The Separation of Powers in A Post-Apartheid South Africa

American University International Law Review
Volume 8
Article 11
Issue 2 Vol 8. #2/3 Winter/Spring 92/93
1993
The Separation of Powers in a Post-Apartheid

South Africa
Stephen Ellmann
Follow this and additional works at: http://digitalcommons.wcl.american.edu/auilr

Part of the Law Commons
Recommended Citation
Ellmann, Stephen. "The Separation of Powers in a Post-Apartheid South Africa." American University International Law Review 8 no.
2/3 (1993): 455-482.
This Symposium or Conference is brought to you for free and open access by the Washington College of Law Journals & Law Reviews at Digital
Commons @ American University Washington College of Law. It has been accepted for inclusion in American University International Law Review by
an authorized administrator of Digital Commons @ American University Washington College of Law. For more information, please contact
[email protected].
THE SEPARATION OF POWERS IN
A POST-APARTHEID SOUTH AFRICA
Stephen Ellmann"
My task this afternoon is a large one: to describe the lessons of the

United States' experience with the separation of powers and suggest
what relevance those lessons might have to the shaping of a South
African government. This is not a project that will produce a short,
simple set of answers. The fact is that scholars and citizens in this
country continue to debate the nature and the value of our system of
separation of powers, and so many of the lessons one observer may
draw will be considered errors by another.' More importantly, our con-
stitutional structure works, or does not work, as part of our nation-a
particular country with a particular history and culture, none of it identi-
cal by any means to South Africa's. Thus, the question of what you can
learn from us is inevitably a contestable one. Finally, just as we debate
the value of our own system of government, and disagree in part be-
cause we have differing political perspectives on what we want our
government to do, so the lessons you may draw from our experience
will no doubt depend on the priorities you hold in designing your new
nation. Yet none of this should discourage us completely: the Framers
of the United States Constitution, like you, acted in a quasi-revolution-
ary context; acted with imperfect knowledge; and acted with good inten-
* Associate Professor of Law, New York Law School. I thank Philip Kurland,
Louis Henkin, Nancy Rosenbloom, and Peter Strauss for their comments on an earlier
draft of this paper.
1. Debate continues in this country over such issues as whether our existing
governmental arrangements separate powers too thoroughly, or not thoroughly enough;
whether the constitutional role of the president is being undermined by Congress, or
the other way around; whether the primary mission of the courts in adjudicating sepa-
ration of powers cases should be to adjust our venerable constitution to modern reali-
ties, or instead to rescue the insights of the founders from the distortions of modern
governmental maneuvering; and whether the courts should intervene in separation of
powers cases at all, or should rather leave the political branches of the government to
resolve their conflicts through political struggle. The resolution of these American
debates is not my concern here, but I will explore certain elements of these ongoing
arguments in the course of this essay.
456 AM. U.J. INT'L L & POL'Y [VOL. 8:455
tions, sober reasoning, political priorities and even prejudices. Nonethe-

less, what they wrought has lasted 200 years. I wish you equal good
fortune, and I am very pleased to have the opportunity to assist you
with my perception of the lessons of the American experience.
Let me begin with a lesson that may scarcely need repeating: struc-
ture is important. Those who framed our Constitution believed that lib-
erty could be adequately guaranteed against the possible depredations of
the federal government by structure alone-by specifying the
government's powers and properly shaping the processes by which it
could exercise them.' To that end, they divided the powers of govern-
ment among the legislative, executive, and judicial branches. But, in
certain respects, they also blended these powers together; the result can
be seen in one light as a "separation of powers," and in another as a
system of "checks and balances" whereby each branch of the govern-
ment constrains, and is constrained by, the others? Many citizens of the
2. See DANIEL A. FARBER AND SUZANNA SHERRY, A HISTORY OF THE AMER-

ICAN CONSTrrUTION 221-22 (1990) (noting that the delegates at the constitutional
convention specifically rejected a proposal to create a committee to prepare a bill of
rights). In addition, the constitution, as submitted by the convention to the nation,
contained only scattered provisions that directly protected individual rights.
3. As is well known, the separation of powers allocates legislative authority to
Congress, executive responsibility to the President, and the judicial power of the Unit-
ed States to the federal courts. U.S. CONST. art. I, § 1 (legislative power); U.S.
CONST. art. II, § 1 (executive power); U.S. CONST. art. III, § 1 (judicial power).
In many important respects the Constitution expressly blends powers rather than
prying them apart. See U.S. CONST. art. I, § 7 (outlining the president's power to
veto legislation, a grant of a form of legislative power to the executive branch); U.S.
CONST. art. II, § 2 (establishing the Senate's power of "advice and consent" with re-
spect to the president's appointments of other executive officials, arguably an instance
of executive power in the bands of one house of the legislature); U.S. CONST. art.
M, § 2 (specifying Congress' power to make exceptions from the Supreme Court's
appellate jurisdiction, an authorization for Congress to limit, though not itself to
wield, the judicial power).
To James Madison, a particularly influential participant in the framing of our
Constitution and in its early years of operation, the strategies of separating powers
and blending them together were not opposed but mutually supporting. As he wrote,
the maxim of separation of powers:
does not require that the legislative, executive, and judiciary departments should
be wholly unconnected with each other. I shall undertake . . . to show that
unless these departments be so far connected and blended as to give to each a
constitutional control over the others, the degree of separation which the maxim
requires, as essential to a free government, can never in practice be duly main-
tained.
THE FEDERALIST No. 48, at 308 (James Madison) (Clinton Rossiter ed. 1961).
19931 POST-APARTHEID SOUTH AFRICA 457
new United States lacked the Framers' confidence that this system by
itself would preserve their liberties, and their insistence on a Bill of
Rights led to its speedy adoption in the first years of the Constitution's
life. But the principles of the separation of powers, and of checks and
balances, have remained important themes of American constitutionalism.
It has never been our constitutional strategy to protect liberty simply by
relying on a Bill of Rights and an independent judiciary to block the
wrongdoing of politicians.
Our experience offers good reason to believe in the importance of
judicial enforcement of a Bill of Rights, but it confirms the Framers'
belief that structural restraints are also needed. Even in a constitutional
state, lawmakers have tremendous discretion. Our Bill of Rights is large-
ly a negative one, forbidding certain government actions while requiring
few and authorizing many-but the point would be essentially the same
even if South Africa were to adopt the kinds of affirmative socioeco-
nomic rights that the ANC's draft bill of rights proposes. How the gov-
ernment acts within the immense zone of the "permissible but not re-
quired" is extremely important to human liberty broadly understood.
Lawmakers who escape the perils of judicial statutory interpretation
(roughly as potent a source of judicial authority in this country as in
yours) are free to act in this zone without constitutional constraint. In
this country, for example, Congress is largely free to decide that welfare
benefits will not be increased in proportion to the increasing number of
children in a family receiving those benefits-or to decide the opposite.'
What Congress decides makes a great difference to the families receiv-
ing those benefits, and perhaps to other people as well, but the Constitu-
tion as currently read will not normally prescribe an answer.s
In addition, the dimensions of what is or is not unconstitutional will
not be subject to wholly apolitical adjudication. The decisions of the
United States Supreme Court, or of South Africa's Appellate Division,
offer ample evidence of the breadth of the judges' discretion and the
degree to which the judges' perspectives on their world affect their deci-
sions.' This is not to disparage the ideal of an impartial judiciary, but
4. See Dandridge v. Williams, 397 U.S. 471, 482 (1970) (holding that a state
welfare regulation limiting the maximum family grant does not violate the Equal
Protection Clause of the Fourteenth Amendment, so long as it is "rationally based and
free from invidious discrimination"). A comparable standard would apply to federal
action, which would be evaluated in this respect under the Fifth Amendment.
5. Ma
6. See STEPBEN ELLAtkNN, IN A TiLE OF TROUBLE: LAW AND LBERrY' iN
SOUTH AFRICA'S STATE OF EMERGENCY (1992) (discussing the influences on the
to recognize the limits on its attainability. It is difficult even to imagine

