1045

CHANGING RELATIONSHIPS AMONG

EUROPEAN CONSTITUTIONAL COURTS

members of the Supreme Court of the United States visited

several European courts with “constitutional” responsibilities. We

met many of our European counterparts; we heard them describe

how their courts worked; and we learned, among other things,

something about the changing relationships among national, and

“Europe-wide,” constitutional courts. In this paper, I shall report

a few of my own reactions to that visit. Although we discussed

legal topics ranging from court procedures to affirmative action, I

shall focus primarily upon the matter of institutional

San Francisco on August 15, 1938. He married Joanna Hare in 1967. They have three

children, Cloe, Nell, and Michael. He is a graduate of Stanford University, Oxford

University (Magdalen College), and Harvard Law School. During the United States

Supreme Court’s 1964 Term he was law clerk to Justice Arthur J. Goldberg. In 1965-67 he

worked as Special Assistant to the head of the Justice Department’s Antitrust Division.

From 1967 through 1980 he taught at Harvard University as Professor of Law, and at

Harvard’s Kennedy School of Government. He also worked as an Assistant Watergate

Special Prosecutor (1973), as a Special Counsel to the Senate Judiciary Committee (1975),

and as the Judiciary Committee’s Chief Counsel (1979-80). In 1980 he was appointed

Judge of the United States Court of Appeals for the First Circuit. He became the circuit’s

Chief Judge in 1990. He has also served as a Member of the Judicial Conference of the

United States and of the United States Sentencing Commission. He has written books and

articles in the fields of administrative law and government regulation. President Clinton

nominated him as an Associate Justice, and he took office in August 1994.

1 The delegates included Hon. Sandra Day O’Connor, Associate Justice, Supreme

Court of the United States; Hon. Anthony M. Kennedy, Associate Justice, Supreme Court

of the United States; Hon. Ruth Bader Ginsburg, Associate Justice, Supreme Court of the

United States; Hon. Stephen G. Breyer, Associate Justice, Supreme Court of the United

States; Hon. Thomas R. Phillips, Chief Justice, Supreme Court of Texas; Hon. Richard S.

Arnold, Circuit Judge, United States Court of Appeals for the Eighth Circuit; Prof. Paul

Gewirtz, Special Representative for the Presidential Rule of Law Initiative, Potter Stewart

Professor of Constitutional Law, Yale Law School; Dean Paul R. Verkuil, Dean and

Professor of Law, Benjamin N. Cardozo School of Law, Yeshiva University; Thomas M.

Susman, American Bar Association Representative; Prof. Michael Rosenfeld, Justice

Sydney L. Robins, Professor of International Law and Human Rights, Benjamin N.

Cardozo School of Law, Yeshiva University.

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

“relationships” within Europe.

I recognize that these reactions are informed by a brief twoweek

visit during which we saw several different courts.

Nonetheless, they may stimulate discussion and perhaps lead to

further study of matters that I believe are important not simply to

Europe, but to our own legal system as well. I shall first set forth a

few “hornbook” European Community (“EC”) institutional

principles, as I understand them. Then I shall describe, city by

city, related observations and questions that arose during our visit.

At the outset, it may help to set forth some of the EC legal

structure for those of us who, like myself, are not experts. I will

provide a crude summary of a few relevant EC institutional

principles that I have drawn from the more sophisticated articles

which, technically speaking, is now a portion of a somewhat

broader entity called the European Union (“EU”). The EC’s four

basic institutions are the Council, the Commission, the Parliament,

and the Court.

EC’s primary policy-making and legislative body. It has fifteen

members, one representing each state, although the individual

representative of each member state varies depending upon the

subject matter to be discussed. When the subject is finance, for

example, each state will likely send its finance minister to Brussels

for the discussion. When the matter relates to particularly

important economic issues, including the move toward a single

currency, each state sends its Prime Minister.

executive authority. It has twenty members, each of whom is

responsible for one or more of the twenty-six “directorates”—

departments that administer various activities (e.g., competition,

external tariffs, transport, the environment, etc.). Together they

3 The six founding members include Belgium, France, Germany, Italy, Luxembourg,

and the Netherlands. In 1973 Denmark, the United Kingdom, and Ireland joined the EC;

Greece joined in 1981; and Portugal and Spain joined in 1986. Finally, in 1995, Austria,

Finland, and Sweden brought the total number to 15 member states.

