OVERVIEW OF ADMINISTRATIVE LAW

1 Sources of law

The main sources of administrative law are as follows: 1. The Constitution of the Republic of South Africa, 1996. As we know, the Constitution is the source of all legal authority in South Africa. It is thus fundamental to administrative law. Section 33 is particularly important, because it enshrines the right to just administrative action. The right of access to information, contained in s 32, may be understood as a complementary right. The remainder of the Bill of Rights contains other normative standards which constrain public functionaries. The same is true, in fact, of constitutional provisions outside the Bill of Rights. The Constitution also establishes a number of state institutions to further the goal of lawful and efficient administration, most notably the Public Protector and the Auditor General. The fundamental constitutional value of the rule of law (s 1(c)) has also played a major role in our post-1994 administrative law. Other constitutional virtues of relevance are the promotion of public participation and accountability, and openness and responsiveness of government. 2. Statute law The most important statute here is the Promotion of Administrative Justice Act 3 of 2000 (PAJA), which enables the s 33 right. The Promotion of Access to Information Act 2 of 2000 (PAIA) was enacted contemporaneously and enables the s 32 right. 3. Common law The common law is of course voluminous. In the specific area of administrative law, South Africas common law is overwhelmingly English in origin; the Roman-Dutch influence has been virtually non-existent.

Judicial review

Judicial review is one legal mechanism constraining administrative action. (We will see later that it is important to distinguish review from appeal.) Judicial review can be subdivided into two kinds: review of the passing of legislation, and review of administrative actions. We are very familiar from previous courses with judicial review of legislation. Judicial review of administrative action, the subject of this course, looks at the legality of the exercise of powers given by the legislature to the executive. (Sometimes the powers are exercised by private persons, but we ignore this for now.) So we can see that the separation between judicial review of legislative actions and judicial review of administrative actions reflects the familiar separation of powers between the legislature and the executive. A judicial body may review the actions of either the legislature in passing legislation (legislative action), or the executive in carrying out such legislation (administrative action). Administrative law is concerned only with the latter.

There is a big emphasis on judicial review in our law; review by the superior courts is by far the most important way in which our law constrains administrative actions. The same is true of other Commonwealth jurisdictions. The preeminence of judicial review is to the detriment of what is (sometimes pejoratively) known as nonjudicial review. But it must always be remembered that there are numerous extra-judicial ways of holding public officials to accountjudicial review is not the only, nor even the best, way of doing so. Non-judicial means of holding public power to account include: the right of access to information ombudsmen Chapter 9 institutions Parliament itself, in its oversight role free media civil society administrative appeals tribunals (tax courts, the Rental Housing Tribunal, Consumer Complaints Tribunal, etc.) And the ultimate protections against abuses of public power are democratic elections and the rule of the motherfucking law. 3 Grounds for review

How would one challenge an administrative action? The following is a rough overview of the requirements that an administrative action would have to meet to be lawful. Suppose that we are dealing with empowering legislation which allows the Minister to do X if he is of the opinion that requirements Y are met. A person seeking to challenge the Ministers action, purportedly taken in terms of that legislation, could then argue that any of the following requirements has not been met: A. Common-law requirements 1. The correct form or medium There are often formal or procedural requirements stipulated in the empowering legislation. For example, it may be stipulated that any ministerial proclamation made in terms of that legislation must be published in the Government Gazette in order to be effective. If the proclamation was not published timeously in the Gazette then it would be unlawful. 2. Undertaken by an authorized agent Of course, the administrative action must be undertaken by the person actually authorized to do so by the empowering legislation, in this case the Minister. Traditionally the person to whom Parliament delegated power could not himself sub-delegate that power. Now s 238 of the Constitution allows sub-delegation in limited circumstances. 3. All jurisdictional facts were satisfied

Jurisdictional facts are those circumstances which are necessary to establish the relevant functionarys jurisdiction. In our example, the empowering legislation specifies the Ministers being of the opinion that requirements Y are met as the relevant jurisdictional fact; only if he is of the opinion that requirements Y are met does the Minister have jurisdiction to do X. If this was not the case, the administrative action would be unlawful. Of course, it would be very hard indeed to show that the Minister was not of the opinion that requirements Y were met. The phrase is intensely subjective; if the functionary swears in affidavit or in oral testimony that he was of the opinion that X, courts will be very loath indeed to question this. This is exactly why the apartheid government often used that language. If the phrase used was has reason to believe then the functionary would have to not only believe the requirements were met but also do so on the basis of certain real, objective information. This would make it easier for a court to invalidate the action. If the phrase used was is satisfied that then it is not clear how objectively or subjectively this standard would be interpreted. The basic point is that the precise language used in the empowering legislation, and the rules of statutory interpretation, are very important. 4. Intra vires the empowering legislation Any exercise of public power has to be authorized by law. Thus the public official must act intra vires (within the powers of) legislation passed by Parliament, as opposed to ultra vires such legislation. This rule is based in democratic theory: unelected members of the executive may exercise public power only when authorized to do so by the elected members of Parliament; were such members of the executive to act otherwise, their actions would be severed from any democratic accountability. During apartheid, the government subscribed very carefully to the doctrine of ultra vires and ensured that all administrative action was cloaked in the legitimacy of law. But, because the sovereign Parliament was not democratically elected, there was no true accountability. Now the Constitution is sovereign. But the doctrine of ultra vires persists. The example of apartheid makes clear that the ultra vires doctrine is a very low hurdle for the executive to overcome. Fortunately there are more substantial ways to challenge administrative action: 5. Bona fides All administrative actions must be undertaken in good faith or bona fide. So, for example, if the Minister proclaims that all the prerequisites of the empowering legislation are met, one may argue that this proclamation was insincere or otherwise mala fides. But, since the onus of proving mala fides rests on the person challenging the legality of the administrative action, this is a very difficult ground to satisfyhow does one prove that the Ministers subjective state of mind was not one of good faith? If the Minister submits a sworn statement as to his bona fides, the courts will almost invariably accept that. 6. Rationality Rationality requires that all administrative actions be coherent or logically connected to a certain public objective. It is a lower standard than reasonableness (see below). All it requires is that the action not be clearly 3

irrational. (Reasonableness was definitely not a pre-constitutional ground for review. It is not even clear that rationality was.) This is another relatively unexacting standard. 7. Procedural fairness There are certain principles of natural justice which require that procedures relating to the exercise of public power are fair. The two key ones are: audi alteram partem (hear both sides), which requires that the person(s) substantially affected by the administrative action be consulted or permitted to make representations before such action nemo iudex in causa sua (no-one should be a judge in his own cause), which requires that the person executing the action be unbiased and disinterested The following requirements would be relevant to whether a court was able to declare the action unlawful: 8. Standing to sue The person coming before a court to challenge the administrative action must have standing to sue, i.e. he must a direct personal interest in the matter. 9. Ouster clause The legislation may contain an ouster clause (known in foreign jurisdictions as a privative clause) which ousts the courts jurisdiction to hear disputes arising from that legislation. For example, the legislation may say, No court may inquire into the legality of any executive action undertaken in terms of this Act. If so, judges will be very reluctant to review such executive actionalthough dangerous mofos like Didcott J reasoned that they were allowed to inquire whether the executive action was in fact undertaken in terms of the Act, thus restoring their jurisdiction.

B. Constitutional era requirements Now that we have the Constitution, it provides the most important constraints on administrative action. But to a large extent the Constitution simply makes more robust the powers of review which had existed under the common law. Section 33(1) of the Constitution states: Everyone has the right to administrative action that is lawful, reasonable and procedurally fair. Each ground of review can now be categorized under one of those three criteria. The lawfulness requirement means that the action was lawfully authorized. This is nothing new, but now we can understand this in terms of the Constitution. Specifically, the Constitution is the supreme law (s 2) and indeed one of South Africas founding values is, according to the super-entrenched s 1(c), the supremacy of the Constitution and the rule of law. The implication is that any exercise of public power must be authorized by the Constitution in order to be lawful. This authorization may be direct (i.e. the Constitution specifically grants that power) or indirect (i.e. legislation grants that power, and that legislation was validly enacted in terms of the Constitution). The upshot, then, is that the exercise of power is: 1. Based on legislative authority 4

In short, the Minister must be able to refer to some legislation which has given him power. Since the Minister will virtually always purport to rely on some actually-existing legislation, what this requirement often means in practice is that a person seeking to challenge the administrative action must apply to have the empowering legislation struck down. The bases upon which such a challenge could be undertaken are varied (although ultimately they are all in the Constitution); we are familiar with them from studying judicial review of legislation. The effect of a successful challenge to the legislation would be that the empowering legislation is taken not to exist, and so any administrative action supposedly taken in accordance with that legislation will necessarily become unlawfulthe law which supposedly authorized the action is no longer valid. 2. Intra vires the empowering legislation Besides the fact that the empowering legislation must validly exist, the exercise of public power must be in accordance with that empowering legislation. This is what was covered above under the common law. The next s 33 requirement is: 3. Reasonableness This is now a powerful ground of judicial review of administrative action, and one which was not available under the common law. We will discuss it in more detail later. Finally s 33 requires: 4. Procedural fairness Again we dealt with this above. Interested parties must be given adequate notice, a fair hearing, and so on. 4 Sachs v Minister of Justice

In this section we look in detail at the case of Sachs v Minister of Justice; Diamond v Minister of Justice 1934 AD 11. Solly Sachs, father of Albie, was a dirty Commie bastard. His job was to be a crazy, shit-stirring, tradeunionizing motherfucker in the greater Johannesburg area. The Minister of Justice, acting in terms of s 1(12) of the amended Riotous Assemblies Act 27 of 1914,1 sent Sachs a notice barring him from entering various magisterial districts in which he worked. Section 1(12) stated:
Whenever the Minister is satisfied that any person is in any area promoting feelings of hostility between the European inhabitants of the Union on the one hand and any other section of the inhabitants of the Union on the other hand, he may by notice under his hand, addressed and delivered or tendered to such person prohibit him, after a period stated in such notice being not less than seven days from the date of such delivery or tender, and during a period likewise stated therein, from being in any area defined in such notice.

And s 1(13):
1

This was the first post-Union piece of legislation to deal with public disturbances. It was enacted to deal specifically with the ongoing miners strikes and the Maritz Rebellion.

If any person to whom a notice has been delivered or tendered under sub-sec. (12) requests the Minister in writing to furnish him with the reasons for such notice, and with a statement of the information which induced the Minister to issue such notice, the Minister shall furnish such person with a statement in writing setting forth his reasons for such notice and so much of the information which induced the Minister to issue such notice as can, in his opinion, be disclosed without detriment to public policy.

Sachss attorney promptly wrote to the Minister asking for reasons and information regarding Sachss barring. The Minister supplied reasons, but he did not supply information, saying that to do so would be detrimental to public policy. The attorney wrote back, saying that more information must be provided or his client will apply to court for an order restraining the Minister from enforcing the notice. And so the dispute went to the High Court, which found against Sachs, who then appealed to the Appellate Division. Note firstly the various jurisdictional facts required before the Ministers action can be lawful: the Minister must be satisfied that Sachs was promoting feelings of hostility between the races; he must send a written letter to Sachs personally which the Minister has signed (under his hand); the areas from which Sachs is barred must be defined in the letter; and seven days notice must be provided. And, since Sachss attorney asked for reasons and information, these must be provided subject to s 1(13). Note secondly the exact language used by the Act: the requirement is that the Minister is satisfied that the person is fomenting hostility (s 1(12)) and that it be in *the Ministers+ opinion that disclosure of the information would be detrimental to public policy (s 1(13)). As we have suggested previously, the degree of subjectivity of these particular phrases is controversial and has a big impact on the reviewability of pursuant administrative actions. Note thirdly that South Africas constitution was at that stage the South Africa Act of 1909 (a British Act of Parliament which had created the Union).2 This constitution entrenched voting rights for all races (or at least their male members with wealth above a certain threshold) as well as language rights (Dutch and English were made South Africas official languages).3 But there were no rights upon which Sachs couldve relied in challenging the administrative action or the empowering legislation upon which it was based. (a) Grounds of review Sachs argued that the ministerial notice must be set aside on the following grounds: 1. Unlawfulness The Minister had acted unlawfully, for the following reasons: a. The Minister was mala fides in concluding that he was promoting hostility. b. The Minister failed to apply his mind, the symptoms of which are:

Corder listed South Africas various constitutions: first, the South Africa Act of 1909; second, the Republic of South Africa Constitution Act of 1961, which marked South Africas departure from the Commonwealth; third, the Republic of South Africa Constitution Act of 1983, which introduced the tricameral Parliament and marked the beginning of the end of apartheid; fourth, the Interim Constitution of 1993; and fifth, the 1996 Constitution. 3 Both of these entrenchments could be amended only by a two-thirds majority of the Senate and House (as the legislative bodies were at that time known) sitting unicamerally. (These entrenched rights, while limited, suggest that Parliament was not truly sovereign, as is often claimed.) Ultimately the black vote was restricted by the process just described; and later the coloured vote was restricted too.

i. Gross unreasonableness, in that the reasons given were insufficient to justify the conclusion that he was promoting hostility; and ii. The Minister made a mistake of law in failing to follow the interpretation given to similar legislation in the case of Rex v Bunting. 2. Audi alteram partem He was not given a hearing prior to being barred. Nowadays we would say that these grounds span the three requirements of s 33, namely lawfulness (grounds 1(a) and 1(a)(ii) above), reasonableness (ground 1(b)(i)) and procedural fairness (ground 2). But, of course, Sachs was heard decades before the new Constitution, when the approach was quite different. Litigants had to characterize almost all their allegations as falling under the ultra vires doctrine, which is why lawfulness subsumes three of Sachss arguments. The only remaining argument is that the Ministers conduct was contrary to the principles of natural justice, specifically audi alteram partem. (b) Reasoning The court dealt with the four above arguments as follows: o Mala fides

Stratford ACJ dismisses this argument very quickly. He simply upholds the finding of the court a quo, saying that the facts disclose no foundation whatever for *a charge of mala fides+. This reminds one how difficult it is to prove mala fides; it requires proof of the Ministers subjective state of mind, which by its nature is unlikely to leave a trail of evidence. One would probably need the testimony of an insider in the Ministers department, Wikileaks, or some very stupid public statement by the Minister. o No hearing

Stratford ACJ begins by noting that the purpose of s 1(12) of the Act is aimed at preventive justice, i.e. stopping at the earliest possible stage the fomentation of feelings of hostility between the races. Given the necessity of prompt action, it would defeat the whole object of the particular provision if the Minister has to hold consultations with interested parties before taking action. The judge later deals specifically with audi alteram partem maxim:
Sacred though the maxim is held to be, Parliament is free to violate it. In all cases where by judicial interpretation it has been invoked, this has been justified on the ground that the enactment impliedly incorporated it. When on the true interpretation of the Act, the implication is excluded, there is an end of the matter.

Thus Stratford ACJ says that it is up to Parliament to include or exclude audi alteram partem as it sees fit. It is in fact false that all cases which read in the maxim did so because the enactment impliedly incorporated it; both before and after this case the maxim was presumed to be a part of any Act. But anyway, Stratford ACJs 7

remarkable account of the law means that it is then a simple question of deciding whether the Act does in fact give expression to the maxim. He finds that it does not, for the following reasons:
Now, however desirable it may be from the wide point of view of what is called natural justice, to hold an enquiry before issuing the notice under the section, the implication in this Act is clearly against it. Not only would such an enquiry defeat the cardinal purpose of prompt and preventive action, but sub-sec. (13) could not then be given a sensible meaning.

So he first says that inclusion of the maxim would defeat the purpose of the Act (which hed described previously) and then says s 1(13)which requires the giving of reasons after a persons barringwould be nugatory if a hearing had to be given before the barring. Stratford ACJ then says that audi alteram partem is not ignored altogether by the Act: it is recognised to a limited extent in s 1(13). The judge appears to be relying on the expressio unius rule4 of statutory interpretation to argue that such recognition as is given in s 1(13) is the only recognition the Act gives to the maxim. o Insufficient information

This is the most interesting ground advanced by Sachs, because it prompts the court to spell out its remarkably executive-minded approach to judicial review of administrative action:
[O]nce we are satisfied on a construction of the Act, that it gives to the Minister an unfettered discretion, it is no function of a Court of law to curtail its scope in the least degree, indeed it would be quite improper to do so. The above observation is, perhaps, so trite that it needs no statement, yet in cases before the Courts when the exercise of a statutory discretion is challenged, arguments are sometimes advanced which do seem to me to ignore the plain principle that Parliament may make any encroachment it chooses upon the life, liberty or property of any individual subject to its sway, and that it is the function of courts of law to enforce its will.

Despite this extreme attitude, Stratford ACJ goes on to spell out two grounds on which a court may validly interfere in exercise of a discretion conferred on the executive: (a) the exercise has not been honest (mala fides), or (b) the discretionary power has been exceeded (ultra vires). Stratford ACJ then comes to the specific argument Sachs has advanced, saying that:
This argument entirely ignores the words Whenever the Minister is satisfied which leaves the selection of the individual on whom he serves notice entirely to his discretion. If he is satisfied that such individual is promoting feelings of hostility he can validly serve the notice upon him whether in fact he is promoting hostility or not. The only question of fact with which the Court is concerned is whether the Minister was satisfied.

The point, in other words, is that the Minister is given a totally unfettered discretion to do what he subjectively believes is appropriate; and the court may not subject this belief to any objective scrutiny.

The rule is literally that expression of one thing excludes all others. So if a statute specifically provides for procedural fairness in one context but omits it in another, very similar context, then that omission was presumably deliberate.

Mistake of law

In advancing this argument, Sachs relied upon the judgment in Commissioner of Inland Revenue v City Deep, Ltd 1924 AD 298 to the effect that members of the executive must obey not only the provisions of a statute but also the law as laid down by a competent court of law. This principle is rooted in the separation of powers: the judiciary has the responsibility of deciding what the law is, and so its decisions must be obeyed by the other branches of government. The City Deep judgment held further that a failure by a government official to obey the decisions of a competent court of law would, in the absence of explanation, be evidence that he had not applied his mind to the matter. (Note by the way that a mistake of law is still a ground of review.) So Sachs was saying that the Minister had failed to comply with an interpretation of the law given by our courts. Specifically, he was relying on an interpretation given in Rex v Bunting 1929 EDL 326. But it is not even necessary to know what was decided in that case because Stratford ACJ summarily dispenses with Sachss argument. He basically qualifies the City Deep rule by giving conditions for the rules applicability:
Again it seems necessary to draw some distinctions; if, for example, a court of competent jurisdiction, interpreting an Act which confers a discretion, defines its nature and extent or lays down the conditions of its exercise, a person thereafter exercising that discretion is bound to observe the judicial interpretation of the enactment for the simple reason that whilst the judgment stands, the meaning given by the Court is the meaning of the Act. On the other hand judgments not relating to the abovementioned two matters may be disregarded if they are honestly thought to be erroneous.

Stratford ACJ says then that Rex v Bunting did not deal with such a matter, so the City Deep rule has no application here. (Gardiner AJA appears to disagree with Stratford ACJs take on the law, saying that he does not wish to expresses an opinion on the question of whether an honest belief that a judgment is erroneous entitles a member of the executive to ignore it.) (c) Argument in the alternative Thus far we have said nothing of Sachss argument in the alternative, which relates to s 1(13). He argued that this section was infringed because the reasons and information which the Minister provided on his request were not sufficient to constitute compliance with that subsection. Reasons are the Ministers justifications for acting (e.g. Sachs was fomenting hostility); information refers to the factual circumstances he was relying on to formulate his reasons (e.g. My sources tell me hes a stinky Communist who publicly claims that poor people have rights). In temporal terms, one must gather information first and then formulate reasons for acting on the basis thereof.5 Anyway, subsection (13) says that the Minister must give reasons, but that he only needs to give the information the disclosure of which in his opinion will not be detrimental to public policy. The phrase in his opinion is extremely subjective, so it is perhaps unsurprising that Stratford ACJ says the following:
The Minister complied with the request for reasons and information in a letter of the 10 November. We must, of course regard that letter as one written in perfect good faith. We are, therefore, bound to accept his statement that disclosure of information will be detrimental to public policy. *E+ven if the reasons seemed insufficient to other minds including that of the Court, the [reasonableness] test is not the one the Act requires. A credulous or over-apprehensive Minister may be satisfied of the existence or imminence of a
5

th

A good culture of justification will require all functionaries to informally note their reasons simultaneously with the making of all decisions, so that the reasons can always be provided if an interested party challenges their decision.

danger by reasons which would not move another to act. But so long as the reasons he gives are truly the reasons for his issuing the notice, he is not bound to have others or to give others.

Again this is a breathtaking abdication by the Court of its review role.

(d) Issues We have already touched on many of the major issues which emerge from the case. Nevertheless the following should be noted: o The influence of the implied incorporation test for audi alteram partem

As we saw, Stratford ACJ held that audi alteram partem is invoked only when it is impliedly incorporated by Parliament into the statute itself. This implied incorporation test was exploited by a narrow majority of the Appellate Division during the apartheid years so as to ban an organization which provided legal representation to criminal defendants. o Parliamentary sovereignty

There were no fundamental rights in the Constitution of that time, so Parliament was totally supreme.6 We will see during the course how important these issues relating to the separation of powers prove to be. In particular, judicial policy, i.e. a courts attitude to its own role, is extremely important in the development of administrative law. Administrative law is really the interface between the different branches of government. o Review and appeal

In Sachs the court viewed its own role as being limited to the law and not the facts. In other words, it had no power to inquire into the facts upon which the Minister based his decision; all it was empowered to do was see that the Minister was carrying out the laws which Parliament had made. This means they considered themselves to be a court of review in the narrowest sense.

Well, we saw that voting and language rights were included, but those dont have much relevance to administrative law.

10

INTRODUCTION TO ADMINISTRATIVE LAW

1 History (a) Administrative and constitutional law Administrative law is a branch of public law because it deals with the exercise of public power. But this does not mean that public power can be wielded only by public bodies. For a long time administrative law was not seen as distinct from constitutional law, because it was just one, relatively unremarkable way in which the constitutional principles of a legal system could be used to regulate the States conduct. There are two reasons why it took so long for it to emerge as an independent and important branch of the law. Firstly, executive discretion simply was not at all common until fairly recently. This was because governing a country was much easier in the past; political and commercial life was much slower and so legislatures were able to keep up with all important developments. This was the heyday of Parliamentary sovereignty. But, starting in the early 20th century, there was an explosion in the power of the executive, and a concomitant need for a branch of law which regulated the exercise of such power. (b) Dicey the mal poes The second reason is the thoroughly negative influence of a leading British constitutional lawyer of the late 19th century, Albert Venn Dicey. Diceys famous account of the rule of law in England consisted of three elements. Firstly, everyone must be subject to the law. This was a fairly novel idea at the time, since for much of the preceding millennium the monarch and certain noblemen were not subject to law. Secondly, everyone must be subject to the same, ordinary law; nobody must be given special treatment in special courts. At that stage, the military and clergy were tried in special courts, but Dicey nevertheless aspired to a fully unitary court system. Thirdly, and relatedly, the common law had been developed through ordinary Englishmen insisting on their common-law rights before the courts. Thus Dicey believed that the common law had developed organically and was part of the very fiber of the populaces collective being. This strengthened his conviction that the ordinary common law must be preserved, and that all must be made subject to it. Anyway, Diceys pernicious influence on administrative law was two-fold. Firstly, his conviction that everyone must be subject to the same, ordinary law meant that he rejected the French system which regulated the exercise of public power by setting up a separate court system solely to review the conduct of the executive (which is still the system used throughout continental Western Europe). Secondly, his account of the rule of lawappropriately, given that he was writing in the heyday of parliamentary sovereigntydid not include things which nowadays we think are fundamental: basic civil rights, democracy, an independent judiciary, and so on. There was no contradiction for Dicey between parliamentary sovereignty and the rule of law. In fact, he wrote at length about their symbiotic relationship. Because there was universal suffrage in England at the time (unless you were a woman), Diceys account of parliamentary sovereignty and the rule of law did make some sense in that context. But it became clear both how right and how wrong Dicey was when the English system was exported to South Africa in 1910. Parliamentary sovereignty was exported, but the rule of law was not: the franchise was very limited (only

11

white males and a select group of wealthy non-white males could vote) and our Parliament had no scruples about doing whack shit. 2 Accountability (a) Accountability and empowerment When we think of administrative law we think mostly about accountability, i.e. holding to account those who exercise public power by providing justiciable remedies for maladministration. This is reflected in the fixation upon judicial review. But there is another side to administrative law, namely empowerment. The law must give public functionaries the power to do a proper job of governing us. This empowerment aspect can be an important force for good, just like accountability. The French, for example, did not just set up courts to review the exercise of public power after the French Revolution; they also empowered bureaucrats to govern effectively so that they could replace the feudal order with an egalitarian one. During apartheid it was natural that accountability became the essence of South African administrative law. But surely government must now, under the new constitutional dispensation, be properly empowered so that it can go about redressing past injustices. Not enough attention is given to this aspect of administrative law, neither by the courts nor academe. (b) Mechanisms of accountability Nevertheless, how does accountability occur? Primarily it occurs through judicial review of administrative action. But there are also non-judicial forms of review. Parliament has extensive review powers, as do the Chapter 9 Institutions and certain appeal tribunals (e.g. the CCMA and the tax courts). (c) The culture of justification Etienne Mureinik, in a very influential journal article (discussed below) written during the transition, said that South Africa needs to move from a culture of authority, where you obey me because Im the boss, to a culture of justification, where you obey me because my decisions are well justified. Drop a Mureinik-bomb in every essay in this course for a guaranteed extra 5%.

Development of our administrative law (a) Judge-made/common law

Our system of administrative law is derived almost exclusively from English law, and as a result our pre-1994 administrative law was 100% judge-made. This meant that the law developed casuistically, i.e. on a case-bycase basis, rather than on a principled basis. The result of this was firstly that most unjust exercises of public power went unchallenged, because there was no interested party wealthy enough to bring the matter to court. And, secondly, the outcome of any particular case depended largely on the judge(s) who were assigned to it and/or the broad judicial policy of the time. This brings us to:

12

(b) The track record of the courts One can get a sense of a judges convictions by looking back at his historical track record of decisions. Obviously during apartheid courts were generally very executive-minded and formalistic in their application of the law. This was especially so when judicial activism was needed the most, viz. in the late 80s. This is a generalization, of course: every era had some great judges, and around the coloured vote crisis of the 1950s the Appellate Division as a whole was anti-executive and struck down a shitload of apartheid legislation (but in 1948 the mechanism for judicial appointments was changed, so the Court became more compliant). One important aspect of a judges track record is: (c) Judicial policy Even though there is never any set of policy guidelines circulated amongst judges, judicial policy is still a real phenomenon. Judges have certain views about the separation of powers and the proper function of a judge, shaped by the dominant intellectual trends of the time and by the judges personal convictions. These have a major impact on his decisions, especially his decisions in administrative law, which lies at the interface between the three branches of government. (d) The stretching of administrative law If one excludes the era of the coloured vote crisis, virtually every challenge to the exercise of public power came through judicial review of administrative action. So, for example, challenges to imprisonments without trial under the Suppression of Communism Act, the Sachs case, and challenges to race classifications were all challenged via administrative law (rather than, say, an application to invalidate the legislation, as often happens now). We will see repeatedly in this course how administrative law had to be stretched to try and fill the void left by the absence of a bill of rights. Taking all of what we have said together, we can see how the regulation of public power was developed in a haphazard manner. (e) Public and private power The balance between public and private power is always shifting. In the early 20th century the States powers were widespread: schools, hospitals, railways, ports, etc. were all private. But, largely due to the Great Depression, the State began taking control of more and more parts of the economy so that it could coordinate a stimulus. This trend was exacerbated by the two World Wars. By the 1960s, the influence of the State had reached its peak, with extensive welfare programs and many State jobs. In South Africa, of course, nationalization, the creation of a new labour regime, and the deployment of the State machinery to provide for its citizens were all skewed to benefit whites at the expense of other races. Anyway, privatization began during the 1970s and accelerated during the reigns of Reagan and Thatcher. South Africa picked up this trend shortly afterwards, with Telkom, Eskom, the Iron and Steel Corporation (Iscor), the Airport Company of South Africa (ACSA) and many others being placed in private hands. The process began towards the end of apartheid and continued seamlessly post-1994. What happens when privatization occurs? Does administrative law lose its right to interfere in functions which were privatized? And, if so, what takes its place? Contract and delict? If so, will those areas of law reform 13

themselves so as to take account of the immense, often monopolistic, power exerted by the private body? These are difficulties which administrative law needs to deal with and which we discuss in later topics. The UK government under Thatcher was wise to these problems. It therefore set up various administrative offices to regulate the private bodies which were now exercising powers previously held by the State. It also tried to ensure a degree of market competition in the privatized sectors, requiring for example that British Telecom allowed other companies to use its infrastructure at reasonable rates. Ultimately numerous mechanisms were created for review of a non-judicial kind. During the process of commercialization, i.e. the privatization of a public body, the company in question will begin to respond to the profit motive. The result is frequently that unprofitable branches or services of the company will be shut down and lots of employees will be fired. This is a particularly challenging period for administrative lawyers. Another issue we might want to consider is, What is public power? Of course this is vital because only exercises of public power are reviewable. There is simply no clear test to determine whether an exercise of power is public, although there are various indicators, e.g. whether it affects the public, whether it was undertaken in terms of authority given by Parliament, whether the person undertaking it is a member of the executive, whether its proclaimed in the GG, etc. But the basic point to take away is that the shifting public/private divide is very important in administrative law. (f) Growth of discretionary power As we have observed above, governance became immensely more complicated and public power rapidly expanded in the early 20th century. As Parliament was unable to keep up, it became necessary to empower other public actors to carry out the States functions. This led to a massive increase in conferrals by Parliament of discretionary powers to the members of the executive and other administrators. What we would now call framework legislationlegislation which creates the basic structure of governance which the executive is responsible for giving content tobecame very common. Thus Diceys vision of administrative law, based on Parliamentary pre-eminence, was rendered obsolete. The question for administrative lawyers is how does one regulate or keep accountable such exercises of discretion? Judges were initially very suspicious of any conferral of discretion, but ultimately adjusted to the reality that it was an essential aspect of governance. Are there limits to the exercise of discretion? Can it ever be truly unfettered, in the words of Stratford ACJ in Sachs? Even in the Sachs case limits were recognised, namely that the discretion had to be exercised honestly and intra vires. Post-1994, it is definitely not possible to have an unfettered discretion; the law will always constrain it. Dworkin describes the discretion as the hole in the center of a doughnut of restrictions. The question for us is How thick is the sweet, sticky ring of donutty goodness?

