My Vote Counts NPC v Speaker of the National Assembly and Others (CCT121/14)  ZACC 31 (30 September 2015)
Download original files
Links to summary
Bookmark/share this page
CONSTITUTIONAL COURT OF SOUTH AFRICA
Case CCT 121/14
In the matter between:
MY VOTE COUNTS NPC Applicant
SPEAKER OF THE NATIONAL ASSEMBLY First Respondent
CHAIRPERSON OF THE NATIONAL COUNCIL
OF PROVINCES Second Respondent
PRESIDENT OF THE REPUBLIC OF SOUTH
AFRICA Third Respondent
DEPUTY PRESIDENT OF THE REPUBLIC OF
SOUTH AFRICA Fourth Respondent
MINISTER OF JUSTICE AND CORRECTIONAL
SERVICES Fifth Respondent
MINISTER OF HOME AFFAIRS Sixth Respondent
AFRICAN NATIONAL CONGRESS Seventh Respondent
DEMOCRATIC ALLIANCE Eighth Respondent
ECONOMIC FREEDOM FIGHTERS Ninth Respondent
INKATHA FREEDOM PARTY Tenth Respondent
NATIONAL FREEDOM PARTY Eleventh Respondent
UNITED DEMOCRATIC MOVEMENT Twelfth Respondent
FREEDOM FRONT PLUS Thirteenth Respondent
CONGRESS OF THE PEOPLE Fourteenth Respondent
AFRICAN CHRISTIAN DEMOCRATIC PARTY Fifteenth Respondent
AFRICAN INDEPENDENT CONGRESS Sixteenth Respondent
AGANG SA Seventeenth Respondent
PAN AFRICANIST CONGRESS Eighteenth Respondent
AFRICAN PEOPLE'S CONVENTION Nineteenth Respondent
Neutral citation: My Vote Counts NPC v Speaker of the National Assembly and Others  ZACC 31
Coram: Mogoeng CJ, Moseneke DCJ, Cameron J, Froneman J, Jappie AJ, Khampepe J, Madlanga J, Molemela AJ, Nkabinde J, Theron AJ and Tshiqi AJ.
Judgments: Cameron J (minority):  to 
Khampepe J, Madlanga J, Nkabinde J, Theron AJ (majority):  to  Order: 
Heard on: 10 February 2015
Decided on: 30 September 2015
Summary: Section 167(4)(e) of the Constitution — exclusive jurisdiction
Section 32(2) of the Constitution — Parliament required to enact national legislation — Parliament has enacted legislation — Promotion of Access to Information Act 2 of 2000 — principle of constitutional subsidiarity applied
Section 32 of the Constitution — right of access to information — private funding of political parties — information required for the exercise or the protection of any right — section 19(3) of the Constitution — right to vote
The application is dismissed.
CAMERON J (Moseneke DCJ, Froneman J and Jappie AJ concurring):
 At issue is whether Parliament has failed to fulfil an obligation the Constitution imposes on it. The specific question is whether information on private funding of political parties is information that is required to exercise the right to vote. If it is, the further question is whether Parliament has passed legislation that gives effect to the right of access to this information. If not, Parliament is in breach of its constitutional obligation, and the applicant asks this Court to require Parliament to remedy the breach.
 The applicant, represented by its director, Mr Axolile Notywala, is My Vote Counts NPC (My Vote Counts). It is a non-profit company campaigning for a more inclusive, transparent and accountable political system in South Africa. It invokes the Court’s exclusive jurisdiction under section 167(4)(e) of the Constitution – or, in the alternative, direct access to this Court – to compel Parliament to pass legislation that obliges political parties to disclose the sources of their private funding. The core of its case is that information about political parties’ private funding is essential to an informed exercise of the right to vote that section 19(3) of the Bill of Rights confers on all citizens. It relies on section 32, “Access to information”, to give proper effect to section 19(3).
 The first and second respondents, the Speaker of the National Assembly and the Chairperson of the National Council of Provinces, representing Parliament, make common cause in opposing the application. They recognise the obligation section 32(2) imposes, but say Parliament has fulfilled it by enacting the Promotion of Access to Information Act (PAIA). They argue that the application should be dismissed.
 The third, fourth, fifth and sixth respondents are the President, Deputy President and the Ministers of Justice and Correctional Services and of Home Affairs. The remaining 12 respondents are all the political parties currently represented in Parliament. Though the Minister of Justice and Correctional Services (fifth respondent) and the Democratic Alliance (eighth respondent) initially filed notices to oppose the application, both were withdrawn. Hence the two houses of Parliament are the sole respondents participating in the proceedings; I refer to them collectively as “Parliament”.
 I have had the benefit of reading the judgment by Khampepe J, Madlanga J, Nkabinde J and Theron AJ (majority judgment). We agree that only this Court has jurisdiction to determine the matter. But beyond that we part. The differences between us concern whether form should prevail over substance when a litigant enforces a constitutional right. More importantly, they concern the extent to which this Court is duty-bound to exercise an adjudicative power the Constitution explicitly confers on it.
 The fundamental difference between the two judgments lies in the distinction between the constitutional process for finding a statute constitutionally invalid and for holding that Parliament has failed to meet a constitutional obligation. This Court has exclusive jurisdiction under section 167(4)(e) of the Constitution to determine whether Parliament has “failed to fulfil a constitutional obligation”. Section 2 of the Constitution requires that all constitutional obligations “must be fulfilled”. It would be wrong, and would impoverish our existing case law, to step back from exercising a power the Constitution imposes on this Court.
 It is correct to emphasise that ordinary challenges to statutory provisions must go through the “usual procedural hoops”. But it does not follow that this Court is precluded from exercising the jurisdiction section 167(4)(e) specifically confers on it. This requires the Court to evaluate the extent to which an obligation has been fulfilled. A proper appreciation of this Court’s task entails a broader embrace of the range of remedies and procedural routes the Constitution affords litigants, and requires this Court to adjudicate.
Previous efforts to secure transparency on private party political funding
 Political parties receive money from public and private sources. The law deals differently with the two types. No legislation requires systematic and proactive disclosure of private funding of political parties. Consequently, political parties are under no express legal obligation to disclose the sources of their private funding, at elections or other times. The applicant seeks a change to that.
 Public funding, by contrast, has already been dealt with in legislation. Section 236 of the Constitution provides that to “enhance multi-party democracy, national legislation must provide for the funding of political parties participating in national and provincial legislatures”. Parliament passed this legislation in 1997 when it enacted the Public Funding of Represented Political Parties Act.
 In doing so, Parliament had also considered the issue of private funding. The Speaker, Ms Baleka Mbete, details this history in her answering affidavit on behalf of Parliament. She describes the question whether political parties’ private funding should be made accessible and, if so, how and to whom, as “a complex policy matter which has been discussed in Parliament since 1997”. She relates that in August 1997, the Promotion of Multi-Party Democracy Bill was introduced.  On 31 October 1997, the Portfolio Committee on Constitutional Affairs reported that the passing of the Bill “represents a very significant step in the ongoing process of consolidating and entrenching a multi-party democracy in South Africa”. The key to the success of our new emerging democracy, it reported, “is the role of strong, resilient, democratically elected political parties”.
 The Bill was to be seen, the Report recorded, “as the first stage of the process”. There are, it stated, “other issues relating to the funding of political parties that will have to be addressed in the near future”. The main one was “the need for public disclosure of the private funding received by political parties, and the form and scope of this disclosure.” On 27 November 1997, Parliament adopted the Bill. It came into force on 1 April 1998. The Speaker emphasised that the Act was not intended to “deal with all questions that may arise in regard to the funding of political parties”. Hence it remained for Parliament “as a follow up” to consider “whether or not there is a need to regulate other aspects of political party funding” – including the disclosure of parties’ private funding.
 The issue of private funding arose again later – in litigation rather than Parliamentary discussions. In 2003, the Institute for Democracy in South Africa (Institute) requested the five political parties with the largest representation in Parliament to disclose records of donations they had received in the run-up to the 2004 general elections. Save for the African Christian Democratic Party, all refused. The Institute then applied to the then Cape Provincial Division of the High Court (High Court) for an order declaring that PAIA and section 32(1) of the Constitution obliged political parties to disclose the requested records. The Institute contended that it enjoyed an unqualified right to access the records of their donations on the basis that they were public bodies under section 11 of PAIA. Alternatively, if they were private bodies, the Institute sought the information under section 50 of PAIA read with sections 19(1) and 19(2) of the Constitution.
 All the political parties cited in IDASA initially resisted. While none accepted that they could be characterised as “public bodies” under PAIA, they supported public debate on the question and took the view that the regulation of private funding of political parties would be best achieved through legislation, rather than piecemeal litigation. The governing party, the African National Congress (ANC), through its deponent, then Secretary-General Kgalema Motlanthe, sought the dismissal of the application or a stay of the proceedings. He said this would “allow the political and legislative process to follow the proper course necessary for the adoption of a national policy through legislation regulating the funding of political parties”.
 On 20 April 2005, the High Court dismissed the application. Griesel J held that, under PAIA, political parties are private bodies for purposes of their “fundraising activities”. Hence the Institute had to link the donation records it sought to the exercise or protection of a right, in particular to section 19(1) and (2) of the Constitution. The Institute had not adequately explained how and why the donations records would assist them in exercising those rights. There was no appeal.
 Before the IDASA litigation, the United Nations (UN) General Assembly adopted the UN Convention against Corruption on 31 October 2003. Article 7(3) requires each State Party to consider taking appropriate legislative and administrative measures “to enhance transparency in the funding of candidatures for elected public office and, where applicable, the funding of political parties”. The African Union Convention on Preventing and Combating Corruption is more specific. It was adopted at a session of the African Union in Maputo, Mozambique, on 11 July 2003. Unlike the UN Convention, it contains a provision separately and expressly addressing the funding of political parties.  And, while the UN Convention requires state parties to “consider” certain measures, the AU Convention uses imperative language. Parliament ratified the UN Convention, without material reservation, on 22 November 2004. Parliament ratified the AU Convention, again without material reservation, on 11 November 2005, after the High Court dismissed the Institute’s application.
 The proposal for the regulation of private funding to political parties lay dormant until 8 November 2012 when the applicant wrote to Parliament. It claimed that—
“appropriate legislation ensuring transparency and accountability in the funding of political parties is a constitutional imperative, as required by sections 1, 7, 32, 33 and 195 of the Constitution . . . [and] that each of these provisions imposes a specific obligation on Parliament to enact national legislation to give effect to these duties, rights and principles.”
 Parliament took the stance that those provisions of the Constitution do not create justiciable rights, but are general obligations. On 10 December 2012, Parliament responded to the letter of 8 November 2012. It stated that it had given effect to its obligation in section 236 of the Constitution by enacting the Public Funding of Represented Political Parties Act, and that the matter of private funding had been referred to the Chief Whips’ Forum in Parliament. After a six-month silence, the applicant’s attorneys wrote to Parliament requesting a timetable for the parliamentary process for passing the legislation, which the letter called “a constitutional imperative”. In the alternative, the applicant asked for reasons justifying Parliament’s decision not to enact the legislation.
 Now, the Speaker took the view that the enactment of legislation regulating the private funding of political parties was “a party-political matter” and that the Speaker and the Chairperson of the National Council of Provinces do not “play a role in the initiation of such legislation”. The Speaker suggested that the applicant take the matter up with the Executive or any member of Parliament. The Speaker also records that the proposed disclosure legislation was deemed “not feasible” and was “not to be proceeded with”. She declined to make any undertakings to enact the legislation.
Unique nature of this application
 This application differs from that dismissed in IDASA. Far from requesting political parties to grant access to private funding records under PAIA, the applicant says the problem is precisely that PAIA does not require disclosure of party political funding. Since the relief the applicant seeks is not contemplated at all in PAIA, this Court is called upon to interpret the ambit of the section 32(1) right and the extent to which Parliament has fulfilled its obligation under section 32(2). As in IDASA, the applicant claims that the information is required for the exercise and protection of the rights in section 19 of the Bill of Rights. But it relies more specifically on the right to vote in section 19(3). It does not seek ad hoc information from any or each political party. Rather, it seeks an order requiring Parliament to enact national legislation regulating the disclosure of private funding records as a matter of continuous course, rather than once-off upon request.
 The first question is jurisdiction. The applicant seeks to bring directly before this Court its assertion that Parliament has failed to fulfil a constitutional obligation by not passing legislation the Constitution obliges it to enact in terms of section 32(2). Does this Court have competence under section 167(4)(e) of the Constitution to consider the claim? On 30 September 2014, the Chief Justice issued directions inviting written argument on this. In response, both the applicant and Parliament submitted that the Court has exclusive jurisdiction to determine the claim.
 The applicant’s approach reiterated the core components of its case. It submitted that section 32(2) of the Constitution imposes an obligation on Parliament to enact legislation that provides for access to information pertaining to the private funding held by political parties that is required for the right to vote. Parliament has not enacted this legislation. It has failed to fulfil a constitutional obligation. Since the validity of existing legislation is not challenged, lower courts do not have jurisdiction. This Court’s exclusive jurisdiction under section 167(4)(e) is engaged.
 Parliament’s response reached the same conclusion by a sparser route. It pointed out that jurisdiction is not determined by the merits of a claim, by whether it must succeed or not, but by how the claimant pleads it. The pleadings contain the legal basis of the claim under which the applicant seeks to invoke the Court’s competence. As the applicant’s claim is based solely on the averment that Parliament has failed to fulfil a constitutional obligation, the application falls within this Court’s exclusive jurisdiction.
 The parties are right (and the majority judgment agrees). This Court’s exclusive jurisdiction is engaged. But, exclusive jurisdiction is too important to be resolved by concession, as here, by consensus.  The Court’s competence, which springs from the sensitive political nature of the separation of powers, must be scrutinised. Previous decisions establish that, despite their broad wording, the exclusive jurisdiction provisions must be narrowly construed. This is because a broad construction of exclusive jurisdiction under section 167(4)(e) may “negate or improperly attenuate the jurisdiction of the Supreme Court of Appeal and the High Court”.
 An over-broad interpretation of exclusive jurisdiction would obviate the need for section 172. So for harmonious interpretation we ought to make “a clear distinction between law and conduct on the one hand and obligations on the other”. More pertinently, we have held that the jurisdictional competence conferred by the words “fulfil a constitutional obligation” must be narrowly read, both in relation to the President and to Parliament.
 In Women’s Legal Centre Trust, the applicant asserted that the President and Parliament had failed to fulfil an obligation the Constitution imposed on them by failing to prepare, initiate, enact and implement a statute providing for the recognition of all Muslim marriages. The Court held that, if there was a constitutional duty to enact the legislation the applicant sought, it was one the Bill of Rights required the state and its organs – including the national Executive, Chapter 9 institutions, Parliament and the President – to perform collaboratively or jointly. The obligation did not fall within the ambit of section 167(4)(e). The provision envisages only constitutional obligations imposed specifically and exclusively on the President and on Parliament, and on them alone. It does not embrace the President when he or she acts as part of the national Executive, nor Parliament when it is required not to act alone, but as part of other constituent elements of the state.
 The pleaded claim here rests on the specific obligation created by section 32(2) only. This requires that “national legislation must be enacted” to give effect to the right of access to information. This wording contrasts with the language the Bill of Rights uses elsewhere to impose duties. There, “the state” is required to fulfil a range of constitutional obligations, either by passing legislation or by other means. More tellingly, the Bill of Rights requires “the state” to take reasonable legislative and other measures to fulfil a range of social and economic rights. These include the duty to foster conditions enabling access to land, as well as to achieve the progressive realisation of the rights to adequate housing, to health care services, sufficient food and water and social security. In addition, the Bill of Rights requires the state to take “reasonable measures” to make further education progressively available and accessible.
 These formulations contrast with four other provisions of the Bill of Rights. Section 9(4), section 32(2) and section 33(3) specify that “national legislation must be enacted” in relation to a particular right. This formulation is akin to that of the fourth, section 25(9), which provides that “Parliament must enact the legislation referred to in subsection (6)”. The formulation of these provisions contrasts with that of those requiring “the state” to take certain actions, or to realise rights through legislative and other measures.
 These four provisions are distinct, in two ways, from those rights requiring progressive realisation through a range of unspecified measures that include legislation. First, they all require the enactment of legislation as an express minimum, although of course the Constitution does not preclude other measures that enhance access to and enjoyment of these rights. Second, the Bill of Rights specifically identifies the legislation to be enacted. While section 32(2) and section 33(3) are cast in passive grammatical form – unlike section 25(9), they do not specify the agent that must enact the legislation – this carries little moment because the obligation is to enact “national legislation”. The enacting agent is necessarily Parliament, which the Constitution makes the sole repository of national legislative power. The grammatical form does not detract from the responsibility placed solely on Parliament to fulfil the obligation.
 In fulfilling the obligations sections 9(4), 25(9), 32(2) and 33(3) create, Parliament will of necessity enlist the participation and assistance of other state organs and institutions that are obliged to fulfil the rights in the Bill of Rights. But that does not diminish the sole responsibility the Bill of Rights places on it. It follows that the applicant’s claim under section 32(2) implicates an obligation on Parliament alone, and engages the exclusive jurisdiction of this Court.
 It is apparent from the provision that “national legislation must be enacted” that section 32(2) creates an obligation. Identifying Parliament as the sole bearer of the constitutional obligation means this Court has exclusive jurisdiction. But the question of the interrelation between section 167(4)(e), which grants jurisdiction to this Court alone, and section 172(1)(a), which empowers the Supreme Court of Appeal and the High Court, subject to this Court’s confirmation, to make orders “concerning the validity of an Act of Parliament”, remains.
Is information on political parties’ private funding “required” for the exercise and protection of the right to vote?
 The foundation of the applicant’s case is that the right to vote requires, for its exercise, access to information about political parties’ private sources of funding. Is this so? “Required” in the context of section 32(1)(b) does not denote absolute necessity. It means “reasonably required”. The person seeking access to the information must establish a substantial advantage or element of need. The standard is accommodating, flexible and in its application fact-bound. The section 19(3) right to vote is among the rights contemplated by section 32(1)(b). So the question is whether information about political parties’ private funding is reasonably required for citizens to be able to exercise their right to vote.
 The founding premise of the applicant’s argument is the unique role of political parties in our constitutional democracy. This is difficult to dispute. The electoral system the Constitution creates pivots on political parties and whom they admit as members. In the First Certification judgment, this Court noted that, “[u]nder a list system of proportional representation, it is parties that the electorate votes for, and parties which must be accountable to the electorate.”
