
more than an ordinary majority, and more than two-thirds
majority.
24
Lijphart finds that a plurality of the countries in his
sample—fifteen of the thirty-six countries—entrench formal
amendment rules requiring a two-thirds majority or its equivalent.
25
He finds that the next most popular are approval by supermajority
greater than two-thirds and by supermajority greater than an
ordinary majority but lower than two-thirds,
26
and that the least
20
. See infra Subpart I.A.
21
. See infra Subpart I.B.
22
. See infra Subpart I.C.
23
. AREND LIJPHART, PATTERNS OF DEMOCRACY 47–48 (2d ed. 2010).
24
. Id. at 207. In an addendum in the new edition of his study, Lijphart
notes that fourteen countries use referenda as an absolute requirement of, or an
optional alternative to, formal amendment rules. Id. at 219.
25
. Id. at 208.
26
. Id.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 919
common voting threshold for formal amendment is approval by
ordinary majority.
27
Lijphart’s classification is helpful to constitutional designers for
three primary reasons. First, it orders three dozen constitutions
whose formal amendment rules vary in many ways. Second, it
demonstrates how voting thresholds contribute to a constitution’s
flexibility and rigidity.
28
Lijphart contrasts how authorizing regular
majorities to formally amend the constitution leads to constitutional
flexibility, whereas limiting formal amendment to extraordinary
majorities fosters constitutional rigidity.
29
Third, Lijphart also
connects constitutional malleability with the strength of judicial
review. He posits that “judicial review can work effectively only if it
is backed up by constitutional rigidity and vice versa,”
30
meaning
that a judicial ruling is more likely to be durable where the rules of
formal amendment are difficult. He also uses his classification to
suggest that “completely flexible constitutions and the absence of
judicial review permit unrestricted majority rule.”
31
His
classification is instructive.
Nevertheless, Lijphart’s classification cannot serve as a
complete guide for constitutional designers planning to design their
own formal amendment rules. It does not account for the complexity
of formal amendment rules, and its generality misses important
distinctions even within the various voting thresholds according to
which he classifies formal amendment rules. For example,
Lijphart’s classification does not consider limits to formal
amendment rules such as unamendable constitutional provisions or
periods of time during which the formal amendment rules are
suspended—namely in emergencies or during the formal
amendment process itself. Constitutional designers cannot fully
understand how to structure formal amendment rules without
appreciating how these and other specifications fit within the
amendment process.
Moreover, Lijphart acknowledges that his classification does not
reflect nuances within particular voting thresholds. Specifically,
Lijphart concedes that his classification does not account for
different rules of formal amendment in the same constitution.
32
Lijphart has a two-part solution for this problem: “first, when
alternative methods can be used, the least constraining methods
should be counted,”
33
and “second, when different rules apply to
different parts of constitutions, the rule pertaining to amendments
27
. Id.
28
. Id. at 206–11.
29
. Id. at 204.
30
. Id. at 219.
31
. Id.
32
. Id. at 209.
33
. Id.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
920 WAKE FOREST LAW REVIEW [Vol. 49
of the most basic articles of the constitution should be counted.”
34
Lijphart’s classification therefore does not classify the multiple
routes of formal amendment entrenched in a single constitution, nor
does it classify the different formal amendment rules that apply to
different sections of a constitution. Yet alternative methods of
amendment and exclusive or specially assigned amendment rules
are fundamental distinctions among formal amendment rules that
must be reflected in any classification if the purpose of the
classification is to serve as a complete guide for constitutional
designers in designing formal amendment rules of their own.
Lijphart’s second solution—to apply the amendment rule for “the
most basic articles”—compounds the problem with either
subjectivity or error. Sometimes formal amendment rules entrench
alternative, exclusive, or specially assigned amendment rules. That
Lijphart does not classify these rules is problematic for identifying
what he refers to as “the most basic articles” of the constitutions in
his classification.
Edward Schneier has developed a similar classification based on
voting thresholds.
35
Schneier classifies 101 constitutions into five
categories and a total of nineteen subcategories. His five main
categories represent the voting thresholds pursuant to which the
legislature may initiate a formal amendment: simple majority, sixty
percent, sixty-five percent, two-thirds, and seventy-five percent.
36
Schneier divides each of these five categories into one of six
subcategories that represent the methods and requirements for
ratification (for instance, no further action, executive approval, or
referendum).
37
His classification, which is presented in a useful
table, therefore allows us quickly to see that formal amendment in
Argentina, for example, requires a legislative vote by a two-thirds
majority with no further action, or that Russia requires a legislative
vote of sixty percent followed by a constitutional convention.
38
Schneier’s objective is both to summarize how constitutions may
be formally amended and to identify patterns in formal amendment.
His classification only partly achieves its objectives. Schneier
concedes that his classification is limited because it “glosses over
important nuances” in formal amendment rules.
39
Although the
classification does summarize methods of formal amendment, it is
not complete. As Schneier acknowledges, his classification does not
reflect the full range of formal amendment methods: “many
constitutions provide alternative methods including, most
34
. Id.
35
. EDWARD SCHNEIER, CRAFTING CONSTITUTIONAL DEMOCRACIES: THE
POLITICS OF INSTITUTIONAL DESIGN 222–25 (2006).
36
. Id. at 224–25.
37
. Id.
38
. Id.
39
. Id. at 223.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 921
frequently, referendums. This table reflects what I believe to be
common practice of the method most frequently used in each
country.”
40
This raises a second limitation to Schneier’s
classification: it is time bound. That one method of formal
amendment may today be common practice does not mean that it
will remain common practice. Infrequently used methods may
become more frequently used, just as more frequently used methods
may lapse into disuse.
41
Relatedly, the implication of excluding
infrequently used methods of formal amendment is to exclude
certain formal amendment methods from his classification.
The consequence of abstracting from specific rules in search of
larger patterns is sometimes to elide over important details.
Schneier’s classification is vulnerable to this criticism. For example,
Schneier classifies the United States under the two-thirds category
with ratification by three-quarters of states.
42
But this classification
does not reflect Article V’s prohibition on formal amendments to the
Equal Suffrage Clause without the consent of the affected state.
43
Similarly, Schneier classifies Canada and South Africa under the
simple majority vote and seventy-five percent categories,
respectively, and under the subcategories of approval by provincial
legislatures or constitutional convention and approval by majority of
provinces, respectively.
44
Yet this classification does not reflect the
escalating structure of both Canada and South Africa’s respective
formal amendment rules, whose intricate design distinguishes it
from many other constitutions.
45
In addition, Schneier includes
sham constitutions in his classification.
46
Although sham
constitutions entrench formal amendment rules, those rules do not
actually bind political actors, nor do citizens accept them as accurate
40
. Id. at 225.
41
. I have elsewhere explored whether constitutional provisions may expire
as a result of nonuse. See Richard Albert, Constitutional Amendment by
Constitutional Desuetude, 62 AM. J. COMP. L. (forthcoming 2014) (developing the
theory that written constitutions may be informally amended by desuetude);
Richard Albert, Constitutional Disuse or Desuetude: The Case of Article V, 94
B.U. L. REV. 1029 (2014) (inquiring whether Article V has fallen into disuse or
desuetude as a result of its unsuccessful uses over the past generation).
42
. SCHNEIER, supra note 35, at 225.
43
. U.S. CONST. art. V.
44
. SCHNEIER, supra note 35, at 224–25.
45
. See infra Subpart II.B.
46
. Schneier includes countries such as Iran, Syria, Eritrea, Pakistan, and
Sudan in his classification. See SCHNEIER, supra note 35, at 224–25. Some of
the countries in Schneier’s study qualify as “competitive authoritarian regimes”
where “formal democratic institutions are widely viewed as the principal means
of obtaining and exercising political authority,” but “[i]ncumbents violate those
rules so often and to such an extent, however, that the regime fails to meet
conventional minimum standards for democracy.” Steven Levitsky & Lucan A.
Way, Elections Without Democracy: The Rise of Competitive Authoritarianism,
13 J. DEMOCRACY 51, 52 (2002).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
922 WAKE FOREST LAW REVIEW [Vol. 49
and legitimate reflections of how power is actually exercised and
constrained.
47
Schneier’s classification nevertheless identifies useful patterns
in formal amendment. Schneier finds that most formal amendment
procedures require a confirmatory vote, do not involve the executive
branch, and establish a modal legislative vote of two-thirds.
48
But
he concludes that “the most interesting pattern” from his analysis
“is that there are few discernible patterns.”
49
Schneier observes that
“[c]ommonwealth countries excepted, the rules seem neither more
nor less restrictive in parliamentary as opposed to presidential
systems, in older as against newer democracies, or by regions of the
world.”
50
That Schneier finds few deployable patterns of formal
amendment design weakens the usefulness of his classification for
constitutional designers seeking to understand how to structure
their own rules of formal amendment.
B. Voting Thresholds and Nonvoting Criteria
Formal amendment rules may also be classified under a
combination of voting thresholds and nonvoting criteria. These
classifications are more instructive to constitutional designers than
those that classify formal amendment rules according only to voting
thresholds because they offer a fuller view of amendment rules and
illustrate the interrelations between voting thresholds and
nonvoting criteria. In his classification of formal amendment rules,
Jon Elster identifies what he calls six “main hurdles” to formal
amendment: absolute entrenchment, supermajority approval, higher
quorum requirements, delays, subnational ratification, and
referenda.
51
These six categories represent both voting thresholds,
which include supermajority approval, higher quorum requirements,
subnational ratification and referenda, as well as nonvoting criteria
that include absolute entrenchment and delays. Elster’s
classification therefore differs from Lijphart and Schneier’s
respective classifications, neither of which uses a nonvoting criterion
as a principal category in its classification of formal amendment
rules.
Elster constructs his classification to demonstrate the features
that make constitutions more difficult to change than ordinary
laws.
52
Each of the six categories he identifies serves the function,
either by design or effect, of controlling the pace of constitutional
47
. David S. Law & Mila Versteeg, Sham Constitutions, 101 CALIF. L. REV.
863, 880 (2013).
48
. SCHNEIER, supra note 35, at 223.
49
. Id.
50
. Id.
51
. JON ELSTER, ULYSSES UNBOUND: STUDIES IN RATIONALITY,
PRECOMMITMENT, AND CONSTRAINTS 101 (2000).
52
. Id. at 100–01.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 923
change. For instance, some constitutions absolutely entrench rights
or structures to prohibit formal amendment.
53
Similarly, both
supermajority approval and higher quorum requirements complicate
formal amendment, although unlike absolute entrenchment they do
not make it impossible.
54
Inserting delays in the process of formal
amendment—for example, by requiring successive votes or imposing
time limitations—likewise makes constitutions more difficult to
formally amend than ordinary laws, as does requiring either
subnational or referendal ratification.
55
A similar effort to classify formal amendment rules is Jan-Erik
Lane’s six-part classification.
56
Five of Lane’s six criteria overlap
with Elster’s criteria, the only exception being Lane’s confirmatory
votes in the place of Elster’s higher quorum requirements.
57
Lane’s
classification shows that the frequency of formal amendment is
indirectly related to the constitutionally entrenched mechanisms
favoring constitutional inertia.
58
The more specific the rules of
formal amendment and the more numerous the amendment
mechanisms entrenched in the text, the greater the likelihood of
constitutional inertia, argues Lane.
59
These mechanisms are the
criteria according to which Lane classifies formal amendment rules:
unamendable rules, referenda, delays, qualified majorities,
subnational ratification, and confirmatory votes.
60
In addition, Donald Lutz has classified formal amendment rules
in his larger effort to measure amendment difficulty in thirty-two
countries.
61
Lutz identifies four general formal amendment
strategies around which constitutional designers structure their
formal amendment rules. The first, legislative supremacy, reflects
“unbridled dominance of the legislature by one legislative vote
sufficient to amend the constitution.”
62
The second, which he calls
intervening election, requires the national legislature to vote to
approve a formal amendment in two separate sessions divided by an
election.
63
The third, legislative complexity, is “usually
characterized by multiple paths for the amending process, which
53
. Id. at 102.
54
. Id.
55
. Id. at 103.
56
. JAN-ERIK LANE, CONSTITUTIONS AND POLITICAL THEORY 41 (2d ed. 2011).
57
. Compare id., with ELSTER, supra note 51, at 102.
58
. LANE, supra note 56.
59
. Id.
60
. Id.
61
. DONALD S. LUTZ, PRINCIPLES OF CONSTITUTIONAL DESIGN 170 (2006).
Lutz compares national constitutions with state constitutions, but I hesitate to
compare them insofar as subnational constitutions are associated with weaker
government structures and rights as well as weaker formal amendment rules.
See Tom Ginsburg & Eric A. Posner, Subconstitutionalism, 62 STAN. L. REV.
1583, 1601 (2010).
62
. LUTZ, supra note 61, at 174.
63
. Id. at 175.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
924 WAKE FOREST LAW REVIEW [Vol. 49
features the possibility of a referendum as a kind of threat to bypass
the legislature.”
64
The fourth amendment strategy is required
referendum, which is used in constitutional states that
institutionalize a referendum as a method of formal amendment.
65
Constitutional designers will find Elster’s, Lane’s, and Lutz’s
classifications useful. Elster explains with examples how those six
categories may be combined to construct formal amendment rules.
66
He also advises designers that supermajority approval and delays
are the most important categories for constitutional
precommitment.
67
Lane’s classification is a particularly useful
resource for constitutional designers because it demonstrates how
countries use multiple amendment mechanisms in many different
combinations for different purposes, namely to protect minorities or
to single out certain laws as special.
68
Lutz’s classification evaluates
formal amendment rules with an interest in understanding what
contributes to amendment ease and difficulty, how to balance a
written constitution’s flexibility and stability, and the extent to
which one amendment strategy affects the amendment rate.
69
Each
of these classifications is a valuable resource for the design of formal
amendment rules.
But all three classifications remain incomplete. Although they
identify both voting thresholds and nonvoting criteria, they do not
identify the larger structures around which formal amendment rules
are built. Voting thresholds and nonvoting criteria are only part of
the architecture of formal amendment rules insofar as they operate
within deeper amendment foundations and frameworks that arise
prior to the constraining effect that voting thresholds and nonvoting
criteria exert. Formal amendment rules are therefore anchored in
underlying foundations and frameworks that Elster, Lane, and Lutz
do not identify, and which would be helpful for constitutional
designers to understand before designing their formal amendment
rules. My objective in this Article is to illustrate and explain how
formal amendment rules are structured, both in the ways that
Elster, Lane, and Lutz identify and in those that they do not.
C. Conceptual Categories
Scholars have also ventured beyond the formal rules of
constitutional amendment to posit broader conceptual categories of
constitutional change. Xenophon Contiades and Alkemen Fotiadou
have developed a comprehensive conceptual classification of
constitutional amendment comprising five models of constitutional
64
. Id. at 176.
65
. Id. at 364.
66
. ELSTER, supra note 51, at 103.
67
. Id. at 104.
68
. See LANE, supra note 56, at 41–43.
69
. LUTZ, supra note 61, at 176–77.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 925
change: elastic, evolutionary, pragmatic, distrust, and direct-
democratic.
70
These models are descriptive, neither normative nor
mutually exclusive, and reflect the basic features of amendment in
constitutional regimes for the purpose of comparing their functional
advantages and limitations.
71
As models, they are particularly
useful in illustrating the interrelation between formal and informal
amendment.
Contiades and Fotiadou describe the elastic model of
constitutional change as “operating under an unentrenched
constitution, which may be altered through the normal lawmaking
process, having no procedural limits and no eternity clauses.”
72
Anchored in parliamentary supremacy, the elastic model makes the
legislative branch all powerful because “no obstacles to revision exist
other than self-restraint flowing from legal culture, tradition, and
political accountability.”
73
The United Kingdom’s unwritten
constitution is the paradigm of the elastic model.
74
In contrast, the
evolutionary model is characterized by a strong judiciary, a high
incidence of informal change, and rigid formal amendment rules.
75
Under this model, the difficulty of formal amendment prompts and
legitimates informal methods of amendment ranging from
constitutional revolutions to incremental alterations: “Dynamic
interpretation lies at the heart of that model, where constitutional
change is meticulously construed through legal reasoning as befits
judicial justification. Judge-made change may not be attributed
exclusively to the judge; political elites or the people might be the
driving force behind judicial constitutional evolution.”
76
Canada and
the United States are two leading examples of this model.
77
Constitutional regimes under the pragmatic model generally
resort to formal amendment for constitutional change. Contiades
and Fotiadou explain that “[t]he pragmatic model allows
constitutional change to take place smoothly, with efficiency of
formal change being the most striking feature of the way the system
works.”
78
Formal amendment rules are not usually difficult under
this model. But even if they are stringent, the consensual political
culture facilitates necessary formal amendments: “The amending
formula may be demanding, designed to secure constitutional
stability; nevertheless, constant change is feasible due to a
70
. Xenophon Contiades & Alkmene Fotiadou, Models of Constitutional
Change, in ENGINEERING CONSTITUTIONAL CHANGE: A COMPARATIVE PERSPECTIVE
ON EUROPE, CANADA AND THE USA 417, 440–57 (Xenophon Contiades ed., 2013).
71
. Id. at 441.
72
. Id.
73
. Id.
74
. Id. at 442.
75
. Id. at 442–43.
76
. Id. at 443.
77
. Id.
78
. Id. at 445.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
926 WAKE FOREST LAW REVIEW [Vol. 49
consensual constitutional ethos.”