14

Some important miscellaneous concepts (a) The administration versus the executive

We often talk about administrative actions, executive actions and the like. Sometimes we use administrative and executive almost interchangeably, but they are distinct terms. There are many ways to conceive of this distinction. We can begin by noting that the members of the executive, i.e. ministers, directorgenerals, MECs, etc. are only a small sub-group of persons exercising administrative powers (the super-group includes the cunts at Home Affairs and other lowly functionaries). The executive members are the important mofos who the average citizen does not have much contact with; the administrators are the sorry fucks at Home Affairs. Another way is to think of a chain of delegation: Parliament delegates power via legislation to the executive, which in turn delegates power via regulation to the administration, which then gets on with the nuts and bolt of daily governance. A final way to think about it is in terms of the amount of discretion given: the executive gets a lot of discretion, because its (indirectly) elected members ought to be able to take account of policy goals in the formulation of broad objectives; but the administration gets virtually no discretion because its (unelected) members just do the humdrum tasks. Nugent JA picks up on this policy dimension to executive action by loosely defining administrative action in Greys Marine Hout Bay (Pty) Ltd v Minister of Public Works [2005] ZASCA 43 as follows. Whereas executive action tends to involve the formulation of policy, administrative action is:
the conduct of the bureaucracy in carrying out the daily functions of the state which necessarily involves the application of policy usually after its translation into law, with direct and immediate consequences for individuals or groups of individuals.

The public administration typically includes all the government departments and their employees, the public service (s 197 of the Constitution), local government administrations, the security forces, and parastatals, and excludes the Cabinet and the President. Nevertheless, there is no clear dividing line between the administration and the executive; there is a spectrum or continuum, with varying degrees of discretion. (b) Prerogative powers A more specific issue deserving attention here is that of prerogative powers. These are certain grand powers held by the members of the executive as a hangover of the royal prerogatives that existed in monarchies. A prominent example would be presidential pardons. Prior to 1994 these powers were unwritten, but now they are encapsulated in s 84(2) of the Constitution. Again, the issue for administrative law is How do we regulate the exercise of these powers? The case of President of the Republic of South Africa and Another v Hugo 1997 (4) SA 1 (CC) is a major milestone in this regard, because it established that the exercise of these powers is reviewable. The case dealt with Mandelas decision to pardon all female prisoners with children under 12 years old. Hugo was a male prisoner with young children, and he claimed that the pardon unfairly discriminated on the basis of gender. The majority of the Constitutional Court denied that this was the case, although the key for our purposes is that the Court was willing to review the exercise of a prerogative power despite the fact that prior to 1994 such powers were unreviewable.

15

(c) Classification of functions One will often see in administrative law cases evidence of the classification of functions approach. This is an extension or application of the separation of powers doctrine. A court, when exercising its review power, will look at the authority or power that is being exercised and classify it (whether expressly or impliedly) as being judicial or legislative (or quasi-judicial or quasi-legislative if you want to get fancy). It might seem strange to classify administrative action into judicial or legislative kinds. But the former just means what happens, for example, at an administrative tribunal; and the latter refers to cases where a Minister has issued regulations or something. Having made its classification, the court will then tailor its review procedure and the resulting remedy to the class of action it is dealing with. Nominally the classification of functions approach to administrative law has now been abandoned, but old habits die hard and one sees its ghost emerging in contemporary judgments. (d) Ultra vires, ousters, standing, grounds of review, and reasonableness Here follows a cursory glance at some other key concepts in administrative law (most of which weve encountered previously). The ultra vires doctrine is the foundation of judicial review of administrative action before 1994. As we saw in the Sachs case, traditionally all the grounds of review were subsumed into this doctrine. Nowadays we recognise that the various grounds are quite separate. Ouster clauses (a.k.a. privative clause) are clauses in legislation which (attempt to) oust the jurisdiction of courts to review administrative actions undertaken in terms of that legislation. They are a feature of all Commonwealth systems, including our own. Standing, i.e. the issue of who is entitled to sue in respect of a particular administrative action, is a crucial issue in administrative law. If a person is prevented from challenging an administrative action in court, that is of course a massive obstacle to his achieving administrative justice. This is exactly what the stringent direct personal interest standard did in the past. But now the Constitution s 38 has blown this out of the water. Reasonableness was not a ground of review until 1994. There was also no common-law obligation on any administrator to give reasons for his decisions (although statute would sometimes alter this, e.g. the Riotous Assemblies Act as described in the Sachs case). This is still the common law of England! 5 Review versus appeal (a) The JCI case Administrative law is preoccupied with judicial review.7 But it never deals with appeal, which is considered to be an unjustified infringement by judges of the separation of powers. But what is the difference between review and appeal? This brings us to the case of Johannesburg Consolidated Investment Co. v Johannesburg Town Council 1903 TS 111.

Remember that in administrative law we are not really interested in judicial review of legislation, so that is ignored in this discussion.

16

The facts were as follows. Valuation courts had been set up by the Transvaal government under Proclamation 38 of 1902 to determine the tax which various legal persons had to pay, based on the property they owned. The valuation courts were not really courts; they were administrative tribunals to which persons could go to contest their tax valuations. (It turned out to be a crucial fact that the valuation courts did not keep a record of their proceedings.) JCI, the applicant, was not happy with the assessment which the valuation court had made. It therefore went to the Transvaal Supreme Court under s 10 of the Proclamation and asked it to review the valuation courts decision, set it aside, and replace it with a lower valuationin other words, JCI wanted the Court to hear the matter de novo. The key question was what the Proclamation meant when it said that a court could review the decisions of the valuation courts: did the Proclamation mean to empower the Transvaal Supreme Court to hear the matter de novo? Innes CJ for the majority reasoned that review could mean one of three things: 1. Review of a lower courts proceedings We are very familiar with this, where a higher court checks on the validity of the proceedings of a lower court. 8 Innes CJ describes review in this sense as follows:
In its first and most usual signification it denotes the process by which, apart from appeal, the proceedings of inferior Courts of Justice, both Civil and Criminal, are brought before this Court in respect of grave irregularities or illegalities occurring during the course of such proceedings.

So the grounds of review here are grave irregularity or grave illegality. 2. Judicial review of administrative action This is of course what we are focusing on in this course. Innes CJ describes it as follows, in a passage which is to this day quoted to describe the process of judicial review of administrative action:
Whenever a public body has a duty imposed upon it by statute, and disregards important provisions of the statute, or is guilty of gross irregularity or clear illegality in the performance of the duty, this Court may be asked to review the proceedings complained of and set aside. This is no special machinery created by the Legislature; it is a right inherent in the Court, which has jurisdiction to entertain all civil causes and proceedings arising within the Transvaal.

Here the grounds of review are gross irregularity or clear illegality. Note that Innes CJ stresses that the right to review administrative action is not conferred by statute; it is a consequence of courts inherent common-law jurisdiction. The ultra vires doctrinewhich, as we will see, is based on a particular understanding of the separation of powers and the rule of lawis the justification for this inherent power. 3. Statutory review This is a special kind of review, which arises not from the courts inherent jurisdiction but from an express conferral by Parliament. Says Innes CJ:
8

Nowadays we can divide such review into two kinds: automatic review, where a statutory trigger makes the lower courts decision automatically subject to confirmation by a higher court (e.g. the imposition of a life-sentence by an inferior court must be confirmed by a superior court); and review initiated by one of the parties, alleging bias, corruption, or some other irregularity.

17

[T]he Legislature, when it conferred powers of review upon the Court, meant by the use of that word to denote a wider exercise of supervision and a greater scope of authority than those which the Court enjoyed under either of the other existing species of review.

What exactly is the difference between appeal and review? Innes CJ had previously dealt with the difference in the context of review of a lower court, viz. review of the kind labelled 1 above. He said in that context that:
[Review] differs in several respects from the procedure by way of appeal. Notably there is this distinction between the two, that an appellant comes into court upon a record of the case in the court below, and by that record, he is bound; he cannot take advantage of any circumstance which does not appear upon or cannot be deduced from the record. The litigant who seeks to have a case reviewed depends upon irregularities which need not necessarily appear upon the face of the record. If they do not so appear, he is at liberty by affidavit to bring the facts upon which he relies to the notice of the Supreme Court. He is not bound by the record in the way in which an appellant is.

In short, then, in an appeal one is bound by the record in the sense that the court a quos findings on the facts will not readily be overturned. But in a review the litigant may bring evidence of irregularities not evident upon the record. The difference between the two, then, is the extent to which one is bound by the record. But review of lower courts is not what we are really concerned with; we want to know the difference between appeal and review in the context of administrative law. Innes CJ does not say anything about the record under the second type of review, although the record is clearly important in judicial reviews of administrative action: the applicant must notify the administrator whose decision is being reviewed, and he has 15 days to provide the record, i.e. the entire body of relevant documentation, so that it can be scrutinised by the court. As to the third type of review, Innes CJ says essentially that this broad kind of review can almost amount to an appeal, and then proceeds to explain what review means by bearing this in mind:
So employed the expression "review" seems to mean "examine" or "take into consideration." And when a court of law is charged with the duty of examining or considering a matter already dealt with by an inferior court, and no restrictions are placed upon it in so doing, it would appear to me that the powers intended to be conferred upon it are unlimited. In other words it may enter upon and decide the matter de novo. It possesses not only the powers of a court of review in the legal sense, but it has the functions of a court of appeal with the additional privileges of being able, after setting aside the decision arrived at by the lower tribunal, to deal with the whole matter upon fresh evidence as a court of first instance.

So we see that Innes CJ regards this special form of review as being very close to appeal (although he stresses later that a real distinction remains),9 in that the court is given the sweeping powers to hear the matter de novo and to replace the previous decision. Having established the three possible definitions of review, Innes CJ reasons as follows. Firstly, the Proclamation definitely does not mean review in the first sense, because the Valuation Court is not actually a court of law; it is an administrative tribunal. That leaves review of types 2 and 3. He concludes that it cannot be the broad review powers implied by type 3. Why? Basically Innes CJ says that if the Proclamation meant to

Such broad review powers, by the way, are conferred by statute; courts may not, at common law, engage in this sort of review (and nowadays Parliament would be very unlikely to grant courts such powers).

18

confer these broad powers then it would put in place measures so as to make the exercise of such powers practically possible. But in fact:
[U]nder the Proclamation as it stands the Legislature has deliberately inserted provisions which make an appeal practically impossible . [W]e find it laid down that the Valuation Court need not keep any record of its proceedings other than the assessment, objection and finding in regard to each objection. As the law stands the Valuation Court is absolutely justified in simply noting the assessment and objection thereto and its own decision; that is to say in practically keeping no record at all of the evidence or the proceedings generally.

So the Proclamation allows the Valuation Court to dispense with the keeping of a record of its proceedings. Innes CJ then, crucially, says the following:
Now an appeal10 without a record is a contradiction in terms. I cannot think that the Legislature, after doing away with the necessity for keeping a record, could have intended to provide for an appeal from the Court's finding.

So Innes CJs essential line of reasoning is this: If the Proclamation meant to confer the broadest form of review power and thus to allow the Court to hear the matter de novo, it would give the Court enough information on the basis of which to do so; but by dispensing with the record-keeping requirement the Proclamation makes this impossible. Therefore the Proclamation could not have meant to confer the broadest form of review power. This leaves us with the conclusion that the Proclamation meant to confer the second type of review power, which does not allow the Court to provide the applicant with the relief claimed. On this basis JCIs claim was dismissed. (b) Wide and narrow appeal Hoexter discusses basically this same issue with reference to the position in modern South African law. She draws a distinction between wide and narrow appeal. Wide appeal consists of the complete rehearsing and redetermination on the merits of a case, with or without additional evidence or information. Here the appellate body will not be confined to the record of the proceedings of the body being reviewed. In a narrow appeal, on the other hand, the appellate body will be restricted to the record of those proceedings, and is empowered only to investigate irregularities appearing from that record. Basically, wide appeal is like the normal meaning of the word appeal, and a narrow appeal is like review. So this is just the mirror-image of what Innes CJ had said: narrow review (type 2 above) is like the normal meaning of the word review, and a wide review (type 3 above) is like appeal. The reason Hoexter refers to both as appeal is because she, like Innes CJ, is trying to address a very specific problem: legislation sometimes allows a court or other body to hear an appeal from an administrative decision. This requires a court to interpret this term (in the same way that Innes CJ had to interpret review) so as to determine how extensive the powers are which the legislation confers. Basically, the appellate body needs to determine whether its allowed to rehear the decision de novo and pronounce on its merits. The following factors point to wide appeal:

10

The reason he is referring to an appeal is because he has already noted that the third meaning of review almost amounts to an appeal.

19

1. Lack of a record: where no record is provided to the appellate body, that is an indication that the appellate body must rehear the matter de novo rather than simply investigate irregularities appearing from the record. Note that this is precisely the opposite of what Innes CJ said in JCI. He held that because there was no record the Court had no evidence on which to base a de novo tax valuation, and therefore it only had narrow review powers. Hoexters point is thatassuming the appellate body does enough evidence to hear the matter de novothe absence of a record is an indication that that body can do a lot more than just scrutinize the record for irregularities. 2. Wide procedural powers: where the powers of inquiry of the appellate body are identical to those of the administrator, that is an indication that it can hear the matter de novo. 3. Wide decisional powers: where the appellate bodys decision is deemed to be that of the administratoras opposed to its substitution, confirmation, variation or setting asidethat is an indication that it can hear the matter de novo. (c) Summary The distinction between appeal and review is mutable, contested and difficult. But the distinction is crucial in administrative law: judges still say that they are not prepared to hear appeals from administrative actions, so it becomes vitally important to the whole practice of administrative law to know where review ends and appeal begins. The basic distinction, as implied by Innes CJs judgment, is that reviews deal with process or form and appeals deal with facts or substance. This is why appeal is more intrusive and why classically-trained lawyers are sceptical of itit involves basically second-guessing the substantive merits of administrators decisions. Post1994 our administrative law occupies an intermediate ground between the two. Although it is spoken of (largely for good reason) as judicial review, the reality is that as soon as reasonableness enters the picture then review has to include an engagement with the facts.

20

ADMINISTRATIVE LAW AND DEMOCRACY

1 Mureinik, A Bridge to Where?

Mureinik builds on the bridge metaphor which had been used in the epilogue to the Interim Constitution. He asks where that bridge is supposed to lead. His answer is that it is a bridge from a culture of authority, in which decisions are respected because the authority of the decision-makers is feared, to a culture of justification, in which decisions are respected because they are well justified. Judges, even in foreign jurisdictions, have gone wild for Mureiniks lucid transformative vision and have repeatedly cited this article in their judgments. The remainder of the article assesses the contents of the transitional Bill of Rights against the ideal of a culture of justification. We are concerned specifically with his critique of s 24 (the text of which is given in the next topic) in bolstering a culture of justification. He argues that justifiability review (as in s 24(d)) is equivalent to reasonableness reviewalthough Corder, as is indicated in the next topic, thinks that because the administrator is able to choose the reasons which must be held to justify the decision, this gives the administrator an advantage and separates justifiability review from reasonableness review. Mureiniks point is really to persuade courts that they ought to interpret justifiable broadly so as to essentially mean reasonable. In a footnote, he explains that there were serious fears during the negotiations that reasonableness review would be used to usurp and frustrate the powers of the executive. He suggested the term justifiable to the drafters as a term which was essentially equivalent to reasonable but without the latters worrying connotations. Mureinik also notes that the right to access to information is an important adjunct to the right to administrative justice and a cornerstone of the achievement of a culture of justification: A government which can close its files, he says, will be under much weaker pressure to justify its decisions than one which has to open them. 2 Davis, Administrative justice in a democratic South Africa

At the outset Davis notes that the major challenge of the State post-1994 will be the redress of socio-economic disparities. This, he says, will likely lead to a rapid expansion of administrative power and a concomitant need for mechanisms of accountability He identifies two broad conceptions of democracy. Firstly, there is majoritarian democracy, or democracy as process, which sees democracy as simply having free and fair elections. Secondly, there is primary goal democracy, which sees democracy as the achievement of a just distribution of resources. Davis prefers primary goal democracy, which he sees as more integrative. It is based, he says, on three Dworkinian principles: participation (everyone has the opportunity to get involved in decisions), equal stake (everyones preferences are weighted equally), and independence (everyone has moral responsibility for the decisions they prefer). Majoritarian democracy sees judicial review as counter-majoritarian and therefore undesirable. Primary goal democracy sees judicial review as desirable insofar as it assists in the achievement of the goal of a just distributionand the remainder of the article basically spells out when this will be so.

21

The elephant in the room throughout this article is the widespread fear when Davis was writing that the wealthy and white would privatize apartheid post-1994 by exploiting the guarantees of administrative law so as to derail or delay RDP. Thus there is a tension in Davis, between the need to redistribute (and quickly) and the need for reasonableness, due process, and integrative democracy. This is why Daviswhile on the one hand clearly thinking administrative law was importantis unable to embrace administrative law and judicial review unequivocally. Dicey argued that a democratically elected Parliament was all one needed to ensure perfect participation and accountability, because then Parliament would enact laws in conformity with the will of the people and would control the executive. Diceys model of democracy regards administrative agencies with suspicion, since their regulation dilutes the supremacy of the ordinary, democratic law. So, much like majoritarian democracy, there is little room for administrative law or judicial review in such a model. But Davis argues that a radical shift away from Dicey and Westminster is required, thus carving out room for administrative law to operate. He argues that the task of a system of administrative law is not to eradicate but rather to strengthen a popular form of accountability, and thereby to ensure greater public participation and control over the exercise of public power. It must promote the three principles of democracy outlined above. Judicial review must also promote participation such that our democracy is one that maximises the space for debate around social, political and moral issues. He says that an entrenched bill of rights is a good foundation upon which to build such a form of democracy. Judicial review can ensure that administrative decisions are justified, that they foster participation, that they are rational, and that they are committed to the values entrenched in the bill of rights. Davis then points to the limitations of the judicial role, even on this enlightened model of democracy: firstly, the boundary between law and politics is not clear, and so judges might find themselves entering into political decision-making; and, secondly, although reasonableness review is important judges must beware the separation of powers and their boundary with other branches of government. He says that judges are not well placed to interfere in the content of administrative decisions, especially because in modern governance such decisions are so incredibly complicated and rely on technical expertise. (One can detect a similarity here with the common problem of polycentricity in the adjudication of socio-economic rights. Davis, not coincidentally, was opposed to the inclusion of socio-economic rights in the Bill of Rights.) He suggests that the obvious legalistic solution to these problems would be to set up administrative tribunals which can review decisions but have access to more specialised knowledge than judges. But, he argues, this is not enough. The real key is not ex post review of administrative decisions by legalistic means; the key is ensuring that citizens participate in decision-making ex ante and are able to hold government accountable through democratic organizing and civil society. 3 Mureinik, Reconsidering review: participation and accountability

This Mureinik article overlaps substantially with Davis, although its less overtly political. Up front Mureinik says that snapshot democracy is not enough, and that government must be responsive at all times, not just once every five years. Responsiveness has two components: participation (being able to affect the content of decisions which affect one) and accountability (the justification of decisions by government to those who they affect). Administrative laws basic purpose is to ensure these two goals.

22

As to participation, he talks about the deprivation and determination theories. In terms of the former, one only has a right to be heard if one can prove that one will be deprived of ones existing rights by administrative action. In terms of the latter, one only has a right to be heard if one can prove that ones rights will be determined by administrative action. The latter is better, because the deprivation theory is so narrow and thus deprives too many people of a right to be heard and to participate. But the unrefined determination theory is perhaps too broad, having the potential to be too onerous and thus to frustrate effective government. It must therefore only be the starting point: it must furnish one with a prima facie right to be heard, subject to the governments showing good grounds not to hear you. Mureinik calls this alternative approach the provisional determination theory. The courts have been trying to find a similar middle ground using the doctrine of legitimate expectations. Given our past, much can be done to make government more participatory. The key thing is notice-andcomment procedures prior to the enactment of all subordinate legislation. The danger, though, is that the wealthy will capture the benefits of this processso again there is the elephant in the room. The solution, for Mureinik, is to have effective Chapter 9 Institutions and a vibrant civil society. Generally, we must create organs that can intervene on behalf of the under-resourced and poorly-mobilised constituencies. As to accountability, the key is that the administration must give reasons for its decisions. This is to foster the oft-mentioned culture of justification. What is also needed is rationality review, for which there is a threefold test:
(a) the decision-maker has considered all the serious objections to the decision taken, and has answers which plausibly meet them; (b) the decision-maker has considered all the serious alternative to the decision taken, and has discarded them for plausible reasons; and (c) there is a rational connection between premises and conclusion: between the information (evidence and argument) before the decision-maker and the decision that it reached.

Obviously review for rationality allows really bad decisions to be invalidated. But, more importantly, the decision-maker will likely invest greater effort during the decision-making process if he anticipates being reviewed later on grounds of rationality. This makes for better decisions and is good for participation, as decision-makers seek to foster public participation before making them. On the basis that rationality review is more restricted than simply asking, Is the administrations decision correct? Mureinik defends rationality review from those who argue that it intrudes too much onto the executives terrain. Ultimately there is a tight connection between administrative law and democracy. Also note the fears and hopes for administrative law in 1993. Have these materialised since then? Can we say in hindsight that there were more important things to worry about? 4 Corder, Administrative justice: a cornerstone of South Africas democracy

This is a more modest retrospective on how our administrative law had developed between 1994 and 1998 (when the article was written). It tried to build on the notions of participatory democracy, now safe in the knowledge that a right to administrative justice had been included in the final Constitution. Corder considers his most important points to be the following. 23

Firstly, he quotes Hutchinson, who says that there is no atheoretical way to do administrative law. Everyone has certain theories of State which impact on their understanding of administrative law, and when these are unconscious they are only more potent. Secondly, he refers to the core values of administrative justice: the entrenchment of openness, responsiveness and accountability as values basic to our constitutional democracy; and the strengthening and broadening of the entrenchment of the fundamental rights of access to information and to administrative justice. Then in the second section of his article, Corder attempts to shape the drafting of the legislation meant to enable the s 33 right (what became PAJA) by setting out the key issues which need to be addressed by such legislation. Firstly, the legislation must define administrative action, which has the effect of determining the ambit of the right to administrative justiceonly if the decision amounts to administrative action do any of the protections kick in. (Because Parliament ultimately rejected the SALRCs draft definition and used an incredibly convoluted one, the definition of administrative action is now a massive issue in our administrative law.) Secondly, administrative tribunals must be created to ease access to justice and take a load off the high courts. (This suggestion was also rejected by Parliament.) Thirdly, the administration (and even the population in general) must be trained to understand the rights and requirements laid down by administrative law. The final section deals with scrutiny of subordinate legislation. Corder argued that PAJA must prescribe minimum standards for subordinate and delegated legislation in relation to administrative justice. If you lay good guidelines, you can pre-empt many later administrative justice issues. One of the best foreign developments in this regard is having a standing Parliamentary committee to scrutinise every single piece of subordinate legislation (of sufficient importance). Usually an academic or independent lawyer is employed to do the scrutiny. Such a mechanism was included in the SALRCs draft legislation, but Parliament just deleted it. 5 Davis, To defer and then when?

The notion of deference is fundamental to all of administrative law. As we have said, administrative law occupies the interface between the branches of government, and so a judges views on the separation of powers and the extent to which he should defer to the other branches will always influence his administrative law decisions. (Deference should also be shown by the other branches of government towards the judiciary, but sadly certain politicians have been utterly contemptuous towards the judiciary and as such threaten to undermine our entire constitutional projectwhich of course relies on the separation of powers, as do almost all modern constitutions. Respect is a two-way street bru!!!!!!!!!) Under the Westminster system, Parliament was the ultimate lawful authority. Over the course of the 20th century, the executive rose to prominence in fact if not in law. Prior to 1994, the judiciary tended to defer to these two branches of governmentdeferring in the sense of submission rather than respect. This obviously had serious negative consequences, and as a result there is controversy and indeed resistance to the idea of deference even post-1994. This is the context in which Davis wrote this article. The central question which Davis tries to address is, What are the proper boundaries between the different branches of government, and who decides these? It is clear from the title of his article that Davis certainly doesnt think that deference is always appropriate. Davis approvingly cites Hoexters definition of deference:

24

A willingness to appreciate the legitimate and constitutionally-ordained province of administrative agencies; to admit the expertise of those agencies in policy-laden or polycentric issues; to accord their interpretations of fact and law due respect and to be sensitive in general to the interests legitimately pursued by administrative bodies and the practical and financial constraints under which they operate.

Judges need a theory of deference to inform their approach to the other branches of government: deference is sometimes appropriate but sometimes not, and the distinction must be drawn on a principled basis. Davis takes issue with the two judgments handed down by the Constitutional Court in Bato Star. Two judgments were delivered, by ORegan J and Ngcobo J, in which all the other members of the Court concurred. He is especially critical of ORegan Js understanding of deference, which she expresses as follows:
In treating the decisions of administrative agencies with the appropriate respect, a court is recognising the proper role of the executive within the Constitution. In doing so a court should be careful not to attribute to itself superior wisdom in relation to matters entrusted to other branches of government.

He suggests that her judgment is indicative of the absence of a coherent theory of judicial review. Her understanding of deference lacks substance and that it is too deferential to the executive. Insofar as she is more specific, it amounts to, basically, Dont interfere in complex decisions. Ngcobo J, on the other hand, focuses on the constitutional framework and the transformative objectives of the Act under discussion. This suggests that he was taking a more robust and well-theorized approach to deference but, in application, he lamely says there is not enough evidence to make a proper assessment. The upshot is that no coherent theory of review emerges from the judgment. He then tries to set up a better model of deference, in reliance on Dyzenhaus, who identifies three separate approaches: 1. Law as authority (a perversion of Bentham). On this view deference is almost invariably desirable because majoritarianism (realised, of course, through Parliament) is seen as the ideal. 2. Law as neutrality. On this classical liberal view deference is appropriate to protect individuals rights from State interference. 3. Law as justification (Mureinik). On this view deference is appropriate where it serves the goals of participation and accountability. Davis prefers approach 3, and understands the judiciarys role accordingly. Judges should not be too deferential, but the reason is not because rule by juristocracy is desirable, but rather because judges have an indispensible role to play in realising a culture of justification. While the separation of powers is important, it should not be allowed to undermine the values of participation and accountability. Davis says that the primary challenge is not to determine the level of scrutiny to be applied in each case. Rather, the fundamental question is how are we even to make any such determination? What are the criteria we should be considering when we think about whether to defer? The criteria, according to Davis, emerge from the constitutional enterprise as a whole. This first of all means that judges must foster participation and accountability. He quotes with approval Justice Breyer of the US Supreme Court, saying that the judiciary must not just preserve civil liberties, but also afford citizens the freedom to participate in self-government as equal citizens. Secondly, judges, as the guardian of all the fundamental rights of the citizens, must not defer when such rights are at issue. 25

On the question of accountability, Davis analyses two cases. The first is Pharmaceutical Manufacturers, which requires a decision to be rationally related to the purpose for which the power was given. Secondly, Grootboom requires the higher threshold of reasonableness. Davis says courts need to decide on a threshold or level of scrutiny appropriate to review the decision before them. If the rights at issue are fundamental, then the courts must intensify their scrutiny. But courts must at the same time be careful not to impose their own view of the best decision. He then goes on to stress that the principle of participation should yield far less to deference than the principle of accountability. He says what is required of the court is to evaluate the weight of the affected right, the nature of the evidence provided to justify the determination or demarcation of the right, and from this preliminary conclusion examine the extent to which its role in the constitutional scheme justifies interference. He says this is in line with a determination theory as opposed to a rigid deprivation theory. Davis then examines the attitude of Canadian courts, who have developed a nice, rigorous account of deference. This is that a court should only defer where the reasons provided for the policy can be justified on the basis that the measure must be consistent with democratic values and it must be necessary in order to maintain public order and the general well- being of citizens. In other words, any legislation is an expression of certain social policies, and judges must hold that legislation and its underlying policies to the fundamental commitments of their society as expressed in the Constitution. Deference is justified only where the policy goals provided for in the legislation are in conformity with constitutional values. Davis repeats that ORegan J is not doing the good kind of deference in Bato Star. She seems to simply be saying that judges must defer to government when the government has a lot of expertise. But Daviss account of deference does not permit this: it could be that although the decision requires expert knowledge and skills, it engages rights issues and therefore the courts are not justified in deferring. For example, the courts may be poorly placed to assess the effectiveness of the mechanisms that the government is using, but they can still interrogate the permissibility of the social policies which underlie those mechanisms, the values of which those mechanisms are an expression. Those policies and values must be reviewed for compatibility with constitutional commitments. Then Davis turns to the case of New Clicks (which we will do in painstaking detail later in the course), especially the judgment of Chaskalson CJ. At the start of his judgment, Chaskalson CJ spells out what seems to be a weak, formalistic account of deference, suggesting that he is going to pass the buck and point to the separation of powers to justify inaction. But in application he actually begins to weigh up the validity of the relevant administrative action with reference to constitutional imperatives, and on this basis finds the action invalid. The judgment of Moseneke J is quite lame, and doesnt interrogate the States constitutional duties. He concludes by saying that, given the case law, there is uncertainty regarding the proper role of the courts and the intensity of review that is justified.

