Questioning the Strength of Protective Clauses in Wealth Management Contracts

A PDF version of the article can be found here.

---

QUESTIONING THE STRENGTH OF PROTECTIVE CLAUSES IN WEALTH MANAGEMENT CONTRACTS

Desmond Chye*

I.        INTRODUCTION

Wealth management contracts often contain protective clauses attempting to shield the wealth manager from liability arising from their wrongful conduct, such as negligence and misrepresentation. It is often claimed that these clauses continue to present clients with ‘insuperable difficulties’ in litigating claims at common law and otherwise. While this was true under Singapore law, it is no longer so. Today, it is possible for clients to challenge these clauses under common law or to outflank them by using statutory remedies. However, even though litigating claims is no longer insuperable, it remains admittedly difficult to do so. Ultimately, an unsophisticated retail client would stand a better chance of successful litigation and obtain an adequate remedy than a corporate or sophisticated one. In the final analysis, the difficulty faced by clients in litigating their claims is inappropriate. The law is too biased towards the wealth manager and needs reform.

II.     TYPES OF PROTECTIVE CLAUSES

Protective clauses in wealth management contracts fall broadly into two types: entire agreement clauses (‘EAC’) and non-reliance clauses (‘NRC’). EACs restrict the parties’ contractual relationship to the confines of the written document so as to prevent term implication and incorporation of collateral oral agreements unfavourable to the wealth manager. It is also to prevent tortious duties from arising as contractual disclaimers deny the requisite proximity with the client. NRCs, on the other hand, protect the wealth manager from misrepresentation liability by basically rewriting history using contract to say that the client did not rely on pre-contractual representations. It is therefore necessary for the client to invalidate these two clauses otherwise his claim for breach of duty is doomed.

III.  CHALLENGING PROTECTIVE CLAUSES AT COMMON LAW

Unfortunately for the client, the common law would generally uphold EACs and NRCs.^[1] Starting with EACs, courts uphold them due to the parol evidence rule.^[2] This rule basically stipulates that where a contract is recorded in a written document, no extrinsic evidence may be adduced to vary or contradict the terms recorded. An EAC in the written contract therefore cannot be refuted to find collateral oral terms.

The common law’s logic is however flawed. The parol evidence rule first requires parties to use a ‘written contract’ but not all written documents, such as receipts, qualify.^[3] Parties are also free to construct their contract part written, part oral.^[4] These factors must mean that a ‘written contract’ arises only when the parties intended a contractual instrument to contain their entire contract, if not then the parol evidence rule is inapplicable.^[5] An EAC should therefore be ineffective if the parties did not actually intend the written document to comprise the entire contract.^[6] Consequently, the parol evidence rule should only create a rebuttable presumption of fact that a written document is the whole contract and so a client should be allowed to adduce extrinsic evidence to prove that there were collateral oral terms.^[7] This is especially desirable for EACs as these are normally boilerplate clauses that the parties do not pay much attention to.^[8]

Notwithstanding the parol evidence rule, there are also some limits on EACs to potentially assist clients. Terms normally implied by usage or course of dealing can be implied into the contract^[9] and terms can still be added through rectification.^[10] Unfortunately, these exceptions are not particularly useful to clients. It is not considered customary for wealth managers to owe fiduciary or advisory duties to clients and so courts are unwilling to imply such duties.^[11] Wealth management contracts are also normally comprehensively drafted with no gaps left for term implication.^[12] Rectification is also difficult as the client must show that the parties made a mistake in expressing their true agreement which infected the EAC.^[13] This is further restrictively construed to cover only situations where an errant term is corrected and not situations where extra terms are inserted.^[14] Rectification is thus only available where the parties wrote an inaccurate clause and not where they intended to be bound by that clause and another one.^[15] Although the standard wealth management contract expressly disclaims many duties that clients would like the wealth manager to owe, wealth managers generally do not intend to owe any duties whatsoever at all times and thus rectification is not possible.

Alternatively, the client may use contra proferentum to restrict the protective clauses’ scope^[16] but this is not possible if the wealth manager stipulated their exclusions thoroughly, which is often so.

For NRCs, there are two possible common law methods to enforce the clause: estoppel by representation and contractual estoppel.^[17] Although the former is useless, the latter presents an insuperable difficulty to clients claiming misrepresentation.

Estoppel by representation first requires an unequivocal representation by party B that he did not rely on any statements, which B intended party A to act on. A must then have believed it to be true and acted in reliance of it.^[18] It is however virtually impossible for A to prove he believed B’s declaration of non-reliance.^[19] Estoppel by representation is thus impotent in practice.

Contractual estoppel can however effectively shield the wealth manager from misrepresentation liability. Under English law, NRCs would be enforced simply to uphold the parties’ agreement, notwithstanding that the client actually relied on the misrepresentations.^[20] Singapore follows this harsh expression of documentary fundamentalism.^[21] The Singapore High Court (‘SGHC’) however recently opined that it can be relaxed if an unsophisticated client was persuaded to sign the clause in ignorance of its nature.^[22] Unfortunately, the Singapore Court of Appeal (‘SGCA’) declined to decide on whether the suggestion was acceptable.^[23]

However, assuming that the SGHC’s position is adopted, since it was not overturned,^[24] there still remains the potential conflict with an earlier case^[25] where contractual estoppel was applied without qualification. It is suggested that the two cases can be distinguished based on the client’s sophistication: the client in the earlier case^[26] was found to be sophisticated while the client in the latter case was arguably not.^[27] If so, then an unsophisticated client would now have a higher chance of invalidating NRCs.

It is submitted that contractual estoppel is unsound and should be abandoned completely. Firstly, the cases^[28] establishing it were arguably decided per incuriam. The English Court of Appeal already rejected the doctrine in an earlier case.^[29] As English courts are bound by the past decisions of same-level courts, the subsequent cases that established contractual estoppel^[30] violated the doctrine of precedent and were hence wrongly decided.^[31] Although Singapore is not bound by English cases, their illegitimacy under English law should still diminish its persuasiveness.

Secondly, contractual estoppel is fundamentally incongruous with established estoppel principles.^[32] Traditionally, “estoppels at root require detriment”^[33] because they are treated as independent sources of rights outside of the contract and so extra-contractual factors such as reliance and unconscionability are required to justify its invocation.^[34] Contractual estoppel, however, lacks these requirements of detriment and is moreover practically indistinguishable from just enforcing the contract’s terms.^[35] The doctrine, being a black sheep, is thus heretical^[36] and superfluous.^[37]

Thirdly, just because a clause appears in a written contract does not necessarily mean that the parties actually agreed to it.^[38] Without true agreement, it is merely a non-promissory statement of past fact.^[39] This is especially pertinent to the wealth management context where the parties’ close relationship makes it likely that protective clauses were not truly agreed to. Unlike the banking sector which is becoming increasingly de-personalised and hence treated more cautiously by clients, the wealth management sector is becoming increasingly personalised. The close ‘familial’ relationships created with clients give rise to a relationship of trust and confidence in the wealth manager which encourages clients to have blind trust in the wealth manager’s intentions. Protective clauses contradicting earlier assurances are thus likely to be treated lightly or ignored completely.^[40] While this is less relevant to institutional clients who might have legal advisers to fall back on to caution them,^[41] it would be for retail clients who do not have such safety nets. Nevertheless, contractual estoppel should be eschewed for all client types since its monolithic nature completely fails to consider the nuances of the client’s especial relationship with the wealth manager.

For both EACs and NRCs, it is claimed that contractual estoppel applies to them on the basis of the signature rule. This rule stipulates that, aside from situations of fraud, misrepresentation or non est factum, a person who signs a contractual document is bound by it, even if it was not read or understood.^[42] A client failing to understand or read the protective clauses (not uncommon)^[43] will therefore not render the clauses invalid.

The signature rule reason is however unconvincing as it contravenes the fundamental contract principle that there must first be an objective meeting of the minds to incorporate a term.^[44] The rule essentially allows a mere signature to substitute for an objective mutual agreement to contract which cannot be right doctrinally.^[45] A term should only be conclusive if the parties truly intended it so.^[46]

Furthermore, the signature rule’s locus classicus^[47] is weak authority for the principle. It did not consider whether an EAC could exclude collateral oral agreements as it was not alleged that there were unrecorded terms, only that the document could not exclude implied terms.^[48] It was therefore not held that a signed EAC conclusively proves the parties’ intention for it to embody the whole contract.^[49] A signed document is thus no more than strong evidence of contract conclusiveness^[50] and so clients should be allowed to prove collateral oral agreements and to challenge the validity of signed terms.

Although abandoning contractual estoppel undermines commercial certainty and would hence affect Singapore’s status as a leading financial hub, the price paid for good industry practices is worthwhile. As protective clauses shield wealth managers from the consequences of their wrongdoing, a dangerous moral hazard is created: wealth managers would not be deterred from acting irresponsibly but their incentive to seek as many clients as possible to earn commissions remains. While such harm may be outweighed by the benefit of promoting economic growth through providing contractual certainty, it is only so where the protective clauses were genuinely agreed to. Unfortunately, protective clauses are frequently foisted on clients through standard forms that clients have little say over. Applying contractual estoppel would therefore licence wealth managers to act irresponsibly in the financial market which can potentially damage the economy significantly^[51] and ultimately undermine Singapore’s position as a leading financial hub. It is thus better to use a rebuttable presumption that protective clauses were not truly agreed to by clients, which applies to unsophisticated but not sophisticated clients.^[52]

IV.  CHALLENGING PROTECTIVE CLAUSES USING STATUTE

Notwithstanding the client’s litigation difficulties at common law, there are still statutory methods to obtain remedies. These statutory avenues are however all unsatisfactory in one way or another.

A.     UCTA & MA

The Unfair Contract Terms Act 1977^[53] (‘UCTA’) and Misrepresentation Act 1967^[54] (‘MA’) may assist the client to invalidate EACs and NRCs, but only in limited circumstances.

