is there an international environmental constitution

Electronic copy available at: http://ssrn.com/abstract=1281007

UNIVERSITY OF GEORGIA

SCHOOL OF LAW

RESEARCH PAPER SERIES

Paper No. 08-008 October 2008

Is There an International Environmental Constitution?

DANIEL M. BODANSKY Emily and Ernest Woodruff Chair

in International Law University of Georgia School of Law

[email protected]

Forthcoming in Indiana Journal of Global Legal Studies

This paper can be downloaded without charge from the Social Science Research Network electronic library at

http://ssrn.com/abstract=1281007


Emily & Ernst Woodruff Chair of International Law, University of Georgia School of Law.*

Many thanks to the help of my student, Logan Pool, who helped me research and formulate thearguments in this paper. Thanks also to Dan Coenen, Harlan Cohen, Jeffrey Dunoff, AnneHerbert, Lori Ringhand, and the participants at the Kandersteg Workshop, who provided manyuseful comments and suggestions.

1

Draft 1.0: October 6, 2008

IS THERE AN INTERNATIONAL ENVIRONMENTAL

CONSTITUTION?

Daniel Bodansky*

ABSTRACT

The surge of interest among international lawyers in “constitutionalism”represents one of several efforts to reconceptualize internationalgovernance; others include the research projects on global administrativelaw and legalization. The paper applies the constitutionalist lens tointernational environmental law – one of the few fields of international lawto which constitutionalist modes of analysis have not yet been applied.Given the protean quality of the terms “constitution” and“constitutionalism,” the paper begins by unpacking these concepts. Bydisaggregating these concepts into a number of separate variables, whichhave a more determinate, unambiguous meaning, we can answer thequestion, “is there an international environmental constitution?”, in a morenuanced way – not in an all or nothing fashion, but by considering theextent to which international environmental law has constitutionaldimensions. The paper concludes that, although individual treaty regimeshave constitutional features, international environmental law as a wholelacks the hallmarks of a constitutional order.


Bardo Fassbender, The United Nations Charter as Constitution of the International1

Community, 36 COLUMBIA J. TRANSNAT’L L. 529 (1998); Christian Tomuschat, InternationalLaw as the Constitution of Mankind, in INTERNATIONAL LAW ON THE EVE OF THE TWENTY-FIRST CENTURY 37 (UN 1997).

Erika De Wet, The International Constitutional Order, 55 INT’L COMP. L.Q. 51 (2006).2

Anne Peters, Compensatory Constitutionalism, 19 LEIDEN J. INT’L L. 579, 582 (2006).3

See Benedict Kingsbury, Nico Krisch, Richard B. Stewart & Jonathan B. Weiner, eds.,4

Symposium: The Emergence of Global Administrative Law, 68 L. & CONTEMP. PROBLEMS 1(2005); Nico Krisch & Benedict Kingsbury, eds., Symposium: Global Governance and GlobalAdministrative Law in the International Legal Order, 17 EUR. J. INT’L L. 1 (2006).

Judith Goldstein et al., Legalization and World Politics, 54 INT’L ORG. 385 (2000) (special5

issue on legalization). Like constitutionalism and global administrative law, legalization hasboth a descriptive and a normative dimension. Descriptively, scholars have focused on actualprocesses of legalization: for example, greater reliance on hard sources of law such as treaties,more precise rules, and legalized methods of dispute resolution. Id. Normatively, somescholars have argued for strengthening the rule of law in international relations. SimonChesterman, An International Rule of Law?, 56 AM. J. COMP. L. 331 (2008); Jeremy Waldron,The Rule of International Law, 30 HARVARD J. L. & PUB. POL’Y 15 (2006).

2

INTRODUCTION

In recent years, there has been a surge in interest among internationallawyers in “constitutionalism.” Writers have variously argued that:

! An international constitutional order already exists– reflected, forexample, in the UN Charter.1

! An international constitutional order is in the process of emerging.2

! An international constitutional order should emerge.3

These proposals to analyze international law in constitutional termsrepresent one of several efforts to reconceptualize international governance.Another is the research project on global administrative law, which, like4

constitutionalism, draws on domestic legal models to identify emergingpatterns of international governance. Both reflect a broader emphasis on thephenomenon of legalization in international relations – the idea thatinternational relations are increasingly governed by rules rather than merelypower and interests.5

This paper applies the lens of constitutionalism to internationalenvironmental law – one of the few areas of international law to which

3

constitutional modes of analysis have not yet been applied. The subject ofinternational environmental constitutionalism raises three principalquestions:

! First, as a conceptual matter, what is constitutionalism? Whatwould it mean to say that there is an international environmentalconstitution?

! Second, as a descriptive matter, to what extent is there, in fact, aninternational environmental constitution? Or, to put it lesscategorically, to what extent does international environmental lawhave constitutional aspects or dimensions?

! Finally, as a normative matter, would it be desirable to have aninternational environmental constitution and, if so, why? In otherwords, what are the benefits of constitutionalism as a mode ofgovernance?

Although the paper focuses primarily on the conceptual and descriptivequestions, it also comments briefly on the normative issues.

Given the Babel-like quality of the discourse about internationalconstitutionalism, I begin in Section 1 by unpacking the terms,“constitution” and “constitutionalism.” By disaggregating these conceptsinto a number of separate variables, which have a more determinate,unambiguous meaning, we can answer the question, “is there aninternational environmental constitution?”, in a more nuanced way – not inan all or nothing fashion, but by considering the extent to whichinternational environmental law has constitutional dimensions. Myconclusion in Section 2 is that, although individual treaty regimes haveconstitutional features, international environmental law as a whole lacks thehallmarks of a constitutional order.

Does this matter? Should it be a source of concern? As I explore inSection 3, there is no magic to the term, “constitution,” so debates aboutwhether something amounts to a “constitution” are, to some degree, adistraction. Nevertheless, the various dimensions of constitutionalgovernance play an important legitimating role. By ordering and limitinggovernment, they help justify an institution’s authority to adopt rules,decide cases and perform other governance functions. So the question ofwhether, and to what extent, international environmental law has

4

constitutional elements is important. And it will become even moreimportant in the future, as collective action problems such as climatechange demand the exercise of greater authority by internationalinstitutions. To justify such authority, the legitimating power of moreconstitutional forms of governance could help to compensate for thedeclining role of state consent.