26

THE CONSTITUTION, PAJA AND THE COMMON LAW

1 The Interim Constitution (a) Drafting history In the Multiparty Negotiating Process, aimed at drafting the Interim Constitution, it was agreed by all parties that in the spirit of never again a right to administrative justice was necessary. From the outset, therefore, all parties agreed that such a right must be included, along with a right to access to information, in the Interim Bill of Rights. But the rights content was exceedingly controversial. The result was a prolix section which bore the hallmarks of design by a discordant committee. (b) Section 24 Section 24 of the Interim Constitution reads:
24 Administrative justice Every person shall have the right to(a) lawful administrative action where any of his or her rights or interests is affected or threatened; (b) procedurally fair administrative action where any of his or her rights or legitimate expectations is affected or threatened; (c) be furnished with reasons in writing for administrative action which affects any of his or her rights or interests unless the reasons for such action have been made public; and (d) administrative action which is justifiable in relation to the reasons given for it where any of his or her rights is affected or threatened.

This right was the basis of administrative law in South Africa from 27 April 1994 to 3 February 2000 (for reasons which will be explain below). (c) Threshold requirements One of the most glaring complexities in s 24 is that of thresholds, the prerequisites which must be met before the right to administrative justice applies. Each subsection has its own threshold requirements. The effect of these are summarised below. Bear in mind that interests are clearly weaker than rights; that legitimate expectations occupy an intermediate ground between the two; and that when ones rights etc. are threatened then they have not yet been affected but are potentially going to be affected. If ones rights are affected one can rely on subsections: If ones rights are threatened one can rely on subsections: If ones interests are affected one can rely on subsections: If ones interests are threatened one can rely on subsection: If ones legitimate expectations are affected: (a) (a) (a) (a) (a) (b) (b) (c) (c) (b) (c) (d) (d)

Clearly the thresholds which apply can have a very serious impact on administrative law; they present a formidable obstacle to the accessing of the right to administrative justice. Although (as we will see below) the final Constitution has dispensed with these complicated threshold requirements, they are not of merely 27

historical or academic interest: terms like interests, legitimate expectations and so on continue to be used post-2000. (d) Key terms There are some terms which would need to be defined in the application of the section. Some of these continue to be relevant under the final Constitutions right to administrative justice. o Lawful

There are wide and narrow concepts of ultra vires. The narrow version holds that lawfulness review is concerned exclusively with the language of the authorising statute. Thus the administrative action will be lawful if it complies with the language of that statute. The wide version holds that lawfulness review is concerned also with common-law requirements like audi alteram partem and certainty. This version therefore encompasses far more bases for lawfulness review, many of which come from the common law rather than the instructions of Parliament. The question, then, is Does s 24(a) of the Interim Constitution refer to the wide or narrow concept of lawfulness? The answer to this question will of course fundamentally determine the nature of judicial review of administrative action. One could argue that a narrow concept is meant, because procedural fairness and justifiability were covered separately in ss 24(b) and 24(d). o Administrative action

Determining the meaning of administrative action is of course absolutely crucial to administrative law; only if something is defined as administrative action will administrative law have any application. We can see that s 24 indeed has no application except where administrative action is involved. What is this term likely to mean? By looking at the legislative history of the sectiona legitimate method of constitutional interpretation according to S v Makwanyanewe can see that the term administrative action was meant to be as broad as possible. o Procedural fairness

What rules, principles and requirements does this term encompass? It is a very flexible term which would need to be delineated in application. o The right to reasons

Subsection 24(c) creates a right to receive reasons for an administrative action. But when exactly would one be entitled to such reasons? It is unlikely that one would always be entitled to them. Note, for example, the escape clause in the subsection relating to cases where the reasons have been published. And would an individual be entitled to such reasons if an application, e.g. for a tender or government job, has been rejected?

28

Justifiability

Section 24(d) requires the administrative action to be justifiable in relation to the reasons given for it. This raises two questions. Firstly, what reasons are these: those given by the administrator at the time the decision was made, those furnished upon application in terms of subsection (c), or those given at a later stage (e.g. when a review is argued in court)? Secondly, is justifiability simply the same as reasonableness? Mureinik argued that it was the same (with the differing terminology a crude sop to those who thought reasonableness had worrying connotations). But, on the other hand, the section requires justifiability in relation to the reasons given for the administrative action, thereby allowing the administrator to choose the scope of any justifiability review; he gets to decide on the reasons which must be held to justify the action. This would then be a basis for saying that justifiability is not the same as reasonableness (which is not confined to the reasons actually given for the action). In this regard, it is worth noting that there was a real concern at the negotiations that the introduction of s 24 would impose an unreasonable burden on the public administrationespecially given that it was historically closed and unaccountable, and likely to be resentful of any regulation by the new government. (e) Miscellaneous Note that the Interim Constitution had an entrenched right of access to courts, making ouster clauses invalid. Mureinik reasoned that the same was implied by s 24(a). Note also the incorporation of liberalised rules of standing to sue (now embodied in s 38 of the final Constitution). (f) Impact on the common law In early judgments, courts were generally very reluctant to engage with the new s 24 right. Courts only began to map out in earnest the scope of the section in about 1998. 2 The final Constitution (a) Drafting history The big issue for administrative lawyers during the drafting of the final Constitution was, simply, Would the right to administrative justice be received from the Interim Constitution? The Constitutional Principles required that universally recognised rights be entrenched in the final Constitution. The ANC-dominated transitional government, which had begun to find the right to administrative justice burdensome, reasoned that the right to administrative justice was not universally recognised (the Namibian constitution was, in 1990, the very first to include it) and therefore ought to be excluded. But, following an outcry from civil society and opposition parties, the ANC reluctantly agreed to include itin a somewhat compromised form. (b) Section 33 The text of s 33 of the final Constitution reads:

29

33 Just administrative action (1) Everyone has the right to administrative action that is lawful, reasonable and procedurally fair. (2) Everyone whose rights have been adversely affected by administrative action has the right to be given written reasons. (3) National legislation must be enacted to give effect to these rights, and must(a) provide for the review of administrative action by a court or, where appropriate, an independent and impartial tribunal; (b) impose a duty on the state to give effect to the rights in subsections (1) and (2); and (c) promote an efficient administration.

(c) The requirement for the drafting of legislation The compromise referred to earlier is reflected in s 33(3): The right is ultimately to be defined in national legislation, giving the ANC government added control over this process. Schedule 6 Item 23 made transitional arrangements as follows: Parliament had to enact the legislation referred to in s 33(3) within three years of the date on which the new Constitution took effect, that is, by the 4th February 2000. 11 Until that legislation is enacted, ss 33(1) and (2) must be read not as they appear in the text of the final Constitution, but rather exactly as s 24 of the Interim Constitution reads. Thus s 33 was suspended until either PAJA was adopted and in force or 4 February 2000 arrived. That is why we said earlier that the right in the Interim Constitution was operative from 27 April 1994 to 4 February 2000. (d) The promotion of an efficient administration The real sting in the tail is perhaps not the mere fact that national legislation was to define the right, but rather the contents of s 33(3)(c). This makes the efficiency of administration the litmus test for the constitutionality of the enabling legislation. There are two ways one could interpret this. The first would be to say that an administration which acts in accordance with the principles of administrative lawtaking actions which are lawful, reasonable and procedurally fair; facilitating participatory democracy; providing reasons for their actions; and so onis for that reason efficient. The second would be to recognise that the public administration is grossly overburdened and so the drafters of the enabling legislation must be given some leeway to limit the right to administrative justice so as not to ask too much of the administration. Unsurprisingly, it is this second rationale which in fact motivated the drafters of the Constitution to include s 33(3)(c). 3 PAJA (a) Drafting history In November 1998 the South African Law Reform Commission was asked to draft the enabling legislation, which later came to be known as the Promotion of Administrative Justice Act (PAJA). The SALRC appointed a team consisting of Jeremy Gauntlett, Cora Hoexter, Phillip Iya, and our very own Uncle Hughie. With the assistance of ze German Technical Co-operation, the draft was completed by August 1999. This draft was workshopped for comment in four cities in the country, and finally submitted to Parliament.
11

The same was true of the legislation referred to in s 9(4) (now PEPUDA) and s 32(2) (now PAIA). These three pieces of legislation, which were specifically required by the Constitution, are sometimes said to occupy a higher tier in the legislative hierarchy than normal legislation. The merits of this view are disputed.

30

Parliament had various hearings after which it threw out the most brilliant aspects of the draft legislation and introduced some stupid complicating things. It was then signed into law on the 3 February 2000, with less than 24 hours before the constitutionally-imposed deadline. But PAJA only came into force on 30 November 2000 (primarily because time was needed to draft the associated ministerial regulations). The result was an peculiar interregnum where ss 33(1) and (2) constituted our administrative law: s 24 of the Interim Constitution had fallen away in accordance with Schedule 6 Item 3, but PAJA was not yet operational. (b) PAJA section by section As we have noted, the definition of administrative action is the gateway to all the protections of administrative law; only if the conduct in question amounts to administrative action are any administrative justice rights applicable. Thus PAJAs definitions section, s 1, is vitally important. According to the nauseatingly long definition of administrative action provided therein:
Administrative action means any decision taken, or any failure to take a decision, by(a) an organ of state, when(i) exercising a power in terms of the Constitution or a provincial constitution; or (ii) exercising a public power or performing a public function in terms of any legislation; or (b) a natural or juristic person, other than an organ of state, when exercising a public power or performing a public function in terms of an empowering provision, which adversely affects the rights of any person and which has a direct, external legal effect

So what then is a decision? For this we must look elsewhere in the definitions section:
"decision" means any decision of an administrative nature made, proposed to be made, or required to be made, as the case may be, under an empowering provision, including a decision relating to: [a nonexhaustive list of examples which Parliament stole straight from Australian case law].

Obviously this is circular. What is an empowering provision? We must again refer to another part of the section:
"empowering provision" means a law, a rule of common law, customary law, or an agreement, instrument or other document in terms of which an administrative action was purportedly taken

Returning to the definition of administrative action, we see that the decision must be taken by any organ of State. This term is defined in turn in s 239 of the Constitution:
(a) any department of state or administration in the national, provincial or local sphere of government; or (b) any other functionary or institution(i) exercising a power or performing a function in terms of the Constitution or a provincial constitution; or (ii) exercising a public power or performing a public function in terms of any legislation,

31

but does not include a court or a judicial officer;

This definition would capture the SABC, UCT, School Board, ACSA, and the thousand or so other officially-listed organs of state. The decision must adversely affect the rights of a person and have a direct external legal effect. This latter phrase means, for example, that the recommendation of a junior administrator to his senior will not itself be reviewable. It is only once the decision starts affecting members of the public that it is administrative action. So the definition is not that long when you stick to what is essential. And, fortunately, our courts have provided a nice, succinct, and now well-used definition with five components:
Administrative action means any decision of an administrative nature made...under an empowering provision [and] taken ... by an organ of state, when exercising a power in terms of the Constitution or a provincial constitution, or exercising a public power or performing a public function in terms of any legislation, or [taken by] a natural or juristic person, other than an organ of state, when exercising a public power or performing a public function in terms of an empowering provision, which adversely affects the rights of any person and which has a direct, external legal effect.

On paper, the definition is a lot longer, mainly because there are nine categories which are specifically excluded from this definition. These appear as follows:
(aa) (bb) (cc) the executive powers or functions of the National Executive ; the executive powers or functions of the Provincial Executive ; the executive powers or functions of a municipal council;

Recall the distinction we drew above between the executive and the administration. The former are not reviewable under PAJA, as the above three exemptions show. They remain reviewable by other means, though (as we will see later).
(dd) the legislative functions of Parliament, a provincial legislature or a municipal council;

Parliament does have other, non-legislative functions. Decisions taken in terms of these functions are not excluded from the definition. So, for example, an MP who was unlawfully suspended could challenge her removal under PAJA.
(ee) the judicial functions of a judicial officer of a court and the judicial functions of a traditional leader under customary law or any other law;

Again, the judiciary does have non-judicial functions, for example granting a search warrant or warrant of arrest.
(ff) a decision to institute or continue a prosecution;

The rationale for this subsection is that, once a prosecution is instituted, the case is going to be argued by the parties and decided before a judge. It would therefore make no sense to allow the institution itself to be judicially reviewed. This subsection does not preclude the review under PAJA of a decision not to institute or not to continue a prosecution. 32

(gg)

a decision relating to any aspect regarding the nomination, selection or appointment of a judicial officer or any other person, by the Judicial Service Commission in terms of any law;

This is a very important exercise of public power, but it is reasonable that it shouldnt be reviewablebecause the judiciary itself would have to do the reviewing!). But when the JSC carries out its disciplinary functions then its decisions are reviewable under PAJA. That was exactly what occurred following the decision not to impeach Judge Hlophe.
(hh) any decision taken, or failure to take a decision, in terms of any provision of the Promotion of Access to Information Act, 2000; or

The reason for this is simply that PAIA has its own review structure.
(ii) any decision taken, or failure to take a decision, in terms of section 4(1);

Section 4(1) (see below) relates to the administrators decision about what to do when he needs to provide for public participation in the taking of an administrative action affecting the public at large. On to the remaining sections of PAJA: Section 2 (Application of Act) provides for a department of functionary of State to be exempted from PAJA. No department or functionary has ever even applied to be exempted under this section. This means either that no department or functionary has anything to hide, or that they dont think administrative law is effective enough to be a concern. Section 3 (Procedurally fair administrative action affecting any person) sets out the minimum content for procedural fairness when the administrative action affects individuals. Section 4 (Administrative action affecting public) does the same in respect of actions affecting the public at large. The administrator must choose either a notice-and-comment procedure, a public enquiry, or neither; but if he chooses neither he must then justify his decision in terms of s 4(4)(b). Section 5 (Reasons for administrative action) provides for the giving of reasons. Time periods are prescribed within which the individual must apply for reasons (within 90 days from the date on which that person became aware of the action or might reasonably have been expected to have become aware of the action) and within which the functionary must provide adequate reasons (within 90 days after receiving the request). The time periods reflect the tension between the right to administrative justice and the need for administrative efficiency: if people could come along way after an administrative decision has been implemented and then try and nullify it, efficiency would be seriously impaired. What would constitute adequate reasons will of course depend on the circumstances. A failure timeously to provide reasons raises the presumption in any review proceedings that the functionary had no good reasons for his decision. This places a serious onus on the administrator in the event of a future review, and thus provides a strong incentive for him to give good reasons in the first place. Note that the threshold or standing requirement to be entitled to be request reasons is that the persons rights must have been materially and adversely affected. This raises the threshold from what is in s 33 of the Constitution.

33

Some have argued that this is an unconstitutional narrowing of the right to administrative justice, but no court has had to deal with this. Section 6 (Judicial review of administrative action) (partially) codifies the common law grounds of review of administrative action in an attempt to clarify what constitutes lawfulness, procedural fairness and reasonableness. There are three points to make at this stage (of course we will consider this section in much more detail later). Firstly, the grounds of review are bizarrely muddled up amongst the various subparagraphs, with no discernable rationale. Secondly, consider s 6(2)(f)(ii), which reads as follows:
(2) A court or tribunal has the power to judicially review an administrative action if (f) the action itself (i) contravenes a law or is not authorised by the empowering provision; or (ii) is not rationally connected to: (aa) the purpose for which it was taken; (bb) the purpose of the empowering provision; (cc) the information before the administrator; or (dd) the reasons given for it by the administrator

The SALRC draft legislation had proposed that in addition to these groundsall based on rationality of processthere be added a requirement that the outcome of the administrative action must be proportional in the circumstances. Thus there wouldve been two legs to PAJAs reasonableness enquiry: the administrative action would need to be internally rational and have a proportionate outcome. But Parliament abandoned the proportionality requirement and replaced it with what appears in s 6(2)(h), which requires that the action must not be so unreasonable that no reasonable person could have so exercised the power or performed the function. This test is known as the Wednesbury test, after the English case in which it was laid down, namely Associated Provincial Picture Houses v Wednesbury Corporation [1947] 1 KB 223. The case involved the challenge on the grounds of unreasonableness to a law which prohibited the screening of movies on Sundays. The Wednesbury test was adopted all over the Commonwealth, except in South Africa. The test is clearly circular: saying that an action is unreasonable is the same as saying that no reasonable person could do it. Thus England abandoned it in 1964, and shortly thereafter the rest of the Commonwealth followed suit. But our Parliament, in its wisdom, decided to resurrect the Wednesbury test and enshrine it in PAJA. Fortunately the Constitutional Court has basically circumvented the clause by interpreting it so as to mimic the proportionality of outcome requirement which the SALRC committee had recommended. Thirdly, we consider s 6(2)(i), which allows a court to review an administrative action if the action is otherwise unconstitutional or unlawful. This is an omnibus clause, allowing courts to develop new grounds of review not mentioned specifically in PAJA. Courts may develop the new grounds on the basis of the Constitution or the common law. Two common-law grounds which are not in PAJA but have already been used in terms of s 6(2)(i) are review for vagueness (where executive legislation is so poorly expressed that its subjects do not know what is required for compliance) and acting under dictation (where the administrator didnt apply his mind and simply gave effect to the orders of his superior). Section 7 deals with the procedure for instituting judicial review. Such proceedings must be instituted without unreasonable delay and not later than 180 days after the date upon which all internal remedies are exhausted or, where no such remedies exist, on which the person became aware of the administrative action or ought reasonable to have become aware of it. The 180 day time period may be problematic given that it could take

34

180 days just to request and receive reasons (each of these acts having a 90 day time period). But see s 9, below. Section 8 deals with remedies for unjust administrative action. A court can order any just and equitable remedy. Section 9 provides for the variation by a court or tribunal of the prescribed time periods in the Act where the interests of justice so require. Section 10 requires the Minister to make certain regulations, and says that he may make certain others. So far the Minister has only made the regulations which are required. 4 The Constitution and the common law: Pharmaceutical Manufacturers

The case of Pharmaceutical Manufacturers Association of South Africa and Another: In re Ex Parte President of the Republic of South Africa and Others [2000] ZACC 1 is widely regarded as the most important case in South Africas administrative law. (a) Facts The Medicines and Related Substances Control Act 101 of 1965 regulated the manufacture and sale of all medicines and related substances. Then Parliament drafted new legislation to replace this, namely the South African Medicines and Medical Devices Regulatory Authority Act 132 of 1998. The Act was to come into force when assented to by the President. The President assented in May 1999. Unfortunately the Act was totally useless because the Department of Health had not published the regulations which gave essential content to the Act. Ordinarily this gross error might have been corrected by Parliament, but Parliament was not in session, as its members were busy preparing for the upcoming national election. Therefore various interested parties (including the Pharmaceutical Manufacturers Association) applied to court to have the Presidents assenting set aside. (b) Proceedings in the High Court In the Transvaal High Court Fabricius AJ dismissed the application on the grounds that the President had acted within his powers and in good faith. This was an appropriate decision for a single High Court judge to make on the law then existing; but the feeling was that a higher court would have the legitimacy to be more pioneering. And indeed, on appeal to a Full Bench of the High Court, the Presidents action was struck down on the grounds that it had not been exercised within the limits of the authority that was conferred to him. (c) Proceedings in the Constitutional Court But the interested parties wanted further answers to the matters of law involved, so they went before the Constitutional Court. This was important, because it gave the Court an opportunity to resolve the relationship between the Constitution and the common law in the area of administrative law (and its understanding of this relationship would later affect the status of PAJA).

35

(d) The major, jurisdictional issue The Court had to decide firstly whether the matter was a constitutional one. For if it was not a constitutional matter, the Supreme Court of Appeal (or, more accurately, the Appellate Division, as it still was at that stage) wouldve been the highest court in the country in respect of that matter and therefore it, rather than the Constitutional Court, would be the court with jurisdiction to hear the appeal. This jurisdictional issue was part of a broader jurisdictional battle between the Constitutional Court and the SCA. The SCA had no jurisdiction between 1994 and 1997 to hear constitutional issues. This was for political reasons: the sentiment amongst the drafters of the Interim Constitution was that the Appellate Division needed to be punished for its pro-executive judgments during the 1980s; and that the new Constitutional Court needed an opportunity to shine and develop a transformative jurisprudence without competition from another court of appeal. This created a spat between the two courts, and resentment still lingers. Anyway, there is a long history to the SCAs attitude to its jurisdiction (or lack thereof) in administrative law matters, which is dealt with in the Constitutional Courts judgment in Pharmaceutical Manufacturers. Chaskalson P summarises the SCAs position in this regard at para 21 ff. The gist is as follows. In the three early cases (Rudolph, Fedsure and Premier, Mpumalanga) in which an administrative law matter had gone to the SCA, the SCA held that it had no jurisdiction to hear constitutional matters under the Interim Constitution, and this included administrative law matters (which engaged s 24 of the Interim Constitution). Or, as the SCA said in Fedure, the question of whether the SCA did have jurisdiction was itself a matter reserved by the Interim Constitution for the CC. On this basis the SCA refused to hear each of the matters, and referred them to the CC. In the CCs Fedsure judgment, the CC made very clear that they agreed that indeed the SCA had no jurisdiction to hear administrative law matters under the Interim Constitution. But Fedsure was actually heard once the 1996 Constitution had been enacted. The 1996 Constitution had renamed the Appellate Division the SCA and restored constitutional jurisdiction to it (although of course the Constitutional Court was higher than it in respect of constitutional matters). The CC encouraged the SCA to use the constitutional jurisdiction given to it under the final Constitution to hear administrative law disputes which had arisen during the currency of the Interim Constitution. Although this subverted what had been written in the Interim Constitution, it was justified by the interests of justice. In essence, the Constitutional Court was trying to avoid a recurrence of what had happened in Rudolph, Fedsure and Premier, Mpumalangait didnt want the Interim Constitutions denial of SCA jurisdiction to continue to hamper litigants even now that the SCAs jurisdiction had been restored by the final Constitution. The first administrative law matter coming before the SCA after this was Commissioner of Customs and Excise v Container Logistics (Pty) Ltd 1999 (3) SA 771 (SCA). Hefer JA agreed in essence that the SCA was now empowered to hear administrative law disputes arising under the Interim Constitution, of which the matter in casu was one. One wouldve thought that the SCA would therefore apply s 24 of the Interim Constitution to the facts before it. But Hefer JA said that it was not necessary to decide whether the administrative action in issue had infringed s 24. His reasoning was, basically, that the common-law grounds of review of administrative action still existed, and that the administrative action in issue was invalid on the basis of these common-law grounds. Whether it was also invalid for infringing s 24 was therefore irrelevant. In this regard, Hefer JA wrote:

36

Judicial review under the Constitution and under the common law are different concepts. In the field of administrative law constitutional review is concerned with the constitutional legality of administrative action, the question in each case being whether it is or is not consistent with the Constitution and the only criterion being the Constitution itself. Judicial review under the common law is essentially also concerned with the legality of administrative action, but the question in each case is whether the action under consideration is in accordance with the behests of the empowering statute and the requirements of natural justice.

On this basis the SCA held that it could review administrative action under the common law, without engaging in any way with the Constitution. (e) The relationship between the Constitution and the common law In Pharmaceutical Manufacturers one of the applicants sought to exploit the SCAs finding in Container Logistics to argue that this was a non-constitutional matter and so the Constitutional Court had no jurisdiction to entertain it. The applicant argued, in other words, that, just like in Container Logistics, the High Court had found the administrative action to be invalid on the basis of the common law; that the common-law grounds of review are separate from the constitutional grounds; and therefore that the Constitutional Court had no jurisdiction to hear the appeal. But Chaskalson P, responding to this argument, writes as follows at para 33:
I take a different view. The control of public power by the courts through judicial review is and always has been a constitutional matter. Prior to the adoption of the interim Constitution this control was exercised by the courts through the application of common law constitutional principles. Since the adoption of the interim Constitution such control has been regulated by the Constitution which contains express provisions dealing with these matters. The common law principles that previously provided the grounds for judicial review of public power have been subsumed under the Constitution, and in so far as they might continue to be relevant to judicial review, they gain their force from the Constitution. In the judicial review of public power, the two are intertwined and do not constitute separate concepts.

And then, at para 44, he savagely obliterates the SCAs view and urinates on its grave:
I cannot accept this contention which treats the common law as a body of law separate and distinct from the Constitution. There are not two systems of law, each dealing with the same subject matter, each having similar requirements, each operating in its own field with its own highest court. There is only one system of law. It is shaped by the Constitution which is the supreme law, and all law, including the common law, derives its force from the Constitution and is subject to constitutional control.

(f) Wank about the rest of the judgment In para 45 Chaskalson P deals with the public/private distinction, making the important point that administrative law is the only legal means of challenging state action and is therefore a vital component of a just legal system. In paras 46-57 Chaskalson P restates the basic principles of post-1994 administrative law. In paras 58-68 he applies these to the facts. His finding is basically that the Presidents actions in assenting to the legislation were bona fide but mistaken. And the mere fact that the action was bona fide is not by itself enough to stop the Court from reviewing the action and potentially invalidating it. Then he refers to foreign law, and then to 37

the Constitutional Courts own previous judgments, most notably SARFU, which had used a robust approach to the review of Presidential powers. At para 79 Chaskalson P says that the Presidents actions did not constitute administrative action. So, does that mean the Court cannot review them? No! Every exercise of public power, whether or not it amounts to administrative action, is subject to the foundational principle of the rule of law. It therefore follows that every exercise of power must be intra vires and undertaken in good faiththis much was recognised as long ago as Shidiack v Union Government (Minister of the Interior) 1912 AD 642. But the Constitution requires more. The Court cites the Mureiniks culture of justification as well as the following dictum from Ackermann J in Makwanyane:
We have moved from a past characterised by much which was arbitrary and unequal in the operation of the law to a present and a future in a constitutional State where State action must be such that it is capable of being analysed and justified rationally. The idea of the constitutional state presupposes a system whose operation can be rationally tested against or in terms of the law.

On this basis the Court holds that the rule of law also requires that every exercise of power must be rationally related to the purpose for which the power was given. The Court finally concludes that the fact that the Presidents assenting to the Act was based on a fundamental error, resulting in the whole Acts being unworkable, means that the Presidents actions had no objectively rational basiswhether or not he acted in good faith being irrelevant to this issue. The Court thus confirmed the Full Benchs order, with the result that the new legislation had not actually been assented to and therefore the old legislation continued to regulate the industry. (g) Issues What do we take out of Pharmaceutical Manufacturers? o The Constitution and the common law

Prior to this judgment, there was a degree of parallelism in our administrative law. This was because the SCA was applying residual common-law grounds of review distinctly from the constitutional grounds. Pharmaceutical Manufacturers unifies these parallel streams and says the Constitution is fundamental to all law. This is the major reason why it is so important. o Administrative and executive actions

There is another kind of parallelism which the case deals with. This is the parallel between administrative action and executive action. The Court found that although the Presidents action was not administrative actionand thus not subject to s 33it is nevertheless reviewable on the basis of fundamental constitutional requirements for the exercise of all public power, namely legality and rationality. So the Court unifies the parallel common law and constitutional avenues for review, but then immediately recognizes that there is a parallelism of another kind: review of administrative action and review of executive action. PAJA, by having such a grotesque definition of administrative action, has inadvertently increased the importance of review of executive action. Courts are increasingly steering clear of PAJAs definition and just 38

saying, It doesnt really matter whether the conduct comes under PAJAs definition, because the principle of legality allows us to review all exercises of public power. Therefore well just deal with the matter in the latter way. It is still vital to distinguish between executive action and administrative action, though, because the grounds of review are different for each. If you can establish that the conduct is administrative action you get all the additional protections of PAJA. 5 Currie, What difference does the PAJA make?

Curries basic answer to the question he poses in the title is, Not very much at all. Although PAJA of course tries to change the common law, it largely fails to do so because the courts have reverted to the principle of legality (as just explained). Currie first of all explains what PAJA is trying to do to the common law. He concludes that it is largely a restatement of the common law of judicial review of administrative action, but does also aim to reform some areas (most obviously the sections on procedural fairness). What PAJA does: 1. The Act defines administrative action (s 1). It does so exhaustively, i.e. the definition is not open-ended, and in considerable detail. The definition is narrower than the constitutional conception of administrative action and is therefore, to that extent, a limitation of the constitutional rights (although possibly a justifiable one). 2. PAJA gives effect to the right to procedurally fair administrative action by setting out express, minimum, mandatory procedures for ensuring that individuals likely to be affected by administrative action have received notice and opportunity to be heard, a clear statement of the administrative action, and a notice of rights to review or appeal and to request reasons for a decision (s 3(2)(b)). The Act also suggests a number of additional procedures beyond the minimum that may be necessary in exceptional cases to ensure procedural fairness (s 3(3)). 3. PAJA gives effect to the right to procedurally fair administrative action by setting out procedures for ensuring that in the case of administrative action likely to affect the public at large an appropriate form of notice and consultation takes place (s 4). 4. PAJA places a duty on administrators to provide reasons for administrative action on request and sets out procedures for making such requests and for responding to them (s 5). To the extent that the duty is triggered by a request and is not automatic and to the extent that PAJA envisages administrators not providing reasons in some cases, the Act is a limitation of the constitutional right. 5. The Act gives indirect effect to the rights to lawful, reasonable and procedurally fair administrative action by providing a list of grounds of judicial review of administrative action (s 6) a set of procedures for review (s 7) and a set of remedial powers for courts or tribunals in judicial review proceedings (s 8). 6. The Act gives general regulatory powers and promotional functions to the Minister of Justice, in particular allowing the making of regulations to establish rules and standards for ensuring procedural fairness (s 9). 39

So why does it do so little in practice? Because the definition is so fucking dumb. As Nugent JA says of it in Greys Marine Hout Bay (Pty) Ltd v Minister of Public Works [2005] ZASCA 43:
The cumbersome definition of that term *administrative action+ in PAJA serves not so much to attribute meaning to the term as to limit its meaning by surrounding it within a palisade of qualifications.