UCTA section 2(2) prevents wealth managers from contractually excluding or restricting liability for their negligence if it is unreasonable within the meaning in section 11. EACs have however been treated as falling outside UCTA’s purview under English law. This is as protective clauses have been interpreted as not exempting liability but instead as merely defining the parties’ relationship (ie. acting as a ‘basis clause’) and so UCTA is inapplicable.^[55] Basically, the court ignores the substantive effect of a clause exempting liability in favour of what the clause claims to be.

MA section 3 invalidates any term excluding or restricting liability for misrepresentation if it is unreasonable within the meaning in UCTA section 11. Unfortunately, NRCs are rarely treated as exemption clauses. Under English law, the key requirement for invoking section 3 is whether the NRC sought to ‘rewrite history’ (ie. to retrospectively alter the ‘character and effect’ of pre-NRC representations).^[56] This criterion was however butchered by subsequent cases^[57] where there was an overemphasis on whether the contract states there was no reliance, as opposed to looking at what the client understood at the time of the misrepresentation.^[58] Consequently, whether a NRC survives section 3 depends on whether contractual estoppel applies – which it would in most cases. Singapore unfortunately followed this English law for both types of protective clauses.^[59]

English law is however becoming more pro-client. A recent English Court of Appeal case^[60] favoured a substance over form approach, albeit in a non-financial context. The court held that if the party subject to a NRC relied upon what they reasonably considered to be a representation before the NRC was incorporated, then the NRC is an exclusion clause under MA section 3.^[61] The ‘basis clause’ and contractual estoppel arguments thus no longer apply to pre-NRC representations unless the client was actually aware that they should not be relied upon.^[62] Presumably this approach also extends to EACs. While no Singapore case has decided on the new English position, the SGCA’s past statements suggest future approval: they opined that courts should focus on the substantive effect and not form of a clause to determine if it is an exemption clause under UCTA.^[63] If so, clients can now use the MA and UCTA to challenge protective clauses.

Ultimately the ‘basis clause’ reasoning misconstrues the MA and UCTA’s true scope and should be rejected. Although the statutes’ plain words confine their effect to only ‘exemption’ and ‘restriction’ of liability clauses, a purposive interpretation^[64] reveals that parliament actually intended them to apply broadly to any clause which has the effect of preventing a duty of care from arising.^[65] This means duty defining clauses would be captured as well.^[66] While this would severely restrict the wealth manager’s freedom to contractually define the scope of its obligations,^[67] it is justifiable on the basis that the statutes only bite where the duty exclusion is unreasonable.

Notwithstanding UCTA and the MA applying to protective clauses, proving the requisite unreasonableness under UCTA section 11 is still difficult for clients. This is as courts generally assume that the parties desire commercial certainty and already reflected the risk allocation in the price paid.^[68] Unreasonableness has thus been confined to exceptional situations where the clause was not clearly drafted^[69] or poorly explained to the client,^[70] even where the client is unsophisticated.^[71]

While the assumptions are true for sophisticated clients, especially where the client is more knowledgeable than the wealth manager,^[72] it is not for unsophisticated ones. Unsophisticated clients often lack equal bargaining power and so struggle to avoid such draconian clauses from being imposed on them. They also frequently lack access to professional advice to fully understand the protective clauses’ nature. The court’s fear of unjustly allowing clients to escape a bad bargain here is thus misplaced.

However, not all unsophisticated clients deserve the same protection. ‘Sophistication’ is a nebulous concept under common law which can encompass transaction familiarity,^[73] financial expertise^[74] or whether the client was institutional.^[75] Amongst such clients, institutional investors^[76] who lack transaction familiarity or financial expertise should generally remain bound to protective clauses since they have equal bargaining power and can access professional advice.

B.      FAA

The Financial Advisers Act 2001^[77] (‘FAA’) may assist clients in obtaining remedies. Section 34(1) obliges financial advisors^[78] to disclose all material information relating to the financial product recommended to a client. Section 35(1) prohibits the financial adviser from making misrepresentations whether fraudulently, recklessly or negligently. This applies broadly as all statements “in connection with the provision of any financial advisory service” are included. Section 36(1) requires financial advisers who make recommendations relating to any investment product, where a client would reasonably expect to rely on the adviser’s recommendation, to have a reasonable basis for the recommendation. This arguably includes considering the client’s suitability for the product per MAS Notice FAA-N16 and Financial Advisers Regulations^[79] Reg 18B.^[80] As the FAA now allows one to obtain statutory damages from a financial adviser’s breach of the aforementioned sections, clients can obtain remedies in those situations (if there was reasonable reliance).^[81]

However, the FAA may be practically ineffectual. Firstly, unlike UCTA or the MA, the FAA neither expressly provides that terms inconsistent with it are void nor gives any guidance for such situations.^[82] It is thus possible that protective clauses would exclude the operation of the FAA.^[83] Secondly, the FAA does not cover all client types. It only covers individuals and not corporate clients. Thirdly, not all individuals enjoy similar protection. ‘Accredited investors’ and ‘High Net Worth Individuals’ (essentially, individuals with high net worth or income) are excluded from sections 34, 36 and Reg 18B’s protections. This is problematic because wealth is not a reliable proxy for financial expertise. Although an opt-in regime was instituted in 2019 to allow such investors to choose their preferred classification,^[84] protection might still be lacking in practice. This is as wealth managers often restrict the sale of lucrative financial products to only investors of a certain classification and so investors are strongly incentivised to choose a higher classification despite the risks involved.^[85]

C.     CPFTA

Since 2009, financial services fall under the Consumer Protection (Fair Trading) Act 2003^[86] (‘CPFTA’) and so clients can now use it to challenge protective terms. As the CPFTA invalidates any term inconsistent with it,^[87] terms inserted through sharp practice^[88] (such as by pressure selling or misrepresentation) and terms unconscionably restricting the wealth manager’s liability can be invalidated.^[89]

However, the CPFTA is not particularly useful to clients in reality. Firstly, only consumers are protected.^[90] Secondly, it only applies to individuals and not corporate consumers. Thirdly, CPFTA claims are limited to only SGD 30,000 which is far too low for virtually all wealth management contracts. Thus, while getting a claim under CPFTA may be better than nothing, it is a hollow victory for the client.

The CPFTA should therefore have its claim limits increased for financial services. This is especially necessary amidst the increase in consumer investment participation. Recently, online investment services have become ubiquitous amongst everyday Singaporeans, enabled by the aggressive marketing of ‘one-stop’ investment smartphone applications.^[91] The market is however crowded and rival applications intensely compete against each other.^[92] This cutthroat business environment, combined with the business model of such tech products typically requiring a sufficiently large number of trades by users to be profitable,^[93] creates a real temptation for sharp practice (such as deceptive marketing).^[94] The CPFTA, being specialised in tackling sharp practice, and is more protective of consumers than the other statutes,^[95] is thus needed to handle this emerging situation. Unfortunately, its low claim cap renders it impotent. Although these investment applications are ostensibly only for small trades, it is easy to see how repeated usage can spiral into larger trades as the applications are structured like games to promote increasing, and intense user participation. A larger cap is therefore necessary to protect consumers adequately.

V.     OTHER REMEDIES

Despite the protective clauses, the client may still get relief at the damages stage by pleading contributory negligence. Essentially, the client pleads contributory negligence on the part of the wealth manager to reduce the damages claimed against him. This method prima facie appears useful because wealth management litigation often arises from the bank suing the client to obtain compensation for losses incurred during trading.^[96] It is however of limited effectiveness in reality. First, the wealth manager must have been negligent but this is difficult to make out in light of the court’s general reluctance to find duties of care (both in tort and contract) owed by the wealth manager.^[97] Second, the client must not have broken the chain of causation and so the client must not have continued to trade anyway or acted unreasonably post-breach.^[98] More problematically, contributory negligence only operates as a shield and not a sword and so the client cannot make a claim independently of the lawsuit against him or to obtain damages from the wealth manager.

Alternatively, the client may use extra-judicial remedies. The Financial Industry Disputes Resolution Centre Limited (‘FIDReC’) was set up in Singapore to provide a low-cost consumer-friendly dispute settlement mechanism. FIDReC offers to mediate and then adjudicate disputes. The process is simple, fast and cheap. No external lawyers are allowed at hearings. FIDReC is very helpful in situations where the client faces practical difficulties in litigating his claims. This is likely where a low net worth, unsophisticated retail client challenges a big institutional wealth manager as the resource and legal talent imbalance between the two might make it practically insuperable for the client to litigate. The pro-consumer features of FIDReC thus helpfully level the playing field between a financially and legally challenged client and their wealth manager.

VI.  DO WEALTH MANAGERS NEED STRONG PROTECTION?

It is feared that without strong protective clauses, wealth managers might be unjustly liable for extreme losses arising from an intervening unforeseen event, such as a financial crisis, that was not particularly related to their wrongdoing.^[99] This fear is however misplaced as there are already adequate safeguards present to protect wealth managers from unfair liability.

Without NRCs and EACs, a contractual duty of care may be implied into the contract or it may be possible to find sufficient proximity to establish a tortious duty of care in executing the wealth account or advising the client.^[100] A contractual duty to act as a fiduciary might also be implied. Although breaching these duties can potentially impose substantial liability on the wealth manager, it would not be to an unjust extent.

In tortious and contractual negligence cases, a client can potentially claim the entire loss but it is subject to the client proving causation and remoteness. Although causation uses the ‘but for’ test which has a low threshold, it can still provide fair and effective protection. There is no causation where the client would have still made the loss-making trades anyway,^[101] would have used their independent judgement instead, or acted unreasonably post-breach as those would break the chain of causation.^[102] Moreover, causation shields the wealth manager from liability where the loss was due to poor market conditions not of their own making (such as by selling defective products).^[103] Remoteness also provides fair protection. In contract, a loss is remote if it does not arise ‘naturally’ from the breach in question and is not one that could have been reasonably contemplated by the parties.^[104] In tort, a loss is remote if it was of a type not reasonably foreseeable by the tortfeasor,^[105] with courts being more stringent in applying remoteness for such ‘pure economic loss’ cases.^[106] Moreover, any losses claimable are limited to only those arising from the tortfeasor’s scope of duty.^[107] Although the MA might grant excessive remedies for negligent misrepresentation by adopting the ‘fiction of fraud’ approach (where even remote losses are claimable) this is only applicable to English and not Singapore law.^[108] The wealth manager’s liability, when imposed, is therefore proportionate.