I. WHAT DO WE MEAN BY CONSTITUTIONALISM?

A. Three Initial Distinctions

The terms “constitution” and “constitutionalism” are used in manydifferent ways. When writers claim that international law has a“constitutional” character, for example, they may mean:

! International law has basic constitutive rules about how other rulesare created, interpreted, changed, and enforced.

! International law contains fundamental substantive rules, such as theprohibition on the use of force and human rights protections.

! International law constitutes a “system,” rather than merely acollection of discrete regimes.

! International law has a hierarchical structure – that is, some normsare entrenched and superior to others in case of conflict.

! International law has a non-consensual dimension, reflecting morethan simply a series of commitments by states.

! International law reflects process values regarding participation,transparency and a separation of powers.

! International law is “democratic.”

Given the protean quality of the term, “constitution,” when scholarsdisagree about whether some area of international law is a “constitution,”it often isn’t clear whether the disagreements are real or merely reflectdifferences about what writers mean by the term. In this regard, the current

See, e.g., Jeffrey L. Dunoff, Constitutional Conceits: The WTO’s “Constitution” and the6

Discipline of International Law,” 17 EUR. J. INT’L L. (2006) (analyzing constitutionalism asan intellectual movement).

Both Rawls and Dworkin make use of this distinction between “concepts” and7

“conceptions”. See JOHN RAWLS, A THEORY OF JUSTICE (1971); RONALD DWORKIN, LAW’S

EMPIRE 70-71 (1986).

Jeff Dunoff masterfully surveys the range of conceptions of constitutionalism put forward8

5

debates about constitutionalism are reminiscent of questions about “hard”vs. “soft” law or about “law” vs. “non-law,” where it is also frequentlydifficult to determine whether people are arguing about substance orterminology.

In considering the discourse on international constitutionalism, it isuseful to draw three initial distinctions.

First, we need to distinguish constitutionalism as a mode of politicalgovernance from constitutionalism as an intellectual movement.Constitutionalism as an intellectual movement within the scholarship oninternational law is interesting in its own right. What accounts for the6

increasing focus among international lawyers on constitutional governance?Does it reflect actual changes in international relations – an actual growthin constitutionalism as a mode of governance? Is it a reaction to increasingconcerns about fragmentation and illegitimacy in international law – anattempt to put international law on a stronger normative footing? Can it beexplained as an effort by European lawyers to extrapolate or generalizefrom the EU experience to global politics more generally? Does it serve anideological function? Is it a power play by international lawyers – a meansof self-aggrandizement – by giving greater importance to judicial rather thanpolitical processes? Or is it simply academic talk – a way of giving afashionable label to more prosaic features of international law? Althoughthese questions are interesting, they are not the primary focus of this article.Instead, the article will address constitutionalism as a potential mode ofenvironmental governance.

Second, we need to distinguish between the general concept of aconstitution and particular conceptions of constitutions. The general7

concept of a constitution, explored in subsection B below, allows a varietyof more specific conceptions, which define constitutions in thicker, moresubstantive ways. For example, Western liberal constitutionalism8

by international trade lawyers in his article, Constitutional Conceits, supra note 6; see also JoelTrachtman, The Constitutions of the WTO, 17 EUR. J. INT’L L. 623 (2006),

Peters, supra note 3, at 582.9

Martti Koskenniemi, Constitutionalism as Mindset: Reflections on Kantian Themes about10

International Law and Globalization, 8 THEORETICAL INQUIRIES IN LAW 9 (2007).

See generally VICKI C. JACKSON & MARK TUSHNET, COMPARATIVE CONSTITUTIONAL LAW11

(2d ed. 2006) (including chapters entitled “ constitutions without constitutionalism” and“constitutionalism without a constitution”).

See generally K.C. WHEARE, MODERN CONSTITUTIONS (Oxford Univ. Press, 2d ed. 1966);12

cf. Anne Peters, supra note 3, at 581 (defining a constitution as “the sum of basic (materiallymost important) legal norms which comprehensively regulate the social and political life of apolity”). Because a constitution consists of the “rules” of a political community, it does notsimply consist of a description of how a community is, in fact, governed, but necessarily hasa normative aspect. See Thomas Cottier & Maya Hertig, The Prospects of 21 Centuryst

6

emphasizes such important features as democracy, separation of powers,checks and balances (including, in particular, judicial review), andprotections of fundamental civil and political rights. This more particularconception of a constitution has many virtues, but its features are notintrinsic to the general concept of a constitution.

Finally, we need to distinguish between a constitution, on the one hand,and constitutionalism more generally. A constitution consists of thefundamental rules governing a political community, generally but notalways in written form. In contrast, constitutionalism refers to the broaderset of qualities or values associated with the liberal conception of aconstitution – limited government, the value and dignity of the individual,9

and so forth – what Koskeniemmi has referred to as a constitutional“mindset.” Because constitutions and constitutionalism are distinct, it is10

possible to have, on the one hand, constitutions without constitutionalism(as was true in former communist countries and in some developingcountries), and, on the other, constitutionalism without a constitution.11

B. The Concept of a Constitution

How should we understand the concept of a constitution? At the mostgeneral level, a constitution is a higher body of law, typically of an enduringnature, setting forth the fundamental rules of a political community.12

Constitutionalism, 7 MAX PLANCK YBK. U.N. L. 261, 279 (2003) (distinguishing “descriptive”and “normative” concepts of a constitution). It should be noted that Cottier and Hertig’sdistinction between “descriptive” and “normative” differs from the usage in this paper: forthem, the term “normative” refers simply to the rule-like character of constitutions.