79
Whereas the evolutionary model
relies less on formal amendment than judicial interpretation to keep
the constitution current, the pragmatic model relies less on the
judiciary than formal amendment: “A strong judiciary is a feature of
this model; nevertheless, there is no need for constitutional review
to operate as a substitute for formal amendment.”
80
In other words,
“[J]udicial interpretation plays a complementary role and is not the
primary vehicle of change,”
81
conclude Contiades and Fotiadou.
Germany illustrates this model.
82
The distrust and direct-democratic models of constitutional
change are opposites. The distrust model incorporates unamendable
constitutional provisions, complicated formal amendment rules, and
elite ratification into its procedures for constitutional change:
“Demanding and complex amending formulas, political elite-driven
change, and difficulty in reaching compromises on constitutional
issues are the basic ingredients of the distrust model.”
83
When
entrenched within a polarized political culture, the distrust model
exhibits conflict—and ultimately dysfunction—manifested by the
near impossibility of formal amendment.
84
Contiades and Fotiadou
ascribe this model to Belgium, Greece, Portugal, Spain, and the
Netherlands.
85
The direct-democratic model differs in several ways:
it grants citizens the power both to initiate and to have the last
word on constitutional change; it makes constitutional referenda
mandatory; it eschews un-amendability; and it privileges popular
participation over elite decision making.
86
As Contiades and
Fotiadou explain, “The design and qualities of referendums are of
great importance, while the role of political elites and courts is
influenced by the fact that the ultimate amending power lies with
the people.”
87
Switzerland is the leading example of this model.
88
Contiades and Fotiadou are not alone in developing a
classification of constitutional change. Carlo Fusaro and Dawn
Oliver have advanced a theory of constitutional change that
embraces both formal and informal amendment.
89
They categorize
the drivers of constitutional change (the people, the people and the
courts, legislative assemblies, the courts, governments and their
79
. Id. at 446.
80
. Id.
81
. Id.
82
. Id. at 446–47.
83
. Id. at 450.
84
. Id. at 451–52.
85
. Id. at 451–54.
86
. Id. at 454.
87
. Id.
88
. Id. at 455.
89
. See Carlo Fusaro & Dawn Oliver, Towards a Theory of Constitutional
Change, in HOW CONSTITUTIONS CHANGE: A COMPARATIVE STUDY 405, 405–33
(Dawn Oliver & Carlo Fusaro eds., 2011).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 927
leaders, and supranational institutions
90
), several legitimizing
theories of constitutional change (sovereignty, parliamentary
sovereignty, the constituting and constituted powers, representative
versus direct democracy, grundnorm setting, and majoritarian and
counter-majoritarian theories
91
), as well as the legal doctrines
according to which we justify or approach constitutional change
(positivism and neo-constitutionalism
92
). Their classification cuts
across formal and informal amendment and provides useful
abstractions about constitutional change. Yet they do not focus on
the actual design of formal amendment rules.
93
I have also developed elsewhere a classification of constitutional
amendment accounting for both formal and informal methods of
constitutional change. Drawing from written democratic
constitutions, I posited three models of constitutional change: the
textual model, the political model, and the substantive model.
94
I
classified Australia, Canada, and Switzerland as examples of the
textual model,
95
which holds that “the constitutional text enshrines
the necessary and sufficient conditions for amending the
constitution.”
96
The political model, represented by the United
States,
97
recognizes that “amendments may spring from expressions
of popular will that manifest themselves in dialogic exchanges
among the political branches and the citizenry” and that
amendments therefore “do not abide by the constitutionally
enshrined procedures for amending the constitution.”
98
The
substantive model, in contrast, “chooses instead to elevate
constitutional substance over political process, in so doing
contemplating the possibility of invalidating constitutional
amendments for departing from the spirit of the constitutional
text—even if those amendments satisfy the textual requirements for
constitutional entrenchment.”
99
I referred to Germany, India, and
South Africa as substantivist regimes.
100
90
. Id. at 414–16.
91
. Id. at 416–21.
92
. Id. at 421–23.
93
. Fusaro and Oliver do, however, discuss some important details of
formal amendment in an earlier chapter that is valuable and interesting, yet
even this chapter does not explain how formal amendment rules are structured.
See Dawn Oliver & Carlo Fusaro, Changing Constitutions: Comparative
Analysis, in HOW CONSTITUTIONS CHANGE: A COMPARATIVE STUDY, supra note 89,
at 381, 391–97.
94
. Richard Albert, Nonconstitutional Amendments, 22 CAN. J. L. &
JURISPRUDENCE 5, 12–31 (2009).
95
. Id. at 47.
96
. Id. at 12.
97
. Id. at 16.
98
. Id. at 12.
99
. Id. at 12–13.
100
. Id. at 21.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
928 WAKE FOREST LAW REVIEW [Vol. 49
My analysis stressed that these categories “are neither
exhaustive nor mutually exclusive, by which I mean not only that
constitutional states may exhibit practices indicative of a fourth or
nth constitutional amendment model but also that one
constitutional state may exhibit amendment practices aligned with
more than one model.”
101
The classification was designed to serve a
narrow purpose: to identify “sufficiently distinguishable” models in
order “to strike instructive contrasts that may help illuminate the
competing theories of sovereignty that underpin the amendment
practices of constitutional states.”
102
In retrospect, I would revise
my classification, but not in material ways that would undermine
my two conclusions: first, that formal amendment rules “either
conceal much about the actual practice of constitutional amendment
or simply do not accurately reflect the political norms that shape
and inform the practice of constitutional amendment,”
103
and
second, that “the theory and practice of constitutionalism is at once
rooted in constitutional texts, public institutions, judicial
interpretation, political practice, extratextual customs, and citizens
themselves.”
104
Nevertheless, my classification gave insufficient
attention to the design of formal amendment rules. Contiades and
Fotiadou’s classification exhibits the same weakness as Fusaro and
Oliver’s own, and neither their classifications nor mine can serve as
a complete guide for constitutional designers tasked with designing
formal amendment rules.
II. THE THREE TIERS OF FORMAL AMENDMENT RULES
Existing classifications do not offer directions for the design of
formal amendment rules. We can facilitate this design task only by
explicating the actual structure of formal amendment rules. In this
Article, I construct a new classification intended to guide
constitutional designers through the process of constructing formal
amendment rules. I demonstrate that formal amendment rules in
democratic constitutions are structured around three tiers, with
options within each of these tiers: one of two fundamental
foundations, one of six operational frameworks, and a combination
of supplementary specifications. I analyze formal amendment rules
in the world’s highest-performing constitutional democracies as
measured by the 2012 Democracy Index.
105
I focus only on full or
flawed democracies earning scores above 7.5 out of ten, and I
exclude democratic regimes without a master-text written
constitution. However, I include democratic regimes with strong
101
. Id. at 12.
102
. Id.
103
. Id. at 47.
104
. Id.
105
. THE ECONOMIST INTELLIGENCE UNIT, supra note 16.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 929
traditions of written and unwritten constitutionalism, and I also
include democratic regimes where the constitution consists of
constitutional statutes. I also include democratic regimes where
formal amendment occurs via constitutional act or fundamental law.
These criteria yield a study sample of thirty-six countries ranked in
the 2012 Democracy Index’s top forty.
TABLE 1: STUDY SAMPLE FROM THE 2012 DEMOCRACY INDEX
(WITH COUNTRY RANKING)
A. The Foundations of Formal Amendment
Formal amendment rules are anchored in the foundational
distinction between amendment and revision. The distinction holds
that amendment alters the constitution within the existing
framework of government while revision amounts to a fundamental
change that departs from the presuppositions of the constitution and
may even reshape its framework.
106
As Thomas Cooley observed in
the late nineteenth century, “[An amendment m]ust be in harmony
with the thing amended, so far at least as concerns its general spirit
and purpose. It must not be something so entirely incongruous that,
instead of amending or reforming it, it overthrows or revolutionizes
106
. WALTER F. MURPHY, CONSTITUTIONAL DEMOCRACY: CREATING AND
MAINTAINING A JUST POLITICAL ORDER 498 n.4 (2007); Miriam Galston,
Theocracy in America: Should Core First Amendment Values Be Permanent?, 37
HASTINGS CONST. L.Q. 65, 92–93 (2009); Jason Mazzone, Unamendments, 90
IOWA L. REV. 1747, 1751 (2005). This seems consistent with the etymological
root of the word “amend,” which derives from the Latin verb emendare, meaning
“to free from fault” or “to put right.” See BRYAN A. GARNER, MODERN AMERICAN
USAGE 41 (3d ed. 2009).
1
2
3
4
Norway
Sweden
Iceland
Denmark
5
6
7
8
New Zealand
Australia
Switzerland
Canada
9
10
11
12
Finland
Netherlands
Luxembourg
Austria
13
14
15
17
Ireland
Germany
Malta
Czech Republic
18
18
20
21
Uruguay
Mauritius
South Korea
United States
22
23
24
25
Costa Rica
Japan
Belgium
Spain
26
26
28
28
Cape Verde
Portugal
France
Slovenia
30
31
32
33
Botswana
South Africa
Italy
Greece
34
35
36
38
Estonia
Taiwan
Chile
India
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
930 WAKE FOREST LAW REVIEW [Vol. 49
it.”
107
More recently, John Rawls has defined an amendment as
“adjust[ing] basic constitutional values to changing political and
social circumstances”
108
and “adapt[ing] basic institutions in order to
remove weaknesses that come to light in subsequent constitutional
practice.”
109
1. Formal Amendment and Constitutional Revision
Some democratic constitutions use the term “revision” but do
not entrench the conceptual distinction between amendment and
revision pursuant to which different procedures apply to amending
or revising the constitution.
110
These constitutions use the term
“revision” to mean “amendment.”
111
Other democratic constitutions
may be read as acknowledging or suggesting a distinction between
amendment and revision but do not specify different procedures for
each process.
112
Still other democratic constitutions make explicit
the conceptual distinction between amendment and revision and
accordingly entrench procedures for amendment that differ from
those that must be used for revision.
113
These constitutions
illustrate one of the two foundations in the design of formal
amendment rules: constitutions that entrench rules for both
amendment and revision. Finally, some constitutions adopt a
middle position between entrenching and leaving unstated the
distinction between amendment and revision: they insist that formal
107
. Thomas M. Cooley, The Power to Amend the Federal Constitution, 2
MICH. L.J. 109, 118 (1893).
108
. JOHN RAWLS, POLITICAL LIBERALISM 238 (2d ed. 2005).
109
. Id. at 239.
110
. See 1975 SYNTAGMA [SYN.] [CONSTITUTION] 2, art. 110 (Greece);
CONSTITUTION OF LUXEMBOURG, art. 114; CONSTITUIÇÃO DA REPÚBLICA
PORTUGUESA, arts. 286–89.
111
. Some constitutions use the term “reform” instead of “amendment” or
“revision.” For example, the Chilean Constitution refers to the “reform of the
Constitution.” CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE CHILE [C.P.] arts. 127–
29. Where constitutions use “reform” without distinguishing it from another
type of alteration, I interpret it to refer to “amendment.” The analysis would
differ if, for example, the constitutional text distinguished between “partial
reform” and “general reform,” in which case the former would mean
“amendment” and the latter would mean “revision.” Textual inquiry is useful to
distinguish among different formal amendment designs, but it is useful,
perhaps necessary, to look beyond the text into constitutional practice and
interpretation in order to understand the domestic significance of these various
terms.
112
. See CONSTITUIÇÃO DA REPÚBLICA DE CABO VERDE, art. 309, para. 1;
CONSTITUCIÓN DE 1967, art. 331 (Uru.).
113
. See BUNDES-VERFASSUNGSGESETZ [B-VG] [CONSTITUTION] BGBl No.
1/1920, as last amended by Bundesverfassungsgesetz [BVG] BGBl I No. 2/1983,
art. 44, ¶ 2 (Austria); CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE COSTA RICA,
arts. 195–96; CONSTITUCIÓN ESPAÑOLA [C.E.], B.O.E. nn.166–68, Dec. 29, 1978
(Spain); BUNDESVERFASSUNG [BV] [CONSTITUTION] Apr. 18, 1999, SR 101, arts.
192–95 (Switz.).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 931
amendments conform to substantive standards but do not specify
separate procedures for adopting a formal change that violates those
substantive standards.
TABLE 2: AMENDMENT AND REVISION IN DEMOCRATIC CONSTITUTIONS
Constitution suggests both
amendment and revision, but
does
not
entrench the distinction.
Cape Verde
Uruguay
Constitution expressly recognizes
both amendment and revision, and
does
entrench the distinction.
Austria
Costa Rica
Spain
Switzerland
Constitution establishes substantive
standard that formal amendment
must respect but does not specify
consequence of adopting amendment
in violation of those standards.
Czech Republic
Norway
For example, the Swiss Constitution entrenches the conceptual
distinction between amendment and revision. It distinguishes
between “total” and “partial” revision, the former referring to
revision and the latter to amendment: “The Federal Constitution
may at any time be subjected to a total or a partial revision.”
114
Total revision “may be proposed by the People or by one of the
Chambers, or may be decreed by the Federal Parliament”
115
but
“[t]he mandatory provisions of international law may not be
violated.”
116
In contrast, partial revision “may be requested by the
People, or be decreed by the Federal Parliament,”
117
but “[a] partial
revision must respect the principle of the unity of subject matter; it
may not violate the mandatory provisions of international law.”
118
The Swiss Constitution establishes a further restriction on partial
revision: “A popular initiative for partial revision must, moreover,
respect the principle of the unity of form.”
119
Therefore whereas
international law is the only textual restriction on total revision, the
Constitution constrains partial revision in more ways—according to
subject matter, unity of form, and international law—consistent
114
. BUNDESVERFASSUNG [BV] Apr. 18, 1999, SR 101, art. 192, para. 1
(Switz.).
115
. Id. at art. 193, para. 1.
116
. Id. at art. 193, para. 4.
117
. Id. at art. 194, para. 1.
118
. Id. at art. 194, para. 2.
119
. Id. at art. 194, para. 3.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
932 WAKE FOREST LAW REVIEW [Vol. 49
with the use of amendment for only narrow, not sweeping,
changes.
120
Similarly, the Costa Rican Constitution distinguishes between a
“partial amendment” and a “general amendment.” The Constitution
specifies that “the Legislative Assembly may partially amend this
Constitution complying strictly with the following provisions,” going
on to list eight requirements for effecting a partial amendment,
including who may initiate a partial amendment and the requisite
voting thresholds, as well as quorum requirements and time
limits.
121
The Constitution also outlines a special procedure for
making a general amendment that does more than merely fine tune
the text, which states: “A general amendment of this Constitution
can only be made by a Constituent Assembly called for the purpose.
A law calling such Assembly shall be passed by a vote of no less than
two thirds of the total membership of the Legislative Assembly and
does not require the approval of the Executive Branch.”
122
That the
Constitution sets strict requirements for partial amendment is
consistent with the limited uses of amendment, as opposed to the
more transformative changes possible with revision, which may be
authorized only by an extraordinary body or procedure—namely a
Constituent Assembly in Costa Rica or the people in Switzerland.
2. Formal Amendment and Implicit Limitations
But most democratic constitutions—thirty out of the thirty-six
in this study—leave unstated the distinction between amendment
and revision.
123
These constitutions neither recognize nor imply
120
. For helpful context to the international law restriction in the Swiss
Constitution, see Yaniv Roznai, The Theory and Practice of “Supra-
Constitutional” Limits on Constitutional Amendments, 62 INT’L & COMP. L.Q.
557, 591 (2013).
121
. CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE COSTA RICA, art. 195.
122
. Id. at art. 196.
123
. See U.S. CONST. art. V; AUSTRALIAN CONSTITUTION s 128; 1994 CONST.
art. 195 (Belg.); CONSTITUTION OF BOTSWANA § 89; Constitution Act, 1982, being
Schedule B to the Canada Act, 1982, s. 38–48 (U.K.); CONSTITUIÇÃO DA
REPÚBLICA DE CABO VERDE, arts. 309–12, 314; CONSTITUCIÓN POLÍTICA DE LA
REPÚBLICA DE CHILE [C.P.] arts. 127–29; GRUNDLOVEN [GRL] [CONSTITUTION] Lov
nr. 88 (Den.); PÕHISEADUS [CONSTITUTION], ss 161–68 (Est.); SUOMEN
PERUSTALSKI [CONSTITUTION], 6 luku 73 (Fin.); 1958 CONST. art. 89 (Fr.);
GRUNDGESETZ FÜR DIE BUNDESREPUBLIK DEUTSCHLAND [GRUNDGESETZ] [GG]
[BASIC LAW], May 23, 1949, BGBl. VII, art. 79 (Ger.); 1975 SYNTAGMA [SYN.] 2,
art. 110 (Greece); STJÓRNARSKRÁ LÝÐVELDISINS ÍSLANDS [CONSTITUTION], art. 79
(Ice.); INDIA CONST. art. 368; IR. CONST., 1937, art. 46; Art. 138 Costituzione
[Cost.] [Constitution] (It.); NIHONKOKU KENPŌ [KENPŌ] [CONSTITUTION], art. 96
(Japan); CONSTITUTION OF LUXEMBOURG, art. 114; KOSTITUZZJONI TA’ MALTA
[CONSTITUTION], art. 66; MAURITIUS CONST., art. 47; Grondwet voor het
Koninkrijk der Nederlanden [GW.] [CONSTITUTION] arts. 137–42 (Neth.);
Constitution Act, 1986, pt. 3 s 15 (N.Z.); CONSTITUIÇÃO DA REPÚBLICA
PORTUGUESA, arts. 285–86; USTAVA REPUBLIKE SLOVENIJE [CONSTITUTION] arts.