Breaches of fiduciary duties do not impose unfair liabilities on wealth managers. This is as the breach must have caused the losses claimed (a safeguard present in Singapore,^[109] unlike the UK).^[110] Although the remoteness and mitigation rules do not apply^[111] and so the liability here may exceed that for negligence breaches, it is justifiable on the basis of giving effect to the strong public interest in deterring fiduciaries from acting in their own self-interests.^[112]

Even if we assume that the liability imposed is excessive, wealth managers can still avoid them easily. To avoid misrepresentation, wealth managers can just avoid making representations they do not wish to be liable for. To avoid liability from collateral terms or implied duties, the wealth manager can just avoid making collateral oral agreements and not act in a way that attracts fiduciary or tortious duties. These are not difficult asks. There is therefore no injustice in making wealth managers liable for unnecessarily causing loss to clients.

VII. CONCLUSION

While it is not insuperable for clients to litigate their claims against wealth managers when there are protective clauses, it is still difficult to do so. Ultimately, the enforceability of protective clauses should not be treated dogmatically such that they are either insuperable or powerless. To strike the right balance, the circumstances of each case, in particular client’s sophistication and the parties’ intentions, should be used to decide whether protective clauses are enforceable.

---

* LLB (Candidate), National University of Singapore, Class of 2023. All errors and views expressed in this article remain my own. An earlier version of this article was submitted for the NUS Law Module LL4191 Wealth Management Law.

[1] See e.g. Deutsche Bank AG v Chang Tse Wen [2013] 4 SLR 886 (SGCA) [Deutsche Bank AG SGCA]; Springwell Navigation Corp v JP Morgan Chase Bank [2010] EWCA Civ 1221 [Springwell Navigation]; AXA Sun Life Services Plc v Campbell Martin Ltd [2011] EWCA Civ 133 [AXA Sun Life].

[2] Shogun Finance Ltd v Hudson [2004] 1 A.C. 919; Common law position codified in Singapore in Evidence Act 1893 (2020 Rev Ed Sing), ss 93–102.

[3] D McLauchlin, “The Entire Agreement Clause: Conclusive or a Question of Weight?” (2012) 128 Law Q. Rev 521 at 527 [McLauchlin].

[4] Ibid.

[5] Ibid.

[6] Ibid.

[7] McLauchlin, supra note 3 at 528; Edwin Peel, Treitel: The Law of Contract, 15^th ed (London, UK: Sweet & Maxwell, 2020) at para 6-022.

[8] McLauchlin, supra note 3 at 523, 531.

[9] Novoship (UK) Ltd v Mikhaylyuk [2015] EWHC 992 (Comm) at para 32.

[10] JJ Huber (Investments) Ltd v Private DIY Co Ltd [1995] NPC 102 (Ch).

[11] Go Dante Yap v Bank Austria Creditanstalt AG [2011] 4 SLR 559 (SGCA); S Booysen, “Financial Advice and the Duty to Advise”, in S Booysen, ed, Financial Advice and Investor Protection (UK: Edward Elgar, 2021) at para 4.04 [Booysen].

[12] See e.g. UBS AG v Ng Kok Qua [2010] SGDC 509, Orient Centre Investments v Société Generale [2007] 3 SLR 566 (SGCA) [Orient Centre Investments] (for examples of no gaps); See e.g. Go Dante Yap v Bank Austria Creditanstalt AG [2011] 4 SLR 559 (SGCA) (for rare example of a gap present due to shoddy contract drafting).

[13] Surgicraft Ltd v Paradigm Biodevices Inc [2010] EWHC 1291 (Ch) at para 75.

[14] Procter & Gamble Co v Svenska Cellulosa Aktiebolaget SCA [2012] EWHC 498 (Ch).

[15] McLauchlin, supra note 3 at 528.

[16] See e.g. Jiang Ou v EFG Bank AG [2011] 4 SLR 246 (SGHC).

[17] L Mason, “Precluding Liability for Pre-contractual Misrepresentation: the Function and Validity of Non-reliance Clauses” (2014) J Bus L 313 at 314.

[18] E A Grimstead & Son Ltd v McGarrigan [1999] EWCA Civ 3029.

[19] Watford Electronics Ltd v Sanderson CFL Ltd [2001] EWCA Civ 317.

[20] Peekay Intermark Ltd v Australia and New Zealand Banking Group Ltd [2006] EWCA Civ 386 [Peekay Intermark]; affirmed by Springwell Navigation, supra note 1 & AXA Sun Life, supra note 1.

[21] Orient Centre Investments, supra note 12; Tradewaves Ltd v Standard Chartered Bank [2017] SGHC 93 [Tradewaves].

[22] Als Memasa v UBS AG [2012] SGCA 43 at para 29; Deutsche Bank AG v Chang Tse Wen [2013] 1 SLR 1310 (SGHC) [Deutsche Bank AG SGHC].

[23] Deutsche Bank AG SGHC, supra note 22.

[24] The SGHC decided to eschew contractual estoppel in Deutsche Bank SGHC, supra note 22 on the basis of the client’s unsophistication. This decision was overturned on appeal by the SGCA in Deutsche Bank AG SGCA, supra note 1 as they found Dr Chang to be sophisticated. The SGHC’s approach of no contractual estoppel when the client is unsophisticated was however not addressed by the SGCA.

[25] Orient Centre Investments, supra note 12.

[26] See Orient Centre Investments, ibid (case involved a corporate client).

[27] Deutsche Bank AG SGHC, supra note 22.

[28] Peekay Intermark Ltd, supra note 20; Springwell Navigation, supra note 1.

[29] Lowe v Lombank Ltd [1960] 1 WLR 196 (CA).

[30] Peekay Intermark, supra note 20 , Springwell Navigation, supra note 1.

[31] G McMeel, “Documentary Fundamentalism in the Senior Courts: The Myth of Contractual Estoppel” (2011) LMCLQ 185 at 191 [McMeel].

[32] Ibid.

[33] Wilken & Ghaly, The Law of Waiver, Variation, and Estoppel, 3rd ed (Oxford: Oxford University Press, 2012) at para 13.22.

[34] McMeel, supra note 31 at 206.

[35] Kelry CF Loi, “Contractual estoppel and non-reliance clauses” [2015] LMCLQ 265 at 366 [Loi].

[36] McMeel, supra note 31 at 206.

[37] Loi, supra note 35 at 357.

[38] McLauchlin, supra note 3 at 536.

[39] Loi, supra note 35 at 351-353.

[40] This occurred in Deutsche Bank AG SGCA, supra note 1.

[41] See Springwell Navigation, supra note 1 (where even corporate clients can occasionally be too trusting and not read the contract carefully).

[42] L’Estrange v F Graucob Ltd [1934] 2 K.B. 394 [L’Estrange].

[43] See e.g. Deutsche Bank AG SGHC, supra note 22.

[44] McLauchlin, supra note 3 at 532; JR Spencer, “Signature, Consent, and the Rule in L’Estrange v Graucob” (1973) 32(1) Cambridge LJ 104 at 117.

[45] Ibid.

[46] Ibid.

[47] L’Estrange, supra note 42.

[48] McLauchlin, supra note 3 at 533.

[49] Ibid.

[50] Ibid.

[51] See e.g. European Systemic Risk Board, “Report on misconduct risk in the banking sector” (June 2015) at 6 (redress costs for bank mis-selling amounted to EUR 100 billion globally from 2010 to 2015).

[52] Adopting a suggestion made in Booysen, supra note 11 at paras 4.45–4.46.

[53] Unfair Contract Terms Act (2020 Rev Ed Sing).

[54] Misrepresentation Act (2020 Rev Ed Sing).

[55] IFE Fund SA v Goldman Sachs International [2006] EWHC 2887 (Comm) [IFE Fund SA]; Crestsign Ltd v National Westminster Bank plc [2014] EWHC 3043 (Ch) [Crestsign].

[56] IFE Fund SA, supra note 55; Raiffeisen Zentral Bank v Royal Bank of Scotland plc [2010] EWHC 1392 (Comm); Springwell Navigation Corp, supra note 1; H Beale and G Palmer, “Non-reliance Clauses, Entire Agreement Clauses and Contractual Estoppel” in Booysen, supra note 11 at paras 5.38–5.42 [Beale & Palmer].

[57] Crestsign, supra note 55 at para 106–108; Thornbridge Ltd v Barclays Bank plc [2015] EWHC 3430 (QB) at para 109; Beale & Palmer, supra note 56 at paras 5.42–5.43.

[58] Beale & Palmer, supra note 56 at paras 5.43–5.45.

[59] Orient Centre Investments, supra note 12; Tradewaves, supra note 21..

[60] First Tower Trustees Ltd v CDS (Superstores International) Ltd [2018] EWCA Civ 1396 [First Tower].

[61] Ibid.

[62] Beale & Palmer, supra note 56 at para 5.48.

[63] See Deutsche Bank AG SGCA, supra note 1 at para 63 (for obiter dicta supporting Smith v Eric S Bush [1990] 1 A.C. 831 & Phillips Products Ltd v Hyland [1987] 1 W.L.R. 659)

[64] UK, Law Commission, Exemption Clauses Second Report (London: Her Majesty’s Stationery Office, 1975) at paras 36, 139; First Tower, supra note 60 at para 51.

[65] L Ho & T Mathias, “Basis Clauses and the Unfair Contract Terms Act 1977” (2014) 130 Law Q. Rev 377 at 380.

[66] Ibid.

[67] Norman Palmer & David Yates, “The Future of the Unfair Contract Terms Act 1977” (1981) 40(1) Cambridge LJ 108 at 127–128.

[68] National Westminster Bank Plc v Utrecht-America Finance Company [2001] 3 All ER 733 (CA).

[69] Camerata Property Inc v Credit Suisse Securities (Europe) Ltd [2011] 2 BCLC 54 (Com Ct).