H.L.A. HART, THE CONCEPT OF LAW (1961); Bardo Fassbender, The Meaning of13

International Constitutional Law, in TOWARD WORLD CONSTITUTIONALISM: ISSUES IN THE

LEGAL ORDERING OF THE WORLD COMMUNITY 842 (R. St. John MacDonald & Douglas M.Johnston eds, 2005). See also Alfred Verdross, who saw an international constitution asestablishing "those norms which deal with the structure and subdivision of, and the distributionof spheres of jurisdiction in a community.” Quoted in id. at 842.

Thomas Gray, Constitutionalism: An Analytic Framework, in, CONSTITUTIONALISM 189,14

at 192-93 (J. Roland Pennock & John W. Chapman eds., 1979).

Of course, this dichotomy is an over-simplification, since in most cases constitutions15

involve a combination of enacted and customary norms. For example, although the U.S.constitution is primary enacted, some of its norms have had a customary character (the two-term limit on the presidency, for example, arose through a customary process, before beingultimately enacted through the Twenty-Second Amendment). Conversely, while much of theBritish constitution arose through a customary process, Her Majesty’s Stationery Office, as of1995, listed 38 Acts of Parliament in its volumes on “constitutional law,” including the Articlesof Union with Scotland and Northern Ireland. See S.E. Finer, Vernon Bogdanor & BernardRudden, On the Constitution of the United Kingdom, in COMPARING CONSTITUTIONS (2d rev.ed. 1995).

7

These rules can be of two types. First, constitutions set forth the proceduralrules that establish the basic political institutions of a community and definehow other norms are created, interpreted, changed and enforced – whatH.L.A. Hart referred to as “secondary rules” (and, in particular, the mostfundamental secondary rule, that is the “rule of recognition,” which servesas the ultimate test of legal validity) and what Fassbender calls“metarules.” Second, constitutions may, in addition, contain substantive13

rules, such as basic human rights protections. Although these substantiverules are not as obviously constitutive as the procedural rules that establishthe basic framework of government, they also help constitute a politicalcommunity by defining its core values.

In his excellent article, “Constitutionalism: An Analytic Framework,”Thomas Gray notes that constitutions can vary along many dimensions.14

They can be enacted (as in the case of the U.S. constitution) or arise througha customary process (as in the case of the British constitution). They can15

have an extralegal status, or have the status of ordinary or fundamental law.They can be enforced through political or judicial processes. They can befederal, unitary or confederate. And they can be defined narrowly to include

WHEARE, supra note 12 (distinguishing broad and narrow meanings of a “constitution”).16

The Thirteenth Amendment to the U.S. Constitution, outlawing slavery, is a notable17

exception.

In Roman law, the “constituo” was “imperial legislation that transcended all other law.”18

Douglas M. Johnston, “World Constitutionalism in the Theory of International Law,” inTOWARD WORLD CONSTITUTIONALISM, supra note 13, at 17.

8

just the formal rules in a particular document, or more broadly to include allof the fundamental rules of a political community, both formal andinformal, written and unwritten.16

Nevertheless, constitutions typically share several features:

First, they aim primarily to regulate governmental rather than privateactors – they represent a type of public rather than private law. Mostattempts to adopt constitutional limits on private conduct – such as theTwenty-First Amendment in the United States, which prohibited the sale ofalcoholic beverages – have foundered.17

Second, because constitutions address fundamental issues and areintended to provide a stable framework of governance, of indefiniteduration, they typically are entrenched – they are more difficult to changethan ordinary laws. In some cases, entrenchment is formalized by requiringspecial procedures for amendment, such as super-majority voting or evenconsensus. In other cases (such as the British constitution), theentrenchment is informal. It represents a social norm rather than a legalrequirement. In either case, entrenchment serves to take constitutionalissues outside the zone of normal politics.

Third, by the same token, constitutions typically take precedence incase of conflict with other norms – they represent a hierarchically superiorform of law, which trumps ordinary legislation. Analytically, superiority18

and entrenchment are distinct – a norm could be superior but notentrenched, or entrenched but not superior. But generally, constitutionsshare both features, because both derive from the same source: thefundamental nature of the rules contained in a constitution.

Of course, these features are not shared by all constitutions – they arenot necessary conditions of a constitution, even in its “thin” sense.” Mostconstitutions have a special legal status, for example, but some have thestatus of ordinary statutes, and indeed, in some respects, the British

Gray, supra note 14, at 191.19

For a classic articulation of the view that constitutions serve as limitations on government,20

see Giovanni Sartori, Constitutionalism: A Preliminary Analysis, 56 AM. POL. SCI. REV. 853(1962).

See Graham Maddox, A Note on the Meaning of “Constitution,”76 AM. POL. SCI. REV. 80521

(1982) (challenging Sartori’s view, and arguing instead that constitutions are a “combinationof ...power and its control”).

The Latin term “constituere” means to set up, establish, construct. See id. at 806.22

The Contracts Clause is one of the comparatively few substantive rights provided for in the23

U.S. Constitution itself.

9

constitution does not even have a legal character, let alone a supra-legalcharacter. Instead, the concept of a constitution is better seen as a familyresemblance, reflecting various features that systems of governance that19

we characterize as “constitutional” share to greater or lesser degrees. Forthis reason, attempts to define the necessary and sufficient conditions of aconstitution, and then to measure systems of governance that aspire to thisstatus against these defining features, seem misguided. Instead, we needto employ a less categorical approach in considering whether aninternational environmental constitution has emerged. The constitutionalcharacter of governance is not a matter of yes or no, but of more or less.

C. What Functions Do Constitutions Serve?

Constitutions are sometimes depicted as limitations on public power.20

But they, in fact, serve a dual function, both constituting and constrainingpower. As the term “constitution” suggests, constitutive rules defining the21

basic values, institutions and decisionmaking processes of a politicalcommunity are central to the concept of a constitution. Of course, in22

modern liberal constitutions, substantive limits (such as human rightsprotections) are also common. But it is worth remembering that, in theUnited States case, the Bill of Rights was not part of the originalconstitution. In its original form, the US constitution was almost entirelyconstitutive, defining the basic structure of government – in particular, the23

separation of powers among the three branches of the federal governmentand the relationship of the federal government to the states. It was adoptednot so much to limit public power – the federal government under the

Massachusetts Constitution, Declaration of Rights, art. XXX (1780). The phrase was24

originally used by Adams in the 1774 Novanglus Papers, no. 7.