168–71 (Slovn.); S. AFR. CONST., 1996, sec. 74; DAEHANMINKUK HUNBEOB
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 933
that amendment and revision entail different consequences and
outcomes. The standard democratic design instead defines formal
alteration exclusively with regard to amendment. Yet it does not
follow from the non-entrenchment of the distinction between
amendment and revision that the distinction does not exist or that
its non-entrenchment will foreclose its emergence from other
sources. For example, the natural right of revolution is not usually
entrenched in constitutions,
124
yet it remains an implicit restriction
against which formal amendment rules and constitutions
themselves are defenseless.
125
I will later show why we should not
likewise presume that the distinction between amendment and
revision does not exist where it is not entrenched.
126
TABLE 3: UNSTATED DISTINCTION BETWEEN AMENDMENT AND
REVISION IN DEMOCRATIC CONSTITUTIONS
Australia
Belgium
Botswana
Canada
Cape Verde
Chile
Denmark
Estonia
Finland
France
Germany
Greece
Iceland
India
Ireland
Italy
Japan
Luxembourg
Malta
Mauritius
Netherlands
New
Zealand
Portugal
Slovenia
South
Africa
South Korea
Sweden
Taiwan
Uruguay
United
States
There are risks to leaving implicit what is subject to
amendment and revision. Disagreement is bound to occur regarding
which constitutional changes work a fundamental redesign to the
constitution and which are less fundamental. As Laurence Tribe
has written in the context of the U.S. Constitution,
[HUNBEOB] [CONSTITUTION] art. 98 (S. Kor.); REGERINGSFORMEN [RF]
[CONSTITUTION] 8, arts. 14–16 (Swed.); MINGUO XIANFA [CONSTITUTION] art. 174
(1947) (Taiwan); CONSTITUCIÓN DE 1967, art. 331 (Uru.).
124
. Roughly twenty percent of the world’s constitutions entrench the right
of revolution in some form. See Tom Ginsburg et al., When to Overthrow Your
Government: The Right to Resist in the World’s Constitutions, 60 UCLA L. REV.
1184, 1217–18 (2013). But note that the Mexican Constitution contemplates
restricting this right by establishing a formal constitutional rule against
revolution. See Constitución Política de los Estados Mexicanos [C.P.], as
amended, Diario Oficial de la Federación [DO], 5 de Febrero de 1917, art. 135
(Mex.).
125
. JOHN LOCKE, TWO TREATISES OF GOVERNMENT: THE SECOND TREATISE OF
CIVIL GOVERNMENT 246–47 (Thomas I. Cook ed., Hafner Publ’g Co. 1947) (1690).
126
. See infra Subpart III.C.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
934 WAKE FOREST LAW REVIEW [Vol. 49
Not only is the list of fundamental constitutional norms open
to debate, but the very identity of the Constitution—the body
of textual and historical materials from which the norms are to
be extracted and by which their application is to be guided—is
itself a matter that cannot be objectively deduced or passively
discerned in a viewpoint-free way.
127
Melissa Schwartzberg echoes this point, observing that
“scholars dispute what constitutes these substantive limits, which
suggests the broader contestability of these sorts of claims[,]”
128
and
concluding that “[e]fforts at restricting the boundaries of
constitutional amendment are bound to be challengeable, and
reasonable people are likely to disagree about what constitutes an
unalterable principle.”
129
The challenge for constitutional designers
is therefore to find the right balance between giving clarity to their
successors and accommodating the future contestability of
constitutional identity.
130
Some constitutions adopt a middle position between
entrenching different procedures for amendment and revision and
leaving the distinction unstated. These constitutions insist that
formal amendments must conform to certain substantive standards,
but they do not state the consequences of adopting a formal
127
. Laurence H. Tribe, A Constitution We Are Amending: In Defense of a
Restrained Judicial Role, 97 HARV. L. REV. 433, 440 (1983). More recently,
however, Professor Tribe has acknowledged that
it may well be that some properly adopted formal amendments could
themselves be deemed “unconstitutional” because of their radical
departure from premises too deeply embedded to be repudiated
without a full-blown revolution. Thus, for instance, an amendment
repealing the Article IV guarantee of a “republican” form of
government and simultaneously making membership in Congress a
matter of heredity, rather than of election by “the People,” might well
be deemed void regardless of its process of adoption, as might an
amendment that repudiates the rule of law or abandons the
indissoluble character of the Union.
LAURENCE H. TRIBE, THE INVISIBLE CONSTITUTION 33–34 (2008).
128
. MELISSA SCHWARTZBERG, DEMOCRACY AND LEGAL CHANGE 148 (2007).
129
. Id.
130
. This challenge for constitutional designers also touches upon subject-
matter restrictions. See infra Subpart II.C.2. Where constitutional designers
designate a particular provision as unamendable, political actors could
nevertheless use the regular procedures of formal amendment to amend that
nominally unamendable provision. This could work an implicit constitutional
revision. Analogously, where constitutional designers do not designate a
particular provision as unamendable, yet that provision is or becomes central to
the regime’s political culture, one could argue that such a change would
likewise work an implicit constitutional revision. Akhil Amar has suggested
that using Article V to repeal the First Amendment would have this effect. See
Akhil Reed Amar, Philadelphia Revisited: Amending the Constitution Outside
Article V, 55 U. CHI. L. REV. 1043, 1044–45 n.1 (1988). John Rawls has taken a
similar view of the First Amendment, which he views as having been “validated
by long historical practice.” RAWLS, supra note 108, at 239.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 935
amendment that violates those standards, and they also do not
entrench a separate procedure for revision. This suggests three
points: formal amendment cannot be the vehicle to adopt a provision
in violation of those standards; formal alteration of those standards
would amount to something more than an amendment; and those
standards may be violated, but only by revision, which would result
in adopting an altogether new constitution.
For instance, the Norwegian Constitution entrenches formal
amendment rules but insists that an amendment “must never,
however, contradict the principles embodied in this Constitution,
but solely relate to modifications of particular provisions which do
not alter the spirit of the Constitution.”
131
The Constitution insists
that formal amendments cannot violate this substantive standard,
but it does not state a procedure for adopting formal alterations that
do. It leaves implicit the power of political actors to revise the
Constitution in contravention of this standard. Were political actors
to “alter the spirit of the Constitution,” the result would be
something more than an amendment, and indeed could not be
achieved through the formal amendment process.
132
The only
alternative appears to be the adoption of a new constitution.
Similarly, the Czech Republic’s Constitution entrenches the
middle position: it neither specifies different alteration procedures
for amendment and revision nor does it leave the distinction
unstated. It instead suggests that the regular formal amendment
process is insufficient to formalize changes that will work something
more than an amendment to the constitution. Its text states that
“the substantive requisites of the democratic, law-abiding State may
not be amended,”
133
and “interpretation of legal rules may not be
used as authorization to eliminate or imperil the foundations of the
democratic State.”
134
It is possible to imagine an alteration that
would conform to these standards of democracy. But where an
alteration would violate those standards, it would be properly
described as a revision, not an amendment. Neither the Czech nor
the Norwegian Constitutions specify what the revision process
entails, but we can infer that their constitutional designers intended
those changes to occur only in a constitutional redesign.
131
. GRUNNLOVEN [GRL.] [CONSTITUTION] § 112 (Nor.). This absolute
prohibition on amendments violating the spirit of the Constitution does not
grant interpretive authority to courts; it is directed to the national legislature,
which holds controlling authority in interpreting the Constitution. See Yaniv
Roznai, Unconstitutional Constitutional Amendments—The Migration and
Success of a Constitutional Idea, 61 AM. J. COMP. L. 657, 670 n.91 (2013).
132
. GRUNNLOVEN [GRL.] § 112 (Nor.).
133
. Ústavní zákon č. 1/1993 Sb., Ústava České Republiky [Constitution of
the Czech Republic], art. 9, para. 2.
134
. Id. at art. 9, para. 3.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
936 WAKE FOREST LAW REVIEW [Vol. 49
B. The Frameworks of Formal Amendment
Written democratic constitutions generally embed one of six
formal amendment frameworks into their formal amendment
foundations. The frameworks of formal amendment vary according
to the number of procedures available for formally amending the
constitution and the range of the constitutional provisions open to
formal amendment by those procedures. The first important
observation about formal amendment frameworks is that formal
amendment rules may entrench either one or more procedures for
formally amending the constitution. I divide these procedures into
two categories: single-track, for formal amendment rules
entrenching only one procedure for formal amendment; and multi-
track, for formal amendment rules entrenching more than one
procedure for formal amendment. Formal amendment procedures
may differ with respect to the institutions authorized to initiate a
formal amendment, to amend an amendment proposal, and to ratify
an amendment.
135
The second important observation about formal amendment
frameworks is that formal amendment rules may also authorize the
use of all, some, or one of these procedures of formal amendment to
amend all, some, or one of the provisions entrenched in the
constitution. I divide the range of constitutional provisions open to
formal amendment into three categories: comprehensive, under
which all amendable provisions are susceptible to amendment by all
available procedures for formal amendment; restricted, pursuant to
which each amendable provision is made amendable by a designated
procedure for formal amendment; and exceptional, which creates
one amendment procedure of general application and a second
procedure exclusively for one constitutional provision or a set of
related provisions. These criteria generate six possible
combinations demonstrable in a 2x3 matrix.
135
. For purposes of my classification, I do not distinguish the power to
initiate an amendment within the same branch of government (for instance as
between the two houses of the bicameral national legislature) but I do
distinguish the power to initiate an amendment among branches of the national
government (for instance as between the president and the legislature) and
between national and subnational government institutions (for instance as
between the national and subnational legislatures).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 937
TABLE 4: FORMAL AMENDMENT FRAMEWORKS IN CONSTITUTIONAL
DEMOCRACIES
Single-Track
Multi-Track
Comprehensive
Comprehensive Single-
Track
(Germany)
Comprehensive Multi-
Track
(France)
Restricted
Restricted Single-Track
(South Africa)
Restricted Multi-Track
(Canada)
Exceptional
Exceptional Single-
Track
(Australia)
Exceptional Multi-
Track
(United States)
C. Formal Amendment Procedures (Single- or Multi-Track), the
Scope of Their Use (Comprehensive, Restricted, or Exceptional), and
Representative Democratic Constitutions
Below, I describe these six formal amendment frameworks with
reference to currently-in-force democratic constitutions. I also
categorize each of the thirty-six democratic constitutions in this
study. To summarize, thirteen adopt the comprehensive multi-track
framework (Costa Rica, Finland, France, Greece, Italy, Luxembourg,
Netherlands, Slovenia, South Korea, Sweden, Switzerland, Taiwan,
Uruguay), ten adopt the comprehensive single-track framework
(Belgium, Cape Verde, Czech Republic, Denmark, Germany,
Ireland, Japan, New Zealand, Norway, Portugal), five adopt the
restricted single-track framework (Botswana, India, Malta,
Mauritius, South Africa), four adopt the restricted multi-track
framework (Canada, Chile, Estonia, Spain), three adopt the
exceptional single-track framework (Australia, Austria, Iceland),
and one adopts the exceptional multi-track framework (United
States).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
938 WAKE FOREST LAW REVIEW [Vol. 49
TABLE 5: AMENDMENT FRAMEWORKS IN DEMOCRATIC CONSTITUTIONS
Comprehensive Multi-Track
Amendment Framework
Costa Rica
France
Italy
Netherlands
South Korea
Switzerland
Uruguay
Finland
Greece
Luxembourg
Slovenia
Sweden
Taiwan
Comprehensive Single-Track
Amendment Framework
Belgium
Czech Republic
Germany
Japan
Norway
Cape Verde
Denmark
Ireland
New Zealand
Portugal
Restricted Single-Track
Amendment Framework
Botswana
India
Malta
Mauritius
South Africa
Restricted Multi-Track
Amendment Framework
Canada
Chile
Estonia
Spain
Exceptional Single-Track
Amendment Framework
Australia
Austria
Iceland
Exceptional Multi-Track
Amendment Framework
United States
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 939
1. Comprehensive Single-Track Amendment
TABLE 6: COMPREHENSIVE SINGLE-TRACK FRAMEWORK IN
DEMOCRATIC CONSTITUTIONS
Belgium
Cape
Verde
Czech
Republic
Denmark
Germany
Ireland
Japan
New
Zealand
Norway
Portugal
Of the thirty-six democratic constitutions, ten entrench the
comprehensive single-track framework.
136
This framework has the
virtue of clarity: there exists only one formal amendment procedure
(it is single-track) and it applies to all amendable constitutional
provisions (it is comprehensive). The German Basic Law illustrates
the paradigmatic example of the comprehensive single-track
framework. The German Basic Law’s text allows amendments “only
by a law expressly amending or supplementing its text”
137
and
specifies that there is only one way to make a formal amendment:
“Any such law shall be carried by two thirds of the Members of the
Bundestag and two thirds of the votes of the Bundesrat.”
138
The
rule is therefore clear: to formally amend the German Basic Law,
two-thirds of both houses of the national legislature must approve
the amendment.
139
Japan likewise follows the comprehensive
single-track framework.
140
To formally amend the Constitution, the
bicameral national legislature must initiate the amendment process
with a two-thirds vote, the proposal must be ratified in a
referendum by a majority vote, and the Emperor must then
promulgate the amendment.
141
This single-track framework is
comprehensive insofar as it applies to all amendable provisions.
136
. See 1994 CONST. art. 195 (Belg.); CONSTITUIÇÃO DA REPÚBLICA DE CABO
VERDE, arts. 309–12, 314; Ústavní zákon č. 1/1993 Sb., Ústava České Republiky
[Constitution of the Czech Republic], art. 39, para. 4; GRUNDLOVEN [GRL] Lov
nr. 88 (Den.); GRUNDGESETZ FÜR DIE BUNDESREPUBLIK DEUTSCHLAND
[GRUNDGESETZ] [GG] [BASIC LAW], May 23, 1949, BGBl. VII, art. 79 (Ger.); IR.
CONST., 1937, art. 46; NIHONKOKU KENPŌ [KENPŌ], art. 96 (Japan); Constitution
Act, 1986, pt. 3 s 15 (N.Z.); GRUNNLOVEN [GRL.] § 112 (Nor.); CONSTITUIÇÃO DA
REPÚBLICA PORTUGUESA, arts. 285–86.
137
. GRUNDGESETZ FÜR DIE BUNDESREPUBLIK DEUTSCHLAND [GRUNDGESETZ]
[GG] [BASIC LAW], May 23, 1949, BGBl. VII, art. 79 (Ger.).
138
. Id. at art. 79, para. 2.
139
. Id.
140
. NIHONKOKU KENPŌ [KENPŌ], art. 96 (Japan).
141
. Id.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
940 WAKE FOREST LAW REVIEW [Vol. 49
2. Comprehensive Multi-Track Amendment
TABLE 7: COMPREHENSIVE MULTI-TRACK FRAMEWORK IN
DEMOCRATIC CONSTITUTIONS
Costa Rica
Finland
France
Greece
Italy
Luxembourg
Netherlands
Slovenia
South
Korea
Sweden
Switzerland
Taiwan
Uruguay
The comprehensive multi-track framework differs from its
single-track counterpart only insofar as it offers political actors more
than one procedure to make a formal amendment to the
constitution. I include in this framework all alternative amendment
procedures, whether they are genuine alternatives or whether they
authorize alternative routes in the event of an amendment failure
under another route. Under the comprehensive multi-track
framework, political actors may deploy any of the available formal
amendment procedures to make a formal amendment to any
amendable provision in the constitutional text. All formal
amendment procedures are equally useable, and any may be used to
amend any amendable constitutional provision. Thirteen of the
thirty-six democracies in this study have adopted the comprehensive
multi-track framework of formal amendment.
142
The French Constitution is a prominent example. Amendments
may be initiated by either the President or members of the national
legislature.
143
An amendment proposal may take the form of a
governmental or private member’s bill,
144
the former initiated by a
member of the cabinet and the latter by a noncabinet
parliamentarian. Both houses of the national legislature must then
approve the amendment proposal, which must subsequently be
ratified in a national referendum.
145
142
. See CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE COSTA RICA, arts. 195–
96; SUOMEN PERUSTALSKI, 6 luku 73 (Fin.); 1958 CONST. art. 90 (Fr.); 1975
SYNTAGMA [SYN.] 2, art. 110 (Greece); Art. 138 Costituzione [Cost.] (It.);
CONSTITUTION OF LUXEMBOURG, art. 114; Grondwet voor het Koninkrijk der
Nederlanden [GW.] arts. 137–42 (Neth.); USTAVA REPUBLIKE SLOVENIJE, arts.
168–71 (Slovn.); DAEHANMINKUK HUNBEOB [HUNBEOB] art. 98 (S. Kor.);
REGERINGSFORMEN [RF] 8, arts. 14–16 (Swed.); BUNDESVERFASSUNG [BV] Apr.
18, 1999, SR 101, arts. 192–95 (Switz.); MINGUO XIANFA art. 174 (1947)
(Taiwan); CONSTITUCIÓN DE 1967, art. 331 (Uru.).
143
. 1958 CONST. art. 90 (Fr.).
144
. Id.
145
. Id. But not all amendment proposals must be ratified by referendum.