[70] Crestsign, supra note 55.

[71] Cassa di Risparmio della Repubblica di San Marino SpA v Barclays Bank Ltd [2011] EWHC 484 (Comm); AXA Sun Life, supra note 1.

[72] See e.g. Springwell Navigation, supra note 1.

[73] Wingecarribee Shire Council v Lehman Brothers Australia Ltd [2012] FCA 1028 [Wingecarribee].

[74] Deutsche Bank AG SGCA, supra note 1; Springwell Navigation, supra note 1.

[75] Australian Securities and Investments Commission v Citigroup Global Markets Australia (No 4) [2007] FCA 963.

[76] See e.g. Wingecarrihee, supra note 73.

[77] Financial Advisors Act (2020 Rev Ed Sing) [FAA].

[78] Ibid, s 2(1) (“financial adviser” means a person who carries on a business of providing any financial advisory service, but does not include any person specified in the First Schedule).

[79] Financial Advisors Regulations (Cap 110, Reg 2, 2004 Rev Ed Sing)

[80] D Neo, “Singapore: Boosting Regulation to Protect Vulnerable Investors” in Booysen, supra note 11 at para 11.10 [Neo].

[81] FAA, supra note 77, ss 34(6), 35(3), 36(3).

[82] Neo, supra note 80 at para 11.56.

[83] Ibid.

[84] Securities and Futures (Classes of Investors) Regulations 2018 (S 665/2018, Sing), regs 3(2)–3(3).

[85] See e.g. Deutsche Bank AG SGCA, supra note 1.

[86] Consumer Protection (Fair Trading) Act 2003 (2020 Rev Ed Sing).

[87] Ibid, s 35.

[88] Ibid, s 4.

[89] Ibid, Second Schedule.

[90] Ibid, s 3.

[91] Yahoo News, “Super-app, moomoo, is the only investment platform to break through top five most downloaded finance apps in Singapore in 2021 since its launch” (March 8, 2022) , online: <https://finance.yahoo.com/news/super-app-moomoo-only-investment-060000424.html?guccounter=1&guce_referrer=aHR0cHM6Ly93d3cuZ29vZ2xlLmNvbS8&guce_referrer_sig=AQAAAHsCPBDVzFGtIhHvpTNnDm4K3xRridHHv5Jda6JlFhvYtWQZopfo-wDweJCujr-srCIqwMBtWL-CrGeyUZ5btZ_UtKwj8wu-XaOIIGvhoQAi_VLjaYTSGgQT-ikAPaGZAF-9WQ_je1Akr4FSapghSFFauFl9k5HTb_jcRy1AIZ0C>.

[92] Ibid (for moomoo’s fight to be the top five most downloaded finance app in Singapore in 2021).

[93] See e.g. for the United States: Robinhood’s business model. See Investopedia, “How Does Robinhood Make Money?” (08 October 2022), online: <https://www.investopedia.com/articles/active-trading/020515/how-robinhood-makes-money.asp> .

[94] See e.g. for the United States: Robinhood. See CNBC, “Robinhood to pay $70 million for outages and misleading customers, the largest-ever FINRA penalty” (30 June 2021), online: <https://www.cnbc.com/2021/06/30/robinhood-to-pay-70-million-for-misleading-customers-and-outages-the-largest-finra-penalty-ever.html>.

[95] See the discussion on the UCTA, MA and FAA in the earlier sections.

[96] See e.g. JP Morgan Chase Bank v Springwell Navigation Corp [2008] EWHC 1186 (Comm) ; Deutsche Bank AG SGCA, supra note 1.

[97] Ibid.

[98] Bank Leumi (UK) plc v Wachner [2011] EWHC 656 (Comm) [Bank Leumi].

[99] Elise Bant & Jeannie Paterson, “Limitations on Defendant Liability for Misleading or Deceptive Conduct Under Statute: Some Insights from Negligent Misstatement” in K Barker, R Grantham & W Swain, eds, The Law of Misstatements: 50 Years on from Hedley Byrne v Heller (London, UK: Hart Publishing, 2015) at 162; K Barnett, “Causation, Remoteness and Calculation of Damages for Financial Mis-Selling” in Booysen, supra note 11 at para 7.065 [Barnett].

[100] See comments by judges in Deutsche Bank AG SGCA, supra note 1.

[101] Bank Leumi, supra note 98.

[102] Elders Trustee and Executor Co Ltd v EG Reeves Pty Ltd (1987) 78 ALR 193 (FCA); Barnett, supra note 99 at para 7.084.

[103] Bank Leumi, supra note 98; Zaki v Credit Suisse (UK) Ltd [2011] 2 CLC 523 (Com Ct). Compare Wingecarribee, supra note 73; Rubenstein v HSBC Bank [2012] EWCA Civ 1184 [Rubenstein].

[104] Hadley v Baxendale (1854) 9 Ex 341; Note Transfield Shipping v Mercator Shipping Inc (The Achilleas) [2009] 1 AC 61 (HL) was rejected in Out of the Box Pte Ltd v Wanin Industries Pte Ltd [2013] 2 SLR 363 (SGCA).

[105] Overseas Tankship (UK) Ltd v Morts Dock & Engineering Co Ltd (The Wagon Mound (No 1)) [1961] AC 388 (PC); Overseas Tankship (UK) Ltd v The Miller Steamship Co Pty Ltd (The Wagon Mound (No 2)) [1967] 1 AC 617 (PC).

[106] Barnett, supra note 99 at para 7.076.

[107] South Australia Asset Management Corporation v York Montague Ltd [1997] AC 191; Manchester Building Society v Grant Thornton UK LLP [2021] UKSC 20; Rubenstein, supra note 103.

[108] Royscot Trust Ltd v Rogerson [1991] 2 QB 297 (CA) was rejected in RBC Properties Pte Ltd v Defu Furniture Pte Ltd [2015] 1 SLR 997 (SGCA).

[109] Sim Poh Ping v Winsta Holding Pte Ltd [2020] 1 SLR 1199 (SGCA) at para 238.

[110] See Brickenden v London Loan & Savings Company of Canada [1934] 3 DLR 465 (PC) (for the ‘Brickenden rule’).

[111] Hodgkinson v Simms [1994] 3 SCR 377 (SCC).

[112] Ibid; Peter Millett, “Equity’s Place in the Law of Commerce” (1998) 114(2) Law Q. Rev 214 at 225.

Does Copyright Last for Too Long in Singapore?

A PDF version of the article can be found here.

Does Copyright Last For Too Long In Singapore?

Desmond Chye*

ABSTRACT

It is conventionally accepted that copyright needs to last beyond the death of the author in order to incentivise authors to create works, and that the lengthy protection period is but just a small price to pay to promote greater creativity. This orthodox logic is however riddled with flaws. Principally, copyright actually provides very little economic incentives for the author to create to begin with. Moreover, an author’s non-monetary desire to publicly associate with his or her work does not readily apply after the author’s death. Times have also changed such that public policy concerns now militate against having an extensive copyright duration. Finally, copyright law as-a-whole has not sufficiently mitigated the impact of a long copyright on the public’s access to the work to the point where the long duration can be justified. This article therefore submits that the right duration should not exceed the life of the author.

I.                    INTRODUCTION

The conventional narrative is that copyright for authorial works should last a long time to economically incentivise authors to create creative works that would ultimately benefit society. In fact, so persuasive is this view that the global trend over the last two hundred years, of which Singapore is of no exception,[1] is to dramatically increase the length of the copyright term to truly biblical proportions: the life of the author plus 70 years after his or her death.[2] If we consider the average global life expectancy of a human today, at 72 years,[3] the average copyright term over an authorial work would easily exceed a hundred years! While this is a very good thing for the owner of the copyright and his or her heirs, the same cannot be said of the public at large as a generation of people would be deprived of the work’s benefits. The question is thus whether such an extensive copyright term strikes the right balance between promoting intellectual creation and society’s need for a free flow of information. This article finds that the current copyright duration does not achieve the right balance at all. To permit copyright to extend beyond the death of the author, let alone for another 70 years, is excessive. It is therefore proposed that copyright should be shortened to the life of the author at most.

II. EVOLUTION OF THE COPYRIGHT TERM

To understand how Singapore got to the point where the copyright term for authorial works is ‘life plus 70 years’, we need to look at the evolution of copyright in the common law world, starting from its origins in the United Kingdom (“UK”). It was not always the case that the copyright duration was very lengthy, or even pegged to the life of the author. In fact, when copyright began life as a creature of statute in the UK with the passing of the Statute of Anne 1709, it only lasted for a very short copyright period of time: a maximum of 28 years.[4] It was only in 1814 that the 28 year term was substituted for the life of the author. This lasted for a while until the Copyright Act 1842 which lengthened the term to 42 years from publication or until 7 years after the death of the author, whichever is longer. The 42 years period remained the status quo until the enactment of the Copyright Act 1911 (and then in Singapore through a proclamation made by the Governor of Singapore) which then extended the copyright term to ‘life plus 50 years’. This British duration then became the universally accepted minimum period across the world via the Berne Convention of 1948 (“Berne Convention”). The Berne Convention period was subsequently extended by another 20 years through the TRIPS Agreement of 1995.[5] Singapore adopted this 20 year extension on 1 July 2004 which has been retained in the new Copyright Act 2021.[6]

III. ANALYTICAL FRAMEWORK FOR THIS ARTICLE

As there are a myriad of arguments that may favour the current length of copyright, this article will split the analysis into four parts. The first part will deal with the economic arguments for the current copyright term. Here, the article will disprove the conventional narrative that authors are predominantly money-minded and thus need a copyright term that lasts beyond their deaths to economically incentivise them to create. The second part will explore the possible non-economic arguments for a long copyright and find that these are not sufficiently compelling to justify a multi-generational copyright. The third part will then argue that the policy arguments for the long copyright duration are outdated and unsound. In the final part, this article will posit that copyright law as-a-whole (subsistence, infringement and defences) has not sufficiently mitigated the impact of a long copyright on the public’s access to the work to the point where the long duration can be justified.