a judge so "impartial" as to have no views on fundamental issues affect-
ig his or her society. Moreover, we must recognize limits on
impartiality's desirability as well. It is possible to imagine a judge who
profoundly disagrees with the basic premises of the state in which he or
she serves, but those who believe their state is a decent one will not be
likely to select its adversaries, however talented, as its judges. (The
Weimar republic in Germany provides a grim example of the conse-
quences of ignoring this rule.)7 Even if South Africa adheres to a sub-
stantially depoliticized system of judicial appointments, judges' own
politics and perspectives will matter. In addition, the more the system of
appointments becomes responsive to political criteria, as has happened
widely in this country and to a significant extent in South Africa, the
more the meaning of the Constitution will tend to resemble the prefer-
ences of the party holding the appointment power.
What our Constitution prescribes to guide the resolution of questions
not governed by the Bill of Rights is a structure. The structure is not by
any means completely spelled out in the Constitution, and it has
changed over time. Nonetheless, I believe the system shaped by the
Framers continues to profoundly influence our politics and our govern-
mental decisions. As a result, perhaps unlike my panel colleague Pro-
fessor Kurland, I believe the separation of powers continues to play a
role in protecting liberty in America.8 It is not a precision tool for this
South African Appellate Division's decisions in cases dealing with the state of emer-
gency).
7. Ingo Miller, in his fierce history of German law under the Nazis, notes that
the judges of the German Empire "whose conscience would not permit them to serve
the republic instead of the kaiser were offered early retirement by the government,
with full consideration of their material needs. Less than 0.15 percent of judges took
advantage of this opportunity, however." INGO MOLLER, HITLER'S JUSTICE: TIlE
COURTS OF THE THIRD REICH 10 (Deborah Lucas Schneider trans. 1991). Even
though the remaining judges swore their loyalty to the nation, Mifller maintains that
the judges "kept their distance from the republic and continued to take what remained
of the old values as their point of orientation." Id Miller also argues that as the
Nazis rose to prominence during the Weimar republic years, "[i]n many trials...
the courts continued to take the Nazi side of the ongoing political struggle, sometimes
openly, sometimes behind the fagade of legal maneuvers." Id. at 17. See also David
A.J. Richards, Terror and the Law, 5 HuM. RTs. Q. 171, 179 (1983) (discussing the
"authoritarian bias" of the Weimar judiciary, including its opposition to "legal action
against . . . the Nazi authoritarian terror of the right").
8. See Philip B. Kurland, The Rise and Fall of the "Doctrine" of Separation of
Powers, 85 MICH. L. REv. 592, 604-13 (1986). Kurland argues that "[w]ith the
1993] POST-APARTHEID SOUTH AFRICA 459
purpose, however, for it does not prevent bad governmental action by

identifying it as such, but by impeding all governmental action. In the
process it undoubtedly also impedes some desirable governmental action
that might enhance liberty. But with all its imprecisions and costs, the
separation of powers is still a supplement to the Bill of Rights as a
limit on the danger of tyranny.
What I have just said is an argument for imposing limits on power
that go beyond those generated by judicial enforcement of a bill of
rights. The next lesson, however, is that power can be too limited, or
too compromised. Our history provides painful examples of the conse-
quences of unwise restraints on governmental authority. The Constitution
itself is in large measure a product of the failure-or so the Framers
saw it-of the Articles of Confederation, under which our nation was
governed for the first years of its existence9 The Articles bound the
former colonies, now states, together, but by no means very tightly. It
was difficult for the national government to act, for it had very little
coercive authority over its constituent states,'" and the Articles estab-
lished no separate executive branch." The Constitution we have was
the result of the conviction that the Articles did not work.
But the Constitution we have did not work either, and in a fundamen-
tal way: it foundered on the issue of slavery. The Constitution was a
growth of government, the safeguard thought to be inherent in separation of powers

has largely failed." Id. at 612.
9. See Farber & Sherry, supra note 2, at 23-26.
10. Id, The Articles also specifically deprived the national government ("the Unit-
ed States in Congress assembled") of power to make a wide range of critically im-
portant decisions, from engaging in war to borrowing or appropriating money, without
the consent of nine of the thirteen states. ARTICI.E OF CO.NFEERATION, art. IX.
11. The Articles envisaged government very much under the control of the legis-
lature. The Continental Congress, or the "United States in Congress assembled," had
not only legislative but also some judicial authority. It It also had executive authori-
ty;, the Articles empowered it to establish a "Committee of the States," consisting of
one delegate from each state, with power "to appoint such other committees and civil
officers as may be necessary for managing the general affairs of the United States
under their direction." I&
While the Articles did not themselves establish an executive branch, the Conti-
nental Congress did take steps in this direction. Geoffrey Miller writes that "the pat-
tern was one of movement from congressional towards executive government. Con-
gress, although loath to cede any of its powers to an executive, was compelled by
the exigencies of war and the needs of government to do so." Geoffrey Miller, Inde-
pendent Agencies, 1986 SUP. Cr. REv. 41, 69. It seems very likely, as Miller sug-
gests, that this pattern "inform[ed]" the Framers' subsequent deliberations on the con-
stitution we now have. See id. at 68.
compromise between the slave states and the free, or freer, states, and
this compromise proved untenable." The nation could not hold together
while the institution of slavery divided it. The resolution of this issue
led the nation through secession and civil war to a revised Constitution
which declared the abolition of slavery, commanded the states no longer
to deny to their people, black and white alike, the equal protection of
the laws, and laid the groundwork for the federal government's enforce-
ment of these principles. 3 I do not mention this piece of history in
order to attribute our Civil War to particular defects in the allocation of
power, though I suspect the elements of our separation of powers did
play their part in shaping the politics of the antebellum years. Rather,
my point is a larger one: a society with intolerable injustices enshrined
or protected by its constitution will not survive. Put more bluntly, unless
a new South African government has the constitutional power to right
the wrongs of apartheid, there will be little reason for optimism about
South African stability or freedom.' 4
These are very broad lessons indeed: that checks on power besides a
bill of rights are needed, and that ineffectual or unjust government must
be avoided. But they do have concrete relevance to South Africa. They
counsel against a system of government in which the only constraints on
majority will are those of the bill of rights as interpreted by the courts.
They also counsel against a government so hamstrung by the need for
consensus among people of sharply different views that it is unable to
respond to the problems South Africa faces. Put more concretely still,
they counsel against both pure majoritarianism and thoroughgoing
12. See DON FEHRENBACHER, THE DRED ScoTr CASE: ITS SIGNIFICANCE IN
AMERICAN LAW AND POLITIcS 26-27 (1978) (noting that though the Constitution
never uses the word "slavery," it nonetheless plainly "acknowledg[es] the legitimate
presence of slavery in American life"). Perhaps the most notorious of its euphemistic
acknowledgements was the "three-fifths clause," which provided for the apportionment
of congressional representation and of direct taxes among the states based on figures
determined by the states' free population and "three fifths of all other Persons." U.S.
CoNsT. art. I, § 2.
13. These guarantees are the promise of the Thirteenth and Fourteenth Amend-
ments.
14. Herman Cohen, then Assistant Secretary of State, made a similar point when
he reported to a congressional subcommittee that a future South African constitution
"'should avoid overly complex arrangements intended to guarantee a share of power
to particular groups which will frustrate effective governance. Minorities have the right
to safeguards. . . . They cannot expect a veto.'" Jim Lobe, South Africa: U.S.
Against White Veto in New Government, INTER PREss SERv., July 23, 1992, available
in LEXIS, Nexis Library, Wires File.
consociationalism. What they counsel in favor of, broadly, is splitting

the difference, so as to enable the government to govern but prevent it
from riding roughshod over opposing views and opposing citizens. How
to split this difference remains a matter of debate, here and elsewhere,
and this is a debate to which I hope South Africa will contribute its
own answers. But the American answer, which I believe contains some
part of the truth on this score, lies in good measure in our system of
separation of powers and checks and balances. Let us look more closely
at what American experience argues for and against, by examining four
more specific points: first, the structure of the legislature; second, the
separation of the executive and legislative branches; third, the structure
of the executive branch; and, fourth, the challenges of preserving in
practice whatever system you agree upon on paper.
1. The Structure of the Legislature: Today the phrase "separation of
powers" calls to mind for Americans the divisions between the three
branches of our government, the executive, legislative, and judicial
branches. But the separation of powers can, and does, operate in the
United States not only between, but also within, branches. Because
South Africans are now considering the possibility of a bicameral legis-
lature, a legislature within which power is in some measure separated or
divided between the two houses, it is important to look at American
experience on this score. I want to focus here on two aspects of con-
gressional structure. The first is bicameralism per se; the second is the
system of allocating representation in each House. The Framers of the
United States Constitution chose bicameralism, and decided as well to
give the states equal representation in the Senate, while making repre-
sentation in the House proportional to population. But there is no intrin-
sic link between the choice of bicameralism and the decision to select
one house of the legislature in a way that is not based on population.
Indeed, every American state has its own legislature, and all but one of
these are bicameral, but in these bicameral state legislatures today both
houses are selected by methods that make representation proportionate to
population."
15. See Reynolds v. Sims, 377 U.S. 533, 568 (1964) (holding "that, as a basic
constitutional standard, the [Fourteenth Amendment's] Equal Protection Clause requires
that the seats in both houses of a bicameral state legislature must be apportioned on
a population basis"). That representation in bicameral legislatures is proportional to
population is probably more the result of the series of Supreme Court reapportionment
decisions beginning with Reynolds than it is the product of decisions initiated by the
states themselves. Nonetheless, the judicial imposition of a duty of equal representa-
tion has not triggered any movement among the states to abandon bicameralism.
The National Party has endorsed bicameralism, while the ANC is

plainly open to the idea." There are several reasons for adopting a bi-
cameral system. First, it can encourage greater reflection on governmen-
tal action."' The sheer fact that two houses, rather than one, must con-
sider each piece of legislation can generate greater reflection. Second, it
can modestly increase the barriers to governmental action of any sort. If
the consent of both houses is required for legislation to be enacted (as it
is in the United States)," then the possibility always exists that for
whatever reason a majority in one house will not be accompanied by a
majority in the other. Third, it can give citizens greater access to the
lawmaking process. This is not simply, or even primarily, because it
multiplies the number of legislators and so the number of interactions
that citizens can have with their legislators. Citizen access is also en-
hanced if the electoral sceme assures that the members of the two hous-
es will hear public opinions most piercingly at different moments; in the
United States Congress, for example, representatives are elected every
two years while senators, one third of them at a time, are elected for
six-year terms. If, as is also the case in the United States Congress,
senators are elected state-wide while representatives are elected from
districts within the states, the result is that the two houses will reflect
different bodies of voters as well. The chance that an individual voter's
opinions will be taken seriously goes up when that voter gets to vote
for more than one legislator, and does so at more than one time and
16. See CONSTITUTIONAL RULE IN A PARTICIPATORY DEMOCRACY: THE NATIONAL

