The European Journal of International Law Vol. 15 no.5  EJIL 2004; all rights reserved
...........................................................................................
EJIL (2004), Vol. 15 No. 5, 907­931 doi: 10.1093/ejil/chh503
........................................................................................................................................................
The Legitimacy of International
Law: A Constitutionalist
Framework of Analysis
Mattias Kumm*
Abstract
Does international law suffer from a legitimacy crisis? International law today is no longer
adequately described or assessed as the law of a narrowly circumscribed domain of foreign
affairs. Its obligations are no longer firmly grounded in the specific consent of states and its
interpretation and enforcement is no longer primarily left to states. Contemporary international
law has expanded its scope, loosened its link to state consent and strengthened compulsory
adjudication and enforcement mechanisms. This partial emancipation from state control
means that domestic accountability mechanisms are becoming ineffective as a means to
legitimate international law. Correspondingly, the legitimacy of international law is increasingly
challenged in domestic settings in the name of democracy and constitutional self-
government. This article addresses this challenge. It develops a constitutionalist model for
assessing the legitimacy of international law that takes seriously the commitments
underlying constitutional democracy. At the heart of this model are four distinct concerns,
each captured by a distinct principle. These principles are the formal principle of international
legality, the jurisdictional principle of subsidiarity, the procedural principle of adequate
participation and accountability as well as the substantive principle of achieving outcomes
that are not violative of fundamental rights and are reasonable. Such a framework provides
a middle ground between national and international constitutionalists. Whereas the former
sometimes suggest that any law not sufficiently connected to domestic legal actors is suspect
legitimacy-wise, the latter tend to underplay what is lost democracy-wise as decision-making is
ratcheted up from the national to the international level.
Some 18 years ago, before the fall of the Berlin wall and the end of the Cold War,
Tom Franck observed that no one seemed to be asking fundamental questions
about the legitimacy of international law.1
This observation no longer holds true
today. International law's legitimacy has become a central concern. Contemporary
* Professor, New York University School of Law.
1
Franck, `Why a Quest for Legitimacy?', 21 UC Davis L Rev (1987) 535.
908 EJIL 15 (2004), 907­931
critiques of international law take many forms.2
Particularly influential in recent
years are critiques made in the name of commitments of constitutional self-government
and democracy. Armin von Bogdandy's contribution to this symposium provides
an illuminating overview of an array of conceptions of democracy and their impli-
cations for the normative assessment of international law, many of them critical of
the current state of international law.3
Furthermore, claims connected to democracy
and constitutional self-government have recently been made by national courts
and constitutional scholars to challenge certain aspects of contemporary international
legal practice.4
Leading international law scholars are beginning to suggest that
international law is suffering from a legitimacy crisis.5
Questions concerning the legitimacy of international law are not just relevant for
deciding what the future of international law should be.6
They can be reframed as
questions about the moral force of international law or, to use more traditional
jurisprudential vocabulary, the duty to obey international law. The very idea of legiti-
macy develops clearer contours when connected to questions of obedience. Only if
and to the extent that international law is legitimate is there a moral duty to obey
international law.7
Such a duty, to the extent it exists, is of central concern for
citizens governing themselves through the framework of a national constitution.
International law may generally be addressed to the states, but in constitutional
democracies the state is merely the institutional framework through which citizens
govern themselves.
This makes it possible to refocus the discussion of the legitimacy of international
law by asking the following questions: To what extent should citizens regard themselves
as morally constrained by international law in the collective exercise of constitutional
2
These critiques range from critiques of the structure of legal discourse, e.g.. M. Koskiennemi, From
Apology to Utopia (1989) to feminist critiques of international law, e.g. H. Charlesworth, C. Chinkin and
S. Wright, Feminist Approaches to International Law, available at 85 AJIL 613 (1991) to post-colonial cri-
tiques, e.g. T. Mahmud, Postcolonial Imaginaries: Alternative Development or Alternatives to Development,
available at 9 Transnat'l L. & Contemp. Probs. 25 (1999).
3
See Bogdandy, `Globalization and Europe: How to Square Democracy, Globalization, and International
Law', EJIL (in this volume).
4
In the constitutional context, arguments from democracy tend to serve one of three purposes. They serve
to justify claims that a rule of international law does not pre-empt the application of a conflicting norm of
domestic law in specific contexts. They are used to justify the claim that little or no weight should be
given to international law as persuasive authority when interpreting national constitutions. And they
can serve as an argument to limit the scope of the Treaty-making power and the participation in certain
kind of international institutions. The infamous Maastricht decision by the German Constitutional Court
is perhaps the best known judicial exemplification of this trend: see BVerfG (1993) 89, 155. In the US so-
called `revisionist' scholars, John Yoo, Jack Goldsmith and Curtis Bradley among them, have articulated
arguments of this sort. For an overview of US constitutional concerns with international law see Young,
`The Trouble with Global Constitutionalism', 38 Tex Int'l L J (2003) 527. See also Rubenfeld, `The Two
World Orders', Wilson Quarterly (Autumn 2003).
5
Weiler, `The Geology of International Law ­ Governance, Democracy and Legitimacy', [2004] ZaöRV 64.
6
In fact it may make more sense to frame discussion on what international law ought to become in terms of
an ideal of international justice, rather than legitimacy.
7
See J. Raz, The Authority of Law (1979) (linking the duty to obey the law to the law actually having the
legitimate authority it claims to have).
The Legitimacy of International Law in Question 909
government? Are they morally obligated to always comply with international law?
Should they design domestic institutions with a view to ensuring that international
law is enforced, even when domestic majorities disagree with the policies it imple-
ments? Or are citizens morally free to democratically reject and disobey international
law as a legitimate exercise of constitutional self-government, if they are strong
enough and willing to `face the music' internationally? Should the way citizens
structure domestic institutions be guided by a strong idea of constitutional self-
government and democracy, suggesting that domestically accountable institutions
need to independently assess and democratically endorse the policies embodied in
rules of international law, before international law can be enforced domestically?
What is the appropriate framework for thinking about these questions? What are the
relevant normative concerns?
The purpose of this article is to present an analytical framework for thinking about
these questions. It hopes to clarify the range and nature of the normative concerns
that any discussion of the legitimacy of international law needs to incorporate to be
plausible. The article consists of three parts. The first provides a historical context and
explains why legitimacy challenges occur now and why they take the form that they
do (Section 1). The second part suggests that the idea of democracy is unlikely to ade-
quately capture the core range of relevant legitimacy concerns. Instead the legitimacy
of international law ought to be assessed using a richer constitutionalist framework
(Section 2). The constitutionalist framework proposed consists of four principles.
These principles address formal concerns relating to the idea of international legality,
jurisdictional concerns relating to subsidiarity, procedural concerns relating to parti-
cipation and accountability, and substantive concerns relating to individual rights
and outcomes. The legitimacy of specific norms of international law can only be
assessed on the basis of an overall judgment that takes into account the whole range
of these concerns. The third part describes implications of adopting a constitutionalist
model for assessing the legitimacy of international law and describes how it is related
to other frameworks of analysis (Section 3).
1 The Legitimacy Challenge to International Law in Context
After World War II the positivist stranglehold on legal and political thought gradually
loosened. The Nuremburg and Tokyo Trials and the Universal Declaration of Human
Rights adopted by the General Assembly in December 19488
signalled a return to a
more normatively focused jurisprudence in the service of anti-Nazism.9
New constitutions
in Western Europe typically contained judicially enforceable constitutional rights.
With the maturation of national courts' rights jurisprudence in many jurisdictions,
jurisprudential theories became more strongly engaged with issues of principle, justice
and legitimacy. Furthermore, rights-based political morality was a liberal political
8
For an insightful account of the drafting of the Declaration and its political context see M.A. Glendon,
A World Made New (2001).
9
The impact on rights culture in the US is illuminatingly described by R. Primus, The American Language of
Rights (1999).
910 EJIL 15 (2004), 907­931
morality. It morally distinguished the West from Communism and provided normative
significance to the Cold War that went beyond the idea of a competitive struggle for
military, technological and economic superiority. By the 1980s the focus of norma-
tive theorizing had expanded from jurisprudential accounts of rights and questions of
legal reasoning to constitutional theory more generally. Not only were legal rights
embedded in a rights-based conception of liberal justice, but the national constitution,
as a whole, was reconceived as embodying a commitment to a coherent ideal of
constitutional justice, the specifics of which remain very much in dispute. But there is
a consensus today that legitimacy of domestic law is predicated on it being justifiable
in terms of a commitment to liberal constitutional democracy, properly understood.