The French Constitution authorizes the President to unilaterally bypass the
referendum requirement for governmental amendment bills; in those cases, the
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 941
This framework is multi-track because it creates multiple ways
to formally amend the constitution. Under the comprehensive
multi-track framework, political actors have the choice of at least
two paths to propose and/or ratify a formal amendment. Recall, in
contrast, that a constitution entrenching the comprehensive single-
track framework (for instance, the German Basic Law) gives
political actors no such choice because the comprehensive single-
track framework offers only one procedure to formally amend the
constitution. Yet the comprehensive multi- and single-track
frameworks share the feature of general applicability. Just as the
single amendment procedure within the comprehensive single-track
framework may be used to amend all amendable constitutional
provisions, the multiple amendment procedures within the
comprehensive multi-track framework may likewise be used to
amend all amendable constitutional provisions. The multiple
amendment procedures within the multi-track frameworks are
therefore equal insofar as they are equally deployable to amend any
of the amendable constitutional provisions entrenched in the written
constitution.
146
The Italian Constitution similarly exhibits the comprehensive
multi-track framework.
147
Formal amendments must be proposed
and adopted in the bicameral national legislature and approved by
an absolute majority of each house in two consecutive votes held
within three months.
148
If the houses approve the amendment
proposal by a supermajority of two-thirds, the amendment becomes
effective.
149
But if the amendment proposal fails to secure
supermajority approval in both houses, it may be subject to the
additional step of a referendum. The proposal must be ratified in a
national referendum if such a request is made by one-fifth of one of
the two houses, 500,000 voters, or five of the autonomous regional
councils.
150
The formal amendment rules require the amendment
proposal to win majority support in the referendum to become
effective.
151
Like the formal amendment procedures in the French
Constitution, these procedures provide more than one way to
President must submit the approved governmental amendment bill to
Parliament convened in Congress, where it must be ratified by a three-fifths
majority. Id.
146
. But the 1962 amendment authorizing the direct election of the
President by national vote was adopted in apparent violation of the French
Constitution’s formal amendment rules. See Martin A. Rogoff, Fifty Years of
Constitutional Evolution in France: The 2008 Amendments and Beyond, JUS
POLITICUM, No. 6, 2011, at 1, 13 (Fr.), available at http://papers.ssrn.com
/sol3/papers.cfm?abstract_id=1793210.
147
. Art. 138 Costituzione [Cost.] (It.).
148
. Id.
149
. Id.
150
. Id.
151
. Id.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
942 WAKE FOREST LAW REVIEW [Vol. 49
formally amend the constitution, and they may each be used to
formally amend any of the amendable constitutional provisions.
3. Restricted Single-Track Amendment
TABLE 8: RESTRICTED SINGLE-TRACK FRAMEWORK IN DEMOCRATIC
CONSTITUTIONS
Botswana
India
Malta
Mauritius
South
Africa
In contrast, the restricted single-track framework creates more
than one amendment procedure but expressly designates which of
these amendment procedures must be used to formally amend
specific provisions of the constitution. Although it consists of more
than one formal amendment procedure, it is restricted because each
amendment procedure is deployable only for specifically enumerated
provisions of the constitution. It is single track only insofar as it
provides a single procedure for amending specifically enumerated
constitutional provisions. The restricted single-track framework is
distinguishable from the comprehensive single-track framework
according to the number of amendment procedures it entrenches:
the comprehensive single-track framework provides only one
amendment procedure to formally amend all amendable
constitutional provisions, whereas the restricted single-track
framework creates more than one formal amendment procedure to
amend designated provisions. The difference between the restricted
single-track framework and the comprehensive multi-track
framework turns on how their various amendment procedures may
be deployed. Under the restricted single-track framework, the
various amendment procedures are expressly keyed to specific
constitutional provisions. In contrast, the comprehensive multi-
track framework contemplates the use of any of the multiple
amendment procedures to formally amend any amendable
constitutional provision.
Five democratic constitutions entrench the restricted single-
track framework.
152
Unlike the multiple amendment procedures
under the French Constitution’s comprehensive multi-track
framework, the multiple amendment procedures under the South
African Constitution’s restricted single-track framework cannot be
used to amend all constitutional provisions. The Constitution
creates three amendment procedures, each of which is linked to
different classes of provisions.
153
The lowest amendment threshold
152
. See CONSTITUTION OF BOTSWANA § 89; INDIA CONST. art. 368;
KOSTITUZZJONI TA’ MALTA, art. 66; MAURITIUS CONST., art. 47; S. AFR. CONST.,
1996, sec. 74.
153
. S. AFR. CONST., 1996, sec. 74.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 943
requires two-thirds approval in the National Assembly.
154
It is the
Constitution’s general amending formula and may be used to
formally amend all provisions not specially assigned to a higher
amendment threshold. The intermediate amendment threshold
requires two-thirds approval in both the National Assembly and the
National Council of Provinces.
155
This procedure must be used to
formally amend the Bill of Rights, as well as any amendment that
concerns the National Council of Provinces, modifies provincial
rights or prerogatives, or changes a constitutional provision relating
specifically to a provincial matter.
156
The most exacting formal
amendment requirements require approval by three-quarters and
two-thirds, respectively, in the National Assembly and the National
Council of Provinces.
157
This amendment threshold governs
amendments to the Constitution’s statement of constitutional values
as well as to the highest amendment threshold itself.
158
We can perceive in the South African Constitution the two
defining characteristics of the restricted single-track framework of
formal amendment—multiplicity and rigidity. First, the restricted
single-track framework entrenches more than one procedure to
amend the constitution. But unlike the comprehensive multi-track
framework, whose multiple procedures may be used to amend any of
the constitution’s amendable constitutional provisions, the restricted
single-track framework’s multiple procedures may not be used to
freely amend all of the constitution’s amendable provisions. Second,
we must note the rigidity of the framework. Its multiple
amendment procedures are designated for specific constitutional
provisions. Where, for instance, a constitution entrenching the
restricted single-track framework creates three amendment
procedures, each of the three procedures is expressly connected to a
particular class of constitutional provisions and those provisions
alone.
154
. S. AFR. CONST., 1996, sec. 74, para. 3.
155
. Id. at sec. 74, para. 2.
156
. Id. at sec. 74, paras. 2–3.
157
. Id. at sec. 74, paras. 1–2.
158
. Id. at sec. 74, para. 1. The Constitution’s statement of constitutional
values proclaims that
[t]he Republic of South Africa is one, sovereign, democratic state
founded on the following values: (a) Human dignity, the achievement
of equality and the advancement of human rights and freedoms. (b)
Non-racialism and non-sexism. (c) Supremacy of the constitution and
the rule of law. (d) Universal adult suffrage, a national common voters
roll, regular elections and a multi-party system of democratic
government, to ensure accountability, responsiveness and openness.
Id. at sec. 1.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
944 WAKE FOREST LAW REVIEW [Vol. 49
4. Restricted Multi-Track Amendment
TABLE 9: RESTRICTED MULTI-TRACK FRAMEWORK IN DEMOCRATIC
CONSTITUTIONS
Canada
Chile
Estonia
Spain
Of the thirty-six democratic constitutions in this study, four
entrench the restricted multi-track framework.
159
This framework
combines the restrictions of specifically designated amendment
procedures with multiple procedures for amending those specific
provisions. Whereas the restricted single-track framework creates
only one formal amendment procedure for amending specifically
enumerated constitutional provisions, the restricted multi-track
framework provides more than one procedure for amending those
specifically enumerated provisions.
The Canadian Constitution provides an illustration. Its text
entrenches five amendment procedures.
160
First, the method of
provincial unilateral amendment authorizes a provincial legislature
to formally amend its provincial constitution by passing a law in the
normal course of the legislative process.
161
This procedure may be
initiated only in the subnational legislature concerned. Second,
under the method of unilateral federal amendment, Parliament may
pass a simple law to amend the Constitution.
162
Either house of
Parliament may initiate the amendment.
Each of the following three procedures may be initiated by
either house of the national legislature or a subnational
legislature.
163
First, under the method of parliamentary
amendment with the approval of the affected provincial legislature,
both houses of Parliament are required to adopt a resolution by
majority vote followed by a majority vote of the legislature whose
subnational unit is affected by the amendment.
164
Second, under
the general amending formula, both houses of Parliament are
required to adopt a resolution by majority vote followed by
resolutions from two-thirds of provincial legislatures whose
aggregate population represents at least fifty percent of the national
159
. See Constitution Act, 1982, being Schedule B to the Canada Act, 1982,
s. 38–48 (U.K.); CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE CHILE [C.P.] artS.
127–29; PÕHISEADUS, ss 161–68 (Est.); CONSTITUCIÓN ESPAÑOLA [C.E.], B.O.E.
nn.166–68, Dec. 29, 1978 (Spain).
160
. See generally, Constitution Act, 1982, being Schedule B to the Canada
Act, 1982, s. 38–48 (U.K.).
161
. Id. at. s. 45 (“Subject to section 41, the legislature of each province may
exclusively make laws amending the constitution of the province.”).
162
. Id. at s. 44.
163
. See id. at s. 46.
164
. Id. at s. 43.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 945
population.
165
Finally, under the method of unanimity, both houses
of Parliament are required to adopt a resolution by majority vote
followed by majority resolutions from each of the provincial
legislatures.
166
These are the five distinguishable formal
amendment procedures under the Canadian Constitution.
167
Each of Canada’s five formal amendment procedures is specially
designated for amending specific constitutional provisions. The first
method—unilateral provincial amendment—applies to all provincial
matters except those specifically assigned to higher amendment
thresholds.
168
Canadian political actors may use the method of
unilateral federal amendment to make formal amendments to the
federal executive government or to either of the two houses of
Parliament.
169
They may use the third mode of amendment—
Parliament and the affected province—to alter provincial
boundaries or to change the use of English or French within a
province.
170
The fourth mode of amendment—Parliament and a
supermajority of provinces—is the Constitution’s default
amendment procedure and must be used to make a formal
amendment to such provisions, including provincial representation
in the House of Commons and the Senate, the powers of the Senate
and the way Senators are selected, and the creation of new
provinces.
171
The fifth mode of amendment—unanimous consent—
must be used to make formal amendments to provisions like the
monarchy, the composition of the Supreme Court of Canada, and the
procedures for amending the Constitution.
172
The subject-matter
restrictions associated with these five amendment procedures
combined with the multiplicity of initiators show why the Canadian
Constitution is an example of the restricted multi-track framework.
The Spanish Constitution also demonstrates the restricted
multi-track framework. The government, either house of the
national legislature, or the legislatures of the country’s autonomous
communities may initiate a formal amendment proposal.
173
The
default formal amendment rule then requires each house of the
legislature to approve the proposal by a three-fifths vote.
174
Alternatively, where both houses are unable to approve the
165
. Id. at s. 38, para. 1. Territorial population does not appear to count
toward national population. See id. at s. 38, para. 2.
166
. Id. at s. 41
167
. The Canadian Constitution also authorizes provinces to dissent from,
and thereby avoid being bound by, certain formal amendments, as well as to opt
out from others. See id. at s. 38, paras. 2–4, s. 40.
168
. Id. at s. 41.
169
. Id. at s. 44.
170
. Id. at s. 43.
171
. Id. at s. 42, para. 1(f).
172
. Id. at s. 41.
173
. CONSTITUCIÓN ESPAÑOLA [C.E.], B.O.E. n. 87, Dec. 29, 1978 (Spain).
174
. Id. at B.O.E. n. 167, para. 1.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
946 WAKE FOREST LAW REVIEW [Vol. 49
amendment proposal by the required two-thirds threshold, the
amendment may pass with simple majority approval in the Senate
and two-thirds approval in the Congress.
175
The formal amendment
process also includes an optional final step: a referendum, if
requested by one-tenth of the members of either house within fifteen
days of approval in the national legislature.
176
For formal
amendments to the Crown, the Constitution’s fundamental
principles and its entrenchment of rights and liberties, the power to
initiate an amendment similarly rests with the government, either
house of the national legislature, or the legislatures of the
autonomous communities. Ratification requires a different
procedure: two-thirds approval in each house of the national
legislature, followed by the new elections for the national legislature
and two-thirds approval from each of the newly constituted
chambers, followed by ratification by referendum.
177
5. Exceptional Single-Track Amendment
TABLE 10: EXCEPTIONAL SINGLE-TRACK FRAMEWORK IN DEMOCRATIC
CONSTITUTIONS
Australia
Austria
Iceland
The exceptional single-track and exceptional multi-track
frameworks are similar to the restricted single-track and multi-
track frameworks, respectively. Both exceptional frameworks differ
from the restricted frameworks only insofar as each of the
exceptional frameworks entrenches only two amendment
procedures: one default amendment procedure that applies
generally to all amendable constitutional provisions, and one special
amendment procedure that applies exclusively to a single
constitutional provision or a set of related provisions. Under both
exceptional frameworks, the special amendment procedure
incorporates the first procedure within it. Amending the specially
entrenched provision or set of related provisions therefore requires
the successful completion of one of the available amendment
procedures for making a formal amendment and an additional
procedure that is not otherwise required for amending other
constitutional provisions. This additional procedure is what is most
significant about the exceptional frameworks of formal
amendment—it is reserved exceptionally for only one constitutional
provision or one set of related provisions.
175
. Id. at B.O.E. n. 167, para. 2.
176
. Id. at B.O.E. n. 167, para. 3.
177
. Id. at B.O.E. n. 168.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 947
Of the thirty-six constitutions in this study, only three entrench
the single-track exceptional framework.
178
A useful example of this
exceptional single-track framework is the Icelandic Constitution.
179
The constitutional text provides a general amendment formula,
180
under which amendment proposals must be introduced in the
unicameral national legislature. If the legislature adopts the
proposal, the legislature must then be dissolved for new legislative
selections. Once it is newly reconstituted, the legislature must then
once again adopt the amendment proposal unchanged, and the
President must subsequently confirm the amendment.
181
This
formal amendment procedure applies to all amendable
constitutional provisions except the following: “The Evangelical
Lutheran Church shall be the State Church in Iceland and, as such,
it shall be supported and protected by the State.”
182
Any formal
amendment to this provision requires an additional step: a national
referendum.
183
This exceptional procedure for amendments related
to the establishment of the Evangelical Lutheran Church illustrates
the exceptional dimension of Iceland’s formal amendment
procedures. Exclusivity is an important design feature with useful
applications, as I will show in the pages below.
184
6. Exceptional Multi-Track Amendment
TABLE 11: EXCEPTIONAL MULTI-TRACK FRAMEWORK IN DEMOCRATIC
CONSTITUTIONS
United States
Only one democratic constitution entrenches the exceptional
multi-track framework: the U.S. Constitution. The text establishes
four ways to amend the Constitution: (1) proposal by two-thirds of
each house of the national legislature and ratification by three-
quarters of the subnational legislatures; (2) proposal by two-thirds
of each house of the national legislature and ratification by three-
quarters of subnational conventions; (3) petition by two-thirds of the
178
. AUSTRALIAN CONSTITUTION s 128; BUNDES-VERFASSUNGSGESETZ [B-VG]
BGBl No. 1/1920, as last amended by Bundesverfassungsgesetz [BVG] BGBl I
No. 2/1983, arts. 34–35, 44, ¶ 2 (Austria); STJÓRNARSKRÁ LÝÐVELDISINS ÍSLANDS,
art. 79 (Ice.).
179
. Residents of Iceland have recently voted in a national referendum in
favor of rewriting their Constitution in light of the 2008 banking crisis. See
Robert Robertsson, Voters in Iceland Back New Constitution, More Resource
Control, REUTERS (Oct. 21, 2012, 10:20 AM), http://www.reuters.com/article
/2012/10/21/iceland-referendum-idUSL5E8LK1TE20121021.
180
. STJÓRNARSKRÁ LÝÐVELDISINS ÍSLANDS, art. 79 (Ice.).
181
. Id.
182
. Id. at art. 62.
183
. Id. at art. 79.
184
. See infra Subpart III.A.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
948 WAKE FOREST LAW REVIEW [Vol. 49
subnational legislatures for a convention to propose one or more
constitutional amendments, and ratification by three-quarters of the
subnational legislatures; and (4) petition by two-thirds of the
subnational legislatures for a convention to propose one or more
constitutional amendments, and ratification by three-quarters of
subnational conventions.
185
Any of these four formal procedures
may be used to amend any provision in the U.S. Constitution except
the Equal State Suffrage Clause, which requires that each state
hold equal voting power in the Senate.
186
The rule for amending the Equal State Suffrage Clause
illustrates why the U.S. Constitution exhibits the exceptional multi-
track framework of formal amendment. In order to amend the
Equal State Suffrage Clause provision, the Constitution requires
adherence to one of these four formal procedures in addition to the
consent of the state or states whose voting power is affected by that
amendment.
187
The additional step of securing state consent is
required only if political actors also successfully complete one of the
four formal procedures.
188
If they fail in that attempt, securing state
consent alone will not result in a formal amendment. Both steps are
necessary for amending the Equal State Suffrage Clause:
successfully completing one of the four formal amendment
procedures and securing state consent. This higher amendment
threshold applies exceptionally to the Equal State Suffrage Clause.
D. The Specifications of Formal Amendment
Formal amendment rules are therefore anchored in one of two
foundations and structured around one of six frameworks. Yet these
foundations and frameworks are neither self-executing nor do they
provide the entire blueprint to formal amendment. They must be
supplemented by specifications that set into motion their operation.
