REFLECTIONS ON THE BIOWATCH DISPUTE – REVIEWING THE FUNDAMENTAL RULES ON COSTS IN THE LIGHT OF THE NEEDS OF CONSTITUTIONAL AND/OR PUBLIC INTEREST LITIGATION

Using as a case study the recent decision on costs in the Biowatch matter, this article critically examines the traditional fundamental rules on costs in the light of the needs of constitutional and a fortiori public interest litigation. The fundamental rules on costs are taken to include the two traditional principles (that costs are a matter of judicial discretion and that to a successful party should be awarded his costs), the requirement that the discretion be exercised judicially, the test for interference in costs orders in a court of appeal, and the characterisation of costs orders as requiring the exercise of only a narrow discretion on appeal. In the light of the decisions in the Biowatch matter it is argued that the current rules do not meet the new needs of constitutional and/or public interest litigation as regards access to justice, equal protection and benefit of the law, proportionality, and the accountability of the judiciary. Suggestions are made for possible reform.


Introduction
This article is concerned with the application of the "fundamental rules" 1 governing the awards of costs to litigation of a constitutional and/or public interest character. 2My interest in this topic arose from the decision on costs in As my use of the sign 'and/or' implies, I am of the view that 'public interest' litigation could be coextensive with constitutional litigation, or not.Of course, at a certain level all constitutional litigation is in the public interest, but not all litigants who raise a constitutional issue are acting chiefly with the public interest in mind, but may rather be seeking to advance their own commercial or private interests -the reliance on the environmental right in the Constitution by the Fuel Retailers' Association in Fuel Retailers Association of Southern Africa v Director-General Environmental Management, Department of Agriculture, Conservation and Environment, Mpumalanga Province 2007 (6) SA 4 (CC) is a case in point.The question as to what constitutes a "public interest matter" or the criteria by which one determines that a party is acting in the "public interest" is a difficult one (as the arguments put forward in the Biowatch matter attest) that merits further elaboration.
For the purposes of this article I can suggest a definition that emerged from a consultation process around the English Access to Justice Act, viz: "The potential of the proceedings to produce real benefits for individuals other than the client (other than benefits to the public at large which normally flow from proceedings of the type in question)".See Lord Justice Brooke 2006 JEL 348.This definition, further, resonates with the Constitutional Court's observation in Van der Merwe v Road Accident Fund 2006 (4) SA 230 (CC), 2006 (6) BCLR 682, where the court noted that while the applicant is the "immediate beneficiary of the outcome of the case; … it is also true that there are similarly situated people who are not before us.The outcome of this litigation has a wide reach and is clearly in the public interest" (par 78).96/166 delivered the majority judgment (in which Molopa J concurred) 5 and Poswa J the minority judgment. 6An appeal against the decision of the majority has since been lodged by Biowatch with the Constitutional Court. 7 main argument is that the fundamental rules governing costs awards are not sufficient to meet new needs which arise in constitutional and/or public interest litigation, as the Biowatch matter illustrates, and that aspects of these rules need to be reframed in the light of the value framework established by the Constitution.While one of the primary issues in the Biowatch matter is ostensibly whether the courts' approach to costs in constitutional litigation should be characterised as a "trend" or a "rule", 8 my view is that the real issue is the appropriateness to constitutional and/or public interest litigation of the complex of existing rules that vest a discretion to award costs in the presiding officer, that articulate the standards by which that discretion must be exercised, and that prescribe the circumstances under which a higher court may interfere in that discretion.
Section 2 below outlines the content and provenance of the fundamental rules that form the focus of the discussion, while Section 3 considers how they have been applied in the jurisprudence of the Constitutional Court to date.Section 4 subsequently looks at how these rules were applied in the Biowatch matter, both in the court a quo and on appeal.In Section 5 I set out my reasons for holding that in the case of constitutional and/or public interest litigation these During the process of peer review, the question was raised if it would not make more 'sense' for this article to be written after the Constitutional Court judgment in the case.This view seems to assume that the role of the academic lawyer is confined to analysis, consolidation and clarification of settled law.While these are certainly important functions, I believe that the academic lawyer should also seek to be at the vanguard of current legal developments and to participate in the debates which these raise.The judgment of the Constitutional Court, whilst the final authority in our legal system, is not necessarily the final or only authority in the world of ideas.8 Biowatch II (Minority) par 92.
T HUMBY PER/PELJ 2009(12)1 97/166 fundamental rules are not "fit for purpose" and set out some suggestions for their possible reformulation within a constitutional value framework.
The issue at the heart of this paper -the question of legal costs in relation to access to justice -is not unique to South Africa.In this article, however, I eschew a comparative analysis of other jurisdictions as I believe it is important to provide a qualitative, in-depth analysis of how and why the rules and principles governing the judicial discretion to award costs were applied with such different outcomes in the decisions of the Biowatch matter.My aim, here, is to articulate the assumptions and values underlying the exercise of the discretion, and to question whether or not these are in line with constitutional values. 9

Fundamental Rules Governing Costs
For the purposes of this article, the complex of rules governing costs comprises four dimensions: (1) the traditional principles applicable to costs orders (the first of which is that an award of costs is in the discretion of the court of first instance) and their relationship; (2) the requirement that the court's discretion be exercised judicially; (3) the test for interference in the discretion of the court a quo; and (4), closely related to this, the characterisation of the nature of the judicial discretion to award costs in both the court a quo and court of appeal and the underlying rationale thereof.A closer examination of the rules and principles within each area is required.98/166

Two principles and their relationship
The two principles which have governed costs orders in South African law since the earliest time 10 are, firstly, that the court of first instance has a judicial discretion to award costs and secondly, that costs follow the event in that the successful party is usually awarded costs.Cilliers refers to these principles as the "basic rule" and "general rule" respectively, 11 but following the nomenclature used in the Biowatch cases I will refer to them as the first and second principle.
As Cilliers himself concedes, these principles exist in a "curious" relationship.
Lying at the heart of this curious relationship is the paradox of holding that a judicial officer of first instance has the discretion to make a costs award whilst at the same time apparently prescribing how that discretion should be exercised.The tension has been managed in two ways.Firstly, there is an extensive body of precedent in support of the rule that the second principle yields to the first. 12As a result, the nature of the judicial discretion in the court a quo has been described as "very wide" 13 or "overriding". 14However, it is not surprising to find judicial officers then maintaining that the discretion is wide and unfettered 15 although there is also a large body of case law maintaining that the discretion is not unfettered.

99/166
The second approach towards reconciling the two principles was put forward in Levben Products (Pvt) Ltd v Alexander Films (SA) (Pty) Ltd 18 where Murray CJ proposed that where the first principle must be exercised "upon grounds which a reasonable man could have come to the conclusion arrived at", the second principle should not be departed from "without the existence of good grounds for doing so". 19The judge did not put forward any other criteria by which to evaluate "good grounds", although his statement is authority for the proposition that a departure from the second principle must be justified.The question of who is the successful party in a case, a question that is not without considerable complexity at times, is therefore primarily relevant for determining if there has been a departure from the second principle.Apart from the "good grounds" referred to by Murray CJ there are many High Court decisions in which it was held that a judge may depart from the second principle only if "special circumstances" are present. 20Because these precedents have not emanated from the Appellate Division or Supreme Court of Appeal, however, the standard of accountability that applies where a judge departs from the second principle is not clearly specified in our law.
Further, given that the larger body of authority sits behind the rule that the second principle yields to the first, the proposition that ultimately emerges is that costs are in the judicial discretion of the judge of first instance.Even if the courts' present approach to costs in constitutional and/or public interest matters was acknowledged as a new, third principle of the same status as the second principle, it would not trump the first principle but would rather have to yield to it, if the current understanding of the relationship between the first and second principle is retained.
therefore, is not unlimited, and there are numerous cases in which courts of appeal have set aside judgments as to costs where such judgments have contravened the general principle that to the successful party should be awarded his costs" (Ibid 357-358).And Solomon JA, noting the 'rule' towards the award of costs in claims of convention and reconvention in the Transvaal and English courts, stated: "That rule, however, is not a rigid one, but may be departed from in special circumstances.And it is only, in my opinion, where there are such special circumstances, that there is room for the exercise of any discretion by the judge or magistrate' (Ibid 360-361

Judicial exercise of discretion
The most frequently employed descriptor of the nature of the discretion is that it must be exercised "judicially" which essentially means "not arbitrarily". 21In Fripp v Gibbon De Villiers JP held that "[The presiding officer] should take into consideration the circumstances of each case, carefully weighing the various issues in the case, the conduct of the parties and any other circumstance which may have a bearing upon the question of costs and then make such order as to costs as would be fair and just between the parties". 22The judicial exercise of a discretion thus entails taking into account circumstances "which may have a bearing upon the question of costs" (although no parameters for identifying relevant from irrelevant circumstances are established) and trying to act in a manner that is fair to both sides.