Liberal democratic constitutionalism, and the rich set of theories that have been
developed since World War II to legitimate it,10
coexisted with international law during
the Cold War, but did not engage it. As Tom Franck observed in 1987, `philosophy of
international normativity . . . [tended] . . . to appear, at most, as the scraggly tail on
elegant, fully developed theories intended to explain national legal phenomena'.11
Such a lack of interest is puzzling. It cannot be explained by either the fact that inter-
national law generally addresses states rather than individuals or that international
law leaves it to states to determine how domestically international law should be
implemented (1). What then accounted for this lack of interest (2)? And why is the
legitimacy of international law emerging as a question of central concern today (3)?
1. From the perspective of citizens in liberal constitutional democracies trying to
assess what makes political and legal institutions that affect them legitimate, the lack
of interest in international law is more puzzling than it may appear at first. The
answer to the puzzle is certainly not that international law is and, to a significant
extent, continues to be mostly addressed to states rather than individuals. The state,
from the perspective of citizens in a constitutional democracy, is just the institutional
framework within which citizens govern themselves. Anything that imposes con-
straints on states also imposes constraints on citizens and how they govern them-
selves. Furthermore, the reasons do not lie in the soft nature of the demands that
international law makes on states. International law does not consist of a set of
suggestions that states are encouraged to take to heart, depending on domestic political
and legal constellations. Instead, international law makes a claim to authority. It is a
trite proposition of international law that domestic law, even domestic constitutional
law, does not serve as a justification for non-compliance with international legal
obligations.12
States do enjoy considerable discretion to determine how to internally
organize their institutions so as to ensure compliance. International law does not generally
10
Perhaps the most influential of such accounts is J. Rawls, A Theory of Justice (1971). In Europe the writings
of Habermas have similar influence, see in particular J. Habermas, Between Facts and Norms (German
1992, English 1996).
11
Franck, supra note 1, at 537.
12
See Art. 27 of the Vienna Convention of the Law of Treaties. For earlier jurisprudence with regard to
international law more generally see also Polish Nationals in Danzig (1931), PCIJ Ser. A/B, no.44, at 24:
`It should . . . be observed that . . . a State cannot adduce as against another State its own Constitution
with a view to evading obligations incumbent upon it under international law . . . '.
The Legitimacy of International Law in Question 911
require, for example, that domestic courts must have jurisdiction to enforce international
law. But if a provision of international law prescribes a certain behaviour, no feature of
the domestic legal system can be invoked as a justification for non-compliance. Demo-
cratic discontent with specific rules of international law or more pressing domestic
priorities are not valid justifications for non-compliance under international law.
2. The reasons for the lack of interest in international law's legitimacy during the
Cold War lie elsewhere. Simplifying somewhat, international law in the period
between World War II and the end of the Cold War appeared to citizens of Western
constitutional democracies and domestic lawyers in one of three guises: it was ineffective
and unreliable, addressed to other people, or occupied a highly specialized and
narrow field that raised few legitimacy concerns.
First, international law was widely believed to be ineffective and unreliable as a
guarantor of international peace and security. Before the backdrop of the Cold War
and the nuclear arms race, it is plausible for many non-professional international
lawyers to have had sympathy with `realist' assessments of `high politics' in inter-
national affairs. It may not implausibly have seemed as if those well versed in the
study of Thucydides, Machiavelli and Metternich or the lay mathematicians with a
knack for multi-level iterated prisoner's dilemmata were more likely to have a firm
handle on the major problems of the world than `peace through law' enthusiasts.
Legal institutions, procedures and rules may have had a role to play during the Cold
War. But there is something deeply misleading in describing the Cold War order as an
order established by international law.
Second, there were areas of international law and practice that flourished during
the Cold War and had a significant impact on people's lives. But to the extent that
international law was perceived as a social force, it was a social force affecting the
lives of other people. There could be little doubt that, for better or for worse, the Inter-
national Monetary Fund (IMF) and the World Bank were central actors in developing
countries and that the United Nations played an important role in the process of
decolonization. But from the perspective of citizens in comparatively rich Western
democracies, international law had serious effects primarily for other people. From the
perspective of the Western democracies that had triumphed in or had learnt their
lessons from World War II, other people were the ones in need of decolonization and
development and needed to be taught how to respect human rights by international
law. Citizens in Western democracies, on the other hand, could rely and did rely on
the resources of domestic legal systems. Compare the legal value of the non-binding
Declaration of Human Rights with the embodiment of human rights in national con-
stitutions as the supreme law of the land. Compare the thick practice of constitutional
rights adjudication in many jurisdictions by the early 1970s with the comparatively
modest impact of the International Covenant on Civil and Political Rights.13
Compare
the political struggles, legislative activities and constitutional debates relating to the
13
In part the significance of the ICCPR was that it served as a model for a national constitutional bill of
rights: see International Law and Constitution-Making, available at, 2 Chinese J. Int'l L. (2003) 467.
912 EJIL 15 (2004), 907­931
establishment of a welfare state to the discussions of the New Economic Order. Inter-
national law had few contributions to make to the great domestic post-WWII struggles
in liberal constitutional democracies.
Third, to the extent that international law was effective and established meaningful
obligations on Western constitutional democracies, they were largely the kind of
narrow and specific Treaty-based obligations characteristic of the classical conception of
international law. Diplomatic and consular relations, international mail delivery,
aviation etc. are the subject of innumerable generally effective bilateral and multilateral
treaties. But these kinds of treaties do not raise serious legitimacy questions from the
perspective of citizens governing themselves within a liberal democratic constitu-
tional framework. For the most part, these treaties pertain to the field of foreign affairs
narrowly conceived, often involving technical issues of low political salience. Here
international law addresses questions for which the national legal process is unlikely
to have produced independent and potentially conflicting results. Furthermore, these
treaties impose relatively specific obligations on states. Given the consent requirement,
all that is necessary to ensure constitutional legitimacy is to establish a constitutional
framework that ensures that whoever is authorized constitutionally to give that
consent is subjected to adequate democratic controls domestically. And even if some
treaties were more ambitious, the flexibility ensured by the general absence of
compulsory dispute resolution and the possibility of states to interpret international
law for themselves provided further guarantees that ultimately international law
would not impede constitutional self-government.
2. In the 15 years following the end of the Cold War, developments in international
law have brought serious legitimacy issues to the fore. The moment of triumph for the
post-World War II model of liberal constitutional democracy at the end of the Cold War14
is increasingly feared to have been the prelude to its decline. This decline is linked to
the emergence of an international legal order that increasingly serves ­ if not as an
iron cage ­ certainly as a firmly structured normative web that makes an increasingly
plausible claim to authority. It tends to exert influence on national political and legal
processes and often exerts pressure on nations not in compliance with its norms. Actors
in constitutional democracies are increasingly engaging seriously with international
law's claim to authority.15
What they find once they seriously engage international
law gives rise to concern. Citizens find themselves in a double bind: the meaning of par-
ticipation in the democratic process on the domestic level is undermined as interna-
tional law increasingly limits the realm in which national self-government can take
14
See F. Fukuyama, The End of History and the Last Man (1991).
15
The debates about international law's legitimacy in constitutional democracies arises from the assump-
tion that international law's claim to authority raises legitimacy problems. Once that claim is not taken
seriously and engaged, there are no reasons to engage in a legitimacy debate. Without international
law's claim to authority in the background, international law discourse is just one more tool that in
some very limited domestic contexts advocates for certain policy outcomes latch onto. For the claim that
international law provides merely a new language to engage in policy advocacy, see Tushnet, `Trans-
national/Domestic Constitutional Law', 37 Loyola of Los Angeles L R (2003) 239. This article takes the
perspective of citizens that engage international law's claim to authority seriously.
The Legitimacy of International Law in Question 913
place. At the same time, there are no comparable democratic institutions and practices
established on the international level. What has happened?
Political choices concerning the desirability of economic globalization and the
assessment of new security threats have given shape to a new kind of international
law. This international law is, largely, built on institutions, treaties and the substantive
international law established after WWII. On the surface, the transformation of inter-
national law after World War II, with the establishment of the UN, the IMF, World
Bank and the GATT, was significantly more impressive than changes that occurred in
the 1990s, notwithstanding major institutional innovations such as the International
Criminal Court and the quasi-judicialization of global trade law within the World
Trade Organization (WTO). Yet, subtle institutional changes, both formal and informal,
have transformed the practice of international law across a wide range of issues.16
These changes have occurred along three dimensions.