PARTY'S FRAMEWORK FOR A NEW DEMOCRATIC SOUTH AFRICA 11-12 (Sept. 4, 1991)
[hereinafter cited as NATIONAL PARTY FRAMEWORK]; AFRICAN NATIONAL CONGRESS
CONSTITUTIONAL COMMITrEE, DISCUSSION DOCUMENT: CONSTITUTIONAL PRINCIPLES
AND STRUCTURES FOR A DEMOCRATIC SOUTH AFRICA 4, 22-23 (Apr. 8, 1991) [here-
inafter cited as ANC PRINCIPLES].
17. See Immigration and Naturalization Service v. Chadha, 462 U.S. 919, 951
(1983) (emphasizing that "[t]he division of the Congress into two distinctive bodies
assures that the legislative power would be exercised only after opportunity for full
study and debate in separate settings").
18. See id. at 951 (ruling that every exercise of "the legislative power of the
Federal government" must rest on bicameral approval of the legislation in question). It
is of course possible to establish a bicameral legislature in which one House alone
can enact legislation, while the other House can only insist on reconsideration or
delay. The British House of Lords today generally exercises only such subordinate
authority, although until as late as 1911 its power was much greater. See James G.
Wilson, Altered States: A Comparison of Separation of Powers in the United States
and in the United Kingdom, 18 HASTINGS CON. L. Q. 125, 135-36 (1990) (discussing
the legislative power of the House of Lords).
with more than one set of fellow voters.

These, it seems to me, are useful functions. Yet it might be argued
that the result of this multiplication of voices, and of opportunities for
reflection, delay, and dissensus, is to undermine the government's ability
to carry out its program. I do not want to ignore that point, but I want
to postpone it until I take up the issue of the separation of the legisla-
ture and the executive, a context in which this danger of hamstringing
the government is even more squarely posed.
Instead, let us turn from bicameralism in the abstract to the particular
bicameralism of the United States Constitution, in which one house, the
Senate, gives every state the same number of senators (two). As a con-
sequence, small states enjoy a level of representation wildly out of pro-
portion to their populations. This system is, undeniably, a form of mi-
nority protection. Half the population of the United States lives in the
nine largest states, represented by 18 senators out of the total of 100.'
Meanwhile, the seventeen smallest states, in which well under a tenth of
the population lives, are represented by 34 senators-mathematically
enough to control the outcome of any vote in which a two-thirds ma-
jority is required." On some issues, such as the ratification of treaties,
only the Senate and not the House of Representatives is entitled to
participate, and so the blow to the voice of the citizens of the larger
states is all the greater.' Yet this feature of our system does not gener-
ate public outcry, and although it surely does give disproportionate
power to some citizens as against others, it would be hard to say that as
a result the United States is not a democracy. There is at least some
room for unequal representation in democracy, for the protection of
relatively smaller interests-here, the smaller states-against the will of
the whole, and for the protection of subdivisions of the government-the
states, large or small-against the central, national government.
It is important, however, not to romanticize this system. It was not
inevitable. In fact, James Madison, sometimes called the Father of our
19. U.S. BUREAU OF THE CENsus, STATISTICAL ABSTRACT OF THE UNrrED

STATES 20 (1991).
20. lt A two-thirds majority in each House is required to override a Presidential
veto. U.S. CONST. art. I, § 7. In addition, a two-thirds majority in the Senate is
required to ratify a treaty. U.S. CONST. art. II, § 2.
21. U.S. CONST. art. ]I, § 2. The Senate is also the only House with authority to
block a Presidential appointment. Id.The Constitution assigns one function exclusively
to the House of Representatives, namely the authority to originate bills for raising
revenue. Id. art. I, § 7. The Senate, however, has the same power to amend or reject
such bills as it would with any other proposed legislation.
Constitution, vigorously opposed it, as did others at the Constitutional

Convention.' Madison even resisted compromise proposals, which
would have resulted in some, but less, disproportionality.3 The states
have two senators each because the delegates from the small states at
the Convention dug in their heels, and forced a compromise on this
point.2 ' Moreover, this compromise has had its costs. The extraordinary
power of Southern senators, who tended to serve many terms and accu-
mulate seniority, and who were deeply opposed to equal rights for
blacks, contributed to the slow response of the political branches of the
national government to the civil rights revolution of the 1950s and
thereafter. Equal representation for the states in the Senate is not integral
to bicameralism, nor to the structure of checks and balances as James
Madison first envisaged it. Rather, equal representation is the result of
sheer political conflict. In South Africa, the prices of compromise, for
any side, may be similar departures from cherished principles.
Finally, it is important not to confuse this system with the system
proposed by the National Party in September, 1991 in its pamphlet,
Constitutional Rule in a Participatory Democracy.' This proposal en-
visaged a bicameral parliament, in which each region would have equal
representation in the upper ("Second") house." But it also envisaged
that "[e]ach political party which . . . gained a specified amount of
22. See Farber & Sherry, supra note 2, at 112-34 (tracing and offering excerpts
from the Convention debates on this issue).
23. Farber & Sherry, supra note 2, at 119.
24. The framers' debates on this score depart dramatically from the image of
gentlemanly colloquy which Americans often take to have been the reality of the
Constitutional Convention. In particular, George Bedford, a delegate from the small
state of Delaware, declared that:
[The large states] insist that although the powers of the general government will
be increased, yet it will be for the good of the whole; and although the three
great States form nearly a majority of the people of America they never will
hurt or injure the lesser states. I do not, gentlemen, trust you. If you possess
the power, the abuse of it could not be checked; and what then would prevent
you from exercising it to our destruction?
Id.at 125 (quoting Bedford, as reported in the notes of fellow delegate Robert Yates)
(emphasis in original). Perhaps the lowest point was reached when Bedford moved
from distrust to threats:
The Large States dare not dissolve the Confederation. If they do the small ones
will find some foreign ally of more honor and good faith, who will take them
by the hand and do them justice.
Id.
25. NATIONAL PARTY FRAMEWORK, supra note 16.
26. NATIONAL PARTY FRAMEWORK, supra note 16, at 11.
support in the election in the region's legislative body will be allocated

an equal number of the seats for that region" in this House." This rule
does not empower states but rather parties within states. Depending on
where the threshold is set, the disproportionate empowerment of minori-
ty parties could be dramatic.' American government sets the winners
of elections in various jurisdictions, and at various moments, against
each other; it does not, characteristically, give seats to the losers. Our
government, as distrustful as it is of unchecked majority power, is much
less distrustful than the National Party's proposal, which is, I take it, not
modeled on American separation of powers but on the theories of
consociationalism. It is hard not to fear that this proposal would gener-
ate precisely the sort of minority veto that consociationalism endorses."
It is also difficult not to fear that such a veto will operate to thwart the
attainment of democratic rule in South Africa?
27. NATIONAL PARTY FRAMEWORK, supra note 16, at 11-12.

