Holmcroft v KPMG – Can a Firm of Accountants be a Public Body?

The ‘Big Four’ accounting firms are commercial organisations par excellence. And they are highly successful. They could be the poster children for globalised capitalism in the Twenty-first Century.

In that capacity, from time to time, their collective strength in certain product markets engages the attention of the competition authorities – as it did, for instance, in the UK Competition Commission’s inquiry into statutory audit services.

But competition law is about preventing the abuse of commercial power, and public law is about preventing the abuse of governmental power. These legal disciplines come from the opposite ends of the public-private spectrum. Are there any circumstances in which an organisation as intrinsically commercial as a major accounting firm can also be regarded as a public body and subject to the requirements of public law?

This was the question addressed by the Divisional Court in R (Holmcroft Properties) v KPMG. The case is revealing as to the courts’ approach to applying public law in a complex public-private environment, and in particular their failure to form a coherent view of how regulation operates.

The Background

The case arose from one of the many mis-selling scandals which have dogged the major UK banks over the last few years. It related to ‘interest rate hedging products’ (IRHPs). These are complex financial products which offer businesses protection against adverse movements in interest rates, while also exposing them to liabilities if rates move in an otherwise favourable direction.

In May 2012, the Financial Services Authority – the statutory predecessor to the Financial Conduct Authority (FCA) – decided that there had been unacceptable practices by a number of banks in relation to IRHPs, especially in their sale to relatively unsophisticated buyers.

Having done so, it had various options open to it, including taking enforcement action and commencing litigation. The decision it made, however, was not to impose any formal sanction, but to seek undertakings from the banks that they would establish ‘voluntary’ review and redress schemes.

One of these banks was Barclays. The scheme it agreed with the FCA required it to assess claims for compensation from its customers and pay basic redress comprising the cost of the IRHP to the customer together with interest. Compensation for consequential loss would also be available if a customer could prove that it had suffered additional detriment through buying the product.

In order to ensure that their was proper oversight of this process, the FCA required each bank to agree that it would appoint an independent reviewer. In the case of Barclays, the chosen reviewer, approved by the FCA, was KPMG. Every payment of compensation would have to be approved by KPMG, and its brief was to agree only to those that it considered ‘appropriate, fair and reasonable’. KPMG would be required to report regularly to the FCA, but would otherwise act independently. One of the purposes of these arrangements was to distance the FCA from individual cases.

Under the scheme, over 14,000 customers have been compensated by Barclays. KPMG’s role as reviewer has been substantial. Over 400 of its staff have worked on the project; at one stage 150 of them were working on it at the same time.

Holmcroft Properties was the owner of a care home and had been sold an IRHP by Barclays, its regular bankers. It was one of the unsophisticated customers about whom the FCA was concerned. It bought the product shortly before interest rates collapsed to their current low level, and the cost to it of the payments that it had to make as a consequence was in excess of £300,000. This was a significant sum in the context of its business.

That business came under severe financial stress, and eventually Barclays, as mortgagor of its properties, called in the receivers. In its later submissions as part of the redress scheme, Holmcroft argued that the £300,000 cost of the IRHP had tipped it over the edge, and that its business, in spite of the financial pressure it was under, would otherwise have been viable. It claimed £5.2m in consequential loss.

Barclays rejected this claim, and KPMG approved the rejection. It was against this approval that Holmcroft sought judicial review.

The Judgment

The first question, and the only one of general importance in the case, was whether KPMG in these circumstances was subject to public law and capable of being judicially reviewed – a question (in the accepted jargon) of its ‘amenability’ to public law challenge.

The test now applied by the courts is that summed up by Dyson LJ in R (Beer) v Hampshire Farmers Market Ltd

…the question whether the decision of a body is amenable to judicial review requires a careful consideration of the nature of the power and function that has been exercised to see whether the decision has a sufficient public element, flavour or character to bring it within the purview of public law.’ [16] (emphasis added)

In that case, giving a clear idea of how expansionist an approach the courts can take when they are so minded, a private limited company running a farmers’ market in Hampshire was considered to be a public body and amenable to judicial review.