First, the subject matter of international law has expanded significantly. Today
there is significant overlap between the kind of questions that traditionally have been
addressed by liberal democracies as domestic concerns and the kind of questions that
international law addresses. Globalization has not led to a world in which borders are
irrelevant. But it has led to a world in which decisions on how borders are relevant are
increasingly made outside of the national democratic process. Who and what can
cross a border and under what conditions exactly is increasingly circumscribed by
rules of international law. Furthermore, the negotiation of these rules creates pres-
sures to harmonize other regulatory choices. In order to ensure agreement that certain
consumer goods can freely cross borders, an agreement on technical specifications
concerning adequate consumer protection may be necessary. Freedom of capital
movement creates pressures to ensure that states enforce effective money-laundering
controls, to prevent it becoming a tool for organized crime, etc. . . . Trade issues
addressed in the context of the WTO are no longer conceived as involving exclusively
economic questions. There are pressures to link it to environmental concerns and
human rights. In the context of Chapter VII of the UN Charter international peace
and security encompasses concerns relating to money laundering and national criminal
law,17
as well as violations of human rights. It seems as if the dictum of an eminent
authority of European law today applies equally to international law: `There simply is
no nucleus of sovereignty that the . . . States can invoke, as such, against (it).'18
The
idea of `matters essentially within the domestic jurisdiction of any state'19
has little
practical significance for cabining in the domain of international law. International
law, then, has been the handmaiden of denationalization by having generated an
increasingly dense set of substantive rules that directly concern questions traditionally
decided by national legal processes.
16
J. Weiler has used the suggestive idea of geological layering to describe international law's complex evolution.
In the last decade or so a regulatory governance layer has been added to previously existing layers. See
Weiler, supra note 5.
17
UN/SC/Res 1373 (2001) and 1566 (2004).
18
See Lenaerts as cited in J. Weiler, A Constitution for Europe (1999), at 43.
19
See Art. 2 VII UN Charter.
914 EJIL 15 (2004), 907­931
Second, the procedure by which international law is generated increasingly attenuates
the link between state consent and the existence of an obligation under international
law. Traditionally international legal obligations arose either because of specific
treaty obligations assumed by states ratifying the treaty or as a matter of customary
international law reflecting longstanding customary practice of states. Treaties
today, though still binding only on those who ratify them, increasingly delegate pow-
ers to treaty-based bodies with a quasi-legislative or quasi-judicial character. Within
their circumscribed subject-matter jurisdiction, these bodies are authorized under the
treaty to develop and determine the specific content of the obligations that states are
under. This means that, though states have consented to the treaty as a framework
for dealing with a specified range of issues, once they have signed on, the specific
rights and obligations are determined without their consent by these treaty-based
bodies.20
The link to state consent is further attenuated by the expansive interpretation
of the jurisdiction of these bodies, perhaps most dramatically exemplified by the
quasi-legislative actions undertaken by the Security Council post-September 11.21
Furthermore, it is doubtful that much legitimating value can be placed on a state's
consent to a treaty, when the state is confronted with a take it or leave it option and
the costs of not participating are prohibitively high. Finally, the link between a particular
legal obligation of a state and its consent has not just been attenuated by the prolifer-
ations of regulatory multilateral treaties. Customary international law (CIL), too, is
no longer thought to require a long general and consistent state practice followed by
states from a sense of legal obligation. Instead, so-called `modern' CIL significantly
discounts the requirement of general and consistent state practice in favour of an
approach that focuses primarily on statements. Particularly in the area of human
rights, declarations made by representatives of states either in international fora such
as the General Assembly or in the context of multilateral treaty-making are central
to the inquiry whether a rule of CIL has developed or not.22
Here too the effect is to
further disconnect the creation of an international legal obligation from a state's
specific consent.
Third, states have less flexibility in the interpretation and enforcement of international
law. In part, this is because of the relative specificity of the obligations involved. In part it
is because of the spread of compulsory third party dispute resolution. Both are connected
20
There are a number of such treaties presently in force, and they cover a vast number of subject areas.
The following is a sampling of such treaties: in the realm of environmental law (the Vienna Convention
for the Protection of the Ozone layer, the Montreal Protocol on Substances that Deplete the Ozone layer,
the United Nations Convention on the Law of the Sea, and the Convention on Biological Diversity); in the
realm of commercial law (the World Trade Organization, the North American Free Trade Agreement,
and the United Nations Convention on the Recognition and Enforcement of Foreign Arbitral Awards); in
the realm of chemical weapons control (The Chemical Weapons Convention); in the realm of collective
security (The United Nations and The North Atlantic Treaty); and in the realm of international criminal
courts (the creation of ad hoc tribunals for the Former Yugoslavia and Rwanda as well as the Inter-
national Criminal Court). See generally, T.M. Franck (ed.), Delegating State Powers: The Effect of Treaty
Regimes on Democracy and Sovereignty (2000).
21
See Szasz, `The Security Council Starts Legislating', 96 AJIL (2002) 901.
22
For a description and evaluation of that transformation see A. E. Roberts, `Traditional and Modern
Approaches to Customary International Law: A Reconciliation', 95 AJIL (2002) 757­791.
The Legitimacy of International Law in Question 915
to the proliferation of international courts and tribunals. General principles of inter-
national law or abstract norms of CIL do not in themselves provide the kind of guidance
that courts tend to insist on or national politicians are likely to respect when they are
asked to assess whether a domestic law is compatible with international law. Besides
the steady codification of customary international law generally throughout the past
century, the panoply of courts and tribunals that have been established on the inter-
national level in the last decade23
have played a central role in specifying the obligations
that states are under in the context of various treaty regimes. Because of the greater
specificity of international legal obligations, discrepancies between international law
and domestic law tend to become increasingly visible and more difficult to gloss over.
In conjunction with more effective monitoring and higher reputational and other
costs to non-compliance, international law has in many areas and in many jurisdictions
developed into a serious constraint on national political and legal processes.
International law that shares these three features will be referred to as international
law as governance. International law as governance blurs the distinction between
national and international law. Both with regard to the scope of its subject matter
and the processes used to generate, interpret and apply it, it is no longer apparent
what structurally distinguishes international law from national law, except, of
course, for one central point: international law is not generated within the institu-
tional framework of liberal constitutional democracy and does not allow for a central
role for electoral supervision. In this sense it lacks democratic pedigree.
Of course representatives of states, authorized by national constitutions are still
the major actors on the international level, even though non-state actors such as
international courts and tribunals, transnational bureaucracies, NGOs, MNCs and
even individuals have emerged as significant participants in the international legal
process. Networks of national governments, national administrators and national
courts remain at the very heart of global governance.24
As Anne Marie Slaughter
aptly puts it: the state has not disappeared, it has become disaggregated.25
But that does little to provide the governance process with democratic legitimacy.
State actors as an integral part of a global network, interacting with transnational
bureaucracies and interest groups, cannot meaningfully be connected to an ideal of
national self-government, either with regard to procedures followed or outcomes gen-
erated. Furthermore, state actors that are deeply embedded in transnational networks
are notoriously difficult to hold democratically accountable on a national level. The
great institutional loser in the shift from classical international law to international
23
See Boon, `Instances of Criminal Courts', in T. M. Franck (ed.), Delegating State Powers: The Effect of Treaty
Regimes on Democracy and Sovereignty (2000), 171­208 (describing the emergence of the ad hoc tribunals
for the Former Yugoslavia (ICTY) and Rwanda (ICTR) and the International Criminal Court, and analysing
the constitutional questions involved). For an excellent resource for information on these and other
international courts and tribunals, see the website of the Project of International Courts and Tribunals
(PICT), available at http://www.pict-pcti.org/ (visited on 28 April 2003).
24
A.-M. Slaughter, A New World Order (2003).
25
Ibid.
916 EJIL 15 (2004), 907­931
law as governance is the national legislature and the national electoral process, the
very institutions believed to be at the heart of liberal constitutional democracy.