Peters, supra note 3, at 583.25

The relationship between constitutionalism and legalization is even closer to the extent that26

the rule of law is conceived as having substantive as well as procedural aspects.

Philip Allott, Intergovernmental Societies and the Idea of Constitutionalism, in THE27

LEGITIMACY OF INTERNATIONAL ORGANIZATIONS 69, at 90 (Jean-Marc Coicaud & VeijoHeiskanen eds., 2001) (characterizing constitutionalism as “more than the rule of law and lessthan natural law”). Particular conceptions of constitutions may contain additional legitimatingfeatures, such as a separation of powers and human rights protections.

10

Articles of Confederation had few powers to begin with – but rather tocreate a new government with stronger powers. This is true of mostconstitutions: they are conceived as creative (even revolutionary)instruments, either establishing a new political order or providing a freshstart for an existing one.

Nevertheless, the common emphasis on the constraining nature ofconstitutions is well-founded, since a central feature of constitutionalgovernance is its limited character. This feature is most obvious in the caseof constitutions that impose substantive constraints on power – for example,by prohibiting ex post facto laws or by protecting human rights. But evenwhen a constitution is purely constitutive, it is still constraining, since theprocedural rules set forth in a constitution do not simply create institutions,but also define and thereby limit those institutions, in much the same waythat the rules of a game both create the game and limit the moves that canbe made.

By limiting governmental authority, constitutions serve an importantlegitimating function. A constitution creates, as John Adams put it, “agovernment of laws, not of men” (or in the case of international law, not24

of states). In this respect, constitutionalism is closely related to25

legalization. But it goes beyond legalization by limiting the legislature26

itself through a body of rules that are fundamental, entrenched andhierarchically superior to ordinary law. Such constitutional limits make27

governance less democratic and flexible, but provide compensating benefits– most importantly, stability, predictability, and security against the threatof unrestricted government power.

Cf. John Marshall’s famous phrase in McCulloch v. Maryland, 17 U.S. 316 (1819), “we28

must never forget that it is a Constitution we are expounding.” See also BENJAMIN N.CARDOZO, THE NATURE OF THE JUDICIAL PROCESS 83 (1921) (“A constitution states or oughtto state not rules for the passing hour, but principles for an expanding future.”). It should benoted that many leading constitutional scholars, particularly in the United States, reject theview that constitutions should be interpreted in a more teleological manner than ordinarystatutes, and instead advocate an originalist or strictly textualist approach. As Jan Klabbersnotes, the tension between a constitution’s constraining function and its open-ended, “living”nature is nicely captured in Wheare’s observation that courts “may choose to treat aconstitution as a living instrument, but they must treat it first of all as a constitution.” JanKlabbers, Constitutionalism Lite, 1 INT’L ORG. L. REV. 31, 52 n.101 (2004) (quoting K.C.WHEARE, FEDERAL GOVERNMENT (1947)).

Peters refers to this as “micro-constitutionalist” analysis. Peters, supra note 3, at 594.29

Cf. Joseph Raz, On the Authority and Interpretations of Constitutions: Some Preliminaries,30

in CONSTITUTIONALISM: PHILOSOPHICAL FOUNDATIONS 152, at 153 (Larry Alexander ed.,1999) (distinguishing “thin” and “thick” senses of a constitution).

11

That said, the limits imposed by constitutions are often pliable in waysthat ordinary laws are not. Because they establish ongoing systems ofgovernance, intended to persist indefinitely into the future (but difficult toamend), constitutions must be framed at a broad level of generality. Theymust allow considerable flexibility if they are to avoid becoming outdated.This open-ended quality of constitutions is manifested in the doctrine ofimplied powers – which is stated explicitly in the “necessary and proper”clause of the US Constitution and has been inferred in the case ofinternational organizations – as well as in the more spacious approach oftenemployed to interpret constitutions as compared to ordinary laws.28

II. IS THERE AN INTERNATIONAL ENVIRONMENTAL CONSTITUTION?

As the preceding discussion makes plain, claims about internationalconstitutionalism can mean several things. At the most minimal level, theycall attention to the constitutive aspect of many international agreements,which establish ongoing systems of governance. Numerous international29

environmental agreements have a constitutional dimension in this “thin”sense. Rather than elaborate a set of static commitments by states, they30

are dynamic arrangements, which establish ongoing systems of governanceto address particular issue areas like climate change, ozone depletion or

International Convention for the Regulation of whaling, Dec. 2, 1946, 161 UNTS 7231

(hereinafter 1946 Whaling Convention).

For example, the 1931 Whaling Convention prohibited the taking of right whales, as well32

as immature whales and female whales accompanied by calves or suckling whales.International Convention for the Regulation of Whaling, Sept. 24, 1931, arts. 4, 5, 49 Stat.3079, T.S. 880, 155, LNTS 349.

1946 Whaling Convention, supra note 31, at arts. III, V.33

Montreal Protocol on Substances that Deplete the Ozone Layer, Sept. 16, 1987, 26 ILM34

1541 (1989).

See EDWARD PARSON, PROTECTING THE OZONE LAYER: SCIENCE AND STRATEGY (2003).35

12

whaling. Although they differ in their particulars, each performs basicconstitutive functions: they establish institutions, specify the rules that guideand constrain these institutions, and entrench these rules through theiramendment procedures.

Consider, for example, the 1946 International Convention for theRegulation of Whaling (ICRW). In contrast to earlier whaling agreements,31

which had consisted solely of specific regulatory limits on whaling, the32

1946 Convention served a constitutive function as well, establishing theInternational Whaling Commission (IWC) and authorizing it to adoptregulations on an ongoing basis by a three-quarter majority vote, which areset forth in a schedule to the Convention. The particular regulatory limits33

imposed by the 1946 Convention are now long forgotten. But the ongoingsystem of governance it established has proven remarkably robust,eventually leading to the adoption by the IWC of a moratorium oncommercial whaling in 1982, which continues to this day.