Test for interference in the discretion of the court a quo
The test for interference in the discretion of the court a quo was foreshadowed in the dictum of De Villiers JP in Fripp v Gibbon where he held that if the presiding officer indeed weighs all issues in the case and makes an order as to costs that is fair and just between the parties "and does not act capriciously or upon any wrong principle, I know of no right on the part of a court of appeal to interfere with the honest exercise of his discretion". 23e most recent formulation of the test which subsequently crystallised 24 occurred in Naylor & another v Jansen 25 (commonly known as Naylor II) where Cloete JA held: Where the law has given a Judge an unfettered discretion, it is not for this Court to lay down rules which, while purporting to guide the Judge, will have the effect only of fettering the discretion.If, therefore, there are factors which the trial Court, in the exercise of its discretion does decide to take into account so as to reach a different 101/166 result, a Court of appeal is not entitled to interfere -even although it may or probably would have given a different order.The reason is that the discretion exercised by the Court's giving the order is not a "broad discretion" (or a "discretion in the wide sense" or a "discretion loosely so called") which obliges the Court of first instance to have regard to a number of features in coming to its conclusion, and where a Court of appeal is at liberty to decide the matter according to its own view of the merits and to substitute its decision for the decision of the Court below, simply because it considers its conclusion more appropriate.The discretion is a discretion in the strict or narrow sense (also called a "strong" or a "true" discretion).In such a case, the power to interfere on appeal is limited to cases in which it is found that the Court vested with the discretion did not exercise the discretion judicially, which can be done by showing that the Court of first instance exercised the power conferred upon it capriciously or upon a wrong principle, or did not bring its unbiased judgment to bear on the question or did not act for substantial reasons.

26
Further on in the judgment Cloete J emphasised that there is no "normal rule" applicable to the exercise of a discretion by a trial judge to award costs.The fact that a judge follows a particular approach to the award of costs creates no precedent binding on judges called upon to exercise such a discretion in exactly the same set of circumstances in future.The exercise of a narrow discretion involves a "choice between permissible alternatives" and, therefore "different judicial officers, acting reasonably, could legitimately come to different conclusions on identical facts". 27

The characterisation of the discretion
The distinction between a "broad" and a "narrow" judicial discretion which Cloete J recalled in Naylor II is a deeply-rooted feature of the law governing the relationship between appeal courts and courts of first instance and is one that is not confined to the question of costs.In Tjospomie Boerdery (Pty) Ltd v Drakensburg Botteliers (Pty) Ltd 28 Stegmann J traced the differences between, and casuistic development of, the two types of discretion.The issue at hand in the Tjospomie case was whether an order made by the court of first instance for the winding-up of a company was "just and equitable" as contemplated in one where its function was not simply to substitute its discretion for that of the court below, but rather to consider the prior question whether the court below had or had not exercised its discretion judicially.Stegmann J held that there was in fact no essential antithesis between these two cases and that a correct statement of the law was as follows: Courts are called upon to exercise discretions which vary infinitely in their nature and it is impossible to formulate a general principle to govern all cases where the exercise of a discretion by a court is brought on appeal. 31There is therefore no general principle to the effect that a court of appeal is always precluded from substituting its own decision for that of the trial court unless the latter can be shown to have been vitiated by a failure to exercise the discretion judicially.Rather, the question is to be approached casuistically and "there are particular categories of cases in which interference by a court of appeal with the exercise of a discretionary power by a court below is conditional upon the appellant's establishing a failure on the part of the court below to exercise the discretion judicially". 32Although it is merely a technical point, it should be noted that in the case of the law of costs, the discretion is therefore wide on the part of the presiding officer, whilst the discretion to interfere in that discretion on the part of an appeal court is narrow.
According to Greenberg JA in Ex Parte Neethling, the question of costs (together with postponements, an amendment of pleadings and alteration of sentence) had previously been held to fall within this category of cases. 33aling with the contention that the courts' approach in such cases reduced the 103/166 right of appeal to that of a review, Greenberg JA held that "the true position is that there is a right of appeal but in cases of this kind the Appeal Court, because of the nature of the case, has only a limited power of correction". 34e identification of the question of costs with this category of cases took place in Rex v Zackey 35 where Greenberg JA had referred to, in particular, De Villiers JP's dicta in Fripp v Gibbon. 36Here De Villiers' JP had held that "[q]uestions of costs are always important and sometimes complex and difficult to determine, and in leaving the magistrate a discretion the law contemplates that he should take into consideration the circumstances of each case …". 37The underlying rationale for characterising the question of costs in this manner is implied in this statement but stated more explicitly in Neugebauer & Co Ltd v Herman 38 where Innes CJ, in refusing to make a special order as to costs, said that "The matter was one with which [the presiding officer] was in the best position to deal". 39It is because the judge of first instance sits at the coalface of the dispute and experiences the complexities of the suit and the conduct of the parties (and where relevant their witnesses at first hand) that he or she is best positioned to make a fair order as to costs.

Constitutional Litigation 40
The Constitutional Court had occasion in Ferreira v Levin NO (2) 41 to consider the application of the traditional principles relating to the award of costs to constitutional litigation.Here Ackermann J stated: The Supreme Court has, over the years, developed a flexible approach to costs which proceeds from two basic principles, the first being that the award of costs, 104/166 unless expressly otherwise enacted, is in the discretion of the presiding judicial officer and the second that the successful party should, as a general rule, have his or her costs.Even this second principle is subject to the first.The second principle is subject to a large number of exceptions where the successful party is deprived of his or her costs.Without attempting either comprehensiveness or complete analytical accuracy, depriving successful parties of their costs can depend on circumstances such as, for example, the conduct of parties, the conduct of their legal representatives, whether a party achieves technical success only, the nature of the litigants and the nature of the proceedings.I mention these examples to indicate that the principles which have been developed in relation to the award of costs are by their nature sufficiently flexible and adaptable to meet new needs which may arise in regard to constitutional litigation.They offer a useful point of departure.If the need arises the rules may have to be substantially adapted; this should however be done on a case by case basis.It is unnecessary, if not impossible, at this stage to attempt to formulate comprehensive rules regarding costs in constitutional litigation. 42 The overriding import of this pronouncement is to uphold the primacy of the first principle -the judicial discretion to award costs -without binding that discretion in any way.Thus although the second principle is recognised, it is subject to the first.Further, the reasons upon which a judge may depart from the second principle are, seemingly, not subject to any moral order; i.e. a reason relating to the conduct of the parties is just as good as one relating to the nature of the litigants or the nature of the proceedings.Ackermann's reluctance to formulate '"comprehensive rules" that would impede the casuistic development of the traditional principles on costs in the context of constitutional litigation is perhaps understandable, however, given that this case was heard very near the start of the Constitutional Court's existence.
Over time, the Constitutional Court has developed a fairly consistent position on the award of costs in constitutional litigation.For instance, where the applicant -the individual or entity contending for the protection of constitutional rights -is successful, costs follow the result.Where, however, the respondent is successful, a government entity will not receive its costs if the applicant raised an important constitutional issue against it.In Motsepe v Commissioner for Inland Affairs 43 Ackermann J stated that "one should be cautious in awarding costs against litigants who seek to enforce their constitutional right against the state, particularly where the constitutionality of a statutory provision is attacked, or on the basis that mulcting the unsuccessful party with costs would undermine another important objective. 48e problem with these "principles" for the potential constitutional litigant, however, is their very flexibility, because there have also been cases in which the Constitutional Court has not refrained from making a costs order against litigants who unsuccessfully sought to ventilate a constitutional issue -the Motsepe case being one in point.Here Ackermann had continued as follows: 44 Ibid par 30.45 Friedman (n 40) 6-2 with particular reference to the case-law quoted in note 6. See also where the court found that while the school, as one of the appellants, had been unsuccessful, it had been at the centre of a difficult constitutional issue and had played an important role in ventilating a constitutional issue.Moreover, if ordered to pay the costs of the respondents, the funds would be diverted from being otherwise spent on learners.As such, the Department was required to bear the full burden of costs towards the respondents (par 118).