 Our constitutional order places the key to elective office and executive power in the hands of political parties. Members of the National Assembly and provincial legislatures are not directly elected. Nor is the President or the Deputy President. The same applies to provincial and national executives. Under the current electoral system, it is political parties, and parties alone, that determine which persons are allocated to legislative bodies and to the executive. If you cease to be a member of the party that nominated you, you lose your membership of that legislature. The President is in turn elected from amongst the members of the National Assembly and the President appoints the Deputy President and the members of the Cabinet bar a maximum of two, from among the members of the legislature.
 These compelling considerations led this Court in Ramakatsa to highlight the centrality of political parties. The judgment’s key findings are that they are “the veritable vehicles the Constitution has chosen for facilitating and entrenching democracy”, and that they are the “indispensible conduits for the enjoyment of the right given by section 19(3)(a) to vote in elections”. The joint majority judgment of Moseneke DCJ and Jafta J noted:
“In the main, elections are contested by political parties. It is these parties which determine lists of candidates who get elected to legislative bodies. Even the number of seats in the National Assembly and provincial legislatures are determined ‘[b]y taking into account available scientifically based data and representations by interested parties.’” (Footnotes omitted.)
 The Court explained:
“Our democracy is founded on a multi-party system of government. Unlike the past electoral system that was based on geographic voting constituencies, the present electoral system for electing members of the national assembly and of the provincial legislatures must ‘result, in general, in proportional representation’. This means a person who intends to vote in national or provincial elections must vote for a political party registered for the purpose of contesting the elections and not for a candidate. It is the registered party that nominates candidates for the election on regional and national party lists. The Constitution itself obliges every citizen to exercise the franchise through a political party.” (Footnotes omitted.)
 Crucially, Ramakatsa’s reasoning elucidates the link between the democratic role of political parties and their funding. Participation in parties’ activities, the judgment explains, is critical to social progress, through the policies they adopt and put forward to address problems facing communities. And it is to enhance multi-party democracy that the Constitution enjoins Parliament to enact national legislation providing for funding of political parties represented in national and provincial legislatures:
“Public resources are directed at political parties for the very reason that they are the veritable vehicles the Constitution has chosen for facilitating and entrenching democracy.”
 Ramakatsa’s reasoning on the public funding of political parties applies pointedly to the question whether information about parties’ private funding is required for the right to vote. Political parties receive public resources because they are the vehicles for facilitating and entrenching democracy. This entails a corollary: that the private funds they receive necessarily also have a distinctly public purpose, the enhancement and entrenchment of democracy, as well as a public effect on whether democracy is indeed enhanced and entrenched. The flow of funds to political parties, public or private, is inextricably tied to their pivotal role in our country’s democratic functioning. There is a further corollary: given parties’ emphatically public role, any notion of privacy attaching to their private funding must be significantly attenuated.
 The applicant submitted that the right to vote is a right to cast an informed vote. This must be correct. The reason was stated by Ngcobo CJ, on behalf of a unanimous Court, in M & G Media Ltd:
“In a democratic society such as our own, the effective exercise of the right to vote also depends on the right of access to information. For without access to information, the ability of citizens to make responsible political decisions and participate meaningfully in public life is undermined.” (Footnote omitted.)
 Section 19(1) of the Constitution envisages that every citizen is “free to make political choices”. This includes forming a political party, participating in a political party’s activities, and campaigning for a political party or cause. It also includes, of course, the freedom to choose one’s leaders. But that choice, like all others, is valuable only if one knows what one is choosing. It loses its value if it is based on insufficient information or misinformation. This the Constitution recognises by insisting that government is not only democratic but openly accessible. That is why its Preamble speaks of a “democratic and open” society; why its fundamental rights are to be interpreted to promote the values underlying an “open and democratic” society, and limited only on that same basis; and why the founding values of universal suffrage and democratic elections are tied to “openness” of government.
 The Bill of Rights also confers the right to freedom of expression. This Court has held that this right is what “makes [the right to vote] meaningful”: only if information is freely imparted, and citizens are kept informed, are their choices genuine. As Mogoeng CJ has also noted on behalf of the Court, “the public can only properly hold their elected representatives accountable if they are sufficiently informed of the relative merits” of the issues at stake. The same is necessarily true when the public decides which representatives to elect by exercising the right to vote.
 So the right to vote does not exist in a vacuum. Nor does it consist merely of the entitlement to make a cross upon a ballot paper. It is neither meagre nor formalistic. It is a rich right – one to vote knowingly for a party and its principles and programmes. It is a right to vote for a political party, knowing how it will contribute to our constitutional democracy and the attainment of our constitutional goals.
 Does this include knowing the private sources of political parties’ funding? It surely does. Private contributions to a political party are not made thoughtlessly, or without motive. They are made in the anticipation that the party will advance a particular social interest, policy or viewpoint. And political parties, in turn, depend on contributors for the very resources that allow them to conduct their democratic activities. Those resources keep flowing to the extent that they meet their contributors’ and funders’ expectations. There can be little doubt, then, that the identity of those contributors, and what they contribute, provides important information about the parties’ likely behaviour. As the United States Supreme Court explained in Buckley v Valeo, disclosure of political funding—
“provides the electorate with information ‘as to where political campaign money comes from and how it is spent by the candidate’ in order to aid the voters in evaluating those who seek federal office. It allows the voters to place each candidate in the political spectrum more precisely than is often possible solely on the basis of party labels and campaign speeches. The sources of a candidate’s financial support also alert the voter to the interests to which a candidate is most likely to be responsive and thus facilitate predictions of future performance in office.
Second, disclosure requirements deter actual corruption and avoid the appearance of corruption by exposing large contributions and expenditures to the light of publicity. This exposure may discourage those who would use money for improper purposes either before or after the election. A public armed with information about a candidate's most generous supporters is better able to detect any post-election special favours that may be given in return.” (Footnotes omitted.)
 For the reasons Ramakatsa sets out, the first two considerations noted in Buckley v Valeo have particular edge in our democracy. This is because political parties hold the key to elective and executive office. They are the indispensable conduits through which the Constitution’s vision of our democratic functioning is to be attained. It follows that information about political parties’ private funding is required for the exercise of the right to vote.
 The applicant claims that PAIA does not confer the right of access to information about political parties’ private funding to which the Constitution entitles voters. Since the Constitution obliges Parliament to create that right of access, the applicant argues, this Court has the power to, and should, order Parliament to do so. Parliament’s response is that this approach is wrong-directional. The correct starting point is not the Constitution, but PAIA, since Parliament enacted it expressly to give effect to the constitutional obligation in section 32(2). The result, Parliament contends, is that the applicant must first seek the right of access it asserts in PAIA.
 Parliament argues that PAIA in fact confers that right – in which case, there is no breach of its constitutional obligation. But, if PAIA doesn’t, Parliament says the applicant’s remedy is to challenge the constitutionality of PAIA in the High Court. It may not circumvent PAIA by relying directly on the constitutional provision the legislation seeks to embody. So the applicant must start again in the High Court. Parliament says the applicant finds itself in a logical trap: whether it is right or wrong about PAIA, the application must be dismissed.
 Parliament’s argument brings to the fore the principle of subsidiarity in our constitutional law. Subsidiarity denotes a hierarchical ordering of institutions, of norms, of principles, or of remedies, and signifies that the central institution, or higher norm, should be invoked only where the more local institution, or concrete norm, or detailed principle or remedy, does not avail. The word has been given a range of meanings in our constitutional law. It is useful in considering the scope of subsidiarity, and Parliament’s reliance on it – to have them all in mind.
 “Subsidiarity” has been used, in assessing the constitutional validity of a statutory provision licensing the use of reasonably necessary force in effecting an arrest, to indicate the necessity for tempering the amount of force. Force is permitted only where there are no lesser means of achieving the arrest. Using force is, in other words, subsidiary to all other means.
 In international law, subsidiarity is employed to resolve a clash of jurisdictions. It determines which state should act when multiple states have jurisdiction over the same events constituting an international crime. Under our Constitution, it signifies that the duty of the South African Police Service to investigate international crimes, including crimes against humanity, is subsidiary to that of the foreign state in which the crimes were committed.
 “Subsidiarity” has also been used to describe the principle that overlap in functional areas of concurrent constitutional competence should be resolved by assigning the power to the sphere of government where the specific function is most appropriate. Within the Bill of Rights, subsidiarity entails that where the Constitution contains both a specific right, like the right of access to housing, and a more general right, like the right to human dignity, which informs the right to housing, the litigant must first invoke the specific right. The more general right is subsidiary.
 But the most frequent invocation of subsidiarity has been to describe the principle that limits the way in which litigants may invoke the Constitution to secure enforcement of a right. Under the interim Constitution, where the Appellate Division had no constitutional jurisdiction, and this Court had constitutional jurisdiction only, this Court laid down as a general principle that, where it was possible to decide a case, civil or criminal, without reaching a constitutional issue, that should be done. This entailed the subsidiarity of the interim Constitution to other judicial approaches to rights enforcement.
 Of course, this approach has long since been abandoned under the final Constitution in favour of its opposite, namely the primacy of constitutional approaches to rights determination. Far from avoiding constitutional issues whenever possible, the Court has emphasised that virtually all issues – including the interpretation and application of legislation and the development and application of the common law – are, ultimately, constitutional. This is because the Constitution’s rights and values give shape and colour to all law.
 But it does not follow that resort to constitutional rights and values may be freewheeling or haphazard. The Constitution is primary, but its influence is mostly indirect. It is perceived through its effects on the legislation and the common law – to which one must look first.
 These considerations yield the norm that a litigant cannot directly invoke the Constitution to extract a right he or she seeks to enforce without first relying on, or attacking the constitutionality of, legislation enacted to give effect to that right. This is the form of constitutional subsidiarity Parliament invokes here. Once legislation to fulfil a constitutional right exists, the Constitution’s embodiment of that right is no longer the prime mechanism for its enforcement. The legislation is primary. The right in the Constitution plays only a subsidiary or supporting role.
 Over the past 10 years, this Court has often affirmed this. It has done so in a range of cases. First, in cases involving social and economic rights, which the Bill of Rights obliges the state to take reasonable legislative and other measures, within its available resources, to progressively realise, the Court has emphasised the need for litigants to premise their claims on, or challenge, legislation Parliament has enacted. In Mazibuko, the right to have access to sufficient water guaranteed by section 27(1)(b) was in issue. The applicant sought a declaration that a local authority’s water policy was unreasonable. But it did so without challenging a regulation, issued in terms of the Water Services Act, that specified a minimum standard for basic water supply services. This, the Court said, raised “the difficult question of the principle of constitutional subsidiarity”. O’Regan J, on behalf of the Court, pointed out that the Court had repeatedly held “that where legislation has been enacted to give effect to a right, a litigant should rely on that legislation in order to give effect to the right or alternatively challenge the legislation as being inconsistent with the Constitution”. The litigant could not invoke the constitutional entitlement to access to water without attacking the regulation and, if necessary, the statute.
 Second, the Court has applied the principle to legislation Parliament adopts with the clear design of codifying a right afforded by the Bill of Rights. After Parliament enacted the Labour Relations Act (LRA), the High Court in NAPTOSA refused to allow a litigant to rely directly on the fair labour practices provision in the Bill of Rights. It had to rely instead on the unfair labour practice provisions in the statute, or challenge the statute itself. Conradie J said he could not “conceive that it is permissible for an applicant, save by attacking the constitutionality of the LRA, to go beyond the regulatory framework which it establishes”. He also stated that it was inappropriate, in a highly regulated statutory environment like labour law, to ask a court to fashion a remedy “which the legislature has not seen fit to provide”.
 This approach was first quoted with approval in this Court in a context unrelated to employment rights, then adopted and endorsed unanimously in a case about labour relations, SANDU. Even though national regulations had been enacted providing for collective bargaining, the applicant sought to rely directly on the provisions of section 23(5) of the Bill of Rights to found a more encompassing duty to bargain. The Court disallowed this. It held that where legislation has been enacted to give effect to a constitutional right, “a litigant may not bypass that legislation and rely directly on the Constitution without challenging that legislation as falling short of the constitutional standard”. If the legislation is wanting in its protection of the right, then that legislation “should be challenged constitutionally”.
 Third, the Court has applied the principle of subsidiarity to those provisions of the Bill of Rights that specifically oblige Parliament to enact legislation: sections 9(4), 25(9), 33(3), and 32(2) – the lattermost section at issue in this case. The Court has held that unfair discrimination cases must be brought “within the four corners” of the Promotion of Equality and Prevention of Unfair Discrimination Act, rather than under the Bill of Rights. In Pillay, Langa CJ, on behalf of the majority, citing New Clicks, SANDU and NAPTOSA, held that “a litigant cannot circumvent legislation enacted to give effect to a constitutional right by attempting to rely directly on the constitutional right”.
 In Bato Star, the application of the Promotion of Administrative Justice Act was at issue. Neither the High Court nor the Supreme Court of Appeal considered the applicant’s claim to administrative review in the context of PAJA. This Court held that they had erred. The Court held that “[t]he provisions of section 6 divulge a clear purpose to codify the grounds of judicial review of administrative action as defined in PAJA”. The cause of action for the judicial review of administrative action now ordinarily arises from PAJA, not from the common law as in the past. And the authority of PAJA to ground such causes of action rests squarely on the Constitution.
 In New Clicks, the applicability of PAJA was also at issue, though the Court was divided on whether it applied to the regulations in issue. Chaskalson CJ affirmed that a litigant “cannot avoid the provisions of PAJA by going behind it, and seeking to rely on section 33(1) of the Constitution or the common law”. Ngcobo J, expressly endorsing the High Court’s approach in NAPTOSA, said that our Constitution “contemplates a single system of law which is shaped by the Constitution. To rely directly on section 33(1) of the Constitution and on common law when PAJA, which was enacted to give effect to section 33 is applicable, is in my view inappropriate”. He proceeded:
“Where, as here, the Constitution requires Parliament to enact legislation to give effect to the constitutional rights guaranteed in the Constitution, and Parliament enacts such legislation, it will ordinarily be impermissible for a litigant to found a cause of action directly on the Constitution without alleging that the statute in question is deficient in the remedies that it provides.”
 In PFE International, the “heart of the matter” was “the determination of the legislative regime regulating the exercise of the right of access to information held by the state after the commencement of legal proceedings”. Jafta J, on behalf of a unanimous Court, said—
“PAIA is the national legislation contemplated in section 32(2) of the Constitution. In accordance with the obligation imposed by this provision, PAIA was enacted to give effect to the right of access to information, regardless of whether that information is in the hands of a public body or a private person. Ordinarily, and according to the principle of constitutional subsidiarity, claims for enforcing the right of access to information must be based on PAIA.”
 These instances explain the powerful, inter-related reasons from which the notion of subsidiarity springs. The principle is concerned in the first place with the programmatic scheme and significance of the Constitution. In New Clicks, Chaskalson CJ said that allowing a litigant to rely on section 33(1) of the Constitution, rather than on PAJA, “would defeat the purpose of the Constitution in requiring the rights contained in section 33 to be given effect by means of national legislation”.
 A second concern is Parliament’s indispensable role in fulfilling constitutional rights. Ngcobo J in New Clicks pointed out that “[l]egislation enacted by Parliament to give effect to a constitutional right ought not to be ignored”. The Constitution’s delegation of tasks to the legislature must be respected, and comity between the arms of government requires respect for a cooperative partnership between the various institutions and arms tasked with fulfilling constitutional rights. As this Court has said, “the courts and the legislature act in partnership to give life to constitutional rights”. The respective duties of the various partners and their associates must be valued and respected if the partnership is to thrive. In SANDU, the Court pointed out that not to apply the principle “would be to fail to recognise the important task conferred on the legislature by the Constitution to respect, protect, promote and fulfil the rights in the Bill of Rights”.
 A third interest the principle protects is the development of a consistent and integrated rights jurisprudence. Our Courts have held that allowing reliance directly on constitutional rights, in defiance of their statutory embodiment, would encourage the development of “two parallel systems” of law. In other words, coherence in developing and applying rights within a unitary system of norms is a further reason for requiring litigants to rely on, or challenge, legislation that gives effect to a provision in the Bill of Rights.
 This approach prevailed in IDASA. There, the applicant sought to rely directly on section 32 of the Constitution, but failed to challenge PAIA. The High Court held that it could not proceed in that way. It found that section 32 was “subsumed” by PAIA, which regulates the right of access to information. Hence, in the absence of a challenge to the constitutional validity of PAIA, the provision in the Constitution could not serve as an independent legal basis or cause of action to enforce rights of access to information. The applicants accordingly had to seek their remedy “within the four corners” of the statute, for to hold otherwise would encourage the development of two systems of law.
 Parliament contends that the approach in IDASA is correct. The applicant cannot invoke section 32 for its claim. It must challenge the constitutionality of PAIA first. Otherwise, Parliament says, the application thwarts the principle of subsidiarity. The application seeks direct resort to the Constitution, even in the face of legislation that is designed to give effect to a fundamental right which the application ignores or subverts.
 The test the Court must apply, Parliament says, is to ask whether PAIA was designed or purports to give effect to the right of access to information in section 32(1). The question is not whether PAIA in fact gives proper effect to that provision. It is precisely when legislation purports to give effect to a right, but fails to do so properly, that subsidiarity requires a constitutional challenge to the deficient legislation. Parliament did not enact PAIA in mere partial fulfilment of the obligation in section 32(2). The statute purports to fulfil the obligation completely. It “covers the field”. Here, Parliament acknowledges, of course, that the applicant’s complaint is precisely that PAIA does not provide the remedy it claims should exist. But, Parliament contends, because PAIA purports to cover the field, subsidiarity prescribes that the applicant must go to the High Court first to challenge PAIA’s constitutional validity in the ordinary way.
Subsidiarity does not apply
 The majority judgment contends that the principle of subsidiarity applies and that the application should be dismissed. I do not agree. Subsidiarity does not apply for a potent reason: the validity of legislation is not at issue. The question is not whether PAIA is valid legislation, but whether Parliament has adequately fulfilled its section 32 obligation. This includes not only PAIA, but any and all of the range of legislation it has enacted in fulfilment of section 32. The applicant says Parliament has not done enough. So this Court’s job is to determine the extent to which Parliament has met its obligation to enact legislation that gives effect to section 19(3) read with section 32(2). That clearly requires an assessment of the reach of existing legislation, though not its validity.