Many other environmental treaties serve a constitutive as well as aregulatory function. For example, the 1987 Montreal Protocol onSubstances that Deplete the Ozone Layer not only set forth targets and34

timetables to reduce the use of certain ozone-depleting substances; it alsoestablished procedures to amend and adjust these regulatory requirements.This has led to a progressive tightening of the reduction schedule for CFCsand halons, the two classes of chemicals originally addressed, as well as theinclusion of numerous other ozone-depleting substances in the Protocol’sregulatory scheme, including carbon tetrachloride, methyl chloroform andHCFCs.35

Cottier & Hertig, supra note 12, at 270-71.36

Bardo Fassbender, Sovereignty and Constitutionalism in International Law, in37

SOVEREIGNTY IN TRANSITION 115, 142 (Neil Walker ed., 2003). On this view, norms of juscogens represent a higher substantive law with constitutional status.

DEBORAH Z. CASS, THE CONSTITUTIONALIZATION OF THE WORLD TRADE ORGANIZATION:38

LEGITIMACY, DEMOCRACY AND COMMUNITY IN THE INTERNATIONAL TRADING SYSTEM (2005).

13

As constitutive instruments, treaties like the ICRW and the MontrealProtocol typically represent just the tip of the normative iceberg within theregimes they establish. The majority of the norms are created not by theoriginal treaty itself, but through the more flexible and dynamic governancearrangements constituted by the treaty. International environmental regimesusually distinguish between these constitutive and non-constitutive elementsby placing them in different instruments. The constitutive elementsestablishing the basic institutions and decisionmaking procedures of theregime are entrenched in the treaty text itself, which like a constitution isdifficult to amend. Meanwhile, the regulatory elements are placed inprotocols, annexes, schedules, or decisions of the parties, which are easierto change.

If characterizing international law in constitutional terms merelyreferred to the constitutive function of many treaties, then suchcharacterizations would be unproblematic. But claims about theconstitutional status of international law usually serve as a short-hand fora stronger set of claims – some putatively descriptive, others more openlynormative – that international law is, or should be, a constitutional order ina “thicker” sense. For example, some scholars argue that international lawis developing a constitutional character because it:

! represents the law of an emerging international community, withshared responsibility and solidarity,36

! has a “higher” status, not dependent on state consent,37

! represents a unified, comprehensive system, rather than a series ofdiscrete regimes,

! provides for a process of judicial review.38

See, e.g., Ernst-Ulrich Petersmann, How to Constitutionalize International Law and Foreign39

Policy for the Benefit of Civil Society?, 20 MICH. J. INT’L L. 1 (1998-1999).

See, e.g., Pierre-Marie Dupuy, The Constitutional Dimension of the Charter of the United40

Nations, 1 MAX PLANCK YBK. U.N. L. 1 (1993); Fassbender, supra note 1; Thomas M. Franck,Is the UN Charter a Constitution?, in VERHANDELN FÜR DEN FRIEDEN – NEGOTIATING FOR

PEACE: LIBER AMICORUM TONO EITEL 95 (Jochen Abr Frowein et al eds., 2003).

Fassbender, supra note 37, at 142.41

UN Charter, art. 108.42

14

Others argue that, if international law does not, as a descriptive matter,already have these features, then, as a normative matter, it should developthem.39

From a descriptive standpoint, the UN Charter is often cited as thepreeminent example of an emerging international constitution in this thickersense. Like a constitution:40

! The UN Charter articulates the fundamental rules of theinternational system (such as the principles of sovereign equalityand self-determination, the prohibition on the use of human force,and the obligation to promote and respect human rights).

! These fundamental rules, although adopted in a treaty requiring stateconsent, represent more than simply bargains negotiated by states.Instead, they represent core values of an emerging internationalpolitical community, which are outside of politics and thus “nolonger dependent on the capriciousness of sometimes well-meaning,sometimes egoistic states.”41

! The rules of the UN Charter are intended to provide a stableframework of governance, of indefinite duration. Accordingly, theyare entrenched, requiring for amendment ratification by two-thirdsof the member states, including all of the permanent members of theSecurity Council.42

! Finally, Article 103 makes the Charter hierarchically superior toother treaties.

CASS, supra note 38; ERNST-ULRICH PETERSMANN, CONSTITUTIONAL FUNCTIONS AND43

CONSTITUTIONAL PROBLEMS OF INTERNATIONAL ECONOMIC LAW (1991); John H. Jackson, TheWTO “Constitution” and Proposed Reform: Seven “Mantras” Revisited, 4 J. INT’L ECON. L.67 (2001). As Dunoff has explored, writings on WTO constitutionalism focus on differentaspects of the WTO regime: its general institutional structure, its normative commitments, andits mechanisms of judicial review. Dunoff, supra note 6.

See Shirley V. Scott, The LOS Convention as a Constitutional Regime for the Oceans, in44

STABILITY AND CHANGE IN THE LAW OF THE SEA: THE ROLE OF THE LOS CONVENTION, at 9(Alex G. Oude Elferink ed., 2005). At the end of the negotiations, the president of thenegotiating conference, Tommy Koh of Singapore, argued that UNCLOS represented “aconstitution for the oceans” because of its comprehensiveness, wide participation, coherencewith the wider system of international governance, and promotion of global interests andequity. See David Anderson, “‘Constitutionalism’ and the Law of the Sea” (unpublishedmanuscript) (noting that Koh’s term has an “element of hyperbole or metaphor”).

Even the United Nations Charter does not create institutions with general law-making or45

enforcement powers, and many of the most fundamental secondary rules of international laware found in other instruments, such as the Vienna Convention on the Law of Treaties and theILC Articles on State Responsibility. Moreover, the UN Charter lacks many of the featuresof the Western liberal conception of a constitution, including democratic decisionmaking, astrong separation of powers, and judicial review. See Jose Alvarez, ConstitutionalInterpretation in International Organizations, in LEGITIMACY OF INTERNATIONAL

ORGANIZATIONS, supra note 27, at 106-09 (describing “unconstitutional” aspects of the UNCharter). For a similarly skeptical view of the constitutional status of UNCLOS, see Anderson,supra note 44.