106/166
This cautious approach cannot, however, be allowed to develop into an inflexible rule so that litigants are induced into believing that they are free to challenge the constitutionality of statutory provisions in this court, no matter how spurious the grounds for doing so may be or how remote the possibility that this court will grant them access.This can neither be in the interests of the administration of justice nor fair to those who are forced to oppose such attacks.

49
Notwithstanding that the overwhelming majority of the Constitutional Court's decisions affirm the principle that a party who raises an important constitutional issue but is unsuccessful should not be mulcted in costs, a precedent like Motsepe indicates that, ultimately, the decision still lies squarely with the presiding officer, who is not bound to follow this principle; i.e. the "constitutional principle", if you like, stands in a similar relationship to the first principle on costs as does the second principle.Further, the Constitutional Court has not expressly indicated the formal or substantive criteria that should apply when a judge departs from the "constitutional principle".unsuccessful party who raises an important constitutional issue should not be mulcted in costs applies equally to other courts and has interfered in costs orders made by the Supreme Court of Appeal or High Courts.Importantly, the Constitutional Court's interference in these costs orders was not dependant on a prior enquiry according to the traditional test for judicial interference and the justification given was usually simply a reference to the demands of justice and equity, if any justification was given at all.In the Goedgelegen case, the Constitutional Court for instance simply stated: [T]he Supreme Court of Appeal ordered the claimants and the Department to pay costs jointly and severally including costs attendant upon the use of two counsel.The applicants have raised an important matter of land restitution and have succeeded in this Court.I can find no reason why this Court should not set aside the costs order of the Supreme Court of Appeal.

59
And in the Olivia Road case the court's words were: This is an appropriate case in which the City should be ordered to pay the costs of the applicants.The proceedings would have been obviated if there had been meaningful engagement before the case had been started.In the circumstances the City should also pay the applicants' costs in the High Court and in the Supreme Court of Appeal.

60
It is notable that in these cases the Constitutional Court found no difficulty in setting aside the costs orders in the lower courts.No reference, for instance, was made to the traditional test for interference in these orders and no elaborated justification was given when interference actually took place.The Constitutional Court's boldness in setting aside these costs orders is commendable but the lack of an account weakens the jurisprudential basis of this development.

108/166
While the Constitutional Court has not made an express pronouncement on the propriety of the reasons which may be put forward to justify costs orders, it is interesting to note that in Swartbooi v Brink 61 the court found that the High Court's approach and purpose towards the costs order was improper because the costs order was apparently aimed at teaching the councilors involved in the litigation a lesson and attempting to influence the way in which they made similar decisions in future.It impinged upon the executive domain and thus on the separation of powers.This decision could therefore be used to justify the proposition that in making costs orders certain reasons could be improper or inappropriate in the light of the value framework established by the Constitution.

4
Costs issue in the Biowatch matter

Nature of the litigation
It is firstly important to note that the issues raised in Biowatch I were of a constitutional nature.Biowatch has, in my view, established that it has a clear right to some of the information to which access was and is now requested; that the Registrar's failure to grant it access to such information as it was legally entitled to, constituted a continued infringement of Biowatch's rights under section 32(1)(a) of the Constitution; that Biowatch had no alternative remedy to enforce its rights; that Biowatch should not be non-suited for the inept manner in which the information sought in its fourth request, as well as in its notice of motion, is formulated, and that the Registrar would be entitled to refuse access to certain records, or parts thereof, in terms of the grounds for refusal contained in ch 4 of Part 2 of PAIA.(My emphasis.) Dunn AJ's use of the word "some" implies a partial victory on the part of Biowatch and does not reflect the fact that in respect of the legal issues raised in the case, Biowatch was indeed the overwhelmingly successful party. 72 See Biowatch I par 24.70 Dunn AJ points to the range of sub-issues encompassed by this question in note 26 of the judgment.71 Biowatch I par 66. 72 A summary of Dunn AJ's decisions on these issues and sub-issues (together with an assessment of the winning party in parentheses) is as follows: • In regard to the issue of the applicability of PAIA, he found that PAIA did not operate retrospectively and that its requests for information should not be invalidated simply because they were not made in terms of, and did not comply with, PAIA -Biowatch (Biowatch I par 29).Biowatch's success on the legal issues, however, was clouded by the issue that arose in connection with their actual requests for information -namely, that the requests, both in their original formulation and in the notice of motion, were apparently vexatious and made in an inept manner.Prior to launching the litigation, Biowatch had submitted four requests for information to the Registrar pertaining to the use of GMOs in South Africa, which were not fully granted. 73 • He rejected the argument, advanced by Monsanto, that Biowatch's right of access to information crystallised only when it launched its court application (on 22 August 2002 by which time PAIA had entered into force), finding that they crystallised on each of the occasions Biowatch submitted a request for information to the Registrar -Biowatch (Biowatch I par 30).

•
He rejected the argument that Biowatch should have exhausted the internal appeal procedures in PAIA -Biowatch (Biowatch I par 32).

•
He rejected the argument, advanced by the first to third respondents and Monsanto, that Biowatch should have exhausted the internal appeal procedures in the GMO Act -Biowatch (Biowatch I par 38).

•
On the argument advanced by Stoneville that s 31 of NEMA limited Biowatch's right of access to information under s 32 of the Constitution, he found that Biowatch's attempts to rely on s 31 of NEMA were misplaced because these provisions in NEMA ceased to apply the moment PAIA was promulgated and that for this reason alone s 31 could not constitute a permissible limitation on Biowatch's rights under s 32 -Biowatch (Biowatch I par 36).

•
Similarly, on the argument that Biowatch's s 32 right was limited by s 18 of the GMO Act, he found -contrary to the submissions advanced by Stoneville -that interpreting s 18(1)(a) in a manner that did not limit the right of access to information was in keeping with the approach to interpretation required by the Constitution -Biowatch (Biowatch I par 37).

•
On the commercial confidentiality of some of the information requested, he found that PAIA could apply retrospectively to the degree that the Registrar could rely on the provisions of Part 2, Chapter 4 as grounds for refusing access to the information -Respondents (Biowatch I par 41).