 Parliament’s argument is mistaken. It misconceives the nature of the applicant’s challenge to PAIA. Subsidiarity is inapplicable because PAIA’s constitutional validity is not in question. The principle of subsidiarity does not assist Parliament, for it simply has no application:
(a) Subsidiary applies only when a statute’s validity is at stake. Here, validity is not at stake.
(b) PAIA is not circumvented because there is no attempt to bypass the provisions of PAIA by invoking the Constitution.
 The principle provides that one may not rely directly on the Constitution in the face of legislation designed to give effect to it; one must treat the Constitution as subsidiary to the legislation. But the crucial point is that the principle operates only if the legislation is not under constitutional attack. This Court has already noted, in Doctors for Life, that validity of legislation can only be impugned in two circumstances: when the content or substance of the legislation does not comply with the Constitution, or because there is a procedural defect in its enactment. By contrast, when a litigant does attack the legislation, as here, saying that it falls short of a standard embodied in the Constitution itself, then they are free to invoke the Constitution directly. That, indeed, is the essence of constitutionalism: it allows all legislation to be subjected to constitutional scrutiny. So a litigant may invoke the Constitution to gauge the extent to which legislation meets a constitutional obligation – but the litigant may not evade addressing that legislation.
 The principle of subsidiarity puts litigants to a choice. It says that, “where legislation has been enacted to give effect to a right, a litigant should rely on that legislation in order to give effect to the right, or, alternatively challenge the legislation as being inconsistent with the Constitution”. But where the legislation is challenged for not meeting a constitutional obligation, the principle does not apply.
 This reflects the principle’s rationale, which is the cooperation the courts, under the separation of powers, owe a fellow actor that is striving to give life to constitutional obligations. Because the courts act in partnership with Parliament in fulfilling the Bill of Rights, comity between the arms of government requires, if the relationship is to be successful, a measure of respect for what the other partner is trying to achieve. Parliament’s role in operationalising constitutional rights by enacting legislation must be respected – and the courts should not, therefore, allow litigants to “circumvent” or “bypass” that legislation.
 But these considerations do not apply where the litigant relies on the restricted ambit of the legislation. Where the litigant does this, there can be no complaint that Parliament’s legislation is being ignored. On the contrary, Parliament is afforded a full and formal opportunity to defend its fulfilment of its legislative obligation, and to show that it has done all that the Constitution requires. The decisions show that in every case where this Court has applied the subsidiarity principle, the litigant has entirely omitted or failed to challenge the constitutionality of legislation enacted to fulfil the right the litigant seeks to enforce by invoking the Constitution directly. It is this that subsidiarity precludes.
 Here, by contrast with IDASA, and in contradistinction to every case where this Court has applied subsidiarity, the applicant says the ambit and purport of PAIA is insufficient. It is not seeking to evade or circumvent PAIA. It is not ignoring PAIA. It is confronting it. Its central contention, in its affidavits and arguments, is that PAIA does not reach the right, because Parliament, in breach of an obligation, has failed to enact legislation embodying the right of access to information it seeks to enforce, namely information about political parties’ private funding. This afforded Parliament the opportunity to defend its legislation – which it fully did. The Speaker’s opposing affidavit strenuously contends for the constitutional adequacy of PAIA. It claims from the outset that PAIA “fully satisfies the requirements of section 32(2)”. Thereafter, Parliament’s response seeks to substantiate this claim in detail.
 So the constitutional reach of PAIA has always been squarely on the table. Far from bypassing or ignoring the legislation, the applicant has confronted it head-on, and invoked the Constitution only as a means to show that PAIA’s reach falls short of fulfilling the obligations of Parliament under section 32(2). The principle of constitutional subsidiarity finds no application here. Subsidiarity cannot be a barrier to a challenge of the kind the applicant brings when its complaint is precisely that there is simply no appropriate legislative regulation. Simply put, subsidiarity does not work in a vacuum. Parliament’s invocation of subsidiarity is therefore fundamentally misconceived.
The application is not a constitutional challenge under section 172
 Parliament contended, and the majority judgment agrees, that the applicant must be sent to the High Court to start again. This misconceives the application. It seeks to recast the application as a constitutional challenge to PAIA under section 172(2)(a), over which the High Court has jurisdiction. That sets up a misplaced procedural bar to the applicant’s claim. In fact, the application is brought under section 167(4)(e) of the Constitution; this Court alone has jurisdiction.
 The distinction is subtle but fundamental. In Doctors for Life, this Court recognised and underscored the difference between a constitutionally invalid statute and an unmet constitutional obligation.
 The applicant does not challenge a statute under section 172(2)(a) on the basis that its provisions are in conflict with the Bill of Rights or because it was adopted in a manner inconsistent with the Constitution. The section 172 route affords procedural safeguards for determining the validity of legislation. These include allowing the member of the Executive responsible for implementing the legislation to justify any limitation under section 36.
 But section 36 applies only where a right in the Bill of Rights is limited by a law of general application. The applicant’s complaint is that there is no legislation, nor legislative provision, that requires political parties to disclose the sources of their private funding. It is in precisely this absence of legislation that the applicant locates Parliament’s failure to give effect to a right in the Bill of Rights. Had Parliament in fact enacted legislation that required disclosure of political parties’ private funding, and had the applicant’s complaint been that the provisions were under or over-broad, or in some other way deficient, the courts’ inquiry would indeed fall within section 172.
 The applicant chose deliberately not to proceed via section 172(2)(a) because that route was not suited to its case. The question the application raises is whether this Court has jurisdiction. If the Court does, it is not for this Court to bar a litigant from a pathway the Constitution provides to it.
 Moreover, the applicant clearly affirmed that it does not seek an order declaring that PAIA or any of its provisions are invalid. Parliament was never called upon to meet that case. What is before us is the content and scope of Parliament’s obligation. That is what the applicant pleaded, and what Parliament accepted. That poses a different question. And Parliament sought to answer by invoking PAIA. It said, Yes, we agree we have this obligation, but we have fulfilled it: look at PAIA. So we must look at PAIA, though only for the purpose of assessing the extent of Parliament’s constitutional obligation and its fulfilment.
 The majority judgment concludes that, because of subsidiarity, this Court is precluded from evaluating the extent to which Parliament has met the obligation. But this runs against the express powers the Constitution confers under section 167(4)(e). The Constitution deliberately affords litigants both options, though the section 167(4)(e) route is available only in the limited instance where the Constitution prescribes that a constitutional obligation must be fulfilled.
 The two options should not be conflated. Nor should either be squeezed out. To shut down the route the applicant has chosen to enforce its right to information risks impoverishing the Constitution and this Court’s jurisdiction to interpret it. Our decisions on subsidiarity have not had to address this distinction, since, in each of them, the legislation at issue was not challenged at all. This was because the litigant sought to derive a right directly from the Constitution without addressing the extent to which the legislation that applied in fact provided for the claimed entitlement.
 It is true that the applicant contends that PAIA alone is insufficient to fulfil Parliament’s section 32 obligation. But it does not follow that the application is simply an application for the judicial review of legislation. The applicant has no complaint about PAIA on that statute’s own terms. It does not demand that the statute be amended (though Parliament, of course, will be free to amend PAIA in response to the applicant’s challenge). This it emphasised in its heads of argument:
“The applicant thus raises no constitutional challenge against PAIA, and seeks no reading in, reading down or striking down of any of its provisions.”
 The question whether Parliament has fulfilled the section 32(2) obligation is not contingent upon the validity of PAIA. Nor is Democratic Party on point, for that case did not pivot on the content of a constitutional obligation, nor was it brought within this Court’s exclusive jurisdiction. Rather, it concerned the validity of a statute and was thus plainly a section 172 challenge.  A finding that Parliament has failed to fulfil an obligation does not impact on the constitutional validity of PAIA. That is a different question. The question is only whether Parliament has fulfilled an obligation the Constitution obliges it to fulfil. Once it chose this way of formulating its case, the only route available to it lies through this Court, invoking section 32(2) directly.
 The applicant’s position was, indeed, that PAIA pursues a “constitutional imperative”; and that its provisions, and the requirements and processes that they embody, “are logical and legitimate for PAIA to serve its purpose”. And rightly so. Everything in PAIA is, in the absence of a challenge to its validity, consistent with section 32. Indeed, PAIA does not stand alone here. It is like many other legislative provisions Parliament has enacted in fulfilment of section 32. It, together with the other legislation, constitutes an indispensable measure to fulfil the provision’s promise. The applicant’s point is that PAIA is not all that section 32 requires; it fails to exhaust the obligation the provision creates. Other legislation is needed too. PAIA is constitutionally necessary, but not sufficient.
 In the face of this, Parliament now insists, and the majority judgment accepts, that the applicant’s case actually is to test the constitutional validity of PAIA in the High Court. But that is not the applicant’s case. It contends that legislation must be enacted, in addition to PAIA (and in addition to all other legislation purported to give effect to section 32). The additional legislation must deal specifically with the disclosure of political parties’ private funding. So the question is not whether PAIA is the legislation envisaged in section 32(2). Both Parliament and the applicant agree that it is. The question is whether Parliament has adequately fulfilled the obligation that provision imposes.
 In this way, the applicant confronts PAIA, but does so only to the extent that Parliament claimed that enacting PAIA meets its constitutional obligation under section 32(2) read with section 19(3).
 And only this Court can make the order the applicant seeks. Were the High Court to be approached, the only competent order it could grant would be one declaring PAIA inconsistent with the Constitution. But that is not what the applicant seeks. It takes no issue with the rights PAIA and the other specialised access to information legislation Parliament has enacted under section 32(2) confer. Its complaint is that all this legislation, together, is not enough. Parliament leaves an unconstitutional void in regard to political parties’ private funding. Hence the applicant seeks a more powerful, direct and trenchant remedy: an order in terms of section 32(2), read with section 167(4)(e). Once this Court’s exclusive jurisdiction is engaged, as it is, only it has power to grant that remedy – and should we conclude, in what follows, that PAIA does not afford access to the information the applicant seeks, then the applicant has established that it is entitled to the relief it seeks.
 What is more, if a court granted the applicant a declaration of constitutional invalidity of PAIA under section 172, that would imply that Parliament erred in enacting PAIA, and hence that Parliament must amend PAIA. But this is not so. PAIA’s enactment was constitutionally necessary. The majority judgment endorses this point. The question is whether it was enough. To portray the applicant’s argument as saying Parliament was wrong to enact PAIA misconceives it.
 It is also at odds with the separation of powers. The applicant’s case is that the scheme and operation of PAIA – though perhaps suited to their current task, of requiring the ad hoc disclosure of specified records upon application by an interested party – are entirely inapposite for the comprehensive disclosure to the public at large of all political parties’ funding. The Court’s order should not prescribe to Parliament that it must amend PAIA. Parliament should be free to meet its obligations under section 32 however it chooses – whether by amending PAIA, or by enacting new legislation, additional to PAIA, that specially targets the disclosure of political parties’ private funds.
 So sending the applicant back to start again in the High Court would force that court to adjudicate a case the applicant does not make, and to grant an order the applicant has never sought. The applicant does not ask Parliament to amend its existing legislation. Its argument is deeper-going: it is the richness of section 32’s promise that requires a manifold legislative response, of which PAIA is only part.
 It is precisely the kind of argument, one going beyond a critique of existing legislative provisions, and invoking the true depth of the Constitution’s vision, that section 167(4)(e) gives this Court special jurisdiction to hear.
 In summary: Parliament’s argument misconceives the applicant’s case, and does not take into account the reach and complexity of the rights and remedies section 32 and section 167 of the Constitution afford. Parliament’s formal defence should not impede this Court from reaching the questions of substance. The central issue is whether PAIA adequately fulfils the promise of section 32. Parliament claims it does. It has had full opportunity in these proceedings to make its case on the merits. Nevertheless, it now asks this Court to avoid determining the merits, and instead to send the case to the High Court on technical grounds. That Court has no jurisdiction to grant the order the applicant seeks, but can only make, in its stead, an inapposite order – one that would have to come to this Court anyhow for confirmation. It would be futile, and circuitous, to require the applicant to re-start in the High Court. This Court’s powers are properly invoked, and the applicant’s claim to relief must be determined.
Does PAIA afford a right of access to information about political parties’ private funding?
 As we have seen, the fundamental right to vote, read with section 32, requires that political parties’ private funding be disclosed. The question now is whether PAIA does this. The answer is No. The reasons are two. The first is that PAIA’s mechanisms and processes are inherently limited. They serve a valuable purpose, but that purpose is narrow. Second, they are not capable of affording citizens their right to be properly informed about political parties’ funding.
 First, PAIA operates pairwise. It requires one “requester” of information to address a request to another entity. PAIA compels disclosure only upon application. Moreover, that application must provide sufficient particulars to identify the record the requester seeks. In sum, as the applicant rightly contended, PAIA affords only the right to gain access, upon specific request, to specific records held by specific bodies at specific times.
 That right of access to information is important. But it is not capable of affording the electoral citizenry the information to which they are entitled about the way political parties vying for their votes are funded. That is a context with unique demands, to which PAIA does not address itself. Most obviously, the relationship is not pairwise. It is a relationship between dozens of political parties and many millions of voters. The right of individuals to apply to receive individual records, furnished on request, could never keep the electorate as a whole meaningfully informed. For that to be achieved, records must be made publicly available to all. And this would have to be done systematically and regularly, not only upon application. It is not possible for each voter to apply to each political party at each election to obtain the specified records he or she seeks. The difficulty reveals the disjunct between the purpose PAIA is designed to serve and the purpose of the legislation the applicant seeks to have enacted.
 In addition, PAIA cannot guarantee that political parties’ private contributions would be available for request at all. It would not require that these be documented. PAIA affords a right of access to “records”. It does not define “information”. It contains only a definition of “record”. This limits the operation of the statute to information that is recorded in some form or medium. Oral communications containing or constituting information are excluded. Also not contemplated are situations that may require physical access to a place in order to obtain information that is yet to be reduced to material form, such as a meeting of a parliamentary portfolio committee, a court hearing or inspecting the site of past happenings.
 Are these omissions serious? It would appear so. Depending on the nature of the information, and the possible disincentives to preserving it, the absence of an encompassing definition, underscores PAIA’s limited ambit. This is because a contract, undertaking, understanding, agreement or donation may all be orally concluded. In that event, as far as PAIA is concerned, there is no “record” – and hence no right of access to that information. This limited ambit creates obvious risks that some deal-doers will want to keep their transactions spoken, so that they are not “recorded”.
 PAIA also imposes no obligation on a record-holder to preserve recorded information until a formal application is made. This fits logically with the nature of the obligation PAIA creates. That is to provide access to records, after a request for them has been made. A body, private or public, can wipe out records as they are created, without falling foul of PAIA. Subject, of course, to any specifically applicable legislation, it could even design its systems so that records are methodically wiped out. There will be no breach of PAIA. The statute creates no proactive duty to preserve or disclose any category of records: the obligation is to provide access only once an information-seeker asks. Its obligations are entirely reactive.
 It is correct that the statute, while not defining “information”, extensively uses the concept. From short title to Schedules, PAIA uses the word “information” over 250 times. Most pivotally, the word “information” is used within the definition of “record”, which means “any recorded information”. But the proliferation of the word “information” makes the absence of a definition, together with the sharply limited definition of “record”, only the more striking. The impact is both conceptual and operational. The statute confines its operation and effect to recorded information in the form of “records” only.
 The upshot is that private contributions, and their amount and provenance, could be left unrecorded – and therefore incapable of being requested in terms of PAIA. This, together with the narrow pairwise relationship PAIA envisages, between individual requesters and individual entities holding the records, whose disclosure is compelled only upon application, means that it cannot fulfil the demand of section 19(3), or the promise of section 32.
 But there is a second, even more obtrusive reason for concluding that PAIA does not afford access to the information the applicant seeks and that it consequently does not meet the constitutional obligation section 32(2) imposes. This springs from the bodies to whose records it provides access. They fall within two categories only – public bodies, and private bodies. Both are defined. Subject, as is usual, to contra-textual indicators, a “public body” in PAIA means—
“(a) any department of state or administration in the national or provincial sphere of government or any municipality in the local sphere of government; or
(b) any other functionary or institution when—
(i) exercising a power or performing a duty 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.”
 This closely echoes the Constitution, which contains no definition of “the state”, but defines “organ of state”, in terms PAIA’s definition of “public body” appropriates. PAIA defines “private body” in section 1 as meaning:
“(a) a natural person who carries or has carried on any trade, business or profession, but only in such capacity;
(b) a partnership which carries or has carried on any trade, business or profession; or
(c) any former or existing juristic person, but excludes a public body”.
 These two definitions create a dichotomy that appears to leave a large gap. Section 32(1) confers a right of access to any information held “by the state”, plus to any information held “by another person”, provided it is “required for the exercise” and “protection of any rights”. Although neither “the state” nor “person” is defined in the Constitution, section 32 creates a single, all-encompassing, dichotomous category within which the right applies. The wide definition of “organ of state” in the Constitution means that in the first instance section 32(1) of the Bill of Rights gives a right of access to all information held by departments of state at any level of government, as well as by any other functionary or institution that exercises a power or performs a function under the national or a provincial constitution or that exercises a public power or public function in terms of any other legislation. This PAIA closely reflects.
 But that right is residually conferred in respect of information held by “any person”, if required for the exercise or protection of any rights. Despite the absence of a definition, the word “person” is plainly very wide. It is not limited to natural persons, for the Bill of Rights binds also 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. What is more, the Bill of Rights specifies which “persons” may enforce the rights it confers. This Court has repeatedly held that the ambit of the standing provision is very wide. So “person” includes any individual or association or community or group. It would certainly include a political party.
 Hence the right section 32 confers operates within a wide and potently encompassing field – the anvil on which its hammer falls is the entire state, and, outside the state, any person who holds information that is required for the exercise or protection of any rights. The obligation section 32(2) imposed on Parliament was therefore to enact legislation to give effect to the right of access to information held by anyone else (“another person”) that is required for the exercise or protection of any rights.