See Helfer, Constitutional Analogies, 37 LOYOLA L.A. L. REV. 193, 230-31 (2003-2004)46

(discussing role of exit in international treaty design).

15

Similar types of claims are made for the two other leading candidates forconstitutional status in international law: the WTO Uruguay RoundAgreements and the UN Convention on the Law of the Sea.43 44

Whether even the UN Charter, or UNCLOS, or the WTO agreementsestablish constitutions in a thick sense is debatable, and beyond the scope45

of this paper. But multilateral environmental agreements clearly do not doso. First, they do not typically create systems of governance with significantindependence from states. Multilateral environmental regimes addressingclimate change, ozone depletion, hazardous chemicals or endangeredspecies are still very much state-driven. States generally retain the right ofexit, by withdrawing from the treaty. The secretariats established by46

multilateral environmental agreements are, in most cases, comparativelyweak, and it is unclear whether the agreements even establish international

This issue has been the subject of dispute in the case of the UN Framework Convention on47

Climate Change. See generally Robin Churchill & Geir Ulfstein, Autonomous InstitutionalArrangements in Multilateral Environmental Agreements: A Little-Noticed Phenomenon inInternational Law, 94 AM. J. INT’L L. 623, 648-49 (2000).

1946 Whaling Convention, supra note 31, at Art. V(2)(c).48

In 1993, the Chairman of the IWC’s Scientific Committee, Philip Hammond, resigned,49

saying that he could no longer continue to serve as “spokesman for a Committee whose workis held in such disregard by the body to which it is responsible.” See William Aron, WilliamBurke & Milton M.R. Freeman, The Whaling Issue, 24 MARINE POLICY 501 (2000).

Jan Klabbers makes the same point about international organizations more generally.50

Klabbers, supra note 28, at 43-44.

Vienna Convention for the Protection of the Ozone Layer, March 22, 1985, art. 9(5), 151351

UNTS 324.

16

organizations with an independent legal personality. Although non-state47

actors play a significant role in the standard-setting and compliance process,this influence is generally extra-constitutional, because it takes placethrough informal processes, outside of the formal arrangements provided forby the regime’s foundational instrument.

Second, multilateral environmental agreements do not effectively limitthe institutions they establish. Consider, for example, the InternationalWhaling Convention. In theory, the Convention sets limits on how theInternational Whaling Commission may operate. According to theConvention, the IWC should act on a scientific basis, and should have asits purpose the “orderly development,” not the abolition, of the whalingindustry. But these requirements did not prevent the IWC from adopting48

a moratorium on commercial whaling, which the chairman of the IWC’sown scientific committee argued was without scientific basis, and which49

is arguably preservationist rather than conservationist in orientation.

As the whaling case illustrates, international environmental agreementsare better at constituting than constraining. They serve primarily to create50

new political arenas, rather than to take issues outside of politics. Theadoption of the 1990 London Amendments to the Montreal Protocolprovides another illustration. Because the Montreal Protocol does not setforth any rule concerning its own amendment, the relevant rule is found inits parent instrument – the Vienna Convention for the Protection of theOzone Layer – which provides that amendments of protocols must beratified by two-thirds of the protocol parties in order to enter into force.51

London Adjustments and Amendments to the Montreal Protocol on Substances that Deplete52

the Ozone Layer, June 29, 1990, art. 2, 30 ILM 537.

Similarly, some argue that the Fourteenth Amendment to the US Constitution was adopted53

in an extra-constitutional manner. See BRUCE ACKERMAN, WE THE PEOPLE:TRANSFORMATIONS (2000).

Of course, in exceptional circumstances, even domestic constitutions do not necessarily54

constrain.

I use the word “phenomenon” rather than “problem” deliberately, since it is not clear to me55

that fragmentation is problematic in the international environmental context. Cf. MarttiKoskenniemi & Päivi Leino, Fragmentation of International Law: Postmodern Anxieties? 15LEIDEN J. INT’L L. 553 (2002).

17

Despite this “constitutional” constraint, when the parties to the MontrealProtocol adopted the London Amendments in 1990, they decided to avoidthis rule by providing in the London Amendments themselves that theamendments would require ratification by only twenty states in order toenter into force. In essence, the attitude of states was – “this is our treaty,52

and if we want to adopt an amendment, we can do so using whatever ruleswe like.” This attitude may be justified in certain circumstances – the USConstitution was itself adopted by a procedure contrary to the then-existingconstitution, the Articles of Confederation – but it is hardly an attitude53

characteristic of those who view the existing legal order as constitutional incharacter.54

Due to these weaknesses, if multilateral environmental agreementsqualify as constitutions at all, they are at most “constitutions” with a little“c” rather than a big “C”. They are constitutions in the thin sense that theyestablish ongoing systems of governance and default rules for how thosesystems of governance will operate, but not in the thicker sense ofestablishing institutions with significant independence from states, or oftaking issues outside the zone of politics by imposing effective constraints,either procedural or substantive, on how international environmentalinstitutions behave. Nor do they establish a constitutional order forinternational environmental law as a whole. Instead, they establish abalkanized system, with a multiplicity of regimes, each with their own“constitutions.” In essence, they contribute to, rather than counteract, thephenomenon of fragmentation.55

Is there an international environmental constitution more generally,which applies to international environmental law as a whole? Certainly,

Daniel Bodansky, Does One Have to Be an International Lawyer to Be an International56

Environmental Lawyer?, 100 PROC. AM. SOC. INT’L L. 303 (2006); Daniel Bodansky, JuttaBrunnée & Ellen Hey, International Environmental Law: Mapping the Field, in OXFORD

HANDBOOK OF INTERNATIONAL ENVIRONMENTAL LAW 1 (Daniel Bodansky, Jutta Brunnée &Ellen Hey eds., 2007).