As an example of the courts refusal to engage with PAJA, Currie refers to the judgment of Ngcobo J (as he then was) in Zondi v MEC for Traditional and Local Government Affairs [2004] ZACC 19. 12 The case concerned a challenge on the grounds of procedural fairness to the Pound Ordinance of Kwazulu-Natal, basically provincial legislation, which provided for the immediate seizure and impoundment of trespassing animals by a landowner without notice to the livestock owner. The applicant was an impoverished widow whose cattle had been impounded under the Ordinance. Ngcobo J decided the matter by going directly to the Constitution and seeing whether the Ordinance could be read consistently with the s 33 right to administrative justice, and the guarantee of procedural fairness contained therein. But one wouldve thought the whole point of requiring Parliament to enact legislation to enable a constitutional right is so that the legislationrather than the constitutional right itselfis always the first port of call when understanding the content of the right. The problem is that Ngcobo Js approach denies PAJA such a role. The way Currie thinks PAJA should be applied is in terms of his supplementation thesis, which is (unsurprisingly) that PAJA ought to supplement existing legislation. In other words, PAJAs procedural fairness requirements must constrain administrative actions undertaken in terms of other laweven when that other law does not itself contain those same procedural fairness requirements. So, since PAJAs procedural fairness requirements must be read as applying to the exercise of the powers granted by the Ordinance, any impoundment under the Ordinance which occurred without a hearing would be invalid, whether or not the Ordinance actually required this. But the Ordinance itself would not be invalid, because it would be supplemented by PAJA. Such an approach would be consistent with the rule that legislation must be read down so as to avoid a constitutional conflict. Indeed, one of PAJAs most important attempts at reform, says Currie, was to create a set of general procedural fairness requirements for administrative action taken in terms of any law. Ngcobo Js judgment therefore undermines the purpose of those sections. So Currie is very angry about the Zondi case, which neutered PAJA in relation to a constitutional challenge to a statute. He then considers two cases in which subordinate legislation was challengedobviously a slightly different situation, with some unique difficulties. The first case is Minister of Home Affairs v Eisenberg & Associates In re: Eisenberg & Associates v Minister of Home Affairs 2003 (5) SA 281 (CC),13 in which certain immigration regulations, made in terms of the Immigration Act, were challenged. The argument was
12

I have reworked what Corder said here in class quite substantially. I just could not reconcile what he said about Zondi with what the case says, and what Currie says about it. But note that Kerrys approach to the case (in second semester) is a bit different to Curries. While Currie said the case rejects the supplementation thesis, Kerry says that Zondi confirms it. I think the problem is that Ngcobo J lays out an approach which confirms it, but then never actually does what he says one should (and thus seems to undermine it). I think the way this is explained in my second semester notes brings this out a bit more clearly, so perhaps refer to that. 13 The discussion of Eisenberg and New Clicks in these notes is again very different to what Corder said in class. Currie is using the cases to make a very specific point, which I didnt think Corder explained at all. Anyway, I think what I have here is correct, but since it is very tricky and Corder did not deal with it, perhaps it is worth ignoring the detail.

40

essentially that there should have been a notice and comment procedure before these regulations were promulgated. The Immigration Act did not itself require any such procedure, but the applicants said s 4(1) of PAJA required this. The problem is that there are two reasons for thinking that s 4(1) of PAJA cant be applied to rule-making. The first is that rule-making arguably doesnt qualify as administrative action at all, because it is not a decision. The Constitutional Court did not decide this question, saying it is not clear whether rule-making constitutes administrative action for the purposes of PAJA. The second reason is that, even if rule-making is administrative action, action taken under s 4(1) is specifically excluded from the definition by s 1(i)(b)(ii). It would arguably be dubious, therefore, to say that a failure to take action as required by s 4(1) could be the subject of judicial review. This would mean that s 4(1) would be incapable of enforcement, and as such would merely act as a guideline for administrators. The Court again didnt decide the matter, saying that s 4(1) anyway leaves it open to the administrator to not have any notice and comment procedure at all, and that his doing so here would be justifiable. Nevertheless, the Court did seem to accept Curries supplementation thesis:
In each case it is a question of construction whether a statute making provision for administrative action requires special procedures to be followed before the action is taken. In addition, whether or not such provisions are made, the administrative action must ordinarily be carried out consistently with PAJA.

Indeed, it is hard to see why the Court wouldve discussed the issues above at all if they didnt think the applicants basic argumentthat PAJAs procedural fairness requirements supplemented the Immigration Acts requirementsmade sense. The final case Currie considers is the disgustingly long Minister of Health v New Clicks South Africa (Pty) Ltd 2006 (2) SA 311 (CC). This involved a challenge (inter alia on grounds of procedural fairness) to certain regulations which dealt with the pricing of pharmaceuticals. Currie says that the judgments of each court (the High Court, the SCA, and the CC) sketch very graphically the difficulty the courts are having with the relationship between s 33 of the Constitution and PAJA. Much ink has been spilled on this issue since 1994, and the obscene length and prolixity of the Constitutional Courts judgments reflects thisand yet the Court could offer no consensus on the essential question of whether the promulgation of the subordinate legislation in issue was administrative action (or merely some other exercise of public power). As we have said, the distinction matters because of the different grounds of review available for each. New Clicks challenged the regulations firstly on the grounds of an alleged breach of the provisions of PAJA, and secondlyin the event of PAJAs being inapplicableon the basis of the principle of legality and/or s 33 of the Constitution. The first argument meant that each court which heard the matter had to discuss the applicability of PAJA to rule-making. Both the majority of the High Court and the SCA held that PAJA was basically optional, in that the regulations were in any event reviewable under the principle of legality. Both courts therefore applied the principle of legality rather than PAJA. The Constitutional Court majority, by contrast, investigates at length the definition of administrative action and holds that rule-making is included:

41

The implementation of legislation, which includes the making of regulations in terms of an empowering provision, is therefore not excluded from the definition of administrative action. [The regulations] accordingly constitute administrative action within the meaning of PAJA.

The majority therefore distances itself from the High Court and SCAs attempt to sidestep PAJA, and properly investigates its provisions. Their finding that rule-making is administrative action decides the first of the issues left undecided in Eisenberg. But that is not the end of the matter, for there was a second issue left undecided in Eisenberg, and which is relevant here too. The issue is whether rule-making, even if it is administrative action, could be susceptible to judicial review on the grounds that it failed to comply with s 4(1) of PAJA. This important question remains in the same place as after Eisenberg: undecided. This is because the Court split 5-5-1. The first five gave various concurring judgments which interact strangely but indicate some agreement, led by Chaskalson CJ, on the fact that judicial review can be undertaken in respect of rule-making. The next five per Moseneke J said that the issue wasnt really argued and so it wasnt appropriate for the Court to decide it. Finally, Sachs J held that rulemaking is not reviewable under PAJAbut that this was not a problem because the principle of legality would step in and thus cause the rule-making to be reviewable anyway. Back to the specific matter of the supplementation thesis: Currie holds that New Clicks is irreconcilable with Zondi, because it (like Eisenberg) presupposed that PAJAs procedural fairness requirements could supplement those of the empowering legislation at issue. This amounts, says Currie, to a silent overruling of the Zondi case. Currie thus regards New Clicks as a valuable corrective to the courts disregard for PAJAalthough it still leaves much undecided, and the courts are unlikely ever to dig PAJA. 6 Hoexter, The principle of legality in administrative law

The gist of Hoexters article is that the principle of legality is being relied on a lot in our administrative law (instead of PAJA), and this has some benefits and some costs. The principle of legality was established in our administrative law by the Constitutional Court in Pharmaceutical Manufacturers. It was described as an incident of the rule of law, a foundational principle of our constitutional dispensation in terms of s 1(c). The principle applies to all exercises of public power. It can therefore be used even when the conduct in question is not administrative actionor to avoid having to decide in earnest whether it is administrative action. What elements comprise the principle of legality? Hoexter sets out to answer this question by considering each familiar ground of review in turn. She starts with lawfulness. She considers three fairly early cases in our post-1994 legal history: Fedsure, SARFU, and Hugo. In Fedsure Life Assurance Ltd v Greater Johannesburg Transitional Metropolitan Council [1998] ZACC 1714 the appellant was challenging certain regulations of the local government which taxed richer areas of Johannesburg more highly than poorer areas. More particularly, they were challenging both the rates payable to the local government and the additional levy which the richer areas had to pay to cross-subsidize the poorer areas. This was well before PAJA was enacted. Indeed, the dispute arose before the final Constitution was enacted. So Fedsure argued on the basis of s 24 of the Interim
14

Corder only discussed this case in detail in the very last lecture of the term, to do with the definition of administrative action. It adds almost nothing to that topic and is highly relevant here, so I have put everything here.

42

Constitution that this was unlawful administrative action or, in the alternative, that it was reviewable on other grounds. The action failed in the High Court, and the SCA passed the buck because this was during the interregnum when it had no constitutional jurisdiction. The Constitutional Court had to decide two issues. Firstly, is it administrative action? The Court unanimously answered that it was not: the local government was exercising its legislative authority in a budgetary matter, which was certainly not administrative action. Secondly, is it nevertheless reviewable? The Court unanimously answered that it was, on the basis of the principle of legality, held to be a fundamental principle of even the Interim Constitution (which did not actually have an equivalent of the final Constitutions s 1(c), on which the principle of legality is now based). As to the facts, the Court was unanimous in saying that the rates were lawful. But on the question of whether the additional, redistributive levy was lawful, the Court split evenly, 5-5. The joint judgment of Chaskalson P, Goldstone J, and ORegan J (concurred in by two others) held that the levy was lawful, and the other five judges per Kriegler J held that it was not. If an appeal court is deadlocked like this, the decision of the lower court stands. Since the SCA had no jurisdiction, this meant the High Courts dismissal of the action was effectively confirmed. So Fedsure is an important milestone in establishing that all exercises of public power are reviewable. The Constitutional Court has taken a similarly robust approach in respect of presidential powers. In SARFU the Presidents power to create a commission of inquiry to investigate South African rugby was challenged. And Hugo also dealt with the Presidents powers, this time to pardon prisoners. Although the Presidents conduct was in each case held not to be administrative action, it was reviewable on the ground of lawfulness. Thus, in short, the three cases just mentioned tell us that lawfulness is the first ground of review implied by the principle of legality. Turning next to rationality, Hoexter says that this was clearly established as an aspect of the principle of legality in Pharmaceutical Manufacturers, where the Court described it as a minimum threshold which all exercises of public power must satisfy. Hoexter says that rationality is not the same as s 6(2)(f)(ii) of PAJA (the rational connection test). Corder adds that it isnt the same as reasonableness either. In Pharmaceutical Manufacturers the Presidents assenting to an utterly ineffectual Act was held to be irrational. So this is the uncontroversial second ground of review implied by the principle of legality. Is procedural fairness part of the principle of legality? This is a much trickier question. In the case of Masetlha v President of the Republic of South Africa 2008 (1) SA 566 (CC) the appellant, the former Minister of Intelligence, challenged his dismissal by President Mbeki on the grounds that he shouldve been given a prior hearing. Mbekis actions were held not to be administrative action, meaning the review had to occur in terms of the principle of legality. As to the issue of whether Mbekis actions could be challenged on the basis that no hearing was given (that is, on grounds of procedural fairness), the majority said that there was no requirement when the President fires Ministers of Intelligence that he give them a hearing first; he was entitled in terms of the NIA legislation to fire them unilaterally. Ngcobo J disagreed, saying that procedural fairness was a requirement of the principle of legality, and as such the President had to consult with Mr Masetlha before firing him. Then there was the recent case of Albutt v Centre for the Study of Violence and Reconciliation [2010] ZACC 4. This concerned some stinky assholes who committed politically motivated crimes in 1995. Because they 43

seemed to meet all the requirements for amnesty in terms of the TRC Act except that their crimes were committed after the cut-off date of 27 April 1994, President Mbeki created a Parliamentary committee to advise him on whether the perpetrators should receive a presidential pardon. Some victims and surviving relatives of the crimes wanted to be consulted before the decision was made, but both the Parliamentary committee and President Mbeki said No thanks. It was this failure to take their representations which the victims wanted to challenge. This was not easily construed as an administrative action, so again the principle of legality was in play. This seemed to require the Court to decide whether procedural fairness is a requirement of the principle of legality, but it didnt really do so in general terms. Ngcobo CJ held that because the potential pardon was a follow-up to the TRC, the decision must be taken in terms of the TRCs process and rationale, i.e. the affected parties must come together and have a chance to express themselves so that reconciliation can be achieved. On this basis Ngcobo CJ held that the applicants did have a right to be heard. But, crucially, Ngcobo CJ said expressly that the unique circumstances of the matter were decisive and that the Court was not deciding that procedural fairness is a general requirement under the principle of legality. Ultimately, then, the courts have not held that procedural fairness is a requirement of the principle of legality, but arguably theyve partly recognised that it is or could be in certain circumstances.15 Hoexter then briefly deals with the giving of reasons. This, she says, is definitely not a requirement of the principle of legality yet, because no court has confronted the matter. But it should be a requirement because reasons must be provided before rationality can be judged. To summarise, the different grounds of review are (at least at this stage of our legal development) as follows: PAJA Lawfulness Rationality Procedural fairness Right to reasons Principle of legality Fedsure, SARFU, Hugo Pharmaceuticals Masetlha dissent, Albutt

So why have the courts used the principle of legality at all? The answer is that it removes the threshold requirement (of both s 33 of the Constitution and PAJA) that the conduct in question must constitute administrative action, thus allowing a broader range of conduct to be reviewed. There are certain benefits to using the principle of legality: 1. judges get additional flexibility in reviewing conduct 2. it avoids the difficulties of whether the conduct amounts to administrative action and goes straight to the more essential question, namely, What does administrative justice actually require? The costs associated with this approach, according to Hoexter, are that: 1. it subverts or undermines PAJA because if the principle of legality is available for all exercises of public power then there is hardly any point in trying to apply PAJA (although there is still some point in that PAJA has broader grounds of review)

15

Corder suggested in second semester that, as far as bias is concerned (usually regarded as a part of procedural fairness), you could argue that lawfulness requires acting in good faith which requires acting without bias. So you could smuggle the rule against bias into the principle of legality under lawfulness.

44

2. the additional avenue for review which the principle of legality creates means that certain procedural issues are raised (especially if the new procedural rules come in which would create separate procedures for each kind of review) 3. litigation becomes more unpredictable because judges effectively have discretion to decide which set of rules (PAJA or principle of legality) they want to apply So, strictly speaking, the principle of legality must be used to subject executive actionas distinct from administrative action, which is covered by PAJAto review. But, because PAJAs definition of administrative action is so difficult to work with, the temptation is to use the principle of legality as a means to avoid dealing with this distinction altogether. The Constitutional Court, by the way, has never actually dealt with the issue of whether PAJAs definition of administrative action is constitutional. They avoided the issue in Albutt, and no other litigant has been willing to waste his dollaz on it. 7 The upshot of the section (a) The two parallelisms revisited Okay, so the basic relationship between the common law, PAJA and the Constitution is as follows. Between 1994 and 2000 there was the unseemly jurisdictional spat between the Constitutional Court and the SCA which had the effect of inappropriately extending the life of the common law (used by the SCA) alongside s 24 of the Interim Constitution (used by the CC). This parallelism was snuffed out by the CC in Pharmaceutical Manufacturers. The result was that there was then just one system of administrative law, deriving its force from the Constitution. When PAJA was enacted, it should, in theory, as the enabling legislation of the constitutional right, have then formed the basis for this new unitary system. But the disgusting definition of administrative action in PAJA led to a new bifurcation in our administrative law: on the one hand there is review in terms of PAJA, with its onerous administrative action threshold requirement; and on the other there is review in terms of the principle of legality (sourced in s 1(c) of the Constitution), without such a threshold and applying to all exercises of public power. What has happened to the common law? Well, since PAJA for the most part codifies the pre-existing common law, the common law will be used in the interpretation of PAJA. The Constitutional Court specifically said in Pharmaceutical Manufacturers that the common law supplements the provisions of the written Constitution although, of course, it derives its force from it and must be developed accordingly. And, according to Hoexter, the common law will remain a free-standing channel of review in limited circumstances. This is where an indisputably private body exercises what is not actually a public power but is like a public power and as such is subject to certain administrative law principles. For example, if a chess club or stokvel kicks out one of its members, it must invite him to a hearing. This is a requirement of the common law principles on administrative law and natural justice, which remains relevant because neither PAJA nor the principle of legality have any application to private bodies exercising private powers.

45

Hoexter does also say that special statutory review is possible nowadays: legislation could specifically provide for review of actions taken in terms of that legislation. Now that administrative law has developed so much in other areas, though, this is fairly rare. (b) The role of the ultra vires doctrine The justification for judicial review pre-1994 was the common law ultra vires doctrine, based in parliamentary sovereignty. Now the justification is the Constitution, which expressly empowers courts to control administrative action. But the ultra vires doctrine does still play a role, by lending a logical explanation to that empowerment; it explains why the Constitution empowers the courts in this way. Administrators nowadays must not just act intra vires parliamentary legislation: they must also act intra vires the Constitution. The post-1994 ultra vires doctrine is therefore as extensive as the Constitution will allow. It must be understood not narrowly in terms of parliamentary sovereignty but widely as founded on respect for the rule of law.

(c) What a litigant must do We have seen repeatedly the critical gate-keeping role or threshold constituted by the definition of administrative action. This has caused a bifurcation in our administrative law. Anyway, a litigants first port of call when he is pissed off with the exercise of a public power is judicial review of administrative action under PAJA (the interpretation of which will be informed in part by the common law). But if he cant satisfy PAJAs definition of administrative action (or wants to avoid this nauseating issue!) then he must refer to the principle of legality, which consists of lawfulness, rationality and (maybe) procedural fairness. This comes at a cost in that there is lesser scope and depth of this form of review as compared to review under PAJA. Finally, if he is pissed off with the exercise of a public-like power by an obviously private body then he can rely per se on common-law principles of natural justice.

46

THE SOURCES OF ADMINISTRATIVE AUTHORITY

In this topic we take a brief look at the sources of administrative authority. This is central to the empowerment aspect of administrative law. 1 The Constitution

Unsurprisingly, the Constitution is one of the key sources of administrative authority. The Constitution deals with the structure and powers of national government (Chapter 5), provincial government (Chapter 6), local government (Chapter 7), the judiciary (Chapter 8), the Chapter 9 institutions (Chapter 9!!!!!!!!!) and the public administration (Chapter 10). We will briefly look at the relevance to administrative law of the oft-neglected Chapter 10, specifically s 195. Section 195(1) sets out the values and principles by which the public administration must be governed. There are several of these, including a high standard of professional ethics; the efficient and economic use of resources; being development-oriented; providing services impartially, fairly, and equitably; and responding to peoples needs. The key question raised by this section is: Are these standards justiciable? In other words, can administrative action be challenged on review on the grounds that it falls short of these standards? Initially this issue was seriously neglected, but s 195 is increasingly being relied upon by litigants. 2 Legislation

The next key source of administrative authority is legislation. We must distinguish between two types: Primary a.k.a. original legislation a.k.a. empowering legislation, i.e. statutory instruments enacted by Parliament, a provincial legislature or local government16 Secondary a.k.a. delegated a.k.a. subordinate legislation, i.e. statutory instruments made by Ministers, Director-Generals, or anyone else who derives their law-making authority from primary legislation There are limits on the powers of a legislature to delegate, as made clear in Executive Council of the Western Cape Legislature v President of the Republic of South Africa 1995 (4) SA 877 (CC). Prior to the first democratic local government elections, there were certain political disputes relating to the demarcation of the constituencies in the Western Cape. Parliament had recently amended the Transition Act (which governed the procedures for setting up a non-racial local government) so as to allow the President to amend the Act itself by promulgation. The President used this power to resolve the disputes and authorise a certain demarcation. The National Party, which at that stage was in power in the Western Cape, was pissed off, so it challenged the validity of the conferral upon the President of the power to amend the Act itself. This was obviously a politically charged issue and was a big moment for the legitimacy of the new judiciary. Mandela did a televised wank about the fact that the government would respect the Courts decision. The majority per Chaskalson P held that it was unconstitutional for the President to be able to amend legislation. Whatever powers a legislature does (validly) delegate, the subordinate legislation of course cannot overstep these powers; the subordinate legislation must be strictly consistent with the empowering legislation.

16

It was confirmed in Fedsure that local government is allowed to enact legislation.

47

Prerogative powers

The Presidents prerogative powers derive from those of the British monarch, but are now constitutionally enshrined. The difficulty with these powers pre-1994 was that they were not reviewable. The question arose in President of the Republic of South Africa v Hugo 1997 (4) SA 1 (CC) whether they were reviewable under our new constitutional dispensation. As we know, the Courts answer was Yes: although they are not administrative action they are exercises of public power and as such are reviewable under the principle of legality. 4 Common law

The common law isnt really an independent source of authority now that, according to Pharmaceutical Manufacturers, all law derives its force from the Constitution. 5 African Customary law

Traditional leaders are dealt with in Chapter 12 of the Constitution. Section 211(1) states that:
The institution, status and role of traditional leadership, according to customary law, are recognised, subject to the Constitution.

Thus African customary law is clearly a source of administrative authority, but it is subject to the Constitution. This section hasnt really been applied yet, and it isnt clear whether traditional leaders are subject to ordinary administrative law standards. Bennett speculates in the prescribed reading (written in 1993) as to how administrative justice might be applied to traditional leaders, specifically in the context of the power of a tribal authority to evict the holder of an allotment from his or her land (trekpas). He begins by noting that, although the powers of a chief originally derived from his peoples consent (he is a chief *only+ by the grace of his people), the colonial and then apartheid governments corrupted this system and essentially made chiefs part of the local government: they became salaried functionaries who actually played a prominent role in enforcing apartheid through indirect rule. Bennett describes them as a backward and politically conservative faction, and more seriously they were widely regarded as stooges of the government. Obviously trekpas has very serious consequences for its subject. Technically a chief may only remove someone from their property if it is in the public interest and various other conditions are met. But the only way to ensure that a chief actually complies with these conditions is that he is obliged to consult his all-male council and even then, he is not bound by their advice. The major shortcoming of this procedure is that it lacks the inclusion of a hearing. But it was held in Mokhatle v Union Government 1926 AD 71 that a hearing was not necessary, making it very clear that courts were at that stage not willing to apply the ordinary rules of administrative law in these circumstances.

48

Bennett goes on to consider the possible remedies now available. Firstly, in terms of procedural fairness, in the case of Administrator, Transvaal v Traub 1989 (4) SA 731 (A) the Appellate Division held that audi alteram partem applies wherever legitimate expectations are affected. This means, according to Bennett, that Mokhatle no longer applies and trekpas orders are subject to audi alteram partem. The case was silent on whether the same was true of the nemo iudex rule. Secondly, there is the possibility of reviewing the order on the grounds of symptomatic unreasonableness, which allows courts to infer from the unreasonableness of the decision that the chief had acted dishonestly, mala fides or from an ulterior motive. Lastly, Bennett discusses the notion of ultra vires. But this is difficult to apply as there are no clear boundaries of a chiefs authority or powers. This points to some of the difficulties in taking common law remedies and just transplanting them into customary law. On the other hand, norms of good governance must apply to all areas of law. Bennett argues that Chiefs are still officials acting in the public interest; their office is part of the state administration; they are paid their salaries from public funds; and they are under the control of the Minister. Therefore they should not be given a privileged position in the national system of justice by not subjecting them to administrative law remedies. Given s 211(1) it seems that such actions would have to be subject to review in line with what was held in Pharmaceutical Manufacturers. Sections 30 and 31, granting rights to language, culture, etc. may also be relevant here. Both are subject to other provisions in the Bill of Rights. 6 Custom

Could a public bodys acting in a certain manner since time immemorial, i.e. in accordance with a custom, create a source of public authority? So, for example, if local government has been regulating water rights in an area since 1850 does this custom give it the power to continue to do so? The answer is surely No, in light of Pharmaceutical Manufacturersthe power must come ultimately from the Constitution. Any customary exercise of power would therefore be open to a challenge on the grounds of unlawfulness.

49

THE PUBLIC/PRIVATE DIVIDE

1 Introduction

We have in the previous section discussed at length the gatekeeping requirement that the conduct in question amount to administrative action. This requirement is defined in s 1(1) of PAJA to be applicable to exercise of public power or the performance of a public function. This reflects the same centrality of public power in s 239 of the Constitution. And, even when PAJAs gatekeeping requirement is not met, the conduct will be reviewable (albeit to a lesser extent) under the principle of legality provided it is an exercise of public power. So clearly the distinction between the public and the private is fundamental to administrative law. The meaning of this distinction is the subject of this topic. The dividing line between public power and private power is not a bright one, especially given the developments of the past thirty years, in a rapidly globalizing world. In practice, the public/private distinction is relevant in relation to two main questions. Firstly, whether an action taken by a public entity in a seemingly private law setting is subject to administrative law; and, secondly, whether and when private entities are capable of exercising public powers or performing public functions and thus subject to administrative law. Corder focused on the former question in class, but Hoexter says lots of exciting things about the latter. 2 Aman, Administrative law in a new century

Alfred Aman is an American administrative lawyer, who sought to understand the effects of globalization, privatization, etc. on administrative law. He begins by looking at these trends in broader historical perspective. In the early 20th century, government took a laissez-faire approach to the private sphere, with the red light model of administrative law putting a stop to any government intervention in the private sphere. With the Great Depression and the world wars, government took an increasingly interventionist role, creating public works programs, establishing welfare and social services, creating government agencies, and aggressively regulating the economy. The green light model of administrative law sought to empower government to serve the public interest, rather than merely to restrict it. Judicial review became more dynamic, trying to encourage participation and accountability rather than just declaring unlawful interventions invalid. Then in the 1980s, with the Reagan/Thatcher revolution there was an aggressive move towards deregulation and privatization in the name of greater efficiency. At the same time, with the growth of global markets and the rise of supranational actors, the authority of the State is declining. Firstly, companies are free to take their operations elsewhere if any particular State has an unfavourable approach to business. This forces governments to involve business in negotiated rule-making (see below) and to have business-friendly regulatory schemes which are sympathetic to businesss demand for cost-minimization. Secondly, where States were previously an absolute authority over their citizens, States are increasingly having to take orders from a higher authority: the global market. Like Cockrell (see below), he clearly sees that in the face of these trends the traditional public/private boundary is collapsing. This change in the structure of administrative law is tiedas it often isto changes in theories of State. Aman quotes a leading administrative lawyer, Stewart, who writes that two fundamental 50

criticisms have been leveled at the traditional model of the public: Firstly, the traditional models protection of the private sphere from State interference is inappropriate given the seemingly inexorable expansion of government power over private welfare. Secondly, government agencies have failed to protect the public interest in their spheres of operation, and traditional administrative law hasnt solved this problem. But, more than this, Stewart describes a general change in the function of administrative law: Where it once simply tried to stop government from unlawfully interfering in the private domain:
Increasingly, the function of administrative law is not the protection of private autonomy but the provision of a surrogate political process to ensure the fair representation of a wide range of affected interests in the process of administrative decisions.

Aman discusses at length certain means through which the US government has privatized formerly public functions, specifically prisons and the welfare system. Both examples resonate with the South African context, where we have some private prisons and a (nominally public) welfare agency. How can the constitutional rights of prisoners be secured within the private context? How can fairness in the distribution of welfare grants be ensured when the body in charge is driven exclusively by profit? Generally, how do we translate essential public law values like participation and accountability into private areas? And what is lost in translation? Aman is obviously concerned about these difficulties in the American context. The starting point is that it would be morally unacceptable to allow the form of a company (i.e. the fact of whether it is public or private) to determine whether or not it is bound by the substance of fundamental human rights. It should be noted that South Africa has one major advantage over America in ensuring that public law values will be properly translated into the private domain following privatization: section 8(2) of the Constitution provides that our Bill of Rights can be applied horizontally. Our Constitutional Court has also recognised that the Constitutions definition of an organ of state in s 239 includes any entity that performs a public function and thus ensures that government cannot be released from its human rights and rule of law obligations simply because it employs the strategy of delegating its functions to another entity: see AAA Investments (Proprietary) Limited v Micro Finance Regulatory Council [2006] ZACC 9. A wholly private body could therefore, if it is performing a public function, constitute an organ of State and therefore be subject to the right to administrative justice. Aman then discusses corporatizing government, which is essentially where the government uses a corporate model to achieve public goals. This was done with the US Postal Service and the Federal Aviation Administration. The American courts have held that administrative law standards must be applied to such organizations if the corporation was created by government to perform a public function and if the government has some degree of control over them. Aman then discusses negotiated rulemaking, which is essentially the making of subordinate legislation by negotiations between government and those who will be affected by the legislation. It amounts to a more vibrant form of notice-and-comment. The purpose is to avoid disputes over the content of the regulations and to increase their legitimacy; and to respect the interests of the regulated parties and to enjoin their expertise. The need for such processes is especially pressing nowadays because manufacturers are free to relocate their operations to the most favorable jurisdiction. The key question, though, is Who sits at the negotiating table? Obviously it must be all interest groupsnot only the industry being regulated but also its consumers. Finally, Aman proposes three models of administrative law which could be adopted to deal with governments changing role and its tendency to privatize public functions and use market regulation. Firstly, the back to the future model simply tries to return to 19th century State minimalism and laissez-faire, with the free market 51

seen as an end in itself and government seen as having no legitimate role to play in the private sphere. This model demands a clear distinction between public and private. Perhaps one sees shades of this approach in the big society mantra of Tory Britain. Secondly, the corporatizing government model is similar, in that it thinks free markets are desirable. But it sees these not as ends in themselves but a means to growth and prosperity. Thirdly, the market cooptation model sees government as legitimately adopting a highly interventionist rolebut through market-based regulation rather than command-and-control. It is in this model that the public/private divide is must blurred. The US perspective is obviously quite different to our own, but it does tell us important things about global trends relevant to South Africa. What model have we adopted? With BEE, the NCA, the CPA, justiciable socioeconomic rights, and a large welfare State, we seem to have adopted the third option. Our economy, while broadly in line with neoliberal principles, is still subject to a lot of regulation. 3 Cockrell, Can you paradigm? Another perspective on the public law/private law divide

Cockrell begins by describing the shifting and blurring of the traditional public/private divide. The traditional distinction was based on certain key tenets: that the State should stay out of private life; that public law must deal with an unequal relationship between the State and private parties, while in private law the parties are equal; and that public law must deal with politics whereas private law is neutral, uncontested and free of politics. But this traditional view is under attack, on two fronts. Firstly, there is the recognition by private lawyers that private law cannot be hermetically sealed off from public law, that private parties can be unequally situated, and that the structure of private transactions is an expression of a particular political moralityand accordingly the State has a legitimate role to play in private law. Secondly, judicial review has reached more and more into the private domain in an attempt to hold private (or nominally private) power accountable. The result is a surrendering of the seemingly-impregnable Maginot Line between public and private law. Second, Cockrell discusses the separate sets of rules of liability for public and private parties in contract and delict. He notes a few emerging trends, but it is clear that not that much has been done by private lawyers to accommodate the realizations described above. Third, he discusses procedural issues relating to the public/private divide. He notes that in English law there are distinct procedural routes depending on whether one is suing a public or a private body. His point is that we must remember the interrelationship between the procedure and substance of a legal system. Finally, Cockrell challenges private lawyers to develop rules so as to achieve the values of participation and accountability in the private sphere. He suggests that the creation of new, distinct rules for State actors is better than trying to shoehorn State liability into existing private law categories. Since Cockrell wrote his article in 1993 we have seen some such changes (see next topics). 4 The SARFU factors17

The judgment in President of the Republic of South Africa v South African Rugby Football Union 2000 (1) SA 1 (CC) is the most important case in our law on the classification of exercises of public power. More generally, it
17

In light of the Vula notice Corder sent out about this, and the fact that Ive added in stuff he said about the case in second semester, it doesnt make much sense to include this in the public/private divide topic at all. But I couldnt think of anywhere better to put it.