•
He further found that the strict standard that applies to the intelligible identification of documents during a process of litigation, as embodied in Rules 35(14) and 38(1) of the Uniform Rules of Court, did not apply to a request for information under s 32 of the Constitution and that Biowatch should not be non-suited on the excessive breadth of its requests for information alone -Biowatch (Biowatch I par 43-44).73 The nature of Biowatch's requests and the Registrar's response thereto are set out in detail in Biowatch I par 18-23.The information Biowatch sought related to: Risk assessments accompanying requests for trial and commercial releases of GMOs, permits and approvals granted in terms of the GMO Act including information in such authorisations on the methods and plans for monitoring GMOs and for emergency measures in the case of an accident; details of public participation relating to the granting of authorisations in terms of the GMO Act; the location of field trials and commercial 112/166 Reflecting, perhaps, Biowatch's frustration with the fact that many aspects of these requests for information were either ignored or overlooked by the first respondent, the fourth request was particularly exhaustive as regards the information to which access was sought. 74Rather than precisely listing the information to which it sought access in its notice of motion, Biowatch simply attached the records of its four attempts to correspond with and elicit information from the statutory respondents.Although Dunn AJ did not non-suit Biowatch because of the excessive breadth of its requests for information, he found some substance in the respondents' submissions that its requests for information were inherently vague, vexatious, oppressive and amounted to a "fishing expedition".He nevertheless acknowledged that requesters for information under section 32(1)(a) of the Constitution would often not have knowledge of the precise description of the record to which access is sought, and that a duty rested on the Registrar to adopt an active role in ascertaining the records to which access was sought. 75In regard to the 'inept' manner in which Biowatch had approached the court, he complained: 76 Unfortunately, Biowatch also did not engage in the task of specifying in its notice of motion the precise list of information it seeks access to.Its approach seems to have been to expect the respondents and the court to read through all the correspondence and to divine precisely what information is requested and what information is outstanding.
He nevertheless proceeded with the task of determining which of the items in the first to fourth requests were valid and justifiable.Essentially, he found that all of the items in the first to third requests had either already been granted or 113/166 were encompassed in the fourth request and that reference to the first to third requests in the notice of motion was therefore entirely unnecessary. 77He then examined the eleven items of information listed in a Schedule to the fourth request, essentially finding that three of them were impermissibly vague or overly broad.Biowatch accordingly had a right to the other eight items, subject to the right of the Registrar to refuse access on the basis of Part 2, Chapter 4 of PAIA. 78 appeal, the majority and minority decisions differed on the extent of Biowatch's success.While both Mynhardt J and Poswa J seemed to agree that Biowatch was the successful party as against the statutory respondents (although Mynhardt J placed hardly any emphasis on this whereas Poswa J expressly affirmed it 79 ), they differed on Biowatch's success vis-à-vis Monsanto.
Mynhardt J went to some lengths 80 to argue that Monsanto was the successful party against Biowatch on the basis of Dunn AJ's finding on the retrospective application of Chapter 4, Part 2 of PAIA, referring to this fact no less than three times in the introductory part of his judgment. 81In contrast, Poswa J argued that Biowatch was the successful party against Monsanto, relying on Dunn AJ's own summary of the judgment 82 and Biowatch's victory on the question on whether it was acting in the public interest.Because the majority and minority decisions differed on Biowatch's success against Monsanto, they also differed in this regard on the extent to which there had been a departure from the 77 Ibid par 45-50.78 Ibid par 52-65.Was Biowatch successful then as regards its actual requests for access to information?On the basis of the judge's finding on those requests that were necessary for purposes of citing in the notice of motion, they were successful on eight out of eleven items (a success rate of some 72%).However, a more formalistic approach would take into account those items of information formulated in the first to third requests, which were dismissed as unnecessary by the judge.If one construes dismissal as failure, then Biowatch's success rate as regards its actual information requests was lower.However, given that Dunn AJ found that most of the items in the first to third requests were incorporated in the fourth request, this approach -in my view -is overly technical and inaccurate.In the case of its actual requests for information, Biowatch was successful in establishing a clear right to most of the information to which it sought access, but this victory was qualified by the respondents' one victory on a question of law, namely that relating to the retrospectivity of Part 114/166 second general principle applicable to costs, as the next section shows.In the opinion of Mynhardt J there had in fact been no departure from the general rule, while in the view of Poswa J there had been an extraordinary departure from the principle in that a successful party had been required to pay the costs of an unsuccessful party.

Decision on costs in the court a quo
Notwithstanding Dunn AJ's own summary assessment that Biowatch's application had met with some degree of success, he decided that no costs order against the statutory respondents should be made in favour of Biowatch ("the first costs order"), and that Biowatch should be ordered to pay the costs of Monsanto ("the second costs order").His reasons for both costs orders were contained in and confined to the following paragraph: 83 As far as costs are concerned, the general rule in litigation is that the costs should follow the result.However, although Biowatch has been partially successful in obtaining some of the relief sought, the manner in which some of its requests for information were formulated, as well as the manner in which the relief claimed in the notice of motion was formulated, has convinced me that it should not be granted a costs order in its favour in these circumstances.Furthermore, the approach adopted by it compelled Monsanto, Stoneville and D&PL SA to come to court to protect their interests.The issues were complex and the arguments presented by them were of great assistance.Stoneville and D&PL SA did not seek any costs order against the applicant.On behalf of Monsanto its counsel sought an order for costs against the applicant.In my view the applicant should be ordered to pay Monsanto's costs.No other order as to costs is warranted in the circumstances of this case.
Dunn AJ demonstrated an awareness that he was departing from the second generally accepted principle relating to costs and offered three reasons in justification: (i) The manner in which Biowatch had formulated its requests for information and the relief claimed in the notice of motion -this reason, presumably, being associated with the judge's earlier comments as regards the vexatious, oppressive and inept nature of Biowatch's conduct; (ii) that Biowatch's approach to the matter had compelled the fourth to sixth respondents to come to court to protect their interests; and (iii) that the arguments presented by these respondents were of great assistance.As it 115/166 appears from the judgment, these were the parameters that guided the exercise of Dunn AJ's discretion as to costs.
On appeal, two disputes of a factual nature relating to Dunn AJ's reasoning on the costs award emerged between the majority and minority judgments.Firstly, there was disagreement on whether or not Monsanto had in fact been compelled to come to court.Mynhardt J simply agreed with Dunn AJ, finding no merit in the submissions that Monsanto could or should have engaged the statutory respondents and indicated to them which documents it regarded as confidential; that Monsanto could have relied on the statutory respondents to protect its confidential information, and that Biowatch was under no duty to approach Monsanto beforehand to ascertain if it was amendable to the order being sought. 84Poswa J, on the other hand, regarded it as an "anomaly that the Secondly, there was a sharp difference of opinion as to whether Dunn AJ had in fact taken the constitutional/public interest nature of the litigation into account and, related thereto, the "trend" or "principle" or "rule" relating to costs in constitutional litigation.From the judgment itself, it appears that Dunn AJ did not, but in paragraph 15 of his judgment on the application for leave to appeal he maintained that "a failure to expressly articulate all the grounds in favour of not making such an order certainly do (sic) not mean that they were not considered". 86He pointed out that he was acquainted with the case law concerning costs in constitutional litigation, not least because he had been involved in at least one of them.87He therefore implied that these factors were where a party comes to Court to protect its constitutional rights, did I not, you know, treat that too lightly?"The thrust of counsel's response was to reassure the judge that the test was whether he had "judicially applied his mind" to the exercise of the discretion and not if another court would come to a different decision based on those facts.
Mynhardt J essentially dismissed the significance of this exchange by holding that "I do not think that the impecuniosity of Biowatch, if that is a fact, provides a ground or reason for holding that Dunn AJ misdirected himself.To the extent that a costs order against an NGO might have a 'chilling' effect that is something that is common knowledge". 89e basis of Biowatch's appeal against the two costs orders to a full bench of the Transvaal Provincial Division was that Dunn AJ had misdirected himself in a number of respects: Firstly, Biowatch argued that Dunn AJ acted capriciously in making the first and second costs orders in that the decision was based on the manner in which the relief had been formulated. 93Secondly, it argued that Dunn AJ misdirected himself by failing to take relevant considerations into account, including the "trend" in judgments delivered by the Constitutional Court and the other courts of South Africa not to make costs orders in matters in which a party seeks to establish an important constitutional principle or in matters where the protection of the environment is relevant. 94As a result of these misdirections, it accordingly maintained that the appeal court was entitled to determine the issue of costs afresh.As regards the "test" for judicial interference in the award of costs by the court a quo, Biowatch maintained that the "traditional test" as recently re-formulated by Cloete J in Naylor II applied only where the court had followed the general principles applicable to costs.But where the trial court had departed from the general rule and had deprived the successful party of his costs, "a court of appeal will ordinarily interfere and apply its own judgment as to whether or not there are any grounds to depart from the general rule". 95 This article focuses on how the fundamental rules on costs were applied in the majority and minority decisions on appeal.As such, it does not deal with two of the other important legal issues raised in the case, namely, whether the appeal was barred by the provisions of s 21A of the Supreme Court Act 59 of 1959 (in terms of which an appeal court has a discretion to dismiss an appeal if it has no practical effect other than costs); and whether the provisions of s 32 of the NEMA were applicable to the judge's discretion.In respect of the latter it is my opinion that the provisions of s 32 were in fact applicable and that this in itself, on the authority of Swartbooi v Brink (n 51) would have been enough to establish that Dunn AJ had misdirected himself as to costs.93 Biowatch II (Majority) par 24.94 Ibid par 30.95 Biowatch II (Majority) par 21.Counsel for Biowatch had obtained this formulation from Merber v Merber 1948 (1) SA 446 (A) at 453G-H where Greenberg JA had stated: "[W]hen a successful party has been deprived of his costs in the trial court, an appeal court will enquire whether there were any grounds for this departure from the general rule and if there are no such grounds, then ordinarily it will interfere".

