Part 5: Effect of the Public Audit Act 2001 ― new public entities

Central government: Results of the 2005/06 audits.

The Public Audit Act 2001 (the Act) resulted in a clearer definition of the Auditor-General’s mandate. The Auditor-General is the auditor of every public entity, and of any entity controlled by one or more public entities under the test for “control” contained in the Act (the control test).

We have examined and made a decision on the status of about 700 entities since the Act was passed. The application of the control test has increased the number of entities audited by the Auditor-General by about 500. Many of these new public entities are trusts associated with public entities. This has caused concerns for some trusts that have not previously been subject to public audit.

In this Part, we highlight some issues that have arisen in applying the control test in the central government sector. A small number of trusts have not yet accepted that they are public entities subject to the Auditor-General’s mandate. We consider it important to advise Parliament that we are not auditing a small number of entities that we consider should be subject to public audit.

The control test

Under section 5 of the Act, the Auditor-General is the auditor of every public entity and of every entity that is controlled by one or more public entities.

The Act uses both legal and accounting definitions of control. Section 5(2) says that an entity is controlled by one or more other entities if:

(a) the entity is a subsidiary of any of those other entities; or

(b) the other entity or entities together control the entity within the meaning of any relevant approved financial reporting standard; or

(c) the other entity or entities can together control directly or indirectly the composition of the board of the entity within the meaning of sections 7 and 8 of the Companies Act 1993 (which, for the purposes of this paragraph, are to be read with all necessary modifications).

The two legal limbs of the control test in paragraphs (a) and (c) above are reasonably straightforward. The definition in paragraph (a) applies where a public entity owns a majority of shares of an incorporated subsidiary and/or has the right to appoint a majority of directors. The definition in paragraph (c) applies where one or more public entities have the right, directly or indirectly, to appoint a majority of the governing body of an entity (whether incorporated or not).

Analysis of control under the accounting test in paragraph (b) is often more difficult. This Part focuses on some of the issues that have arisen in applying the accounting test for control.

Control under the accounting test

The relevant approved financial reporting standard, for the purpose of the control test, is Financial Reporting Standard No. 37: Consolidating Investments in Subsidiaries (FRS-37).1 We have used this standard to determine whether an entity is a subsidiary of another public entity (that is, a controlled entity).

The effects of being assessed as a controlled entity under FRS-37 are that the controlled entity must be consolidated into the parent entity’s group financial statements, and the Auditor-General is the auditor of the controlled entity.

For financial reporting periods beginning on or after 1 January 2007, a New Zealand equivalent to an international financial reporting standard (NZ IAS 27: Consolidated and Separate Financial Statements) will apply for the purpose of the control test. That standard also uses the concepts of control, power, and benefit that apply under FRS-37, and refers to FRS-37 as a source of additional guidance when applying NZ IAS 27. FRS-37 is, therefore, still relevant to determining control for New Zealand public entities. We do not anticipate major changes to the Auditor-General’s portfolio arising from the adoption of New Zealand equivalents to international financial reporting standards.

We discuss in paragraphs 5.112-5.127 how we have applied FRS-37 in determining whether a public entity controls another entity since the enactment of the Public Audit Act.

The approach under FRS-37 is to consider the substance of the relationship between two entities to determine whether one controls another. Control is defined in FRS-37 as:

“Control” by one entity over another entity exists in circumstances where the following parts (a) and (b) are both satisfied:

(a) the first entity has the capacity to determine the financing and operating policies that guide the activities of the second entity, except in the following circumstances where such capacity is not required:

(i) where such policies have been irreversibly predetermined by the first entity or its agent; or

(ii) where the determination of such policies is unable to materially impact the level of potential ownership benefits that arise from the activities of the second entity.

(b) the first entity has an entitlement to a significant level of current or future ownership benefits, including the reduction of ownership losses, which arise from the activities of the second entity.

Part (a) of the definition is referred to in FRS-37 as the “power” element, and part (b) is the “benefit” element. These elements are linked, as ownership benefits are derived from the policies that guide the activities of a subsidiary. Both elements must be present for control to exist, unless one of the exceptions to the power element in subparagraphs (i) or (ii) applies.

Power element

Under FRS-37, an entity is presumed to control another entity if it appoints a majority of members of the second entity’s governing body or controls a majority of voting rights at a meeting.2 FRS-37 overlaps with the legal limbs of the control test in this respect. However, FRS-37 goes further than the legal tests by setting out other indicators of power that are not solely related to appointment of the governing body or voting rights. Examples of other indicators of power include where an entity has a direct or indirect ability to:

  • determine the revenue raising, expenditure, and resource allocation policies of another entity, including an ability to modify or approve the entity’s budget; and
  • veto, overrule, or modify decisions of the governing body other than for the purpose of protecting existing legal or contractual rights or restrictions.

The exceptions to the power element (subparagraphs (i) and (ii) in the FRS-37 definition of control) are also a significant extension of the legal tests of control. These are discussed in paragraphs 5.119 and 5.120.