52

is one of the major post-1994 administrative law decisions. As such we discuss it in detail here. The case concerned the challenge by SARFU to President Mandelas exercise of prerogative power in appointing a commission of inquiry to investigate its affairs. There are three Constitutional Court judgments in the SARFU matter. SARFU had, together with its president, Louis Luyt, argued that the Court had no jurisdiction to hear the matter because its members were biased in favour of President Mandela: Chaskalson P had represented Mandela during apartheid, Sachs J had been a leading ANC member, etc. The first judgment related to this issue, with the Court rejecting SARFUs allegations of bias. We are concerned with the third judgment (now generally referred to as SARFU III), which relates to the substance of SARFUs challenge to the Presidents conduct. The matter was obviously politically sensitive because it called into question the Presidents actions. Steve Tshwete, then Minister of Sport, had announced at a press conference that Mandela was going to appoint such a commission of inquiry; and Tshwete had led the (ultimately unsuccessful) negotiations with SARFU prior to the commissions appointment. SARFUs basic argument was that the President had failed to exercise his mind and had improperly delegated his prerogative power to the Minister, contrary to s 84(2) of the Constitution, which specifically says that the President is responsible for appointing commissions of enquiry. They also argued that the terms of the commission of inquiry were invalid because they were too vague (basically the commissions objective was to investigate South African rugby). Generally, the Court was called upon to discuss the principles underlying the review of the Presidents prerogative power. But what is important for our purposes is how the Court dealt specifically with the question of whether the Presidents conduct constituted administrative action as opposed to executive action. The Court begins by noting that:
*T+he test for determining whether conduct constitutes administrative action is not the question whether the action concerned is performed by a member of the executive arm of government. What matters is not so much the functionary as the function. The question is whether the task itself is administrative or not. The focus of the enquiry as to whether conduct is administrative action is not on the arm of government to which the relevant actor belongs, but on the nature of the power he or she is exercising.

The Court goes on to say that the formulation of policy is characteristic of executive action, and the implementation of legislation is characteristic of administrative action. But:
Determining whether an action should be characterised as the implementation of legislation or the formulation of policy may be difficult. It will, as we have said above, depend primarily upon the nature of the power. A series of considerations may be relevant to deciding on which side of the line a particular action falls. Difficult boundaries may have to be drawn in deciding what should and what should not be characterised as administrative action for the purposes of section 33. These will need to be drawn carefully in the light of the provisions of the Constitution and the overall constitutional purpose of an efficient, equitable and ethical public administration. This can best be done on a case by case basis.

The considerations which the Court provides here are known as the SARFU factors for determining whether conduct amounts to administrative action: The source of the power; The nature of the power; 53

Its subject matter; Whether it involves the exercise of a public duty; and How closely it is related to policy matters, which are not administrative, and on the other hand the implementation of legislation, which is.

In SARFU the Court was clearly concerned with deciding whether the conduct was administrative action as opposed to executive action. Our courts have subsequently referred to the SARFU factors whenever they need to decide whether conduct amounts to administrative action. Note that even if the conduct is not administrative action, it will still be reviewable under the principle of legality. The question of whether it is administrative action determines whether it is reviewable specifically under s 33 of the Constitution and PAJA. The all-important last factor was taken up in Permanent Secretary of the Department of Education of the Government of the Eastern Cape Province v Ed-U-College (PE) (Section 21) 2001 (2) SA 1 (CC). The Eastern Cape government disbursed subsidies to independent schools. Ed-U-College was one of these, but it was unhappy with the disbursement. It therefore challenged three different stages in the process: firstly, the decision of the Eastern Cape legislature to allocate a certain amount of money to education generally; secondly, the decision of the same legislature to allocate a certain proportion of that amount to the private sector; and, thirdly, the MECs decision to use a certain formula to determine the subsidy to be given to each school (based on the number of students at the school). The important question, in respect of each of these decisions, was whether they amounted to administrative action and were thus capable of being reviewed under s 33. Specifically, the argument was that they were not administrative but, rather, legislative actions. ORegan J unsurprisingly found, relying on Fedsure, that the first two decisions were indeed legislative actions. The third decision was less clear; it fell close to the borderline between the formulation of policy and the implementation of legislation. However:
I am persuaded that the source of the power, being the Legislature, the constraints upon its exercise, and its scope point to the conclusion that the exercise of the power constitutes administrative action, not the formulation of policy in the broad sense as suggested by the applicants.

Anyway, the next issue for consideration in the SARFU case was the reviewability of the exercise of so-called prerogative powers of the President, those powers formerly held by the sovereign. The Presidents power to appoint commissions of inquiry is one such power, and that is why the Court must discuss these within their general discussion of administrative action. The Court, reaffirming Hugo, says that exercises of prerogative powers are reviewable. The appointment of commissions of inquiry is not administrative action, but it is nevertheless reviewable under the principle of legality. Having established that the Presidents actions in casu are in principle reviewable, the Court addresses the specific grounds of review advanced by SARFU. Their first three argumentsthat the commission did not relate to matters of public concern, as required by the Commissions Act; that the President had unlawfully delegated his powers of appointment to the Minister of Sport; and that the Minister had made an assurance to SARFU obliging him not the appoint the commissionare dealt with in detail by Danwood in second semester. The fourth argumentthat SARFU had a legitimate expectation and as such ought to have been consulted prior to the appointmentis dealt with by Kerry, also in second semester. The final argument was that the terms of the commission of inquiry were so broad (basically, to inquire into South African rugby) that its appointment must be invalidated for vagueness. In other words, this exercise of public power was not undertaken with the reasonable clarity required such that the public can know how to 54

comply. This argument was rejected on the facts, but importantly the Court did hold that vagueness was a ground of review. PAJA has been enacted since the SARFU decision and, although s 6(2) of PAJA is essentially a codification of the existing common-law grounds of review, it does not mention vagueness at all. Nevertheless, it should be read in as a ground of review under s 6(2)(i), the omnibus or catch-all provision.18 5 Powers exercised by public bodies in private law settings

This section considers the application of the SARFU factors in two subsequent cases, where the difficulty was that the State claimed it had been acting in its capacity as a private owner. Note that very similar issues are raised in the next topic on administrative law and contract: the issue there is where the State was exercising a public power specifically when it concluded a contract or exercised a contractual right. In Bullock NO v Provincial Government of North West Province [2004] ZASCA 2119 the Premier had decided to grant a servitude over property on the banks of Hartebeestpoort Dam to the owner of the adjacent property to the exclusion of other persons. The State argued that this decision was made in its position as landowner, involving the exercise of a property right just like any private owner; and therefore was not administrative action. The Court unsurprisingly said FUCK YOU! This is obviously administrative action. In the course of doing so, the Court applied the SARFU factors. The Courts finding that the granting of the servitude constituted administrative action was justified in particular by the fact that *t+he dam is a valuable recreational resource available to the public at large. In Greys Marine Hout Bay (Pty) Ltd v Minister of Public Works [2005] ZASCA 43 the SCA had to consider the application of the Bullock precedent in the post-PAJA era. The State had granted a lease over part of the Hout Bay quayside to Bluefin, a BEE fishery. The appellants were the long-term lessees of the neighbouring sections of the quayside, who felt that they were negatively impacted by the granting of the lease to Bluefin, and as such were entitled to a hearing. In other words, they argued that the States decision was procedurally unfair. They relied primarily on s 3 of PAJA in this regard. Reliance on the terms of PAJA required them first of all to establish that the States conduct amounted to administrative action. It is clear that Nugent JA thinks that PAJAs definition sucks balls, as he writes (in a now-famous passage):
The cumbersome definition of that term in PAJA serves not so much to attribute meaning to the term as to limit its meaning by surrounding it within a palisade of qualifications.

He then sets out a handy, abbreviated form of the definition which captures its essential elements:
Administrative action means any decision of an administrative nature made...under an empowering provision [and] taken...by an organ of state, when exercising a power in terms of the Constitution or a provincial constitution, or exercising a public power or performing a public function in terms of any legislation, or [taken by] a natural or juristic person, other than an organ of state, when exercising a public power or performing a public function in terms of an empowering provision, which adversely affects the rights of any person and which has a direct, external legal effect.
18 19

This was confirmed in New Clicks (see second semester). This case was not discussed in class but it is an important link between SARFU and Greys Marine and is mentioned in Hoexter, so I thought Id include it.

55

But he rejects a literal reading of the phrase which adversely affects the rights of any person, saying that it should be taken to require the lower threshold of having the capacity to affect legal rights. He justifies this conclusion firstly by relying on the nature and purpose of s 33 of the Constitution; and secondly by noting that s 3(1) of PAJA applies when a persons rights or legitimate expectations are affectedthus acknowledging that administrative action has occurred even when rights are not affected. One of the States arguments (of most relevance to this topic) was that it was not administrative action because they were acting in their private rather than their public capacity: they were simply leasing their property, like any private owner would do. The SCA shat explosively on this argument, basically holding that Bullock had already rejected this sort of buffoonery. The Court also quoted SARFU at length20 to explain the test for administrative action. The point Nugent JA made about the interpretation of s 3(1) is important in the resolution of the case, because clearly the appellants rights werent being affected. Perhaps their potential rights or legitimate expectations were threatened, but Nugent JA was not even persuaded of this. The applicants merely had an interest in the granting of the lease to Bluefin, but this interest was not enough to get them over the threshold. On this basis the appeal was dismissed. 6 Private bodies exercising public powers21

We have already acknowledged that with the advent of privatization more and more private entities are exercising the type of public power that the government typically exercises. PAJA explicitly includes such actorsit says any decision taken by an organ of state (as defined in s 239 of the Constitution) or any other natural or juristic person exercising a public power or performing a public function in terms of an empowering provision can amount to administrative actionbut it does not elaborate on what makes powers or functions public. The courts have been developing this concept in an incremental fashion. The natural tendency is to try and include as many private bodies as possible in this definition, and ensure that the State can never evade legal intervention by privatization. This tendency is reflected in certain judicial decisions. But it must sometimes be restrained in the knowledge that subjecting private entities to onerous administrative rules and procedures may undermine the very benefits that privatization is trying to achieve. In the case of Dawnlaan Beleggings (Edms) Bpk v Johannesburg Stock Exchange 1983 (3) SA 344 (W) the Court held that they were prepared to review the decision of the JSE (not actually a statutory body), on the basis that it was under a statutory duty to act in the public interest. The Court also stressed that it exercises a public function. By contrast, in Cronje v United Cricket Board of South Africa the UCB had banned Cronje from membership of or an association with the UCB for life and it was held that this action was not reviewable. The Court differentiated the UCB from the JSE on the grounds that the former had no statutory duty to act in the public interest. The Court also drew support from Coetzee v Comitis, in which it was held that the National Soccer league was performing a public function and its activities were in the public interest, and that therefore its actions did constitute administrative action.

20

I find this very naughty because PAJA had been enacted since SARFU was decided. So PAJAs definition of administrative actionnot the test courts were using before PAJA was enactedshould be used. 21 This section is based completely on Hoexter. Corder did not deal with it at all in class.

56

The most relevant case for this discussion is AAA Investments (Proprietary) Limited v Micro Finance Regulatory Council [2006] ZACC 9. The Constitutional Court concluded that the Micro Finance Regulatory Council, a nominally private entity, exercised public power when it made rules to regulate an exemption from the Usury Act 73 of 1968. The High Court had also concluded that the actions did constitute an exercise of public power. The SCA, favouring a form over substance approach, found that they did not. Yacoob J, in the Constitutional Court majority judgment, wrote that its duty to act in the public interest and the ministerial control over it meant that the Council was exercising public power:
The Councils composition and mandate show that although its legal form is that of a private company, its functions are essentially regulatory of an industry. These functions are closely circumscribed by the ministerial notice. I strain to find any characteristic of autonomy in the functions of the Council equivalent to that of an enterprise of a private nature. The Council regulates in the public interest and in the performance of a public duty.

On this basis it held that the Council was an organ of state within the meaning of s 239 of the Constitution and, as such, it was subject to the principle of legality and the constitutional right to privacy. It was not at issue whether the Councils conduct amounted to administrative action, but remember that PAJAs definition of that term makes use of s 239s definition of organ of state. ORegan J in her separate concurring judgment offered further guidelines:
In determining whether rules are public in character, although made and implemented by a nongovernmental agency, several criteria are relevant: whether the rules apply generally to the public or a section of the public; whether they are coercive in character and effect; and whether they are related to a clear legislative framework and purpose. This list is not exhaustive, nor are any of the criteria listed necessarily determinative.

Hoexter discusses AAA Investments in the context of the relevance of state control over a purported private entity. Although the Court touched on the element of state control it did not refer to the control test which had been used in previous cases. In Directory Advertising Cost Cutters v Minister for Posts, Telecommunications and Broadcasting 1996 (3) SA 800 (T) the High Court held that organs of state are:
institutions which are an intrinsic part of government and those institutions outside the public service which are controlled by the State i.e. where the majority of the members of the controlling body are appointed by the State or where the functions of that body and their exercise is prescribed by the State to such extent that it is effectively in control. In short, the test is whether the State is in control.

On the basis of this test the Court held that Telkom was an organ of state. This control test was later cited by the SCA in Mittalsteel South Africa Ltd v Hlatshwayo [2006] ZASCA 93 to decide whether Mittalsteel (formerly ISCOR) was acting, in the matter giving rise to the dispute, as a public body for the purposes of PAIA (so obviously a slightly different question to whether its an organ of state). It was held that indeed it was. Finally, in Hoffmann v South African Airways 2001 (1) SA 1 (CC), it was held that:
Transnet is a statutory body, under the control of the state, which has public powers and performs public functions in the public interest. It was common cause that SAA is a business unit of Transnet. As such, it is an organ of state and is bound by the provisions of the Bill of Rights in terms of section 8(1), read with section 239, of the Constitution.

57

So state control is by no means the only factor relevant to deciding whether a private body was exercising a public function, but it is a useful one to discuss. 7 Private bodies exercising public powers: Trengove guest lecture 22 (a) Introduction The question here is the extent to which conduct of private bodies is subject to administrative-law review. It has always been possible to review the conduct of private bodies, but only according to private-law standards. The libertarian starting point is that private bodies are not at all accountable to the public for how they behave; they are free to do whatever they want provided it is consistent with private and criminal law. The law does recognize that when private persons club together they assume accountability to one another, and to the organization which they have formed. But the rules which bind such persons are the rules that they have chosen themselves, i.e. the rules of the contract into which they have entered or of the constitution of the club of which they are voluntary members. Such persons are still free to act in their own interest and are not accountable to the public at large. It is true that the courts sometimes read implied rules into the contract or constitution, for example requiring procedural fairness at disciplinary hearings. But these are always subject to contrary agreement by the members. An important consequence of the application of contractual rather than public-law standards is that the only people whose rights are protected are the members themselves. Examples of this private-law review are the various Jockey Club cases and Marlin v Durban Turf Club 1942 AD 112. But administrative law has always had a different set of rules, which are directed at ensuring accountability to the public at large and which do not originate from agreement between the parties. The powers exercised by administrators are bestowed by society at large (classically by means of legislation) and therefore those administrators must be accountable to the public at large. It is these rules that we are considering here, not the rules sourced in the private law of contract. So, to what extent do these public-law rules bind nominally private parties? The short summary is this: The Constitution and PAJA mark a sea change in the answer to this question: where previously one looked at the source of the power and the nature of the body exercising it, we now look at the nature of the power. Let us look at this development in more detail. (b) Common law At common law the focus was on the source of the power in deciding whether it was reviewable. In particular, the source had to be legislation if administrative law was to have application. The rationale was suggested above: when people are empowered by legislation it is presumed that the legislature intended them to exercise those powers in the public interest, and accountability must be imposed in accordance with this intention. Of course, this rationale does not extend to private bodies which get their powers from their own constitution, and so these bodies were free of public-law review. So, to emphasize, the source of the power

22

Obviously this lecture was given in second semester. But it fits perfectly in at this point in the material, so that is where I put it. I have not merged it with the preceding section (based on Hoexter) but Im sure you guys can draw the necessary connections.

58

was the key. If the power was sourced in legislation, it was subject to public-law review. But if it was sourced in non-legislation (basically, contract), it was subject to private-law review only. English administrative law was built entirely upon the ultra vires doctrine. The courts worked on a fiction, namely that Parliament intended its legislation to be implemented lawfully, reasonably and fairly. The courts job, when applying administrative law standards, was supposedly to enforce Parliaments will. The point is that Parliaments intention is absolutely fundamental to all of administrative-law review. Of course, South African administrative law is largely based on this same idea. The most interesting discussion in our law is the case in which Hefer JA tried to break away from the ultra vires doctrine, Staatspresident v UDF 1988 (4) SA 830 (A).23 Van Heerden JA applied the traditional ultra vires doctrine to hold that vague regulations were void; Parliament could not have intended for regulations made under its legislation to be vague. But Hefer JA wanted to uphold the regulations, so he denied that our administrative law rested on the ultra vires doctrine as it was then understood (which allowed review on grounds of vagueness). On this basis he was able to say that the Court had no power to set aside vague regulations unless the empowering legislation specifically prohibited the making of vague regulations (which of course it didnt in this case). Anyway, the point is basically that administrative-law review was clearly based on the ultra vires doctrine (despite Hefer JAs strategic denial of this fact), and this doctrine rests on the idea of Parliaments intention. This doctrine self-evidently cannot be transplanted into a contractual context, because Parliament is not involved in contracts. If our law was going to impose duties of fairness on private parties acting in terms of a contract, therefore, it had to read in implied contractual terms. The identification of legislation as the source of all administrative-law review created serious anomalies, because whether a particular power had its genesis in statute or contract was often a pure accident of history. For example, the Boxing Board happened to have legislation underpinning them. But the rugby, soccer and horseracing boards happened not to have any statutory authority. So if you were a boxer you got administrative-law protection, but if you were a rugby player you did not. In Dawnlaan Beleggings v JSE 1983 (3) SA 344 (W) Goldstone J had to fish about frantically for a piece of relevant legislation upon which he could pin his jurisdiction to review the conduct of the JSE. It was completely fortuitous that he could find some (barely) relevant legislation. Should the applicant really have gotten no relief at all if Goldstone J had not managed to find such legislation? We see similarly arbitrary results with universities, airlines and so on, some of which are public and some of which are private. Why distinguish these? It seems completely arbitrary given that a university or airlines capacity to effect society is in no way dependent on its being public. Private bodies can have just as much of an impact. To take an extreme example, consider the ANC Polokwane conference. The occurrences there led to a change of governmentabout the biggest effect an exercise of power could possibly have on the publicand yet these were not reviewable because the ANC is a nominally private organization. Returning to the sports boards, these have a total national monopoly over an entire sport, and yet they (or at least many of them) are supposedly private and as such immune from administrative law review. In Cronje v United Cricket Board of SA 2001 (4) SA 1361 (T) Hansie Cronje sought the review of the UCBs decision to ban him for life without giving him a hearing. The problem was that Cronje had already resigned from the UCB and so he had no contractual rights against them. Kirk-Cohen J therefore dismissed his application out of hand. This is a particularly stark illustration of the public/private divide (though a correct decision on the law then
23

We will discuss this case again later, though only briefly.

59

existing): simply because the UCB was constituted contractually rather than legislatively, the applicant has no rights at all and as such must fuck off. If he had been an attorney, auditor or boxer, he would have been entitled to review. So, in short, it was patently ridiculous to attach so much significance to the legislation/contract distinction. (c) The Constitution As we said previously, the Constitution marked a sea change. We can see the changes it wrought simply be looking at the text of its provisions. Section 33(1) is of course nice and simple: Everyone has the right to administrative action that is lawful, reasonable and procedurally fair. But the section doesnt say who is under the corresponding duty, and this is what we are really interested in here. We must therefore refer to ss 8(1) and 8(2), which answers this question for all rights in the Bill of Rights. These read as follows:
(1) The Bill of Rights applies to all law, and binds the legislature, the executive, the judiciary and all organs of state. (2) A provision of the Bill of Rights binds a natural or a juristic person if, and to the extent that, it is applicable, taking into account the nature of the right and the nature of any duty imposed by the right.

These provisions embody the two bases upon which private bodies can be reviewed in administrative law. The first is through the use of the term organs of state in s 8(1). As we know, this term is defined (in s 239 of the Constitution) to include any functionary or institution exercising a public power or performing a public function in terms of any legislation. The fact that this section specifically includes functionaries and institutions other than those ordinarily thought of as being governmental seems helpful. But, of course, the phrase in terms of any legislation means that we still need a statutory underpinning, just like under the common law. So this is not the provision which marks the sea change. It is s 8(2) which does this. The formulation used by the drafters (in particular, Halton Cheadle) is extraordinarily vague; the provision is an empty vessel into which we have to pour meaning. The first part of the section says that a right applies to a private person if it is applicable. This gives no content at all. The second part does tell us that whether it is applicable depends on the nature of the right and the corresponding duty. This is still non-specific, but the crucial point is that there is no mention of any statutory underpinning; there is no requirement that the natural or juristic person be acting in terms of any legislation. So s 8(2) read with s 33 is where the decisive break from the past is made. (d) PAJA PAJA also has these two routes by which private bodies can be subject to administrative-law review. The distinction is made in the definition of administrative action. Section 1(i)(a) basically repeats the organ of state definition, thus embodying s 8(1) of the Constitution. Then s 1(i)(b) refers to a natural or juristic person, other than an organ of state, when exercising a public power or performing a public function in terms of any empowering provision. This embodies the s 8(2) basis. The term empowering provision might lead one to think that actually there must be a statutory underpinning even here, but in fact this term is defined very broadly so as to include any statutory law, any rule of common or customary law, or any agreement. This makes clear that the legislative underpinning is no longer essential, and creates the space for holding private bodies accountable even where their powers are not derived from legislation. 60

The upshot, then, is that we dont now need to look at the source of the power at all, nor find a statutory underpinning to justify review. Rather, we simply ask: Is the power being exercised of a public nature? We now consider the way the courts have answered this question. (e) Case law The first case we discuss is AAA Investments v Micro Finance Regulatory Council 2007 (1) SA 343 (CC), with which we are familiar.24 The SCA had said the respondent (the Council) was private, but the Constitutional Court had overturned this decision and said it was public. The Court didnt make clear, however, whether it was conducting its review under the principle of legality or s 33. But anyway, what really concerns us here is that Yacoob J for the majority said the following at para 39:
[T]he exercise of all public power in South Africa is constrained by the legality principle. It is therefore not necessary for the purpose of determining whether the legality principle applies to decide whether the power is governmental.

So the Court is making very clear that the publicness of an exercise of power in no way depends on whether the power is governmental. The approach of Nugent JA in Calibre Clinical Consultants v National Bargaining Council for the Road Freight Industry 2010 (5) SA 457 (SCA) therefore came as a surprise. The case concerned a decision of the Bargaining Council to appoint a particular service provider to manage one of its projects. Calibre was one of the consortiums which had applied for the role, but had been rejected. It tried to challenge the Councils decision, thus prompting the Court to decide whether the Councils actions had been of a public nature. Nugent JA held that the decision was not of a public nature, therefore did not constitute administrative action, and therefore was not reviewable in public law.25 The outcome reached by Nugent JA is probably correct, but Wim is critical of his reasoning. Nugent JA held (predictably) that there are no strict rules applicable when deciding whether an exercise of power is public. But he does list a number of relevant factors. One factor which he identifies as important is whether the power is governmental. He says this without referring to AAA Investments, which had of course said precisely the opposite. Nugent JA just looked at foreign law in deciding that this was an important factor:
Thus in cases concerning the scope of public law judicial review in other countries and most often in this country as well courts have consistently looked to the presence or absence of features of the conduct concerned that is governmental in nature.

The problem with this is that Nugent JA is looking back into the past to see what courts have historically considered important. This loses sight of the fact that the Constitution and PAJA mark a deliberate and decisive break from the past. We must now look at the Constitution to see what public/private means. We see this problem again in the first part of the following quotation:

24 25

This case is discussed much more in second semester, under lawfulness. This is not to say that the provision of services by the Council was not public. The point is simply that the decision on who to appoint to provide the service (the decision at issue in these proceedings) was not public.

61

While curial pronouncements from other jurisdictions are not necessarily transferrable to this country they can nonetheless be instructive. I do not find it surprising that courts both abroad and in this country including the Constitutional Court in AAA Investments have almost always sought out features that are governmental in kind when interrogating whether conduct is subject to public law review. Powers or functions that are public in nature, in the ordinary meaning of the word, contemplates that they pertain to the people as a whole or that they are exercised or performed on behalf of the community as a whole (or at least a group or class of the public as a whole), which is pre-eminently the terrain of government.

Why should only governmental functions be reviewable? Why should we ask only upon whose behalf the powers are performed? Surely the really important question to ask is: On whom do they impact? Nugent JA commits another, related error later on in his judgment, when he tries to justify his focusing on the governmental nature of the power:
It has been said before that there can be no single test of universal application to determine whether a power or function is of a public nature and I agree. But the extent to which the power or function might or might not be described as governmental in nature, even if it is not definitive, seems to me nonetheless to be a useful enquiry. It directs the enquiry to whether the exercise of the power or the performance of the function might properly be said to entail public accountability and it seems to me that accountability to the public is what judicial review has always been about. It is about accountability to those with whom the functionary or body has no special relationship other than that they are adversely affected by its conduct and the question in each case will be whether it can properly be said to be accountable notwithstanding the absence of any such special relationship.

Here Nugent JA seems to be saying that a power is public if exercises thereof are held accountable to the public. But it is when the power is public that it must be accountable. In other words, the whole point of the enquiry into the publicness of the power is to determine whether exercises of that power should be held accountable to the public. To say, therefore, that the power is public if it is (already) publicly accountable is to turn the question on its head, and to undermine its very purpose. Nugent JA therefore seems insufficiently sensitive to the shift which has been occasioned by the Constitution and PAJA. The House of Lords in YL v Birmingham City Council [2008] 1 AC 95, which Nugent JA had cited extensively, did no better. Lord Bingham and Baroness Hale agreed on the set of factorsincluding whether the power is governmentalto be used in deciding whether it is public or private. But they reached different conclusions when applying these factors. The difference is essentially that Baroness Hale attached less weight to the governmental factor. This is better, of course, because asking whether its governmental in nature or the sort of power typically exercised by the government is to foreclose any progressive extension of administrative-law review to nominally private bodies. The Constitution tells us that we must look beyond what acts are, or historically have been, done by the State. One cant look in the rearview mirror to understand the future that is envisaged by the Constitution. We must instead go back to s 8(2), and do what it tells us: look at the nature of the s 33 right and the nature of the duty that it imposes. We must not look at the source of the power or the nature of the body which exercises it. Wim thinks we should adopt the following criteria. A power is public in nature (and thus public-law reviewable) if it is:

62

1. A legal power. This is already captured by PAJA, which requires the power to be sourced in an empowering provision, i.e. any law. So we are not concerned with, say, the exercise of purely commercial powers (e.g. whether company A releases its new product on date x or y). 2. Which materially affects the public (or sectors thereof). We are not concerned with small-scale powers within families or tiny community groups. This criterion does not mean that each exercise of the power affects the public at large. The point is simply that it is the kind of power which has the potential in aggregate to materially affect the public. So each exercise of the power to issue passports affects only one individual, but the power still materially affects the public. 3. Without the affected persons will. By will we do not mean consent in the formalistic contractual sense. When you register with the United Cricket Board you have exercised a choice in the strict sense, but since the UCB regulates the entire sport of cricket in the country you have no real option but to register with them if you want to be a professional cricketer. These are of course Wims tentative suggestions, which definitely are not the law. The last criterion is most interesting. Let us consider again what happened to Hansie Cronje. Under private-law review of course Hansie had no case: since he had left the UCB he had no contractual rights to assert, and that was the end of the enquiry. But nowadays wed have to see whether the power to ban him is public in nature. And one feels that it would be, because in South Africa you cant be a professional cricketer at all unless youre a member of the UCB; the UCB is the sole regulator with a complete monopoly. So, although each cricketer joins the UCB voluntarily, his being subject to the UCBs power is not really in accordance with his will. For this reason, the absence of a statutory underpinning doesnt stop the exercise of power from being public. In National Horseracing Authority of SA v Naidoo 2010 (3) SA 182 (N) Wallis J (who, having been counsel for Hansie in Cronje v UCB, probably felt that he had a score to settle) advocated precisely this line, saying that public-law review should potentially be available in such circumstances:
The contention that there is a blanket exclusion of sporting bodies from the definition of administrative action is not in my view justified by the wording of the provision or, on a comprehensive basis, by the nature of such bodies. The statutory test is whether they are exercising a public power or performing a public function. It seems to me that when this question arises in relation to specific conduct by a sporting body it requires a close examination of the functions of that sporting body. There may well be sport organisations operating in relation to relatively minor sports, which have virtually no regulatory powers. I have in mind sports that attract little public interest, where participation is not dependent on association with the national body and where there is no significant commercial interest. On the other hand larger sporting bodies operating in relation to major sports such as horseracing, football, cricket and rugby stand on an entirely different footing. They exercise a virtually monopolistic control over all aspects of those sports from junior to national levels and are active in the international sphere. Public interest in those sports is massive and the amounts of money generated by these sporting activities are very considerable. A person excluded by one of these sporting bodies from participation in their sport is effectively deprived of their livelihood. I can see no reason why Parliament should have overlooked such bodies in enacting the provisions of PAJA much less deliberately excluded them without saying as much.