See Churchill & Ulfstein, supra note 47.57

Marti Koskenniemi, Breach of Treaty or Non-Compliance? Reflections on the Enforcement58

of the Montreal Protocol, 2 YBK. INT’L ENVTL. L. 123 (1992).

On international environmental principles, see generally NICOLAS DE SADELEER,59

ENVIRONMENTAL PRINCIPLES: FROM POLITICAL SLOGANS TO LEGAL RULES (2002).

De Wet, supra note 2, at 53.60

18

international environmental law has a number of characteristic features,which serve to distinguish it from classical international law – widespread56

use of the framework convention/protocol approach, rapid amendmentprocedures, a distinctive system of treaty bodies, and non-compliance57

procedures that are more political than judicial in nature. But these58

distinctive features of international environmental law do not amount to a“constitution” in any meaningful sense of the term. They do not establishunitary requirements with regard to the creation and operation ofinternational environmental law – “secondary” rules about howinternational environmental law is developed and enforced. (Indeed, someprominent features of international environmental law – such as the use ofpolitically-oriented non-compliance procedures – cut strongly against theconcept of constitutionalism.) Rather, they represent a toolbox that statescan use when addressing new problems, much like a carpenter’s toolbox ofhammers, screwdrivers, drills, and measuring tapes. We do not consider thecarpenter’s collection of instruments to be a “constitution” for carpentry.By the same token, the toolbox of international environmental lawyers doesnot represent a constitution of international environmental law.

A more promising candidate to fill the role of an internationalenvironmental constitution are the general principles of internationalenvironmental law, such as the duty to prevent transboundary harm, thepolluter-pays principle, the precautionary principle, the principle ofcommon but differentiated responsibility, and the principle of sustainabledevelopment, which cut across the various treaty regimes we examinedearlier. Do these general principles represent a “core value system” for59 60

See, e.g., Christopher D. Stone, Is There a Precautionary Principle? 31 ENVTL. L. REP.61

10790 (2001); Daniel Bodansky, Deconstructing the Precautionary Principle, in BRINGING

NEW LAW TO OCEAN WATERS 381 (D.D. Caron & H.N. Scheiber eds., 2004).

19

the international community – a set of fundamental, substantive rules thatmight be analogized to a constitution?

Perhaps. But if they do represent an incipient internationalenvironmental constitution, they are a weak and vague one. Althoughnorms such as the precautionary principle and the principle of sustainabledevelopment serve to structure international discourse on environmentalproblems, they have so many different meanings that they exercise little realconstraint on the behavior of states or international institutions.61

Moreover, even to the extent that they do serve a “constitutional”function by articulating a set of overarching, shared values, this is only oneof several constitutional functions. International environmental law stilllacks a general system of governance – a world environmental organization,for example – which could provide a common framework for themultiplicity of existing international environmental regimes. It lackscommon rules that cut across and serve to organize its different regimes.And it lacks other features often associated with constitutionalism, such asjudicial review or a separation of powers more generally.

III. DOES IT MATTER WHETHER

INTERNATIONAL ENVIRONMENTAL LAW HAS A CONSTITUTION?

Discussions about the constitutional status of international law assumethat the answer matters – that something important is at stake. But is thistrue? What, if anything, is the value-added of “constitution talk”?

Consider, for example, the descriptive claims regarding theconstitutional status of international law. Why can’t these stand on theirown bottom, so to speak? Why can’t we consider whether internationalenvironmental law is a comprehensive system, or is developing a non-consensual character, or reflects the law of an emerging internationalcommunity, without regard to the vexing question of whether it hasassumed a constitutional character? For example, having described the UN

See Scott, supra note 44.62

Alec Stone Sweet & Jud Mathews, Proportionality Balancing and Global63

Constitutionalism, 47 COLUMBIA J. TRASNAT’L L. (forthcoming 2008).

20

Convention on the Law of the Sea as an entrenched, stable system ofgovernance, of indefinite duration, intended to address ocean issuescomprehensively, with aspirations for universal membership, what does thelabel “constitution” contribute? What additional insights does it generate?62

Similarly, with respect to the normative claims regarding constitutions andconstitutionalism, why can’t we consider whether judicial review or adiminished role for state consent would be a “good” thing, withoutintroducing the concept of “constitutionalism”? What, if anything, does theconcept of constitutionalism add to our analysis?

To my mind, the constitutionalist movement in international law hasnot sufficiently addressed these questions. One possible answer is thatconstitutional discourse serves as an efficient shorthand to indicate acollection of more specific attributes, such as hierarchy, entrenchment, andso forth – a shorthand, in essence, for limited government. The problemwith this response is that the meaning of “constitutionalism” is so unclear– the term is used in so many different ways – that it seems more confusingthan simplifying.

Another possible argument for constitutional modes of analysis is thatthe various features of constitutionalism that we examined earlier are notdisparate phenomenon, but go together. Thus, we need to study them in aholistic way, rather than in the disaggregated manner suggested earlier.

In principle, the holistic nature of constitutionalism could have eitheran empirical or a normative character, or both. Empirically, if the differentelements of constitutionalism (for example, entrenchment, superiority, andjudicial review) are interrelated – if they stand in some kind of causalrelationship with one another – then presence of one “constitutional” featurewould suggest the presence of others. Examining international law throughthe lens of constitutionalism would thus help us see aspects of theinternational legal system that might otherwise be overlooked orinsufficiently appreciated – the role of proportionality balancing ininternational dispute resolution, for example, which Alec Stone Sweetargues is a characteristic feature of constitutional adjudication. This is63

This seems to be the view of Peters, supra note 3, and Klabbers, supra note 28, at 47-48.64

21

often the benefit of domestic analogies: they allow us to understandinternational law in new and useful ways.