It seems that the introduction of the argument of democratic legitimacy into the
world of constitutional discourse to address just these kinds of concerns occurred as
recently as November 1993 when the German Constitutional Court pronounced itself
on the constitutionality of the Treaty of Maastricht.26
The complainant, ironically
a former high-ranking official in the European Commission, had invoked a violation
of his constitutional right to vote on the national level as a cause of action. He claimed
that the right to participate in the democratic process on the national level was in
substance eroded as further regulatory authority was delegated from the national to
the European level by the Treaty of Maastricht. Decisions made under those delegated
powers on the level of the European Union take precedence over national law and
pre-empt any parliamentary actions to the contrary. The Court, to the surprise of
many, held the action to be admissible but then went on to issue the Maastricht Treaty a
constitutional bill of health. But as it did so, the Court drew lines in the sand with regard
to limits and the direction of future developments of European integration. The Court
made its understanding of democracy the central pillar of its argument and a guiding
concern for fashioning doctrines relating to the enforcement of supranational law.
It is not surprising that arguments from democracy were first marshalled by the
German Constitutional Court adjudicating issues of European integration at the time
that it did. On the one hand, European law had long established itself as hard law. As
a general matter, litigants in national courts could effectively enforce EC Law, even in
the face of conflicting national law.27
Furthermore, regulatory authority exercised by
EU institutions on the European level was not just expansive with regard to its juris-
diction. Since 1987 it had also, to a large extent, become non-consensual. Since the
Single European Act, representatives of the most powerful nations in Europe could be
outvoted in the European jurisgenerative process. Yet, even though the Treaty of
Maastricht had introduced the idea of European citizenship, electoral politics has only
an insignificant role to play on the European level (notwithstanding the existence of a
European Parliament). There is no government promoting a legislative agenda that
can be voted out of office.
Even though there are many features of European integration that suggest that
global processes are unlikely to simply replicate European processes with a time lag,
there are striking structural similarities between contemporary international law and
European law that go right to the legitimacy issue. Both international law and Euro-
pean law are no longer limited jurisdictionally to a relatively clearly circumscribed
domain. In both European and international law the link between state consent and
the emergence of legal obligations is often attenuated. Certainly, the proliferation of
non-consensual international obligations created by diffuse sets of actors has spread
significantly. And both in international and European law the cost of non-compliance
26
BVerfGE (1993) 89, 155.
27
For a general overview of the reception of EU law in the domestic systems of European nations see,
generally, A.M. Slaughter, A. Sweet and J. H. H. Weiler, The European Courts and National Courts (1998).
The Legitimacy of International Law in Question 917
for national actors is increasingly high, as monitoring and enforcement mechanisms
are strengthened. On the one hand globalization is feared or hoped by many to proceed
through the establishment of forms of governance loosely comparable to those estab-
lished in the European Union.28
On the other hand, constitutional arguments invoking
democracy are beginning to be articulated in Western liberal democracies to challenge
the legitimacy of international law, in much the same way as discussion of the democratic
deficit has dominated European debates in the last decade.
2 The Constitutionalist Model: A Framework
of Four Principles
How then should citizens in liberal constitutional democracies engage international
law? To what extent should they see themselves constrained by it and design
domestic institutions so that compliance with international law is assured? To what
extent should they see themselves free to disregard it and design institutions to ensure
the desired flexibility? If a duty to obey international law is a function of its legitimacy,
how should the legitimacy of international law be assessed?
In the following I will present a constitutionalist model for thinking about the
legitimacy of international law. At the heart of the model are four distinct concerns,
each captured by a distinct principle. These principles are the formal principle of
international legality, the jurisdictional principles of subsidiarity, the procedural
principle of adequate participation and accountability, as well as the substantive
principle of achieving outcomes that are not violative of fundamental rights and are
reasonable.
The principle of international legality establishes a presumption in favour of the
authority of international law. The fact that there is a rule of international law gov-
erning a specific matter means that citizens have a reason of some weight to do as
that rule prescribes. But this presumption is rebutted with regard to norms of inter-
national law that constitute sufficiently serious violations of countervailing normative
principles relating to jurisdiction, procedure or outcomes. To put it another way: citi-
zens should regard themselves as constrained by international law and set up
domestic political and legal institutions so as to ensure compliance with international
law, to the extent that international law does not violate jurisdictional, procedural
and outcome-related principles to such an extent that the presumption in favour of
international law's authority is rebutted. When assessing concerns relating to jurisdic-
tion, procedure and outcome each of the relevant principles can either support or
undermine the legitimacy of international law.
When citizens in constitutional democracies accept the constraints imposed by an
international law that is legitimate as assessed under this approach, they are not
compromising national constitutional commitments. Instead, such respect for inter-
national law gives expression to and furthers the values that underlie the commitments
to liberal constitutional democracy, properly understood.
28
See generally Slaughter, `The New Real World Order', 76 Foreign Affairs (1997) 183.
918 EJIL 15 (2004), 907­931
Given their pivotal role, the content of these principles deserves some further clari-
fication. Such clarification would ideally occur both in the form of a rich set of examples
that illustrates the practical usefulness of the framework in concrete contexts and a
more fully developed theoretical account of each of these principles. But here a brief
further description of each of these principles will have to suffice.
A Formal Legitimacy: The Principle of International Legality
The first principle is formal and establishes a prima facie case for the legitimacy of
international law. The principle of international legality generally requires that
addressees of international law should obey it.29
International law is prima facie
legitimate and deserves the respect of citizens in liberal constitutional democracies
simply by virtue of it being the law of the international community. International law
serves to establish a fair framework of cooperation between actors of international
law30
in an environment where there is deep disagreement about how this should
best be achieved. In order for international law to achieve its purpose, those who are
addressed by its norms are morally required to generally comply, even when they
disagree with the content of a specific rule of international law.31
There is a prima facie
duty of civility to comply with even those norms of international law that the majority
of national citizens believe to be deficient.32
Otherwise international law has no
chance of achieving its purpose.
The benefits of a sufficiently well-established international legal system for the
international community are well rehearsed and include the following.33
First, an international community that makes do without the resource of a well-
developed legal system in which the authority of law is generally recognized is impov-
erished. An effective international legal system is an asset to the international community
as a whole. Law is an effective instrument that enables and fosters the establishment of
welfare-enhancing cooperative endeavours between various actors. Law can help
reduce transaction costs for setting up trans-border cooperative schemes. It is a tool
that helps build trust between international actors and thus facilitates engagement in
mutually beneficial cooperative endeavours, thereby enhancing global welfare. Law
then can be a tool that helps foster the development of transnational communities,
internalize externalities, prevent prisoner-dilemma-based misallocation of resources,
realize efficiency gains, etc....
29
For a discussion of the idea of the International Rule of Law as an argument for national courts to
enforce international law over national law see Kumm, `International Law Before National Courts: The
International Rule of Law and the Limits of the Internationalist Model', 44 Virginia Jl of Int'l L (2003)
19­32.
30
For an extensive discussion see T. Franck, Fairness in International law and Institutions (1995).
31
The idea of a duty to support the International Rule of Law is in some sense analogous to what Rawls
has called the natural duty to support a just constitution. See J. Rawls, A Theory of Justice (1971), at
333­342.
32
According to Rawls there is a `natural duty of civility not to invoke the faults of social arrangements as a
too ready excuse for not complying with them': Rawls, supra note 31, at 355.
33
The following draws on Kumm, supra note 29.
The Legitimacy of International Law in Question 919
Second, the international rule of law also contributes to the protection of domestic
groups within a particular state who are protected by international law. International
law contributes to the checks and balances of a constitutional system, complementing
domestic separation of powers and federalism as another means of achieving this.
An effective institutionalization, international legality also has the tendency to limit
the options of the executive branch to claim foreign affairs prerogatives and thereby
shift power to the executive branch in a way that endangers and potentially destabi-
lizes democracy on the national level. In these ways, the international rule of law
has the tendency to lock in and stabilize liberal constitutional democracy on the
domestic level.
Third, an effective international system also provides predictability and enhances the
freedom of individual actors. The rule of law helps secure fixed points of reference by
stabilizing social relationships and providing a degree of predictability for them. In
this way the international rule of law protects and enhances the freedom of various
actors by creating a predictable environment in which they can make meaningful
choices, just as the rule of law does on the national level.34
Finally, an effective international legal system has the potential to curtail the illegal
abuse of political power. It prevents powerful actors who would otherwise have the
capacity to exercise political power to the detriment of those protected by law from
doing so. More specifically, there are three kinds of problems that an effective inter-
national legal system can help mitigate. The first is related to asymmetries in power and
the potential for political abuse that such an asymmetry entails. This abuse can take the
form of unjustified coercive intervention or other forms of unilateral impositions.