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

employ about 20,000 civil servants, most of whom work in

Brussels. Each member state selects at least one, but no more

term and each, in principle, owes his or her loyalty to the EC

simple majority.

members. It has significant power over the EC’s budget, but it has

little direct legislative power. It can participate in the legislative

process, but it cannot enact “laws.” Typically, it votes by simple

majority and its members normally split along “party,” not

“national,” lines.

in greater detail below.

The EC’s “constitution” consists of treaties (including Rome,

not only over trade and commerce, but also in such areas as the

environment, health, safety, consumer protection, and, to a degree,

education. The treaties are highly detailed; together they contain

more than 500 articles, many of which are open to court

application and interpretation. The EC institutions also make law

in the form of directives, regulations, and decisions by the Council

and the Commission in the exercise of their law-making authority

157(1), 298 U.N.T.S. 71-72 [hereinafter EC Treaty].

5 The Commission is responsible for ensuring member state compliance with the

treaties and therefore can initiate an action before the ECJ against a member state which

fails to fulfill its obligation.

6 In addition to the European Court of Justice, there is one other EC court: the Court

of Auditors, based in Luxembourg, carries out the EU’s audit, “examin[ing] the accounts

of all revenue and expenditure of the Community,” and “examin[ing] whether all revenue

has been received and all expenditure incurred in a lawful and regular manner and

188C(1)-(2).

7 Technically speaking, the Treaty of Rome created the EC; the Maastricht Treaty

expanded the authority of the EC and also created a constitutional structure, or “three

pillars,” for what is now called the European Union. The first, and central, pillar is the

existing EC and its law. The second pillar is a common foreign and security policy. The

third pillar is cooperation in the fields of justice and home affairs. Because the Court’s

basic concern is the first pillar, or EC law, I shall not go into these latter complicating

features.

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

The basic system for making laws involves the Commission

writing the text of a law and then proposing it to the Council for

adoption. The Council then decides whether or not to enact the

most important and far-reaching legislative acts, the Council could

act only by unanimity. More recently, EC members have

introduced a system of weighted majority voting in certain subject

areas, including health, safety, the environment, consumer

protection, and education. This voting system permits Council

enactment by, very roughly speaking, two-thirds of a set of

The Parliament plays a secondary role in the legislative

process. Before the Maastricht Treaty, its basic formal power was

that of presenting to the Council a nonbinding opinion on

force the Council to act unanimously, instead of by qualified

majority, by disapproving a Council action. And sometimes the

Parliament can ultimately threaten to veto a Council action and

thereby insist upon Council/Parliament mediation to reconcile

9 The Commission retains considerable control over the text of the legislation

because it may amend the proposal at any time, while the Council may amend the

10 The Maastricht Treaty expanded the subjects for which the Council can use

qualified voting. Some of the subjects now using qualified majority voting include

workplace safety, education, health, consumer protection, and most measures in

130S. Qualified majority voting has not been introduced in the most sensitive fields,

however, such as fiscal harmonization. Under the revised EC Treaty, unanimity is also

still applicable to most measures concerning competition; citizens’ freedom to move and

reside; social security for migrant workers; reform of structural funds; and budgetary

L.J. 2403, 2429-30 (1991).

12 More specifically, the Parliament is now involved in the enactment of legislation in

four different ways: consultation (providing a nonbinding opinion, just as before the

Maastricht Treaty), cooperation, co-decision, and assent. Cooperation requires the

Commission to submit a proposal to the Council, which then votes on it by a qualified

Parliament must then approve or vote, by an absolute majority, to amend or reject the

proposal; the Council may overturn this only by unanimity. If the Parliament does

nothing, the Council may adopt the proposal. Co-decision is similar but, if Parliament

rejects the proposal, the Council and 15 members of Parliament assemble a Conciliation

Committee, akin to our Conference Committees, and attempt to reconcile the differences.