The majority decision
Mynhardt J commenced by emphasising a fairly broad interpretation of the first principle applicable to costs.In order to succeed, Mynhardt J held, Biowatch needed to show that the court a quo failed to exercise a judicial discretion in making the costs orders, affirming a line of authority that stretches back to the English decision in Ritter v Godfrey 96 where it was said: "The discretion must be judicially exercised and therefore there must be some grounds for its exercise, for a discretion exercised on no grounds cannot be judicial.If, however, there be any grounds, the question of whether they are sufficient is entirely for the Judge at the trial and this court cannot interfere with his discretion". 97Mynhardt J assumed that "any grounds" in this passage meant "any grounds on which a reasonable person could come to the conclusion arrived at". 98As such, a judge had a "wide unfettered discretion" to make a costs order after taking into account all of the relevant factors or circumstances of the case.There was no "normal rule" or "general rule".The mere fact that an appellate judge would have given more weight to the grounds did not mean that the trial judge had acted arbitrarily, that is, without a judicial discretion. 99nhardt J aligned himself with the test for interference in a costs order made by a court a quo formulated by Cloete J Naylor II and pointed out that the principle that an appeal court will interfere with the exercise of a narrow or "true" or "strong" discretion only in limited circumstances had also been  ).He did not point out that the discretion here related to s 173 of the Constitution according to which Supreme Court of Appeal and High Courts are required "to regulate their own processes".This case concerned an application to televise the proceedings of the Shaik trial, and the Constitutional Court, while cautioning that the exercise of the powers in s 173 may well not be capable of a single characterisation for purposes of determining the correct approach on appeal, found that the nature of the discretion the Supreme Court of Appeal was asked to exercise in this particular case was a narrow one (par 41).Accordingly, the question was not whether the Constitutional Court would have reached a different conclusion on permitting live radio and television broadcasting of the proceedings, but if the Supreme Court of Appeal had acted judicially in In applying these principles to the facts in Biowatch II, Mynhardt J firstly disposed of the argument advanced by Biowatch that the "traditional test" for interference in costs orders, as formulated in Naylor II, applied only where a court had followed the general rule that a successful litigant should be awarded costs, by pointing out that in Naylor II the Supreme Court of Appeal was dealing with a case where the general rule had indeed not been followed, and yet the court affirmed the traditional test, rather than holding that it would "ordinarily interfere" on the authority of Merber. 103He failed, however, to take into account the constitutional jurisprudence where the Constitutional Court interfered in the costs orders of lower courts without resorting to this test.
In response to Biowatch's argument that Dunn AJ had acted capriciously in making the first and second costs orders, Mynhardt J noted that in essence Dunn AJ had disapproved of the manner in which the relief was formulated in the four letters written by Biowatch and in the notice of motion and had decided to deprive Biowatch of its costs against the statutory respondents "as a mark of his disapproval".In Mynhardt J's view Dunn AJ was entitled to do so and he could find no fault with his decision. 104He therefore did not consider the propriety of this rationale within a constitutional context.
Mynhardt J also dismissed Biowatch's contentions that Dunn AJ had failed to take relevant considerations into account as regards the "trend" relating to costs orders in constitutional litigation.Mynhardt J's position was that the exercising its discretion.It would thus interfere only where the Supreme Court of Appeal had not acted judicially, or had based the exercise of its discretion on wrong principles of law or a misdirection on the material facts (Ibid).101 1999 ( 4 120/166 courts' tendency not to make costs orders against parties litigating on the basis of an important constitutional principle was a trend and not a "rule" that fettered the discretion of the trial judge to make an order as to costs.In this regard he relied in particular on Ackermann J's statements in Motsepe 105 (where, as noted above, the applicant had been unsuccessful in obtaining the constitutional relief sought) and rejected Biowatch's reliance on the Sanderson case 106 finding that the dictum here related only to criminal proceedings.

Minority decision
Contrary to Mynhardt J, Poswa J emphasised both the first and second principles on costs.He noted that according to the "traditional test" for interference in the award of costs by the court of first instance, an appeal court will not readily interfere.However, while the trial court has a discretion, this is not unlimited, being subject to the general principle stated in Fripp v Gibbon 1913 AD 354 that 'to the successful party should be awarded his costs '. 107 Acknowledging Cloete J's restatement of the traditional test for judicial interference in Naylor II, the judge then emphasised that while the case was a good example of how a successful party (in casu the defendant) was denied costs contrary to the general principle stated in Fripp v Gibbon, it also illustrated the need to indicate reasons why this had been done. 108Poswa J pointed out how Cloete J had found that the trial judge in the Naylor matter, Willis J, had been acutely aware that he was exercising a discretion as well as the parameters thereof.Willis J had justified his decision on costs with reference to the nature of the action (that in defamation actions the quantum essentially takes the form of a solatium), the motive of the plaintiff (that the plaintiff needed to persist with the application in order to vindicate his reputation), and the attitude of the defendant in making the without-prejudice tender (that the tender contained no acknowledgement that a defamatory statement had been made and no apology). 109These, Poswa J found, were substantive reasons as to why 121/166 there was a departure from one of the two general principles relating to costs. 110While Willis J was fully conscious of the fact that he was exercising a discretion and gave reasons for his departure from the general rule, "Dunn AJ did not, in his judgment, when dealing with costs, mention the question of public interest action, and give reasons, therein, why he was disregarding that factor" (my emphasis). 111swa J comprehensively reviewed the approach towards costs in constitutional and public interest litigation in foreign jurisdictions as well as in the Constitutional Court, Supreme Court of Appeal and High Courts, 112 and came to the conclusion that the Constitutional Court adopts the trend, in vogue in numerous foreign jurisdictions, of not awarding costs against applicants in public interest litigation.This approach amounted to more than a "trend" and could be characterised as a "flexible rule". 113swa J essentially found that Dunn AJ had done more than depart from this "flexible rule" in an unjustifiable manner -he had completely failed to take it into account.Poswa J examined the exchange that had taken place between Dunn AJ and counsel for Monsanto in respect of Biowatch's affidavit indicating the effect of an adverse costs order upon it and found that as a matter of fact Dunn AJ had not taken into account that Biowatch was an NGO acting in the public interest when making the costs order.The judge therefore concluded: "I do not, therefore, agree with the majority judgment that, that omission on Dunn AJ's part does not show 'that [he] had committed a demonstrable blunder'.In my view, it demonstrates just that".122/166 holding that no third principle relating to costs in constitutional litigation had been recognised.On the other hand, Poswa J held that the judicial discretion as to costs was subject to both the second principle and, evidently, a third principle relating to costs in constitutional litigation which had, by now, been recognised as a "flexible rule".Quite clearly, these judges held different views on the relationship between the first principle and the second and third principle (assuming the latter has such a status for this discussion) and it is difficult to say who was correct according to the existing law.As noted above, the weight of authority lies behind the proposition that the second principle yields to the first, but does not indicate the circumstances in which this may occur, while the view put forward in Levben Products is that the second principle may be departed from only on "good grounds".
However, for Mynhardt J, it is clear that "good grounds" would be equivalent to "any rational ground".He was satisfied that Dunn AJ had exercised his discretion as to costs judicially because there were some grounds for its exercise, which a reasonable person could accept.In the majority opinion, the cryptic 115 manner in which Dunn AJ justified his departure from the principle that a successful party should be awarded his costs was sufficient.The judicial exercise of the discretion did not require that Dunn AJ should have expressly taken into account the public interest motive of the litigant and constitutional nature of the litigation or justified his departure from the trend toward costs in these matters.By contrast, in highlighting cases in which judges had given express reasons for their departure from the fundamental rules on costs, Poswa J implied that the judicial exercise of a discretion on costs requires, at least , taking into account the second and third principle and expressly providing substantive reasons when departing from them.
While both judges affirmed the traditional test for interference in costs orders (and thereby also the characterisation of costs as broad in the court a quo and narrow in the court of appeal) their application of this test to the facts yielded different results.Bringing in the test for interference in the exercise of such 123/166 discretion, as formulated by Cloete J, the aspect of the test which seems most applicable to the judges' reasoning is the failure by the judge a quo to act for "substantial" reasons.Again, for Mynhardt J what counted as a "substantial reason" was equated with a reason which a reasonable person could accept, viz the "unreasonable" manner in which Biowatch had formulated its requests for information and notice of motion.Any rational reason would have sufficed.
In contrast, Poswa J evidently regarded the reasons Dunn AJ did put forward as insufficient and held that his failure both to take into account and to expressly justify his departure from the trend toward costs in constitutional litigation vitiated the discretion as "judicial".In this view, there are potential reasons which a judge should take into account and, where they are relevant, he or she should expressly account for failing to apply them in the making of a costs order.A "substantial" reason is therefore not simply any rational reason, but the reasons one would expect a judge to take into account in the particular circumstance.
Neither judge expressly considered the propriety, in a constitutional context, of the reasons Dunn AJ did present in justification of his award on costs, in particular the propriety of an order that essentially sought to punish Biowatch, as a mark of the court's disapproval, for the manner in which they had conducted the litigation.