The problem with the test is that it is circular – a body is a public body if its functions are sufficiently public to bring it within the ambit of public law. This is hardly helpful. The mist clears a little if we understand that what is meant by the rather vague word ‘public’ is ‘governmental’ (Wallbank v PCC of Aston Cantlow at [10], addressing the related question of who is a public body under the Human Rights Act 1998). But even so this is an opaque formulation which offers minimal guidance in individual cases, and the courts at all levels have struggled to apply it.

In Holmcroft, the Divisional Court consisted of two experienced public law judges – Elias LJ and Mitting J. On the one hand, they looked at what role KPMG was performing under the Barclays redress scheme. They accepted that this was essentially ‘regulatory’ in nature – This was more than a mere private arrangement and the bank would never have conferred the veto power upon KPMG unless required to do so by the FCA as part of its regulatory functions’ [39].

In particular there was ‘a clear public connection between [KPMG’s] functions and the regulatory duties carried out by the FCA’ [40]. KPMG was ‘woven into’ the regulatory function [39]. These were all ‘powerful pointers in favour of amenability’ [41].

On the other hand, and weighed against this, the judges considered the circumstances in which KPMG came to be performing these functions. Here they listed five factors.

First, the FCA could have done something different, but chose to adopt an essentially voluntary redress scheme [42]. Second, KPMG’s powers were conferred by contract [43]. Third, the arrangements between Barclays and KPMG were private, even though they secured a public objective [44]. Fourth, the FCA was under no duty to carry out the role that KPMG did, and indeed would have lacked the resources to do so [45]. Fifth, the FCA had reserved its right to step in and take a more active role in particular cases [46].

Looked at in the round, the court considered that these five factors outweighed the other features in favour of amenability. But this was a finely balanced case, and the judgment was reached ‘not without some hesitation’ [41].

In the light of this, it was academic whether KPMG would have been in breach of public law in its substantive decision, but the court concluded – to immunise itself against appeal – that the decision would have withstood challenge.

Amenability – Public Law’s Achilles Heel

At one time, a clear dividing line separated public and private activities. This depended simply on whether or not the person performing the action was a State body – in particular where its authority derived from a public law source (statute or the prerogative powers), though even this was not essential.

However, the law took an important turn in R v Panel on Take-overs and Mergers, ex p Datafin [1987] 1 All ER 564), when the Court of Appeal recognised a self-regulatory body (the Take-over Panel) as being amenable to judicial review. Following that judgment, it is the nature of a function, rather than the type of person performing it or its precise legal source, that is ultimately decisive of whether that function is ‘public’.

The effect of Datafin was to recognise that in an increasingly complex and highly-regulated society, governmental power occurs in many different forms, not all of them deriving from traditional sources or presenting in a traditional guise. It has been used to bring a wide range of bodies within the scope of judicial review.

The problem is that, having been untethered from its previous moorings in the nature of a body performing a function and the source of powers, the test of whether any function is public or private has struggled for clarity.

As any detailed review of the case law shows, it is impossible to find a consistent pattern of decisions or to predict with accuracy on which side of the line a given situation will be deemed to fall.

The Problem with Holmcroft

This is a significant problem. To the Divisional Court in Holmcroft, the principles for determining amenability are ‘tolerably clear, albeit stated at a high level of abstraction’ [23]. However, the second part of this statement contradicts the first. In the real world – where people need to know when law applies to them so as to guide their actions – principles existing only in highly abstract form are never ‘tolerably clear’ as a guide to either conduct or outcomes.

Of course the courts are routinely required to make finely-balanced judgments in factually complex cases, often deriving specific conclusions from general principles. This is part of the normal judicial process.