IV. ANALYSIS OF THE ECONOMIC ARGUMENTS

The argument here is that copyright “provides the economic incentive that is essential to the creation of new works”.[7] It goes like this: copyright grants authors exclusive rights in their works,  allowing them to gain financial rewards by monetising those rights.[8] A longer copyright term would thus mean a greater incentive to create.[9] If so, then the copyright term should last until the incentive value of a longer term diminishes to insignificance. To the world at large, this ‘vanishing point’ of marginal utility is determined to be somewhere between one to two generations after the author’s death.[10] The ‘economic incentive’ theory however rests on several flawed assumptions, which will be elaborated on below.

A.                 Creators are not ‘rational profit maximisers’

It is conventionally assumed that authors are ‘rational profit maximisers’ who would only be willing to expend time, energy and resources in proportion to the expected monetary gains from their work. This is however false as creators can be motivated by reasons besides money.[11] Examples of creators working for free would be those in the open source movement where programmers create and share software for free and Wikipedia where anonymous volunteers contribute content pro bono.[12] Closer to home, Singaporeans creators from all kinds of backgrounds have worked for free. Former Prime Minister Lee Kuan Yew donated royalties from his bestselling books to charitable causes.[13] Local artist Peter Kiew sketches strangers on the MRT without charge and would vehemently reject any attempt by contented recipients to pay him for his work.[14] In perhaps the most extreme example, vandals Andreas Von Knorre and Elton Hinz sprayed graffiti on Singapore’s MRT trains despite knowing full well that it would have meant hefty fines (and lengthy imprisonment).[15] These examples show that the call to create need not be from monetary incentives only. The desire to do a good deed, self-satisfaction from creating something or garnering social reputation (good or bad) may be just as incentivising as a pot of gold to the author.

B.                  Copyright gives too little economic incentives to create

Even if we buy the ‘rational profit maximisers’ theory, any money to be gained from a multi-generational copyright is too uncertain to be sufficiently consequential in incentivising people to create. The biggest issue is that the chance of a creative work being successful is not very high to begin with.[16] A writer’s odds of success are dismal. According to the Huffington Post, only 2% of all books will sell beyond 5,000 copies.[17] Digitalisation has also enabled many to be self-publishers, saturating the market and putting success out of reach of all but the lucky few.[18] A similar story is also unfolding for artists. Only 19% of US artists made over $50,000 USD a year in 2016, well below the median US household income of $58,000 USD. This is likely to be worse in Singapore as our society does not regard art as particularly essential.[19] On the music front, songwriters are also being pummelled. In a recent article by the UK’s Guardian newspaper, it was estimated that only one in 10 artists who sign to labels go on to achieve commercial success.[20] The research sector also does not appear to give much profit to its authors. The famous economist Joseph Stiglitz observed that the bulk of important research originated from non-profit governmental or educational institutions, which pay researchers a salary to continue researching, rather than from commercial entities who compensate the researchers with intellectual property instead.[21] While YouTube content creators appear to be earning substantial coin from their work,[22] this is likely an illusion. A recent German study found that 96.5% of aspiring YouTubers fail to earn enough money to pass the US poverty line.[23] These examples illustrate that a long copyright would be hardly motivational for creators when their works can barely support themselves, let alone their successors.

C.                 The lottery theory fails to justify a long copyright

Perhaps the ‘economic incentive’ from copyright is not from assured returns but rather from the hope of getting great rewards. Under the ‘lottery’ theory, creators are willing to toil away despite the low chances of financial success because they seek the chance to strike the proverbial pot of gold at the end of the rainbow.[24] Copyright would be instrumental here as it protects the chance of winning the lottery. The longer the term, the more chances the creator will get to win big.[25] While this may be a plausible reason in more risk-taking nations, it is unlikely to be so for Singapore because we are notoriously risk-adverse.[26] In fact, a key concern for anyone joining the Singapore art industry is the lack of a stable income to offset the considerable time and monetary investment required to join it.[27]

Aside from whether the lottery theory adequately explains why authors create, there are also serious doubts as to whether copyright law should promote risk-taking when the chances of success are so low.[28] This is because people who gravitate towards this kind of risk taking may be afflicted by an ‘optimism bias’ that clouds their ability to weigh the pros and cons of their actions properly and ultimately encourages them to make unwise decisions or act imprudently.[29] It is thus submitted that copyright should not be too long as it would pander to such dangerous behaviour.

D.                Copyrighted works generate most of their income at the start of their term

Another problem with the conventional narrative is the assumption that a copyrighted work would generate earnings throughout its entire copyright term, which is false for the vast majority of works. The bulk of the income from a copyrighted work actually comes from the immediate years post publication. For books, how long it stays on the shelf is up to the ‘vagaries of popular taste’.[30] If a book is no longer popular, the publisher would stop selling it to make way for more in-demand books.[31] The same applies to other works like art, music and movies in today’s consumerist society. This phenomenon is especially problematic for the author because we live in an era of short attention spans where the purchasing public’s interest in a work can surge rapidly but also fade at a similar pace.[32] In essence, the author’s work must be in demand now or it may never be so. While research articles may be more evergreen,[33] it is often the case that their contents cannot be monetised for a significant amount of time post publication,[34] if they can even be monetised at all.[35]

E.                  Authors do not necessarily benefit from the copyright

There are also numerous situations where an author may not benefit from the copyright in his or her work at all. Although the default position in the Copyright Act 2021 is that the author is the first owner of the copyright in his or her work,[36] the exceptions to this rule can encompass a large proportion of copyright work creators. An example is the employer-employee exception. Unless one is lucky, we will all be employees at some point in our lives, and in predominantly white-collar Singapore,[37] this means we will write many literary works for the organisation we are serving. These works may be commercially valuable to us, the white-collared author, but the copyright would be almost always vested in the employer as the default position is that the employer automatically acquires the copyright in the employee’s works if the works were done in the course of the author’s employment.[38] This is made worse by how ‘course of employment’ is liberally interpreted to also include things that the author ought to have created for the employer and not the employer’s rivals.[39] It is thus rare for the public at large to ever benefit from the copyright in their work despite contributing so disproportionately to the overall volume of copyrighted works.

F.                  More monetary rewards do not equal more creativity

Lastly, the conventional narrative’s assumption that there is a causal link between monetary rewards and greater creativity is also suspect. Dangling monetary rewards may instead undermine the quality of works produced.[40] To explain why this is so, we need to understand how creators are motivated to create. A creator has two sources of motivation: the first is an intrinsic one to get self-satisfaction from expressing one’s creativity, and the second is an external one to create for some rewards, usually monetary.[41] While having an excess of the first is not problematic, having too much of the second is. If the extrinsic motivator becomes so strong that the creator perceives it as controlling, then creativity drops.[42] This is because the creator would treat the creative activity as “a means to an end rather than an end in itself”, making him less personally invested in the work at hand, resulting in a less creative product.[43] It is thus said that “[t]ruly creative people respond most strongly to some innate drive to solve problems or to produce art and are unlikely to be encouraged to make a greater effort by the promise of profit if their work is successful” .[44]

The correlation between lower creativity and higher monetary incentives is backed by empirical evidence. Experiments by various psychologists have shown that “higher monetary incentives [lead] to worse performance” when cognitive work is being done.[45] Outside of the university context, we see this play out regularly in the arts scene, especially in the movie sector. Some of the most commercially successful movies have been slammed for being ‘unimaginative’. Take the famously panned yet commercially successful third movie of the third trilogy of the popular Star Wars genre as an example. A quick look at the influential movie review aggregation website Rotten Tomatoes reveals that the plot was generally criticised as ‘bland’ and ‘derivative’ by the audience, with much of the problem caused by the studio’s desire to stick to tried and tested tropes to maximise sales.[46] Such are the perils of using money to encourage more creative works.

V. ANALYSIS OF THE NON-ECONOMIC ARGUMENTS

There are two possible non-economic arguments for a long copyright duration: to protect the creator’s moral rights, and to cater to the creator’s familial instincts to provide for his or her family. This section will look at each in turn and conclude that both are weak justifications for a long copyright term.

A.                 The ‘moral rights’ argument is weak

The moral rights argument goes like this: authors have a personal connection with their work. Thus, they want to protect their personality as expressed in it.[47] Conferring on authors ‘moral rights’ in their work, such as the rights of disclosure, attribution and integrity, enable them to remain publicly associated with their work.[48] Without such rights, authors would not be motivated to create new works as their expression would just be appropriated by someone else post publication.[49] In the context of the copyright duration, moral rights enable the author (and his or her immediate successors) to continue to require attribution as a condition precedent to further reproduction of his or her work. This would then allay the author’s concerns that his or her work would be passed off by someone else, thereby encouraging the author to create more works.

However, the moral rights rationale does not justify copyright lasting beyond the author’s death. As perceptively observed by Ricketson, the “authors’ natural concern to protect their moral rights during their lifetime becomes less natural after the author’s death”.[50] Moreover, there are also no practical benefits to be gained by authors from the way their work is treated post death.[51] Perhaps the preservation of one’s legacy after death is more compelling on those who believe in the existence of an afterlife. To these authors, the prospect of looking down from above to see their work still being attributed to them might be the motivation to create while they still mingle with the living. This kind of motivation is however unlikely to be prevalent amongst Singaporean authors as only 34.5%[52] of the population subscribes to a religion that believes in an afterlife.