So the rationale underlying Wallis Js thinking (and Wims third criterion) would not extend to little neighbourhood soccer clubs. This means a line must be drawn somewhere between these and SAFA. But that line cannot be made rigid; it is inescapably a matter of degree. Requiring strictly that the body have a national monopoly would not be fair: a provincial or even city-wide monopoly may in certain circumstances be enough to deprive people of any will in the matter.

63

We can make similar points with respect to other organisations, like NGOs (which was exactly what was at issue in the YL case). Some NGOs might be one of many performing a similar function. But others may have monopoly power. For example, NICRO helps rehabilitate convicted criminals and, if one is accepted into their program, one will almost invariably get parole earlier. NICRO is nominally private, but if its the only body doing this sort of thing then it has a highly material impact on your life without your will. Even if its just operating under an internal constitution with no official relationship with the prison service, it should be subject to review. In conclusion, Wim doesnt claim to know the answer to the question we posed at the outset. But he does know that the answer doesnt lie in the rearview mirror, motherfucker.

64

STATE LIABILITY: ADMINISTRATIVE ACTION AND CONTRACT

1 Introduction to the following topics

In this topic we are discussing the interface between two parts of the law that are ordinarily kept separate: private law (specifically, contract and delict) and public law (specifically, administrative law). There are differences in the approach to and legal consequences of contracts involving public bodies as opposed to ones involving only private ones. And there are different legal consequences for wrongful behaviour by State bodies or their employees as opposed to private persons. So we need to be very aware, throughout this topic, of the set of rules which we are applying. We also need to bear in mind courts obligation under s 39(2) to interpret legislation and develop the common law in light of the spirit, purport and objects of the Bill of Rights. Also, in this regard, recall Cockrells challenge to private lawyers to develop rules to foster participation and accountability within the domain of private law. Indeed, everything we said about the public/private divide has application here. 2 Public procurement

One key section of the Constitution, which we havent encountered before, is s 217:
217 Procurement (1) When an organ of state in the national, provincial or local sphere of government, or any other institution identified in national legislation, contracts for goods or services, it must do so in accordance with a system which is fair, equitable, transparent, competitive and cost-effective. (2) Subsection (1) does not prevent the organs of state or institutions referred to in that subsection from implementing a procurement policy providing for(a) categories of preference in the allocation of contracts; and (b) the protection or advancement of persons, or categories of persons, disadvantaged by unfair discrimination. (3) National legislation must prescribe a framework within which the policy referred to in subsection (2) must be implemented.

This set of requirements provides the starting point in the law of public procurement. Is it permissible for ministers to stay in the most expensive hotel in Cape Town given the cost-effective requirement of s 217(1)? Note that BEE and similar schemes are constitutionally permitted in terms of s 217(2). The Constitutional Court decided relatively early on that the tender process (advertising, considering applications, and awarding tenders) did amount to administrative action. Accordingly it is subject to PAJA. A lot of case law has emerged on this issue, with rejected parties challenging the decision to award the tender to someone else. But what about the contracts the State actually concludes with the winning tenderer? Is the actual act of contracting, and he States subsequent performances under the contract, subject to special rules given that the State is one of the parties? Or do the general principles of the law of contract apply unaltered?

65

State liability

Historically, citizens had no right at all to sue the British monarch. This immunity influenced our own commonlaw rules. The result was that State liability had to be made possible by the enactment of a statute. Thus we have the State Liability Act 20 of 1957. It is a short act, which has three noteworthy provisions. Firstly (s 1), the State can now be sued. Secondly (s 2), the minister of the relevant department must be cited as the defendant. Thirdly (s 3), no attachment of State property may be sought in order to satisfy any judgment against the State. This last section is obviously the most interesting one. The case of Nyathi v MEC for the Department of Health Gauteng 2008 (5) SA 94 (CC) arose in response to this naughty third section. Mr Nyathi had successfully been awarded damages against the State because of the injuries he had suffered due to negligent medical treatment at a State hospital. The State never paid up, but he couldnt, because of s 3 of the State Liability Act, have recourse to attachment to satisfy the judgment debt. Thus he went to the Constitutional Court to have that provision declared unconstitutional. The Court held as follows:
Section 3 of the State Liability Act is declared to be inconsistent with the Constitution to the extent that it does not allow for execution or attachment against the state and that it does not provide for an express procedure for the satisfaction of judgment debts.

But, cognisant of the time it takes to make legislative changes, the Court suspended the operation of the order of invalidity for 12 months. But when the deadline came around, on the 1 June 2009, the Minister of Justice asked for an extension. He received an extension, until 31 August 2009. When that date came, Parliament still hadnt gotten its act together, so a further extension was granted, with the deadline now set back a further two years (!). But the Constitutional Court was pissed off with Parliament, and there was a catch: the Court required the Minister, in cooperation with the National Treasury, to put together an expedited procedure for people in Mr Nyathis position to have recourse pending the amendment of the Act. The State Liability Amendment Bill of 2011 provides, most importantly, that one can now attach the States movable property, provided firstly that it is not property the attachment and execution of which would severely disrupt service delivery, threaten life or put the security of the public at risk (so e.g. you cant attach vital medical supplies or all the police forces guns); and secondly that the sheriff gives 30 days notice of his intention to attach the specified property before he actually does so. This gives the State an opportunity to pay up first. The upshot, then, is that the State can be held liable, and one can get an attachment order against them. Soon (hopefully) this will be in terms of the amended State Liability Act. In the meantime it is in terms of the special interim procedure. 4 Hoexter: reviewing contractual disputes involving public bodies

Insofar as we discuss the specific issue of State contracts, we will rely almost exclusively on Hoexters article prescribed under this section. Most of her discussion surrounds the termination of a contract entered into between a public body and a private entity.

66

She begins by distinguishing between the purely contractual and public law approaches which a court may take in the adjudication of contractual disputes involving the State. She criticizes the purely contractual approach for being an example of formalism, which she defines as follows:
A judicial tendency to attach undue importance to the pigeonholing of a legal problem and to its superficial or outward characteristics; and a concomitant judicial tendency to rely on technicality rather than substantive principle or policy, and on conceptualism rather than common sense.

Nevertheless she cautions against seeing formalism as the specific province of conservative or executiveminded judges, noting that during apartheid Didcott J strategically used formalism to render ouster clauses ineffective. She also says that although the public law approach is better than the purely contractual approach, both are flawed and permit formalistic reasoning. This obscures the substance of the matter and prevents courts from addressing directly the fundamental question of what administrative justice demands on the facts before them. Hoexter suggests that South Africas attempt to move towards substantive justice must occur in conjunction with purposive and policy-oriented modes of legal reasoning. As Karl Klare says:
[T]he drafters cannot have intended dramatically to alter substantive constitutional foundations and assumptions, yet to have left these new rights and duties to be interpreted through the lens of classical legalist methods.

The culture of justification, accountability, transparency and administrative justiceHoexter arguesare not served through a formalistic approach. Moreover, the purely contractual approach depicts the legal relationship between parties as a simple matter of consensus, with freedom of contract unduly elevated and no recognition that one of the parties is public in nature, having certain public duties and immense bargaining power. The traditional law on State liability in contract, says Hoexter, is overwhelmingly formalistic; most courts have in the past adopted the purely contractual approach. She refers to a number of cases, including the wellknown Mustapha v Receiver of Revenue Lichtenburg 1958 (3) SA 343 (A). In this case, a trader classified as Indian had his license summarily terminated. He challenged the termination on the grounds of procedural fairness. But the majority of the Appellate Division held that this was a simple contractual relationship, so the State had no obligation to accord with principles of procedural fairness. Although the Court conceded that the terms and conditions of the permit were prescribed by legislation (and certain regulations made under it), the terms and conditions were non-existent unless and until a permit [was] granted and accepted in relation to some particular site. But, once the terms were accepted by the licensee, the parties relationship was grounded in consensus and thus their rights were then defined by contract law alone. Hoexter is very critical of this approach, saying that the Courts reasoning is wholly artificial and subverts the substance or reality of the situation to form and doctrine. She says there is nothing consensual about the licensing process: the applicant could not negotiate the terms nor could the strictures dictated by the State constitute an agreement. The Courts reasoning also severs altogether the contract from the requirements contained in the legislationeven though the contracts existence and validity depends on the legislation. The better, contrary approach is the public law approach, which focuses on the public nature of one of the contracting parties and particularly to the statutory source of its contractual powers. The underlying purpose 67

of such an approach is that the contractual rights of public bodies may be exercised only in certain ways: lawfully, reasonably and fairly. This approach is illustrated by Schreiner JAs dissenting judgment in Mustapha. He reasons that the States powers must be exercised within the express or implied limitations of the relevant legislation and the regulations made under it. But, towards the end of apartheid, two cases emerged in which the Appellate Division indicated that certain of its judges (both majority judgments were written by Hoexter JA, Prof Hoexters father) favoured the public law approach. Both cases involved litigants who, in the absence of a bill of rights, sought the protection of administrative justice. In Administrator, Transvaal v Zenzile 1991 (1) SA 21 (A), healthcare workers at Baragwanath were striking. The black Africans among them were all temporary workers, no matter how long they had worked for the hospital, because apartheid laws prohibited their being permanently employed. This meant that, in terms of their temporary work contracts, they could be fired on 24 hours notice. The Administrator of the Transvaal (the equivalent of a Premier) said that if the workers did not return to work within 24 hours they would be fired. This would be perfectly lawful in terms of the contract of employment, had the employer been private. But, since the employer was the State, Zenzile and his co-applicants argued that procedural fairness had to be observed, specifically that a hearing must be given before they were fired. This argument was upheld in both the High Court and the Appellate Division, evidencing the public law approach. Hoexter JA found that the relations could not be defined as purely contractual and indeed that this was no ordinary contract: the employer was a public authority whose decision to dismiss the applicants involved the exercise of a public (that is, statutory) power. Thus due regard for the principles of natural justice was required. This was taken even further in Adminstrator, Natal v Sibiya 1992 (4) SA 532 (A), on trickier facts. Again the litigants were State employees, this time in Natal and on annual26 rather than temporary contracts. The State decided not to renew the workers contracts when the one year period expired, citing a lack of funds. The litigants challenged this action on the groundsas in Zenzilethat they were entitled to a hearing. This was a trickier argument for the Court, because no punitive action was being taken: the workers were not being fired; their contracts just werent being renewed. Nevertheless, Hoexter JA boldly found for the workers. He came very close to recognizing a right to work, saying that the non-renewal of their contracts, by depriving them of their expected wages, affected their property rights. Because this right was negatively affected, the workers were entitled to procedural fairness. Hoexter says that in both these cases the Courts essential question was, What does administrative justice require? This makes them examples of the public law approach, in contrast to the formalist approach which asks simply, What do the existing doctrines of contract law require? She goes on to argue that the public law approach is preferable (albeit imperfect). The public law approach is more than just Hoexters preferred approach: thanks to Logbro Properties CC v Bedderson NO [2002] ZASCA 135 it appears to be fixed in our post-apartheid administrative law. The KwaZuluNatal provincial government had awarded a tender for the sale of a valuable piece of property in Richards Bay. Logbro, one of the rejected tenderers, successfully had the award invalidated by a court on the grounds that the accepted offer did not comply with the tender conditions. When the tender committee reconvened to award the tender anew, Logbros bid for the property was now the highest. But, because of the increase in the value of property in the Richards Bay area since it had last convened, the committee decided to call for
26

Black African employees were not given contracts lasting more than 365 days, because to do so would allow them to claim urban residence rights on the grounds that theyd worked for an uninterrupted period of X years.

68

completely new tenders. Logbro objected to this in the High Court, which rejected its argument on the grounds that the terms of the previous court order permitted the committee to reconsider the matter completely. Logbro appealed to the SCA. Cameron JA began by saying, with reference to numerous cases, that the tender process was administrative action under the Constitution. It seemed prima facie, therefore, that administrative justice principles would apply to the exercise by the State of contractual rights, where the contract in question was concluded pursuant to the award of a tender. But Cameron JA had to deal with the States argument to the contrary, namely that Mustapha and the much more recent case of Cape Metropolitan Council v Metro Inspection Services (Western Cape) CC 2001 (3) SA 1013 (SCA) had shown that the State had a contractual right to withdraw any property from a tender without having to pass the scrutiny of lawful administrative action. In Cape Metro the SCA had to decide whether the cancellation of the agreement amounted to administrative action, to which they answered No. The Court listed the SARFU factors in order to answer this question. It held that, although the Council had derived its power to enter into the contract from statute, it had derived its power to cancel the contract from the terms of the contract and the common law. Thus the Court held that the termination of the contract was neither the performance of a public duty nor the implementation of legislation; and thus was not administrative action. The fact that the termination itself was the exercise of a purely contractual power was the basis upon which the Court distinguished Zenzile and Sibiyain both of which a statute was relevant to the termination. There was in fact a set of regulations that had obvious application to the Councils termination of the contract. The Court admitted that if the Council had purported to cancel the contract on the basis of those regulations then their actions would have constituted administrative action. As it was, they chose to terminate the contract on the basis of the contracts own terms, so their actions were not administrative action. That is, the mere fact that they chose to terminate on the basis of the contract and not the regulations meant that public law was excluded. Hoexter understandably says this reasoning is bizarre. She says that although the Court purported to apply the factors laid down in SARFU, application of those factors actually pointed to the use of public power. Moreover, there was relevant legislation which made provision for the termination of such an outsourcing agreement on grounds including fraud. That is, the Council could have relied on this legislative authority to terminate the contract. The fact that they chose to construe their cancellation as the exercise of an ordinary contractual remedy should be neither here nor there. In a nutshell Hoexters problem is that the Court applied a purely contractual approach. Anyway, back to Logbro: Cameron JA distinguished Cape Metro on the grounds that it dealt with unique facts, which the Court had carefully distinguished from other tender cases. Moreover, and crucially:
The case is thus not authority for the general proposition that a public authority empowered by statute to contract may exercise its contractual rights without regard to public duties of fairness. On the contrary: the case establishes the proposition that a public authoritys invocation of a power of cancellation in a contract concluded on equal terms with a major commercial undertaking, without any element of superiority or authority deriving from its public position, does not amount to an exercise of public power.

So Cameron JA is basically saying that it was crucial to the Courts reasoning in Cape Metro that there was no inequality of bargaining power between the contractual parties, and therefore no reason to apply administrative law. Where this is not soas on the facts before itCape Metros finding on this score has no relevance; and as such the Province was therefore burdened with its public duties of fairness in exercising the

69

powers it derived from the terms of the contract. So Cameron JA has taken the sting out of Cape Metro by distinguishing it. As to Mustapha, Cameron JA unsurprisingly vomits all over the majority judgment for doctrinal and historical reasons. He concludes by saying:
[I]it is necessary for Mustapha now to be overruled, and for the dissenting judgment of Schreiner JA to be recognised as correct.

Thus the Court emphatically states that the general principles of contract law must yield, where the State is involved, to principles of administrative justice:
Even if the conditions constituted a contract , its provisions did not exhaust the provinces duties toward the tenderers. Principles of administrative justice continued to govern that relationship, and the province in exercising its contractual rights in the tender process was obliged to act lawfully, procedurally and fairly. In consequence, some of its contractual rights such as the entitlement to give no reasons would necessarily yield before its public duties under the Constitution and any applicable legislation.

In applying this law to the facts, the SCA held that it was not in a position to order the Province not to take the increase in property values into account. The reason is that the tender committees functions are a prime instance of polycentric decision-making. The Court therefore felt compelled to show due deference to the committee and allow it to reconsider the decision anew in light of the change in circumstances, and, if it chose to do so, to withdraw the tender offer altogether. On the other hand, the Court held that the committee was required, by the dictates of procedural fairness, to give the original tenderers the opportunity to make representations, at least in writing, on any supervening consideration relevant to the committees exercise of its powers in relation to the award or non-award of the tender. This was not a particularly stringent requirementthe committee remained free to ignore the representations and withdraw the tender offerbut it was sufficient grounds for the Court to invalidate the committees previous decision. Hoexter gets dripping wet for the Logbro case, and says that it is admirable in several respects. Firstly, it rejects the purely contractual approach. Secondly, it emphasizes that the Constitution demands that the principles of administrative justice must frame the contractual relationship wherever public bodies are involved. Section 33 of the Constitution gives everyoneincluding parties to a contractthe right to just administrative action. That is, the State cannot, by entering into a contract, escape its public duties, in particular the duties to act lawfully, reasonably and fairly. This does not mean that contractual terms can have no effect: Cameron JA suggests that contractual terms may be able to bear on the exact ambit of these ever-flexible duties. Other considerations that would have an impact are obviously the presence or absence of consensus and the parties relative bargaining power (as stressed in Cape Metro). Hoexter says that what Cameron JAs approach to the Cape Metro precedent was to emphasize the substantive factors which it relied upon, rather than to simply apply a formalistic approach. At the end of her article, Hoexter rides her favourite hobby-horse and says that too much time is spent by courts deciding on whether the impugned conduct is administrative action and not enough time on what administrative justice demands. She says that a focus on drawing up lists of factors that would constitute administrative action leads us to become obsessed with conceptualism (not formalism as such, but an overly 70

rigid attempt to pigeonhole the issues). She argues that this has an all-or-nothing consequence: the conduct is either administrative action, and thus reviewable under PAJA; or it is not. She then goes on to say (as weve remarked many times) that PAJAs narrow definition of administrative action explains why the Constitutional Court uses the principle of legality more and more, as the latter imposes requirements of lawfulness on every exercise of public power. But this isnt ideal given the narrower range of protections offered here. We see thatwith the exception of employment contracts, which are dealt with in numerous conflicting cases which we will discuss in a future topicthe law of contract has been substantially reworked post-1994 in an attempt to handle the State differently to other contracting parties, and thus to adopt the public law approach. We will see that it has been much less progressive in the law of delict. 5 Fairness in contract27

It is manifest in Hoexters argument that she thinks administrative law should be allowed to permeate contract law because administrative law ensures fairness. But it ought always to be borne in mind that the modern law of contract is not completely devoid of the value of fairness. Also remember, in this regard, Cockrells suggestion that private lawyers must take up the challenge to develop appropriate rules on state liability and foster public law values even in private lawadministrative law shouldnt have to step in and do all the work. There is no general presumption in favour of procedural fairness in contract law, but in some instances courts have recognised implied terms to this effect, for example in disciplinary hearings of churches, unions, professional associations, privately owned schools and various other bodies. However, our courts have also recognised that the contract between two parties may explicitly exclude procedural fairness. In the preconstitutional era, even public bodies were free to do this. On this count, then, contract law has not done enough, and administrative law is justified in interfering. We do know that our contract law has distanced itself in the last twenty-five years or so from the libertarian veneration of freedom of contract and moved towards a more interventionist and fairness-oriented view of contractual relations. This can be understood as a breakdown in the traditionally sharp divide between the public and the private spheres: there is increasing recognition that private bodies can wield immense power and must therefore be restrained in much the same way as the State. The most obvious manifestation of this recognition is in interventionist legislation like the Consumer Protection Act. There is also the landmark case of Barkhuizen v Napier, whichalthough regarded by many as a missed opportunity to give a robust role for constitutional values in contract lawdoes gesture towards a more interventionist role for courts. Despite these developments, being labeled private certainly does make your life easier. The real challenge, therefore, is to develop an appropriate and encompassing definition of public body and public power. We noted in the previous topic that our post-1994 law takes a progressive approach to this matter, looking at substance and not form. 6 Fettering by contract

It is a well-entrenched principle of good governance that public bodies should not fetter their future discretion. The rationale for this is that a public body may, if it does fetter its future discretion, prevent itself
27

This was not covered at all in class. These notes are from Hoexter and just my wanking.

71

from acting in the public interest in the future. Thus contracts entered into by government may, on a strict application of this principle, be unenforceable if they fetter the future exercise of statutory powers. This is obviously very unrealistic given that government enters into thousands of contracts all the time that may well have such effects. To mediate this tension, our courts seem to have adopted a test of compatibility. The test asks whether the contractual undertaking is compatible with the purpose of the power that it fetters. Only those contracts that are incompatible with the very purpose for which the power has been conferred will be regarded as void. Our courts consider a number of other factors in order to balance the public and private interests involved: the existence of an express or implied mandate to exercise the power by contract, the importance of the powers allegedly being fettered, the extent of the fettering and the likelihood of its in fact constraining future decision-making. If the contract is indeed invalid on this basis, how ought the victims (i.e. the other party, who thought it had a valid contract with the State) to be compensated? There does not appear to be a clear answer to this question. Hoexter suggests that a remedy should be read into s 8(1) of PAJA, which allows a court to grant any order that is just and equitable.

72

STATE LIABILITY: ADMINISTRATIVE ACTION AND DELICT

1 Introduction

In this topic we consider the circumstances in which negligent State conduct will be delictually wrongful (and thus capable of furnishing delictual liability). Are the circumstances the same as for private parties? In other words, do special obligations rest on the State which create a higher standard of conduct such that delictual liability arises more easily against the State than against private persons? If not, must delict be developed in a public law direction? We are already familiar with certain tensions that one encounters when trying to blend two areas of law. On the one hand, there is some value in keeping the areas separate so as to develop coherent rules within each area. On the other, keeping the areas hermetically sealed can simply lead to unjust results, for example if the State is not treated differently by the rules of private law. The dominant view is that doctrinal purity needs to be sacrificed to make the rules fairer. It is worth noting, in relation to the particular context of administrative law, contract and delict, that administrative law was stretched too thin during apartheid, as it was forced to fill the unjust lacunae in the legal system, especially the absence of a bill of rights. This meant that administrative laws own coherence suffered. It may be that, by developing the principles of contract and delict so as to relieve administrative law of its unduly heavy burdens, we can actually achieve greater overall doctrinal coherence. We already know about the cases of Carmichele, Van Duivenboden and Van Eeden, which developed the common law of delict via s 39(2) of the Constitution insofar as it relates to the State, and thus changed the circumstances under which it would be held delictually liable. But this is only part of the story: we will see, with reference to the two seminal cases of Knop and Steenkamp, that our law in certain respects takes a very narrow view of the States delictual liability. What is the relevant difference between the two sets of cases such as would justify their being handled so differently? The answer is probably that Carmichele and others dealt with physical injury whereas Knop and Steenkamp dealt with pure economic loss resulting from the States commercial dealings. As we know, this difference is a very significant one because delictual wrongfulness is much harder to establish in relation to pure economic loss. Another issue we should flag at the outsetbut which will become much clearer as it plays out in the casesis the answer to the question, Why would a litigant care to sue in delict rather than administrative law? The answer is that the two give very different remedies. Generally, judicial review is an inappropriate remedy for an individual who has suffered loss as a result of administrative action as it is primarily designed to set aside the unlawful action rather than compensate the individuals who it has affected. So the better remedy will often be in delict. Nevertheless, a delictual action will almost invariably require the court to pronounce on the lawfulness of the administrators exercise of power. Only by doing this can the court establish whether the administrator has breached any duty. So this basically amounts to an indirect review of the administrators decision. A breach of duty is a necessary condition for a finding of delictual wrongfulness, but it is not sufficient. That is, even if the court finds that the administrator did act ultra vires, this does not by itself establish delictual wrongfulness. It might be that it is unreasonable to grant the plaintiff a right to compensation notwithstanding the administrators naughty behaviour. Thus the wrongfulness enquiry is the focus of a lot of the case law.

73

Case law

The first relevant case is Knop v Johannesburg City Council 1995 (2) SA 1 (A). Knop was an experienced property developer who bought a piece of land near Joburg on which he planned to build cluster housing. The town planning regulations required that no plot for a cluster house be smaller than 250m2although an exemption could be granted by the Council following a notice period, the hearing of objections from the neighbours, etc. Knop submitted plans for the development which had plot sizes of 200m 2. In what was clearly a failure to comply with the town planning regulations, the Council gave permission in respect of these plans (without any exemption procedures being followed). Knop proceeded to begin work on the development. Sometime later, a complaint was laid about the faulty plans, and the Council ordered Knop to stop building until theyd gone through the exemption process properly (which took 10 months). This allegedly caused Knop a whole lot of problems, because he couldnt transfer the erven hed already sold, he had to keep paying interest on the money hed borrowed to finance the project, his reputation suffered, and so on. Therefore he sued for compensation in the amount of R553 000 on the basis of three causes of action: 1. The Council was negligent in granting him permission despite the flaw in the plans. 2. The Council was negligent in misrepresenting the minimum required size for the erven. 3. The Council acted unreasonably in halting the project for 10 months, and thus negligently caused loss to him. The Council excepted to Knops pleadings on the basis of three essential grounds: 1. The law confers an immunity from claims for negligence in respect of the decision taken by the local authority. 2. There was, in any event, no duty of care owed to the plaintiff. 3. The decision to grant planning permission was unlawful, and therefore a nullity; so Knop was not entitled to rely on that decision as the grounds for a cause of action in law. (This raises the difficult issue, which we will see in Oudekraal next semester, of how to unwind an invalid administrative actioncan legal consequences be attached to that invalid decision?) In the court a quo, MacArthur J employed the classification of functions approach, holding that town planning decisions are quasi-judicial in nature (in that they involve the application of legal rules, the weighing up of competing interests, and the exercise of a discretion). But in our law one cannot sue a judge for his bona fide decision, so Knops claim was dismissed. On appeal to the Appellate Division, much of the preliminary discussion centered around the merits of the classification of functions approach which MacArthur J had applied. Relying on the forceful criticisms of this approach expressed by Corbett CJ in Administrator, Transvaal v Traub 1989 (4) SA 731 (A)that even English law (from whence it came) has abandoned it, and that the distinctions it involves are hard to drawBotha JA held that it is of little value in resolving the matter before them (viz. delictual liability for loss caused by the negligent exercise of public power). Having then discussed the basic elements of a delict, Botha JA notes that the case turns on wrongfulness:
In the present case, if it is assumed that the Council was negligent in exercising its statutory functions, the question whether it is liable in damages to the plaintiff must depend on the answer to the question whether its conduct was wrongful.

74

In the wrongfulness enquiry, the nature of the function being performed is just one factor to be considered. Botha JA then discusses the basic nature of the wrongfulness enquiry, with reference to all the usual cases like Ewels. The point to note is that Botha JA is giving a textbook account of orthodox doctrine; the Court is situating its dialogue squarely within the private law of delict. The Court then applied the law to the facts to decide whether the Council had acted wrongfully. The Court reasons that the town planning regulations provide for a disgruntled applicant to take the matter on appeal at a full hearing; but there is no suggestion in the regulations that a damages claim was also permitted. Thus the crux of the judgment:
As to the intention of the legislature, the fact that it has prescribed a particular form of procedure by which an aggrieved applicant can obtain relief against the refusal of his application shows by necessary implication that it did not intend a negligently incorrect refusal to give rise to an action for damages. As to the broader considerations of policy, on the one hand an aggrieved applicant does not need an action for damages to protect his interests; he has readily at hand the appeal procedure provided within the legislative framework. On the other hand, considerations of convenience militate strongly against allowing an action for damages; the threat of such an action would unduly hamper the expeditious consideration and disposal of applications by the local authority in the first instance. That is not to say that the local authority need not exercise due care in dealing with applications; of course it must, but the point is that it would be contrary to the objective criterion of reasonableness to hold the local authority liable for damages if it should turn out that it acted negligently in refusing an application, when the applicant has a convenient remedy at hand to obtain the approval he is seeking. To allow an action for damages in these circumstances would, I am convinced, offend the legal convictions of the community.

This established, for Botha JA, that if the Council had incorrectly refused planning permission, it could not be delictually liable. To draw the inference that the same was true if the Council had incorrectly granted planning permission, he writes:
A curious feature of [the facts alleged] is that the plaintiff is not complaining about a refusal of his application by the Council. His complaint is that the Council negligently granted it. In effect, his case is that the Council committed a wrongful act by giving him what he was asking for. The answer to this somewhat startling proposition would seem to be quite simply this: if the Council would not have been liable to the plaintiff if it had wrongly refused the application (as I have been at pains to show), it is inconceivable that it could be held liable for having incorrectly allowed it.