189B. Assent is used for only a small number of subjects.

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

The European Court of Justice (“ECJ”) consists of fifteen

judges, one nominated by each member state. Each judge holds

office for a renewable six-year term. The ECJ interprets and

applies both treaty provisions and EC legislation. The Court hears

legal questions posed by national courts about the meaning or

state or an EC institution brings asking the Court, for example, to

review EC administrative actions, set aside EC laws, or force

The bulk of the ECJ’s work consists of preliminary rulings. In

1997, for example, the Court decided 456 cases. Of these, 301

were preliminary rulings, 116 were direct actions, and 34 were

appeals from the Court of First Instance.

The ECJ works in French, but a party may argue in any one of

the eleven official languages of the EC. Oral arguments are

simultaneously translated, and all judgments must appear in each

of the official languages. Thus, all eleven languages of the EC are

sit en banc in important cases, and will do so when requested by a

13 There is also a fourth type of case: a request for an “opinion,” in which a member

state or EC institution asks the ECJ to render its opinion on the compatibility of an EC

This procedure is rarely used; in 1997 the ECJ issued no “opinions.”

14 In preliminary rulings, which comprise the majority of the Court’s workload, the

Court rules on issues of EC law in cases pending before a national court. Most national

courts may, and a national court of last resort must, refer questions of EC law to the ECJ.

reduce the caseload of the ECJ and assist with fact-finding. At first, the court’s

jurisdiction was limited to hearing disputes between employees and the EC, but the EC

expanded the jurisdiction of the Court of First Instance in 1993 and 1994 to include all

actions brought by individuals and corporations. The Court of First Instance does not

have jurisdiction to hear questions referred from national courts for a preliminary ruling.

(1995). A party presents its pleadings in its own language, which must then be translated

into French. All internal documents are circulated in French, but every judgment is issued

in each of the 11 languages. The demands of translation limit the Court’s ability to process

cases quickly. About one-third of the Court’s staff is involved with translation and, of the

average two-year time for processing a case, six months is attributable to the translation

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

before the Court is initially considered by one of eight Advocates

General, who prepares a report and recommends a decision to the

Court. The Court also hears argument from the parties, and

possibly from affected member states as well. The ECJ’s decisions

are not signed; they are issued in the name of the Court, without

dissent.

The ECJ has established the legal supremacy of EC law and

its own legal decisions in several important ways. It has held that

some EC law, such as treaty provisions and EC legislation, is

automatically incorporated into national law without further

national legislation. This principle is known as “direct

applicability.” The Court has held that an individual can invoke

certain EC law—law that is sufficiently precise and unconditional

so as to confer legal rights upon that individual. This principle is

known as “direct effect.” (An individual might be able to assert

the EC law against governments, thus creating a “vertical direct

effect,” or against other private parties, creating a “horizontal

direct effect.”) The Court has made clear that it has the final word

as to which provisions are directly applicable and directly

The ECJ has also made clear that all EC law—including

treaty provisions, directives, regulations, Council and Commission

decisions, and its own rulings—prevails over conflicting national

law. This is the principle of “supremacy.” The Court has further

held that a national court may not examine the validity of EC law,

even when deciding whether a national constitution conflicts with

important EC law, such as a directive, violates a higher EC legal

Amministrazione delle Finanze dello Stato v. Simmenthal S.p.A., 1978 E.C.R. 629, [1978]

3 C.M.L.R. 263 (1978); Case 34/73, Fratelli Variola S.p.A. v. Amministrazione Italiana

delle Finanze, 1973 E.C.R. 981 (1973); Case 6/64, Flaminio Costa v. Ente Nazionale per

L’Energia Elettrica (ENEL), 1964 E.C.R. 585, 594, [1964] C.M.L.R. 425, 455 (1964); Case

26/62, N.V. Algemene Transport- en Expeditie Onderneming Van Gend en Loos v.

note 2, at 349-50.