B.                  Copyright’s appeal to the author’s familial instincts is minimal

The next oft-cited reason for why copyright has to last so long is the need to appeal to the author’s natural instincts to provide for his or her spouse and children. While this can be a strong emotional pull on individuals to create, it simply does not apply to the vast majority of creators. As explained previously, the financial incentives from creating copyrighted works are not usually high enough to be even capable of sustaining the author during his life, let alone his or her next generation and spouse. This theory also does not adequately explain why some of the most creative people did not have any family to leave their works behind for. Leonardo da Vinci, John Locke and Ludwig van Beethoven were all single throughout their life but yet they each produced some of the world’s most influential and timeless works. In Singapore, the iconic singer-songwriter and film director Dick Lee and legal academic cum accomplished playwright Eleanor Wong are also similarly without any offspring to leave their influential works to. Even if we do recognise the power of familial instincts as a core driver of creative content, its relevance to Singapore is diminishing as more and more Singaporeans are shunning children and even marriage.[53]

VI. ANALYSIS OF THE POLICY ARGUMENTS

A.                 A multi-generational copyright fails to advance Singapore’s societal objectives

It is often said that the primary policy reason why copyright lasts beyond the death of the author is to achieve the societal objective of enabling authors to provide for their family after their death,[54] presumably so that the government does not have to pick up the tab. This reason is however flawed on many levels. The most obvious flaw is that this view is hopelessly out of date. While it is true that copyright was necessary to support the immediate descendants of the author when the copyright term was extended to life plus 50 years in 1911, we no longer live in that bygone era. Back then, most families relied on a single breadwinner for income[55]. An author who spent many unpaid years creating his work but perished not long after its publication would have led to his family becoming destitute, and this was not a remote possibility as a person’s life in that era was often brutish, nasty and short.[56] Authors and their families no longer face such a predicament these days as single income families are now the minority in Singapore.[57] When you add in the extra facts that household incomes have ballooned[58] and education levels have skyrocketed,[59] it would be even more ludicrous to think that an inheritable copyright is necessary to stave off family poverty. The reality today is that an author’s premature death and corresponding demise of the copyright would not condemn the author’s family to crippling poverty because the surviving family members would likely be rich enough to care for themselves and highly employable in their own right. Moreover, as the average lifespan has increased dramatically in the last hundred years,[60] there would now be more than enough time for authors to earn enough money to bequeath to their family by producing more works, eliminating the need to rely on continued copyright royalties from any one particular work after the author’s death.

Another flaw with an inheritable copyright is that it goes against the principles of a meritocratic society. This is especially problematic for Singapore because we strongly aspire to achieve such a society.[61] A meritocracy is a social system where people get ahead in life based on their own accomplishments rather than on extraneous factors, such as their parents’ social class or wealth. As such, the core aspect of a meritocratic society is that one must have earned what he has received – an individual’s wealth or status must have been ‘merited’ by his deeds. Inheriting the fruits of someone else’s labour would thus be incongruous with such a society. Unfortunately, this is exactly what an inheritable copyright does. The descendants played no part in the creation of the author’s work but yet are allowed to benefit from it. In short, they did not deserve the copyright proceeds at all.

While this article recognises that the meritocracy argument can be extended to include the banning of all forms of inheritance, it makes no comment on it.[62] It would however suffice to say that an inheritable copyright poses more problems than other kinds of inheritable property because, unlike those, copyright can have a very detrimental impact on society’s technological development if it lasts for too long. Property rights are inherently ‘selfish’ as they deliberately create zones of exclusivity.[63] This is problematic when the property that is fenced up is potentially useful information for society. If society cannot access the information freely for multiple generations, this can produce a significant drag on its technological progress. It is therefore imperative that we treat an inheritable copyright as much more than just a mere bequest to one man.

Conferring upon the descendants of the author a windfall would also have the perverse impact of discouraging them to be as motivated to create their own intellectual property. In a capitalistic society like Singapore’s, the enticement to produce is the core driver of productivity in the economy so the most effective way of enticing people to produce would be to distribute income and wealth according to the productivity of each person.[64] This is however absent in the case of an inheritable copyright because the copyright which is inherited by the author’s successor would be divorced from the successor’s own productivity. As a result, since the author’s successor would be richly rewarded without the need to work anyway, the successor would be naturally discouraged from creating his or her own works.

Furthermore, permitting long copyrights would entrench the dominance of the reigning intellectual property giants of the world at the expense of global economic development. It is a common criticism of intellectual property rights that they are truncheons wielded by the more developed world to club the less developed into delaying their rise up the economic value chain. Although one would no longer think of Singapore as a developing nation these days, we are apparently still being clubbed by more developed nations. Despite the shift towards a knowledge based economy, we remain a significant net importer of intellectual property.[65] To preserve the copyright term would thus perpetuate the imbalanced trade relationship between the intellectual property giants and us. Moreover, as Singapore has itself gained quite a fair repertoire of intellectual property, there is now the moral issue of whether we should also partake in the same clubbing of less developed nations for our own selfish gains.

B.                     A long copyright term is superfluous to protect the author’s privacy

A potentially more convincing policy reason for a long copyright may be that it serves an important role in protecting the author’s privacy. This would be especially pertinent to Singapore as we do not have a statutory or common law right to privacy per se.[66] It might appear odd to use copyright to protect one’s privacy but it can be a powerful tool if the right conditions are met. How it does so is by permitting the copyright owner to restrain any unauthorised reproduction of the copyrighted material which then indirectly prevents its dissemination to the public. An example of how this would work in practice is the fact scenario in Lee Wei Ling v Attorney-General.[67] The dispute there revolved around a contractual arrangement between the Singapore Government and former Prime Minister Lee Kuan Yew regarding the rights in some transcripts of interviews of the latter by the former. Under the arrangement, the former Prime Minister was to have the copyright over the transcripts while the Government was to have only the physical right to possession to them. This legal construct had the effect of restraining the government from reproducing the transcripts without the former Prime Minister’s consent, effectively granting him the right to privacy in those documents.[68]

Interestingly, the previous iteration of the Copyright Act[69] also facilitated this need for greater privacy safeguards in Singapore by enabling copyright to be perpetual in unpublished works.[70] This was possible because the copyright term in an authorial work was pegged to the time when the work is first published or made publicly available.[71] The relevant provisions were however repealed in the Copyright Act 2021 as Parliament found that the need for a bigger pool of works in the public domain outweighed the justifications for privacy.[72]

However, privacy is ultimately not a strong justification for a long copyright. For one, there is a much better way to protect one’s privacy: through the breach of confidence tort. Under this better method, the protection offered would be more comprehensive, there is no need for the information to be fixed in some material form,[73] more types of information can be protected,[74] and there are also fewer ways for a defendant to defend his reproduction of information.[75] Another more fundamental reason is that protecting the author’s privacy is not the purpose of copyright law. Copyright in Singapore has always been viewed as a means to an end – the end being the creation of works for the public to eventually enjoy.[76] It would thus be illogical to increase the copyright term so as to protect works from ever being seen by the public.

C.                 A long copyright term is impractical

Even if we ignore the policy arguments raised thus far, there is still the issue of impracticality arising from copyright lasting multiple generations, in particular, the problem with orphan works. Orphan works are copyrighted works whose owner cannot be identified or located because too much time has elapsed since the author’s death.[77] Such works are highly detrimental to society because “[t]he inability to request for permission to use the work means that it cannot be legally used, even if the prospective user has spent much time and effort to find the owner”.[78] Unfortunately, no adequate solution has been found for this pressing problem. An ‘orphan works registry’ was floated as a possible addition to the Copyright Act 2021 but it was shot down by the law reform committee because it was deemed too administratively costly for the government to operate and would have also added litigation costs for the parties should they fail to come to an agreement.[79] Aside from this copyright registry idea, other methods involving the government or a court determining the appropriate copyright fee were also proposed[80]. These proposals were similarly rejected, on the basis that they would  either lead to undue market interference, would be too costly for the government to administer or insufficient to prevent costly litigation between the copyright owner and the copyright user.[81]

In light of the foregoing, it is submitted that we have made life unnecessarily difficult for ourselves. The ‘Gordian’s knot’ can be easily untied if we just reduce the copyright term to only the author’s life. There would no longer be any orphan works to worry about as it would become very easy to know when the copyright has expired – when the author’s obituary appears in the newspapers.

VII. IS THERE SUFFICIENT MITIGATION?

Notwithstanding the points raised thus far, some may still argue that the nature of copyright as a limited monopoly right can still accommodate the longer ‘life plus 70 years’ copyright term without overly compromising the public’s free access to the work. The thrust of such an argument would be that the legal requirements to find copyright subsistence and infringement are broad enough such that the public can still make use of the copyrighted work freely for useful purposes. It is however submitted that the legal exceptions to copyright’s exclusive power are actually insufficient and even if they may be, they are built on pillars of sand.

To establish copyright infringement, the owner must prove that, firstly, copyright exists in the work allegedly infringed; secondly, there was a substantial taking of the owner’s work; and, thirdly, the alleged infringer does not have any defences that can be raised. This article will look at each step in turn.

A.                 Copyright subsistence

Beginning with the first step, finding copyright subsistence, the key criterion that the copyright owner needs to prove is that the copyrighted work in question had sufficient ‘originality’. This is found when the author applied intellectual effort towards the creation of the work.[82] As a result, only expressions of facts and ideas can give rise to copyright. This requirement per se is however not particularly impactful as anything other than an unorganised dump of raw facts or ideas would have sufficient originality to find copyright. The common law therefore created the additional de minimis threshold which the work must pass to have copyright.[83] Under this rule, the work must be sufficiently ‘consequential’,[84] which is when the work is not ‘commonplace or banal’.[85]

One may argue that it is through this de minimis threshold that the effects of a long copyright can be mitigated. If common phrases could be the subject of copyright then the free flow of information in society would be greatly constricted. It could even go as far as undermining the “evolution of a language and hence the culture of a society”.[86] As such, by providing courts with a legal mechanism to exclude certain kinds of works from falling under the protection of copyright based on how banal or commonplace they are, the law can avoid the problem of allowing such works to have copyright.

However, it would appear that the de minimis threshold is incongruous with the Copyright Act 2021. To see whether an authorial work is sufficiently consequential to enjoy copyright would amount to evaluating the work on its merits but that is not permitted under the Act – as per a plain and purposive reading of it.[87] The common law de minimis requirement was therefore introduced per incuriam and would likely crumble in the face of a direct challenge to its validity.