 But PAIA instead created a different dichotomy – that between public and private bodies, though the statute itself recognises that the dichotomy cannot be absolute. The statute defines “person” as meaning “a natural person or a juristic person”. While its definition of “public body” closely replicates that of “organ of state” in the Constitution, its definition of “private body” is far narrower than the concept of “person” that informs the Bill of Rights. The definition encompasses any former or existing juristic person. But as far as natural persons are concerned, it is confined to those who carry on, or who have carried on any trade, business or profession, “but only in such capacity”. Its application to partnerships is similarly confined. In the case of all private bodies, access is of course, in accordance with the Bill of Rights, restricted to records required for the exercise or protection of any rights.
 There are two problems with this. First, what if a natural person has records that are needed for the exercise or protection of any rights – but is not engaged in any trade, business or profession? PAIA does not give access. The field of natural persons is plainly not covered. But the omission does not stand on its own. There is a second, more telling difficulty with PAIA’s definition of “private body”. What if a body is neither public nor private? What if it straddles PAIA’s definition of public and private bodies, and falls in-between? While it is true that PAIA gives access to the records of juristic bodies, insofar as they are “private bodies”, political parties appear to fall within a category of political actors who may or may not be “juristic persons” for the purposes of PAIA. They fall into a very singular category of “persons”, envisaged in the Bill of Rights, but for whom PAIA doesn’t appear to cater at all.
Political parties and PAIA
 Political parties are neither public bodies nor private. In argument, Parliament contended that the records of political parties are accessible under PAIA, since they fall within the definition of private bodies. But the submission is overbroad. PAIA offers access to the records only of political parties that are juristic persons (or, conceivably, partnerships). What is more, the right of access it affords excludes from regulation all non-juristic persons not carrying on a trade, business or profession.
 Two implications flow from PAIA’s limited definition. First, it is not clear that political parties are in fact all “juristic persons”. It is impossible to say that the constitutions of all political parties constitute them as juristic persons. And whether a body that has capacity to sue and be sued in its own name is also a juristic person depends on its constitution. The mere capacity to sue and be sued does not necessarily entail juristic personhood. Although difficult to proceed in practical terms, the possibility that a political party may not be a common law universitas and hence not a juristic person cannot be excluded.
 There is a second more telling point. This is that our law does not require that political parties be juristic persons. A political party can be simply an organisation or movement, and not a juristic person. Under the Electoral Commission Act, “party” means a party registered under the Act, and includes—
“any organisation or movement of a political nature which publicly supports or opposes the policy, candidates or cause of any registered party, or which propagates non-participation in any election”.
 The Electoral Commission Act requires the chief electoral officer, on the fulfilment of certain prescribed conditions, to register any party. There is conspicuously no requirement in the statute that a political party be a juristic person. It could be any organisation or movement of a political nature, juristic or non-juristic. And, indeed, section 19(1) of the Bill of Rights, which gives every citizen the right to form a political party, lays down no requirement that the party formed must be a juristic person. Whether a requirement that a political party must be a juristic person, if it were imposed, would be constitutionally valid is not in issue now. The point is that PAIA does not cover political parties – whether big or small, predominant or minor – if they are not juristic persons.
 The applicant submitted in its written argument that political parties are properly to be considered part of “the state”, for purposes of disclosure of private funding information under section 32(1)(a). But this contention was pursued only faintly during oral argument. Rightly so. Political parties do not sit comfortably within the Constitution’s definition of “organ of state”, or PAIA’s definition of “public body”. The reason is that, while in certain of their functions they may perform statutory duties (such as when they constitute the national and provincial bodies that elect the members of executive government), it is simply constitutionally inappropriate to call them organs of state. They are not the state, nor are they part of it, even though on occasion they perform statutory functions. Thus, while it is possible to shoe-horn political parties into the definition of “public body”, they cannot sit comfortably there.
 On the other hand, political parties are quite plainly not private bodies, and, as already shown, if not juristic persons, they are not covered by PAIA at all. Even where a political party is a juristic person, and thus falls inside PAIA, the term “private” ill befits it. The reason lies in the nature of political parties, and the critical importance of their functioning to the success of the country’s constitutional project.
 This emerges from this Court’s decision in Ramakatsa, where the appellants sought to set aside as invalid a provincial conference of the ANC and all its outcomes on the basis that there were irregularities in many of the branch meetings that elected delegates to the conference. This Court exercised a robust jurisdiction. The whole Court, minority and majority, concluded that fundamental constitutional rights were implicated. This was because of their importance to the fulfilment of the right to vote. The judgment of Yacoob J explained, on behalf of the Court, that the right to participate in the activities of a political party obliges every political party to act lawfully and in accordance with its own constitution, and, correlatively, every member of a political party has “the right to exact compliance with the constitution of a political party by the leadership of that party”.
 Both the unanimous conclusions of the Ramakatsa Court, and its majority judgment, are antithetical to the notion that political parties are merely private entities. In short, the public/private disjunct in PAIA appears to have been created without having political parties in mind at all. They are a category of “persons”, outside the state, for whom PAIA has failed to make express or any provision. Where political parties should be, there is a gaping hole. This, the applicant submits, was because Parliament had been deliberating whether to regulate political parties separately. But there is no need to speculate; the practical upshot, and the meaning and application of PAIA, are clear.
Parliament has failed to meet its section 32(2) obligation
 PAIA, in other words, does not provide at all for access to the information about political parties’ private funding required for the exercise of the right to vote. It, like the other statutes the legislature has enacted in fulfilment of the right of access to information, constitutes, at best, only a partial fulfilment of Parliament’s obligation.
 Without specific legislation requiring political parties registered for elections to legislative bodies established under the Constitution to disclose their private funding, it follows that the applicant’s attack cannot be repulsed. It has established the constitutional insufficiency of PAIA, and hence that Parliament has not fulfilled its obligation under section 32 to enact legislation to afford access to the information reasonably required to exercise the right to vote. The supremacy clause, in section 2 of the Constitution, requires that an “obligation [which is] imposed by [the Constitution] must be fulfilled”.
 An order granting the applicant the relief it seeks should issue. We should acknowledge, as in Doctors for Life, that—
“[a]n order declaring that Parliament has failed to fulfil its constitutional obligation … and directing Parliament to comply with that obligation constitutes judicial intrusion into the domain of the principle legislative organ of the state. Such an order will inevitably have important political consequences. Only this Court has this power.”
Where a constitutional obligation is impugned, as here, the Constitution itself mandates the intrusion. But Parliament has considerable discretion to determine how best to fulfil its duty. This Court does not seek to prescribe to Parliament that it ought to legislate in a particular manner, as the majority judgment suggests, but Parliament must legislate in a way that gives effect it to its constitutional obligation.
 The applicant asked for an order directing Parliament to enact the required legislation within eighteen months, and to require Parliament to lodge a report on the steps it has taken every three months. Because this is a minority judgment, it is unnecessary to consider further the form of the order, save to say that I would have declared that Parliament has failed to fulfil its constitutional obligation to enact national legislation to give effect to the right of access to information as required by section 32(2) of the Constitution, to the extent that—
(a) information about the private funding of political parties registered for elections for any legislative body established under the Constitution is
reasonably required for the effective exercise of the right to vote in those elections; and
(b) no national legislation currently requires that this information be publicly accessible.
KHAMPEPE J, MADLANGA J, NKABINDE J and THERON AJ (Mogoeng CJ, Molemela AJ and Tshiqi AJ concurring):
 We have had the benefit of reading the judgment penned by our brother, Cameron J (minority judgment). We agree, for different reasons, with his finding regarding this Court’s exclusive jurisdiction. We further agree with the minority judgment’s exposition of the history behind the principle of constitutional subsidiarity.
 Our disagreement with the minority judgment lies in its conclusion that Parliament has failed to fulfil its constitutional obligation to enact the legislation envisaged in section 32(2) of the Constitution. Summarising it, our difficulty with the minority judgment is two-fold. First, insofar as it seeks to have Parliament legislate in a manner preferred by the applicant, the minority judgment violates the doctrine of separation of powers. We elaborate on this below. Second, the minority judgment’s conclusion that “the validity of [PAIA] is not at issue” does not bear scrutiny. The suggestion that PAIA has certain shortcomings is, in fact, an attack on its validity. Because – in that sense – the validity of PAIA is challenged and PAIA is the legislation envisaged in section 32(2), the principle of subsidiarity applies. On these alleged shortcomings, the applicant ought to have challenged the constitutional validity of PAIA frontally in terms of section 172 of the Constitution in the High Court (frontal challenge).
 The points of difference on the merits lead us to a different outcome: a dismissal of the application.
 Our approach makes it unnecessary for us to pronounce on whether information on the private funding of political parties is required for the exercise of the right to vote.
 The minority judgment sets out the background to this application and the parties’ submissions in great detail. We will only deal with the submissions to the extent necessary for our conclusions.
 The applicant accepts that Parliament did enact national legislation to give effect to the right of access to information in the form of PAIA. It contends, however, that the principle of subsidiarity does not apply because PAIA does not cover nor purport to “cover the entire field of legislation [giving] full effect to section 32(2)”, and that “Parliament’s obligation under section 32(2) of the Constitution did not begin and end with the enactment of PAIA”. It argues that PAIA gives effect “only” to one aspect of the right of access to information, namely the right to gain access to specific records held by specific bodies at specific times. The applicant submits that this is not an ordinary case of enforcing the right of access to information, but rather a case of enforcing the duty to enact national legislation required to give effect to the right under section 32(1) of the Constitution.
 In addition, the applicant argues that even though political parties are not “organs of state”, they are a special species of “private actors” with constitutional responsibilities to the voting public. It contends that political parties are part of the state for the purposes of section 32(1)(a). Consequently, everyone is entitled to information regarding the private funding of political parties as information “held by the state”. In the alternative, the applicant contends that citizens are entitled to information on the private funding of political parties as information that is required for the exercise or protection of their right contained in section 19(3) of the Constitution. This entitlement stems from section 32(1)(b).
 Notably, the applicant steadfastly asserts that it is not challenging the constitutional validity of PAIA, even if it was legislation enacted pursuant to section 32(2). Even so, according to it, the legislation required is fundamentally different from PAIA in its nature and purpose as it would require the disclosure of private funding information as a matter of “continuous course, rather than once-off upon request”. The ability to access information pertaining to the private funding of political parties on an ongoing basis, the applicant contends, is necessary if the right to vote is to have meaningful content.
 In essence, Parliament contends: the principle of subsidiarity applies; Parliament met its constitutional obligation by enacting PAIA; in accordance with the principle of subsidiarity, the applicant ought to have challenged the constitutional validity of PAIA in the High Court; and – because it has not done so – the matter ought to be dismissed. In amplification, Parliament maintains that the principle of subsidiarity precludes the applicant from having direct recourse to the section 32(1) right itself. The question is whether PAIA “purports to be the legislation required by section 32(2)”. And if it does, then the applicant is obliged to challenge it directly for failing to give effect to the right in the manner that the applicant contends is constitutionally compliant. What the applicant is not permitted to do, continues the contention, is to demand the enactment of a different piece of legislation that would deal with a matter for which PAIA was enacted.
 Issues that we are going to deal with are:
(a) exclusive jurisdiction;
(b) whether PAIA is the legislation envisaged in section 32(2) of the Constitution;
(c) separation of powers;
(d) circumstances in which the principle of subsidiarity applies and the need for it; and
(e) whether the applicant has challenged the constitutional validity of PAIA.
 On this aspect, our discomfort with the minority judgment lies in the fact that its conclusion is coloured by the finding it ultimately reaches on the nature of this application and the principle of subsidiarity. For that reason, we prefer a shorter route. And it is this.
 A court’s jurisdiction is determined on the basis of the claim in the pleadings. In Chirwa, Langa CJ held that—
“a court must assess its jurisdiction in the light of the pleadings. To hold otherwise would mean that the correctness of an assertion determines jurisdiction, a proposition that this Court has rejected. It would also have the absurd practical result that whether or not the High Court has jurisdiction will depend on the answer to a question that the court could only consider if it had that jurisdiction in the first place. Such a result is obviously untenable.”
 In a unanimous judgment, this Court confirmed Chirwa and held that—
“Jurisdiction is determined on the basis of the pleadings, as Langa CJ held in Chirwa, and not the substantive merits of the case. . . . In the event of the court’s jurisdiction being challenged at the outset (in limine), the applicant’s pleadings are the determining factor. They contain the legal basis of the claim under which the applicant has chosen to invoke the court’s competence. While the pleadings – including in motion proceedings, not only the formal terminology of the notice of motion, but also the contents of the supporting affidavits – must be interpreted to establish what the legal basis of the applicant’s claim is, it is not for the court to say that the facts asserted by the applicant would also sustain another claim, cognisable only in another court.” (Footnote omitted.)
 It follows that “the substantive merits of a claim cannot determine whether a court has jurisdiction to hear it”. We do realise that in certain instances claims of exclusive jurisdiction may be palpably contrived. Needless to say, those will not succeed.
 We conclude thus: the applicant alleges that Parliament has failed to fulfil the obligation imposed by section 32(2) of the Constitution to enact legislation that gives effect to the right contained in section 32(1) of the Constitution. In terms of section 167(4)(e) of the Constitution, only this Court has jurisdiction to answer that question.
Is PAIA the legislation envisaged in section 32(2) of the Constitution?
 The applicant contends that the “required legislation” in terms of section 32(2) has not been enacted and there is “simply no Act of Parliament” to be tested by the Supreme Court of Appeal or the High Court in terms of section 172(2)(a) of the Constitution. In the same breath, it attempts to show the insufficiency of PAIA in regard to the right for which it contends.
[137 Section 32 of the Constitution provides:
“(1) Everyone has the right of access to—
(a) any information held by the state; and
(b) any information that is held by another person and that is required for the exercise or protection of any rights.
(2) National legislation must be enacted to give effect to this right, and may provide for reasonable measures to alleviate the administrative and financial burden on the state.”
 The long title says PAIA was enacted to—
“give effect to the constitutional right of access to any information held by the state and any information that is held by another person and that is required for the exercise or protection of any rights”.
This mirrors the wording of section 32(1) of the Constitution.
 In relevant part, the preamble to PAIA provides that it was enacted to—
“(a) foster a culture of transparency and accountability in public and private bodies by giving effect to the right of access to information; [and]
(b) actively promote a society in which the people of South Africa have effective access to information to enable them to more fully exercise and protect all of their rights.” (Emphasis added.)
This relates to the exercise and protection of “all rights”, and not merely some.
 Section 9 of PAIA contains the objects of the Act, which include the following:
“(a) [T]o give effect to the constitutional right of access to—
(i) any information held by the state; and
(ii) any information that is held by another person and that is required for the exercise or protection of any rights.”
This too adopts the exact wording of section 32(1) of the Constitution.
 Having regard to all this, one can hardly think of any other indications and more plain language to evince the purpose for which PAIA was enacted.
 Indeed, in Independent Newspapers, Moseneke DCJ, writing for the majority, observed:
“At a general level, the right of access to information is entrenched, in the first instance, by the Constitution itself. Section 32 of the Constitution confers on everyone the right of access to any information held by the state or by another person that is required for the exercise or the protection of any rights. That right of access to information is given effect to and regulated through legislation in the form of the [PAIA].” (Footnotes omitted and emphasis added.)
 In the same case, Sachs J wrote:
“It is the right in section 32 of everyone to have access to information. [PAIA], adopted on 3 February 2000, gives effect to this right.” (Footnote omitted.)
 In Brümmer v Minister for Social Development and Others, writing for a unanimous Court, Ngcobo J says:
“Section 32(1) of the Constitution guarantees the right of access to information ‘that is required for the exercise or protection of any rights’. And the declared purpose of PAIA is to give effect to this constitutional right.”
 In a unanimous decision in PFE International, Jafta J held that “PAIA is the national legislation contemplated in section 32(2) of the Constitution”. Later in a dissent in Agri SA, Froneman J reasoned, on a point to which the disagreement did not relate:
“The [Mineral and Petroleum Resources Development Act] is not legislation that explicitly seeks to give effect to and circumscribe a fundamental right in the manner of, for example, [PAJA], [PAIA] or the Labour Relations Act, but in my view its provisions need to be interpreted in a manner that is best consistent with section 25.” (Footnotes omitted.)
 Even the High Court has considered PAIA to be the legislation enacted to give effect to section 32.
 These authorities brook no possibility that PAIA is not the legislation enacted to give effect to the section 32(1) right. That it may have shortcomings in its protection of the right and possibly even be constitutionally invalid does not alter this legal reality.
 Notably, PAIA shares a similar history with PAJA, which gives effect to the section 33(1) and (2) rights. As this Court held in Bato Star, PAJA was enacted to give effect to the right to administrative justice contained in section 33(1) and (2) of the Constitution. Likewise, as we have concluded, PAIA was passed in compliance with section 32(2) of the Constitution. Schedule 6 of the Constitution contained transitional provisions that applied to sections 32 and 33. The Schedule required that national legislation be enacted within three years of the Constitution coming into effect. The President assented to PAIA on 2 February 2000 and to PAJA a day later. Both dates are within three years of the Constitution coming into effect and just shy of missing the deadline. If the legislation had not been enacted, sections 32(2) and 33(3) would have fallen away. It would be ludicrous for anyone to suggest that section 32(2) has lapsed for lack of enactment of the legislation envisaged in that section. Indeed, it is not surprising that no one – to our knowledge – has ever made a suggestion of the sort. This buttresses our view that PAIA is the legislation enacted to give effect to the section 32(1) right.
 The minority judgment makes the point that PAIA is not the only legislation that gives effect to section 32. In this regard, it refers to various other pieces of legislation that make provision for access to information. However, even though those pieces of legislation do make this provision, they are distinguishable from PAIA. The main focus of each is some other subject; not access to information in terms of section 32(1) of the Constitution. That this is so is reinforced by the sparse manner in which the content of each touches on the right of access to information. In each, provision for the right is merely incidental to the legislation’s main focus. On the contrary, PAIA’s focus is one subject: the provision of information in terms of section 32(1) of the Constitution. In short, that there is out there a plethora of other pieces of legislation providing for access to information does not mean all those pieces of legislation are the legislation envisaged in section 32(2) of the Constitution.
Separation of powers
 This Court has expressed itself thus:
“The principle of separation of powers, on the one hand, recognises the functional independence of branches of government. On the other hand, the principle of checks and balances focuses on the desirability of ensuring that the constitutional order, as a totality, prevents the branches of government from usurping power from one another. In this sense it anticipates the necessary or unavoidable intrusion of one branch on the terrain of another. No constitutional scheme can reflect a complete separation of powers: the scheme is always one of partial separation.”