5
Reconceptualising the Fundamental Rules on Costs to meet new needs in constitutional and/or public interest litigation

An unsatisfactory situation
The fundamental rules on costs currently fail to meet new needs which arise in constitutional and/or public interest litigation and the reasons for this are essentially four-fold, based on the right of access to justice, the right to equal protection and benefit of the law, the need for proportionality, and the accountability of the judiciary.124/166

Access to justice
The fundamental rules on costs currently fail to effectively engage with the insight that the rules and principles governing costs orders -including security for costs -have a critical impact on access to justice.In this regard, the discretion on costs differs from the other types of discretion which Greenberg JA identified in Ex Parte Neethling and Rex v Zackey, such as an amendment of pleadings, the granting of a postponement, or the alteration of a sentence on this ground.In the context of environmental litigation, Lord Justice Brooke has written that there are three immediately obvious obstacles to public access to the courts: "The costs of the courts (in terms of high court fees); the cost of lawyers (in terms of even higher professional fees) and, above all, the risk of having to pay one's opponent's costs if one loses, and the uncertainty at the outset of litigation as to how large those costs will be". 116Arguments based on the so-called "chilling" effect adverse costs orders may have on public interest and constitutional litigation are in fact understated because these three obstacles, as recognised in a number of the cases and academic articles cited by Poswa J in his minority judgment, 117 do more than "chill" such actions -they kill them at the root, because they deter the individuals and public interest organisations that would otherwise institute an action from even arriving at the court's doors out of fear that the financial burden of the litigation will effectively destroy them. 118In this instance the use of the word "chill" is a euphemism that covers the true effect of the court's application of costs principles and rules.
Further, the potential of an adverse costs order can be used in a way that undermines the objectives underlying the relaxation of locus standi, as has occurred in respect of constitutional litigation by virtue of section 38 of the 126/166 indeed vital to a society that is based on the rule of law. 125Further (as they were testing the constitutionality of a contractual provision) they held that section 34 "not only reflects the foundational values that underlie our constitutional order, it also constitutes public policy". 126One could contend that as the fundamental rules on costs omit to provide certainty that the nature of the litigation and the position of the litigant will be considered when a judge exercises his discretion on costs, this constitutes an infringement of section 34, and there would be no bar (as was the case in Napier) to subjecting the fundamental rules on costs -as law of general application -to a limitation analysis.Dugard has already shown the extent to which there is a demand for access to justice (and the extent to which this is not being met by direct access to the Constitution Court). 127And in a recent paper, Heywood and Hassim have pointed to the disjuncture between the realisation of section 35 rights and the progressive erosion of section 34 rights, as measured in terms of funds directed to criminal and civil matters respectively by the Legal Aid Board. 128They have also argued convincingly that "unless the need for justice and remedies for injustice are met by the courts and the law, there will be negative consequences for the popular legitimacy of the courts and indeed the Constitution itself". 129 In the light of all this, it is surely time that the courts acknowledged the linkages between costs and access to justice and reformed the laws on costs in a manner that favours access to justice rather than undermines it, as would appear to have been the case in the Biowatch matter.

Equal protection and benefit of the law
If section 34 not only reflects the foundational constitutional values but also reflects public policy, then the same must hold true of section 9(1) of the Constitution -the right of equality before the law and the right to equal 127/166 protection and benefit of the law.But unless there are mechanisms in the process of litigation whereby the inequality of contending parties can be taken into account, the law benefits only those who are most powerful and wellresourced to invoke its protection.The Biowatch matter is typical in this regard.
Biowatch depended upon advocates and attorneys willing to offer their services pro bono or had to undertake public fundraising to secure the services of an able legal representative, yet was pitted against an assembly of lawyersbeing the very best that money could buy -representing the government and each of the biotechnology companies; a situation Lord Brooke has described as "a cadre of Goliaths with their clubs and battle axes" in one corner, and "a squad of Davids, with their slings and peashooters" in the other. 130If the trial judge is to make a costs award that is "fair and just between the parties" (to adopt De Villiers' formulation in Fripp v Gibbon), then this requires closer regard to the actual status of the parties, the effect of an adverse costs order upon them and the differing roles they play in upholding a constitutional democracy.An NGO such as Biowatch fulfills an important constitutional function in both representing and encouraging a diversity of views on the issue of GMOs in South Africa.Is it in the interests of a constitutional democracy that an adverse costs order can be both made and upheld against Biowatch (or any other similarly situated institution) where it raises an important constitutional issue and is at least partially or even substantially successful, and the effect of that order is to end or at least substantially strain its ability to function? 131I think not.
Further, there seems to be little appreciation of the debt owed by society to the individuals and organisations who institute litigation, not for their own personal financial gain, but in the interest of a significant section of the public (if not the public as a whole) and who through their success -and sometimes even their failure -secure a benefit for those who had no part in the litigation.In fact, in 128/166 other jurisdictions the inherent unfairness of this situation has been recognised and attempts to remedy it through making funds available through other means, predominantly legal aid, have been made. 132However, given the de facto situation regarding legal aid in South Africa, a duty is imposed on the courts to do everything in their power to facilitate access and, towards this end, to reduce the burden of litigation on poor people and poorly-resourced organisations.129/166 as a minor error or oversight in any court document would be viewed in a very serious light by a judge and could justify the matter being thrown out of court.