Vorratsstelle für Getreide und Futtermittel, 1970 E.C.R. 1125, 1131, [1970] 9 C.M.L.R. 294

(1970).

The ECJ applies these principles expansively. For example, the ECJ confers on itself the

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

In sum, the ECJ has final legal authority to interpret an EC

imposes upon a national court that reaches a final decision in any

case an obligation to consult the ECJ about any disputed, relevant

issue of EC law. This requirement is the source of the numerous

preliminary rulings. Hence, one might believe, or at least plausibly

argue, that EC law, as interpreted by the ECJ, slowly but surely

will come to dominate national law in many areas of European

life.

Two days of meetings with the ECJ judges in Luxembourg,

and the observation of one oral argument, began to help us

understand how the Court functions. The procedures, obviously,

are somewhat different than our own. Yet the Court itself seemed

far more familiar than foreign. The judges work with a basic

“constitutive” document consisting of the treaties; they try to

reflect that document’s basic purposes in their decisions. They

recognize that their decisions will have important, widespread

practical effects, and they understand the need to enunciate legal

rules that are practically administrable. Thus, the process of

judicial decision making seemed not so different.

The ECJ also faces two problems which together suggest a

likelihood of further significant institutional change. The first

problem, which we can easily understand, is that of a growing

caseload. My own Court considers approximately 7,000 petitions

for certiorari each year. Yet, because review in our Court is

discretionary, we hear fewer than one hundred full cases. But the

ECJ does not have discretionary jurisdiction.

The ECJ now must consider about 450 cases annually,

although the Court does divide the cases among several different

panels. Translation problems slow the process to the point where

consideration of a case takes, on average, about two years. The

anticipated addition of four or five new member states will add to

the docket’s size and translation needs and, despite the additional

right to determine the criteria for establishing whether a provision is sufficiently clear to

1974 E.C.R. 1337, 1342, [1974] 13 C.M.L.R. 347, 358-59 (1974). Similarly, with supremacy,

the ECJ has determined that EC law takes precedence over member state law, regardless

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

judges, slow the process further. And, most importantly, this

rapidly expanding docket is likely to grow because the ECJ is the

sole authoritative interpreter of EC law, and the amount of EC

law may increase exponentially. The basic treaties, as I have

indicated, contain 573 articles that grant the EC considerable

legislative authority, and the EC’s legislative process will create

ever more law. These factors, combined with the “national-courtto-

ECJ” mandatory referral process, make it uncertain whether

the ECJ can avoid a commensurate increase in the size of its

docket.

Consider some of the questions related to possible change:

Will the EU grant the ECJ discretionary jurisdiction? Will the EU

create a set of lower courts, akin to our federal courts? Will the

ECJ itself change its mind about its exclusive authority to interpret

EC law, perhaps coming to rely upon national courts to interpret

EC law in the first instance—thereby making it procedurally easier

to limit full ECJ review to questions where the “lower” courts

“split”? If so, will the EC develop a substantive counterpart to our

“federalism,” finding a kind of state/federal parity in the

enforcement of “federal/EC” rights between national and EC

answers and its approach may prove both interesting and relevant

to us, as we debate the White Commission’s report and suggestions

about the general problem of congestion in our own federal courts

of appeals.

A second problem, which is more difficult for us to

understand, concerns the relationship between the ECJ’s decisions

and the EC’s legislative process. That legislative process, as I

described above, gives the European Parliament only a limited

role in enacting legislation. The procedure specifies that the

Commission can propose, but not enact, legislation. It makes clear

that the Council, which can legislate, must act sometimes with

unanimity and sometimes through a weighted majority. And, of

course, the EC itself can revise the treaties only by acting with

unanimity. The need to obtain unanimity, or even a weighted

majority, means that it is difficult for member states or other EC

institutions to revise ECJ decisions with which they disagree.