28. ARTHUR CHASKALSON, THE NATIONAL PARTY'S CONSTITUTIONAL PROPOSALS
2 (1991). Chaskalson, a member of the ANC's delegation in the constitutional nego-
tiations at the Convention for a Democratic South Africa (CODESA), points out that
the impact of this idea is compounded by the National Party's proposal that legis-
lation amending the constitution, relating to the interests of minorities or regions, or
dealing with matters entrenched in the constitution, would require a "weighted majori-
ty" in this House for approval. Id at 2-3; see also NATIONAL PARTY FRAmEWORK,
supra note 16, at 12.
29. AREND LUPHART, POWER-SHARING IN SOTH AFRICA 6 (1985) (identifying
"[tihe four basic elements of consociational democracy" as: "(1) Executive power-shar-
ing among the representatives of all significant groups; (2) A high degree of internal
autonomy for groups that wish to have it; (3) Proportional representation and propor-
tional allocation of civil service positions and public funds; and (4) A minority veto
on the most vital issues"). The National Party, however, disclaims any intention to
achieve a minority veto. See, e.g., British Broadcasting Corporation, Summary of
World Broadcasts, Official Transcript of Reimarks and Answers to Questions by Presi-
dent de Klerk, Aug. 5, 1992, available in LEXIS, Nexis library, Wires File.
30. Ideally, the minority veto works not as a cudgel with which to bludgeon the
majority but as a fallback device buttressing cooperation among the leaders of diverse
and potentially conflicting groups, cooperation on which consociationalism relies). See
Lijphart, supra note 29, at 8. But such cooperation is not inevitable, as Lijphart rec-
ognizes. IM. at 100-01. It may not even be likely. See DONALD L HoROWnrZ, A
DEMOCRATIC SOUTH AFRICA? CONSTITUIONAL ENGINEERING IN A DIVIDED SOCm'Y
139-43 (1991) [hereinafter HOROWnTZ, A DEiOCRATIC SOUTH AFRICA?]; DONALD L
HOROWIrZ, ETNIC GROUPS IN CONFLICT 571-76 (1985). In South Africa, arrange-
ments that appear to insulate white power and privilege seem particularly unlikely to
win popular support. See HOROWnT, A DEMOCRATIC SOtrH AFRICA? at 143 (citing a
survey in which only one percent of Black respondents approved of "special White
voting privileges").
466 AM. U.J. INT'L L. & POL'Y [VOL. 8:455
2. The Separation of the Executive and Legislative Branches: It is a

central claim of American constitutionalism that preventing tyranny
requires a substantial separation of powers. Madison wrote "the very
definition of tyranny" consisted in "[t]he accumulation of all powers,
legislative, executive, and judiciary, in the same hands, whether of one,
a few, or many." 3 In particular, the Framers were at pains to separate
the legislative and executive branches. This separation was not absolute,
for each branch has various powers over the other. These powers over
each other, however, were not meant to merge the two branches but
rather to give each, in Madison's words, "the necessary constitutional
means and personal motives to resist encroachments of the otherfl32 ....
Ambition," as he put it, "must be made to counteract ambition."
The Convention for a Democratic South Africa (CODESA) Declara-
tion of Intent expresses its signatories' agreement "[t]hat there shall be a
separation of powers between the legislature, executive and judiciary
with appropriate checks and balances. '33 But this broad endorsement
leaves much room for argument about the details, as our Framers' at-
tachment to separation of powers ideas left much room for argument
here.' The ANC leaves open the question of whether the President
should be directly elected by the people or "elected by and answerable
to Parliament." 3' The National Party proposes to make the Presidency
"consist of the leaders of the three largest parties" in the lower, more
representative, House of Parliament, and would empower Parliament to
pass a motion of no confidence which would presumably bring down
the Presidency.' The structure the United States has adopted firmly en-
dorses more separation than that. To understand our structure, and the
possibilities it offers for South Africa, it will be helpful first to contrast
31. THE FEDERALIST No. 47, at 301 (James Madison) (Clinton Rossiter ed.,
1961).
32. THE FEDERALIST NO. 51, at 322 (James Madison) (Clinton Rossiter ed.,
1961).
33. CONVENTION FOR A DEMOCRATIC SOUTH AFRICA, DECLARATION OF INTENT I
5(d) (adopted at the first meeting of CODESA, Dec. 20-21, 1991), reprinted in NE-
GOTIATION NEws, Apr. 24, 1992, at 16 [hereinafter CODESA DECLARATION].
34. See THE FEDERALIST No. 47, at 301 (James Madison) (Clinton Rossiter ed.,
1961) (commenting that during the debates over ratification of our Constitution, "[o]ne
of the principal objections inculcated by the more respectable adversaries to the Con-
stitution [was] its supposed violation of the political maxim that the legislative, exec-
utive, and judiciary departments ought to be separate and distinct").
35. ANC PRINCIPLES, supra note 16, 1 3.2, at 21.
36. NATIONAL PARTY FRAMEWORK, supra note 16, at 13.
1993] POST-APARTHEID SOUTH AFRICA
a very different form of government, British parliamentary rule, with the

system we have adopted.-"
The British Prime Minister is the leader of the majority party in the
House of Commons.' A member of that House himself (or herself), the
Prime Minister appoints the rest of the Cabinet from the ranks of the
other Members of this House (the "M.P.s" or Members of Parliament)
and to some extent from the House of Lords." Armed with tremendous
authority to insist on the voting support of the backbench M.P.'s,. the
Prime Minister is in a position to enact his or her program into
law-more or less without regard to what the opposition may have to
say about it.4 If the Prime Minister ceases to be able to do this, in
particular if he or she loses a vote of no confidence, then executive and
legislative branches are both likely to fall at once,': and a new exec-
utive will emerge from the new Parliament subsequently elected. This is
the British system; at least until the adoption of the 1983 constitution,
South Africa too had such a "Westminster system.' Even now, the
National Party and the State President are constitutionally capable of
wielding comparably decisive governmental authority." So far as this
37. For a detailed analysis of the British system (an analysis on which I rely
here), see Wilson supra note 18.
38. That is, the Prime Minister is "the person best able to form and maintain a
majority in the House of Comrnmons,"-the leader of the majority party when, as is
the case today, a single party does hold a majority.
39. Wilson, supra note 18.
40. Constitutional convention obliges Ministers to "publicly maintain complete
loyalty to the government's policies," and binds the other majority party Members of
Parliament (the "backbenchers") "to comply with their leaders' wishes on specific
votes." Id. Lest these conventions be violated (as can happen, id. at 138 n. 78), the
Prime Minister enjoys, inter alia, "enormous powers of patronage" with which to
bend resistant party members to his or her will. ladat 141-42.
41. See Wilson, supra note 18, at 140-41; Lloyd M. Cutler, To Form a Govern-
ment, reprinted in REFO M NG AmRICAN GOvERNmrES. THE BICENTENNIAL PAPERS
OF THE COMVY1rrrEE ON THE CONSTITUTIONAL SYSTEM 11, 14 (Donald L Robinson
ed., 1985) [hereinafter REFORMING AMERICAN GOVERNMENT].
42. See Wilson, supra note 18, at 139 (explaining that "upon a vote of no confi-
dence, the backbenchers can force the Prime Minister to resign or dissolve Parlia-
ment"). The latter course, the dissolution of parliament, results in the election of a
new Parliament. I&
43. See GRETCHEN CARPENTER, INTRODUCTION TO SOUTH AFRICAN CONST"U-
TiONAL LAW 74 (1987).
44. The 1983 Constitution created a tricameral Parliament for South Africa, in
which whites, "Coloureds," and Indians-but not Africans--each have a House). See
Republic of South Africa Constitution Act of 1983, §§ 37, 52. Whites, however, and
description goes, this is a formula for effective, but also unchecked,