Nevertheless, Knop argued that, whether or not he had asked for the Council to accept his application, the Council was under a statutory duty to ensure that the subdivision did not fall foul of the town planning scheme and that it had obviously breached this duty. The Court agreed that this was sobut that the duty breached was not one owed to the plaintiff. Rather, it was a duty aimed to promote public order by ensuring that property developments take place in an orderly fashion and in the interests of the inhabitants of the area. Thus this duty was of no assistance in establishing that the Council owed a duty to Knop such as would justify the awarding of compensation for his pure economic loss. Knops claim was therefore rejected. Hoexter refers to Telematrix (Pty) Ltd t/a Matrix Vehicle Tracking v Advertising Standards Authority SA 2006 (1) SA 461 (SCA),28 where the Court had to consider the delictual liability of an independent regulatory body

28

This case was not mentioned in lectures.

75

for a negligent decision requiring the plaintiff to withdraw certain of its advertisements (thus causing the plaintiff loss). Harms JA held that litigants were not entitled to a perfect process and that public accountability had not evolved into a general liability for damages for imperfect administrative decisions. Hence, incorrect or negligent adjudicative decisions do not ordinarily give rise to a claim for damages excepting cases where fraud or mala fides is presentbecause otherwise the body will become too afraid to make decisions lest it expose itself to liability. These two major pointsfirstly that bad administration does not always establish delictual liability and secondly that the reasons for this being so are ones of policyare amplified in the next case. That case is Steenkamp NO v Provincial Tender Board of the Eastern Cape 2007 (3) SA 121 (CC). The Eastern Cape government was offering a tender for a contract to provide social assistance grants. Some businessmen hastily submitted papers for the incorporation of a new company, Balraz, through which they applied for and were awarded the tender. In reliance on the tender agreement, Balraz spent R4m in setting up its operations. About a year after the tender had been awarded, one of the rejected tenderers successfully obtained a court order to have the award declared irregular and therefore invalid. The Tender Board then reconsidered the matter, but Balraz had by then gone into liquidation and was unable to apply. Steenkamp, Balrazs liquidator, sought damages from the Tender Board on the basis of two causes of action: firstly that the Board had breached the contract between it and Balraz, and secondly that the Board was delictually liable as a result of its having caused loss to Balraz through its negligent failure to comply with the regularities in the award of the tender. The Tender Board successfully excepted to the contractual claim in the High Court, but the delictual claim was heard on its merits. The High Court said that the Board did owe tenderers a duty of care and could, in principle, be delictually liable for failing to act properly and with due care in awarding tenders. But there was one major problem: when Balraz submitted its tender it was not registered as a companythe registration process had not yet been completedand for that reason it had no legal personhood whatsoever. There was therefore no way that, at the time of the conduct complained of, the Board could have owed any legal duties to Balraz. For this reason the action was dismissed. On appeal to the SCA, the matter was again dismissed. There were two reasons for this. Firstly, the wrongfulness element was not satisfied, because policy reasons precluded the recovery of pure economic loss resulting from the bona fide award of a tender. Secondly, and even if wrongfulness was satisfied, the tender was a nullity at its very inception because on the closing date for submission of tenders Balraz had not been incorporated, and thus it had no standing whatsoever to complain (in other words, it agreed with the High Court). But, undeterred, Steenkamp took the matter to the Constitutional Court, which gave three judgments. Writing for the majority, Moseneke DCJ sums up the stakes at the start of his judgment:
This case raises the complex debate on the proper interface between private law and public law remedies in our constitutional dispensation. The narrow issue is whether financial loss caused by improper performance of a statutory or administrative function should attract liability for damages in delict. On the facts, the issue may be rendered as whether a successful tenderer whose award is later set aside by a court on review may claim delictual damages from the tender board for out-of-pocket expenses incurred subsequent to and in reliance on the award.

76

As Moseneke DCJ notes, the applicants claim is couched in the private law of delict. And yet, the applicant argued that the conduct complained of was administrative action and that the legal duties alleged to have been breached are sourced not in the common law but in our constitutionally-informed administrative law. Moseneke DCJ begins with a discussion of administrative laws typical role:
It goes without saying that every improper performance of an administrative function would implicate the Constitution and entitle the aggrieved party to appropriate relief. In each case the remedy must fit the injury. The remedy must be fair to those affected by it and yet vindicate effectively the right violated. It is nonetheless appropriate to note that ordinarily a breach of administrative justice attracts public law remedies and not private law remedies. The purpose of a public law remedy is to pre-empt or correct or reverse an improper administrative function.

The suggestion is, perhaps, that one ought to phrase your claim in the appropriate terms for the relief you are seeking. Anyway, the key question in this dispute is whether administrative law furnishes a remedy for damagesobviously a typical private law remedy. On this question, Moseneke DCJ writes:
[A] concession that the tender board acted inconsistently with the tenets of administrative justice is neither decisive of the existence of a duty of care nor is it of any avail to the applicants case. In our constitutional dispensation, every failure of administrative justice amounts to a breach of a constitutional duty. But the breach is not an equivalent of unlawfulness in a delictual liability sense. Therefore, an administrative act which constitutes a breach of a statutory duty is not for that reason alone wrongful.

This definitely does not mean to say that South Africa grants delictual immunity to the government:
Unlike in other jurisdictions, this does not mean that the government enjoys delictual immunity when performing its functions, but a negligent statutory breach and resultant loss are not always enough to impute delictual liability. Policy considerations of fairness and reasonableness have to be taken into account when imposing a duty of care and ultimately liability to make good harm suffered by a claimant.

So wrongfulness, as always, is a flexible requirement. Moseneke DCJ discusses the broad test, and notes the various policy factors which inform it, citing Knop, Van Duivenboden and others (including Olitzki, which the minority thought was very weak authority: see below). He concludes by saying:
The ultimate question is whether on a conspectus of all relevant facts and considerations, public policy and public interest favour holding the conduct unlawful and susceptible to a remedy in damages.

Having regard to these policy factors, is the Board delictually liable? Moseneke DCJ answers No. His basic reason is that the tenderer should negotiate with the State about a right of restitution of out-of-pocket expenses should the tender award be set aside, and thus regulate the matter in contract:
A negotiated or contractual remedy of this order is likely to be effective because it would be tailored to the peculiar facts connected to the actual delivery of supplies and services to the state. This avenue is bound to be better suited than a blunt remedy of recognising a generic duty of care in relation to out-of-pocket expenses incurred on the back of a tender award.

He also cites two very familiar policy reasons in favour of a conservative approach: firstly (citing Telematrix) that adjudicators should be immune from damages claims for their honest decisions; and secondly that there is a floodgates risk if any initially successful tenderer who later had its tender set aside by court order could 77

claim damages. Thus, despite his recognition that the common law must be developed in light of s 39(2), Moseneke DCJ provides numerous reasons in favour of judicial conservatism. Langa CJ and ORegan J co-wrote a dissenting judgment, in which they reach a contrary finding on the issue of wrongfulness. The majority had relied on Olitzki Property Holdings v State Tender Board 2001 (3) SA 1247 (SCA), in which Cameron JA held that a rejected tenderer could not recover his lost profits. The majority used this judgment as a basis to hold that an initially successful tenderer could not recover out-of-pocket expenses in reliance on a tender award which turned out to be invalid. But the minority holds that Olitzki is distinguishable, that although Cameron JAs reasoning is correct on those facts, one cannot successfully carry his reasoning over to the facts in casu. There are two reasons for this. Firstly, a rejected tenderer is in a very different position to a tenderer who was initially successful but who then discovered that the acceptance was granted in error. Not only is the latter tenderer unable to seek redress through judicial review, but also he is (or at least appears to be) legally bound to perform under the contract. Moreover, it would undermine the constitutional commitment to efficiency if a successful tenderer was constantly looking over its shoulder in the fear that its contract would be declared void. Secondly, a loss of profits claim (which a rejected tenderer will bring) is typically much larger than an outof-pocket expenses claim (which Steenkamp is bringing), and thus the policy reasons in favour of immunizing the State from loss of profits claims does not translate with equal force to the case of an out-of-pocket expenses claim. Indeed, the result of the inability to recover for out-of-pocket expenses may well be that less financially secure tenderers go insolvent (which is of course exactly what has happened to Balraz). This is an especially important consideration given that tenders are a key mechanism by which government tries to give work to previously-disadvantaged, financially insecure firms. As to Moseneke DCJs argument that the tenderer should simply negotiate in contract for a right of recovery, the minority says that this is simply unrealistic in practice: the State has much more bargaining power than the tenderer and in all likelihood will stipulate standard terms in advance. The dissenting judgment closes by remarking on s 8(1)(c)(ii)(bb) of PAJAone of its noveltieswhich allows courts to not only set aside an invalid administrative action but also to pay compensation to affected parties. They suggest that this section intends to nudge our law towards the development of administrative law principles governing the payment of compensation to vindicate the constitutional right to administrative justice. In the result, Langa CJ and ORegan J say they would award damages to the applicant. Finally, Sachs J concurs in the majority judgment but feels the need to express one qualification, namely that the law of delict must get the fuck off administrative laws turf:
I do not feel that PAJA is irrelevant to the present matter. Even though its terms are not applicable because it was adopted after the events in question had taken place, it does serve to articulate public policy as it was emerging at the relevant time. Both the interim Constitution and the final Constitution envisage a right to just administrative action. The implication is that a constitutionalised form of judicial review is intended to cover the field, both in substantive and remedial terms. To my mind it would not only be jurisprudentially inelegant and functionally duplicatory to permit remedies under constitutionalised administrative law, and remedies under the common law, to function side by side. It would be constitutionally impermissible. The provision in PAJA [s 8(1)(c)(ii)(bb), as above] to the effect that in special circumstances a court reviewing

78

administrative action could award compensation, did not invent the public law remedy it articulates. On the contrary, it gave precise expression to a remedy already implicit in the interim Constitution and, later, in the final Constitution. The existence of this constitutionally-based public law remedy renders it unnecessary and inappropriate to hybridise and stretch the common-law delict of injury beyond its traditional limits in this area.

Sachs J goes on to say that the applicant should have pleaded the matter in purely administrative law terms, asking for a new public law remedy to be developed to deal equitably with the paying of compensation in these types of cases. Sachs Js reasoning in this passage, suggesting that the law should not provide remedies in two separate areas of law for the same conduct, is heavily criticized by Richard Stacey in the prescribed article.

79

INTRODUCTION TO JUDICIAL REVIEW

1 Breitenbach: apartheid era review

In this topic we introduce the concept of judicial review by considering its theoretical justification and constitutional basis. We begin with the work of Breitenbach, who in a 1992 journal article, Justifications for judicial review, describes the status and nature of judicial review as it existed at that stage, just before the end of apartheid. He notes at the outset that the doctrine of ultra vires was the justification for judicial review in Englandand therefore, by inheritance, in South Africa. The basic idea is that those exercising a public power must do so only within the bounds set by the law; and thus any act which is ultra vires must be declared invalid. Traditionally this meant simply that those exercising public power must comply with the parliamentary legislation conferring such power. But there was a fiery dispute in England in the 1980s as to whether courts ought to take a wider and more progressive view of the ultra vires doctrine. That is, there was a debate between the following two approaches to the ultra vires doctrine: the narrow view of ultra vires, which holds that judges may only invalidate administrative actions if such actions were undertaken contrary to the empowering legislation; and the wide view of ultra vires, which holds that judges may invalidate administrative actions if such actions were undertaken contrary to the empowering legislation or contrary to fundamental rules of the common law, such as the prohibition on vagueness and the requirements of natural justice. Clearly the wide view gives judges greater scope for scrutinizing and invalidating administrative action. On the narrow view judges can only invalidate an administrative action if it is inconsistent with the language of the empowering statute. On the wide view judges can draw on a range of additional grounds, ones which are available from the common law, regardless of whether Parliament has chosen to include them in the legislation. Lord Hoffmanns judgment in A v Secretary of State for the Home Department [2005] UKHL 71 exhibits the narrow view of ultra vires. Whereas the majority of the House of Lords invalidated the regulation permitting indefinite detention without trial on grounds of its being irrationalwhich obviously goes beyond the simple application of the requirements of the empowering legislationLord Hoffmann (in his concurring but differently reasoned judgment) invalidated the regulation because it was contrary to what the empowering legislation permitted. Specifically, he said that the legislation only allowed habeas corpus to be suspended if there was a threat to the life of the nation; but terrorism doesnt threaten the very existence of the nation itself, only the life of some of its citizens. On this basis he held that regulation did something that the legislation didnt permit. Lord Hoffmanns judgment was heavily criticised because he distanced himself from the majoritys wide view of ultra vires and applied instead the conservative, narrow view. His judgment provides a notable reminder, however, that intense scrutiny under the narrow view can lead to the same outcome as an application of the wide view. Breitenbach is very critical of the narrow view of ultra vires for one specific and simple reason: South Africas parliament was not (when he wrote the article) democratically elected. As a result, the whole justification (most influentially expressed by Dicey) for caring about what parliament has permitted the administration to donamely that parliament is democratically elected and as such reflects the will of the peoplebreaks down. 80

Breitenbach goes on to consider the case law in light of the two competing conceptions of ultra vires. The first case he discusses is Staatspresident v UDF 1988 (4) SA 830 (A). The case took place when a state of emergency had been declared, which meant challenges to exercises of public power were possible only on very narrow grounds. The UDF was challenging a regulation made in terms of the Public Safety Act which allowed members of the Force to search and seize any media publications which contained subversive statements. The UDF argued that this was too vague to be enforcedit was unclear what would classify as a subversive publication. They pursued their action in the Natal High Court (which activists would often choose in the hope that they would get Didcott J), where they succeeded. On appeal to the Appellate Division, however, the UDF faced the A Team, the 5 flaccid arseholes who Rabie CJ was in the habit of choosing to hear sensitive matters of State securityto the exclusion of any judge who wasnt an executive-minded piece of shit. The leading judgment was delivered by Hefer JA. He acknowledged that subversive was a very vague term which made it almost impossible for people to arrange their conduct so as to comply with the regulations. The implication was clearly that if vagueness was included in the ultra vires doctrine then the regulations would be ultra vires. But Hefer JA reasoned essentially that, because nothing in the Public Safety Act precludes the issuing of vague regulations, this particular regulation was not ultra vires no matter how vague it may be. This, of course, was an application of a very narrow view of the ultra vires doctrine. This view remained fixed in our law, to pernicious effect, until the end of apartheid. As we have said throughout this course, judgments like this must be recognized as the result of judicial policy; they are manifestations of the judges convictions about the proper role of the State and the proper role of a judge. So the UDF case shows that the Appellate Division decisively rejected the wide view of ultra vires insofar as vagueness was concerned. But, almost simultaneously with this decision, two other cases were decided by the Appellate Division: Attorney-General, Eastern Cape v Blom 1988 (4) SA 645 (A) and Administrator, Transvaal v Traub 1989 (4) SA 731 (A). Both judgments were delivered by Corbett JA, who was not a member of the A Team and was South Africas Chief Justice during the transition. The latter case we have already encountered briefly. It established inter alia the doctrine of legitimate expectations in our law, thus expanding the scope of judicial review. In the former case, Blom, an emergency detainee, had been refused bail without a hearing. This action was challenged, requiring Corbett JA to decide whether audi alteram partem was a requirement of the ultra vires doctrine. It had been decided in 1967 that it was not, but Corbett JA abandoned this narrow view of ultra vires and reestablished the prior law, in terms of which anyone whose property, liberty or existing rights were affected by administrative action had a right to be heard. Since this requirement had not been expressly excluded by the empowering legislation, Corbett JA read it in as a common-law principle and invalidated the action accordingly. So we see that, while the Appellate Division took a narrow view insofar as vagueness was concerned, it simultaneously took a wide view insofar as audi alteram partem was concerned. 2 Hoexter: post-apartheid review

What is the constitutional basis for judicial review post-1994? On a superficial level, it is s 33 and the various judicial powers stated in Chapter 8. The scope and depth of scrutiny that is permitted is more unclear. Hoexters article, The future of judicial review in South African administrative law, discusses the status of judicial review in post-apartheid, post-PAJA South Africa. This article has been heavily cited by our courts, especially the Constitutional Court. Hoexter confronts two major issues: firstly, the incompleteness of South 81

African administrative law (in that there is not enough provision made for non-judicial review); and, secondly, judges haphazard and unprincipled understanding of deference. So Hoexters first major point is that the continuing primacy of judicial review as the dominant mechanism for holding public power to account is a major problem for the country; non-judicial forms of review ought to be much better developed. As we have already said many times, our administrative law had to fill the void created by the absence of a bill of rights. (In fact, Lord Hoffmann made a similar point about English administrative law, which was used to enforce civil rights prior to the enactment of the Human Rights Act.) And people looked to the judiciary as the only real way of holding the other, morally despicable branches of government to account. For these reasons, judicial review was revered and virtually any interference with the executive was applauded. Moreover, any form of review which would enjoin the other branches of government or civil society was understandably thought to be ineffective. But, says Hoexter:
Now, of course, we have entered an era of constitutional democracy, an era in which it is no longer possible automatically to equate judicial deference with acquiescence in political repression; an era, indeed, in which executive-mindedness might sometimes by a desirable judicial stance. The problem is that the proreview attitude may be hard to shake off, particularly when it is coupled with the South African lawyers traditional disregard of the limits and limitations of review.

So, basically, government is now legitimate and therefore we ought not to think that judicial review is the best or only way to hold exercises of public power to account. But, unfortunately, our attitudes to judicial review, shaped by many years of apartheid, are hard to change. Hoexter then discusses the limits and limitations of review at length. She identifies the following problems with judicial review: 1. Marginal and peripheral: it ignores altogether matters that are not justiciable in a court of law; there is no assurance that administrators actually change their behaviour in response to what is held by judges in the course of judicial reviewor, worse, they may change their behaviour not to make it better but only insofar as to make it harder to challenge in court; and it often doesnt even help a successful litigant, because the administrative decision will be invalidated but can be taken againthis time with the proper procedures, etc.by the same administrator. 2. Backward-looking: it only points out defects in past decisions rather than deciding what good administrative practices must be in the future. 3. Sporadic and random: it only occurs in relation to administrative actions which somebody has the time and money to take to court. 4. Unsuitability: it is unsuited to many kinds of administrative action, especially polycentric decision-making.

82

5. Access to justice: it is inaccessible to most people because it is slow, expensive and mysterious to a layperson. 6. Undemocratic: it doesnt foster participation; and it permits the judiciary to usurp the powers of the administration, powers which were entrusted to the administration by the democratically-elected legislature. Hoexter thinks the last problem is the most important. It is particularly serious in South Africa, because the government is now responsible for alleviating gross socio-economic disparities and as such must not be interfered with by judges (who are not democratically elected and come overwhelmingly from the wealthy elites) at the instancing of litigants (who, similarly, are those citizens with the most money and power). On the other hand, one could say that this is not an argument against judicial review as such but rather an argument against an overly-invasive way of adjudicating socio-economic rights cases (with an appropriate degree of deference; see below). Also, if government does not have good policies or is not meeting its constitutional imperatives then the assumption that judicial review will hamper efficient service-delivery is misguided: the reality is that government does not always have good policies, so judicial review can still be beneficial in alleviating socio-economic disparities and advancing social justice. Anyway, in light of all these problems, Hoexter flipped the fuck out when she realized that PAJA had made virtually no provision for non-judicial forms of review. She laments the fact that the SALRC committees various recommendations in that regard were simply ignored by Parliament. Ultimately, therefore, South African administrative law remains disgustingly incomplete even after the s 33 right to administrative justice was legislatively enabled. Turning now to deference, Hoexter emphasizes some points weve raised previously. During apartheid, judicial activism and an extensive role for the judiciary within the separation of powers were seen as uncontroversially good things. Now that we have a democratic government with opportunities for participation, a Bill of Rights, and so on, maybe its time for the judiciary to restrain itself and become more passive. Since the executive needs strong control and a degree of flexibility in order to address socioeconomic disparities, the judiciary must be deferential to it. Moreover, s 33 of the Constitution is very broadeveryone has the right to just administrative actionand so it needs to be intelligently restricted by a good theory of deference or the judiciarys powers of review would be almost unbridled. But this policy of deference must be well thought out: deference must occur on a principled and well-justified basis, considering constitutional imperatives and our countrys circumstances. Hoexter, in the remainder of her article, endeavors to provide such a theory of deference in terms of three basic themes. Firstly, she discusses variability, that is, the idea that the grounds of review need not be applied in an all-or-nothing fashion, and that the intensity of judicial scrutiny may vary according to the context. By recognising that different degrees of scrutiny are appropriate in different circumstances, courts are able to be more subtle about their approach to deference: by recognising that one doesnt have to apply either all the grounds of review or none of them, judges are able to chart a moderate course which doesnt require them to interfere unduly or otherwise not at all. Secondly, and relatedly, Hoexter refers to the definition of administrative action. She discusses the various cases (which we know) which said that although the conduct in question was not administrative action it was reviewable under the principle of legality. She approves of this strategy 83

because it allows for exactly the kind of variability she has just spoken aboutadministrative actions are reviewed on more grounds than other exercises of public power, but the principle of legality ensures that the latter are still subject to review. She suggests that this technique should be used in other contexts to avoid the problem that courts are reluctant to review certain conduct at all because they think it would be too interfering to apply all the usual grounds of review to it. Thirdly, she says that the notion of reasonableness must be properly formulated so as to preserve due deference. Of course reasonableness review requires judges to investigate the merits of the administration decisionand this is a good thingbut, she says, they must at the same time avoid invalidating a decision simply because they do not agree with it. She quotes with approval the dictum of Froneman DJP:
As long as the judge determining [the reasonableness of administrative action] is aware that he or she enters the merits not in order to substitute his or her own opinion on the correctness thereof, but to determine whether the outcome is rationally justifiable, the process will be in order.

This recalls the distinction between appeal and review: the point is basically that judges must not adjudicate the matter as if its an appeal. The possibility of reasonableness review, which involves scrutiny of the merits traditionally a matter for appeal, not reviewmakes the distinction harder to draw. Hoexter then considers what contribution PAJA has made under the three themes she has identified. Her overall assessment is unequivocal: The final drafters of PAJA, she says, have given effect to a misguided idea of deference. They have got it wrong. She says that PAJA is decent as far as variability is concerned, in that in ss 3, 4 and 5 it gives different procedural fairness requirements for different contexts. In fact, s 3(2)(a) specifically says that *a+ fair administrative procedure depends on the circumstances of each case. Section 4 gives a choice between notice and comment procedures, public enquiries and other possible procedures. And s 5 requires administrators to furnish adequate reasons unless it is justifiable in the circumstances to not do so. As to the definition of administrative action, it is no surprise that Hoexters criticism is scathing. She suggests that the definition probably renders the Act unconstitutional. But, besides this, it does little to advance the debate about deference. The reason is that it forces lawyers and judges to work out the content of the definition of administrative action instead of working out the factors relevant to judicial intervention and nonintervention. As far as the definition of reasonableness is concerned, PAJA also fails miserably. It resurrects the circular Wednesdbury test (the conduct must be so unreasonable that no reasonable person could have so exercised the power or performed the function) and thus fails to give the term any content at all. Hoexters wry conclusion is that the debate about deference will be cancelled owing to lack of interest. Administrative lawyers will instead be debating the meaning of the various technical terms used in PAJAs definition of administrative action. 3 Dyzenhaus, The past and future of the rule of law in South Africa

Dyzenhaus wrote his article well after Hoexter, in 2007. His article was prompted by the litigation in the New Clicks case, and the issues that raised for the rule of law. 84

He begins by noting that even the apartheid government complied with the rule of law in a certain, narrow sense, namely that all exercises of public power must be authorized by legislation. He says we should not underestimate the value of the rule of law, even in this formalistic senseit was a significant check on the apartheid executive and administration. But clearly this is not enough, and fortunately weve now embraced a more substantive conception of the rule of law. Dyzenhaus then charts the course of the New Clicks litigation and assesses how South Africa is realizing the (substantive) rule of law. He draws a distinction between old-order judges (those appointed before 1994) and new-order judges (those appointed after). He refers particularly to Harms JA, who is an old-order judge who subscribed pre-1994 to a narrow, submissive account of the rule of law but who has since had an epiphany andas exhibited in his judgment in New Clicksadopts a very substantive and activist account of the rule of law. He is also very critical of the High Court (that is, Hlophe JP and Yekiso J), which delayed in giving judgment in what was clearly an urgent matter and inexplicably refused leave to appeal. When Hlophe JP was criticised for this he blamed it on a racist conspiracy. Generally his actions were very dangerous for the rule of law. Dyzenhaus points out various other threats to the rule of law, for example the consistent disregard by the Eastern Cape provincial government of the myriad judgments against it relating to the disbursement of social welfare grants. Corders familiar point about the rule of law is that it is impossible to separate the power of judicial review from the basic principles of the rule of law. Judges have a certain role to playindeed they are our last line of defencein upholding the rule of law, both in scrutinizing legislation and in scrutinizing administrative action. For that reason, understanding what the rule of law requires is essential to understanding the nature of judicial review.

85

NON-JUDICIAL FORMS OF REVIEW

1 Introduction

We have already noted that non-judicial forms of review are essential to creating a complete system of administrative law which can fully realize administrative justice. There is not much to say about this topic, but that is not a reflection on its lack of importance. Rather, it is simply a reflection on the unfortunate primacy of judicial review in South African law. 2 Hoexter, the forms of non-judicial review29

Our most notable non-judicial controls on the administration are as follows. 1. Administrative appeals If one is dissatisfied with an administrative decision it is sometimes possible to have an internal appeal rather than immediately taking the matter on judicial review. Both are aimed at the reconsideration of an administrative decision by a higher authority, of course, but they are different in important ways: Administrative appeal Internal check, by someone within the administration Scrutinizes the merits of the decision The adjudicator will step into the shoes of the original administrator and decide the matter anew Judicial review External check, by members of the judiciary Scrutinizes the way in which the decision was reached The adjudicator will only test the legality of the original administrators decision

Administrative appeal has three main advantages over judicial review. Firstly, the administrative authorities have the necessary specialist expertise and are therefore often the best judges of decisions made by other administrators. Secondly, it keeps the matter in the hands of the elected members of government and their employees, thus avoiding the undemocratic judiciary. Thirdly, administrative appeals are usually (but not always) cheaper and speedier than judicial review. There are various types of administrative appeal bodies which vary in their independence from the executive, their functions and their methods of operation. Examples are internal appeals from the decision of a tribunal or official to a superior departmental official or to the Minister concerned; appeals to an administrative tribunal specifically created to hear administrative appeals (such as licensing boards and town planning appeal boards); appeals to special courts presided over by judges (tax courts and the Competition Appeal Court); and even appeals to ordinary courts (for example, the Health Professions Act allows for an appeal to the appropriate High Court).The absence of a coherent system of administrative appeals will continue to undermine the ability of these controls. 2. Legislative oversight This type of control is not as strong as it theoretically ought to be, because the executive tends to dominate the legislature. Section 55 of the Constitution requires that the National Assembly provide for mechanisms to
29

Corder ran through these very quickly during a lecture. These notes are entirely from Hoexter.

86

ensure that all executive organs of state in the national sphere of government are accountable to it. This manifests in the following ways (although with virtually all of them the shortcomings are obvious): Parliament enacts the legislation which then empowers and constrains the executive. It is a traditional principle of good governance that Ministers must consider themselves to be responsible to Parliament. Parliamentary can require that delegated legislation be laid before it for scrutiny. Parliamentary committees, including the standing committees appointed for every Cabinet portfolio, can keep the executive accountable. Parliament is involved in the preparation and approval of the budget, which obviously has significant implications for the executive. Reports are submitted to Parliament by, for example, the Public Protector and Auditor-General. 3. Public participation The Constitution makes express provision not only for representative and direct democracy but also for participatory democracy, particularly with regard to original law making. There are many different ways that government can encourage participation by its citizens: parliamentary inquiries, surveys, roundtable conferences, workshops and local meetings, consensus conferences, impact assessment procedures, and so on. Section 4 of PAJA is the most innovate and important addition to the law on procedural fairness. It however only applies where an administrative action materially and adversely affects the rights of the public. The section refers to two types of public participation: notice and comment procedures and public inquiries. Notice and comment procedures are involved in rulemaking. It is typically required that the proposed legislation be published for comment, whereafter the comments are assessed and, where persuasive, incorporated. The legislation is then promulgated with (ideally) a concise description of the comments and the administrators reasons for adopting the legislation in its final form. This process is designed to promote fairness, capture expertise from the public in particular areas, retain flexibility and promote a culture of justification. A public inquiry, on the other hand, tends to be more useful in the making of particular decisions (not the drafting of subordinate legislation). But the choice on which procedure to use is left to the members of the executive, and is not reviewable under PAJA as it has been expressly excluded from the definition of administrative action. However like other exercises of public power, it is subject to the principle of legality. Negotiated rulemaking is often considered preferable to the notice and comment procedure. Here it is the interested parties who participate rather than the broader public, and that participation consists of direct involvement in the actual drafting of the rules. 4. Ombudsmen Ombudsmens principle function is to protect citizens from government maladministration. The office of ombudsman is an official one, a constitutional mechanism to hold government to account. Its incumbent is responsible to the legislature, and must be independent. His jurisdiction is to receive and investigate complaints by the public about government maladministration. He is not able to take remedial action himself, but rather issue reports and make recommendations usually to the legislature. So he depends on wider institutional support.

87

The Public Protector has a very wide mandate to investigate allegations of maladministration in connection with the affairs of government at any level. He is meant to be independent but is completely reliant on the executive for fundingwhich has tended to be minimal. The Auditor-Generals mandate is more limited: his job is to ensure the proper use and management of public money, requiring him to audit and report on the accounts, financial statements and financial management of all state departments and administrations at all three levels of government. 5. Access to information Offering citizens access to state-held information is one of the most effective ways of upholding the constitutional values of transparency, openness, participation and accountability. It is a fundamental part of a properly functioning participatory democracy and increases public confidence and enhances legitimacy. As we know, South Africa has a constitutional right of access to information (s 32) which is enabled in PAIA.