Compare that difficulty with the difficulty facing those

persons who disagree with a decision of the United States

Supreme Court. If the decision is based upon the Constitution, it

is very difficult to overturn. But if the decision rests upon

statutory or administrative grounds, revision requires only a new

24 401 U.S. 37 (1971).

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

statute or reconsideration by the relevant agency. In contrast, the

ECJ’s decisions are more difficult to revise; furthermore, in light of

the length and the detailed nature of the ECJ’s “constitution”

(namely, the basic treaties), many more ECJ decisions will likely

rest upon “constitutional” grounds.

This institutional relation suggests several questions. For

example, is the relation seen as a “democratic” or an institutional

problem? Why? Or, why not? Has it led national courts to seek

legal checks upon the ECJ’s decision-making authority? Has it

affected the attitude with which the ECJ’s judges approach their

interpretive task? Is the relationship “stable,” or is it likely to

provoke further institutional change?

Discussion with several members of the European

Commission in Brussels suggested that the problem just

mentioned—that of “legislative revision”—has no easy answer. Of

course, the easier it is for the Council to legislate, the easier it is to

change an ECJ decision, based upon “legislation,” with which the

citizens of member states disagree. And many Europeans argue

for granting the Council greater authority to act by weighted

majorities, or by what could amount to a majority of the EC’s

citizens.

Yet there are strong arguments to the contrary. Some feel

that the Council itself operates at too great a distance from those

citizens who elect the governments that then, in turn, send

representatives to sit as legislators on the Council. Others may

believe that the EC, if left free to legislate by simple majorities,

would act in highly controversial areas—affirmative action, for

example—in ways that majorities of citizens in particular countries

would strongly oppose, regardless of opinion in other nations. Still

others may feel that, in general, member states should retain a

check on the legislative and administrative actions of EU

institutions, lest those institutions become too powerful.

Of course, an American might expect the EC to expand the

European Parliament’s authority to create legislation. Doing so

would create a democratically elected legislative body analogous

to our Congress, and would thereby eliminate what some

Europeans describe as the EC’s “democratic deficit.” But there

are obstacles. A European “Congress” would have to work in

several languages. To create political responsibility—permitting a

meaningful democratic check—this “Congress” would have to

develop cross-boundary political alliances, perhaps leading to

Europe-wide political parties. And it would have to produce

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

institutional systems and understandings within the Parliament

that would lead the member states to entrust it with responsibility

for implementing, through legislation, treaties that at least

arguably delegate legislative power as broadly as does our own

Constitution to Congress.

The impression we received from most of those we met is that

the member states, at present, will focus upon problems of

common currency and member state expansion. They will not

likely increase the lawmaking authority of the Parliament or the

Council, lest EC institutions become too “federal” in nature—i.e.,

too powerful. Indeed, the most popular institutional change seems

to be that embodied in the Maastricht Treaty’s principle of

“subsidiarity,” a principle that would permit the EC to legislate

only as to matters that cannot be better dealt with at a local (e.g.,

enforceable—would give the Court still more “unchecked”

authority.

This creates a dilemma. The EC’s member states have strong

reasons to resist granting the Council or the Parliament added

majoritarian lawmaking authority. But that very fact means that

those same EC institutions find it difficult to enact legislation that

would overturn an ECJ decision. In that sense, the effort to limit

the EC’s authority and to maintain that of member states (often

justified in terms of maintaining member state “democracy”)

ironically provides greater authority to the least democratic EC

institution—namely, the ECJ—to determine the ultimate shape of

the law.

At the same time, the ECJ’s broad, revision-free legal

authority seemed widely accepted. Why? Have the judges

exercised that authority with particular circumspection? Have

they avoided seriously unpopular opinions? Are they particularly

willing to revisit and to revise their own earlier opinions, in the

way that our Court seems more ready to revisit constitutional,

rather than statutory, issues? Are the current institutional

relationships with respect to revision stable ones? Answering

these questions might cast light on our own attitudes towards

decisions by an unelected judiciary.