majority rule.
In contrast, the American President is not normally elected by Con-
gress or by the majority party in Congress."' Cabinet officials also are
not members of Congress; such dual office-holding is explicitly forbid-
den by the constitution.' Nor does the President share a term with
Congress; Representatives are elected for two-year terms, and must face
re-election in the middle of the President's four-year term, while Sena-
tors are elected for staggered six-year terms, so that only one-third face
election campaigns in any Presidential election year. Thus, the
President's electoral fortunes are quite independent of those of the legis-
lators. In fact, during the past 25 years, the White House has almost al-
ways been in the hands of the Republicans, while at least one, and
usually both, houses of Congress have had Democratic Party majorities.
The President has no authority to dissolve a Congress dominated by the
in particular the National Party as the majority party in the white "House of Assem-
bly," remain preeminent. National Party discipline has historically been strong, see
HERMANN GIMOMEE, Afrikaner Politics: How the System Works, in HERIBERT ADAM
& HERMANN GIIOMEE, ETHNIC POWER MOBILIZED: CAN SOUTH AFRICA CHANGE?
200-01 (1979), and so this party controls the results of votes in the House of Assem-
bly. Even without the members it has recently gained in the other houses, the Nation-
al Party's control over the whites' house has been enough to insure its control over
the government as a whole. If the Coloured house (the "House of Representatives")
or the Indian house (the "House of Delegates") rejects proposed legislation which the
House of Assembly has approved (or vice-versa), then the State President may choose
to submit the disagreement for resolution to a body known as the President's Council.
The State President may also choose not to submit the disagreement to the Council,
thereby preventing the bill in question from becoming law. See Republic of South
Africa Constitution Act of 1983, §§ 32, 78. The State President, to whom this deci-
sion is entrusted, is effectively elected by the white majority party, id § 7; HAROLD
G. RUDOLPH, Constitutional Law, in ANNUAL SURVEY OF SOUTH AFRICAN LAW 1983
1, 4 (1985), and is, in fact, the leader of this party. Moreover, the President's Coun-
cil is so composed that the National Party and the State President can readily ensure
that this Council too works the will of the Nationalists-though this party will not
necessarily consider it expedient to override opposition in the other Houses. See Re-
public of South Africa Constitution Act § 70. See Rudolph, supra, at 1-9 (offering
comprehensive treatment of these and other features of the 1983 Constitution).
45. Normally the President is chosen by popular election (more precisely, by
"electors" selected in each state based on the popular vote for President in that state);
only in the event that no candidate wins election by this process does the House of
Representatives select the President. See U.S. CONST. art. H, § 1, amended by U.S.
CONST. amend. XII.
46. Id. art. I, § 6.
47. Id. art. I, §§ 2, 3; Id. art. II, § 1.
other party, nor does Congress have any authority to remove the Presi-
dent from office on the basis of a vote of "no confidence."' Yet while
Congress and the President are so plainly capable of having very differ-
ent perspectives on public policy, the passage of legislation generally
requires the two branches to concur. If the President vetoes a bill passed
by Congress, it will not become law unless a two-thirds majority votes
to override the veto. 9 This system is meant to prevent government ac-
tion-not all government action, to be sure, but that action which is the
product of what Madison called "faction," of segments of the communi-
ty, even majority segments of the community, acting in a way that is
unjust or unwise for the community as a whole.'
There have been powerful criticisms of the American system from
respected observers of, and participants in, our government. You will
hear from one of these critics, Lloyd Cutler, tomorrow. Mr. Cutler can
speak for himself, and from a wealth of experience, but let me outline
part of his critique in order to pursue my own argument here. Cutler
has argued that in our system of divided and mutually resistant authority
the President cannot, in his words, "form a government" and enact a
legislative program. Though he does not urge that this country adopt a
parliamentary system, he has written admiringly of such systems, in
which a slim majority can work its will, subject of course to the ulti-
mate test of the voters' approval or disapproval at the next elections.
Cutler maintains that on many issues of our day consensus is unattain-
able; indeed, even when the President's party holds a congressional
majority, consensus on the President's proposals is far from automatic.
Effective government therefoer requires that action should not await the
consensus that our system makes so important. In essence, Cutler main-
tains that the Framers succeeded too well; at least in our day, the sys-
tem they designed may avoid the dominance of faction only by weaken-
ing the government at a time when we need, as perhaps the Framers did
not, to have the government in action."
48. Congress can only remove the President from office by impeachment for
"Treason, Bribery, or other high Crimes and Misdemeanors." L. art. I § 4. No
President has ever been removed from office this way.
49. Id. art. I, § 7.
50. See THE FEDERALIST No. 10 (James Madison); GEOFFREY R. STONE, Louis
M. SEIDMAN, CASS R. SUNSTEIN, AND MARK V. TUsHNEr, CONSTITUTIONAL LAW 12-
14 (2d ed. 1991).
51. See Cutler, supra note 41; LLOYD N. CUTLER, Political Parties and a Work-
able Government, in A WORKABLE GOVERNMIENT? THE CONSTITUTION AFrER 200
YEARS 49-58 (Burke Marshall ed., 1987). For similar critiques of the American con-
Whether this critique is correct continues to be debated." I remain

agnostic about how much of the weakness of American politics should
be attributed to our governmental structure, and how much should be
attributed to the genuine complexity of our problems and the failure of
our people and our politicians to face them frankly and seriously. Fur-
thermore, as a person whose support for liberal Democrats has been
rather out of fashion lately, I think it is good that the Republicans have
not been able to write all of the policies of earlier Democratic eras out
of the statute books. In short, I believe there is value in inertia, and in
the requirement of considerable consensus to overcome it, in the legis-
lative process in this country.
It is surely true, however, that the separation of executive and legisla-
tive powers makes governmental action harder; the only debatable issue
is how much harder. It is equally clear that governmental action is not
as hard under the American system as it could be under a
consociational system, such as the one proposed by the National Party,
that would give extra representation to minority political parties and
impose heightened majority requirements as a predicate for the passage
of legislation affecting the relations among groups in South Africa." To
set barriers to government action as high as this is a risky step indeed,
for if it is entirely possible, as Donald Horowitz has argued, that the
necessary strong consensus simply will not be achieved.' In this light,
I think the American system has much to recommend it for South Afri-
ca. Certainly it would not give a Parliamentary majority as much power
as the British system; it would give considerable effective authority to a
President who is backed by a Parliamentary majority-as the first post-
apartheid South African President may be-while still offering minority
stitutional system, as compared to systems of parliamentary democracy, see Charles

Hardin, The Crisis and Its Cure, reprinted in REFORMING AMERICAN GOVERNMENT,
supra note 41, at 3-10; C. Douglas Dillon, The Challenge of Modern Governance, in
REFoRMING AMERICAN GOVERNMENT, supra note 41, at 24-29.
52. See Don K. Price, Words of Caution About Structural Change, in REFORMING
AMERICAN GOVERNMENT, supra note 41, at 39-49 (arguing against the application of
the parliamentary model in the United States-though in favor of a range of less
drastic reforms); See also Martin H. Redish and Elizabeth J. Cisar, "IfAngels Were
to Govern": The Need for Pragmatic Formalism in Separation of Powers Theory, 41
DUKE L.J. 449, 465-74 (1991) (suggesting that alternative forms of governing struc-
tures would not protect "popular sovereignty and individual liberty" in the United
States as well as the separation of powers model of government).
53. See supra notes 25-30, 36 and accompanying text.
54. See supra note 30.
parties a distinct capacity to resist policies that they dislike." This does
not make this system perfect for either those who share the views of the
National Party or those who share the views of the ANC. Nonetheless,
it may make it a useful compromise. Furthermore, the details of the
American structure need not be your blueprint; it is open to the parties
in South Africa, as some Americans urge we should do here at home, to
modulate the separation of powers, with a view to enhancing the power
of government to act. No doubt you have already begun looking at the
various solutions that other countries, from Germany to Namibia, have
devised to address these concerns.5
3. The Structure of the Executive Branch: One of the central doctrinal
points of American law governing separation of powers is that our Con-
stitution provides for a "single executive." There is one President, and
55. See HOROWrIZ, A DEMOCRATIC SOUTH AFRICA?, supra note 30, at 205, 214
(endorsing a "separate, directly elected executive" for South Africa in part on the
ground that this structure limits the chance that any single group can "capture the
state permanently by merely capturing a majority in parliament"). Horowitz also main-
tains that if the electoral system is properly designed the election of the President can
be an "arena for intergroup conciliation." Id at 205.
56. See Lloyd N. Cutler, Modern European Constittions and Their Relevance in
the American Context, in REFORMING AMERICAN GOVERNNT, supra note 41, at
299-312 (discussing the structures of government that have evolved in the United
Kingdom, France and Germany). Namibia's recently adopted Constitution establishes a
President who is elected directly by the people, and can only be removed by the
legislature by impeachment, a procedure which requires a two-thirds majority in both
houses of the legislature. See NAMIBIA CONST. arts. 28, 29. In these respects,
Namibia's system closely resembles that of the United States. Namibia ties the Pres-
ident somewhat more closely to Parliament than does the United States, however, for
Namibia authorizes the President to dissolve the National Assembly (the popularly
elected house of Parliament) "if the Government is unable to govern effectively," and
that step must be followed by new elections for both the National Assembly and the
President. Id art. 57. Even when the President exercises this power, however, the link
between the fate of the executive and legislative branches is not complete, for the
second house of Parliament, the National Council, is not subject to dissolution by the
President. See id (making no provision for dissolution of the National Council). Nor
are the Council's members elected directly by the people, as are the President and
the members of the Assembly. I, art. 69. The powers of the National Council are
not as great as those of the National Assembly, but they include the ability, in cer-
tain circumstances, to block the passage of legislation in the absence of a two-thirds
majority in the Assembly. NAMIBIA CONST. arts. 74, 75(5).
57. See Peter L. Strauss, The Place of Agencies in Government: Separation of
Powers and the Fourth Branch, 84 COLUM. L. REv. 573, 599-600 (1984) ("the Con-
vention clearly and consciously chose a single and independent executive over a colle-
gial body subject to legislative direction").
one only, and that person is vested with the "executive Power."58 This
decision, a fundamental choice made by the Framers after considering
alternative structures,59 rests on the belief that vesting power in a single
executive will make government both more effective and more ac-
countable than would a more diffuse system of executing the laws.'
This decision is frankly inconsistent with the National Party proposal.
That proposal stated "that the office of head of state and of government
should be vested in a collective body known as the Presidency," to
"consist of the leaders of the three largest parties" in the lower ("First")
house of Parliament, making decisions "by consensus," with the chair-
manship of this collective entity and the position of "State President" (if
such a position is actually established) changing hands on a rotating
basis."' Since no one member of the Presidency will be able to control
its decisions, this proposal is dangerously likely to impede effective
executive action and dissipate executive accountability.' It will also
deprive South Africa of the symbol of national unity, and source of
national leadership, that a single executive can provide-as we have
seen in this country, and as you have seen, not least under F.W. de
Klerk, in yours.
But American experience by no means counsels that each and every
element of the execution and administration of the law must be under
the control of a single person. This may seem paradoxical, given what I
have just said about the Framers' decision in favor of a single exec-
utive. Whether or not it is paradoxical, it is certainly controversial;
recent years have seen both court opinions and scholarly commentary
focused on explicating just how powerful the President must be. 3 I will
58. U.S. CONsT. art. II, § 1.