These formal specifications are written into the text of the
constitution and are expressly designed as operational restrictions
on the formal amendment process that political actors must
navigate to formally amend the constitution. Of the several types of
formal specifications, five of them appear with relative frequency in
written constitutions: voting thresholds and quorum requirements,
185
. U.S. CONST. art. V.
186
. Id. (“Provided . . . that no State, without its Consent, shall be deprived
of its equal Suffrage in the Senate.”).
187
. It is theoretically possible that the affected state could express its
consent along with other states at the three-quarters ratification stage. This
would obviate the need for a separate step to obtain the affected state’s consent.
188
. U.S. CONST. art. V. The Equal Suffrage Clause may be read as
requiring the unanimous consent of the states. See Sanford Levinson,
Designing an Amendment Process, in CONSTITUTIONAL CULTURE AND
DEMOCRATIC RULE 271, 284 (John Ferejohn et al. eds., 2001).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 949
subject-matter restrictions, temporal requirements, electoral
preconditions, and defense mechanisms.
1. Voting Thresholds and Quorum Requirements
All formal amendment rules entrench thresholds specifying the
quantum of agreement needed to use any of their procedures.
Constitutional designers have wide latitude to tailor these
thresholds to the appropriate specifications in their jurisdictions;
each of the six frameworks of formal amendment therefore
entrenches various voting thresholds. We find examples of simple or
supermajority voting thresholds in the comprehensive single-
track,
189
comprehensive multi-track,
190
restricted single-track,
191
restricted multi-track,
192
exceptional single-track,
193
and exceptional
multi-track frameworks.
194
We also find voting thresholds requiring
at least two of the following in comprehensive, restricted and
exceptional frameworks: supermajority vote, a combination of
national and subnational approval, and referendum.
195
Constitutional designers have also specified quorum
requirements that political actors must meet to validly deploy
formal amendment rules. Where political actors do not achieve
these quorum requirements, the formal amendment rules are
effectively rendered inoperative. In Denmark, for example, the
Constitution requires a referendum to ratify an amendment
proposal approved by the national legislature.
196
In order to validly
189
. See IR. CONST., 1937, art. 46, para. 2 (requiring a referendum);
Constitution Act, 1986, pt. 3 s 15 (N.Z.) (requiring a majority vote in national
legislature); GRUNNLOVEN [GRL.] § 112 (Nor.) (requiring a supermajority vote in
the national legislature).
190
. See SUOMEN PERUSTALSKI, 6 luku 73 (Fin.) (authorizing amendment by a
combination of simple and supermajority vote, or extraordinary supermajority
vote); BUNDESVERFASSUNG [BV] Apr. 18, 1999, SR 101, arts. 192–95 (Switz.)
(authorizing referendum).
191
. See INDIA CONST. art. 368 (authorizing amendment by a combination of
simple majority and supermajority vote).
192
. See CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE CHILE [C.P.] arts. 127–
29 (authorizing amendment by supermajority vote in the national legislature).
193
. See AUSTRALIAN CONSTITUTION s 128 (requiring simple majority vote in
the national legislature and in subnational states).
194
. See U.S. CONST. art. V (requiring supermajority approval in the
national legislature and subnational states).
195
. See AUSTRALIAN CONSTITUTION s 128 (requiring a combination of
referendum and national/subnational approval under exceptional single-track
framework); Constitution Act, 1982, being Schedule B to the Canada Act, 1982,
s. 38–48 (U.K.) (requiring supermajority and national/subnational approval
under the restricted multi-track framework); NIHONKOKU KENPŌ [Kenpō], art. 96
(Japan) (requiring supermajority vote and referendum under a comprehensive
single-track framework); MAURITIUS CONST., art. 47, para. 3 (requiring
referendum and extraordinary supermajority in the legislature under a
restricted single-track framework).
196
. GRUNDLOVEN [GRL] Lov nr. 88 (Den.).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
950 WAKE FOREST LAW REVIEW [Vol. 49
ratify the amendment proposal, at least forty percent of the
electorate must participate, and of those a simple majority must
vote in favor.
197
Under South Korea’s referendum requirement, a
simple majority of eligible voters must approve a proposed
amendment.
198
The Canadian Constitution entrenches a variation
on the conventional quorum requirement: its general amending
formula requires approval from a simple majority in both houses of
the national legislature, as well as simple majority approval from at
least two-thirds of subnational legislatures whose population
represents at least half of the entire provincial population according
to the latest census.
199
Constitutional designers therefore have
many options to design quorum requirements.
200
2. Subject-Matter Restrictions
Democratic constitutions generally entrench two types of
provisions: the first are susceptible to the formal amendment
procedures entrenched in the text; the second are absolutely
immune to them. These unamendable provisions establish explicit
restrictions on the subject matter of formal amendments. Subject-
matter restrictions may preclude formal amendments to
secularism,
201
republicanism,
202
or democracy.
203
In addition, they
may prohibit formal amendments that suppress or diminish
fundamental rights and freedoms,
204
create a single-party state,
205
197
. Id. The final step calls for Royal Assent. Id.
198
. DAEHANMINKUK HUNBEOB [HUNBEOB] art. 130, para. 2 (S. Kor.).
199
. Constitution Act, 1982, being Schedule B to the Canada Act, 1982, c. 38,
para. 1 (U.K.).
200
. Constitutional designers should be attentive to the choice of electoral
procedures as these may have an impact on the representativeness of formal
amendments. Where a regime authorizes formal amendments by constitutional
act or fundamental law and the legislature is constituted by a first-past-the-
post system, the legislative majority may in fact represent only a popular
minority. This is relevant to the popular legitimacy of formal amendments. On
this point, see David Landau, Abusive Constitutionalism, 47 U.C. DAVIS L. REV.
189, 208–16 (2013).
201
. See CONSTITUIÇÃO DA REPÚBLICA PORTUGUESA, art. 288, cl. c.
202
. See CONSTITUIÇÃO DA REPÚBLICA DE CABO VERDE, art. 313, para. 1, cl. b;
1958 CONST. art. 89 (Fr.); Art. 139 Costituzione [Cost.] (It.); CONSTITUIÇÃO DA
REPÚBLICA PORTUGUESA, art. 288, cl. b.
203
. See GRUNDGESETZ FÜR DIE BUNDESREPUBLIK DEUTSCHLAND
[GRUNDGESETZ] [GG] [BASIC LAW], May 23, 1949, BGBl. VII, art. 79, para. 3
(Ger.); id. at BGBl II, art. 20, para. 1.
204
. See CONSTITUIÇÃO DA REPÚBLICA DE CABO VERDE, art. 313, para. 2;
GRUNDGESETZ FÜR DIE BUNDESREPUBLIK DEUTSCHLAND [GRUNDGESETZ] [GG]
[BASIC LAW], May 23, 1949, BGBl. VII, art. 79, para. 3 (Ger.); id. at BGBl. I, art.
1; CONSTITUIÇÃO DA REPÚBLICA PORTUGUESA, art. 288, cl. d.
205
. See CONSTITUIÇÃO DA REPÚBLICA DE CABO VERDE, art. 313, para. 1, cl. g;
CONSTITUIÇÃO DA REPÚBLICA PORTUGUESA, art. 288, cl. i.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 951
redraw national borders,
206
or violate the separation of powers.
207
Other subject-matter restrictions disallow formal amendments that
violate federalism
208
or international law.
209
Yet subject-matter restrictions often fail to actually protect a
provision. First, no degree of constitutional entrenchment can
withstand revolution.
210
Constitutional designers cannot escape this
risk; all constitutions are vulnerable to this nonconstitutional form
of change. Constitutional designers can, however, avoid the faulty
constitutional design that undermines their effort to entrench a
subject-matter restriction. Consider the French Constitution, which
entrenches a subject-matter restriction on republicanism.
211
This
subject-matter restriction originated in 1884, when Jules Ferry, an
opponent of monarchy and proponent of republicanism,
212
proposed
to entrench the following provision: “The Republic form of the
Government cannot be made the subject of a proposed revision.
Members of families that have reigned in France are ineligible to
the presidency of the Republic.”
213
Today, under the French
Constitution, “the Republican form of government shall not be the
object of any amendment.”
214
The design of this subject-matter restriction illustrates the
second reason why its entrenchment fails to protect it: the text
absolutely entrenches republicanism, but it does not absolutely
entrench itself against formal amendment. Constitutional
reformers in France could therefore lawfully deploy the
comprehensive multi-track framework to circumvent the subject-
matter restriction on amendments to republicanism by amending
the entrenching provision. This design flaw is fatal to the effort to
establish a subject-matter restriction. Rewriting the French
206
. See CONSTITUIÇÃO DA REPÚBLICA DE CABO VERDE, art. 313, para. 1, cl. a;
1958 CONST. art. 89 (Fr.).
207
. See 1975 SYNTAGMA [SYN.] 2, art. 110, para. 1 (1975) (Greece);
CONSTITUIÇÃO DA REPÚBLICA PORTUGUESA, art. 288, cl. j.
208
. See GRUNDGESETZ FÜR DIE BUNDESREPUBLIK DEUTSCHLAND
[GRUNDGESETZ] [GG] [BASIC LAW], May 23, 1949, BGBl. VII, art. 79, para. 3
(Ger.)
209
. See BUNDESVERFASSUNG [BV] Apr. 18, 1999, SR 101, art. 194, para. 2
(Switz.)
210
. JEFFREY GOLDSWORTHY, PARLIAMENTARY SOVEREIGNTY: CONTEMPORARY
DEBATES 70 (2010). Relatedly, subject-matter restrictions also fail in the face of
wholesale constitutional replacement that may be democratic in form though
not in fact. See Landau, supra note 200, at 237–41.
211
. 1958 CONST. art. 89 (Fr.).
212
. See Nathalie Droin, Retour sur la loi constitutionnelle de 1884:
contribution à une histoire de la limitation du pouvoir constituent derive, 80
REVUE FRANÇAISE DE DROIT CONSTITUTIONNEL 725, 740 (2009) (Fr.).
213
. Law Partially Revising the Constitutional Laws, Aug. 14, 1884, in
CONSTITUTIONAL AND ORGANIC LAWS OF FRANCE 168, 168 (Charles F.A. Currier
ed. & trans. 1893).
214
. 1958 CONST. art. 89 (Fr.).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
952 WAKE FOREST LAW REVIEW [Vol. 49
Constitution’s entrenching clause in light of this exposed design flaw
would yield a provision that had been minimally, textually revised
yet substantively transformed, stating henceforth that “neither this
clause nor the Republican form of government shall be the object of
any amendment.” Constitutional designers may therefore entrench
both the subject-matter restriction as well as the entrenching
provision at little additional political cost.
3. Temporal Limitations
Formal amendment specifications also limit formal
amendments with respect to the timing of the various steps
comprising the amendment process. We can identify two general
types of these temporal limitations: deliberation requirements, of
which there are two kinds (deliberation ceilings and deliberation
floors)
215
and safe harbor provisions.
216
Temporal limitations of both
varieties are commonly entrenched in written constitutions.
217
TABLE 12: TEMPORAL LIMITATIONS IN DEMOCRATIC CONSTITUTIONS
Deliberation Requirements
Australia
Chile
Costa Rica
Italy
Luxembourg
South Korea
Sweden
Safe Harbor Provisions
Cape Verde
Estonia
Greece
Uruguay
215
. See, e.g., CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE COSTA RICA, art.
195, para. 3; DAEHANMINKUK HUNBEOB [HUNBEOB] art. 129 (S. Kor.).
216
. See, e.g., CONSTITUIÇÃO DA REPÚBLICA DE CABO VERDE, art. 309, para. 1;
PÕHISEADUS, s 168 (Est.).
217
. Democratic constitutions commonly entrench deliberation
requirements. See AUSTRALIAN CONSTITUTION s 128; CONSTITUCIÓN POLÍTICA DE
LA REPÚBLICA DE CHILE [C.P.] art. 129; CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA
DE COSTA RICA, art. 195; Art. 138 Costituzione [Cost.] (It.); CONSTITUTION OF
LUXEMBOURG, art. 114; DAEHANMINKUK HUNBEOB [HUNBEOB] arts. 129–30 (S.
Kor.); REGERINGSFORMEN [RF] 8, art. 14 (Swed.). Safe harbors are also
commonly entrenched. See, e.g., CONSTITUIÇÃO DA REPÚBLICA DE CABO VERDE,
art. 309, para. 1; PÕHISEADUS, s 168 (Est.); 1975 SYNTAGMA [SYN.] 2, art. 110,
para. 6 (Greece); CONSTITUCIÓN DE 1967, art. 331, cl. c (Uru.).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 953
The Costa Rican Constitution provides an instructive example
of a deliberation ceiling.
218
The Constitution requires a special
commission to render advice on the proposed amendment within no
more than twenty working days.
219
This is the upper limit for the
commission to deliberate on the matter before the proposal proceeds
through other steps.
Conversely, the South Korean Constitution offers an example of
a deliberation floor.
220
A deliberation floor establishes a minimum
period for deliberating on a proposal prior to a binding vote or action
on the proposal.
221
In order to formally amend the South Korean
Constitution, the President must make an amendment proposal
public for at least twenty days: “Proposed amendments to the
Constitution shall be put before the public by the President for
twenty days or more.”
222
Deliberation floors are the corollary of
deliberation ceilings. But rather than establishing an upper time
limit for deliberating on an amendment proposal, deliberation floors
require either political actors, the public, or both, to consider an
amendment proposal for a minimum period of time.
Formal amendment rules are also sometimes temporally limited
by a safe harbor. Whereas deliberation requirements compel
political actors to consider the merits and demerits of an
amendment proposal over the course of a defined period of time, safe
harbors do the opposite: they prohibit political actors from making
amendment proposals for a defined period of time. Safe harbors
might, for example, foreclose political actors from reintroducing a
defeated amendment proposal until the passage of a defined period.
The Estonian Constitution adopts this model.
223
Alternatively, safe
harbors might ban amendment proposals for a fixed number of years
beginning immediately upon the ratification of a new constitution.
The Cape Verdean Constitution illustrates this safe harbor.
224
Safe
harbors might also prohibit subsequent formal amendments within
a defined period of time after the successful formal amendment of
218
. CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE COSTA RICA, art. 195, para.
3.
219
. Id.
220
. DAEHANMINKUK HUNBEOB [HUNBEOB] art. 129 (S. Kor.).
221
. See, e.g., id. (“Proposed amendments to the Constitution shall be put
before the public by the President for twenty days or more.”).
222
. Id.
223
. PÕHISEADUS, s 168 (Est.) (stating that an amendment to the
Constitution regarding the same issue shall not be initiated within one year
after the rejection of a corresponding bill by a referendum or by the Riigikogu).
224
. CONSTITUIÇÃO DA REPÚBLICA DE CABO VERDE, art. 309, para. 1 (creating
a five-year safe harbor).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
954 WAKE FOREST LAW REVIEW [Vol. 49
the constitution. The Greek and Portuguese Constitutions entrench
this type of safe harbor.
225
Temporal limitations appear to serve the interest of
constitutional peace. Deliberation ceilings give notice to political
actors that the window for deliberating upon a formal amendment
will not remain open indefinitely. They circumscribe amendment
debates within a finite period of time during which political actors
are compelled to proceed apace to consider the merits and raise any
possible demerits about the proposal before time elapses. Were
there no time limit within which to act, it would be possible to
imagine the amendment debate enduring for years—perhaps even
decades.
226
Relatedly, deliberation floors frustrate political actors
endeavoring to rush the formal amendment process.
By requiring a
minimum period of deliberation, deliberation floors reserve time to
consider the amendment proposal. Safe harbors serve the same
purpose, though from a different angle. Entrenching a safe harbor
to a new constitution allows the document to take root and avoid the
disruption that formal constitutional amendment entails. Likewise,
a safe harbor on reintroducing a failed amendment requires political
actors to redirect their attention elsewhere, thereby giving both
themselves and the constitution a respite. Constitutional designers
should consider whether and when their own constitutional states
might benefit from entrenching temporal limitations in the formal
amendment process.
227
4. Electoral Preconditions
Constitutions also impose electoral preconditions upon formal
amendment rules, often requiring successive votes separated by an
election. For instance, some prohibit the same voting body from
both proposing and ratifying formal amendments without an
intervening national election to reconstitute the body between both
votes.
228
For instance, the Norwegian Constitution requires the
225
. 1975 SYNTAGMA [SYN.] 2, art. 110, para. 6 (Greece) (establishing a five-
year safe harbor); CONSTITUIÇÃO DA REPÚBLICA PORTUGUESA, art. 284, para. 1
(same).
226
. This is precisely what has transpired in the United States. The
Twenty-Seventh Amendment was originally proposed in 1789, but it was not
ratified until two centuries later in 1992. See Res. 3, 1st Cong. (1789)
(proposing amendment); 57 Fed. Reg. 21187 (May 18, 1992) (certifying
amendment).
227
. For a comprehensive analysis of the use of temporal limitations in
constitutional design, see generally Ozan O. Varol, Temporary Constitutions,
102 CALIF. L REV. 409 (2014) (investigating how constitutional designers do and
should consider time in designing constitutions).
228
. See 1994 CONST. art. 195 (Belg.); GRUNDLOVEN [GRL] Lov nr. 88 (Den.);
1975 SYNTAGMA [SYN.] 2, art. 110, para. 3 (Greece); STJÓRNARSKRÁ LÝÐVELDISINS
ÍSLANDS, art. 79 (Ice.); Grondwet voor het Koninkrijk der Nederlanden [GW.] art.