 Why do we have this principle? Langa CJ explains in Glenister I:
“The principle of checks and balances focuses on the desirability that the constitutional order, as a totality, prevents the branches of government from usurping power from one another. The system of checks and balances operates as a safeguard to ensure that each branch of government performs its constitutionally allocated function and that it does so consistently with the Constitution.”
 We are mindful that it is this Court that is the final arbiter on adherence to the Constitution and its values. On this, in Doctors for Life, Ngcobo J says:
“But under our constitutional democracy, the Constitution is the supreme law. It is binding on all branches of government and no less on Parliament. When it exercises its legislative authority, Parliament ‘must act in accordance with, and within the limits of, the Constitution’, and the supremacy of the Constitution requires that ‘the obligations imposed by it must be fulfilled’. Courts are required by the Constitution ‘to ensure that all branches of government act within the law’ and fulfil their constitutional obligations. This Court ‘has been given the responsibility of being the ultimate guardian of the Constitution and its values’. Section 167(4)(e), in particular, entrusts this Court with the power to ensure that Parliament fulfils its constitutional obligations. This section gives meaning to the supremacy clause, which requires that ‘the obligations imposed by [the Constitution] must be fulfilled’. It would therefore require clear language of the Constitution to deprive this Court of its jurisdiction to enforce the Constitution.” (Footnotes omitted.)
 With all this in mind, we proceed to have a close look at what the applicant seeks.
 The true complaint by the applicant is the manner in which Parliament – exercising a power that vests solely in it – has chosen to legislate. Let us demonstrate this in the following manner. Assuming that – besides this complaint – there was no basis for raising the shortcomings that the minority judgment deals with, there would be no question that PAIA does not afford interested voters access to information on the private funding of political parties. The only complaint would be that this information can only be made available at the request of an individual and only to that individual. But would it still be open to that individual to say legislation envisaged in section 32(2) has not been passed? Definitely not. Why do we say so?
 The applicant wants information on the private funding of political parties to be made available in a manner preferred by it. It prefers that the legislation should require the disclosure of the information as a matter of “continuous course, rather than once-off upon request”. According to the minority judgment, what South Africa must have is systematic disclosure. It may well be that this is ideal; who knows? But that is not the issue. It is for Parliament to make legislative choices as long as they are rational and otherwise constitutionally compliant. Crucially, lack of rationality is not an issue in these proceedings.
 Despite its protestation to the contrary, what the applicant wants is but a thinly veiled attempt at prescribing to Parliament to legislate in a particular manner. By what dint of right can the applicant do so? None, in the present circumstances. That attempt impermissibly trenches on Parliament’s terrain; and that is proscribed by the doctrine of separation of powers.
 To the extent that the minority judgment suggests that individual requests would present interested voters with insurmountable problems, this is difficult to grapple with, and indeed, inappropriate to raise in the absence of an irrationality challenge on the choice made by Parliament.
 We must highlight that the minority judgment proceeds from an assumption that all voters require information on the private funding of political parties. The basis for that assumption is not explained. And the minority judgment’s conclusion, on the importance of this information to the exercise of the right to vote, does not give a basis for the assumption. On what basis does the minority judgment discount the possibility that – even if the information were readily available – some people would not have recourse to it before exercising their right to vote? We do not know. We should not be understood to say access to this information may not reasonably be required for the exercise of the right to vote. That is a matter we need not reach. What we take issue with is the unexplained assumption from which the minority judgment proceeds.
 We are mindful that the applicant does complain of, and the minority judgment also points to, other shortcomings in PAIA. Those shortcomings are best dealt with in a frontal challenge. Based on the applicant’s own say so and the minority judgment’s conclusion, this application is not a frontal challenge. For the reasons we give shortly, the application has been brought in breach of the principle of subsidiarity.
Circumstances in which the principle of subsidiarity applies and the need for it
 Contrary to the suggestion in the minority judgment that our insistence on compliance with the principle puts form ahead of substance, this principle plays an important role. The minority judgment correctly identifies the “inter-related reasons from which the notion of subsidiarity springs”. First, allowing a litigant to rely directly on a fundamental right contained in the Constitution, rather than on legislation enacted in terms of the Constitution to give effect to that right, “would defeat the purpose of the Constitution in requiring the right to be given effect by means of national legislation.” Second, comity between the arms of government enjoins courts to respect the efforts of other arms of government in fulfilling constitutional rights. Third, “allowing reliance directly on constitutional rights, in defiance of their statutory embodiment, would encourage the development of ‘two parallel systems of law’”.
 The principle of subsidiarity is a well-established doctrine within this Court’s jurisprudence. The essence of the principle was captured by O’Regan J in Mazibuko, where she held that—
“where legislation has been enacted to give effect to a right, a litigant should rely on that legislation in order to give effect to the right or alternatively challenge the legislation as being inconsistent with the Constitution.”
 The minority judgment says that subsidiarity does not apply because the validity of PAIA is not in issue. This is difficult to follow. If legislation fails to provide sufficiently for the protection of the right contained in section 32(1) of the Constitution, surely it must be invalid to the extent of the insufficiency. Therefore, the assertion of insufficiency puts PAIA’s validity in issue. The two are indistinguishable. For that reason, we say – on this Court’s jurisprudence – subsidiarity must apply.
 Essentially, the applicant’s complaint is that PAIA suffers from certain shortcomings. In that context, this Court has held that the principle of subsidiarity enjoins an applicant to challenge the legislation exactly for its shortcomings. In SANDU, which concerned the right to collective bargaining, this Court remarked:
“If . . . legislation is wanting in its protection of the section 23(5) right in the litigant’s view, then that legislation should be challenged constitutionally. To permit the litigant to ignore the legislation and rely directly on the constitutional provision would be to fail to recognise the important task conferred upon the Legislature by the Constitution to respect, protect, promote and fulfil the rights in the Bill of Rights.” (Footnote omitted and emphasis added.)
 The deficient legislation must be challenged “as falling short of the constitutional standard”.
 According to Ngcobo J, in New Clicks:
“Where, as here, the Constitution requires Parliament to enact legislation to give effect to the constitutional rights guaranteed in the Constitution, and Parliament enacts such legislation, it will ordinarily be impermissible for a litigant to found a cause of action directly on the Constitution without alleging that the statute in question is deficient in the remedies that it provides.”
 Axiomatically, it cannot be that the principle of subsidiarity applies only where the legislation does exactly that which is constitutionally required. If that were the case, there could hardly ever be any meritorious challenges based on constitutional deficiencies or other bases of constitutional invalidity. Unsurprisingly, Parliament argues that – in accordance with the principle of subsidiarity – a frontal challenge is indicated where legislation that seeks to give effect to a constitutional right is believed to fall short of doing so. Van der Walt aptly says:
“In view of the Constitutional Court’s justification of the first two subsidiarity principles, the question is not whether legislation in fact gives effect to a right in the Bill of Rights, but whether it was enacted to do so. In other words, the focus is on the intention of the post-1994 democratic legislature to honour its constitutional obligations and promote the spirit, purport and object of the Bill of Rights through exercise of its legislative powers.” (Emphasis added.)
 Let us make the point that, in context, the authorities referred to in this and the minority judgment on the principle of subsidiarity envisage a frontal challenge.
 The strong reservations expressed by this Court in the comparable case of Democratic Party underscore this. In that case the appellant contended that section 16(5) of the Local Government Transition Act (Transition Act) was inconsistent with section 160(3)(b) read with section 160(2)(b) of the Constitution. The inconsistency relied upon purportedly stemmed from the fact that section 16(5) of the Transition Act required a two-thirds majority of members of council for the approval of a council’s budget. On the other hand, section 160(3)(b) read with section 160(2)(b) of the Constitution provides for the approval of a council’s budget by a simple majority of members of the council. After engaging in an interpretative exercise that, inter alia, involved a consideration of section 16(5) of the Transition Act, the two sections of the Constitution and item 26(2) of Schedule 6 of the Constitution, the Court concluded that there was no merit in the appeal. It then bemoaned the fact that the appellant “did not apply for an order declaring section 16(5) invalid” and that the appellant had instead “relied on the invalidity of the section as the foundation for the relief claimed”.
 The appellant had sought to justify the procedure it had adopted thus. It was interested only in relief that was consequent upon the constitutional invalidity of the Transition Act. It was inconvenient, expensive and time consuming first to engage in the lengthy process that would culminate in a confirmation by this Court in terms of section 172(2) of the Constitution. To subject the appellant to this served no purpose. The Court noted that the adverse conclusion it had reached against the appellant “render[ed] it both unnecessary and undesirable to adjudicate on a preliminary issue which would have otherwise been of some relevance”.
 Unimpressed by the appellant’s attempt at justifying the procedure followed, Yacoob J, writing for a unanimous Court, cautioned:
“[C]onsiderable difficulties stand in the way of the adoption of a procedure which allows a party to obtain relief which is in effect consequent upon the invalidity of a provision of an Act of Parliament without any formal declaration of the invalidity of that provision.
Firstly, such a procedure appears to be incompatible with the Constitution. Section 172(1) obliges a Court to declare a statutory provision which is inconsistent with the Constitution invalid to the extent of the inconsistency. It was conceded by counsel for the appellant that the course chosen is at least inconsistent with the literal meaning of section 172(2)(a) of the Constitution, which provides that a declaration of invalidity of an Act of Parliament by a High Court has ‘no force’ unless it is confirmed by this Court. The grant of any order by a High Court premised on a finding of invalidity of a provision of an Act of Parliament (other than temporary relief contemplated by section 172(2)(b) of the Constitution) is tantamount to that finding being infused with ‘force’ contrary to section 172(2)(a) of the Constitution.
Secondly, the suggested procedure is likely to be a source of uncertainty and confusion about the status of a provision of an Act of Parliament. The purpose of section 172(2) is to provide certainty by requiring confirmation of an order of invalidity of a provision of an Act of Parliament by this Court as a prerequisite for any finding of invalidity being of force. Sanctioning the suggested procedure could nullify that purpose.
Thirdly, the practice that has been urged upon this Court carries with it the distinct danger that Courts may restrict their enquiry into the constitutionality of an Act of Parliament and concentrate on the position of a particular litigant. This might mean that a provision of an Act of Parliament may be held valid for one set of circumstances and invalid for another. As Ackermann J said:
‘The consequence of such a (subjective) approach would be to recognise the validity of a statute in respect of one litigant, only to deny it to another. Besides resulting in a denial of equal protection of the law, considerations of legal certainty, being a central consideration in a constitutional state, militate against the adoption of the subjective approach.’” (Footnotes omitted.)
 The present application is comparable because in it too the relief sought by the applicant “is in effect consequent upon the invalidity of . . . an Act of Parliament”, PAIA in this instance. The essence of the applicant’s assertion that PAIA is deficient in its protection of the section 32(1) right is that PAIA is constitutionally invalid to the extent of the deficiency. No amount of disavowal of that – something we deal with later – can change this reality. As in Democratic Party, here too the applicant similarly is seeking the relief without any formal declaration of the invalidity of PAIA. In criticising our reliance on this case, the minority judgment proceeds from the premise that the instant application is not about the invalidity of PAIA. That premise is mistaken. The shortcomings complained of suggest that PAIA is invalid.
 Also, we have demonstrated that the other basis of distinction, which is that the applicant is seeking relief of a special kind, cannot succeed for the simple reason that what the applicant is asking for flouts the separation of powers doctrine.
 Democratic Party did not concern rights protected in the Bill of Rights. It was about testing the validity of a statutory provision against a non-Bill of Rights provision of the Constitution. The application before us is about a right in the Bill of Rights. In this context, the difficulties of an applicant who – in essence – relies for relief on the constitutional invalidity of an Act of Parliament without seeking a declaratory order to that effect are exacerbated. This is so because of the procedure followed in determining whether an Act of Parliament is inconsistent with a right in the Bill of Rights. That procedure entails: an enquiry whether the impugned legislation – including a deficiency in it – does, in fact, limit a right in the Bill of Rights; if it does, whether the limitation is justified under section 36(1) of the Constitution; and, if it is not justified, a declaration that it is constitutionally invalid. In that process, evidence – especially on the section 36(1) justification analysis – plays a crucial role. Because of the form the application has taken, the evidence that has been proffered is not of a nature that could address all these.
 Or, does none of this matter? Of course it matters. Here lies the fundament of our problem: we cannot bring ourselves to hold that there has been non compliance with a constitutional obligation in circumstances where the shortcomings complained of by the applicant – and amplified by the minority judgment – may well prove to be constitutionally compliant. The issue is not whether they are indeed compliant. Whether they are, is something that may be tested properly in what we have tagged a frontal challenge. Therein lies the jurisprudential value of the principle of subsidiarity.
 On the procedure resorted to by the applicant and the approach adopted by the minority judgment, the usual procedural hoops in a frontal challenge that invokes inconsistency with a right in the Bill of Rights are bypassed. It may well be that Parliament might have been able to demonstrate that what shortcomings there may be are justified in terms of section 36(1) of the Constitution. How do we then reach a conclusion that Parliament has failed to comply with a constitutional obligation? Or, do we simply say, quite plainly, Parliament could never have been able to show justification? How can we say that when – as we seek to demonstrate below – that was not a case that Parliament had to meet and, therefore, not an issue before us? That cannot be so.
 Authority tells us that even in an apparent “open and shut” case, an affected party must be given an opportunity to meet the case advanced by an adversary. Parliament has been denied that opportunity. We cannot resist the eloquence of Megarry J in John v Rees:
“As everybody who has anything to do with the law well knows, the path of the law is strewn with examples of open and shut cases which, somehow, were not; of unanswerable charges which, in the event, were completely answered; of inexplicable conduct which was fully explained; of fixed and unalterable determinations that, by discussion, suffered a change.”
 Some of the contentions made by the applicant rather belatedly – in written and oral argument – also illustrate the inherent problem with the procedure adopted by it. Two examples are the “challenge” on PAIA’s limitation as to the definition of “record” and its exemptions on confidentiality. Parliament was never called upon to meet a case of that nature. We have no idea what it might have said on the constitutional validity of these issues. It is, in any event, imperative that a litigant should make out its case in its founding affidavit, and certainly not belatedly in argument. The exception, of course, is that a point that has not been raised in the affidavits may only be argued or determined by a court if it is legal in nature, foreshadowed in the pleaded case and does not cause prejudice to the other party.
 As we see them, the authorities we cite on the principle of subsidiarity in  to  and Democratic Party are not about a tangential challenge that does not frontally seek a declaration of constitutional invalidity. The challenge they refer to is one that seeks a declaration of constitutional invalidity. The applicant’s case is not a challenge of this nature.
 Apparently, the applicant seeks to distinguish this case from those that insist on the principle of subsidiarity purely on the basis that here the applicant is not seeking to enforce a right in the Bill of Rights; it is rather seeking a mandamus for Parliament to fulfil a constitutional obligation. We cannot agree. The applicant, by the simple stratagem of crying a failure to comply with a constitutional obligation to enact the requisite legislation, seeks to extricate itself from what the principle of subsidiarity demands. Surely, that cannot be. As we have sought to demonstrate above, an insistence on compliance with the principle of subsidiarity in an instance like the present is not idle. It serves a useful jurisprudential and practical purpose. The minority judgment’s approach presents us with a conceptual, if not jurisprudential, difficulty.
 What also reinforces the need to respect the principle of subsidiarity in the context of this application is the minority judgment’s recognition of the multiplicity of other persons or entities – besides just political parties – that are excluded by the definition of “private body” in PAIA. That means PAIA has failed to provide for the enjoyment of an unimaginable number of rights by countless categories of entities and persons. The approach in the minority judgment lends itself to the possibility of multiple and varied complaints that Parliament has failed to legislate for access to information of one type or another from this or the other type of non-public person or entity that does not fit within the definition of “private body”. An approach that is susceptible to these ad hoc and possibly nit-picking individualised claims is problematic. This points us in one direction, and one direction only: if PAIA is the legislation envisaged in section 32(2) of the Constitution, the principle of subsidiarity must definitely apply to this matter. We have already concluded that PAIA is the envisaged legislation.
 For all the above reasons, there is absolutely no reason for the principle of subsidiarity not to apply in this matter.
 We should not be understood to suggest that the principle of constitutional subsidiarity applies as a hard and fast rule. There are decisions in which this Court has said that the principle may not apply. This Court is yet to develop the principle to a point where the inner and outer contours of its reach are clearly delineated. It is not necessary to do that in this case.
 Having concluded that PAIA is the required legislation under section 32(2) of the Constitution, the next question is whether it has been challenged. If it has not been, the applicant is in breach of the principle of subsidiarity.
Has PAIA been challenged?
 The applicant disavows any challenge to the validity of PAIA. It says it “does not direct any challenge against the inherent constraints on the application of PAIA. These constraints are logical and legitimate for PAIA to serve its purpose, which is a deliberately limited one”. We accept that this is not a frontal challenge to the validity of PAIA. If anything, it is an attempt to avoid dealing with PAIA.
 The applicant refers to certain constraints in PAIA that, in its view, make it impossible for anyone to access the sort of information it seeks at the intervals that the applicant would prefer. In its written submissions, the applicant makes clear that—
“the question of whether political parties are public or private bodies, for the purposes of PAIA, does not arise in the present application. This case concerns the proper interpretation of section 32 of the Constitution, which distinguishes between ‘the state’ and ‘another person’.” (Emphasis added.)
Curiously, the minority judgment relies heavily on the very issue the applicant disavows. That is, whether political parties are public or private bodies.
 It is exactly because there has been no frontal challenge to the constitutional validity of PAIA that one sees Parliament’s refrain, said almost ad nauseam and without much more, which is that PAIA is the requisite legislation for the protection of the section 32(1) right. In the circumstances, that refrain is understandable. The challenge is too tangential to expect more of Parliament. What else could it have said? It could not have been expected to answer the oblique suggestion of constitutional invalidity by following the usual steps when a Bill of Rights-based frontal challenge has been brought.
 The time-honoured practice in frontal constitutional challenges is for the Minister responsible for the administration of the impugned Act to be cited so as to make the necessary input expected of her or him on the constitutional validity, or lack of it, of the Act. In Tongoane, this Court reiterated:
“On a number of occasions this Court has emphasised that when the constitutional validity of an Act of Parliament is impugned, the Minister responsible for its administration must be a party to the proceedings inasmuch as his or her views and evidence tendered ought to be heard and considered.” (Emphasis added.)