59. See Farber & Sherry, supra note 2, at 81-86; Strauss, supra note 57, at 599-
600; Miller, supra note 11, at 70.
60. Hamilton sounded both of these themes in THE FEDERALIST No. 70 (Alex-
ander Hamilton).
61. See NATIONAL PARTY FRAMEWORK, supra note 16, at 13.
62. Writing two centuries ago, Alexander Hamilton sharply criticized a system
that provided for "a plurality of magistrates [that is, executives] of equal dignity and
authority, a scheme, the advocates for which are not likely to form a numerous sect."
THE FEDERALIST No. 70, at 427 (Alexander Hamilton) (Clinton Rossiter ed., 1961).
Hamilton feared that such a structure would generate "the most bitter dissensions,"
which "might impede or frustrate the most important measures of the government in
the most critical emergencies of the state. And what is still worse, they might split
the community into the most violent and irreconcilable factions, adhering differently to
the different individuals who composed the magistracy." Ii at 426.
63. See, e.g., Morrison v. Olson, 487 U.S. 654, 685-696 (1988); Bowsher v.
touch on certain parts of this controversy, but I am concerned here not

to resolve this issue but to explicate possibilities that American law
presents for South Africa. Before I do so, let me just add that South
Africans are already investigating related possibilities. The ANC, for
example, has called for "an independent electoral commission" to over-
see every aspect of elections from the printing of ballot papers to the
adoption of regulations for access by parties to the public media and
fairness to all political parties by the public media.6 ' This proposal
would entail a vesting of very substantial executive authority outside the
central executive. Both the ANC and the National Party want to create
an independent Ombud or Ombudsman.' The National Party has pro-
posed greater autonomy as well for the Auditor-General, the Public
Service Commission, and the Reserve Bank' Let us look, therefore, at
the methods American constitutional law suggests for establishing such
quasi-independent executive agencies.
The beginning of wisdom in this sphere is surely the recognition that
even if the president is the head of government, almost all of what
government does will be done by subordinate officials. The exact extent
of the President's power, therefore, depends in large measure on the
extent of his or her ability to hire, fire, and control those subordi-
nates.Y
Limiting the President's power to hire officials is one way of con-
straining his or her authority. As our constitution is currently read, Con-
gress itself cannot appoint any official who exercises significant execu-
tive authority.' But Congress has wielded substantial discretion to de-
Synar, 478 U.S. 714, -721-734 (1986); Richard J.Pierce, Jr., Morrison v. Olson, Sepa-
ration of Powers, and the Structure of Government, 1988 SUP. Cr. REv. 1; Miller,
supra note 11; Strauss, supra note 57.
64. ANC PRINCIPLES, supra note 16, at 26, 5.
65. Id at 31, 14; NATIONAL PARTY FRAmeWORK, supra note 16, at 7.
66. Id.
67. See Wilson, supra note 18, at 129 (maintaining that 'the powers to hire and
fire largely determine constitutional structures"). I use the term "subordinates" some-
what loosely, since, as Professor Kurland suggested in his comments when this paper
was presented, the more the President's power to hire, fur, and direct an official is
constricted the less "subordinate" that official is. Cf. Mordson v. Olson, 487 U.S.
654, 719-23 (1988) (Scalia, J., dissenting) (distinguishing between "subordination" and
mere differences "in station or rank").
68. According to Buckley v. Valeo, 424 U.S. 1, 126 (1976), "any appointee exer-
cising significant authority pursuant to the laws of the United States is an 'Officer of
the United States,'" who must be appointed in accordance with the appointments
clause of the Constitution, U.S. CONST. art II, § 2, cl. 2. This clause by its terms
AM. U.J. INT'L L & POL'Y [VOL. 8:455
cide when appointments can be made. For example, Congress may pro-
vide by legislation that holders of an office serve a term longer than
that of the President, and thereby can delay or even eliminate the
President's chance to fill the office in question.' When appointments
are made, moreover, the Constitution requires Senatorial advice and con-
sent for all appointments of principal officers. The Constitution also
permits Congress to require this procedure for all officers of the United
States- a very wide category of officials." Congress has also enacted
legislation limiting the President's choice of appointees, by requiring the
President to appoint from, or at least consider, lists of nominees pre-
pared by others, or by specifying that those appointed must have partic-
ular credentials (such as political party membership).7' In addition, Con-
gress can decide to vest the appointments of "inferior officers" in the
heads of executive departments rather than in the President." Congress
can also place the appointment of at least some "inferior officers," in-
cluding even officials as powerful as independent prosecutors appointed
to prosecute government wrongdoing, outside of the executive branch
altogether and into the hands of the courts.'" Finally, the Constitution
does not expressly limit Congress' authority to control the appointment
of those other government employees who do not wield enough authori-
ty to count as "officers of the United States."' At the very least, it
authorizes the President to "nominate, and by and with the Advice and Consent of
the Senate .... appoint" the most senior of these officials. Lesser, or "inferior,"
officers can be appointed in the same way, or Congress may choose to alter the pro-
cess, by vesting these appointments "in the President alone, to the Courts of Law, or
in the Heads of Departments"-but Congress' discretion does not extend to making
these appointments itself. Id
69. This device was first used within a few years of the nation's founding. In
Marbury v. Madison, 5 U.S. 137, 162 (1803), Chief Justice Marshall appeared to take
for granted that Congress could constitutionally provide for the appointment of justices
of the peace to hold office for five-year terms-one year longer than the term of the
incoming President).
70. U.S. CONST. art. II, § 2, cl.2.
71. See Metropolitan Washington Airports Auth. v. Citizens for the Abatement of
Aircraft Noise, 111 S. Ct. 2298, 2319 (1991) (White, J., dissenting) (citing instances
in which the Supreme Court voiced no objection to statutory requirements that the
President consider lists of candidates); see also Miller, supra note 11, at 51 (noting
that "the independent regulatory commissions .. . almost uniformly . .. [have] politi-
cal criteria for appointment, with no more than a majority allowed to come from one
party").
72. See supra note 68 (discussing the appointment of officials).
73. See Morrison v. Olson, 487 U.S. 654, 670-77 (1988) (holding that Congress
may vest the authority to appoint an independent counsel to a special federal court).
74. The appointments clause speaks explicitly only to the appointment of "offi-
seems clear that Congress can direct that these lesser officials be hired
through civil service procedures over which the president's control is
restricted."
Similarly, limiting the President's power to fire officials also con-
strains Presidential authority. Again, our Constitution is currently un-
derstood to deny Congress authority to fire executive "officers" except
through the rarely used mechanism of formal impeachment." But that
does not mean that the President has unchallenged power in this sphere.
Where appointment is not vested in the President but in lesser offi-
cials-as would be the case for a great many civil service
appointments-Congress has long been thought to have authority to
regulate the conditions of discharge." Even where the President ap-
points, his or her freedom to discharge may be constrained. At one time,
the Supreme Court seemingly held the view that the President was con-
stitutionally entitled to fire a wide range of officials at will.' Subse-
quent cases, however, including one very recently, have decidedly cir-
cunscribed this notion.' If the President cannot fire "at will" but only
for "cause," presidential power is reduced. If the President cannot treat
as "cause" a subordinate's refusal to comply with presidential policy
cers." See U.S. CONST. art. 1I, § 2, cl.2.