137 (Neth.). Some constitutions establish this rule with exceptions. See, e.g.,
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 955
unicameral national legislature to propose a formal amendment, but
the proposal can only be approved by a newly elected legislature
following a general election.
229
The South Korean Constitution
illustrates a variation on this restriction. It establishes an electoral
precondition designed to frustrate the entrenchment of a formal
amendment from which an incumbent president would profit:
“Amendments to the Constitution for the extension of the term of
office of the President or for a change allowing for the reelection of
the President shall not be effective for the President in office at the
time of the proposal for such amendments to the Constitution.”
230
These electoral preconditions seem intended to prevent self-dealing
that would either benefit incumbent political actors or disadvantage
their adversaries.
5. Defense Mechanisms
Constitutional designers have also entrenched defense
mechanisms within formal amendment rules. Whereas voting
thresholds, quorum requirements, subject-matter restrictions,
temporal limitations, and electoral conditions constrain how formal
amendment rules are used, defense mechanisms disable the formal
amendment process altogether. Spurred by fears that the
amendment process could be hijacked by foreign or nefarious
influences, rushed in the face of a national emergency, or
compromised during times of war or instability, these defense
mechanisms remove the power of formal amendment from political
actors. It is not unusual for constitutional designers to disable the
formal amendment process during a national emergency, martial
law, or a state of siege or war.
231
For example, the Spanish
Constitution orders that “the process of constitutional amendment
may not be initiated in time of war or under any of the states
contemplated in Article 116,”
232
which includes “alarm, emergency
and siege (martial law).”
233
The Estonian Constitution declares that
“amendment of the Constitution shall not be initiated, nor shall the
Constitution be amended, during a state of emergency or a state of
war.”
234
Constitutions may also disable the formal amendment process
during periods of regency or succession. When the monarch is
absent or unable to serve, constitutions prohibit formal
PÕHISEADUS, s 165 (Est.); SUOMEN PERUSTALSKI, 6 luku 73 (Fin.);
REGERINGSFORMEN [RF] 8, art. 16 (Swed.).
229
. See GRUNNLOVEN [GRL.] § 112 (Nor.).
230
. DAEHANMINKUK HUNBEOB [HUNBEOB] art. 128, para. 2 (S. Kor.).
231
. See 1994 CONST. art. 196 (Belg.); CONSTITUIÇÃO DA REPÚBLICA DE CABO
VERDE, art. 315; CONSTITUIÇÃO DA REPÚBLICA PORTUGUESA, art. 289.
232
. CONSTITUCIÓN ESPAÑOLA [C.E.], B.O.E. n. 169, Dec. 29, 1978 (Spain).
233
. Id. at B.O.E. n. 116.
234
. PÕHISEADUS, s 161 (Est.).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
956 WAKE FOREST LAW REVIEW [Vol. 49
amendments,
235
perhaps out of fear that the regent named as
steward would seek to advantage himself via constitutional
amendment during the monarch’s absence. For instance, in
Luxembourg, the Constitution declares that “during a regency, no
change can be made to the Constitution concerning the
constitutional prerogatives of the Grand Duke, his status as well as
the order of succession.”
236
The Belgian Constitution imposes a
similar prohibition: “During a regency, no change can be made in
the Constitution with respect to the constitutional powers of the
King and Articles 85 to 88, 91 to 95, 106 and 197 of the
Constitution.”
237
Borrowing from these monarchical design
strategies, constitutional designers could impose analogous
prohibitions in republican forms of government during periods of
interim government or when a vacancy exists in a high office (such
as the presidency). Constitutional designers could likewise disable
formal amendment rules within a defined period of time prior to an
election. This would help guard against illegitimate, though
perhaps not illegal, efforts to exploit formal amendment rules to
shape the result in change-of-control contests.
III. DESIGNING FORMAL AMENDMENT RULES
Formal amendment rules are therefore structured in three tiers.
At their base, they are anchored in one of two underlying
foundations. The first entrenches a distinction between formal
amendment and constitutional revision. The second makes no
distinction between amendment and revision but, as I will show
below,
238
is susceptible to judicially imposed restrictions that mirror
the distinction between amendment and revision. The second tier in
the structure of formal amendment rules is one of six amendment
frameworks: comprehensive, restricted, or exceptional single track;
or comprehensive, restricted, or exceptional multi-track. Formal
amendment rules also consist of supplementary specifications—for
instance, voting thresholds, quorum requirements, content
restrictions, temporal requirements, electoral preconditions, and
defense mechanisms—that may be used in combination to complete
the structure of formal amendment.
This three-tiered classification of the foundations, frameworks,
and specifications of formal amendment rules helps identify and
understand formal amendment design possibilities, and it is useful
as a guide for constitutional designers to achieve their objectives.
Where, for instance, constitutional designers wish to entrench
235
. See 1994 CONST. art. 197 (Belg.); CONSTITUTION OF LUXEMBOURG, art.
115.
236
. CONSTITUTION OF LUXEMBOURG, art. 115.
237
. 1994 CONST. art. 197 (Belg.).
238
. See infra Subpart III.C.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 957
special protections for a given constitutional provision or a family of
related provisions, they may choose to textually entrench subject-
matter restrictions against formal amendment
239
and they may
likewise specially entrench that provision or family of related
provisions through one of the restricted or exceptional amendment
frameworks.
240
Likewise where, for example, constitutional
designers wish to keep the formal amendment process
uncomplicated, transparent, and subject to little misinterpretation,
they may entrench the comprehensive single-track amendment
framework.
241
In this final Part, I explore and demonstrate how
constitutional designers may use this new classification to reinforce
federalism, express constitutional values, enhance or diminish the
judicial role, and pursue democratic outcomes relating to
governance, constitutional endurance, and amendment ease or
difficulty. I reiterate here what I stated at the outset of this study:
the purpose of this Article is to re-enliven the study of formal
amendment rules by demonstrating its possibilities in constitutional
design.
A. Federalism Safeguards
The longest surviving written constitution—the U.S.
Constitution—entrenches federalism safeguards in its formal
amendment rules using the exceptional multi-track framework. The
Equal Suffrage Clause,
242
described above,
243
may be formally
amended only where political actors satisfy one of the Constitution’s
four amendment procedures and also secure the consent of the state
whose equal suffrage in the Senate is diminished.
244
No other
provision requires this form of particularized consent. We know
from the records of the Philadelphia Convention of 1787 that the
Clause was a necessary bargain between large and small states.
245
Without the protection conferred to small states by the requirement
that a state consent to any diminution of its voting power in the
Senate, small states would not have ratified the Constitution.
246
As
239
. See supra Subpart II.C.2.
240
. See supra Subparts II.B.3–6.
241
. See supra Subpart II.B.1.
242
. U.S. CONST. art. V (“Provided . . . that no State, without its Consent,
shall be deprived of its equal Suffrage in the Senate.”).
243
. See supra Subpart II.A.
244
. U.S. CONST. art. V.
245
. 2 THE RECORDS OF THE FEDERAL CONVENTION OF 1787, at 629–31 (Max
Farrand ed., 1911).
246
. Equal suffrage in the Senate was nonnegotiable for the smaller states.
1 THE RECORDS OF THE FEDERAL CONVENTION OF 1787, at 196, 201 (Max Farrand
ed., 1911). James Madison recollected that the Clause “was probably insisted on
by the States particularly attached to that equality.” THE FEDERALIST No. 43, at
296 (James Madison) (Jacob E. Cooke ed., 1961).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
958 WAKE FOREST LAW REVIEW [Vol. 49
Douglas Smith explains, the Clause was a “constitutional
essential.”
247
The two exceptional amendment frameworks—single track and
multi-track—share two properties that distinguish them from other
frameworks: first, they create two tiers of amendment thresholds;
and second, the higher threshold applies exclusively to one
constitutional provision or one set of related constitutional
provisions, whereas the lower threshold applies to all other
constitutional provisions. The higher threshold is cumulative
insofar as it incorporates the lower threshold within itself. These
are two design features of the exceptional amendment frameworks.
They can also be distinguished on the basis of how the frameworks
are actually used: constitutional designers have entrenched them
predominantly to safeguard federalism.
The federalism-reinforcing function of the exceptional
frameworks is evident in two other democratic constitutions with
one of these frameworks, namely the Australian and Austrian
Constitutions.
248
To formally amend the Australian Constitution,
each house of the bicameral national legislature must adopt a
proposal by a simple majority.
249
Between two and six months
thereafter, the amendment proposal must be presented to all
Australian voters in a national referendum.
250
If a nationwide
majority of voters representing majorities in a majority of states
approve the proposal,
251
it proceeds to the final step—assent by the
Governor-General.
252
But the Australian Constitution’s formal
amendment rules entrench an additional requirement for formal
amendments that change the balance of powers between the
national and subnational states. Where the formal amendment
affects the powers, boundaries, or representation of a state, the
amendment must also be ratified by a majority of voters in that
247
. Douglas G. Smith, An Analysis of Two Federal Structures: The Articles
of Confederation and the Constitution, 34 SAN DIEGO L. REV. 249, 322 (1997).
248
. Of the four democratic constitutions entrenching the single-track or
multi-track exceptional framework of formal amendment—Australia, Austria,
Iceland, and the United States—only Iceland uses the exceptional framework to
grant special protection to a constitutional provision that is not expressly
related to federalism. See STJÓRNARSKRÁ LÝÐVELDISINS ÍSLANDS, arts. 62, 79
(Ice.) (applying the exceptional amendment framework to the established
church).
249
. AUSTRALIAN CONSTITUTION s 128.
250
. Id.
251
. Id.
252
. Id. The office of Governor-General is partly ceremonial, constitutional,
mystical, and practical. For more on the evolution of the office in Australia, see
generally Greg Craven, The Developing Role of the Governor-General: The
Goldenness of Silence, 32 FED. L. REV. 281 (2004) (Austl.) (illustrating the many
functions of the Governor-General).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 959
affected state.
253
This is similar to the structure of the Equal
Suffrage Clause in the U.S. Constitution.
254
The exceptional framework also applies in Austria. The
Austrian bicameral national legislature consists of the nationally,
popularly elected National Council, and the Federal Council, whose
members are chosen proportionately by population by the
subnational legislatures.
255
The Federal Council is therefore where
subnational interests are more directly represented. Consistent
with the exceptional framework, the Austrian Federal
Constitutional Law establishes a formal amendment threshold for a
special class of provisions and another for all other provisions. The
general amendment rule requires approval by two-thirds
supermajority vote of a quorum of only one-half in the National
Council in order to pass a constitutional law or provision.
256
But for
constitutional laws or provisions that restrict the powers and
prerogatives of subnational states, the Federal Constitutional Law
also requires a quorum of one-half in the Federal Council and two-
thirds approval in the Federal Council.
257
The Federal
Constitutional Law also protects the design, election rules, and
eligibility requirements for the Federal Council with a formal
amendment rule not unlike the Equal Suffrage Clause in Article V
of the U.S. Constitution. In order to amend the design, election
rules, or eligibility requirements for the Federal Council, political
actors must also secure the consent of a majority of members in the
Federal Council representing at least four of the nine subnational
states.
258
253
. AUSTRALIAN CONSTITUTION s 128 (“No alteration diminishing the
proportionate representation of any State in either House of the Parliament, or
the minimum number of representatives of a State in the House of
Representatives, or increasing, diminishing, or otherwise altering the limits of
the State, or in any manner affecting the provisions of the Constitution in
relation thereto, shall become law unless the majority of the electors voting in
that State approve the proposed law.”).
254
. The South African Constitution also protects subnational entities by
ensuring that if a formal amendment “concerns only a specific province or
provinces, the National Council of Provinces may not pass the Bill or the
relevant part unless it has been approved by the legislature or legislatures of
the province or provinces concerned.” S. AFR. CONST., 1996, sec. 74, para. 8.
255
. See BUNDES-VERFASSUNGSGESETZ [B-VG] BGBl No. 1/1920, as last
amended by Bundesverfassungsgesetz [BVG] BGBl I No. 2/1983, arts. 26, 34–35
(Austria).
256
. Id. at art. 44, ¶ 1.
257
. Id. at art. 44, ¶ 2.
258
. Id. at art. 35, ¶ 4.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
960 WAKE FOREST LAW REVIEW [Vol. 49
TABLE 13: FEDERALISM REINFORCEMENT IN AMENDMENT DESIGN IN
DEMOCRATIC CONSTITUTIONS
Exceptional Single-Track
Framework
Australia
Austria
Exceptional Multi-Track
Framework
United States
Constitutional designers must often be particularly attentive to
the vertical separation of powers between national and subnational
governments. They can learn from these examples of the
exceptional framework of formal amendment inasmuch as Australia,
Austria, and the United States have been successful to moderately
successful in their management of federalism. Although the
strength of federal relations in these constitutional states has not
hinged exclusively (nor even primarily) on the design of formal
amendment rules, the extent to which formal amendment rules give
voice and representation to subnational states within the larger
federal structure is a consideration that constitutional designers
should not discount, particularly at the early stages of constitution
design where the design of constitutional rules are as important for
the aspiration or intent they reflect as the binding effect they
ultimately have. Granting to subnational states veto power over
formal amendments that affect the distribution and scope of
subnational powers serves both as a signal of good faith in
contentious constitutional design and a useful check on national
government as federalism evolves under the new constitution.
B. Constitutional Values
Constitutional designers commonly express constitutional
values in the preamble. For instance, Japan’s postwar Constitution
expresses its commitment to peace and nonaggression in its
preamble: “We, the Japanese people, desire peace for all time and
are deeply conscious of the high ideals controlling human
relationship, and we have determined to preserve our security and
existence, trusting in the justice and faith of the peace-loving
peoples of the world.”
259
The preamble moreover declares that the
Constitution is founded in the “universal principle” of popular
sovereignty: “Government is a sacred trust of the people, the
259
. NIHONKOKU KENPŌ [KENPŌ], pmbl. (Japan).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 961
authority for which is derived from the people, the powers of which
are exercised by the representatives of the people, and the benefits
of which are enjoyed by the people. This is a universal principle of
mankind upon which this Constitution is founded.”
260
The current
Japanese prime minister has recently undertaken controversial
efforts to amend the Constitution, including the preamble, to reflect
what he argues are Japan’s modern constitutional values.
261
Constitutional designers also express constitutional values
elsewhere. For example, the non-preambular text of the Finnish
Constitution declares that it will “guarantee the inviolability of
human dignity and the freedoms and rights of the individual and
promote justice in society,” and adds, “Finland participates in
international co-operation for the protection of peace and human
rights and for the development of society.”
262
Similarly, the South
African Constitution declares outside the preamble that the state is
founded on the values of human dignity, equality, human rights and
freedoms, non-racialism, non-sexism, constitutional supremacy, the
rule of law, universal adult suffrage, voter registration, regular
elections, accountability, responsiveness, transparence, and multi-
party democratic government.
263
Spain also entrenches a statement
of constitutional values in the non-preambular text of its
constitution: “Spain is hereby established as a social and democratic
State, subject to the rule of law, which advocates freedom, justice,
equality and political pluralism as highest values of its legal
system.”
264
This new classification of formal amendment rules suggests
that constitutional designers may express constitutional values in a
third site: formal amendment rules themselves.
265
Constitutional
designers may deploy special forms of entrenchment in their design
260
. Id.
261
. See Tobias Harris, Shinzo Abe’s Constitutional Quest, WALL ST. J. (May
16, 2013, 1:02 PM), http://online.wsj.com/article
/SB10001424127887323582904578486642338035044.html; Yuka Hayashi,
Japan Leader Charts Path for Military’s Rise, WALL ST. J. (Apr. 24, 2013, 6:10
PM), http://online.wsj.com/article
/SB10001424127887323551004578438253084917008.html; Yuka Hayashi, New
Headwinds for Constitution Campaign, WALL ST. J. BLOG (May 24, 2013, 4:22
PM), http://blogs.wsj.com/japanrealtime/2013/05/24/new-headwinds-for-
constitution-campaign; Tokujin Matsudaira, New Developments on Japan’s
Proposed Constitutional Amendment Process, INT’L J. CONST. L. BLOG (Apr. 23,
2013), http://www.iconnectblog.com/2013/04/new-developments-on-japans-
proposed-constitutional-amendment-process/.
262
. SUOMEN PERUSTALSKI, 1 luku 1 (Fin.).
263
. See S. AFR. CONST., 1996, sec. 1.
264
. CONSTITUCIÓN ESPAÑOLA [C.E.], B.O.E. n. 1, Dec. 29, 1978 (Spain).
265
. I have theorized and illustrated how formal amendment rules may be
used to express constitutional values. See generally Albert, supra note 5
(arguing that constitutional designers may express constitutional values in
their design of formal amendment rules).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
962 WAKE FOREST LAW REVIEW [Vol. 49
of formal amendment rules to identify what matters more or most in
a constitutional text. By incorporating special forms of
entrenchment into their design of certain constitutional provisions,
constitutional designers can thereby signal their importance. There
are obvious and less obvious ways to assign special forms of
entrenchment and to prioritize constitutional provisions in order to
express constitutional values. The more obvious ways include
assigning heightened protection to a constitutional provision within
the restricted frameworks of formal amendment, making a
constitutional provision absolutely unamendable using subject-
matter restrictions, or using both in combination. The less obvious
ways include entrenching the distinction between amendment and
revision, or combining temporal limitations with electoral
conditions.