 The Minister responsible for the administration of PAIA, then known as the Minister of Justice and Constitutional Development was cited as a respondent. It is worth noting that, although he initially filed a notice to oppose the application, he has remained supine - not filing a piece of paper thereafter. We are certain that this Court would have been most displeased with that attitude had this been a frontal challenge of PAIA. We are not aware that a single member of this Court is in any way concerned that the Minister has not participated in these proceedings.
 The Minister’s non-participation does not surprise us. He was cited in proceedings concerning an alleged failure by Parliament to comply with a constitutional obligation, not in proceedings challenging the constitutional validity of an Act for whose administration he was responsible. The relief sought was a battle pre-eminently between the applicant, Parliament and, possibly, political parties. We do not know, unless we were to speculate, what the Minister’s attitude would have been had the fray been one for him to enter.
 Crucially, in written submissions filed in response to directions issued by the Chief Justice on 30 September 2014, the applicant says the relief sought is “directed at Parliament alone”. It adds that the other respondents have been cited “only by virtue of the interest that they may have in its outcome” and that “[n]o relief is sought against them”. That is the nature of application the applicant knows itself to have brought. It is not a challenge that would have required the Minister or Parliament, for that matter, to defend the constitutional validity of PAIA. The Minister and Parliament were not afforded an opportunity to oppose a challenge to PAIA’s constitutional validity.
 The “challenge” that the minority judgment says has been launched exposes us to the risk that we refer to in  above. That is, the possibility of conflicting findings: one that says there has been a failure to comply with a constitutional obligation; and another that says PAIA’s shortcomings are constitutionally compliant.
 In sum, the applicant does not challenge the constitutional validity of PAIA, at least not frontally, as envisaged in section 172 of the Constitution. The principle of subsidiarity requires that it should have done so.
 Although the application falls under this Court’s exclusive jurisdiction, PAIA is the legislation envisaged in section 32(2) of the Constitution. The applicant has not challenged it frontally for being constitutionally invalid. In accordance with the principle of subsidiarity, it ought to have done so as that principle is applicable to this application. The application must fail.
 Biowatch applies. There should be no order as to costs.
 Consequently, the following order is made:
The application is dismissed.
For the Applicant: D Unterhalter SC and M du Plessis instructed by Webber Wentzel Attorneys
For the First and Second Respondents: W Trengove SC, V Ngalwana SC and F Karachi instructed by the State Attorney[Context] [Hide Context]
 Section 167(4) of the Constitution provides for matters that only this Court may decide:
“Only the Constitutional Court may—
(a) decide disputes between organs of state in the national or provincial sphere concerning the constitutional status, powers or functions of any of those organs of state;
(b) decide on the constitutionality of any parliamentary or provincial Bill, but may do so only in the circumstances anticipated in section 79 or 121;
(c) decide applications envisaged in section 80 or 122;
(d) decide on the constitutionality of any amendment to the Constitution;
(e) decide that Parliament or the President has failed to fulfil a constitutional obligation; or
(f) certify a provincial constitution in terms of section 144.”
 Section 167(6) of the Constitution provides:
“National legislation or the rules of the Constitutional Court must allow a person, when it is in the interests of justice and with leave of the Constitutional Court—
(a) to bring a matter directly to the Constitutional Court; or
(b) to appeal directly to the Constitutional Court from any other court.”
 Section 19 of the Bill of Rights is headed “Political rights” and provides:
“(1) Every citizen is free to make political choices, which includes the right—
(a) to form a political party;
(b) to participate in the activities of, or recruit members for, a political party; and
(c) to campaign for a political party or cause.
(2) Every citizen has the right to free, fair and regular elections for any legislative body established in terms of the Constitution.
(3) Every adult citizen has the right—
(a) to vote in elections for any legislative body established in terms of the Constitution, and to do so in secret; and
(b) to stand for public office and, if elected, to hold office.”
 Section 32 of the Bill of Rights is headed “Access to information” and provides:
“(1) Everyone has the right of access to—
(a) any information held by the state; and
(b) any information that is held by another person and that is required for the exercise or protection of any rights.
(2) National legislation must be enacted to give effect to this right, and may provide for reasonable measures to alleviate the administrative and financial burden on the state.”
 2 of 2000.
 In Doctors for Life International v Speaker of the National Assembly and Others  ZACC 11; 2006 (6) SA 416 (CC); 2006 (12) BCLR 1399 (CC) (Doctors for Life) at para 18, this Court noted that “the case of a law that infringes a right in the Bill of Rights . . . concerns the validity of the impugned law and not the failure to fulfil an obligation”.
 Majority judgment at .
 Section 236 of the Constitution provides in full:
“Funding for political parties.—To enhance multi-party democracy, national legislation must provide for the funding of political parties participating in national and provincial legislatures on an equitable and proportional basis.”
 103 of 1997.
 B 67-97 (Bill).
 Annexed to the Speaker’s answering affidavit is a document titled “Announcements, Tabling and Committees Reports” dated 31 October 1997 (Report).
 Section 11 of PAIA provides:
“(1) A requester must be given access to a record of a public body if—
(a) that requester complies with all the procedural requirements in this Act relating to a request for access to that record; and
(b) access to that record is not refused in terms of any ground for refusal contemplated in Chapter 4 of this Part.
(2) A request contemplated in subsection (1) includes a request for access to a record containing personal information about the requester.
(3) A requester’s right of access contemplated in subsection (1) is, subject to this Act, not affected by—
(a) any reasons the requester gives for requesting access; or
(b) the information officer’s belief as to what the requester’s reasons are for requesting access.”
 Section 50 of PAIA provides:
“(1) A requester must be given access to any record of a private body if—
(a) that record is required for the exercise or protection of any rights;
(b) that person complies with the procedural requirements in this Act relating to a request for access to that record; and
(c) access to that record is not refused in terms of any ground for refusal contemplated in Chapter 4 of this Part.
(2) In addition to the requirements referred to in subsection (1), when a public body, referred to in paragraph (a) or (b)(i) of the definition of “public body” in section 1, requests access to a record of a private body for the exercise or protection of any rights, other than its rights, it must be acting in the public interest.
(3) A request contemplated in subsection (1) includes a request for access to a record containing personal information about the requester or the person on whose behalf the request is made.”
 Above n 3.
 IDASA above n 12 at para 7. The Inkatha Freedom Party filed a notice to abide.
 IDASA above n 12 at para 25. See also PAIA’s definitions of “private body” and “public body” as set out at  to  below.
 IDASA id.
 In Mr Motlanthe’s answering affidavit in IDASA, annexed to My Vote Counts’ founding affidavit in this Court, he noted that South Africa is a signatory member of the African Union and, in terms of Article 10 of the African Union Convention on Preventing and Combating Corruption (see n 25 below), it is obliged, inter alia, to adopt legislative and other measures to “incorporate the principle of transparency into funding of political parties”. He added: “Parliament will fulfil this obligation”.
 IDASA above n 12 at paras 30-2.
 Id at para 42.
 Id at para 52.
 United Nations Convention against Corruption resolution 58/4 of 31 October 2003 (UN Convention).
 Article 7(3) of the UN Convention provides:
“Each State Party shall also consider taking appropriate legislative and administrative measures, consistent with the objectives of this Convention and in accordance with the fundamental principles of its domestic law, to enhance transparency in the funding of candidatures for elected public office and, where applicable, the funding of political parties.” (Emphasis added.)
 African Union Convention on Preventing and Combating Corruption of 11 July 2013 (AU Convention).
 Article 10 of the AU Convention provides:
“Each State Party shall adopt legislative and other measures to:
(a) Proscribe the use of funds acquired through illegal and corrupt practices to finance political parties; and
(b) Incorporate the principle of transparency into funding of political parties.”
 Above n 24 and 26.
 The applicant also relied on these international conventions and section 7(2) of the Bill of Rights to show that disclosure of parties’ private funding is a constitutionally required, corruption-fighting measure, but my conclusion makes it unnecessary to consider these further arguments.
 In April 2010, then Independent Democrats Member of Parliament Lance Greyling, successfully lobbied for Parliament’s Joint Rules Committee to establish an ad hoc committee to draft legislation regulating private political party funding. While the Chief Whips’ Forum outright rejected the legislative proposal, it was nevertheless referred for consideration to the Committee on Private Members’ Legislative Proposals and Special Petitions, as well as to the Joint Committee on Ethics and the Presidency. Both committees concluded that the legislation should not proceed, as it was “not feasible”. On 18 August 2011, the National Assembly adopted a report deciding not to pursue the legislative proposal.
 Direct reliance on section 32 of the Constitution may be possible where the basis of the attack on legislation giving effect to the right is “under inclusive” or “over restrictive” and therefore limits the substance of the right. Therefore, it is “consistent with constitutional democratic theory to give Parliament the ability to flesh out the detail of a fundamental right, but not to construct the very meaning of the right.” (Footnote omitted). Klaaren and Penfold “Access to Information” in Woolman et al (eds) Constitutional Law of South Africa Service 3 (2011) at 62-4 to 62-5.
 IDASA above n 12 at para 6.
 Section 167(4)(e) at above n 1 read in contrast with section 172(2)(a) of the Constitution below n 145.
 Majority judgment at .
 Doctors for Life above n 6 at para 9.
 Minister of Police and Others v Premier of the Western Cape and Others  ZACC 33; 2014 (1) SA 1 (CC); 2013 (12) BCLR 1405 (CC) at para 20 and Von Abo v President of the Republic of South Africa  ZACC 15; 2009 (5) SA 345 (CC); 2009 (10) BCLR 1052 (CC) at para 33.
 Doctors for Life above n 6 at para 253.
 Id at para 254.
 Section 167(4)(e) above n 1.
 Doctors for Life above n 6 at para 19.
 Id at para 21.
 Id at para 20.
 The Bill of Rights specifies that the state must “respect, protect, promote and fulfil the rights” in it (section 7(2)); it may not discriminate unfairly (section 9(3)); it must assign a legal practitioner in certain circumstances, if substantial injustice would otherwise result (section 28(1)(h) [civil proceedings affecting a child], section 35(2)(c) and section 35(3)(g) [detained and accused persons]); and the state bears specified duties in relation to those detained under a state of emergency (section 37(6)(h), (7) and (8)).
 Id at section 25(5) reads: “The state must take reasonable legislative and other measures, within its available resources, to foster conditions which enable citizens to gain access to land on an equitable basis.”
 Id at section 26(2).
 Id at section 27(1)(a).
 Id at section 27(1)(b).
 Id at section 27(1)(c), read with section 27(2).
 Id at section 29(1)(b).
 Id at section 9(4) reads:
“No person may unfairly discriminate directly or indirectly against anyone on one or more grounds in terms of subsection (3). National legislation must be enacted to prevent or prohibit unfair discrimination.”
 Id at section 32 is set out in above n 4.
 Id at section 33 provides in part:
“(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.”
 Id at section 25(6) provides an entitlement to legally secure tenure or to comparable redress for persons or communities whose tenure of land is legally insecure as a result of past racially discriminatory laws or practices.
 The grammatical form is the same as that in section 23(5) of the Bill of Rights, which provides in part that “[n]ational legislation may be enacted to regulate collective bargaining”. The provision is permissive, and creates no obligation. The full terms of section 23(5) are set out in n 114 below.
In a similar vein to section 23(5) of the Bill of Rights, section 23(6) provides:
“National legislation may recognise union security arrangements contained in collective agreements. To the extent that the legislation may limit a right in this Chapter the limitation must comply with section 36(1).”
 Section 43(a) of the Constitution provides that, in the Republic, the legislative authority “is vested in Parliament”, as set out in section 44 of the Constitution. While it is true that Parliament is also the enacting agent of the “legislative measures” that the Bill of Rights elsewhere requires, those measures are enacted by Parliament as part of a range of legislative and non-legislative measures that the state, as a whole, must take in fulfilment of the Bill of Rights.
 Women’s Legal Centre above n 42 at para 21. See also section 7(2) of the Bill of Rights.
 Section 19 of the Bill of Rights.
 Ex parte Chairperson of the Constitutional Assembly: In re Certification of the Constitution of the Republic of South Africa  ZACC 26; 1996 (4) SA 744 (CC); 1996 (10) BCLR 1253 (CC) (First Certification judgment) at para 186.
 Section 47(3)(c) of the Constitution provides (section 62(4)(d) being to the same effect in the case of the National Council of Provinces):
“(3) A person loses membership of the National Assembly if that person—
. . .
(c) ceases to be a member of the party that nominated that person as a member of the Assembly, unless that member has become a member of another party in accordance with Schedule 6A.”
In the case of the National Council of Provinces, section 62(4)(d) of the Constitution similarly provides:
“(4) A person ceases to be a permanent delegate if that person—
. . .
(d) ceases to be a member of the party that nominated that person and is recalled by that party.”
In the case of provincial legislatures, section 106(3)(c) similarly provides in relevant part:
“(3) A person loses membership of a provincial legislature if that person—
. . .
(c) ceases to be a member of the party that nominated that person as a member of the legislature, unless that member has become a member of another party in accordance with Schedule 6A.”
 Section 86(1) of the Constitution provides in relevant part:
“At its first sitting after its election, and whenever necessary to fill a vacancy, the National Assembly must elect a woman or a man from among its members to be the President.”
 Section 91 of the Constitution provides in relevant part:
“(3) The President—
(a) must select the Deputy President from among the members of the National Assembly;
(b) may select any number of Ministers from among the members of the Assembly; and
(c) may select no more than two Ministers from outside the Assembly.”
 Id at para 67.
 Id at para 68.
 Id at para 66.
 Id at para 68.
 Id at para 66.
 Id at para 67.
 In Bernstein and Others v Bester and Others NNO  ZACC 2; 1996 (2) SA 751 (CC); 1996 (4) BCLR 449 (CC) at para 67, this Court explained that an integrated approach to interpreting the right to privacy eschews “an abstract individualistic approach”. Because no right is absolute, “each right is always already limited by every other right accruing to another citizen”. Hence—
“[p]rivacy is acknowledged in the truly personal realm, but as a person moves into communal relations and activities such as business and social interaction, the scope of personal space shrinks accordingly.”
 Section 39.
 Section 36.
 Section 1(d).
 Section 16.
 Id at paras 122-3.
 In New National Party of South Africa v Government of the Republic of South Africa and Others  ZACC 5; 1999 (3) SA 191 (CC); 1999 (5) BCLR 489 (CC) at para 11, this Court observed that “the mere existence of the right to vote without proper arrangements for its effective exercise does nothing for a democracy; it is both empty and useless”.
 The principle of subsidiarity derives from Roman Catholic Canon Law, dating back to the First Vatican Council in 1869-70, where it entails that human affairs are handled best at the lowest possible level of management. See Murray “The Principle of Subsidiarity and the Church” (1995) Australasian Catholic Record 163 at 164-5 and 171. In the European Community, subsidiarity entails that Community organs should act only where action cannot be more effectively taken at Member State level. Subsidiarity thus tries to devolve as much power as possible to the constituent states. The principle seeks to recognise the diversity of national traditions with Europe, acknowledging that many matters are best dealt with below Community level. See Critchley Europe and Industry: The Integration of the European Union (e-book 1995 available at: http://mmu.academia.edu/PeterCritchley/Books) vol 1 at 117-34 and Sibanda “Beneath it all lies the Principle of Subsidiarity: The Principle of Subsidiarity in the African and European Regional Human Rights Systems” (2007) 40 Comparative and International Law Journal of Southern Africa 425 at 425 and 431. It is in this sense that the word “subsidiarity” is used in City of Johannesburg Metropolitan Municipality v Gauteng Development Tribunal and Others  ZAGPHC 30; 2008 (4) SA 572 (W);  2 All SA 298 (W) (Gauteng Development Tribunal) at para 53 and fn 68 (recording the argument that the sphere of government where the specific function would be most appropriate must inform the understanding of functional areas of concurrent constitutional competence).
 See Ex Parte Minister of Safety & Security & Others: In Re: S v Walters & Another  ZACC 6; 2002 (4) SA 613; 2002 (7) BCLR 663 (CC) at para 22 and Govender v Minister of Safety and Security 2000 (1) SA 959 (D) at 969C.
 See section 4(3) of the Implementation of the Rome Statute of the International Criminal Court Act 27 of 2002 providing for the jurisdiction of South African courts in respect of crimes committed under international criminal law as codified in the Rome Statute. See also the Arrest Warrant Case (Democratic Republic of the Congo v Belgium)  ICJ 3 at para 59, Separate Opinion of Judges Higgins, Kooijmans and Buergenthal, noting that a state contemplating bringing criminal charges based on universal jurisdiction “must first offer to the national state of the prospective accused person the opportunity itself to act upon the charges concerned”. See Langer “The Diplomacy of Universal Jurisdiction: The Political Branches and the Transnational Prosecution of International Crimes” (2011) 105 American Journal of International Law 1-49.
 National Commissioner of the South African Police Service v Southern African Litigation Centre and Another  ZACC 30; 2015 (1) SA 315 (CC); 2015 (1) SACR 255 (CC) at paras 61-4 and National Commissioner of the South African Police Service and Another v Southern African Human Rights Litigation Centre and Another  ZASCA 168; 2014 (2) SA 42 (SCA);  1 All SA 435 (SCA) at para 68. As this Court explained, ordinarily, there must be a substantial and true connection National Commissioner of the South African Police Service v Southern African Litigation Centre and Another  ZACC 30; 2015 (1) SA 315 (CC); 2015 (1) SACR 255 (CC) at paras 61-4between the subject-matter and the source of the jurisdiction. And once jurisdiction is properly founded, investigating international crimes committed abroad is permissible only if the country with jurisdiction is unwilling or unable to prosecute, and only if the investigation is confined to the territory of the investigating state.
 Schedule 4 of the Constitution, “Functional areas of concurrent national and provincial legislative competence”. See also Gauteng Development Tribunal above n 86; on appeal neither the Supreme Court of Appeal (City of Johannesburg Metropolitan Municipality v Gauteng Development Tribunal and Others  ZASCA 106; 2010 (2) SA 554 (SCA); 2010 (1) BCLR 157 (SCA)) nor this Court (City of Johannesburg Metropolitan Municipality v Gauteng Development Tribunal and Others  ZACC 11; 2010 (6) SA 182 (CC)) used the term “subsidiarity”.