Moreover, even if some legal witchcraft can be conjured up to keep it consistent with the statute, the effect of the de minimis threshold is rather minimal. Only the most banal and commonplace works appear to be deprived of copyright. This is as very basic things such as ‘rudimentary drawings’ of fire doors[88] and telephone directories[89] have been found sufficiently consequential to have copyright. If such common and useful things can remain locked behind a copyright paywall, then it is hard to say whether the de minimis threshold has a more than de minimis impact on preventing works from being unduly fenced up from the public domain.

B. Copyright infringement

Moving on to the next step, the question is whether there was a ‘substantial taking’ of the copyright owner’s work. The legal test to establish this is quite lenient to the copyright owner and thus fails to mitigate the negative impact of a long copyright. Although the test appears difficult to make out as the alleged infringer must have overly appropriated the benefit of another’s skill and labour,[90] it may not actually be so as the copyright owner can avail himself to multiple legal doctrines to make it much easier to find a substantial taking. The first is the prima facie presumption of copying which would arise whenever the defendant’s work is substantially similar to the plaintiff’s work and the defendant had prior access to it.[91] This would then shift the burden of proof to the defendant to then show that he did not copy the plaintiff’s work. As we all know, whoever who bears the burden of proof would be at a disadvantage in court. This is further compounded by the fact that the defendant’s intentions are irrelevant and even subconscious copying would amount to an infringement.[92]

The second stone around the defendant’s neck is the potential for infringement to be found even though the defendant’s work did not share any identical portions with the plaintiff’s work – the ‘altered copying’ situation. In such cases, the defendant’s work did not directly lift any portion of the plaintiff’s work into his, but rather, incorporated the plaintiff’s work with modifications. Intuitively, one would expect no finding of infringement since what is appropriated would be the plaintiff’s idea and not his expression of the idea. Case law however disagrees and finds that an idea can be protected so long as the idea taken was a ‘detailed’ one.[93] In the English case of Designers Guild Ltd v Russell Williams (Textiles) Ltd,[94] the plaintiff claimed that her fabric design, which comprised flowers superimposed onto a red and white striped background, was infringed by the defendant. Both designs however did not share any identical elements: the flowers were of a different type and the stripes were of a different thickness. Only the overall impression was similar. The plaintiff therefore claimed that the defendant appropriated her idea of using ‘flowers superimposed on a red and white striped background’. The House of Lords held that this was sufficient to find infringement by the defendant. In their Lordships’ opinion, there was no need for any identifiable part of the defendant’s work to be the same as the plaintiff’s for there to be infringement as having sufficient similarity between the two would suffice. As such, since the defendant’s and plaintiff’s works were similar, there was infringement. [95] In so holding, the court permitted ideas to be protected. Recognising altered copying poses a big problem for copyright law because it blurs the line between ideas and expressions. Should copyright start protecting the former then the pool of information that the public can use freely would be significantly decreased and this would be greatly exacerbated if the copyright term was biblically long.

C. Defences to copyright infringement

We shall now turn to the last step – whether the defendant can avail himself of any defences to excuse his act of infringement. Numerous permitted uses in the Copyright Act 2021 would excuse an act of infringement.[96] Except for the open-ended general fair use defence in section 190[97] where the only condition to satisfy is that the defendant’s use must be fair, which is to be assessed by reference to all factors, including the Statutory Factors,[98] the other defences are all narrowly defined. The general fair use defence is therefore the only plausible means to mitigate the effects of a long copyright sufficiently, but as will be shown below, it fails to do so.

It is submitted that the most important thing needed to mitigate the effect of a long copyright for the general fair use defence is to permit the public to freely use the copyrighted work so long as the purpose of doing so is to build upon it. This would be possible if the ‘transformative use’ doctrine is used to find fair use. Under the doctrine, fair use would be found if the defendant’s work does not “merely supersede the objects of the original creation” but rather “adds something new, with a further purpose or different character, altering the first with new expression, meaning, or message”.[99] Such an approach would provide the right balance between the protecting the copyright holder’s interests and the public’s because the copyright user would be allowed to benefit society using the author’s works so long as he does not directly compete with the author by using his works for the same purposes as the author.

However, the open-ended general fair use defence is a surprisingly narrowly circumscribed one that has no room for the transformative use doctrine. It remains to be seen whether the local courts would embrace the transformative use doctrine[100] but it is unlikely that they would. For one, Singapore has retreated from a purely utilitarian approach to copyright[101] which would encourage courts to support the doctrine. We are now placing greater emphasis on the ‘moral’ aspects of the copyright, as evidenced by how the new Copyright Act 2021 has an entire section devoted to the author’s moral rights, which include, amongst others, the ‘right to be identified’.[102] Apart from this, the Singapore Government is also trying to brand copyright infringement as an act of ‘stealing’ in the minds of the public.[103] Another reason is that the structure of the Copyright Act 2021 is at odds with the transformative use doctrine. The Statutory Factors expressly require the courts to consider not just the purpose of the infringement, but also the effect the infringement has on the copyright owner’s income and the nature of the work taken[104] which must necessarily mean that the transformative use doctrine by itself, which looks only at the purposes of the works, cannot be decisive.

Notwithstanding the above, even if we treat the general fair use doctrine as sufficient mitigation, there is still the possibility that an open-ended fair use doctrine is incongruent with Singapore’s treaty obligations under the TRIPS Agreement. Article 13 of TRIPS sets out a three-step test to evaluate the legitimacy of any exceptions to copyright: the exception must be confined to certain special cases, the exception must not conflict with the normal exploitation of the work, and the exception does not unreasonably prejudice the legitimate interests of the right holder.[105] A local court has yet to interpret Article 13 but a very persuasive body, The World Trade Organisation (“WTO”), has done so. A WTO panel in the United States interpreted Article 13 strictly: each criterion acts on a cumulative basis, with each step constituting a separate and independent requirement.[106] As Article 13 begins with the criterion that the exceptions to the copyright owner’s rights must be confined to ‘certain special cases’, the overriding obligation imposed by TRIPS on member states must be to ensure that any exceptions to the copyright owner’s rights are narrow in scope and clearly defined – in short, no open-ended exceptions are permitted.[107]

VIII. A COMPROMISE SOLUTION?

Even if we were to ignore all the arguments above and decide to keep the ‘life plus X years’ formulation, the length should still be shortened to the original Berne Convention duration. This is as the recent extension to 70 years from 50 appears to have been made on dubious grounds. Rather disturbingly, the 20 year extension was primarily driven by administrative convenience rather than a rigorous assessment of whether the extension was justified from the standpoint of the author or the public. As observed by Dworkin, the main reason for the harmonisation of the copyright term to 70 years from 50 years is that it was more “convenient to harmonise upwards than downwards”.[108] This was as the European Union (“EU”), a major proponent of the TRIPS Agreement, feared that the process of making ‘70 year’ member states reduce their term to 50 years would have entailed too many transitional provisions and would have set back the EU’s objective of creating a single market by the dawn of the 21^st century.[109] It is thus submitted that the TRIPS copyright extension was an act of a politician’s wants overruling the public’s needs which cannot be justified.

IX. CONCLUSION

Gripes with copyright have existed since the beginning of copyright itself. Achieving the right balance has always been difficult and will continue to be so. It is however worrying that the trend in the last 100 years is to blindly adhere to a mantra of continuous extensions to the copyright term which is increasingly divorced from the reality on the ground. Copyright holders are now akin to ‘dynastic kings’ who control a rapidly expanding domain of intellectual property that they can pass down to multiple heirs before entering the commons. Unfortunately, the great irony is that this system helps neither the author nor the public. We therefore need to shake off our rose-tinted glasses and see the conventional narrative of ‘long copyrights equal great benefits to society’ for what it really is – outdated and unsound reasoning. Ultimately, it is hoped that the points raised in this short article can prod policymakers to pare back the unduly long copyright duration to a more reasonable one that lasts no longer than the life of the author.

---

*  LLB (Candidate), National University of Singapore, Class of 2023. All errors in this article remain my own.

[1] Singapore adopts the ‘life plus 70 years’ duration in the Copyright Act 2021, s 114 [Copyright Act 2021].

[2] Agreement on Trade-Related Aspects of Intellectual Property Rights, 15 April 1994, Marrakesh Agreement Establishing the World Trade Organization, Annex 1C, 1869 UNTS 299, 33 ILM 1197 (1994) art 9(1) (entered into force 1 January 1995, as amended 23 January 2017) [TRIPS Agreement].

[3]The World Bank, “Life expectancy at birth, total (years)”, online: <https://data.worldbank.org/indicator/SP.DYN.LE00.IN> (last accessed 9 Dec 2021).

[4] There was an initial term of 14 years with an additional 14 years if the author was still alive after the initial term’s expiry.

[5] TRIPS Agreement, supra note 2, art 9(1).

[6] Copyright Act 2021, supra note 1, s 114.

[7] Stanley M. Besen & Leo J. Raskind, “An Introduction to the Law and Economics of Intellectual Property” (1991) 5:1 J. Econ. Persp. 3, 5.

[8] Ibid.

[9] Ibid.

[10] 164 countries have signed the TRIPS Agreement which mandates that the necessary copyright duration should be the length of the author’s life plus 70 years after the author’s death. This duration is likely to encompass the life of the author’s immediate successor and the generation after that, depending on the age of the immediate successor when the author created the work.

[11] Diane Leenheer Zimmerman, “Copyrights as Incentives: Did We Just Imagine That” (2011) 12:1 Theoretical Inq L 29, at 43 [Diane Leenheer Zimmerman].

[12] Ibid.

[13] Mary Kwang, “Royalties will go to education in China and Singapore” (9 Oct 1998), Singapore Press Holdings online:<http://www.vrijmetselaarsgilde.eu/Maconnieke%20Encyclopedie/FMAP~1/REFORM/reform3/lee2_1009.html> (last accessed 9 Dec 2021).

[14] Justin Ong, “‘No use for money’: 72-year-old artist sketches strangers on the MRT for free” (29 Dec 2020), The Today online: <https://www.todayonline.com/singapore/no-use-money-72-year-old-artist-sketches-strangers-mrt-free> (last accessed 9 Dec 2021).

