‘Equity’s treatment of third parties is riddles with inconsistencies.

The courts waver

between resort to analogies with trustees, or with fiduciaries, or with common law
accessories, and seem to find it impossible to hold a firm line on the appropriate rationale
for finding these people liable for anything’. Discuss.

The need to find an appropriate rationale for holding third party accessories liable is
apparent. As a matter of legal certainty, we must know when third parties will be liable and
what for. While the standards in Royal Brunei Airlines v Tan (on dishonest assistance) and
BBCI v Akindele (on knowing receipt) may leave much to be desired in the way of clarity, the
crucial emergent point is that the third-party liability is fault-based. Thus the strict liability
seen in unjust enrichment has been rejected. However, the court has still failed to appoint
the appropriate nature of the liability in these cases, inhibiting a consistent application.

A rationale for third party liability

As indicated above, in the seminal cases of Tan and Akindele the Privy Council and Court of
Appeal, respectively, clarified the fault elements for the two ‘limbs’ of third-party liability
from Barnes v Addy. This resulted in terminological changes: ‘dishonest’ assistance and
‘unconscionable’ receipt are the labels to be used in light of the decisions. To this extent the
decisions can be considered unsatisfactory. Both dishonesty and unconscionability suffer
from real difficulties of definition, and both standards have required the courts to make
assessments of the defendant’s level of knowledge, prompting reliance on the Baden
Delvaux scale that was itself seen as unsatisfactory (e.g. in Armstrong v Winnington
Networks). Unconscionability has been described variously as nebulous, vague and unstable
(Birks), and was rejected as the standard for assistance liability by Lord Nicholls in Tan for
these reasons. Despite this, the message from Tan and Akindele is clear: strangers to the
trust will be liable only where they are at fault. This is important because in the case of
strangers to the trust, the justifications which are used in relation to the liability of trustees
and fiduciaries, as well as common law accessories, are inappropriate. As Devonshire
argues, the prophylactic rationale for strict liability for trustees and the good man theory of
equity (which explains, for example, why all bribes must be held on constructive trust for
the principal: FHR European Ventures (Millett)) have no application in relation to strangers
to the trust. Further, it is justifiable to have a different standard of liability for accessories in
equity than common law accessories because of the different nature of equitable property.
This argument is put forward by commentators including Low and Stevens.

Knowing receipt

The appropriate rationale for holding someone liable in knowing receipt should be that a
knowing recipient is a constructive trustee. This is an idea advanced by Mitchell and
Watterson, who hold that knowing recipients have a custodial duty to restore assets, and
the order to account simply gets them to perform that duty. The remedy is directly
enforcing the duty, so there is no element of wrongdoing on the part of the defendant. This
is further rationalised by Chambers and Worthington, who both correctly rationalise that if
there is a continuing interest, there has to be a constructive trust due to the split title. As we
know from Westdeutsche Bank, to be a constructive trustee you have to be aware that you
are holding the property on trust. Here, for knowing receipt, there has to be some degree of
knowledge, which is enough to relevantly engage D’s conscience. This knowledge affecting
D’s conscience, along with the title split is enough to form a constructive trust. This idea also
finds support in the caselaw, with Arthur and Grimaldi both directly support this, with
Buyers acknowledging the existence of these custodial duties that come with being a
constructive trustee.

There has been some argument from both Birks and Lord Nicholls that knowing receipt
should operate as a reversal of unjust enrichment. This would include two claims: one a
restitutionary claim with strict liability and a separate claim for dishonest receipt. However,
as Salmons argues, this is an inappropriate rationale for third party liability. It is difficult to
identify an ‘unjust’ factor that would be consistent with trust principles or practically
workable, as the fact T has the legal power to transfer title to trust property is fundamental
to the trust’s existence and operation. Furthermore, he argues that the aim of the equitable
regime is to recover trust property, which would not be supported by an unjust enrichment
analysis. The desire for an unjust enrichment response to recipients is waning, yet there
remains doubt over the nature of a recipient’s liability and how far it should extend. This has
also been rejected in Akindele and Buyers.

Dishonest assistance

In Baden Delvaux v Société Générale, Peter Gibson J identified four requirements for
establishing X’s liability for what was then known as knowing assistance. It had to be shown
that there was 1) a trust, 2) a fraudulent and dishonest design on the part of the trustee
(‘T’), 3) assistance by X and 4) ‘knowledge’ on X’s part of the breach of trust. Requirements 2
and 4 have now been modified by Tan. There is no longer any requirement that T be
fraudulent or dishonest; the fact that T was innocently in breach of trust will not relieve X of
liability for assisting that breach of trust. Further, the standard of liability for X has been
altered from knowledge to dishonesty.

Here, the nature of the liability cannot be described as a constructive trust. As Mitchell and
Watterson point out, X can only be described as a trustee in the fictional sense, as he has
never owed duties in regard to the property himself. This was affirmed by the rejection of
the term in Dubai Aluminium.

In the recent case of Novoship v Nikitin, a stranger to the trust who had assisted its breach
was held to be liable to account for the profits gained from that conduct. This remedy
indicates that the assister was primarily liable, as Worthington notes. Ridge had argued prior
to this case that X’s liability ought to be primary, reflecting the fact that X commits a wrong
which is distinct from T’s. She notes that as the levels of culpability of T and X can vary
greatly, it is helpful to assess D’s separately. However, X’s liability is contingent on a breach
of trust. Although an honest T may be relieved from liability under s61 of the Trustee Act
1925, his liability still arises. If there is no breach of trust, however dishonest X has been,
he cannot be liable. X’s liability is therefore fundamentally accessorial. This is so even
despite the removal of the second requirement from Baden Delvaux in Tan. If this is true
then X’s liability should not exceed T’s, and the decision in Novoship must be wrong to this
extent: if T would not be liable to account for the profits, nor should X be. Yet, Novoship
represents the current law, and it has been welcomed by some commentators. Devonshire,
on the other hand, argues that the prophylactic rationale which justifies disgorgement of
profits in relation to fiduciaries and the good man theory of equity should have no
application to strangers to the trust. This serves to demonstrate the uncertainty which
continues to pervade the law of dishonest assistance.

Conclusion

The court has been unwilling to assign a consistent and appropriate rationale for third party
liability. They have not yet confirmed, as they should, that a knowing recipient is a
constructive trustee, which has prevented them from applying consistent principles to
accessories who fall in this area. Further, the remedies used by the court in Novoship stops
dishonest assistance from being regarded as secondary liability, which would be a more
appropriate approach than treating it as primary liability. This has led to uncertainty and
inconsistency.