Benefit element

The benefit element requires the parent entity to be entitled to a significant level of ownership benefits from the subsidiary’s activities, or to have a greater entitlement to benefits than any other parent entity. Ownership benefits are benefits that give a return on an investment.

Types of ownership benefits include:

  • benefits from the distribution of earnings or net assets (for example, a right to a significant level of the net assets of an entity in liquidation); or
  • other benefits from control over net assets (for example, synergistic benefits from a parent and subsidiary combining their activities); or
  • benefits from an entity undertaking activities that are complementary to those of the parent.

In our experience, the activities of trusts formed by public entities often complement those of the public entity. FRS-37 states:

A parent’s entitlement to other ownership benefits may also arise in circumstances where there is a supply of goods or services to a third party by the possible subsidiary, which meets an operating objective of the parent. For example, it is common for special entities such as trusts to be established to provide certain services to support the operating objectives of another entity. In such circumstances, a parent may benefit from complementary activities. Because it can be difficult to identify clearly whether a given circumstance establishes an entitlement to receive the benefits resulting from complementary activities, this Standard takes the position that such entitlement arises when all three of the following conditions apply:

  • the supply of goods or services by the possible subsidiary is directly consistent with, and is likely to enhance, the operating objectives of the parent, and
  • determination of the nature of the goods or services to be supplied is a direct consequence of the exercise of the parent’s decision-making ability over the activities of the possible subsidiary, and
  • the parent is relieved, as a result of the activity of the possible subsidiary, of an actual or constructive obligation to provide such supply; or the parent has a right to receive a future service delivery from the possible subsidiary that is not subject to additional funding to be provided by the parent.

Exceptions to the power element

FRS-37 identifies two circumstances where it is not necessary to have the power element to satisfy the definition of control (see subparagraphs (i) and (ii) of the FRS-37 definition of control in paragraph 5.112).

We have found that the first circumstance often applies to trusts formed by public entities. This is where the policies that guide the activities of an entity have been predetermined and are unable to be modified. In such cases, a power element is not necessary, although the benefit element is still required. Any party that has established such an entity, and has ownership benefits, has control. These arrangements are sometimes described as “irreversible predetermined mechanisms” or “autopilots”. This is discussed further in paragraphs 5.122-5.127.

Trusts controlled by public entities

Since the Act was passed, we have identified a number of charitable trusts in the central and local government sectors as being controlled by one or more public entities in terms of FRS-37. In the central government sector, the majority of such trusts are in the health and education sectors. The most common circumstances of control include:

  • a public entity where, given its right to appoint all or a majority of the trustees, control under FRS-37 is presumed to exist in the absence of evidence to rebut that presumption. The presumption is generally not rebuttable where the public entity receives significant ownership benefits from the charitable trust.
  • a charitable trust established by a public entity where the public entity does not appoint a majority of trustees but:
    • where the objects or purposes have been determined by the public entity and cannot be changed; and
    • where complementary activities provide benefits to the public entity (such arrangements are referred to under FRS-37 as autopilots, discussed in paragraphs 5.122-5.127).


In the case of a trust established for charitable purposes, it is reasonably common to find either that the objects or purposes specified in the trust deed cannot be changed or that substantive changes to the terms of the trust cannot be made. In some cases, substantive changes could be made only if it is no longer possible or practicable to achieve the objects and if approved by the High Court.3 Trustees of charitable trusts often have a power to make amendments to procedural or technical aspects of trust deeds in order to better give effect to the purposes of the trust, provided that any such changes do not affect the status of the trust for income tax purposes.

Such trust deeds can be an “irreversible predetermined mechanism” or “autopilot”, in terms of the first exception to the power element in FRS-37. Where that is the case, the power element under the standard does not have to be present and the parent entity does not need to have an ongoing power to appoint trustees or some other form of power.

We have found that many trusts controlled by public entities are in this category – that is, the policies that guide the activities of the trust have been irreversibly predetermined by the public entity at the time the trust was established. Where the public entity is entitled to receive benefits from the trust’s activities, and where the trustees cannot make substantive changes to the objects of the trust that would have an effect on the public entity’s entitlement to receive those benefits, the significant policy direction of the trust is unlikely to change and the public entity therefore controls the trust under FRS-37.

In many cases, public entities have established trusts at arm’s length from the public entity so that the trust would be able to perform its functions independently. Examples that we have considered include fundraising foundations established by schools and universities, and trusts established to operate facilities such as libraries or hostels. In some cases, entities have established the trusts in order to avoid the restrictions that apply to the parent entity.4

Many public entities and trusts have found it surprising to be told that many such trusts are controlled for accounting purposes under FRS-37 and are therefore public entities. In part, this is because the standard did not apply when the trusts were established. The concept of control is not seen as appropriate for a trust, as the trustees are under a legal duty to act independently in accordance with the objects of the trust and do not consider themselves to be controlled in any sense by the organisation that established the trust.

In general trust law, once a settlor has given property to trustees, the settlor has divested themselves of the asset and the trustees must act independently. The trustees do not receive ongoing funding from the settlor and must act independently. The accounting standard does not sit easily with trust law in this respect, but it does acknowledge that entities often form trusts to provide services that support their objectives. The standard-setters were clearly aware of the accounting standard’s possible application to trusts.