Under the principle of international legality, less powerful states tend to be more
effectively protected against impositions by powerful states. Just as the rule of law
became the battle cry for political reformers in much of Europe in the 18th and early
19th centuries to curb what was experienced as the arbitrary exercise of authority by
powerful actors on the domestic level, so the international rule of law has been
embraced by many in the 20th century as a means to reign in the arbitrary exercise
34
This may be of lesser significance for powerful governments, which have the resources and bureaucracies
to process information and negotiate commitments concerning the behaviour of other actors in other
ways. But non-governmental actors, too, increasingly have a planning horizon expanding beyond their
jurisdiction and something real to gain. Their radius of action is not limited to the jurisdiction of the
state they live in. That is most obviously true for multinational corporations (MNCs). It is not surprising
that MNCs have been among those non-governmental actors pushing the Rule of Law both on the
international and on the national level aggressively in the areas that concern the scope of their
decision-making. Corporations want to be able to make strategic decisions knowing that market access
according to international rules will be guaranteed. They want budgets and projects relating to
research and development to be determined knowing that intellectual property will be protected
according to international rules. Investors can make informed investment decisions without having to
factor in the risks and insurance costs relating to confiscation of property in violation of international
standards. But ordinary citizens, too, have something to gain. Citizens have an interest in being able to
rely on being treated in accordance with international legal standards when in a foreign country, and
in having access to diplomatic and consular personnel, should they require it. Citizens can also make
their decisions knowing that they can rely on certain basic guarantees relating to fundamental rights
to a greater extent, wherever they are.
920 EJIL 15 (2004), 907­931
of power by militarily and economically powerful actors on the international level. Of
course, to some extent any asymmetries of power are likely to be reflected in the rules
of the international legal system. But even to the extent that the rules reflect these
power asymmetries, the existence of such rules, if obeyed, provide some degree of
protection and security for the weak, just as they did on the domestic level. Not only
does an effective international legal system instil a habit of obedience, thereby civiliz-
ing the exercise of power, the requirement of consistency would also provide greater
predictability and would thereby provide a more stable international environment.
Furthermore, an environment in which the international rule of law is realized is one
that provides a more solid basis for further principled deliberation, with a view to
reform and improve on those rules. One dimension of having an effective system of
international law, then, is that it furthers the right of a people to govern itself without
inappropriate impositions by other states.
A commitment to the principle of international legality says nothing about the
proper scope of international law. It certainly provides no grounds for some inter-
national lawyer's enthusiasm for expanding the reach of international law to as
many domains as possible. Nor does it make a fetish of legality by suggesting that
legal forms of dispute resolution are superior to other forms. But it does suggest that
once a norm of international law has come into existence, its very existence provides
a reason to comply with it. It establishes a presumption in favour of compliance with
international law.
In the European world at the beginning of the century Max Weber could claim that
formal legality could replace charisma or tradition as the source of legitimacy.35
After
WWII, such a thin notion of legitimacy has been gradually replaced by the considerably
richer idea of constitutional legitimacy. To be fully legitimate more is required of a rule
than just its legal pedigree. Formal legality matters, but it is not the only thing that
matters. More specifically, there is a range of other concerns that provide countervailing
considerations and suggest that under certain circumstances the presumption in favour
of the legitimacy of international law can be rebutted. These concerns are related to a
more substantive commitment to liberal-democratic governance. Concerns about demo-
cratic legitimacy should best be understood as concerns about three analytically distinct
features of international law. These concerns are related to jurisdiction, procedure and
outcomes, respectively. The presumption in favour of compliance with international law
can be overridden, by reason of sufficient weight relating to jurisdiction, procedure or
outcome. Once there are such reasons, citizens in a constitutional democracy ought to
think of themselves as free to deviate from the requirements of international law. In these
cases, citizens have good reasons to conceive of themselves as free to generate and apply
the independent outcomes of the domestic legal and political process.
B Jurisdictional Legitimacy: The Principle of Subsidiarity
The first of those three concerns is captured by the principle of jurisdictional legitimacy or
subsidiarity. Subsidiarity is in the process of replacing the unhelpful concept of
35
M. Weber, Politics as Vocation (1919).
The Legitimacy of International Law in Question 921
`sovereignty' as the core idea that serves to demarcate the respective spheres of the
national and international.36
The principle of subsidiarity is a central principle of
European constitutionalism. It ought to be an integral feature of international law as
well. It was used to guide the drafting of the European Constitutional Treaty signed in
October 2004. It is a principle that guides the exercise of the European Union's power
under the Treaty. And it guides the interpretation of the European Union's laws. As such, it
is a structural principle that applies to all levels of institutional analysis, ranging from the
big picture assessment of institutional structure and grant of jurisdiction to the micro-
analysis of specific decision-making processes and the substance of specific decisions.
The specific wording of the provisions in the European Community Treaty37
and
the Protocol governing the application of subsidiarity38
has done little to clarify the
exact content of the principle. But independent of its exact formulation, at its core the
principle of subsidiarity requires any infringements of the autonomy of the local level
by means of pre-emptive norms enacted on the higher level to be justified by good
reasons. It follows that any norm of international law requires justification of a special
kind. It is not enough for it to be justified on substantive grounds, say, by plausibly
claiming that it furthers the general welfare. Instead, the justification has to make
clear what exactly would be lost if the assessment of the relevant policy concerns was
left to the lower level. With exceptions relating to the protection of minimal standards
of human rights, only reasons connected to collective action problems ­ relating to exter-
nalities or strategic standard-setting giving rise to `race to the bottom' concerns, for
example ­ are good reasons to ratchet up the level on which decisions are made. And
even when there are such reasons, they have to be of sufficient weight to override any
disadvantages connected to the pre-emption of more decentralized rule-making. On applica-
tion subsidiarity analysis thus requires a two-step test. First, reasons relating to the
existence of a collective action problem have to be identified. Second, the weight of
these reasons has to be assessed in light of countervailing concerns in the specific
circumstances. This requires the application of a `proportionality test' or `cost-benefit
analysis' that is focused on the advantages and disadvantages for ratcheting up the
level of decision-making. This means that on application this principle, much like the
others, requires saturation by arguments that are context sensitive and most likely
subject to normative and empirical challenges. Its usefulness does not lie in providing
a definitive answer in any specific context. But it structures inquiries in a way that is
likely to be sensitive to the relevant empirical and normative concerns.
There are good reasons for the principle of subsidiarity to govern the allocation and
exercise of decision-making authority wherever there are different levels of public
authorities. The reasons are related to sensibility towards locally variant preferences,
36
J. Jackson labelled a similar idea `sovereignty-modern'. See Jackson, `Sovereignty-Modern: A New
Approach to an Outdated Concept', 97 Am J Int'l L (2003) 782.
37
Art. 5 ECT in so far as it is relevant states: `In areas which do not fall within its exclusive competence, the
Community shall take action in accordance with the principle of subsidiarity, only if and in so far as the
objectives of the proposed action cannot be sufficiently achieved by Member States and can therefore, by
reason of the scale or effects of the proposed action, be better achieved by the Community.'
38
See ECT Protocol No. 30.
922 EJIL 15 (2004), 907­931
possibilities for meaningful participation and accountability and the protection and
enhancement of local identities that suggest that the principle of subsidiarity ought to
be a general principle guiding institutional design in federally structured entities. But
the principle has particular weight with regard to the management of the national/
international divide. In well-established constitutional democracies instruments for
holding accountable national actors are generally highly developed. There is a well-
developed public sphere allowing for meaningful collective deliberations, grounded in
comparatively strong national identities. All of that is absent on the international
level. To a significant extent it is also absent on the European level. Not surprisingly,
Member States, with the support of the large majority of European citizens, have
spent a great deal of effort and resources on drawing jurisdictional boundaries and
designing procedures that effectively police them. If international law as governance
is no longer grounded in the specific consent of states, jurisdictional concerns
addressed by the framework of analysis provided by the principle of subsidiarity have
a central role to assess, guide and constrain transnational legal practice.
The principle of subsidiarity is not a one-way street, however. Subsidiarity-related
concerns may in certain contexts strengthen rather than weaken the comparative
legitimacy of international law over national law. If there are good reasons for
deciding an issue on the international level, because the concerns addressed are con-
cerns best addressed by a larger community, then the international level enjoys
greater jurisdictional legitimacy. And even though the principle generally requires
contextually rich analysis, there are simple cases. The principle can highlight obvious
structural deficiencies of national legislative processes with regard to some areas of
regulation.