 Nokotyana and Others v Ekurhuleni Metropolitan Municipality and Others  ZACC 33; 2010 (4) BCLR 312 (CC) (Nokotyana) at para 50 stating that “[w]here the Constitution contains both a specific right, and a more general right, it is appropriate first to invoke the specific right” which was quoted and applied in Grancy Property Limited v Gihwala  ZAWCHC 97; 2014 JDR 1292 (WCC) (Grancy) at para 198. The label “subsidiarity” is not used in Nokotyana, but is used in Grancy.
 Section 101(5) of the interim Constitution.
 Section 98(2) of the interim Constitution.
 S v Mhlungu and Others  ZACC 4; 1995 (3) SA 867 (CC); 1995 (7) BCLR 793 (CC) at para 59 approved in Zantsi v Council of State, Ciskei and Others  ZACC 9; 1995 (4) SA 615 (CC); 1995 (10) BCLR 1424 (CC) at para 3.
 See Du Plessis “‘Subsidiarity’: What’s in the Name for Constitutional Interpretation and Adjudication?” ( 2006) Stellenbosch Law Review 207-31, where Du Plessis calls the use of the word in this context “adjudicative subsidiarity”.
 Section 39(2) of the Bill of Rights.
 See Klare “Legal Subsidiarity and Constitutional Rights: A response to AJ van der Walt” (2008) Constitutional Court Review Vol 1 2008 129 at 140, in which he aptly notes:
“When Parliament ‘gives effect’ to a constitutional right it may task itself with giving the right an enforceable floor of protections and implementations. In practice, it may also erect a ceiling and walls around the right. At a certain point ‘giving effect’ to a constitutional right slides into defining the right by setting out its metes and bounds.”
Thus, he continues:
“The constitutional adequacy of the relief afforded to an effect giving statute is not, strictly speaking, a question subsidiarity theory addresses – it is a substantive problem of constitutional law that must be decided by the courts.”
 The doctrine’s emergence may be traced through the decisions of this Court. In Member of the Executive Council for Development Planning and Local Government, Gauteng v Democratic Party and Others  ZACC 9; 1998 (4) SA 1157 (CC); 1998 (7) BCLR 855 (CC) at para 62, where the claimant relied on the invalidity of a statutory provision as the basis for claiming relief, but omitted to seek a declaration that it was invalid, Yacoob J on behalf of the Court pointed out that “considerable difficulties stand in the way of the adoption of a procedure which allows a party to obtain relief which is in effect consequent upon the invalidity of an Act of Parliament without any formal declaration of invalidity of that provision”.
Ingledew v Financial Services Board  ZACC 8; 2003 (4) SA 584 (CC); 2003 (8) BCLR 825 (CC) (Ingledew) at para 22, noted this finding, but held that it was not directly on point. More directly, Ingledew noted at paras 24 and 29 that NAPTOSA and Others v Minister of Education, Western Cape and Others  ZAWCHC 9; 2001 (2) SA 112 (C) (NAPTOSA) and other cases had “cast doubt on the correctness of the proposition that a litigant can rely upon the Constitution, where there is a statutory provision dealing with the matter without challenging the constitutionality of the provision concerned” but left the question open.
Bato Star Fishing (Pty) Ltd v Minister of Environmental Affairs and Tourism and Others  ZACC 15; 2004 (4) SA 490 (CC); 2004 (7) BCLR 687 (CC) (Bato Star) at paras 21-6 followed. It was the first decision to give explicit recognition to the doctrine of subsidiarity, though the word was not used.
In Minister of Health and Another v New Clicks South Africa (Pty) Ltd and Others  ZACC 14; 2006 (2) SA 311 (CC); 2006 (1) BCLR 1 (CC) (New Clicks) the judgments of Chaskalson CJ and Ngcobo J alluded to the principle. Ngcobo J now endorsed NAPTOSA at paras 436-7.
In South African National Defence Union v Minister of Defence and Others  ZACC 10; 2007 (5) SA 400; 2007 (8) BCLR 863 (CC) (SANDU), a unanimous Court held that the approach in NAPTOSA and New Clicks was correct.
In MEC for Education, Kwa-Zulu Natal and Others v Pillay  ZACC 21; 2008 (1) SA 474 (CC) (Pillay) at para 40, Langa CJ, on behalf of the majority, citing New Clicks, SANDU and NAPTOSA, upheld the principle.
In Mbatha v University of Zululand  ZACC 43; (2014) 35 ILJ 349 (CC); 2014 (2) BCLR 123 (CC) at para 172, Jafta J (Moseneke DCJ and Nkabinde J concurring) alluded to “the principle of constitutional subsidiarity” in holding that a claim by an applicant, who alleged he was employed by the respondent, engaged a constitutional issue because he asserted a breach of the duty to pay his salary, a right enshrined in a statute, the Basic Conditions of Employment Act 75 of 1997, which was enacted to fulfil a constitutional right, the right to fair labour practices.
In Sali v National Commissioner of the South African Police Service and Others  ZACC 19; (2014) 35 ILJ 2727 (CC); 2014 (9) BCLR 997 (CC) at para 2 and fn 2, Jafta J, in a minority judgment, pointed out that “[w]here there is legislation giving effect to a right in the Bill of Rights, a claimant is not permitted to rely directly on the Constitution” (footnoting that this is known as the principle of subsidiarity). Jafta J also pointed out that because section 6(1) of the Employment Equity Act 55 of 1998 gives effect to section 9(3) of the Bill of Rights, the applicant was not permitted to rely directly on the Constitution.
 Section 27 of the Bill of Rights provides:
“(1) Everyone has the right to have access to—
(a) health care services, including reproductive health care;
(b) sufficient food and water; and
(c) social security, including, if they are unable to support themselves and their dependents, appropriate social assistance.
(2) The state must take reasonable legislative and other measures, within its available resources, to achieve the progressive realisation of each of these rights.
(3) No one may be refused emergency medical treatment.”
 108 of 1997.
 Mazibuko above n 101 at para 73.
 Id. Even though the applicants challenged the City’s water policy as unreasonable, they omitted to attack the regulation. Since the Court found that the policy was in any event not unreasonable, it did not have to decide whether the principle applied.
 Section 27(1)(b) of the Bill of Rights above n 102.
 Mazibuko is very different from this case. First, there was no challenge to the validity of existing legislation. Second, it invoked no express obligation on a specific organ of state – Parliament – to enact national legislation. Section 27(1)(b) of the Bill of Rights does not contain an obligation of this sort.
 Act 66 of 1995.
 Section 23(1) provides that “[e]veryone has the right to fair labour practices”.
 NAPTOSA above n 100 at para 123.
 Id at para 100.
 New Clicks above n 100 at para 436.
 SANDU above n 100 at para 51.
 Section 23(5) provides:
“Every trade union, employers' organisation and employer has the right to engage in collective bargaining. National legislation may be enacted to regulate collective bargaining. To the extent that the legislation may limit a right in this Chapter, the limitation must comply with section 36(1).”
 SANDU above n 100 at para 51.
 Id at para 52.
 4 of 2000.
 New Clicks above n 100.
 SANDU above n 100.
 NAPTOSA above n 100.
 Pillay above n 100 at para 40.
 Bato Star above n 100.
 3 of 2000 (PAJA).
 In Bato Star above n 100 at para 26, this Court wrote: “[t]o the extent, therefore, that neither the High Court nor the SCA considered the claims made by the applicant in the context of PAJA, they erred”.
 Id above n 100 at para 25.
 Id. The Court added:
“It is not necessary to consider here causes of action for judicial review of administrative action that do not fall within the scope of PAJA. As PAJA gives effect to section 33 of the Constitution, matters relating to the interpretation and application of PAJA will of course be constitutional matters.”
 New Clicks above n 100.
 Id at para 96.
 Id at para 436.
 Id at para 437.
 Id at para 4, citing Mazibuko, Pillay, SANDU and Bato Star above n 100.
 New Clicks above n 100 at paras 96. Hence Chaskalson CJ quoted Hoexter at para 97, relying on the constitutional scheme itself, where it was asserted that it “follows logically from the fact that the PAJA gives effect to the constitutional rights”. Therefore, PAJA cannot simply be circumvented by resorting directly to section 33.
 Id at para 437.
 SANDU above n 100 at para 52, which Langa CJ quoted with approval in Pillay above n 100 at para 40.
 NAPTOSA above n 100 at 123B-C, endorsed by Ngcobo J in New Clicks above n 100 at para 436.
 IDASA above n 12 at para 17.
 Id at para 19.
 Doctors for Life above n 6 at para 16.
 Mazibuko above n 101 at para 73.
 See  to  above. Subsidiarity finds its clearest application in two groups of cases. The first is when the legislation evinces a purpose to codify a right in the Bill of Rights. Instances already established in the jurisprudence of this Court are administrative justice and PAJA, and fair employment rights and the LRA. This Court has held that both of these statutes must be resorted to first, and that, absent an invalidity challenge, the litigant cannot invoke the Constitution.
The second clear instance is where Parliament adopts legislation in fulfilment of social and economic rights. The Bill of Rights obliges the state to take “reasonable legislative and other measures” to fulfil these rights: see sections 24(b), 25(5), 26(2) and 27(2) of the Constitution. In the foreground here is the institutional deference the principle accords to the state or Parliament and their task in the scheme of the Constitution. The reason is plain. Parliament, as part of the state, has adopted a legislative measure it considers reasonable in order to fulfil a social and economic right. The judicial branch owes it to Parliament to require a litigant seeking to enforce the right to rely on the legislation first, or establish through constitutional attack that its legislative measure is not reasonable. For this reason, unless a litigant successfully attacks Parliament’s judgment in enacting the legislation, the legislation must stand as the basis for enforcing the right.
 Pillay and SANDU above n 100.
 The applicant squarely attacks the constitutional breadth of PAIA on the basis that Parliament has failed to fulfil the section 32(2) obligation because it has not enacted national legislation to require transparency in the private funding of political parties. It asserts that PAIA gives effect “only to one aspect of the right of access to information, namely the right to gain access, upon specific request, to specific records held by specific bodies at specific times”. The legislation Parliament has failed to enact is fundamentally different from PAIA – it would replace unregulated secrecy (which PAIA permits) with regulated transparency (which PAIA fails to do). In applying alternatively for direct access, the applicant points out that in IDASA the High Court held that PAIA does not enable citizens to access parties’ private funding records.
In response to this attack, the Speaker’s opposing affidavit squarely defends the constitutional adequacy of PAIA. From the outset, Parliament’s defence is that PAIA “fully satisfies the requirements of section 32(2)”. The Speaker unequivocally supports the finding in IDASA that PAIA does not give access to information about the private funding of political parties, since that information is not required for the exercise or protection of any right. But even if the applicant is correct that this information is required for rights-protection, the Speaker says, this just means that PAIA in fact does afford the required access. Under the heading “PAIA is adequate”, the Speaker asserts that “there already exists legislation that gives effect to the very right that the applicant claims to champion”, namely PAIA. A central theme of the opposing affidavit is the Speaker’s insistence that the applicant is wrong that PAIA “does not enable citizens to access the records of the private funding of political parties”. On the contrary, legislation already exists, in the form of PAIA, for the purpose of requiring political parties to disclose who their private funders are. PAIA is precisely the legislation that requires disclosure of parties’ private funding, though that information is not “required” (as IDASA rightly found) to effectively exercise the right to vote. PAIA is an adequate constitutional tool by which accurate information of the sort the applicant seeks can be obtained. IDASA did not dismiss the application on the basis that the relief sought was incompatible with PAIA – it rejected the section 19 political rights argument. The applicant has thus failed to show any inadequacy in PAIA.
The core proposition of the applicant’s written argument is that Parliament has failed to fulfil its section 32(2) obligation because PAIA does not require disclosure of donations to political parties and without specific legislation regulating the creation and disclosure of such records, the applicant can never obtain the relief it seeks. The applicant attacks PAIA’s distinction between public and private bodies and the uses to which the Speaker puts it in characterising political parties’ obligations. Under the heading: “PAIA does not require disclosure of donations to political parties,” the applicant attacks PAIA’s ambit in detail. It inveighs against PAIA’s limitations as to “records”, its confidentiality exemptions, its confinement to records requested, and the impossible evidentiary burden created by the mandatory disclosure override provisions of section 70. Parliament’s written argument relies on the subsidiarity principle to characterise the applicant’s reading of PAIA as “conveniently self-defeatist” and insists that PAIA gives “ample effect to the right of access to information”. If IDASA was wrong in finding either that political parties are private, or that the information sought is not required for the exercise of the section 19 rights, then the information is accessible under PAIA. Whether political parties are private actors or an extension of the state, PAIA is the legislation Parliament enacted to give effect to the right of access to political parties’ private funding for the exercise of the right to vote. Nowhere in the answering affidavit does the Speaker say that PAIA does not permit disclosure of parties’ private donations – she says only that section 19 does not confer that right. The applicant’s argument on “records” under PAIA is incorrect, because the applicant does not aver that political parties do not keep records of private funding.
 Section 172(2)(a) provides:
“The Supreme Court of Appeal, the High Court of South Africa or a court of similar status may make an order concerning the constitutional validity of an Act of Parliament, a provincial Act or any conduct of the President, but an order of constitutional invalidity has no force unless it is confirmed by the Constitutional Court.”
 Doctors for Life above n 6 at para 18.
 Id at para 16.
 See the majority judgment at ,  and .
 Majority judgment at  and .
 See section 56 of the Higher Education Act 101 of 1997; section 110 of the Labour Relations Act 66 of 1995; sections 2 and 7 of the Legal Deposit Act 54 of 1997; section 71 of the Prevention of Organised Crime Act 121 of 1998; section 142 of the National Water Act 36 of 1998; section 31 of the Nuclear Energy Act 46 of 1999; sections 21, 28 and 30 of the Mineral and Petroleum Resources Development Act 28 of 2002; section 3 of the Collective Investment Schemes Control Act 45 of 2002; section 2 of the Home Loan and Mortgage Disclosure Act 63 of 2000 (whose Preamble expressly alludes to section 32(1) of the Bill of Rights); section 72 of the National Credit Act 34 of 2005; sections 31 and 187 of the Companies Act 71 of 2008; as well as section 2 of the Protection of Personal Information Act 4 of 2013.
 This, again, the applicant made expressly clear in its founding affidavit:
“Parliament’s obligation under section 32(2) of the Constitution did not begin and end with the enactment of PAIA. PAIA gives effect to only one aspect of the right of access to information. . . . The required legislation as set out in this affidavit is fundamentally different from PAIA in its nature and purpose.”
The applicant also quoted approvingly IDASA’s finding at para 58 that “private donations to political parties ought to be regulated by way of specific legislation”. Finally, it noted its application in terms of section 167(4)(e) is entirely separate from one that “test[s] ‘the constitutional validity of an Act of Parliament’” under section 172, as there is simply no Act of Parliament to test.
 Majority judgment at .
 Majority judgment at  to .
 See section 172(2)(a) of the Constitution above n 145.
 See sections 1, 11 and 50 of PAIA.
 Id at sections 11, 18 and 50. Section 15 allows for the “automatic” availability of certain records, but this is exceptional.
 Id at section 18(2)(a)(i).
 Section 15 of PAIA does provide for departments to make categories of records automatically available.
 By contrast, PAJA, which codifies the right to just administrative action, purports to contain a full definition of administrative action. Section 1 of PAJA envisages an exhaustive, careful and minutely detailed definition. In addition, PAJA defines “decision”. There is no administrative action, and there are no administrative decisions, to which PAJA doesn’t apply.
 Section 1 of PAIA provides that, unless the context otherwise indicates, “record” of, or in relation to a public or private body, means—
“any recorded information—
(a) regardless of form or medium;
(b) in the possession of or under the control of that public or private body, respectively; and
(c) whether or not it was created by that public or private body, respectively”.
 Section 21 of PAIA, entitled “Preservation of records until final decision on request” provides:
“If the information officer of a public body has received a request for access to a record of the body, that information officer must take the steps that are reasonably necessary to preserve the record, without deleting any information contained in it, until the information officer has notified the requester concerned of his or her decision in terms of section 25 and—
(a) the periods for lodging an internal appeal, an application with a court or an appeal against a decision of that court have expired; or
(b) that internal appeal, application or appeal against a decision of that court or other legal proceedings in connection with the request has been finally determined, whichever is the later.”
 See, for instance, sections 1, 11(2), 16, 20(1), 21, 28(1) 34-5, 37, 39, 41-3, 59, 63, 64(1)(b) and (2), 65, 68(1)(b) and (c), 69(1) and (2) and 83(1).
 Id at section 1.
 Unless the context indicates otherwise, “organ of state” is defined in section 239 of the Constitution as—
“(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,
but does not include a court or a judicial officer”.
 PAJA provides that, in certain circumstances, private entities may also for its purposes be organs of state. See Allpay Consolidated Investment Holdings (Pty) Ltd and Others v Chief Executive Officer of the South African Social Security Agency and Others  ZACC 12; 2014 (4) SA 179 (CC); 2014 (6) BCLR 641 (CC) (Allpay No 2) at paras 52-3. Section 32 of the Bill of Rights, too, envisages overlaps between private and public entities in a way that PAIA’s definitions do not.
 Section 8(2). See also section 8(3) which provides that:
“When applying a provision of the Bill of Rights to a natural or juristic person in terms of subsection (2), a court—
(a) in order to give effect to a right in the Bill, must apply, or if necessary develop, the common law to the extent that legislation does not give effect to that right; and
(b) may develop rules of the common law to limit the right, provided that the limitation is in accordance with section 36(1).”
Further, section 8(4) provides that:
“A juristic person is entitled to the rights in the Bill of Rights to the extent required by the nature of the rights and the nature of that juristic person.”
 Section 38, entitled “Enforcement of rights”, provides:
“Anyone listed in this section has the right to approach a competent court, alleging that a right in the Bill of Rights has been infringed or threatened, and the court may grant appropriate relief, including a declaration of rights. The persons who may approach a court are—
(a) anyone acting in their own interest;
(b) anyone acting on behalf of another person who cannot act in their own name;
(c) anyone acting as a member of, or in the interest of, a group or class of persons;
(d) anyone acting in the public interest; and
(e) an association acting in the interest of its members.”
 Giant Concerts CC v Rinaldo Investments (Pty) Ltd and Others  ZACC 28; 2013 (3) BCLR 251 (CC) at paras 36-7 and Ferreira v Levin NO and Others; Vryenhoek and Others v Powell NO and Others  ZACC 13; 1996 (1) SA 984 (CC); 1996 (1) BCLR 1 (CC) at para 165.