Proportionality
These rules are not unimportant because they are closely bound up with the integrity of the legal system as a means of resolving disputes between different parties in a manner that is "fair".However, since the advent of democracy even these rules are subject to the overriding objectives of the Constitution to create a society that is open, democratic and diverse, in which ordinary people are able, through the courts, to vindicate their constitutional rights.The necessity and suitability of this reason should therefore be subjected to scrutiny in the light of these broader considerations.

Accountability of the judiciary
The final reason is that currently the fundamental rules on costs do not adequately capture the accountability of the judiciary to fundamental constitutional norms and standards, which includes the norm of accountability itself.In the South African Broadcasting Corporation case, the Constitutional Court stated: "The foundational constitutional values of accountability, responsiveness and openness apply to the functioning of the judiciary as much as to other branches of government". 137The present formulation of the nature of the judicial discretion in awards of costs does not adequately specify whether that discretion is fettered or unfettered and, more importantly, if fettered by principles such as the second principle on costs or a principle on costs in constitutional and/or public interest litigation, the circumstances in which such principles can be departed from and the standards of justification that then apply.130/166

Suggestions for possible reform
The parameters for possible reform have already been foreshadowed in my critique above. 138It seems sound to maintain that a judicial discretion to award costs inheres in the judge of the first instance, on the basis that he or she is well positioned to evaluate the relative success of the parties in the light of their overall behaviour and conduct in court.However, this discretion is not unfettered and is subject to both the traditionally recognised second principle on costs as well as the principle that has emerged in the context of constitutional litigation, which is that where a party has raised an important constitutional principle, an adverse costs order should generally not be made.
In exceptional circumstances, it may even be justified to order the successful party to pay such a party's costs.These considerations are all the more compelling where the litigation is of a public interest nature.A departure from either the second or "third" principle on costs should be justifiable in the light of the values, norms and standards of the Constitution and not merely on the basis of rationality.Moreover, in order to demonstrate accountability to such norms, values and standards, a judge's reasons for departure must be set out expressly.The traditional categorisation of costs as one which vests the court of appeal with only a narrow discretion for interference should be expressly reevaluated in the light of the significant role that costs orders play in access to justice and to account for the de facto practice in the Constitutional Court in certain cases.A test formulated more along the lines of a proportionality analysis may be more appropriate.At the very least, the reference to "substantial reasons" in the existing test should be taken to refer to reasons which are justifiable in the light of the value framework established by the Constitution.
In conclusion, in making costs orders courts wield considerable power.This is a power that operates "in advance," in the sense that it is a critical factor in the decision to launch litigation.Courts should exercise this power in a manner that 138 I am assuming that the power to effect such reform lies with the courts' power to develop the common law so as to promote the spirit, purport and objects of the Bill of Rights, as per s 39(2) of the Constitution.131/166 is conscious of its effects, particularly of its differential effects on different classes of litigant, and should be held accountable for their costs decisions in the light of constitutional values, norms and standards.
Trustees, Biowatch Trust v Registrar: Genetic Resources (Biowatch I) 3 and the subsequent appeal on this issue in Trustees of the time being of the Biowatch Trust v Registrar, Genetic Resources (Biowatch II) 4 in which Mynhardt J * Senior Lecturer, School of Law, University of the Witwatersrand, South Africa.Email: Tracy-Lynn-Field@wits.ac.za. 1 A term used in Cilliers Law of Costs, arguably the most authoritative work on costs in South Africa. 2 parties in the court a quoOn the face of it, the merits of the case related to Biowatch's right to the information to which it sought access, and if the Registrar's failure to furnish such information constituted a continued infringement of Biowatch's rights under section 32(1)(a) of the Constitution.Much of the judgment, however, was concerned with the array of defences brought by the phalanx of lawyers representing the first to sixth respondents, which included:69  (1) The applicability of PAIA to the matter together with Biowatch's failure to comply with such provisions (and a variety of sub-issues relating to this);70  (2) Biowatch's failure to exhaust the internal appeal remedy in the GMO Act; (3) the alleged commercial confidentiality of the information sought by Biowatch; and (4) Biowatch's failure to articulate properly the information to which it sought access.Dunn AJ in the court a quo summarised the outcome of the case in relation to these issues as follows:71 only legal issue on which the respondents achieved any substantial success was the retrospective applicability of Part 2, Chapter 4 of PAIA, which could thus be relied upon by the Registrar as grounds to refuse Biowatch access to the information.
Registrar and Monsanto had failed to communicate about Biowatch's request.If they had communicated Monsanto could have conveyed its fears regarding the protection of confidential information in the custody of the Registrar, who would in turn have communicated this to Biowatch.It would therefore not have been necessary for Monsanto to join as a party in the application".85 ) SA SA 799 (W) at 808B.102 Biowatch II (Majority) par 22. 103 Ibid par 21. 104 Biowatch II (Majority) par 25. 114 and Poswa JJ proceeded from the first principle on costs, Mynhardt J relied upon it heavily, barely mentioning the second principle and 110 Ibid par 29. 111 Ibid.112 Ibid par 43-61.113 Ibid par 61.114 Ibid par 42.
Proportionality is a value implicit in the Constitution.It is specifically associated with the test for limitations of rights, 133 but has over time been applied to discretions exercisable in other areas of law.134Proportionality, in essence, requires a balancing of different interests 135 and, as Mokgoro and Sachs pointed out in the Bel Porto case, a specific understanding of justifiability, i.e. one which requires "more than a mere rational connection between the reasons and the decision …", 136 but rather a substantiation of necessity and suitability in the light of certain factors.According to the present understanding of "judicial discretion", at least as proposed by the majority in Biowatch II in the making of a costs award, one rational reason is as good as any another.Thus, in the reasoning of the court a quo, the first and second costs orders were justified on the basis that Biowatch had acted improperly in the manner in which it had formulated its relief.The unarticulated moral foundation of this reason is the integrity of the legal process and the time-worn professional rules that have developed around "fairness" in litigation.Judges are steeped in the ethos of these rules and guard them vigilantly, such that what would appear to a layman 132 See the Brooke's (n 2) 347-349 description of the English Access to Justice Act 1999 and the work of the Legal Services Commission as well as Preston 1991 JELL 61-62.133 S v Makwanyane 1995 (3) SA 391 (CC), 1995 (6) BCLR 665 par 105.134 For instance, the use of force, in terms of the Criminal Procedure Act 51 of 1977 in effecting an arrest (see Ex Parte Minister of Safety and Security: In re S v Walters 2002 (7) BCLR 663 (CC)) and the right to just administrative action (see in particular Bel Porto School Governing Body v Premier of the Province, Western Cape 2002 (3) SA 265 (CC), 2002 (9) BCLR 891 par 162 ff).135 Makwanyane (n 132) par 104, Bel Porto (n 130) par 162.136 Bel Porto (n 133) par 164-165.As the comments of the minority in this case make clear, aproportionality test is particularly suited to considering the question of costs on appeal because it effects a compromise between a test based purely on a rational connection and one where the judiciary substitutes its decision for that of the decision-maker of first instance.
The cases noted by Cilliers as contra in fact affirm that the discretion is not unfettered.17 This tension was already evident in the very early decision of Fripp v Gibbon where De Villiers JP had held: "Discretion implies latitude, and it cannot affect the matter that the members of this Court, or some of them, might have arrived at a different conclusion … 16 1710 See for instance Lord De Villiers CJ's opening remarks in Fripp v Gibbon 1913 AD [I]n my view it is undesirable to lay down any hard and fast rules for the guidance of magistrates to which they will be expected to conform in the absence of exceptional circumstances.Where the law has given the magistrate absolute discretion, free and unfettered by any rules, it is not for the Court to lay down rules which, while purporting to guide him, will only have the effect of fettering his discretion" (n 10 above at 363-364, my emphasis).In the same matter, Lord de Villiers CJ held that "[t]he discretion of such Court, T HUMBY PER/PELJ 2009(12)1