Imagine that in the year 2010 a UN Security Council Resolution enacted under
Chapter VII of the UN Charter imposes ceilings and established targets for the reduction
of carbon dioxide emissions aimed at reducing global warming. Assume that the case
for the existence of global warming and the link between global warming and carbon
dioxide emissions has been conclusively established. Assume further that the necessary
qualified majority in the Security Council was convinced that global warming pre-
sented a serious threat to international peace and security and was not appropriately
addressed by the outdated Kyoto Protocol or alternative treaties that were open to
signature, without getting the necessary number of ratifications to enter into force.
Finally, assume that a robust consensus had developed that Permanent Members of the
newly enlarged and more representative UN Security Council39
were estopped from
vetoing a UN resolution, if four-fifths of the Members approved a measure.
Now imagine that a powerful constitutional democracy, such as the United States,
has domestic legislation in force that does not comply with the standards established
by the resolution. The domestic legislation establishes national emission limits and
structures the market for emission trading, but goes about setting far less ambitious
39
Assume that current proposals had become law and that it included as new Permanent Members an
African state (Nigeria or South Africa), two additional Asian states (Japan and India or Indonesia), a
South American state (Brazil) and an additional European state (Germany), as well as 5 new non-
permanent members.
The Legitimacy of International Law in Question 923
targets and allowing for more emissions than the international rules promulgated by
the Security Council allow. Domestic political actors invoke justifications linked to
life-style issues and business interests.40
National cost-benefit analysis, they argue,
has suggested that beyond the existing limits it is better for the nation to adapt to
climate change rather than incurring further costs preventing it. After due delibera-
tions on the national level, a close but stable majority decides to disregard the inter-
nationally binding Security Council resolutions and invokes the greater legitimacy of
the national political process. Yet, assume that the same kind of cost-benefit analysis
undertaken on the global scale has yielded a clear preference for aggressively taking
measures to slow down and prevent global warming along the lines suggested by the
Security Council resolution.
In such a case, the structural deficit of the national process is obvious. National
processes, if well designed, tend to appropriately reflect values and interests of
national constituents. As a general matter, they do not reflect values and interests of
outsiders. Since in the case of carbon dioxide emissions there are externalities related
to global warming, national legislative processes are hopelessly inadequate to deal
with the problem. To illustrate the point: the US produces approximately 25 per cent
of the world's carbon dioxide emissions potentially harmfully affecting the well-
being of peoples worldwide.41
Congress and the EPA currently make decisions with
regard to the adequate levels of emissions. Such a process clearly falls short of even
basic procedural fairness, given that only a small minority of global stakeholders is
adequately represented in such a process. It may well turn out to be the case that
cost-benefit analysis conducted with the national community as the point of reference
suggests that it would be preferable to adapt to the consequences of global warming
rather than incurring the costs trying to prevent or reduce it. In other jurisdictions,
the analysis could be very different. In Bangladesh, it could be necessary to relocate a
significant part of the population if global temperature was to rise as expected
without significant cuts in emissions. More importantly, cost-benefit analysis
conducted with the global community as the point of reference could well yield
results that would suggest aggressive reductions as an appropriate political
response. The jurisdictional point here is that the relevant community that serves as
the appropriate point of reference for evaluating processes or outcomes is clearly the global
community.42
When there are externalities of this kind, the legitimacy problem
would not lie in the Security Council issuing regulations. Legitimacy concerns in
these kinds of cases are more appropriately focused on the absence of effective
40
For an argument of this kind in respect of the US position on the Kyoto Protocol, see Yandel and Buck,
`Bootleggers, Baptists and the Global Warming Battle', 26 Harv Envt'l L Rev (2002) 177 (contending
that `the Kyoto Protocol would have been a huge drag on the US economy' while producing minimal
environmental benefits).
41
C. Flavin, Slowing Global Warming: A Worldwide Strategy (1989), at 8 (citing statistics which show that
the US is the largest producer of carbon dioxide in the world).
42
For an elaborate argument ­ grounded in both domestic and international environmental law ­ in sup-
port of this reasoning, see Guruswamy, `Global Warming: Integrating United States and International
Law', 32 Ariz L Rev (1990) 221.
924 EJIL 15 (2004), 907­931
transnational decision-making procedures and the structurally deficient default
alternative of domestic decision-making.43
But now imagine the UN Security Council enacting a binding regulation concerning
ambient standards regarding air pollution. These too are stricter than national
standards. This too is a simple case. Since there are no externality problems or, let us
assume, other good reasons44
for such an issue to be addressed beyond the national
level, the situation is clearly different. There are no reasons why the national com-
munity should not qualify as the relevant community for assessing the legitimacy of
procedures. Measures of such a kind enacted transnationally would be incompatible
with the principle of subsidiarity and suffer from lack of jurisdictional legitimacy.
C Procedural Legitimacy: The Principle of Adequate
Participation and Accountability
One reason why national law is thought to enjoy comparatively greater legitimacy
than anything decided on the international level is the idea that the core depository of
legitimacy is electorally accountable institutions. On the national level, legislative
bodies constituted by directly elected representatives make core decisions. There are
no such institutions on the international level. CIL is generated by an ensemble of
actors ranging from democratically legitimate and illegitimate governments, unelected
officials of international institutions, judges and arbitrators, scholars and NGOs. Trea-
ties, on the other hand, are legitimate to the extent and exactly because they tend to
require national legislative endorsement in some form or another. Problems arise when
treaties create institutions in which unelected officials in conjunction with other actors
may create new obligations, which, at the time the treaty was signed, were impossible
to foresee. National law is superior because it tends to be parliamentary law, that is law
authorized by a directly representative institution.
Many things would need to be said to address this claim. I will confine myself to two
core points. First, even on the national level, parliament as the traditional legislative
forum has lost significant ground in the 20th century in constitutional democracies.
Parliament is no longer considered as the exclusive institutional home of legitimate
decision-making on the domestic level. On the one hand, this is linked to the emer-
gence of the administrative state. For what generally are believed to be good reasons,
the turn to the administrative state in the first half of the 20th century has involved
significant delegation of regulatory authority to administrative institutions of various
kinds. Whether in the area of monetary policy, anti-trust policy or environmental
policy, many of the core decisions are no longer made by parliament. This is generally
justified on diverse grounds ranging from the expertise of decision-makers, the greater
possibilities of participation for the various stakeholders involved, and the like.44
The
43
For a general discussion of global climate policy and regulatory approaches to be taken to it see Stewart
and Wiener, `Comprehensive Approach to Global Climate Policy: Issues of Design and Practicality',
9 Ariz J Int'l & Comp L (1992), 17.
44
See, e.g., Stewart, `The Reformation of American Administrative Law', 88 Harv L Rev (1975) 1667 at
1760­1790 (describing how this promotes interest group competition and representation in the admin-
istrative process itself).
The Legitimacy of International Law in Question 925
argument that this is of little significance because legislatures retain the possibility to
legislate whenever there is the requisite majority to do so is not irrelevant. But as a
matter of institutional practice and of political realism, the effective control over
administrative decision-making that exists by virtue of such a possibility is modest.45
On the other hand, liberal constitutional democracies developed in the second half of
the 20th century to include constitutional courts with the authority to strike down
laws generated by the legislative process on grounds of constitutional principle. And
constitutional courts have engaged in such a practice more or less aggressively in
many jurisdictions. In many jurisdictions, they enjoy more public support than any
other political institution as a result.46
The reasons generally invoked to justify judicial
review of legislative decisions are well rehearsed. They include the comparative
advantage to secure the rights of individuals against inappropriate majoritarian
intervention, concerns that are particularly pertinent with regard to groups disad-
vantaged in the political process as well as other instances in which political failures
of various kinds suggest a comparative advantage for judicial review of other actor's
decisions. It is important to take note of a bad argument for judicial review. Judicial
review is not generally justified because the necessary supermajority for constitutional
entrenchment has determined that a certain specifically circumscribed right ought to
be protected. To the extent that this argument casts constitutional courts as the
mouthpiece and mechanical instrument of legislative self-restraint as defined by the
constitutional legislature, it is misleading at best. In most jurisdictions, a core task of
constitutional courts is to interpret highly abstract constitutional clauses invoking
equality, liberty, freedom of speech, property or due process. Courts in many jurisdic-
tions engage in elaborate arguments of principle about why this or that policy concern
ought to take precedence over competing concerns in a particular context. To that
extent constitutional courts can only be understood as political actors in their own
right. If it is desirable for there to be such an actor, it can only be because of widely
held beliefs about the comparative advantage of the judicial process over the ordinary
political process across the domain that falls within the constitutional jurisdiction of
the court.47
It turns out that any robust version of majoritarian parliamentarianism cannot be
understood as the ideal underlying contemporary political practice in liberal constitu-
tional democracies. Instead, there is a predominance of a more pragmatic approach.