Disputes with controlled entities

We have had protracted debates with trustees of a small number of trusts about whether the trusts are in fact controlled by public entities under FRS-37. The matters that are usually contested are discussed in paragraphs 5.129-5.132.

One issue is whether the objects and purposes of a trust are “the financing and operating policies that guide the activities of the entity” within the meaning of FRS-37.5 In the case of a charitable trust, we consider that the policies that guide the activities of the trust are the objects or purposes of the trust rather than day-to-day administrative matters, such as the particular powers applying to the operational, borrowing, or investment activities of the trust (which, in any case, must be exercised in furtherance of the trust’s objects or purposes).

Another issue is whether the policies that guide the activities of the subsidiary can be modified – that is, whether the trustees can make substantive changes to the objects or purposes of the trust. In our view, it is not possible for trustees to make substantive changes to the terms of the trust in a way that affects the parent entity’s entitlement to ownership benefits. For example, the trustees of a charitable trust established to raise funds for the benefit of a particular entity would be likely to be in breach of their duty if they were to change the objects and purposes of the trust to benefit another entity.

In some instances, trustees have contested whether the public entity established the trust. This is partly a question of fact, and often the trusts and public entities have not been willing or able to make records or evidence of the facts of establishment available to us. In some cases, we have been told that the person such as the chairperson or chief executive of a public entity settled a trust associated with the public entity in their private capacity rather than on behalf of the public entity.

Whether the public entity derives ownership benefits from the activities of the trust is another issue that trustees have contested. In most cases, we consider that the activities of the subsidiary trust are complementary to those of the parent where the three requirements in FRS-37 for complementary benefits apply and the benefit test is met.

In many cases, the real concern of trustees is with the idea that they are controlled by another entity when in legal terms and in practice they are independent. They are also concerned about the possible effect of being consolidated into the group financial statements of a public entity. Some trustees have told us that they believe that consolidation would affect their ability to raise funds from members of the public and other funding organisations, as they would be perceived to be part of a publicly funded entity.

We do not know whether this concern has eventuated for those trusts that have already been consolidated. This would be an unintended consequence of the application of the control test if the trustees’ concern were realised. In our view, being subject to public audit and the greater accountability associated with that may enhance a trust’s appeal to the public and funding organisations.

Trustees tend to be less concerned about the Auditor-General appointing their auditor than about the potential effect of consolidation. The concern about consolidation has proved to be an obstacle to our appointing an auditor in a small number of cases, and some trusts have not been willing to accept that they are subject to the Auditor-General’s mandate.

Where the activities of a subsidiary entity are material to the activities of a parent entity, generally accepted accounting practice requires the parent to consolidate the subsidiary entity into its group financial statements. Where the parent entity is not willing to do so, or is unable to do so because the subsidiary will not provide the necessary information, then the audit opinion on the group financial statements of the parent entity may need to be qualified.

In some cases, the trustees have considered winding up the trust to avoid consolidation and public audit, or resettling the trust fund on a new trust that would not be subject to public audit. We think this is an extreme response to the application of the control test in the Act, and it is one that has involved cost for the trusts concerned. In some cases, trustees have found that they do not have the ability to resettle the trust in the way they seek if the trust deed does not contain an express power to resettle.

We have resolved most disagreements with controlled entities and they have eventually accepted our view that they are controlled under FRS-37. We have explained that the test for control under FRS-37 is relevant for accounting purposes only, and has no effect on the role or independence of the trustees. In many cases, we have been able to appoint the trust’s existing auditor to conduct the audit on our behalf.

We can appreciate why our conclusions are sometimes contentious for trustees who regard themselves as completely independent from the settlor entity, and who are concerned about the implications of control.

The Auditor-General is bound by the Act and the scheme of the Act, which is to ensure that there is public accountability for all public entities, including controlled public entities. We have explained that, in determining control under the Act, we are applying the accounting standard as we understand it. We have suggested to entities that they should raise their concerns about the application of the standard to trusts with the standard-setters.6

If we reach the point where a controlled public entity refuses to accept that the Auditor-General is its auditor, we consider that it would be important to advise Parliament of that fact, including the name of the controlled entity concerned. We do not yet need to take this step, but will do so as necessary in the future.

1: The standard was issued in October 2001 and applies to general purpose financial reports covering periods ending on or after 31 December 2002.

2: Paragraph 5.10 of FRS-37 sets out other circumstances that establish “rebuttable presumptions” that control exists.

3: The Charitable Trusts Act 1957 contains a regime for the variation of charitable trusts where the trust fails in some way.

4: The Crown Entities Act 2004 has introduced requirements for a Crown entity (not including a tertiary education institution or a school) that wishes to settle, or be or appoint a trustee of, a trust – see section 100.

5: The definition of control in FRS-37 refers to “the financing and operating policies that guide the activities of the second entity”.

6: The Financial Reporting Standards Board of the Institute of Chartered Accountants of New Zealand.

page top