Consider the restricted multi-track framework entrenched in
the Canadian Constitution. As the degree of procedural difficulty
increases from the first to the fifth amendment procedure, the
importance of the subject matter of the amendment procedures
likewise increases. Whereas provincial unilateral amendment may
amend a provincial constitution,
266
and parliamentary amendment
may be used to amend a constitutional provision dealing exclusively
with the House of Commons,
267
the Canadian Constitution requires
something more—the consent of Parliament and the affected
province—to amend provincial boundaries.
268
Similarly, the
Constitution requires a higher threshold—the approval of
Parliament and a supermajority of provinces—to amend provincial
representation in the Senate.
269
The Constitution reserves the
highest quantum of agreement—unanimous consent—for changes to
what its drafters regarded as the most important constitutional
provisions, namely Canada’s association with the monarchy.
270
What appears to underpin this framework is a hierarchy of
constitutional importance. This rigid escalating framework
channels political actors toward fixed routes for formal amendment.
These routes range from difficult to even more difficult to nearly
impossible. As the degree of procedural difficulty rises under this
restricted amendment framework, the subject matter of the
provisions assigned to these increasing amendment thresholds
likewise rises in importance, at least in Canada.
271
Constitutional
266
. Constitution Act, 1982, being Schedule B to the Canada Act, 1982, s. 45
(U.K.).
267
. Id. at s. 44.
268
. Id. at s. 43.
269
. Id. at s. 42.
270
. Id. at s. 41.
271
. See PETER W. HOGG, I CONSTITUTIONAL LAW OF CANADA 85 (5th ed. 2007
& Supp. 2010); Stephen A. Scott, Pussycat, Pussycat or Patriation and New
Constitutional Amendment Processes, 20 U. W. ONT. L. REV. 247, 303–04 (1982);
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 963
designers may therefore confer special or lesser status on
constitutional provisions by assigning varying degrees of formal
amendment difficulty and thereby establishing a constitutional
hierarchy in their design of formal amendment rules.
Constitutional designers may alternatively express
constitutional values by entrenching subject-matter restrictions.
Notwithstanding the naivety of a constitutional design that relies on
the force of mere words to protect the constitutional text from
amendment or replacement,
272
subject-matter restrictions
nonetheless reflect the considered judgment of constitutional
designers to set apart some provision or provisions of the
constitution. Whether this special entrenchment is spurred by a
political bargain, Ulyssian self-constraint, or the expression of
constitutional values, the predicate is the same: to designate a
constitutional provision as special—as the German Basic Law
does—by absolutely entrenching the concept of human dignity.
273
Constitutional designers may therefore specially entrench, and
thereby express, constitutional values with subject-matter
restrictions. They may also combine one of the restricted
frameworks with subject-matter restrictions, constructing a
constitutional hierarchy with escalating thresholds of amendment
difficulty that culminate with amendment impossibility.
The distinction between amendment and revision may similarly
be exploited to express constitutional values. By differentiating one
class of provisions subject exclusively to revision from another class
subject more freely to amendment, constitutional designers may
designate the former as not only more important, but also more
fundamental to the constitutional regime than the latter. For
instance, the Spanish Constitution establishes different voting
thresholds and amendment procedures for constitutional changes,
variously defined as amendments or revisions, with the revisionary
changes requiring substantially more difficult procedures.
274
Constitutional designers may adopt this distinction, make
provisions they deem more fundamental subject to revisionary
change, and thereby signal their relative importance.
Constitutional designers may also express constitutional values
using temporal limitations and electoral preconditions. For
instance, constitutional designers may impose deliberation
Walter Dellinger, The Amending Process in Canada and the United States: A
Comparative Perspective, 45 LAW & CONTEMP. PROBS. 283, 300 (1982).
272
. Melissa Schwartzberg, Should Progressive Constitutionalism Embrace
Popular Constitutionalism?, 72 OHIO ST. L.J. 1295, 1308 n.55 (2011).
273
. The German Basic Law holds that “human dignity shall be inviolable,”
GRUNDGESETZ FÜR DIE BUNDESREPUBLIK DEUTSCHLAND [GRUNDGESETZ] [GG]
[BASIC LAW], May 23, 1949, BGBl. I, art. 1, para. 1 (Ger.), and expressly
designates it as unamendable. See id. at BGBl. VII, art. 79, para. 3.
274
. CONSTITUCIÓN ESPAÑOLA [C.E.], B.O.E. n. 168, Dec. 29, 1978 (Spain).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
964 WAKE FOREST LAW REVIEW [Vol. 49
requirements for certain constitutional provisions and not others;
this is a less absolutist strategy than entrenching subject-matter
restrictions yet it conveys a similar message about the specialness of
a constitutional provision or a set of provisions. Constitutional
designers may also choose to impose safe harbors around a
constitutional provision or a set of provisions, even though safe
harbors commonly apply only to entire constitutions. Finally,
subjecting a constitutional provision or provisions to electoral
preconditions—for instance, requiring two successive national
legislative votes separated by national legislative elections—can
highlight constitutional importance. The same is true of combining
temporal requirements with electoral conditions, such as imposing a
safe harbor around a provision and requiring two successive
referenda separated by a reasonable number of years in order to
amend that provision.
C. The Judicial Role
Formal amendment in the United States and India
demonstrates the difficulty of predicting when democratic courts
will take an active role in managing the formal amendment
process.
275
Neither written constitution entrenches the distinction
between amendment and revision, but courts in each country have
taken different approaches on whether the formal amendment
process is bound by unwritten norms on amendability. In the
United States, the rules of formal amendment do not restrict what
may be amended using the normal processes of amendment.
276
The
Supreme Court has rejected claims that constitutional amendments
may be unconstitutional
277
but has agreed that the ratification of a
formal amendment must be reasonably contemporaneous with its
proposal in order to be constitutional.
278
The Court has, however,
disclaimed responsibility for determining contemporaneousness,
ceding this duty to Congress under the political question doctrine.
279
Yet in India the nonentrenchment of the distinction between
amendment and revision has not precluded its emergence. The
275
. Judicial review of constitutional amendments may be procedural or
substantive. For a leading analysis on this distinction as well as the difficulty
of identifying the line separating procedure from substance, see generally
KEMAL GÖZLER, JUDICIAL REVIEW OF CONSTITUTIONAL AMENDMENTS: A
COMPARATIVE STUDY (2008) (inquiring into the practice and theory of judicial
review of formal amendments around the world).
276
. See U.S. CONST. art. V. But the slave-traded and census-based taxation
were made unamendable until the year 1808. See id. Likewise, the composition
of the Senate is made subject to a special consent requirement. See id.
277
. See Leser v. Garnett, 258 U.S. 130, 136 (1922) (upholding Nineteenth
Amendment); Nat’l Prohibition Cases, 253 U.S. 350, 386 (1920) (upholding
Eighteenth Amendment).
278
. Dillon v. Gloss, 256 U.S. 368, 375 (1921).
279
. Coleman v. Miller, 307 U.S. 433, 454–55 (1939).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 965
Indian Supreme Court has developed the distinction by limiting the
national legislature’s power to amend the Constitution, even though
the constitutional text imposes no such substantive limits on the
formal amendment power.
280
In creating what is known as the basic
structure doctrine, the Supreme Court wrote that “[t]he true
position is that every provision of the Constitution can be amended
provided in the result the basic foundation and structure of the
Constitution remains the same.”
281
Neither constitutional
supremacy, republicanism, democracy, secularism, the separation of
powers, nor federalism can be “destroyed” by an amendment,
insisted the Court, suggesting that those foundational features of
Indian constitutionalism can be changed only with a new
constitution.
282
Constitutional designers should therefore not
presume that the non-entrenchment of the distinction will
necessarily lead to either its rejection, as in the United States, or its
judicial creation, as in India.
But there are ways to exclude the judiciary from matters
involving formal amendment. Constitutional designers intent on
diminishing the judicial role may learn from the Netherlands, where
formal amendment occurs through the legislative process with the
proposal and ratification of a law deemed constitutional.
283
Unlike
constitutional states where courts are granted by delegation or
acquiescence the power of judicial review and would therefore
possess the power to review the constitutionality of such laws, the
Netherlands forecloses this power from Dutch courts. The Dutch
Constitution prohibits courts from exercising the power of judicial
review: “The constitutionality of Acts of Parliament and treaties
shall not be reviewed by the courts.”
284
Dutch courts have
interpreted this provision broadly to prohibit them from reviewing
the constitutionality of laws against both written laws and
unwritten principles of natural law.
285
Constitutional designers intent on enhancing the judicial role
may purposely introduce ambiguity in their design of formal
amendment rules.
286
The two restricted amendment frameworks
280
. See INDIA CONST. art. 368.
281
. Kesavananda Bharati v. State of Kerala, (1973) 4 S.C.C. 225, 366
(India).
282
. Id.
283
. See Grondwet voor het Koninkrijk der Nederlanden [GW.] art. 137–42
(Neth.).
284
. Id. at art. 120.
285
. Gerhard van der Schyff, Constitutional Review by the Judiciary in the
Netherlands: A Bridge Too Far?, 11 GERMAN L.J. 275, 277–78 (2010).
286
. This applies in primarily common-law regimes where the judiciary
either asserts or is granted the authority of constitutional control. But courts
are not always endowed with, nor do they always assert, this power; it may
belong to other bodies, for instance legislatures or the constitution-making
body.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
966 WAKE FOREST LAW REVIEW [Vol. 49
appear from experience to be the most vulnerable to interpretive
ambiguity. By design, the restricted amendment frameworks create
more than one formal amendment procedure, and each is deployable
only for specifically enumerated constitutional provisions of the
constitution. Constitutional designers can exploit this restricted
framework by defining only in broad terms which constitutional
provisions are formally amendable by specific amendment
procedures, thereby leaving future political actors to resolve which
amendment procedure they must follow to formally amend specific
constitutional provisions. Though it was unlikely designed with this
intention in mind, the Canadian Constitution’s restricted multi-
track framework has created confusion about which of its five formal
amendment procedures applies to formally amending the long-
standing practice of prime ministerial Senate appointments.
287
This
is a matter of current constitutional controversy; the Government of
Canada recently requested an advisory opinion from the Supreme
Court to clarify these and related questions.
288
D. Democratic Outcomes
Constitutional designers may also use this classification to
measure how the choice of formal amendment rules maps onto
democratic outcomes in constitutional states. The relationship
between formal amendment design and democratic governance,
constitutional endurance, and amendment difficulty is of particular
interest to constitutional designers insofar as their design of formal
amendment rules is informed both by the suitability of a particular
foundation, framework, or specification for their own constitutional
287
. Part of the debate concerns which part of the formal amendment rules
will be engaged by the proposed changes to Senatorial appointments. Under
the Constitution’s formal amendment rules, amendments to “the method of
selecting Senators” must pass the default amendment threshold, requiring
approval from both houses of the national legislature and from two-thirds of the
subnational legislatures whose aggregate population represents at least half of
the population of all provinces combined. See Constitution Act, 1982, being
Schedule B to the Canada Act, 1982, s. 42 (U.K.). However, the Constitution’s
formal amendment rules also state that “subject to sections 41 and 42,
Parliament may exclusively make laws amending the Constitution of Canada in
relation to the executive government of Canada or the Senate and House of
Commons.” See id. at s. 44. The changes proposed are tracked historically and
available on the Canada of Parliament’s website. LEGISinfo, PARLIAMENT CAN.,
http://www.parl.gc.ca/LegisInfo/BillDetails.aspx?Language=E&Mode=1&billId=
5093616&View=3 (last visited Apr. 25, 2014).
288
. The questions referred to the Supreme Court of Canada are archived by
the Privy Council Office. See PC Number: 2013-0070, PRIVY COUNCIL OFF. (Feb.
1, 2013), http://www.pco-bcp.gc.ca/oic-ddc.asp?lang=eng&Page=secretariats
&txtOICID=2013-0070&txtFromDate=&txtToDate=&txtPrecis=
&txtDepartment=&txtAct=&txtChapterNo=&txtChapterYear=&txtBillNo=&rd
oComingIntoForce=&DoSearch=Search+%2F+List&viewattach=27293&blnDisp
layFlg=1. The Court’s advisory opinion was released on April 25, 2014. See
Reference re Senate Reform, [2014] S.C.R. 32 (Can.).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 967
regime, and also by demonstrable correlative or causational
evidence that liberal democracies are more likely to adopt a
particular framework or combination of formal amendment rules.
Among the thirty-six highest-performing constitutional
democracies, the most-to-least common amendment frameworks are
comprehensive multi-track (13), comprehensive single track (10),
restricted single track (5), restricted multi-track (4), exceptional
single track (3) and exceptional multi-track (1).
TABLE 14: MOST-TO-LEAST COMMON AMENDMENT FRAMEWORKS IN
DEMOCRATIC CONSTITUTIONS
(WITH DEMOCRACY INDEX RANKING)
Comprehensive Multi-Track
Amendment Framework
Costa Rica (22)
France (28)
Italy (32)
Netherlands (10)
South Korea (20)
Switzerland (7)
Uruguay (18)
Finland (9)
Greece (33)
Luxembourg (11)
Slovenia (28)
Sweden (2)
Taiwan (35)
Comprehensive Single-
Track
Amendment Framework
Belgium (24)
Czech Republic (17)
Germany (14)
Japan (23)
Norway (1)
Cape Verde (26)
Denmark (4)
Ireland (13)
New Zealand (5)
Portugal (26)
Restricted Single-Track
Amendment Framework
Botswana (30)
India (38)
Malta (15)
Mauritius (18)
South Africa (31)
Restricted Multi-Track
Amendment Framework
Canada (8)
Chile (36)
Estonia (34)
Spain (25)
Exceptional Single-Track
Amendment Framework
Australia (6)
Austria (12)
Iceland (3)
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
968 WAKE FOREST LAW REVIEW [Vol. 49
Exceptional Multi-Track
Amendment Framework
United States (21)
Using the Democracy Index and this distribution of amendment
frameworks, we can draw three additional conclusions. First, the
two comprehensive amendment frameworks account for twenty-
three of the thirty-six amendment frameworks in the world’s top
constitutional democracies.
289
Second, three of the top five
constitutional democracies entrench the comprehensive single-track
framework.
290
Third, the United States is alone among democracies
in entrenching the exceptional multi-track framework, thus further
confirming the exceptionalism that scholars have attributed to the
U.S. Constitution.
291
The Democracy Index arrives at its overall ranking by
measuring subsidiary indicators of democracy, including electoral
procedures and pluralism, the functioning of government, and
289
. Thirteen constitutions entrench the comprehensive multi-track
framework. See CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE COSTA RICA, arts.
195–96; SUOMEN PERUSTALSKI, 6 luku 73 (Fin.); 1958 CONST. art. 89 (Fr.); 1975
SYNTAGMA [SYN.] 2, art. 110 (Greece); Art. 138 Costituzione [Cost.] (It.);
CONSTITUTION OF LUXEMBOURG, art. 114; Grondwet voor het Koninkrijk der
Nederlanden [GW.] arts. 137–42 (Neth.); USTAVA REPUBLIKE SLOVENIJE, arts.
168–71 (Slovn.); DAEHANMINKUK HUNBEOB [HUNBEOB] art. 98 (S. Kor.);
REGERINGSFORMEN [RF] 8, arts. 14–16 (Swed.); BUNDESVERFASSUNG [BV] Apr.
18, 1999, SR 101, arts. 192–95 (Switz.); MINGUO XIANFA art. 174 (1947)
(Taiwan); CONSTITUCIÓN DE 1967, art. 331 (Uru.). Ten constitutions entrench
the comprehensive single-track framework. See 1994 CONST. art. 195 (Belg.);
CONSTITUIÇÃO DA REPÚBLICA DE CABO VERDE, arts. 309–12, 314; Ústavní zákon č.
1/1993 Sb., Ústava České Republiky [Constitution of the Czech Republic], art. 9,
para. 4; GRUNDLOVEN [GRL] Lov nr. 88 (Den.); GRUNDGESETZ FÜR DIE
BUNDESREPUBLIK DEUTSCHLAND [GRUNDGESETZ] [GG] [BASIC LAW], May 23, 1949,
BGBl. VII, art. 79 (Ger.); IR. CONST., 1937, art. 46; NIHONKOKU KENPŌ [KENPŌ],
art. 96 (Japan); Constitution Act, 1986, pt. 3 s 15 (N.Z.); GRUNNLOVEN [GRL.] §
112 (Nor.); CONSTITUIÇÃO DA REPÚBLICA PORTUGUESA, arts. 285–86.
290
. See GRUNDLOVEN [GRL] Lov nr. 88 (Den.) (fourth); Constitution Act,
1986, pt. 3 s 15 (N.Z.) (fifth); GRUNNLOVEN [GRL.] § 112 (Nor.) (first).
291
. See, e.g., BRUCE ACKERMAN, BEFORE THE NEXT ATTACK: PRESERVING
CIVIL LIBERTIES IN AN AGE OF TERRORISM 122–23 (2006); SEYMOUR MARTIN
LIPSET, AMERICAN EXCEPTIONALISM: A DOUBLE-EDGED SWORD 20–23 (1996);
Lorraine E. Weinrib, The Postwar Paradigm and American Exceptionalism, in
THE MIGRATION OF CONSTITUTIONAL IDEAS 84, 85–86 (Sujit Choudhry ed., 2006);
G. Brinton Lucas, Structural Exceptionalism and Comparative Constitutional
Law, 96 VA. L. REV. 1965, 1998 (2010); Miguel Schor, Judicial Review and
American Constitutional Exceptionalism, 46 OSGOODE HALL L.J. 535, 536–38
(2008); Carol S. Steiker, Capital Punishment and American Exceptionalism, 81
OR. L. REV. 97, 97–99 (2002).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 969
political participation.