But while these empirical relationships are interesting, I think theemphasis on constitutional discourse has a primarily normative impetus. Incontrast to the descriptive perspective, which presumes an empiricalcorrelation between the different features of constitutionalism, thenormative argument is that these different features should go together or, incombination, take on special normative significance. If the various featuresof constitutionalism are normatively related, then the presence of some“constitutional” features would seem to suggest the need for others. To theextent that an international regime creates an ongoing system of governancethat limits state consent, then this creates a need for transparency, aseparation of power, judicial review, and so forth. Or, to the extent that theUN Charter serves as a constitution for the international community, thenit can legitimately make demands on non-member states and even non-stateactors; it should be able to develop as a “living instrument,” includingthrough “dynamic-evolutionary” interpretation; it should be open to allstates; it should provide for judicial review; it should have direct effects indomestic law, and so forth.

Whether these arguments are persuasive depends on whether thedifferent features of a constitution are merely additive, or whether, whencombined, a new type of governance emerges with properties of its own.If the former, then we would be better off focusing directly on itscomponent elements rather than speaking in constitutional terms. Weshould consider whether international law has, or should have, judicialreview, or a doctrine of implied powers, or direct effects – or whateverdescriptive or normative implications are drawn from the supposedly“constitutional” status of international law – without wrapping thesearguments in a constitutional gauze, which makes it more difficult to focuson the specific normative relationships between specific features of theinternational legal system. But if the different dimensions ofconstitutionalism are normatively linked, or provide legitimacy in a waythat exceeds the sum of the parts, then it is appropriate – indeed necessary– to study constitutions as a distinctive phenomenon.64

TETSUO SATO, EVOLVING CONSTITUTIONS OF INTERNATIONAL ORGANIZATIONS (1996); cf.65

Helfer, supra note 46, at 204 (distinguishing WTO dispute settlement from constitutionalismbecause of its less teleological method).

Peters, supra note 3. On judicial review, in particular, see Erica De Wet, Judicial Review66

as an Emerging General Principle of Law and Its Implications for the International Court ofJustice, 47 NETH. INT’L L. REV. 181 (2000); James Crawford, Marbury v. Madison at theInternational Level, 36 G.W. INT’L L. REV. 505 (2004).

Cottier & Hertig, supra note 12, at 299 (describing a “five story” constitutional house).67

Daniel Bodansky, The Legitimacy of International Governance: A Coming Challenge for68

International Environmental Law, 93 AM. J. INT’L L. 596 (1999).

22

The normative arguments regarding international constitutionalismproceed in different ways, depending on whether or not, as a descriptivematter, international law has already developed a constitutional character.That is why the answer to the descriptive question matters about whetherthere is an international environmental constitution. On the one hand, tothe extent that international law has already developed into a constitutionalorder, then this has certain normative implications. For example, it mightgive international law greater legitimacy. Or it might mean thatinternational law should be seen as superior to national law, or should beinterpreted liberally, using a teleological approach, as befits a constitution.65

On the other hand, to the extent that the international legal system still fallsshort of a constitutional order, then this suggests the need for furtherconstitutionalization. To become more legitimate and effective, and therebycompensate for the declining authority of states, the international legalsystem needs to become more constitutional – for example, by limitingpublic power through checks and balances, human rights protections, andjudicial review, or possibly through new constitutional structures66

appropriate to multilevel governance and shared sovereignty.67

The latter belief, I think, underlies the recent focus onconstitutionalism. Traditionally, international law has relied for itslegitimacy on state consent. But to the extent that non-consensual norms68

and decisionmaking processes will need to play a larger role in the future,to respond to collective action problems such as climate change, theninternational law will need some new basis of legitimacy. Initially,democracy seemed a potential candidate. But it is difficult to conceive howdemocracy could operate at the global level, in the absence of a global

See Mattias Kumm, The Legitimacy of International Law: A Constitutionalist Analysis, 1569

EUR. J. INT’L L. 907 (2004); Peters, supra note 3. Of course, not all would agree with thenormative appeal of global constitutionalism. See, e.g., Ernst A. Young, The Trouble withGlobal Constitutionalism,38 TEX. INT’L L.J. 527 (2003).

23

demos. So constitutionalism has become an attractive alternative. If the69

various features of constitutionalism – including limited government,separation of powers, judicial review, transparency, and human rightsprotections – represent the touchstone of legitimacy, then the constitutionalstatus of international law matters deeply. To put it crudely, ifconstitutionalism is a “good thing,” then giving international law aconstitutional character would make international law a good thing as well.

I am sympathetic to this general sentiment, and believe that providing“constitutional” limits on governance could help legitimate internationalenvironmental institutions. Yet I remain skeptical that we need to thinkabout these constitutional issues in holistic terms. Given the complexityand vagueness of constitutional discourse, as well as the difficulty ofachieving the different dimensions of a constitution at once, we might dobetter by unpacking the concept into its component elements, and thenconsidering the proper role of each in the distinctive context of internationalgovernance, rather than attempting to ground the legitimacy of internationallaw in its constitutional status tout court.

CONCLUSION

States face common environmental problems, but they do notnecessarily have common interests or values. This perhaps accounts, atleast in part, for their weak commitment to constitutionalism and, indeed,even to legalization. The refusal by the International Whaling Commissionto lift its moratorium on commercial whaling, the reliance in multilateralenvironmental regimes on non-judicial compliance procedures, the adoptionof the London Amendments to the Montreal Protocol by a procedure atvariance with the regime’s requirements – all of these examples reflect thefact that, in today’s world, states still prefer to retain significant flexibilityrather than pre-commit to constitutional constraints. They remain far from

Id.70

24

developing the sense of community that underlies constitutionalgovernance.

Over the long haul, it is unclear whether this state-driven approach willprove adequate to address international environmental problems such asclimate change and loss of biodiversity. As I have argued elsewhere, suchproblems may require the creation of international institutions with greaterauthority. And, if this happens, it could push international environmental70

law in the direction of greater constitutionalization, to address concernsabout the legitimacy of these new, more powerful, institutions.

But any transformation of this kind still lies in the future. For themoment, international environmental law remains a set of commitments bystates, rather than a constitutional order.

is there an international environmental constitution

Documents