In Germany we discussed with members of the German

26 The court is also referred to as the Bundesverfassungsgericht, or the Federal

Constitutional Court.

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

by any national court to the ECJ’s supremacy with respect to EC

law. That challenge arises out of two German constitutional cases.

Constitutional Court held that, so far, EC law seemed to protect

fundamental liberties less fully than did Germany’s own

constitution; for that reason, German law in this area would

court held that EC law had “caught up” with German protections,

not disavow the basic principle of the earlier case—namely that, in

case of conflict, German constitutional law would prevail over EC

decided that the German Constitution’s guarantee of the right to

vote did not conflict with the German government’s decision to

ratify the Treaty of Maastricht. In doing so, the German court

again suggested that it, not the ECJ, was entitled to the last word

about which law—the German Constitution or EC law—would

prevail.

argued that the treaty had delegated so much power to EC

institutions that the German citizens’ right to vote for members of

the now less powerful German legislature had become significantly

less meaningful, to the point where the delegation was

unconstitutional. In rejecting this argument on the ground that the

German legislature continued to exercise meaningful power, the

German Constitutional Court did not simply point to the many

powers still exercised by the German legislature. Rather, it

acknowledged that the German voter would “lose some . . .

influence,” but held that voters would retain enough “influence”—

in part because the treaty “elevated the principle of subsidiarity to

which means “so long as.”

31 33 I.L.M. at 418, 424.

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

to mean that the German court, not the ECJ, has the final word as

to the meaning of much EC law. Suppose the ECJ, for example,

held that a German trade rule hindered trade, violating article 30

reconsider whether the ECJ’s interpretation was proper, or

improper because it went beyond the “true” meaning of those two

articles.

would never hold any such thing; instead, the German court would

always refer a matter of interpretation to the ECJ. In doing so, the

German court could indicate its concerns, and the ECJ itself would

likely respond to those concerns with an interpretation aimed at

avoiding conflict. This kind of approach, however, is judgespecific;

it may, or may not, continue to exist over time.

The result is a degree of uncertainty. Will the German

decisions eventually deprive the ECJ of power analogous to that

which Justice Holmes thought so essential in our own case—the

final authority in a supreme court to hold a state’s law

unconstitutional? Does the absence of any strong EC legislative

check (e.g., ability to revise ECJ decisions) make this problem

different from the problems Holmes foresaw? In the absence of

EC legislative authority, has the German court seen a need to

reserve a check, even if only a theoretical check unlikely to be

exercised in practice?

In Strasbourg, we visited the European Court of Human

Rights (“ECHR”), a court with Europe-wide jurisdiction that has

no direct relation to the European Union. The ECHR is the

creation of a treaty that, in 1949, established the Council of

Europe, now an organization of forty European states. The

Council is comprised of a decision-making body (the Committee of

Ministers) and a deliberative body (the Parliamentary Assembly).

The Council, in a sense, uses the ECHR to enforce the European

34 The Council of Europe authored the European Convention on Human Rights in

1950. The Convention contains 17 articles guaranteeing many rights, some of which

and Fundamental Freedoms, Nov. 4, 1950, 213 U.N.T.S. 221. For example, the rights in

the Convention include the right to life; the right to be free from torture; the right to

liberty and security of person; the right to freedom of thought, conscience, and religion;

the right to freedom of expression; the right to peaceful assembly and association with

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

same forty state signatories. Roughly one-half of these states

automatically enforce the treaty’s human rights provisions, and the

Court’s decisions interpreting those provisions, as part of their

state domestic law.