45
Lowi, `Two Roads to Serfdom: Liberalism, Conservatism and Administrative Power', 36 Am U L Rev
(1987) 295 at 321­322 (criticizing broad, unaccountable discretionary power held by modern adminis-
trative agencies).
46
See generally Tate and Vallinder, `The Global Expansion of Judicial Power: The Judicialization of Politics',
in C.N. Tate and T. Vallinder (eds.), The Global Expansion of Judicial Power (1995), 1, at 5 (describing the
worldwide expansion of judicial power in several jurisdictions and dubbing it `one of the most significant
developments in late twentieth century and early twenty-first century government').
47
Hirschl, `Resituating the Judicialization of Politics: Bush v. Gore as a Global Trend', 25 Canadian J of L & Juris
(2002) 191 (arguing that the availability of a constitutional framework that encourages deference to the
judiciary and the existence of a political environment conducive to judicial empowerment have helped
bring about a growing reliance on adjudicative means for articulating, framing, and settling fundamental
moral controversies and highly contentious political questions).
926 EJIL 15 (2004), 907­931
That approach does take seriously concerns relating to checks and balances,
accountability, participation, responsiveness, transparency, and so on.48
But over the
whole spectrum of political decision-making, constitutional democracies allocate
decision-making authority to a wider range of decision-makers than a robust parlia-
mentarianism is willing to acknowledge. This draws attention to two points of
significance for assessing the comparative legitimacy of international and national
law. First, much of international law that is in potential conflict with outcomes of the
national political process competes with national rules determined either by adminis-
trative agencies or constitutional courts, suggesting that the argument from demo-
cracy has less bite at least in such cases. And even if international law does compete
with the outcomes of the national parliamentary process, the domestic example
suggests that under some circumstances the outcomes of a non-parliamentary proce-
dure may be preferable over the outcome of a parliamentary procedure. Given that
the prerequisites for meaningful electorally accountable institutions on the international
level are missing, the absence of electorally accountable institutions on the international
level is insufficient to ground claims that the international legal process is deficient
procedurally.
On the other hand, the absence of directly representative institutions on the tran-
snational level and the difficulty of establishing a meaningful electoral process on
the global level49
is one of the reasons why the principle of subsidiarity has greater
weight when assessing institutional decision-making beyond the state than within
a national community. It is not surprising that in well-established federal systems
concerns about jurisdictional issues are typically less pronounced. A well developed
national political process involving strong electorally accountable institutions, a
cohesive national identity and a working public sphere on the national level lower
the costs of ratcheting up decision-making. In the European Union, on the other
hand, European elections do not mean much as the Commission in conjunction
with the Council ­ consisting of Members of the executive branch of Member State
governments ­ remain largely in control of the legislative agenda. Limiting the
scope of what the European Union can do is regarded as a core concern. It ought to
be at least as much of a concern when it comes to international law.
But even when international law plausibly meets jurisdictional tests, it could still
be challenged in terms of procedural legitimacy. The principle of procedural legitimacy
focuses on the procedural quality of the jurisgenerative process. Electoral accountability
may not be the right test to apply, but that does not mean that there are no standards
of procedural adequacy. Instead, the relevant question is whether procedures are
sufficiently transparent and participatory and whether accountability mechanisms
exist to ensure that decision-makers are in fact responsive to constituents' concerns.
The more of these criteria are met, the higher the degree of procedural legitimacy. In
48
For such claims in the context of the legitimacy of the EU, see Moravscik, `In Defence of the "Democratic
Deficit": Reassessing Legitimacy in the European Union', 40 JCMS (2002) 603.
49
Those arguing for a global democracy include Held, Democracy and the Global Order (1995). See also
Falk and Strauss, `On the Creation of a Global Peoples Assembly: Legitimacy and the Power of Popular
Sovereignty', 36 Stanford J of Int'l L (2000) 191.
The Legitimacy of International Law in Question 927
many respects mechanisms and ideas derived from domestic administrative law may
be helpful to give concrete shape to ideas of due process on the transnational level.50
Furthermore, principles and mechanisms described by the EU Commission's 2001
White Paper51
could also provide a useful source for giving substance to the idea of
transnational procedural adequacy. Yet it is unlikely that the idea of procedural ade-
quacy as it applies to the various transnational institutional processes will translate into
a standard template of rules and procedures comparable to, say, the US Administrative
Procedure Act. When it comes to assessing procedures as varied as dispute resolution by
the WTO's DSB, UN Security Council decision-making under Chapter VII or prosecutions
under the newly established ICC, a highly contextual analysis that takes seriously the
specific function of the various institutions will be necessary.
D Outcome Legitimacy: Achieving Reasonable Outcomes
The final concern is related to outcomes. Bad outcomes affect the legitimacy of a
decision and tend to undermine the authority of the decision-maker.52
Yet an
outcome-related principle has only a very limited role to play for assessing the
legitimacy of any law. Principles related to outcomes only play a limited role because
disagreement about substantive policy are exactly the kind of thing that legal
decision-making is supposed to resolve authoritatively.53
It is generally not the task
of addressees of norms to re-evaluate decisions already established and legally bind-
ing on them. This is why the legitimacy of a legal act can never plausibly be the
exclusive function of achieving a just result, as assessed by the addressee. Were it
otherwise, anarchy would reign. But that does not preclude the possibility of inter-
national rules that cross a high threshold of injustice or bear a costly inefficiency
being ignored by a national community on exactly the grounds that they are deeply
unjust or extremely costly and inefficient. What needs to be clear, however, is that
any principle of substantive reasonableness is applied in an appropriately deferential
way that takes into account the depth and scope of reasonable disagreement that is
likely to exist in the international community. In particular, where jurisdictional
legitimacy weighs in favour of international law and international procedures were
adequate, there is a strong presumption that a national community's assessment of
the substantive outcome is an inappropriate ground for questioning the legitimacy
of international law.
50
See Stewart, Administrative Law in the Twenty-First Century, available at 78 N.Y.U. L. Rev. (2003) 437.
51
The European Commission's White Paper on European Governance (2001).
52
For a sceptical view that considerations of justice should play a core role in assessing the legitimacy of
international law see T. Franck, Legitimacy Among Nations (1990), at 208.
53
To some extent that is also true about questions of procedure and jurisdiction that the approach
sketched here opens up for evaluation. But there is a difference of degree between them. Questions of pro-
cedure or jurisdiction often provide a focal point for consensus even when an agreement on outcomes
cannot be reached.
928 EJIL 15 (2004), 907­931
3 The Constitutionalist Model: Implications
To what extent then should citizens regard themselves as constrained by international
law in the collective exercise of constitutional government? Are they morally obligated
to always comply with international law? Should they design domestic institutions
with an exclusive commitment to a principle of international legality? Or are citizens
always morally free to democratically reject specific rules of international law as a legit-
imate exercise of constitutional self-government? Should the way citizens structure
domestic institutions be guided by a strong idea of constitutional self-government and
democracy, suggesting that domestically accountable institutions need to independently
assess and democratically endorse the policies embodied in rules of international law,
before international law can be enforced domestically?
If the constitutionalist model is correct, both positions are mistaken. Instead, citizens
in liberal constitutional democracies would do well to adopt a more complex, inter-
mediate approach. On the one hand, the principle of international legality establishes
a presumption in favour of the authority of international law. The fact that there is a
rule of international law governing a specific matter means that citizens have a reason
to do as the rule prescribes. But this presumption is rebutted with regard to norms of
international law that seriously violate countervailing normative principles relating to
jurisdiction, procedure or outcomes. To put it another way: citizens should regard
themselves as constrained by international law and set up domestic political and legal
institutions so as to ensure compliance with international law, to the extent that inter-
national law does not violate jurisdictional, procedural and outcome-related principles
to such an extent that the presumption in favour of international law's authority is
rebutted. When assessing concerns relating to jurisdiction, procedure and outcome
each of the relevant principles can either support or undermine the legitimacy of inter-
national law. As the discussion has shown it is not necessarily the case that jurisdictional
and procedural concerns will weigh in favour of national decision-making, though
often that will be the case. When citizens in a constitutional democracy comply with
legitimate international law, citizens are not compromising constitutional principles.