Nonetheless, the question of amenability is not an ordinary legal question about the interpretation or application of an individual law. It is the question of whether the entirety of a large, complex and sophisticated body of law (contemporary public law) applies to a given entity exercising a specific function. If the purpose of law is to operate normatively – i.e. to establish a set of rules to govern conduct – then it is of importance to be clear, if on nothing else, at least as to when someone is entering the territory in which those rules will apply.

In fact, the test has been described by the Court of Appeal – honestly if not attractively – as being ‘as much a matter of feel as deciding whether particular criteria are met’ (R (Tucker) v Director General of the National Crime Squad at [13]). It is the application of ‘feel’ that leads to the inconsistency of the case law. Holmcroft adds to the undistinguished body of judgments in this area.

Two things in particular strike a jarring note in the judgment.

First, the five factors that are said to weigh against amenability in practice all amount to different ways of saying the same thing – namely that the Barclays redress scheme, and KPMG’s role in it, arose from undertakings given to the FCA rather than being mandated by any formal action on its part. That is true, and may well be a relevant consideration. But it difficult to see how it can gather any strength merely by virtue of being viewed from five different angles.

Second, even assuming that there were five factors which were sufficiently independent of each other to be meaningfully listed as such, the process carried out by the court struggles to find any rational basis. Weighing a set of factors relating to the source of a body’s power against another set of factors describing the nature of that power is to balance two things which are not commensurable. How exactly does this balancing exercise work so as to allow one to be said to tip the scales against the other?

Post-Datafin, it is the nature of powers rather than their source that should determine whether they are ‘public’. It is unclear how the exercise conducted by the Divisional Court in Holmcroft can ever be meaningful.

Regulation, Self-Regulation and Co-Regulation

Holmcroft exemplifies another trend – the failure of the courts to understand or get to grips with the role of regulation.

In YL v Birmingham City Council, the House of Lords was divided as to whether being subject to regulation made an entity more (Lord Bingham) or less (Lord Mance) likely to be a public body, or whether it was an essentially neutral fact which pointed neither one way nor the other (Lord Neuberger).

However, one thing that ought to be clear is that carrying out a regulatory activity in any area of public interest is a public function. On this basis, even pure self-regulation can be a public function, as with the Take-over Panel in Datafin – a decision subsequently applied to a number of other self-regulatory bodies such as the Advertising Standards Authority (R v ASA, ex p Insurance Service plc (1990) 2 Admin LR 77).

The situation in Holmcroft was not an example of self-regulation but of co-regulation – the ‘efforts by government authorities to promote and oversee self-regulation.’*

In the right context, this is a means by which a regulator can achieve regulatory aims while steering a course between (on the one hand) engaging in a formal intervention and (on the other) ceding the territory entirely to self-regulation. It can sometimes be a proportionate solution to a need for regulatory action. The FCA plainly considered that was the case in relation to the IRHP redress schemes.

Co-regulation involves an inherently greater ‘public’ element than self-regulation. It has force precisely because there is a statutorily-empowered regulator standing behind it and able to intervene if it does not achieve its effects. In this sense there was nothing truly ‘voluntary’ about the redress schemes operated by any of the banks.

The acid test point in Datafin was that the Take-over Panel exercised public functions because (in the manner of Voltaire’s dictum about God) if it had not existed it would have been necessary to invent it. If this was correct, it should a fortiori be true that a  function is public where a regulator which the State has already thought it necessary to invent then steps back and temporarily, for a specific purpose, lets part of that regulatory function be performed by another.

In considering that KPMG was not performing a public function, the Divisional Court seems to have fallen into the trap anticipated by Sir John Donaldson MR in Datafin

‘…I should be very disappointed if the courts could not recognise the realities of executive power and allowed their vision to be clouded by the subtlety and sometimes complexity of the way in which it can be exerted.’ ([1987] 1 All ER 564 at 577)

Holmcroft are currently seeking permission to appeal.


*  (Also known, for those who like their post-modernism, as ‘meta-regulation’) – § 8.5, Oxford Handbook of Regulation – Baldwin, Cave, Lodge (eds) – 1st ed 2010.