The Court that we visited is in the process of major

institutional change. Until recently, it effectively consisted of two

separate institutions. The first, called the European Commission

of Human Rights, consisted of one representative from each of the

Convention’s signatories, selected for a term of six years. The

Commission, directly or through staff, would review any petition

sent by any citizen of any member state claiming a human rights

violation. The Commission could dismiss the claim as meritless, or

it could investigate the claim. Investigation, particularly in states

without strong human rights records, might uncover serious

abuses. After investigation, the Commission would try to reach a

friendly settlement. If mediation failed, it could decide that the

state had violated the Convention and refer the matter to the

second institution, the Council, to review the decision and issue a

final judgment. Alternatively, the state could appeal to the court,

consisting of one judge from each member state, nominated for a

term of nine years. The Court would sit in “panels” of nine judges.

It would hear cases brought by a member state or by the

Commission and would reach a decision binding upon the member

state.

In the new, post-Communist world, membership in the

Council has expanded to forty. And, recently, the structure I have

just described has changed radically. The Commission has been

abolished. The Court now has forty members—one from each

nation. The judges are selected by the Council’s Parliamentary

Assembly from a list of three nominees sent by each state. Each

judge serves for a term of six years and judges sit in panels of three

or in an en banc panel of seven judges, with the possibility of an

internal referral to a panel of seventeen judges. The Committee of

Ministers of the Council no longer plays a decision-making role.

The effects of these changes are uncertain. Will the abolition

others, including the right to join trade unions; and the right to freedom from

discrimination. The Convention also provides that “[e]veryone whose rights and freedoms

as set forth in this Convention are violated shall have an effective remedy before a

national authority notwithstanding that the violation has been committed by persons

member state’s national laws, but article 1 requires each member state to “secure to

everyone within their jurisdiction the rights and freedoms defined in . . . this Convention.”

Convention into their national laws, thereby allowing an individual to invoke the rights in

national courts. The other member states give effect to the judgments of the Council.

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

of the Commission, with its investigation and mediation functions,

make enforcement of the Convention more difficult? What are

the consequences of changing the Court’s agenda from a

“structured” agenda, determined by the decisions-to-appeal of

member states, to an “unstructured” agenda, determined by

individual petitioners or by the reviewing judges themselves? In

light of our own system’s generation of more than 20,000 habeas

could be a vast inflow of petitions—for example, from ordinary

prisoners in nations without a history of important procedural

protections in criminal cases? How will the Court assimilate the

many new judges who will be appointed by some of those new

member states?

After discussions with members of several French courts, we

began to understand some of the special ways in which the

development of EC and ECHR law can affect a nation’s own

preexisting system of constitutional law. The French system is

complex. The Cour de cassation, a group of about ninety judges

and forty assistant judges, acts as a final court of appeal in ordinary

civil and criminal cases. It interprets and applies French statutes,

but it does not have the legal power to set aside a statute as

At the same time, members of a different body, the 260-

member Conseil d’Etat, act both as high level administrators and

as members of a court with jurisdiction over administrative

proceedings. The Conseil d’Etat interprets nonstatutory

(administrative) rules and regulations; it determines whether

administrative actions are consistent with law; and it can set aside

STATISTICS DIVISION, ADMINISTRATIVE OFFICE OF THE U.S. COURTS, 1997 REPORT OF

THE DIRECTOR 132.

36 More specifically, the Cour de cassation has jurisdiction to review all decisions

from lower civil and criminal courts. Like our Supreme Court, the Cour de cassation

decides only questions of law and application of law to facts. The Cour de cassation has

ultimate legal authority to interpret French statutes. The judges and assistant judges sit in

six different chambers divided by subject matter (five civil and one criminal).

37 The Conseil d’Etat is the court of last resort for administrative decisions. The

Prime Minister is the president of the body, which has 260 members. The Conseil d’Etat

has ultimate authority to interpret nonstatutory rules and regulations, and to determine

whether administrative actions are consistent with the law. The Conseil d’Etat may

examine the constitutionality of administrative acts, but it may not consider the

constitutionality of acts passed by the Parliament. The Conseil d’Etat also advises the

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

A third body, the Conseil constitutionnel, considers whether

or not a statute is consistent with the constitution. But it may do

become law. And it must decide about a proposal’s

constitutionality within a few weeks after a group of legislators has

Traditionally, the result has been that no one may challenge

as unconstitutional the French national legislature’s statutory

enactments. The Conseil d’Etat’s legal power to set aside state

actions as unconstitutional applies only to administrative acts, not

to statutes. The Cour de cassation has authority only to interpret

statutes, not to determine whether they are unconstitutional. And

the Conseil constitutionnel can consider only statutory proposals

prior to their enactment, not statutes once enacted and applied.