75. Indeed, it would seem that Congress can regulate the hiring and firing of
"inferior officers," as well as employees below the level of "officer," through the
establishment of a civil service system. See generally 5 U.S.C.A. §§ 1201-04 (Supp.
1993) (specifying composition and functions of Merit Systems Protection Board); id.
§§ 2301, 2302; (setting out "merit system principles" and "prohibited personnel prac-
tices" to govern "[flederal personnel management"); United States v. Perkins, 116 U.S.
483, 485 (1886) (holding that Congress may restrict the removal of inferior officers
whose appointment it has vested in the "Heads of Departments"); Jalil v. Hampton,
460 F.2d 923, 926 (D.C. Cir.), cert denied, 409 U.S. 887 (1972) (characterizing the
civil service laws as limiting presidential appointment authority by prescribing quali-
fications for appointees, and stating that "it is now generally accepted that these [civil
service] laws raise no constitutional questions"); Strauss, supra note 57, at 608.
76. See Bowsher v. Synar, 478 U.S. 714, 726 (1986) (deciding "that Congress
cannot reserve for itself the power of removal of an officer charged with the execu-
tion of the laws except by impeachment").
77. See United States v. Perkins, 116 U.S. 483, 485 (1886) ("when Congress, by
law, vests the appointment of inferior officers in the heads of Departments it may
limit and restrict the power of removal as it deems best for the public interests").
78. See Myers v. United States, 272 U.S. 52, 176 (1926) (deciding that Congress
could not constitutionally deny the President exclusive power to fire any executive
officer appointed by the President with the advice and consent of the Senate).
79. See Morrison v. Olson, 487 U.S. 654 (1988); Humphrey's Executor v. United
States, 295 U.S. 602 (1935).
preferences-an undecided question-then the chief executive's power is

even further diminished.'
The power to hire and fire largely determines the president's power to
control the acts of other executive officials, but by no means complete-
ly. On the contrary an array of features of the relationships between
lesser officials on the one hand, and the President and Congress on the
other, will also affect the extent of presidential power.8 Perhaps the
most vivid method of control that a President might have available
would be the prerogative to bypass a subordinate official altogether and
execute the matter in question directly. But this course of action will be
blocked if elements of executive responsibility can be vested in named
subordinate offices, so that only the occupant of the particular office can
carry out the function at issue. Suppose, for example, that the Attorney
General is assigned by statute the responsibility to decide whether to
bring a prosecution or not. Can the President take this decision out of
the Attorney General's hands in a particular case in which the President
is intent on prosecution, but the Attorney General believes none is mer-
ited? Peter Strauss argues that the American Constitution would require
the president to find an Attorney General who would do his bidding.
80. For an argument that the power to remove "for cause" should be interpreted
to allow the removal of policymaking officials when they "fail[] to comply with any
valid policy decision made by the President or his agent," see Pierce, supra note 63,
at 24-35. See also Miller, supra note 11, at 44-45, 50-97 (contending that the Presi-
dent must be able "to remove a policy-making official who has refused an order of
the President to take an action within the officer's statutory authority").
81. For example, it will make a difference whether the President has a right to
be informed of lesser officials' planned actions before they are undertaken; without
such an authority, as Peter Strauss argues, the President is less able to coordinate
policy and less able to bring political pressure to bear on officials who seek to chart
a course with which the President disagrees. See Strauss, supra note 57, at 646-47.
On this score, the Constitution somewhat obscurely provides that the President "may
require the Opinion, in writing, of the principal Officer in each of the executive
Departments, upon any Subject relating to the Duties of their respective Offices." U.S.
CONsT. art. II, § 2, cl. 1. See Strauss, supra note 57, at 646-47 (providing a broad
interpretation of this clause).
So, too, it matters whether the President has the prerogative of communicating
with other officials confidentially, or whether instead the legislature can insist on
being provided with full accounts of internal executive deliberations. See id. at 653-
62. The Constitution contains no express grant of an "executive privilege," but the
Supreme Court has declared that this "privilege of confidentiality of Presidential com-
munications" is "fundamental to the operation of Government and inextricably rooted
in the separation of powers under the Constitution"-although it is not absolute. Unit-
ed States v. Nixon, 418 U.S. 683, 705, 708 (1974).
Although the president might well find such a person, the process of
doing so would-and in comparable circumstances actually did, during
the unraveling of the Watergate scandal-exact a political cost.'
By channeling or limiting the President's appointment and removal
powers, and by vesting particular duties in specified subordinate offi-
cials, legislation in this country has been able to place important ele-
ments of executive power to some degree outside the President's control.
It deserves mention that legislation on these issues also has often at-
tempted to insure that particular areas of government policy were to
some extent beyond partisan control by either political party-in partic-
ular by offering members of both the majority and the minority party
some input into decision-making. Such nonpartisan policy-making seems
to be an aspiration of statutes that divide seats on administrative agen-
cies between the Democratic and Republican parties, although American
political parties are so ideologically inclusive that such systems are not
in themselves likely to deny the President the chance to select people,
from either party, with whose views he or she is largely content.' If
the South African Constitution that is now being written is to limit
executive power, and ensure that minority views are heard in the pro-
cess of policy-making, such steps may prove helpful. So, for example, a
Judicial Service Commission might be required to have representation
from each of the major parties, and the President required to pick judges
from nominees submitted by the Commission.' Such structures could
restrain the power of the single executive without the drastic inroads on
governmental efficacy that the consociational model risks.
4. Preserving in practice what you set out on paper. Benjamin
Franklin, already a venerable statesman when he served as a member of
the Constitutional Convention, supposedly was asked after the Conven-
tion what kind of government the Framers had shaped for the American
82. See Peter L. Strauss, Letter to the Editors of the Harvard Law Review, May
22, 1992 (copy on file with the author); Strauss, supra note 57, at 605 (suggesting
that Congress has considerable discretion to "plac[e] ultimate responsibility for deci-
sion" on particular matters with subordinate officials rather than with the President).
See also Miller, supra note 11, at 58-59 (similarly observing that it was "accepted
early on that Congress could not only create executive departments but could make
action by the agency head a precondition of the effective exercise of the power").
83. I am grateful to Peter Strauss for this point.
84. See NAMIBIA CONST. arts. 82, 85 (providing that the President is to appoint
judges "on the recommendation of the Judicial Service Commission"; two of the five
members of the Judicial Service Commission are appointed by organizations of the
legal profession rather than by the president).
people. He answered, "A republic, if you can keep it."" Keeping your
republic, or democracy, must be as great a concern for you as it was,
and is, for Americans. The record of African states on this score is not
good, and South Africa itself has a dreary record of oppression of hu-
man rights. Designing a system that will hold up over time is by no
means easy.
United States history demonstrates this proposition-if any proof is
needed. The Framers of our Constitution are revered as unusually far-
sighted and wise, but they evidently failed to predict that so fundamen-
tal an institution as the political party would become an important fea-
ture of the new United States.' It was only a few years after the Con-
stitutional Convention when many of them began to form the political
party system which has been a critical part of our polity ever since.
Similarly, many of the Framers viewed the greatest danger to republican
liberty as coming from the legislatureY Today, many Americans are
convinced that the executive branch has long since overmatched the
Congress in the struggle for power. So, too, the Framers apparently
anticipated a rather straightforward, and modest, role for the courts."
Modem Americans live in a world in which concern about judicial
power undermining democratic self-government is a recurrent theme.
And, as you have already heard, the Framers anticipated a relatively
small central government, carrying out relatively confined functions,89
but the reality of modem American life is far removed from this. Not
only does the federal government do vastly more and the states relative-
85. See Kurland, supra note 8, at 613 and n.51 (citing 3 THE RECORDS OF THE
FEDERAL CONVENTION OF 1787 85 (Max Farrand ed., 1911).
86. See Kurland, supra note 8, at 599; see also James MacGregor Burns, THE
DEADLOCK OF DEMOCRACY: FOUR-PARTY POLITICS IN AMERICA 27 (rev. ed. 1963)
(noting that "most Americans in the early 1790s did not want parties").
87. See THE FEDERALIST No. 48 (James Madison). Other delegates to the Consti-
tutional Convention, however, may have seen the executive branch as the greatest
source of peril. See Kurland, supra note 8, at 598 (discussing the fears of some at
the Constitutional Convention about concentrated power in the executive branch).
88. Kurland, supra note 8, at 599. In his defense of the new Constitution's pro-
visions for the judiciary, Hamilton was able both to insist on the courts' power to
declare statutes unconstitutional and to characterize the judiciary as the branch "least
dangerous to the political rights of the Constitution," and "incontestably . . . beyond
comparison the weakest of the three departments of power." THE FEDERALIST No. 78,
at 465-66 (Alexander Hamilton) (Clinton Rossiter ed., 1961).
89. See THE FEDERALIT No. 45, at 292 (James Madison) (Clinton Rossiter ed.
1961) (contrasting the powers of the federal government, which are "few and de-
fined," with the states' "numerous and indefinate" powers).
ly less, but the federal government characteristically acts through admin-