T HUMBY PER/PELJ 2009(12)1
have an unduly inhibiting or 'chilling' effect on other potential litigants in this category". 44 45certain occasions, the court has extended this particular "principle" to costs orders in which the respondents were private companies.46Movingastep beyond a refusal to mulct an unsuccessful party with costs, in certain exceptional circumstances the Constitutional Court has even ordered the state or state institutions to pay the costs of unsuccessful parties, simply on the basis that the unsuccessful parties raised important constitutional issues47 42 Ibid par 3. 43 1997 (2) SA 898 (CC), 1997 (6) BCLR 692.
Occupiers of 51 Olivia Road, Berea Township v City of Johannesburg, 57 Njongi v Member of the Executive Council, Department of Welfare, Eastern Cape 58 the Constitutional Court has indicated or simply implied that the principle that an So far, I have looked at costs granted by the Constitutional Courts in proceedings before it.What of its position on costs orders made by the Supreme Court of Appeal or High Courts?The Constitutional Court has affirmed the traditional test for interference50and, where it has found that the court misdirected itself, has interfered in the costs order.51Where is has done this, it has also implicitly accepted the characterisation of costs issues as narrow in the court of appeal.52However,beginning with Sanderson v Attorney-General, Eastern Cape 53 and subsequently in ANC v Minister of Local Government and Housing , KwaZulu-Natal, 54 Van der Merwe Road Accident Fund, 55 Department of Land Affairs v Goedgeleden Tropical Fruits (Pty) Ltd, 56 49 Motsepe (n 43) par 30.50 See the cases cited in Friedman (n Governing Bodies of State-aided Schools: Eastern Transvaal 1999 (2) SA 91 (CC), 1999 (2) BCLR 151 par 53.For the affirmation of the characterisation in other contexts see Justice O Regan's comments in Giddey (n 46) notes 16 and 17.
Genetically Modified Organisms Act 15 of 1997 (section 18).65Secondly, although this matter was placed in dispute by the respondents, all of the judges in the matter either accepted or concluded that Biowatch was indeed acting in the public interest.66Inmy view Biowatch I was clearly a public interest matter.Biowatch maintained that it sought access to information held by the State in order to further its objective of protecting the right of everyone to an environment that is not harmful to their health or well-being and to have the environment protected for present and future generations." 65 Par 2. 66 On the issue of whether Biowatch had acted in the public interest, Mynhardt J in Biowatch II (Majority) had been prepared, like the court a quo, to "accept" that Biowatch had indeed acted in the public interest in instituting the proceedings (par 38).In contrast, Poswa J, after an exhaustive review of the founding affidavits and the respondents' responses, found that (Biowatch II, Minority, par 12): "[I]t is evident that the appellant was, in approaching the statutory respondents for information, actuated by genuine concerns about public interest.It is also important, in my view, to highlight that the fourth respondent [Monsanto] vehemently and relentlessly disputed that Biowatch was acting in the public interest … I seek further to emphasise and illustrate that, unlike the court a quo and the majority judgment in this Court, I do not merely accept that the applicant acted in the public interest but find that it demonstrated that with convincing evidence.In my view, the fact that the fourth respondent denied that the appellant was acting in the public interest is not The trustees of the Biowatch Trust brought an application for access to information relating to genetically-modified organisms(GMOs)held by the first respondent, being the Registrar: Genetic Resources; and the second respondent, being the Executive Council for Genetically Modified Organisms (the statutory respondents).62TheMinister of Agriculture had been joined as a third respondent, while the fourth to sixth respondents, being Monsanto South Africa (Pty) Ltd, Stoneville Pedigreed Seed Company and D&PL South Africa respectively, had intervened in the application at a later stage.The fourth to sixth respondents were all private biotechnology companies with varying levels of interest in the South African GMO market.The requests were made between the date that the Promotion of Access to67In particular, it sought this information in fulfilment of its role as a 'watchdog' in respect of the use, control and release of GMOs in South Africa.Its capacity to access and use the information sought therefore held potential benefits for consumers who were not before the court as well as for the environment, which factor accords with the Constitutional Court's description of public interest matter in the Van der Merwe case. 6864 S 32(1) of the Constitution provides that "(1) Everyone has the right of access to-(a) any information held by the state; (b) any information that is held by another person and that is required for the exercise or protection of any rights.amatter of mere observation.It went into great detail to support its submission in that regard.A finding that Biowatch has standing and is acting in the public interest is, in my view, a comprehensive success by Biowatch and a comprehensive loss by the fourth respondent".(Judge's emphasis.)67 The rights protected in s 24 of the Constitution.68 See n 2 above.
into account, although they did not work in Biowatch's favour.However, because he admittedly did not expressly articulate all the grounds relating to his decision on costs, what was or was not in his mind at the time he made the order became a moot point.In this Biowatch relied on an exchange between Dunn AJ and counsel for Monsanto in the proceedings for the application for leave to appeal in respect of an affidavit Biowatch had attempted to place before the court indicating the effects of an adverse costs order on its operations.88Inthis exchange, Dunn AJ asked counsel for Monsanto if one could not take into account the fact that Biowatch was an NGO which needed funding from outsiders, to which counsel responded that the judge could have taken this into account.Dunn AJ then stated: "But that is something that I did not take into account.I must say, you know, at face value I treated the applicant as a normal litigant".And later, he stated: "What concerns me is this, did I not, 84 Ibid par 28.85 Biowatch II (Minority) par 82.86 Quoted in Biowatch II (Minority) par 41. 87 Democratic Alliance v Masondo NO 2003 (2) SA 413 (CC).
90In Poswa J's final opinion, Dunn AJ did not take this National Environmental Management Act 107 of 1998.In Wildlife Society of Southern Africa v Minister of Environmental Affairs and Tourism, 119 a case decided before locus standi in environmental matters was broadened through the right to environment in section 24 of the Constitution, case law that invoked this right as the basis for standing and section 32(1), the Pickering J stated:It was not certain that to afford locus standi to a body such as the [Wildlife Society of Southern Africa] in circumstances such as these would open the floodgates to a torrent of frivolous or vexatious litigation against the State by cranks or busybodies.Neither was it certain, given the exorbitant costs of Supreme Court litigation, that, should the law be so adapted, cranks and busybodies would flood the Courts with vexatious or frivolous applications against the State.Should they be tempted to do so, an appropriate order of costs would soon inhibit their litigious ardour.quotation,one has to wonder if the courts, at least in some quarters, view costs orders as a "weapon" that can be invoked against their oftquoted fear of opening the "floodgates" of litigation. 121ut as Mr Justice Toohey, a member of the High Court of Australia, said during his address to a conference of the National Environmental Law Association in 1989:There is little point in opening the doors to the courts if litigants cannot afford to come in.The general rule in litigation that "costs follow the event" is in point.The fear, if unsuccessful, of having to pay the costs of the other side (often a government instrumentality or wealthy private corporation) with devastating consequences to the individual or environmental group bringing the action, must inhibit the taking of cases to court.In any event, if will be a factor that looms large in any consideration to initiate litigation.Barkhuizen 123 the Constitutional Court reaffirmed its position on the importance and purpose of section 34 of the Constitution, 124 the constitutional right arguably most pertinent to the issue of access to justice.They held that the orderly and fair resolution of disputes by courts (or other independent and impartial tribunals) is fundamental to the stability of an orderly society and provides that "[e]veryone has the right to have any dispute that can be resolved by the application of law decided in a fair public hearing before a court or, where appropriate, another independent and impartial tribunal or forum".
120 Based on this 122 In Napier v Smith 2007 (5)SA 250 (CC).Although the applicants were ultimately successful in upholding their constitutional rights to privacy and dignity, this took place "at great personal cost to them and financial cost to the [Aids Law Project]".
130 Brooke (n 2) 344.131See in this regard alsoHeywood and Hassim (n 127)on the effects of the decision in NM v

137
South African Broadcasting Incorporation v National Director of Public Prosecutions 2007 (1) SA 523 par 32.