The establishment of European courts, and the growth of

European law, may begin to undermine this traditional French

system. That is because French judges recognize that they must

apply European law and set aside French statutes insofar as they

conflict with that law. At the same time, the content of the French

Constitution and that of basic European documents will

sometimes overlap. The French Constitution of 1958, for example,

expressly incorporates statements of certain basic human rights,

such as the right of “free communication of thoughts and

French judges, while unable to set aside a French statute that they

believe conflicts with a right guaranteed by the French

Constitution, might set aside that same statute on the ground that

it conflicts with a European document that says approximately the

same thing.

Is this “back door” kind of constitutional judicial review itself

38 The Conseil constitutionnel has limited authority and is in essence an appendage to

the legislature. It determines the constitutionality of proposed, but not enacted,

Citizens may not petition the Conseil constitutionnel; only the President of the Republic,

the Prime Minister, the President of the Senate or National Assembly, or a group of 60

its decisions quickly: generally one month before promulgation or ratification, or eight

39 The French Constitution of 1958 expressly incorporates the 1789 Declaration of the

Rights of Man and the Citizen as reaffirmed and complemented by the Preamble of the

Convention; for example, the French Constitution guarantees each man and woman “free

communication of thoughts and opinions” and the right to “join the union of his [or her]

FR. CONST. of 1946 pmbl.

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

likely to provoke further institutional change within the French

system? It means, after all, that French judges considering the

“constitutionality” of a statute are forced to look only to what

European courts might say about the meaning of the relevant

European treaty. They cannot themselves measure the statute

against the French Constitution, nor can they interpret the

constitution in a way that might save the statute, prove more

consistent with French traditions than the European Court’s

interpretation, and prevent conflict between the French

Constitution and European law. In other words, will it begin to

seem anomalous to have French judges apply only European

“constitutional” law, but not their own? And, if so, will we see the

“constitutional” power of French courts, such as the Conseil

constitutionnel, increase over time?

The questions that I have raised have grown out of a brief

European visit. They concern the emerging structure of a

European constitutional system. They suggest that some of the

major changes that are now taking place will continue to take

place among major constitutional institutions. I mention them for

a special reason. I would suggest that, while the more obvious

comparative study of substantive constitutional law (“free speech”

law, for example) is important, such substantive law is not the only

kind worth serious examination. One must look as well at the

comparative aspect of the structural, or governance-related,

characteristics of constitutional courts.

The foreign environment in which such questions arise is not

always quite so “foreign” as one might think. After all, the nations

we visited, like our own, attach great importance to the rule of law,

representative democracy, the protection of basic human liberty,

the division of power to avoid its accretion in too few hands, and

equality in the sense of requiring some justification for

governmentally-imposed distinctions. And the nations we visited,

like our own, have come to believe that independent judicial

systems can help secure these basic values for their citizens. That

being so, we can learn from their structural efforts, just as they

may learn from ours. How much does it matter, for example, that

a statute like our antitrust or civil rights laws, rather than a

constitution, embodies a basic principle, if that statute is difficult

to change? And evaluations of different judicial approaches to the

“democratically irreversible” decision, or of the potential of the

ECHR as compared to our own habeas corpus jurisprudence, or of

different approaches to problems of “federalism,” can help.

BREYER WEBGALLEYS.DOC 04/03/00 1:20 PM

In short, the academic job of investigating, learning,

comparing, systematically explaining, and evaluating complements

the tasks of those who work in government, including judges. The

work of both is required to build legal systems that work better for

those who live within them. I am certain that comparative

constitutional study of substance and of structure is part of that

task.