istrative agencies in which the functions of law-making, law-enforcing,
and law-adjudicating, however separated they may be in the text of the
constitution, are actually melded together.
Yet, in the midst of this, the structural provisions of our Constitution
remain almost unamended after 200 years. As much as the country and
the government have changed, moreover, much of the structure of our
polity would be recognizable to the Framers as being their handiwork.
Much else would plainly reveal that its ancestry lay in what they had
wrought. To a great extent, moreover, this continuity cannot be attribut-
ed to judicial enforcement of the provisions of the separation of powers.
Until quite recently, the courts rarely adjudicated questions concerning
the relative powers of the executive and legislative branches of the
federal government.' Instead, this continuity has to be attributed to
other factors. Among them would be the Framers' insight as constitu-
tional designers, shaping provisions that offered a fairly clear and fairly
workable structure, and one in which the political branches of govern-
ment did have the resources to challenge each other's pretensions as
well as the capacity, at least sometimes, to work effectively together."
So, too, the country's growing constitutional faith surely helped to en-
courage Americans to adhere to this structure, or at least to seek the
image of adherence.z Ironically, the stability of our written Constitution
no doubt also owes a great deal to the facility Americans have shown
for improvising, on the framework of the constitutional text, structures
90. See JESSE FL CHOPER, JUDICIAL REVIEW AND THE NATIONAL POLITICAL
PROCESS: A FUNCTIONAL RECONSIDERATION OF THE ROLE OF THE SUPREME COURT
287 (1980) (commenting that "at least before the tenure of President Nixon, the re-
cord demonstrates that an effective functioning of the relationship between the legisla-
tive and executive branches has been almost wholly accomplished without the partic-
ipation of the national judiciary").
91. See id. at 275 (arguing that "[e]ach branch-legislative and executive-has
tremendous incentives jealously to guard its constitutional boundaries and assigned
prerogatives against invasion by the other. If either branch perceives a constitutional
violation of this kind, not only will it be encouraged to respond vigorously but each
department possesses an impressive arsenal of weapons to demand observance of
constitutional dictates by the other"). See i& at 275-314 (detailing the "strictures on
ultra vires executive action").
92. See ROBERT M. COVER, JUSTICE ACCUSED: ANTISLAVERY AND TIM JUDICIAL
PROCESS 132 (1975) (commenting on "the growing sanctity of the Constitution" by
the beginning of the nineteenth century and noting that "[f(or most Americans, and,
certainly for the legal profession, the Constitution was supreme law but had a moral
authority beyond its supremacy as positive law").
480 AM.U.J. INT'L L & POL'Y [VoL. 8:455
and principles of which the Framers themselves scarcely dreamt. For a

Constitution to last it must work; but part of its working must be its
flexibility-whether intended by the drafters, or insinuated by those who
come after them.93
How should courts respond to this simultaneous, and almost paradoxi-
cal, need for both structural strength and flexibility? Our experience on
this score has not been a terribly happy one. Sometimes our Supreme
Court has tried to fashion bright-line rules to resolve separation of pow-
ers questions, but these rules have more than once had an arbitrary,
even illogical, flavor to them.' On other occasions the Court has self-
93. At least some of the flexibility of our constitutional structure seems to have
been intended. As Peter Strauss has commented, "[o]ne scanning the Constitution for
a sense of the overall structure of the federal government is immediately struck by its
silences." Strauss, supra note 57, at 597. In particular, the Constitution says scarcely
anything about the structure of the executive branch below the Presidency. Strauss
concludes that "a determination was made to eschew detailed prescription as a means
of underscoring presidential responsibility and preserving congressional flexibility with-
in the constraints of the judgments that had been made." Id.at 599. Moreover, he
argues, "[t]he imprecision inherent in the definition and separation of the three gov-
ernmental powers contributes to the tensions among them," and this very tension helps
protect the liberty of the people. Id. at 603.
Even if such flexibility was not intended, however, our patterns of constitution-
al interpretation have often found room in the constitutional text for the infusion of
modern perspectives. Terrance Sandalow maintains that although the framers "set us
on the path" to our present constitutional understandings, "[tihe entirety of [our) histo-
ry, together with current aspirations that are both shaped by it and shape the meaning
derived from it, far more than the intentions of the framers, determine what each
generation finds in the Constitution." Terrance Sandalow, Constitutional Interpretation,
79 MicH. L. REv. 1033, 1042, 1050 (1981). "Ultimately," he believes, "the values to
which constitutional law gives expression are more nearly those of the present than
those of the past." It.at 1039. It is possible to deny the legitimacy of such reinter-
pretation of the largely unamended language of our Constitution, but very hard to
deny that it has often taken place.
94. See Immigration and Naturalization Service v. Chadha, 462 U.S. 919 (1983)
(holding the congressional veto provision in the Immigration and Naturalization Act
unconstitutional). In Chadha, the Supreme Court insisted on what it saw as the
Constitution's limits on Congress' exercise of legislative power. In the process, it
appeared to say that Congress is exercising legislative power when its action has "the
purpose and effect of altering the legal rights, duties, and relations of persons . . .
outside the Legislative Branch." Id. at 952. As a definition of legislative power, this
formulation is clearly ill-formed, since it would encompass a wide variety of actions
by courts and executive agencies as well. For Critical assessments of this aspect of
Chada's definition of legislative power, see Peter L. Strauss, Was There a Baby in
the Bathwater? A Comment on the Supreme Court's Legislative Veto Decision. 1983
DUKE L.J. 789, 794-98; E. Donald Elliott, INS v. Chadha: The Administrative Consti-
consciously tried to be flexible, or, as we sometimes say, functional-to

avoid bright-line rules in favor of a much more wide-ranging assessment
of whether particular structural innovations do, or do not, fit with the
broad concerns of the constitutional design. This method avoids rigidity,
but perhaps only to approach the other extreme at which, as Justice
Scalia has complained in dissent, there are no lines at all. '
One might respond by embracing the position of Jesse Choper, ftom
whom you will hear in the next panel. Choper has urged courts to stay
out of the business of regulating the constitutional separation of powers
between the executive and legislative branches.' I do not make that
response, however, because I think that judicial monitoring of the sepa-
ration of powers, awkward and imperfect as it will be, is still likely to
help affirm the importance of fidelity to the Constitution's structure,
especially in a new nation whose structural design is untested.9 ' But if
you accept that the desirable structure is one that is both firm and flexi-
ble, and if you look to courts to help you achieve and maintain that
structure, then you will be vesting in the courts a very challenging role
indeed. Rather than simply parsing text, courts concerned with the sepa-
ration of powers will need some quantity of the expertise of political
scientists or even politicians. Just as courts will acquire something of
the role of a national conscience as they enforce the Bill of Rights, they
will play something of the role of a national political engineer as they
monitor the separation of powers.
I do not think, however, that the courts can be relied on to make the
government work. They can refine some decisions, overrule certain
others, but the stability and success of this government will ultimately
depend on the government itself, and the people themselves. That, again,
is why structure is important. South Africa faces profound challenges of
correcting past injustice, achieving economic development, and embrac-
ing a very diverse population in a single nation. The task of shaping a
government that can effectively address these problems, both by acting
on behalf of the majority and by honoring the concerns of minorities, is
tution, the Constitution, and the Legislative Veto, 1983 S. Cr. REv. 125, 134-38.
95. See Morrison v. Olson, 487 U.S. 654, 726 (1988) (Scalia, J., dissenting) (dis-
cussing the danger of further congressional interference with the President's power to
remove subordinates because "[t]here are now no lines" within which Congress must
stay).
96. See Choper, supra note 90, at 263 (stating this position, the "Separation
Proposal"); id. at 260-379 (discussing and defending this proposal).
97. For critical assessments of Choper's proposal, see Redish & Cisar, supra note
52, at 492-94; Strauss, supra note 57, at 620-21, n.194.
yours. I hope this dip into American experience helps you in your work.

The Separation of Powers in A Post-Apartheid South Africa

Uploaded by

Copyright:

Available Formats

The Separation of Powers in A Post-Apartheid South Africa

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Separation of Powers in A Post-Apartheid South Africa

Uploaded by

Copyright:

Available Formats

American University International Law Review

The Separation of Powers in a Post-Apartheid

Follow this and additional works at: http://digitalcommons.wcl.american.edu/auilr

My task this afternoon is a large one: to describe the lessons of the

tions, sober reasoning, political priorities and even prejudices. Nonethe-

2. See DANIEL A. FARBER AND SUZANNA SHERRY, A HISTORY OF THE AMER-

to recognize the limits on its attainability. It is difficult even to imagine

purpose, however, for it does not prevent bad governmental action by

growth of government, the safeguard thought to be inherent in separation of powers

consociationalism. What they counsel in favor of, broadly, is splitting

The National Party has endorsed bicameralism, while the ANC is

16. See CONSTITUTIONAL RULE IN A PARTICIPATORY DEMOCRACY: THE NATIONAL

with more than one set of fellow voters.

19. U.S. BUREAU OF THE CENsus, STATISTICAL ABSTRACT OF THE UNrrED

Constitution, vigorously opposed it, as did others at the Constitutional

support in the election in the region's legislative body will be allocated

27. NATIONAL PARTY FRAMEWORK, supra note 16, at 11-12.

2. The Separation of the Executive and Legislative Branches: It is a

a very different form of government, British parliamentary rule, with the

description goes, this is a formula for effective, but also unchecked,

Whether this critique is correct continues to be debated." I remain

stitutional system, as compared to systems of parliamentary democracy, see Charles

58. U.S. CONsT. art. II, § 1.

touch on certain parts of this controversy, but I am concerned here not

cers." See U.S. CONST. art. 1I, § 2, cl.2.

preferences-an undecided question-then the chief executive's power is

ly less, but the federal government characteristically acts through admin-

and principles of which the Framers themselves scarcely dreamt. For a

consciously tried to be flexible, or, as we sometimes say, functional-to

You might also like