88

THE LABOUR/ADMINISTRATIVE LAW DIVIDE

1 Introduction

In this topic we discuss a messy area of law in which our courts have embarrassed themselves. The difficulty with the labour/administrative law divide reflects a more general feature of administrative law: basically, it has application in so many different areas of the law. Administrative law is engaged whenever public power is exercised, and of course such exercise can occur in a contractual matter, a delictual matter, the enforcement of a statute, or just about anything else. How administrative law interacts with these other areas of law is, as weve already seen in the previous topics, a complete fucking mystery. In practical terms, we want to know how to determine whether labour law or administrative law (or both) applies to a particular dispute. Remember that administrative law will only be at issue if the employer is a public body, so all the cases below involve employers who could be construed as such. 2 The development of our law

Before 1993, labour matters were regulated by the Labour Relations Act of 1956. It didnt apply to public sector employees or black Africans. It also gave no guarantee of fair labour practices. At that stage, therefore, labour rights were virtually non-existent. Only much later were some more comprehensive (but sector-specific acts) introduced, like the Public Service Labour Relations Act. Anyway, administrative law was called upon to fill the void created by the virtual absence of any substantive labour rights. This occurred in the case of Administrator, Transvaal v Zenzile 1991 (1) SA 21 (A), which weve already encountered. It involved the summary dismissalon 24 hours notice and without a hearingof three striking public healthcare workers. The legal position was governed by the contract between the two parties, but the Public Service Act 111 of 1984 was relevant too. If the pre-existing common law had been applied, then the Court wouldve treated this as a matter of pure contract law, with no administrative law rights engaged. And, since the employment contract between the parties did permit such summary dismissal, Zenzile wouldve been fucked. But the Court held that since the Act had application in matters between the two parties, this was an administrative law matter. And therefore audi alteram partem was required, in accordance the rule that this is so whenever property, liberty or rights are affected. This was extremely radical for the time, because traditionally contract law was seen as a pure, private law matter, completely distinct from administrative law. Zenzile was entrenched, though, by Adminstrator, Natal v Sibiya 1992 (4) SA 532 (A), which followed shortly after. The upshot is that at that stage our highest court had clearly said there was no hermetic seal between private law and administrative law; it had opened the door for administrative law standards to be used in labour law. In many ways (as we will see) this was the high-water mark for this public law approach. Then the 1993 Constitution was enacted, with fundamental labour rights included in s 27. The 1996 Constitution includes them in s 23, which reads:
(1) Everyone has the right to fair labour practices. (2) Every worker has the right(a) to form and join a trade union; (b) to participate in the activities and programmes of a trade union; and

89

(c)

to strike.

These rights are enabled by the Labour Relations Act 66 of 1995 (LRA). This is such a comprehensive and detailed piece of legislationespecially relative to the slim PAJAthat one immediately wonders whether it is prudent to allow administrative law to interfere with it. 3 The review of an arbitrators decision: Sidumo

The first issue we deal with is the review of a CCMA arbitrators decision on a labour dispute. Such review must occur in terms of s 145 of the LRA. This requires the matter to go to the Labour Courts first. More importantly, the reviewing court can only set aside the arbitrators order on four, fairly limited grounds (misconduct, gross irregularity, etc.). The issue, then, is whether a dissatisfied party can apply additional grounds of review from administrative law. This is exactly what was at issue in Sidumo v Rustenburg Platinum Mines Ltd [2007] ZACC 22. The facts were as follows. RPM (a private company) suspected that people were stealing platinum from their mines, so they required their security guards to search people leaving the premises. Sidumo was one of these security guards. He was caught on camera doing a really shitty job of searching those leaving the mines. RPM therefore summoned him to a disciplinary hearing, after which he was fired on grounds of misconduct. This decision was confirmed in an internal appeal. Sidumo took the matter to the CCMA, which agreed with the finding of misconduct but regarded dismissal as too harsh. It was therefore ordered that Sidumo be reinstated subject to a three month probation period. RPM was unhappy, so it took the matter to the Labour Court (which is empowered under the LRA to hear appeals from the CCMA arbitration) and, having lost there, to the Labour Appeal Court, where it lost again. Undeterred, RPM went to the SCA. The first question was whether PAJA could in principle apply to the arbitrator, i.e. whether his decisions constituted administrative action. Cameron JAs answer was Yes. But the trickier, second question was whether PAJA was ousted by s 145 of the LRA. This section only lists four grounds upon which an arbitration award can be reviewedand irrationality, which Sidumo was arguing, was not one of them. So the whole dispute turned on the interaction between PAJA and the LRA. Cameron JAs answer was that PAJA extends the grounds of review available to any party seeking to review a CCMA arbitrators decision, thus effectively trumping the s 145 of the LRA to the extent of any inconsistency. This was so, he said, even though the LRA is the more specialized legislation. His justification is basically that PAJA enables the constitutional right to administrative justice, purports to codify all the available grounds of review, and gives a broader set of grounds to review the arbitrators decision than the LRA. On the factsand particularly the fact that Sidumos misconduct went to the heart of the employment relationship and violated the trust the employer placed in himSidumo was found against. Sidumo appealed to the Constitutional Court. Navsa AJ, writing for the majority, agreed that the arbitrators decision is administrative action. But he disagreed with Cameron JA on the question of the two Acts interaction. He applied the well-established rule that the more specific statute trumps the more general. Since the LRA was the specialized attempt to regulate administrative action in the labour sphere, it had to trump PAJA.

90

But, although PAJA itself did not apply, s 33 of the Constitution did: Navsa JA held that the LRA had to be interpreted in accordance with the constitutional right to administrative justice, so s 145 must be understood as infused with the constitutional standard of reasonableness. This standard then meant that the arbitrators decision had to be reasonable or otherwise be invalidated under s 145. On the facts, Navsa JA held that the arbitrators decision was a reasonable one; whether or not the Court agreed with it, the point was that it was the sort of decision which a reasonable person could make. The RPM also raised the argument that the arbitrator should have deferred to the employers decision as to what was a fair punishment for Sidumo. The majority unsurprisingly dismissed this suggestion, saying basically that the LRA and CCMA are not aimed at deferring to employers wishes. Ngcobo J concurred in Navsa JAs order, but reasoned differently. He argued that the arbitrator was performing a judicial function and was therefore not administrative action at all. This meant that neither PAJA nor s 33 could have any role to play. Nevertheless, the arbitrator did exercise a public power and was hence constrained by the principle of legality (and other provisions in the Bill of Rights). Section 145 of the LRA had to be interpreted in these terms, thus requiring an expansive reading of the grounds of review listed there. In short, Ngcobo Js argument is that the gross irregularity ground contained in s 145 must be understood as including a breach by the arbitrator of the doctrine of legality. On the facts, however, he held that even this had not been established. ORegan J (also in a separate, concurring judgment) disagrees strongly with Ngcobo Js finding that the arbitrators conduct was not administrative action. She says that, although it is true that courts ought not to review the decisions of other, well-placed courts, this is a stupid little administrative body and thus judicial review of its proceedings is at the very essence of administrative law. Hoexter raises three points for us to consider on the basis of Sidumo. First, she says that we are again faced with the problem of deciding what is and what is not administrative actioneven when we are considering the broad s 33 rather than the gross PAJA definition. Second, she argues that a new pathway to review has opened up with Sidumo. While the pre-democratic era had two routes (ordinary, common-law review of the decisions of public bodies and the special statutory review specifically conferred by legislation), s 33 created a third route when the Constitution came into force. A fourth was added when the Constitutional Court in Fedsure identified the principle of legality as residual review mechanism for those instances that did not qualify as administrative action. The fifth was via PAJA (with all of the problems of its very limited definition). And now, finally, there is a sixth: special statutory review under the LRAas suffused with s 33s valuesfor labour-related action that also qualifies as administrative action. The importance of this sixth route is that it could be developed by analogy to apply to other statutory regimes too, for example the Liquor Act and the Insolvency Act. Thirdly, Hoexter says that Sidumo reflects a holistic constitutional vision, since the Court envisages overlapping and interconnected constitutional rights. Sections 23 and 33, on this view, cannot be compartmentalized; the one must infuse the other. We will see that the Court has moved drastically away from this vision subsequently.

91

A fourth, more obvious point to note is that PAJA is again getting sidelined. This trend will become even clearer as we discuss more cases. It shows how difficult the statute is to work with. So Sidumo is relatively clear: you use s 145 of the LRA, infused with s 33, and thats that. But shit goes rapidly downhill from here. 4 The application of administrative law standards to public employers (f) Introduction Now we move away from the narrow issue of the review of a CCMA arbitrators decision. We want to see whether public employers are bound by administrative law standards. But more important than the substance of the law in this area is the procedure, especially the issue of jurisdiction. (g) The old view: Fredericks The nature of this difficulty is made clear in Fredericks v MEC for Education and Training, Eastern Cape 2002 (2) SA 693 (CC). This involved teachers who had applied for voluntary retrenchments under a collective agreement. Previous teachers applications had been accepted, but the applicants had not. They therefore challenged the decision on the basis of ss 9 and 33 of the Constitution. PAJA was not operative when the dispute arose, so one mightve thought that things would be straightforward. But the High Court held that it had no jurisdiction to hear the matter. This was because s 24 of the LRA required the CCMA and the Labour Courts to hear disputes relating to the interpretation and application of collective agreements. On appeal to the Constitutional Court, this particular argument was rejected on the grounds that this dispute, although peripherally involving a collective agreement, engaged constitutional rights. But the MEC had other tricks up his sleeve. The first was to refer to s 169 of the Constitution, which provides that:
A High Court may decide any constitutional matter except a matter that is assigned by an Act of Parliament to another court of a status similar to a High Court.

The argument was that Parliament had done exactly that with the LRA, by assigning labour disputes to the CCMA. But the Constitutional Court per ORegan J rejected this argument, saying that the CCMA is not a court of a status similar to a High Court. The obvious counter-argument to the Courts reasoning is that the LRA provides that reviews from the CCMA must go to the Labour Courts, which are of a similar status to High Courts. ORegan J rejects this because the review permissible under the Act is too limited to allow the Labour Courts to hear all matters de novo, so it isnt really sufficient to fulfill the exception created by s 169. As we will see, the Constitutional Court has subtly overturned what ORegan J says here, and apparently now endorses the counter-argument. Anyway, the MECs second argument was based on s 157 of the LRA, which reads:
(1) *T+he Labour Court has exclusive jurisdiction in respect of all matters that elsewhere in terms of this Act or in terms of any other law are to be determined by the Labour Court. (2) The Labour Court has concurrent jurisdiction with the High Court in respect of any alleged or threatened violation of any fundamental right entrenched in [the Bill of Rights] and arising from [a labour dispute].

92

A literal reading of subsection (2), especially the word concurrent would strongly suggest that the High Court did have jurisdiction in this matter. The problem is that a purposive reading yields a different answer: subsection (2) had been inserted in a 1998 amendment in order to fill the lacuna created by the Labour Courts being unable to hear any labour dispute which involved the violation of a constitutional right, so the subsection was purposed at giving the Labour Courts jurisdiction in rights cases rather than giving the High Courts jurisdiction to hear labour disputes. But ORegan J used a literal interpretation and on that basis held that s 157(2) did not oust the Courts jurisdiction. Again, this is a view which the Constitutional Court now rejects. Ultimately ORegan J holds the High Court had had jurisdiction to hear the matter, so the dispute was referred back to it. The law at this stage seemed reasonably clear: litigants were free to bring labour disputes in the High Court if the right to administrative justice was involved. This has been radically altered, though, in later cases, as we will see below. (h) Principle of legality stream: Masetlha The next case is one weve encountered previously: Masetlha v President of the Republic of South Africa 2008 (1) SA 566 (CC). We dealt with this under the principle of legality. That was because, despite seeming to be a quintessential labour matter, intelligence service workers (of which Mr Masetlha was the head), along with employees of the army and secret service, are specifically excluded from the LRA. And Moseneke DCJ held that President Mbekis firing of Mr Masetlha was executive action rather than administration action, so PAJA was not relevant. This meant that the principle of legality was the applicable standard, so the Court was able to take a very different approach to that taken in the other labour cases. Anyway, Moseneke DCJ held that the Presidents actions complied with the principle of legality, so the firing was valid. But the firing did constitute a breach of Mr Masetlhas fixed-term employment contract, so he was awarded compensation. That the Court happily considered both administrative law issues and contract law issues alongside one another is interesting: with administrative law and labour law, the courts are adamant about picking one or the other. The difference is explained by the fact that there is no constitutional right to contract and no equivalent of the LRA which regulates contract law, so the arguments which have prompted courts to keep administrative and labour law separate (see below) have no application to contract. Moseneke DCJ also said obiter that if the dismissal had amounted to an implementation of legislation then such dismissal wouldve amounted to administrative action, and PAJA wouldve been relevant. This is weird given what had been held in Sidumo, but has been overturned subsequently. (i) The new view: Chirwa and Gcaba Next we come to the vomit-inducing travesty of justice that is Chirwa v Transnet Limited [2007] ZACC 23. Chirwa had been summarily dismissed for poor performance from her position as Transnet HR manager. She referred the matter to the CCMA, who were lazy and didnt pursue the matter properly. She therefore went to the High Court, which applied the common law principles recognised in Zenzile (!), thus avoiding all the jurisdictional issues associated with relying on the legislation. The Court held that Chirwa should have been heard and that her dismissal was for that reason invalid. On appeal to the SCA, a majority held (although each judge for different reasons) that the High Court did have jurisdiction. Conradie JA dissented, saying that the 93

Labour Court had exclusive jurisdiction in such matters. Skweyiya J wrote the majority judgment for the Constitutional Court. Ultimately he agrees with Conradie JA, holding that the Labour Court had exclusive jurisdiction. We discuss his abominable judgment at length in what follows. Chirwa had framed her pleadings almost exclusively in terms of administrative law and PAJA, but Skweyiya J seized on one tiny reference to the LRA and on that basis recast her pleadings entirely, saying that in truth this was a labour matter regulated by the LRA. This is not coolthe litigants claims should be assessed as they are brought. Langa CJ duly has a fat kak on this approach in his minority judgment:
While [the LRA reference] alone might have been construed as a disguised reliance on the LRA, in the broader context of her argument, I do not believe that is a fair or correct characterisation. It should be added that it was not a characterisation urged upon us by the applicants counsel in argument; nor one adopted in any of the three judgments in the Supreme Court of Appeal, nor in the High Court judgments. In my view, it is incorrect. Most of my disagreement with the judgment of Skweyiya J flows from this mischaracterisation.

Anyway, Skweyiya J still needs to deal with Fredericks, which had specifically said that the High Court had concurrent jurisdiction with the Labour Courts under s 157(2) of the LRA. Skweyiya J distinguished Fredericks on the grounds that, whereas Fredericks had relied directly upon a constitutional right (which is the sort of claims which s 157(2) refers to), Chirwa was relying on PAJA. This wouldnt be clinically retarded except for that the Constitutional Court had previously held in Bato Star that litigants must rely on PAJA, the enabling legislation, rather than on the s 33 right itself. In any event, it seems bizarre to say that reliance upon PAJA, the enabling legislation for the s 33 right, does not involve a constitutional right. But Skweyiya J wasnt finished yet. Oh heavens no. He went so far as to say that even if a litigant does rely directly on the s 33 right, that still does not give the High Court jurisdiction to hear the matter unless the litigant has first exhausted all his rights under the LRA. The implication is that one has to go through all the LRA structuresthe CCMA, Labour Courts, and Labour Appeal Courtsbefore one can bring a challenge in the High Court on the basis of s 33, even if from the outset one thinks that the LRA is clearly unconstitutional. The only way one could bring such a challenge in the High Court directly is if one alleges that the LRA is inconsistent specifically with s 23 of the Constitution. So how does Skweyiya J explain himself? The basic reason why he is so adamant that labour matters must stay the fuck out the High Court is because the legislature has carefully constructed a set of forums and procedures to deal with labour matters, and so labour matters must jolly well be heard there and not in the High Court:
The purpose of labour law as embodied in the LRA is to provide a comprehensive system of dispute resolution mechanisms, forums and remedies that are tailored to deal with all aspects of employment. It was envisaged as a one-stop shop for all labour-related disputes. The LRA provides for matters such as discrimination in the workplace as well as procedural fairness; with the view that even if a labour dispute implicates other rights, a litigant will be able to approach the LRA structures to resolve the disputes.

So Skweyiya J is bending principle and sidelining administrative law in order to give proper effect to the comprehensive dispute resolution mechanism which the LRA has set up.

94

Ngcobo J, in a weird supplementary judgment in which the majority also concurred, goes into more detail about s 157(2) of the LRA. In essence, he is adopting a purposive reading of the section (where of course ORegan J in Fredericks had applied the plain meaning):
[T]he primary purpose of section 157(2) was not so much to confer jurisdiction on the High Court to deal with labour and employment relations disputes, but rather to empower the Labour Court to deal with causes of action that are founded on the provisions of the Bill of Rights but which arise from employment and labour relations. Given the manifest purpose of section 157(2) the use of the word concurrent is unfortunate.

Skweyiya J, because he has held that the High Court has no jurisdiction, doesnt bother to interrogate the question of whether Chirwas dismissal was administrative action. This is misguided, because the question of whether PAJA is in principle applicable has implications for the jurisdictional questionto decide which statute predominates, we surely need to know which statutes are relevant. Anyway, Ngcobo J does discuss the issue, but says that it was not administrative action because it was the exercise of a private power conferred by the employment contract. The remarkable implication of this reasoning seems to be that weve been taken right back to the apartheid era position, pre-Zenzile, where public employment contracts are hermetically sealed from administrative law. Ngcobo J justifies his conclusion as follows:
Support for the view that the termination of the employment of a public sector employee does not constitute administrative action under section 33 can be found in the structure of our Constitution. The Constitution draws a clear distinction between administrative action on the one hand and employment and labour relations on the other. It recognises that employment and labour relations and administrative action are two different areas of law.

Interesting. The major difficulties with Chirwa, then, are these: Firstly, Skweyiya J deviously reframes the applicants claim. His reasoning on this point has been exploited by many later courts to throw out matters framed in administrative law terms (like in the High Court in Gcaba, immediately below). Secondly, he doesnt properly distinguish Fredericks and ignores what ORegan J said there about s 169 of the Constitution. Recall that ORegan J had held that s 169 didnt permit the CCMA to oust the jurisdiction of the High Courts. The implication is that the legislatureif it had done with s 157 what the Chirwa Court claimed it had donehad acted unconstitutionally. Thirdly, the Constitutional Court has moved to the diametrically opposed position to the one it espoused in Fredericks as to the interconnectedness and fluidity of constitutional rights. Now the Court says in remarkably blunt terms that the structure of the Constitution requires rights to be kept separate and has barred litigants from bringing administrative law matters to the High Court if they arise from a labour dispute. We now come to the case of Gcaba v Minister for Safety and Security 2010 (1) SA 238 (CC). The case concerned the failure to promote a SAPS employee, who thought this was unfair. He framed his action in the High Court exclusively in terms of PAJA, but it was obviously a labour-related matter. The immediate problem was that Chirwa implied that the Court had no jurisdiction, so the matter was duly thrown out. Gcaba appealed to the Constitutional Court. 95

Van der Westhuizen J, writing for the unanimous Court, immediately makes clear what a fuckup this area of law is when he starts his judgment as follows:
One of the purposes of law is to regulate and guide relations in a society. Yet the legislature, courts, legal representatives and academics often create complexity and confusion rather than clarity and guidance. In the case of fairly new legislation based on a young Constitution this is perhaps understandable. Sometimes a jurisprudence needs to develop along with the insight and wisdom emerging from a debate over some time. The legislature may also have to intervene in appropriate circumstances, for example, when incremental development results in uncertainty or an otherwise unsatisfactory situation.

He basically seems to be pleading with the legislature to amend s 157(2) of the LRA. But in the meantime, he says the present dispute gives the Court an opportunity to clarify the issues. He begins by summarising the Courts confusing previous judgments, noting the different schools of thought in the existing jurisprudence. He must then try and bring order to the madness. He does so by trying to identify a few general principles and policy considerations relevant to the law in this area. On the one hand, he says that the same conduct can engage many different areas of law, and rigid compartmentalization should be avoided. Furthermore, human rights are intrinsically interdependent, indivisible and inseparable, and legislation must not be interpreted to limit the remedies for their enforcement. However:
[T]he Constitution recognises the need for specificity and specialisation in a modern and complex society under the rule of law. The legislature is sometimes specifically mandated to create detailed legislation for a particular area, like just administrative action (PAJA) and labour relations (LRA). Once a set of carefullycrafted rules and structures has been created for the effective and speedy resolution of disputes and protection of rights in a particular area of law, it is preferable to use that particular system. This was emphasised in Chirwa by both Skweyiya J and Ngcobo J. If litigants are at liberty to relegate the finely-tuned dispute resolution structures created by the LRA, a dual system of law could fester in cases of dismissal of employees. Following from the previous points, forum shopping by litigants is not desirable. Once a litigant has chosen a particular cause of action and system of remedies (for example, the structures provided for by the LRA) she or he should not be allowed to abandon that cause as soon as a negative decision or event is encountered. One may especially not want litigants to relegate the LRA dispensation because they do not trust its structures to do justice as much as the High Court could be trusted. After all, the LRA structures were created for the very purpose of dealing with labour matters, as stated in the relevant parts of the two majority judgments in Chirwa, referred to above.

And moreover, the doctrine of precedent is very important:

[A] single source of consistent, authoritative and binding decisions is essential for the development of a stable constitutional jurisprudence and for the effective protection of fundamental rights. This Court must not easily and without coherent and compelling reason deviate from its own previous decisions, or be seen to have done so.

You can probably see where this is going: despite paying lip-service to the vision of the Constitution articulated in Fredericks, Van der Westhuizen J seems to be setting the stage to decide that Chirwa must be upheld.

96

Anyway, before he can do that, Van der Westhuizen J must decide whether this is administrative action, i.e. whether PAJA even needs to be considered in the jurisdictional issue. The following passage proved decisive in this regard:
Generally, employment and labour relationship issues do not amount to administrative action within the meaning of PAJA. This is recognised by the Constitution. Section 23 regulates the employment relationship between employer and employee and guarantees the right to fair labour practices. The ordinary thrust of section 33 is to deal with the relationship between the state as bureaucracy and citizens and guarantees the right to lawful, reasonable and procedurally fair administrative action. Section 33 does not regulate the relationship between the state as employer and its workers. When a grievance is raised by an employee relating to the conduct of the state as employer and it has few or no direct implications or consequences for other citizens, it does not constitute administrative action.

This is a very important finding for the interaction between administrative law and labour law. The basic idea is that a labour-related grievance will not be administrative action unless it has an impact on the public beyond the person directly involved. Kerry says is dubious reasoning given the fact that all that the PAJA definition requires is a decision which affects any person. Indeed, PAJA has a whole section (s 3) dedicated to regulating procedural fairness for administrative action affecting any person. But given what Van der Westhuizen J is setting out to dodrive a wedge between labour and administrative law so that they dont overlapthis makes some sense. Nevertheless, on this basis Van der Westhuizen J finds that the failure to promote Gcaba was not administrative action. He says this is consistent with Ngcobo Js finding in Chirwa (which is probably true, but is hardly something to be proud of). Since this means Gcabas claim, despite the way he framed it, could not actually be an administrative law matter, the next question to be answered is the narrow one of whether the High Court had jurisdiction to hear a labour matter. As to this jurisdictional question, Van der Westhuizen J adopts a purposive reading of s 157 of the LRA, affirming the finding in Chirwa and holding that the Labour Court has exclusive jurisdiction to hear labour matters. Finally, lest it be said that there are contradictions in the Courts jurisprudence:
To the extent that this judgment may be interpreted to differ from Fredericks or Chirwa, it is the most recent authority.

Because the Court didnt actually find that the failure to promote was administrative action, it didnt have to consider whether the High Court wouldve had jurisdiction to hear PAJA matters. Gcaba does clarify things a bit. Kerrys problem with it is that the Courts reasoning is still based almost entirely on policy considerations. As we have seen, this has forced it to ride roughshod over the principles involved. It is also an unstable basis for judicial precedent, because good policy can change. 5 Attempt to make some sense

Imagine you have an angry client who was dismissed from her public employment by the Minister acting in terms of an unconstrained statutory power. What would you advise her? 97

Well, obviously you could just do everything in the CCMA and Labour Courts, in terms of the LRA. That is clearly what the LRA envisaged and what the Constitutional Court is desperate for you to do. If you were reviewing the arbitrators decision in terms of s 145 of the LRA, you could rely on Sidumo to infuse the section with s 33 of the Constitution and thus incorporate some additional grounds upon which to review his decision. What use you would be able to get from the infusion is an open question. You could also try and frame your cause of action purely in terms of contractual rights, and thus avoid altogether the admin/labour law issues. But of course this is only going to be possible if your clients contractual rights will vindicate her sufficiently. So lets assume that you want to bring some robust administrative law standards to bear on your clients employer. What then? If you try and use PAJA in the High Court, the judge is going to point and laugh at you. The most daring strategy, which excites Kerry, is to argue PAJA or the principle of legality in the Labour Court. This obviously avoids the difficulty that the High Courts have no jurisdiction. You would presumably have to convince the Labour Court that constitutional rights are at issue. Another thing you could do is go to the High Court, convince it that this is administrative action, and then see how it deals with the jurisdictional question on that basis. In neither Chirwa nor Gcaba was there administrative action. Such litigation would be very uncertain, not just because the jurisdictional issue is a messand it is quite possible the court would just hold that Skweyiya Js reasoning in Chirwa straightforwardly appliesbut also because youd have to pass the test set in Gcaba, namely that the conduct has direct implications or consequences for other citizens. It is a mystery what this means. Perhaps youd have to show that your client was so senior that her dismissal was bad for efficient governance, or something similar. Alternatively, you could perhaps establish that it was administrative action by showing that the employers power to dismiss was sourced directly in statute. Finally, you could rely directly on Masetlha and base a claim on the principle of legality. But this would be possible only on unusual facts, namely where your client was working for the intelligence service, the SANDF or the secret service. Because such workers are excluded from the LRA, the rationale for keeping administrative law and labour law separate breaks down, and so hopefully no jurisdictional issues would arise.

98

THE DEFINITION OF REVIEWABLE ADMINISTRATIVE ACTION30

1 Introduction

We have encountered this definition in s 1 of PAJA many times already, and we saw its application in a very specific context (namely labour law) in the last topic. We will make some brief remarks about it here, based on the prescribed Hoexter article. 2 Hoexter, Administrative action in the courts

The focus of Hoexters discussion in this article is PAJAs definition of administrative action. It is no secret that she thinks its a stinky piece of poo. Not only is it unnecessarily complicated and probably as unfriendly to users as it is possible to be, but the effect it has had on adjudication by the courts suggest problems of a far more fundamental and systemic nature. Hoexter makes the following (mostly very familiar) points about why the definition is an embarrassing failure: 1. It is much narrower than the term had been under s 33 of the Constitutionit has a long list of exclusions, imposes thresholds like the requirement that rights are affected and that the decision has a direct, external legal effect, and so onand is therefore too restrictive and probably unconstitutional. 2. The courts waste a lot of time deciding whether the conduct in question is administrative action rather than dealing with the more important questions: What is the precise content of lawfulness, reasonableness and procedural fairness? What does administrative justice requires in this case, and why? She says that nearly half of the reported administrative law cases in recent years are explicitly focussed on whether the conduct in question amounts to administrative action. 3. The definition is so complicated, with so many requirements and sub-requirements, that the courts categorization of conduct as administrative action tends not only to be controversial but, far worse, misdirected: courts are able to hide behind a highly technical, legalistic screen of reasoning about the definitions various components and thus avoid articulating views as to the substance of the case. A court may feel that a particular applicant should not be entitled to enjoy, say, the benefits of procedural fairness. Instead of saying so, and saying why, the court can simply make the problem disappear by exploiting the many mysterious elements in the definition. This need not be cynical or even conscious, but the point is the definition encourages courts to use code instead of giving explicit recognition to whatever feature is really driving them. 4. The definition has distorted the boundaries between the various streams of review (the Constitution, PAJA and the common law) as courts have tried to avoid PAJA at all costs and relied on some parallel stream. In particular, courts have tended to rely on the principle of legality without engaging with PAJA. As an example of this, Hoexter cites both the High Court and SCA judgments in New Clicks,31

30 31

Corder covered this quickly in the last lecture. It is very repetitive of what weve already covered. The Constitutional Court had not yet decided the matter, although there is a postscript in which Hoexter discusses the CC judgment.

99

where it was held essentially that the principle of legality would in any case step in if PAJA did not apply, so PAJA could be ignored. Hoexter then goes on to argue for a radically different approach. This section of her article is titled One system of administrative law, which gives an indication of what that approach is. Basically, she says we should have a very wide definition of administrative actionin fact, she says s 33 of the Constitution was idealand then moderate the level of scrutiny involved in each case by embracing the variability32 of the requirements of administrative justice. This would mean litigants arent unjustly denied relief simply because they dont meet the threshold requirement, and it would mean further that *t+he focus of every case would be what administrative justice requires in these circumstances. This would also collapse all the silly distinctions between PAJA and the principle of legality and create a simple, unitary system of administrative law. But, given that the legislature is a bunch of indolent plonkers, how could this ever be achieved? One obvious way would be for the courts to expand the principle of legality so as to create a self-sufficient administrative law. Hoexter doesnt actually support this approach, though, because it would totally circumvent both the constitutional right to administrative justice and the legislatures legitimate role in the matter. She therefore prays that someone will bring a constitutional challenge against PAJA so we can take that bitch the fuck down and redo things properly this time, enabling s 33 of the Constitution in its full majesty. Bringing a tear to her readers eyes, she closes her article thus:
*I+t is time we rejoiced in *s 33s+ breadth instead of concentrating so hard on narrowing it down. South Africa already has one system of administrative law in the form of that sectionif only we would realise it.

32

We spoke about her concept of variability earlier.

100