292
Eight of the thirty-six countries earn
perfect scores for electoral procedures and pluralism; of these, six
entrench one of the two comprehensive amendment frameworks.
293
Of the bottom eight scores for electoral procedures and pluralism,
four are held by countries entrenching the restricted single-track
framework.
294
Six of the top twelve scores for functioning of
government belong to countries entrenching the comprehensive
multi-track framework.
295
Moreover, six of the nine restricted
frameworks rank in the bottom ten for political participation,
296
and
the other three rank seventh, twelfth and twentieth, respectively.
297
These are noteworthy, but more country-specific research is needed
to determine whether, and if so why, these relationships are more
than correlative.
Constitutional designers may likewise have an interest in
whether one of the frameworks is associated with constitutional
mortality or endurance.
298
The recent work of Zachary Elkins, Tom
Ginsburg, and James Melton suggests that the rate of formal
amendment, and by implication the formal amendment procedures
themselves, bears some measurable relationship to the life
292
. THE ECONOMIST INTELLIGENCE UNIT, supra note 16.
293
. See GRUNDLOVEN [GRL] Lov nr. 88 (Den.); SUOMEN PERUSTALSKI, 6 luku
73 (Fin.); CONSTITUTION OF LUXEMBOURG, art. 114; Constitution Act, 1986, pt. 3
s 15 (N.Z.); GRUNNLOVEN [GRL.] § 112 (Nor.); CONSTITUCIÓN DE 1967, art. 331
(Uru.).
294
. See CONSTITUTION OF BOTSWANA § 89; KOSTITUZZJONI TA’ MALTA, art. 66;
MAURITIUS CONST., art. 47; S. AFR. CONST., 1996, sec. 74.
295
. See SUOMEN PERUSTALSKI, 6 luku 73 (Fin.); CONSTITUTION OF
LUXEMBOURG, art. 114; Grondwet voor het Koninkrijk der Nederlanden [GW.]
arts. 137–42 (Neth.); REGERINGSFORMEN [RF] 8, arts. 14–16 (Swed.);
BUNDESVERFASSUNG [BV] Apr. 18, 1999, SR 101, arts. 192–95 (Switz.);
CONSTITUCIÓN DE 1967, art. 331 (Uru.).
296
. See CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE CHILE [C.P.] arts. 127–
29; PÕHISEADUS, ss 161–68 (Est.); INDIA CONST. art. 368; KOSTITUZZJONI TA’
MALTA, art. 66; MAURITIUS CONST., art. 47; CONSTITUCIÓN ESPAÑOLA [C.E.],
B.O.E. n. 166–68, Dec. 29, 1978 (Spain).
297
. See CONSTITUTION OF BOTSWANA § 89 (twentieth); Constitution Act,
1982, being Schedule B to the Canada Act, 1982, s. 38–48 (U.K.) (seventh); S.
AFR. CONST., 1996, sec. 74 (twelfth).
298
. Constitutional endurance is not a democratic outcome in the sense that
all long-enduring constitutions are democratic or that all democratic
constitutions are long enduring. For example, constitutional endurance may be
the result of an authoritarian regime’s resistance to constitutional change.
However, in a functioning constitution in a liberal democracy, constitutional
endurance is likely to reflect respect for civil liberties, democratic political
culture, few barriers to political participation, fair electoral procedures, and
properly functioning government. It is of course also possible for constitutional
revision—which leads to constitutional mortality for the superseded
constitution—to occur in regimes where these five democratic indicators are
high.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
970 WAKE FOREST LAW REVIEW [Vol. 49
expectancy of a constitution.
299
They conclude that constitutional
endurance is generally associated with low amendment thresholds,
though they stress that it is not always clear which type of
amendment threshold is easier than another: “For example, it is
difficult to evaluate whether a constitution that requires a two-
thirds vote of the legislature to amend the constitution is more or
less flexible than one that requires an ordinary legislative majority
with a subsequent referendum by the public.”
300
Future research
could build on their work by exploring whether states with high
constitutional mortality rates have ultimately found constitutional
stability by switching from one framework to another, or by
adopting particular specifications.
We can draw three observations from currently-in-force
constitutions. First, three of the five oldest constitutions entrench
one of the two exceptional frameworks: Australia (single-track),
Austria (single-track) and the United States (multi-track).
301
Second, in the twenty years immediately following the Second World
War, eight then-new and still-in-force constitutions entrenched a
comprehensive amendment framework: five entrenched the multi-
track framework
302
and three entrenched the single-track
framework.
303
Three also entrenched the restricted single-track
framework.
304
Third, in the aftermath of the Cold War, six out of
the nine still-in-force constitutions have entrenched one of the two
comprehensive amendment frameworks: three single-track and
305
three multi-track,
306
along with one each under the restricted single-
299
. ZACHARY ELKINS ET AL., THE ENDURANCE OF NATIONAL CONSTITUTIONS
99–100 (2009).
300
. Id. at 100.
301
. See U.S. CONST. art. V; AUSTRALIAN CONSTITUTION s 128; BUNDES-
VERFASSUNGSGESETZ [B-VG] BGBl No. 1/1920, as last amended by
Bundesverfassungsgesetz [BVG] BGBl I No. 2/1983, arts. 34-35, 44 (Austria);.
The 1868 Luxembourgian Constitution and the 1814 Norwegian Constitution
entrench the comprehensive multi-track and single-track framework,
respectively. See CONSTITUTION OF LUXEMBOURG, art. 114; GRUNNLOVEN [GRL.] §
112 (Nor.).
302
. See CONSTITUCIÓN POLÍTICA DE LA REPÚBLICA DE COSTA RICA, arts. 195–
96; 1958 CONST. art. 89 (Fr.); Costituzione [Cost.] (It.); DAEHANMINKUK
HUNBEOB [HUNBEOB] art. 98 (S. Kor.); MINGUO XIANFA art. 174 (1947) (Taiwan).
303
. See GRUNDLOVEN [GRL] Lov nr. 88 (Den.); GRUNDGESETZ FÜR DIE
BUNDESREPUBLIK DEUTSCHLAND [GRUNDGESETZ] [GG] [BASIC LAW], May 23, 1949,
BGBl. VII, art. 79 (Ger.); NIHONKOKU KENPŌ [KENPŌ], art. 96 (Japan).
304
. See INDIA CONST. art. 368; KOSTITUZZJONI TA’ MALTA, art. 66; MAURITIUS
CONST., art. 47.
305
. See 1994 CONST. art. 195 (Belg.); CONSTITUIÇÃO DA REPÚBLICA DE CABO
VERDE, arts. 309–12, 314; Ústavní zákon č. 1/1993 Sb., Ústava České Republiky
[Constitution of the Czech Republic], art. 9, para. 4.
306
. See SUOMEN PERUSTALSKI, 6 luku 73 (Fin.); USTAVA REPUBLIKE
SLOVENIJE, arts. 168–71 (Slovn.); BUNDESVERFASSUNG [BV] Apr. 18, 1999, SR
101, arts. 192–95 (Switz.).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 971
track,
307
restricted multi-track,
308
and exceptional single-track
frameworks.
309
It may also interest constitutional designers to evaluate the
relationship between constitutional endurance and the
entrenchment of temporal limitations like deliberation requirements
or safe harbors.
310
One could reasonably hypothesize that
deliberation ceilings and safe harbors both foster constitutional
endurance. Insofar as deliberation ceilings confine constitutional
amendment debates to a predetermined amount of time, the
hypothesis would be that the constitution avoids being seen as a
catalyst or obstruction to the adoption or rejection of a particular
amendment. And insofar as safe harbors prohibit constitutional
amendment or revision within a given period of time after the
coming-into-force of a new constitution or constitutional provision,
the related hypothesis would be that safe harbors give the
constitution a chance to survive through its first years in force—
perhaps the most important years for achieving popular legitimacy.
Further empirical research on these questions could help illuminate
whether deliberation ceilings and safe harbors actually do serve the
interest of constitutional endurance.
As constitutional designers structure their rules of formal
amendment, they may likewise wish to inquire whether any of the
formal amendment frameworks is associated with unusual ease or
difficulty of formal amendment. The choice of which framework to
adopt will turn on whether constitutional designers wish to privilege
flexibility or stability, or more likely some measure of both. Donald
Lutz has computed and compiled data on the difficulty of formal
amendment in thirty-two constitutional states.
311
Lutz’s valuable
data ranks the easiest and hardest formal amendment processes
among the pool of studied constitutions. His data on amendment
difficulty are not presented in terms of formal amendment
frameworks. One could therefore build on Lutz’s important study of
amendment difficulty by further investigating how the data map
onto the formal amendment frameworks and their related
amendment specifications.
307
. See S. AFR. CONST., 1996, sec. 74.
308
. See PÕHISEADUS, ss 161–68 (Est.).
309
. See STJÓRNARSKRÁ LÝÐVELDISINS ÍSLANDS, art. 62, 79 (Ice.).
310
. See supra Subpart II.C.3.
311
. LUTZ, supra note 61.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
972 WAKE FOREST LAW REVIEW [Vol. 49
TABLE 15: AMENDMENT DIFFICULTY AND AMENDMENT FRAMEWORKS
(WITH LUTZ INDEX OF DIFFICULTY)
Lutz Ranking of Amendment
Difficulty
(In Descending Order)
312
Constitutional Amendment
Framework
United States (5.10)
Exceptional Multi-Track
Australia (4.65)
Exceptional Single-Track
Costa Rica (4.10)
Comprehensive Multi-Track
Spain (3.60)
Restricted Multi-Track
Italy (3.40)
Comprehensive Multi-Track
Norway (3.35)
Comprehensive Single-Track
Japan (3.10)
Comprehensive Single-Track
Chile (3.05)
Restricted Multi-Track
Ireland (3.00)
Comprehensive Single-Track
Belgium (2.85)
Comprehensive Single-Track
Denmark (2.75)
Comprehensive Single-Track
Iceland (2.75)
Exceptional Single-Track
France (2.50)
Comprehensive Multi-Track
Finland (2.30)
Comprehensive Multi-Track
India (1.81)
Restricted Single-Track
Greece (1.80)
Comprehensive Multi-Track
Luxembourg (1.80)
Comprehensive Multi-Track
Germany (1.60)
Comprehensive Single-Track
Sweden (1.40)
Comprehensive Multi-Track
Botswana (1.30)
Restricted Single-Track
Austria (0.80)
Exceptional Single-Track
Portugal (0.80)
Comprehensive Single-Track
New Zealand (0.50)
Comprehensive Single-Track
Using Lutz’s data, we can begin to make some preliminary
observations. The most-difficult-to-amend constitutions that appear
in both Lutz’s study and mine are, in descending order of difficulty,
the constitutions of the United States, Australia, Costa Rica, Spain,
312
. This ranking is drawn from Donald Lutz’s study of amendment
difficulty. See LUTZ, supra note 61. Lutz’s data covers a limited time period; he
includes no constitution drafted after 1976 in his list of thirty-two constitutions.
Id. Some of the constitutions in my study were adopted after 1976. In this
Article, my references to the easiest or most difficult formal amendment
processes identified in Lutz’s study include only those constitutions that were
then and remain today in force. Lutz’s study of thirty-two democratic
constitutions does not include the following democratic constitutions in my list
of thirty-six: Canada, Cape Verde, Czech Republic, Estonia, Malta, Mauritius,
the Netherlands, Slovenia, South Africa, South Korea, Taiwan, and Uruguay.
My study of thirty-six democratic constitutions does not include the following
democratic constitutions in Lutz’s study of thirty-two constitutions: Argentina,
Brazil, Colombia, Kenya, Malaysia, Papua New Guinea, Venezuela, and
Western Samoa.
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 973
Italy, Norway, Japan, Chile, Ireland, Belgium, Denmark, Iceland,
France, Finland, India, Greece, Luxembourg, Germany, Sweden,
Botswana, Austria, Portugal, and New Zealand.
313
Of these twenty-
three shared data points, the two most-difficult-to-amend
constitutions entrench one of the two exceptional amendment
frameworks (United States, Australia); two of the three next-most-
difficult-to-amend constitutions entrench the comprehensive multi-
track framework (Costa Rica, Italy); and the two least-difficult-to-
amend constitutions entrench the comprehensive single-track
amendment framework (New Zealand, Portugal).
314
Among the top fifteen most-difficult-to-amend democratic
constitutions, five entrench the comprehensive single-track
framework (Norway, Japan, Ireland, Belgium, Denmark), four
entrench the comprehensive multi-track framework (Costa Rica,
Italy, France, Finland), one entrenches the restricted single-track
framework (India), two entrench the restricted multi-track
framework (Spain, Chile), two entrench the exceptional single-track
framework (Australia, Iceland), and one entrenches the exceptional
multi-track framework (United States).
315
Of the ten easiest-to-
amend democratic constitutions, three entrench the comprehensive
single-track framework (New Zealand, Portugal, Germany), four
entrench the comprehensive multi-track framework (Sweden,
Luxembourg, Greece, Finland), two entrench the restricted single-
track framework (Botswana, India), and one entrenches the
exceptional single-track framework (Austria).
316
Further country-
specific study could illuminate why exceptional amendment
frameworks are associated with a higher index of amendment
difficulty and correspondingly whether and why comprehensive
amendment frameworks are associated with a lower index.
CONCLUSION
An unamendable constitution is not an option for democratic
constitutional design. It would lack the legitimacy of popular
consent,
317
it would betray the self-assurance constitutional
designers have in themselves and the distrust they bare for
313
. Id. I have excluded from this list the Swiss and Venezuelan
Constitutions, which respectively entrenched the second- and third-most
difficult formal amendment procedures, because the Swiss and Venezuelan
Constitutions that formed the basis of Lutz’s study have since been replaced by
new constitutions.
314
. See supra Subpart II.B.
315
. See id.
316
. See id.
317
. Dellinger, supra note 6, at 386–87. I have elsewhere suggested that un-
amendability is undemocratic. See Richard Albert, Constitutional Handcuffs,
42 ARIZ. ST. L.J. 663, 667 (2010).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
974 WAKE FOREST LAW REVIEW [Vol. 49
others,
318
and its rigidity would risk provoking revolution.
319
Yet
hyper flexibility is as inadvisable as hyper rigidity because it erodes
the distinction between a constitution and a statute. Constitutional
designers must therefore design impermanent yet stable
constitutions and entrench rules for their amendment. Karl
Loewenstein has observed that designers have appropriately sought
to strike the right balance between stability and flexibility: “[T]he
process of constitutional amendment everywhere is kept sensibly
elastic, neither too rigid to invite, with changing conditions,
revolutionary rupture, nor too flexible to allow basic modifications
without the consent of qualified majorities.”
320
Although constitutional scholars have emphasized the
importance of formal amendment rules,
321
they have offered
constitutional designers little guidance on how to actually design
them. In this Article, I have endeavored to illustrate and explain
the structure of formal amendment rules in a way that is useful
both to constitutional designers tasked with creating them and to
constitutional scholars engaged in their study. I have created a new
classification of formal amendment rules in democratic
constitutional states. I have demonstrated that formal amendment
rules are structured in three tiers, and around options within each
of these tiers: one of two fundamental foundations, one of six
operational frameworks, and a combination of supplementary
specifications. I have also shown how constitutional designers may
use this new classification to manage federalism, express
constitutional values, enhance or diminish the judicial role, and
pursue democratic outcomes related to governance, constitutional
endurance, and amendment difficulty.
My corollary purpose has been to re-enliven the study of formal
amendment rules. There are many empirical, historical, and
normative questions to explore in the comparative study of formal
amendment: whether democratic outcomes are more likely to follow
from the entrenchment of any of the foundations, frameworks, or
specifications of formal amendment rules; whether any of the
foundations, frameworks, or specifications are associated with
constitutional mortality or endurance; and whether the ease or
difficulty of amendment is influenced by the choice of amendment
318
. See Sanford Levinson, The Political Implications of Amending Clauses,
13 CONST. COMMENT. 107, 112–13 (1996).
319
. A.V. DICEY, INTRODUCTION TO THE STUDY OF THE LAW OF THE
CONSTITUTION 66 (Liberty Fund 8th ed. 1982) (1915).
320
. Karl Loewenstein, Reflections on the Value of Constitutions in Our
Revolutionary Age, in CONSTITUTIONS AND CONSTITUTIONAL TRENDS SINCE
WORLD WAR II 191, 215 (Arnold J. Zurcher ed., 2d ed. 1951).
321
. I have argued elsewhere that formal amendment rules are so important
to constitutionalism that they should be subject to a higher threshold of
amendment than other constitutional provisions. See Richard Albert, Amending
Constitutional Amendment Rules, 12 INT’L J. CONST. L. (forthcoming 2015).
W03_ALBERT (DO NOT DELETE) 11/28/2014 10:54 AM
2014] CONSTITUTIONAL AMENDMENT RULES 975
foundation, framework, or specification. My larger purpose has
been to contribute to the understanding and evaluation of formal
amendment rules and to suggest that constitutional designers would
benefit from the revived study of formal amendment rules as a
helpful complement to the continuing study of informal amendment.