 See sections 25(6) and (7) and 31 of the Bill of Rights (“community”) and section 38(c) (“group or class”).
 Above n 168. Section 32 thus acknowledges overlaps between private and public in a way that PAIA’s definitions do not.
 Thus, section 8 of PAIA, entitled “Part applicable when performing functions as public or private body”, provides:
“(1) For the purposes of this Act, a public body referred to in paragraph (b) (ii) of the definition of ‘public body’ in section 1 or a private body—
(a) may be either a public body or a private body in relation to a record of that body; and
(b) may in one instance be a public body and in another instance be a private body depending on whether that record relates to the exercise of a power or performance or a function as a public body or as a private body.
(2) A request for access to a record held for the purpose or with regard to the exercise of a power or the performance of a function—
(a) as a public body must be made in terms of section 11; or
(b) as a private body, must be made in terms of section 50.
(3) The provisions of Parts 1, 2, 4, 5, 6 and 7 apply to a request for access to a record that relates to a power or function exercised or performed as a public body.
(4) The provisions of Parts 1, 3, 4, 5, 6 and 7 apply to a request for access to a record that relates to a power or function exercised or performed as a private body.”
 Section 1 of PAIA.
 Id at section 50(1)(a).
 There are three categories of juristic persons: associations established by separate legislation; associations incorporated in terms of special or enabling legislation; and associations that comply with the common law requirements for establishment of juristic persons. The common law requires that the association remains in existence irrespective of a change in membership, functions as a bearer of rights, duties, and capacities separate from its individual members, and its object is not for the acquisition of gain (if so, it must register as a company). See Kruger and Skelton (eds) The Law of Persons in South Africa (OUP, Cape Town 2010) at 17-8.
All three categories of juristic persons require the community, association, or party to have rights and powers different from the individuals that make the community, association, or party. See Wilken v Brebner and Others 1935 AD 175 at 182.
This Court held that “[i]t is trite that a company is a legal entity altogether separate and distinct from its members, that its continued existence is independent of the continued existence of its members, and that its assets are its exclusive property.” (Footnote omitted). See First National Bank of SA Ltd t/a Wesbank v Commissioner, South African Revenue Services and Another; First National Bank of SA Ltd t/a Wesbank v Minister of Finance  ZACC 5 (CC); 2002 (4) SA 768 (CC); 2002 (7) BCLR 702 (CC) at para 43.
In Weare and Another v Ndebele N.O and Others  ZACC 20; 2009 (1) SA 600 (CC); 2009 (4) BCLR 370 (CC) at para 53 this Court held that “[t]he most relevant characteristics of a juristic person are its separate legal personality and the limited liability of the natural persons involved”.
 51 of 1996.
 Id at section 1.
 Id at section 15(1) reads:
“The chief electoral officer shall, upon application by a party in the prescribed form, accompanied by the items mentioned in subsection (3), register such party in accordance with this Chapter.”
 Id at section 16 specifies the “prohibition on registration of [a] party under certain circumstances”. The legal form and nature of the entity is not among them. Section 16 reads:
“(1) The chief electoral officer may not register a party in terms of section 15 or 15A, if—
(a) fourteen days have not elapsed since the applicant has submitted to the chief electoral officer proof of publication of the prescribed notice of application in the Gazette in the case of an application referred to in section 15 or in a newspaper circulating in the municipal area concerned in the case of an application referred to in section 15A.
(b) a proposed name, abbreviated name, distinguishing mark or symbol mentioned in the application resembles the name, abbreviated name, distinguishing mark or symbol, as the case may be, of any other registered party to such an extent that it may deceive or confuse voters; or
(c) a proposed name, abbreviated name, distinguishing mark or symbol mentioned in the application or the constitution of the party or the deed of foundation mentioned in section 15 or 15A contains anything—
(i) which portrays the propagation or incitement of violence or hatred or which causes serious offence to any section of the population on the grounds of race, gender, sex, ethnic origin, colour, sexual orientation, age, disability, religion, conscience, belief, culture or language; or
(ii) which indicates that persons will not be admitted to membership of the party or welcomed as supporters of the party on the grounds of their race, ethnic origin or colour.
(2) Any party which is aggrieved by a decision of the chief electoral officer to register or not to register a party, may within 30 days after the party has been notified of the decision, appeal against the decision to the Commission in the prescribed manner.
(3) The Commission shall in the case of such an appeal enquire into or consider the matter and may, subject to subsection (4), confirm or set aside the decision of the chief electoral officer.
(4) In considering such an appeal against the refusal to register a party in terms of subsection (1)(a) the Commission—
(a) shall take into account the fact that the party which is associated with the name, abbreviated name, distinguishing mark or symbol, as the case may be, for the longest period, should prima facie be entitled thereto;
(b) may, for the purposes of paragraph (a)—
(i) afford the parties concerned an opportunity to offer such proof, including oral evidence or sworn or affirmed statements by any person which, in the opinion of the Commission, could be of assistance in the expeditious determination of the matter; and
(ii) administer an oath or affirmation to any person appearing to testify orally before it.”
 Ramakatsa above n 68.
 Section 19(1)(b) of the Constitution.
 Ramakatsa above n 68 at paras 14-6 where Yacoob J states, with concurrence by all members of the Court, that—
“the Constitution confers upon all citizens the right to participate in the activities of a political party. The appellants contended in the application for leave to appeal that this right has been denied to them or has been infringed because the irregularities that were complained of went so far as to prevent them from participating in the activities of the ANC appropriately and properly. Their argument was that their right to participate in a political party included a right to be governed by properly elected members of the ANC in the province.
The system of proportional representation provided for in our Constitution means that a political party is entitled to representation in Parliament in proportion to the number of votes it obtains in an election relative to the total number of votes cast. In other words, of the 400 members of the National Assembly, a political party that succeeds in securing the vote of, say, 60% of the electorate will have 240 of 400 seats in the National Assembly.
I do not think that the Constitution could have contemplated political parties could act unlawfully. On a broad purposive construction, I would hold that the right to participate in the activities of a political party confers on every political party the duty to act lawfully and in accordance with its own constitution. This means that our Constitution gives every member of every political party the right to exact compliance with the constitution of a political party by the leadership of that party.”
 Ramakatsa above n 68 at para 68.
 IDASA above n 12 at para 23 held that, for purposes of their donations records, political parties are not “public bodies”, but “private bodies”, as defined in PAIA. To the extent that this conclusion overlooks considerations set out in the text, it appears to me mistaken.
 IDASA id.
 See above at .
 Doctors for Life above n 6 at para 27.
 Id at para 124.
 The “failure” is, of course, said to be in the limited sense that there is no legislation that makes it possible for citizens to have access to information on the private funding of political parties for the purpose of exercising the right to vote contained in section 19(3) of the Constitution and that no national legislation currently requires this information to be publicly accessible.
 Minority judgment at .
 We use “alleged” consciously because at this stage we have no idea what results a challenge of the constitutional validity of PAIA might yield.
 Above n 4.
 Our preference for what the minority judgment terms, at , a “sparser” route is, in no way a divergence from the jurisprudence laid down by this Court on the interplay between sections 172 and 167(4) of the Constitution. In this regard, see Women’s Legal Centre Trust above n 42 at paras 11-20; Doctors for Life above n 6 at para 21 and SARFU above n 41.
 Chirwa above n 198 at para 155.
 Independent Newspapers (Pty) Ltd v Minister for Intelligence Services: In re: Masetlha v President of the Republic of South Africa and Another  ZACC 6; 2008 (5) SA 31 (CC); 2008 (8) BCLR 771 (CC) (Independent Newspapers) at para 23.
 Id at para 156.
 PFE International above n 131 at para 4.
“As far as section 32 of the Constitution is concerned, the applicant’s counsel did not provide any authority for the proposition that the applicant is entitled to simply rely on this section in the Constitution, and ignore the provisions of PAIA – which was enacted to give effect to section 32 of the Constitution.”
And Koalane and Another v Senkhe and Others  ZAFSHC 165 at para 7 where the High Court said:
“The national legislation envisaged in section 32(2) is [PAIA]. It is clear from the long title, the preamble and section 9 of PAIA that the object thereof is to give effect to the constitutional right to access to information in terms of both section 32(1)(a) and (b).”
 See majority judgment at  below.
 Section 33 of the Constitution provides:
“(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.”
 Bato Star above n 100.
 In First Certification judgment above n 64, the transitional provisions were tested and this Court, at paras 83 and 86, held that—
“[t]he transitional measure is obviously a means of affording Parliament time to provide the necessary legislative framework for the implementation of the right to information. Freedom of information legislation usually involves detailed and complex provisions defining the nature and limits of the right and the requisite conditions for its enforcement.
. . .
The Legislature is far better placed than courts to lay down the practical requirements for the enforcement of the right and the definition of its limits.” (Footnotes omitted.)
 PAIA was assented to on 2 February 2000 (Promotion of Access to Information Act 2 of 2000, GN 95 GG 20852, 3 February 2000) and PAJA was assented to on 3 February 2000 (Promotion of Administrative Justice Act 3 of 2000, GN 96 GG 20853, 3 February 2000).
 See First Certification judgment above n 64 at paras 82-3 and 86. See also item 23(3) in Schedule 6 which provides:
“Sections 32(2) and 33(3) of the new Constitution lapse if the legislation envisaged in those sections, respectively, is not enacted within three years of the date the new Constitution took effect.”
 Minority judgment at .
 We choose a few examples to illustrate this point. In the main, the Higher Education Act 101 of 1997 deals with the regulation of higher education. The National Water Act 36 of 1998 principally concerns the reform of the law relating to water resources. And the Nuclear Energy Act 49 of 1999 regulates the acquisition, possession, importation, exportation and the use of nuclear materials, the discarding of radioactive waste and the storage of irradiated nuclear fuel.
 First Certification judgment above n 64 at para 109.
 Doctors for Life above n 6 at para 38.
 Here, we are referring to—
(a) the point that the definition of “record” is deficient in certain specified respects and the related issue to the effect that “information” is not defined (minority judgment at  to  and  to );
(b) the conclusion that there is a lack of a proactive duty to preserve records (minority judgment at ); and
(c) the deduction that the definitions of “private body” and “public body” do not accommodate political parties (minority judgment at  to ).
 This is what the minority judgment refers to as a “pairwise” relationship (minority judgment at  to  and ).
 Compare Women’s Legal Centre Trust above n 42 at para 24; Doctors for Life above n 6 at para 37; and First Certification judgment above n 64 at paras 106-13.
 Minority judgment at  to  and .
 See, for example, minority judgment at  to ,  and  to .
 Id at  to .
 Id at .
 Id at .
 Id. New Clicks above n 100 at para 96.
 Id at .
 Id at . NAPTOSA above n 100 at para 123B-C.
 See decisions of this Court referred to in the minority judgment, above n 100.
 Mazibuko above n 101 at para 73. See also Mbatha above n 100 at para 173, where Jafta J said:
“[W]here legislation has been passed to give effect to a right in the Bill of Rights, a litigant is not permitted to rely directly on the Constitution for its cause of action.”
 These being: as to the confinement of its application to “records”; its confidentiality exceptions; the fact that it would apply unequally and arbitrarily; and the impossible evidentiary burden imposed on a requester by section 70 of PAIA.
 SANDU above n 100 at para 52.
 Id at para 51.
 New Clicks above n 100 at para 437.
 Van der Walt Property and Constitution (Pretoria University Law Press, Pretoria 2012) at 40.
 Democratic Party above n 150. We render a lengthy summary of, and quote extensively from, this judgment because we consider what it said to be instructive.
“All questions concerning matters mentioned in subsection (2) are determined by a decision taken by a Municipal Council with a supporting vote of a majority of its members.”
Section 160(2)(b) provides in part:
“The following functions may not be delegated by a Municipal Council:
. . .
(b) the approval of budgets”.
 Section 16(5) of the Transition Act reads:
“(a) [A]ny resolution of any transitional council or transitional metropolitan substructure referred to in subsection (1) pertaining to the budget of such transitional council or transitional metropolitan substructure shall be taken by a two-thirds majority of the members of such council or substructure, and any resolution of any transitional council or transitional metropolitan substructure pertaining to town planning shall be taken by a majority of the members of such council or substructure: Provided that any such transitional council or transitional metropolitan substructure may delegate the power to take any decision on any matter pertaining to town planning to the committee referred to in subsection (6) or to any other committee appointed for this purpose; and
(b) if such transitional council or transitional metropolitan substructure—
(i) on the last day of June in any financial year has failed to approve a budget for the subsequent financial year; or
(ii) on the last day of April in any financial year has failed to take steps to prepare a budget for the subsequent financial year,
the [MEC] may exercise any power or perform any duty conferred or imposed upon such transitional council or transitional metropolitan substructure by this Act or any other law in relation to the approval or preparation of a budget, as the case may be.”
 This Schedule provided:
“Section 245(4) of the previous Constitution continues in force until the application of that section lapses. Section 16(5) and (6) of the Local Government Transition Act, 1993, may not be repealed before 30 April 2000.”
 Democratic Party above n 150 at para 60.
 Id. Section 172(2) of the Constitution stipulates in full:
“(a) The Supreme Court of Appeal, a High Court or a court of similar status may make an order concerning the constitutional validity of an Act of Parliament, a provincial Act or any conduct of the President, but an order of constitutional invalidity has no force unless it is confirmed by the Constitutional Court.
(b) A court which makes an order of constitutional invalidity may grant a temporary interdict or other temporary relief to a party, or may adjourn the proceedings, pending a decision of the Constitutional Court on the validity of that Act or conduct.
(c) National legislation must provide for the referral of an order of constitutional invalidity to the Constitutional Court.
(d) Any person or organ of state with a sufficient interest may appeal, or apply, directly to the Constitutional Court to confirm or vary an order of constitutional invalidity by a court in terms of this subsection.”
 Democratic Party above n 150 at para 60.
 Id at paras 61-4.
 Id at para 61.
 Compare id. In that matter, the applicant was seeking relief without any formal declaration of the invalidity of section 16(5) of the Transition Act.
 See below at  to , when we deal with the role played by a Minister responsible for the administration of an Act that is the subject of a frontal challenge.
 Minority judgment at  to .
 John v Rees and Others; Martin and Another v Davis and Others; Rees and Another v John  Ch 345 at 402D, quoted with approval in Administrator, Transvaal and Others v Zenzile and Other  ZASCA 108; 1991 (1) SA 21 (A) (Zenzile) at 37E-F. Zenzile was in turn quoted with approval by this Court in Lufuno Mphaphuli & Associates (Pty) Ltd v Andrews and Another  ZACC 6; 2009 (4) SA 529 (CC); 2009 (6) BCLR 527 (CC) at para 154.
 This complaint is founded on the definition of “record”, which PAIA defines as—
“of, or in relation to, a public or private body, means any recorded information—
(a) regardless of form or medium;
(b) in the possession or under the control of that public or private body, respectively; and
(c) whether or not it was created by that public or private body, respectively.”
 This complaint is based on section 65 of PAIA, which provides:
“The head of a private body must refuse a request for access to a record of the body if its disclosure would constitute an action for breach of a duty of confidence owed to a third party in terms of an agreement.”
 SAPS v Solidarity obo Barnard  ZACC 23; 2014 (6) SA 123 (CC); 2014 (10) BCLR 1195 (CC) (Barnard) at para 204. See also Director of Hospital Services v Mistry 1979 (1) SA 626 (A) at 635F-636A.
 Barnard id at para 218; Maphango and Others v Aengus Lifestyle Properties (Pty) Ltd  ZACC 2; 2012 (3) SA 531 (CC); 2012 (5) BCLR 449 (CC) at para 109; Barkhuizen v Napier  ZACC 5; 2007 (5) SA 323 (CC); 2007 (7) BCLR 691 (CC) at para 39; Alexkor Ltd and Another v Richtersveld Community and Others  ZACC 18; 2004 (5) SA 460 (CC); 2003 (12) BCLR 1301 (CC) at para 43; Carmichele v Minister of Safety and Security  ZACC 22; 2001 (4) SA 938 (CC); 2001 (10) BCLR 995 (CC) at para 31; Fischer and Another v Ramahlele and Others  ZASCA 88; 2014 (4) SA 614 (SCA) at paras 13-8; Cole v Government of the Union of South Africa 1910 263 (A) at 272 and Kannenberg v Gird 1966 (4) SA 173 (C) at 182A.
 Above n 150.
 Majority judgment at  to .
 Minority judgment at  to .
 In Mazibuko above n 101 at paras 73-4, O’Regan J held that the subsidiarity principle may not apply. This was so because the constitutional obligation imposed on government by section 7(2) is to take reasonable legislative and other measures to achieve the right. She nonetheless held that it was not necessary to decide the question in the circumstances of the case.
 Minority judgment at  to .
 Tongoane and Others v National Minister for Agriculture and Land Affairs and Others  ZACC 10; 2010 (6) SA 214 (CC); 2010 (8) BCLR 741 (CC) (Tongoane). See also Mabaso v Law Society of the Northern Provinces  ZACC 8; 2005 (2) SA 117 (CC); 2005 (2) BCLR 129 (CC) at para 12.
 Tongoane id at para 120. See also rule 5 of the Rules of this Court, which provides:
“(1) In any matter, including any appeal, where there is a dispute over the constitutionality of any executive or administrative act or conduct or threatened executive or administrative act or conduct, or in any inquiry into the constitutionality of any law, including any Act of Parliament or that of a provincial legislature, and the authority responsible for the executive or administrative act or conduct or the threatening thereof or for the administration of any such law is not cited as a party to the case, the party challenging the constitutionality of such act or conduct or law shall, within five days of lodging with the Registrar a document in which such contention is raised for the first time in the proceedings before the Court, take steps to join the authority concerned as a party to the proceedings.
(2) No order declaring such act, conduct or law to be unconstitutional shall be made by the Court in such matter unless the provisions of this rule have been complied with.”
 Now the Minister of Justice and Correctional Services. See Transfer of Administration and Powers and Functions Entrusted by Legislation To Certain Cabinet Members in Terms of Section 97 of the Constitution, GN 47 GG 37839, 12 July 2014, signed by the President transferring the administrative, powers and functions “entrusted by legislation to certain cabinet members in terms of section 97 of the Constitution.” This Proclamation also had the effect of changing the names of certain government ministries.
[Context] [Hide Context]