Instead they are complying with the demands of principle that underlie the best
interpretation of the liberal constitutional tradition they are part of.54
What then are the institutional implications of a constitutional model? How would
citizens committed to a constitutionalist approach structure their domestic institutions
with regard to international law? What should the terms of engagement between
national and international law be? How should the national constitution manage the
interface between national and international law and allocate decision-making
authority between institutions? Here there are no quick and easy answers. In part this
is because each jurisdiction has, as its starting point, its own tradition and institutions
54
With regard to the loss of self-government or sovereignty this entails, Neil MacCormick's point on the
loss of sovereignty applies: sovereignty is not `the object of some kind of zero sum game, such that the
moment X loses it Y necessarily has it. Let us think of it rather more as of virginity, which can in at least
some circumstances be lost to the general satisfaction without anybody else gaining it': MacCormick,
`Beyond the Sovereign State', 56 Modern Law Review (1993) 1, at 16.
The Legitimacy of International Law in Question 929
addressing foreign affairs which would need to be carefully developed within their
own constitutional framework. On application, there is no `one size fits all' solution.
In part it is because a great deal of additional work would need to be done to subject
international human rights law, UN Security Council decision-making, WTO or
NAFTA decision-taking to careful scrutiny. The purpose here was merely to introduce
a general framework of analysis. Such a framework helps ask the right questions and
deal with the right problems. The constitutionalist model provides guidance to
questions of institutional design by clarifying what it is that these institutions should
aim to achieve: They should be designed to ensure that international law is domestically
enforced when it is legitimate. And they should provide constructive avenues for crit-
ical engagement, interpretative amelioration and even outright disobedience, to the
extent that there are serious legitimacy concerns. On application, constitutionalist
analysis involves close and contextual analysis of institutional practices. It does not
lend itself to schematic application.
But there are other features of the constitutionalist model that make it attractive as
an inclusive general framework. First, it avoids one-sidedness by placing on the table
the whole menu of central features that determine the legitimacy of international
law. This holistic focus ­keeping the whole range of factors in view within any given
context ­ is the virtue of a constitutionalist thinking.55
It avoids two pitfalls.
The first danger is to imagine the legal and political world in strictly dualist terms,
suggesting that domestic and international law are radically different and require
radically different conceptual tools of analysis. This is the trap that national constitu-
tional lawyers are prone to fall into. All claims suggesting that there is something
deeply problematic legitimacy-wise about international law because it lacks certain
features that are present on the domestic level ­ the absence of a demos or directly rep-
resentative institutions ­ are exaggerated or mistaken. The legitimacy of international
law does not depend on whether every state has in fact consented or can reasonably
be interpreted to have consented to a specific obligation imposed on it by international
law, or that there is a chain of authorization that connects state consent to a specific
legal obligation. Nor does international law's legitimacy depend on establishing
directly representative institutions on the transnational level. These are all ideas
connected to a statist or nationalist conception of democratic constitutionalism. The
principles of constitutionalism are not, however, tied to the organizational form of the
state, but need to be respected and ought to guide legal and political practice wher-
ever it occurs. National constitutional lawyers are sometimes less open than they
ought to be to incorporate jurisdictional concerns and values relating to the idea of
international legality into their assessment of international law. With the experience
55
For a similar defence of constitutionalism as an approach to the assessment and design of national and
transnational institutions see Walker, `Postnational Constitutionalism and the Problem of Translation',
in J. Weiler and M. Wind (eds.), European Constitutionalism Beyond the State (2003), 27. See also
M. Maduro, From Constitutions to Constitutionalism: A Constitutional Approach for Global Governance
(forthcoming, on file with author). See also C. Joerges, I. Sand and G. Teubner, Transnational Governance
and Constitutionalism (2004).
930 EJIL 15 (2004), 907­931
of European integration shaping the perspective on transnational legal practices,
such an attitude is today less frequently found in Europe than in the US.
On the other hand, there is a growing trend, more pronounced in Europe than in
the US,56
to claim that international law has a constitution. The idea of a constitution
is employed in a number of ways. Some suggest that the UN Charter is to be regarded
as the world's constitution57
or that the WTO is the economic constitution of the
world58
or that there is a global unwritten constitution.59
Once it is clear that a classi-
cal model of international law as the consent-based law between states needs to be
modified ­ among other things to take into account the transitions that characterize
international law as governance ­ the idea that, in one way or another, international
law has a constitution, seems to be an obvious alternative. There is nothing intrinsically
wrong (or new60
) in using the language of constitutionalism to capture more ade-
quately certain features of international law. But it should be clear what exactly is
claimed. The problem is that the language of constitutionalism suggests that, through
law's formalities, dispersed powers and the judicial protection of rights, international
law may be legitimate because the outcome it generates is likely to be fair and just. Yet
as a way to legitimize international law, there is something conspicuously absent in
this kind of international constitutional discourse: the idea of self-government and
electoral accountability. Yet these ideas are at the very heart of challenges to
international law made in the name of national constitutional values. This can make
serious engagement between national and international scholars difficult at times.61
By focusing on the jurisdictional principles of subsidiarity and procedural concerns
and providing a framework that allows for the constructive engagement of these
concerns, the constitutionalist model developed here not only avoids one-sidedness,
but builds a bridge between national and transnational constitutional discourse.
Finally, the constitutionalist approach conceptually reconnects to a tradition of
thinking about national and international law in which the legal and political world
was not yet deeply divided between national and international law. As Tom Franck
puts it:
56
But see, e.g., the Symposium on `The Emerging Transnational Constitution', 37 Loyola of LA L Rev
(2003).
57
See, e.g., B. Fassbender, UN Security Council Reform and the Right of Veto: A Constitutional Perspective
(1998).
58
E.g. Petersmann, `Theories of Justice, Human Rights and the Constitution of International Markets', 37
Loyola of LA L Rev (2003) 407.
59
For an overview of various ways in which the terms `constitution' and `constitutionalization' are used
see Cass, `The "Constitutionalization" of International Trade Law: Judicial Norm-generation as the
Engine of Constitutional Development in International Trade', 12 EJIL (2001) 12.
60
See, e.g., A. Verdross, Die Verfassung der Völkerrechtsgemeinschaft (1926). For an early discussion of the
history the idea of constitutionalism in international law see Opsahl, `An "International Constitutional
Law"?', 10 ICLQ (1961) 760.
61
As a highly respected scholar of US constitutional law put it: `learning what international lawyers were
talking about seemed to me like trying to learn what space aliens were saying. Their concerns appear
completely unconnected to the things I worry about as a domestic U.S. constitutionalist, and yet they
seemed to think that these concerns mattered a lot for domestic constitutional law': Tushnet, supra note
15, 239 at 240.
The Legitimacy of International Law in Question 931
When both the national and international systems of dominance, governance or order were
based on similar divine or natural-order rationales, . . . questions [of legitimacy] were
addressed to both national and international systems, if only because, in the two fields of
inquiry, the respective phenomena of social behavior could not be separated neatly. The same
`natural' or divine validation that was believed to authorize national governance also vali-
dated the rules of interstate behavior. God's writ, or the natural order of things, was no
respecter of the puny lines drawn by men on maps.62
From approximately the mid 19th century until very late in the 20th century
when Tom Franck was writing about international law's quest for legitimacy, the
distinctions between international and national still seemed to be deep. Today citizens
in constitutional democracies would do well to assess and engage the world of legal
and political practice using a conceptual framework that transcends old dichotomies.
The constitutionalist model is committed not to an international constitutional law
but to constitutionalism beyond the state. It suggests that constitutionalist analysis is
usefully undertaken when confronting and assessing legal and political practices that
effect citizens. It is increasingly less obvious that it even makes sense to think about
something called `domestic law' and something else called `international law' relating
to one another. The unity of domestic law is disintegrating at the same time as inter-
national lawyers worry about fragmentation. Simultaneously, deep linkages between
national and international practices in specific areas of regulation are established. It
may be more helpful to think of the legal world consisting of a myriad of legal and
political practices relating to one another in complicated and highly differentiated
ways, ultimately grounded in a commitment to an idea of the human dignity
and autonomy and plausibly assessed by constitutionalism as a method of analysis.
There is no longer a person (the king) or an institution (the parliament), a text (the con-
stitution) or even a source (`We the People') that can plausibly serve as an ultimate
normative point of reference and symbol of the unity and coherence of the legal and
political world. In this sense, the world may be moving beyond constitutions. But it
may just be discovering constitutionalism beyond the state.
62
Franck, supra note 1, at 535.