[15] Elena Chong, “Two Germans who vandalised MRT train in Bishan depot get 9 months, 3 strokes each” (5 Mar 2015), The Straits Times online <https://www.straitstimes.com/singapore/courts-crime/two-germans-who-vandalised-mrt-train-in-bishan-depot-get-9-months-3-strokes> (last accessed 9 Dec 2021).

[16] Diane Leenheer Zimmerman, supra note 11, at 38.

[17] William Dietrichhe, “The Writer's Odds of Success” (4 Mar 2013), The Huffington Post online: <https://www.huffpost.com/entry/the-writers-odds-of-succe_b_2806611> (last accessed 9 Dec 2021).

[18] Ibid.

[19] Prisca Ang & Chelsea Kiew, “Artists defend value of creative work to society after survey sparks debate” (16 Jun 2020), The Straits Times online: <https://www.straitstimes.com/lifestyle/arts/artists-defend-value-of-creative-work-to-society-after-survey-sparks-debate> (last accessed 9 Dec 2021).

[20] Rhian Jones, “Songwriters fight to be heard in streaming revenues debate” (12 Feb 2021), The Guardian online: <https://www.theguardian.com/music/2021/feb/12/songwriters-fight-to-be-heard-in-streaming-revenues-debate> (last accessed 9 Dec 2021).

[21] Joseph E. Stiglitz, “Economic Foundations of Intellectual Property Rights” (2008) 57 DUKE L.J. 1693, at 1697.

[22] Amanda Perelli, “How much money YouTubers make, according to dozens of creators”, Business Insider online: <https://www.businessinsider.com/how-much-money-youtube-creators-influencers-earn-real-examples-2021-6> (last accessed 9 Dec 2021).

[23] Bloomberg, “Why ‘Success’ on YouTube Still Means a Life of Poverty” (27 Feb 2018), Fortune online: <https://fortune.com/2018/02/27/youtube-success-poverty-wages/> (last accessed 9 Dec 2021).

[24] F.M. Scherer, “The Innovation Lottery”, in Rochelle Dreyfuss et al., eds., Expanding The Boundaries Of Intellectual Property: Innovation Policy For The Knowledge Society 3 (Oxford: Oxford University Press, 2001).

[25] Ibid.

[26] Sharon Ong, “Risk aversion among Singaporean youth, survey reveals” (20 Sep 2019), ASEAN Economist online: <https://www.aseaneconomist.com/risk-aversion-among-singaporean-youth-survey-reveals/> (last accessed 9 Dec 2021).

[27] Sandra Davie, “askST: My daughters want to study fine arts. Should I worry?” (18 Apr 2021), The Straits Times online: <https://www.straitstimes.com/singapore/parenting-education/askst-my-daughters-want-to-study-fine-arts-should-i-worry> (last accessed 9 Dec 2021).

[28] Diane Leenheer Zimmerman, supra note 11, at 42.

[29] Diane Leenheer Zimmerman, supra note 11, at footnote 55 citing Christine Jolls, “Behavioral Law and Economics” in Peter Diamond & Hannu Vartiainen, eds, Behavioral Economics And Its Applications (Princeton: Princeton University Press, 2007) 115 and Peter R. Harris, Dale W. Griffin, & Sandra Murray, “Testing the Limits of Optimistic Bias: Event and Person Moderators in a Multilevel Framework” (2008) 95:5 J. Pers & Soc Psychol 1225.

[30] Ian Kilbey, “Copyright duration? Too long!” (2003) 25:3 E.I.P.R. 105 [Ian Kilbey].

[31] Ibid.

[32] Dream McClinton, “Global attention span is narrowing and trends don't last as long, study reveals” (17 Apr 2019), The Guardian online: <https://www.theguardian.com/society/2019/apr/16/got-a-minute-global-attention-span-is-narrowing-study-reveals> (last accessed 9 Dec 2021).

[33] Ibid.

[34] OLR Research Report, “R&D life cycles” <https://www.cga.ct.gov/2015/rpt/2015-R-0207.htm> (last accessed 9 Dec 2021): The length of time between the research & development phase and the monetisation of the intellectual property varies from industry to industry. However, what is common to all research works is that there is usually a significant length of time between the creation of the intellectual property and its commercialisation.

[35] Not all kinds of research can give rise to a valuable copyright. Copyright only protects the expression and not the ideas of the author so the nature of the copyright in a research work can be useless from a commercial standpoint. A case in point is the copyright in a telephone book. Only the selection and arrangement of the telephone numbers listed in the book would be protected and not the numbers per se. This is despite the latter likely taking the most time to compile and is the most valuable aspect of the book. As a result, a rival company can just take the numbers and use them in their book so long as the selection and arrangement are sufficiently different. See Global Yellow Pages Ltd v Promedia Directories Pte Ltd [2016] 2 SLR 165 [Global Yellow Pages Ltd].

[36] Copyright Act 2021, supra note 1, s 133(1)(a).

[37] Ministry of Manpower Singapore, “Labour Force In Singapore 2017” <https://stats.mom.gov.sg/iMAS_PdfLibrary/mrsd_2017LabourForce_survey_findings.pdf> (last accessed 9 Dec 2021) at Chart 19.

[38] Copyright Act 2021, supra note 1, s 134(3).

[39] See Nanoflim Technologies International Pte Ltd v Semivac International Pte Ltd [2018] 5 SLR 956.

[40] Diane Leenheer Zimmerman, supra note 11, at 49.

[41] Edward L. Deci & Richard M. Ryan, Intrinsic Motivation and Self-determination in Human Behavior (Springer, 1985) at 66, 149, 310.

[42] Ibid.

[43] Ibid.

[44] Diane Leenheer Zimmerman, supra note 11, at 54.

[45] Dan Ariely, Uri Gneezy, George Loewenstein & Nina Mazar, “Large Stakes and Big Mistakes” (2008) 76 Rev. Econ. Stud. 451.

[46]Rotten Tomatoes, “Star Wars: The Rise of Skywalker” <https://www.rottentomatoes.com/m/star_wars_the_rise_of_skywalker> (last accessed 9 Dec 2021).

[47] Sam Ricketson, “The Copyright Term” (1992) 23 International Review of Intellectual Property & Competition Law 753 at 771-772 [Ricketson].

[48] Cornish, “Authors in Law” (1995) 58 Modern Law Review 1 at 8-11.

[49] Ricketson, supra note 47, at 771-772.

[50] Ibid, at 773.

[51] Ian Kilbey, supra note 30.

[52] Singapore Department of Statistics, “Singapore Census of Population 2020” <https://www.singstat.gov.sg/-/media/files/publications/cop2020/sr1/findings.pdf> (last accessed 9 Dec 2021) at Chapter 5: Religions that believe in an afterlife are Christianity, Judaism and Islam.

[53] Grace Ho, “Fewer S'poreans marrying and having children: Population census” (16 Jun 2021), The Straits Times online: <https://www.straitstimes.com/singapore/politics/fewer-sporeans-marrying-and-having-children-population-census> (last accessed 9 Dec 2021).

[54] Gerald Dworkin, “Intellectual Property Rights: What Are Appropriate Terms of Protection” (1997) 18 Sing L Rev 553 at 563 [Gerald Dworkin].

[55] Pundarik Mukhopadhaya, “Changing labor-force gender composition and male-female income diversity in Singapore” (2001) 12:4 Journal of Asian Economics 547 at 552, DOI: <10.1016/S1049-0078(01)00102-6>; There is no data for Singapore prior to 1975. However, the data from that point onwards shows a massive increase in female participation in the workforce while male participation remained largely the same which equates to a demise in single income families.

[56] Ministry of National Development Singapore, “Our Early Struggles” <https://www.mnd.gov.sg/our-city-our-home/our-early-struggles> (last accessed 9 Dec 2021).

[57] Justin Ong, “Rising household incomes, more working couples in Singapore over past decade: Census” (18 Jun 2021), The Straits Times online: <https://www.straitstimes.com/singapore/rising-household-incomes-more-working-couples-in-singapore-over-past-decade-census> (last accessed 9 Dec 2021).

[58] Since independence, Singaporeans have become some of the richest people in the world. In 2020, the Gross Domestic Product per capita was $59,797, and that was notwithstanding the COVID-19 pandemic’s catastrophic effect on the local economy (See The World Bank, “GDP per capita (current US$) – Singapore” <https://data.worldbank.org/indicator/NY.GDP.PCAP.CD?locations=SG&most_recent_value_desc=true> (last accessed 9 Dec 2021)).

[59] Goh, C. B., & Gopinathan, S., “Education in Singapore: Development since 1965” in B. Fredriksen & J. P. Tan, eds, An African Exploration of the East Asian Education (Washington, DC: The World Bank, 2008) 80-108.

[60] Singapore’s life expectancy for men and women were 81.4 years and 85.7 respectively in 2019 (See Channel NewsAsia, “Singaporeans' life expectancy among highest in the world: Public sector report” <https://www.channelnewsasia.com/singapore/public-sector-report-life-expectancy-spor-covid-19-570646> (last accessed 9 Dec 2021)). There is no data for Singapore’s average life expectancy pre-1950. However it is unlikely to be more than the UK’s which was 51 years for men and 55 years for women in 1910 (See Jane Kirby, “Life expectancy in Britain almost 30 years higher than a century ago” (1 Sep 2015), The Independent online: <https://www.independent.co.uk/life-style/health-and-families/health-news/life-expectancy-in-britain-almost-30-years-higher-than-a-century-ago-10481491.html> (last accessed 9 Dec 2021)). This would mean a 30-year increase for both men and women’s life expectancy since the Copyright Act 1911.

[61] Although there has been some public dissatisfaction of late with the concept of a meritocracy in Singapore, chiefly because it can ironically promote social ossification and elitism, the political leadership still supports the meritocratic ideal (See, Melissa Heng, “Meritocracy still key principle for recognising individuals in Singapore, says Ong Ye Kung” (27 Jul 2019), The Straits Times online: <https://www.straitstimes.com/singapore/meritocracy-still-key-principle-for-recognising-individuals-in-singapore-says-ong-ye-kung> (last accessed 9 Dec 2021)).

[62] If you are interested, this article talks about whether inheritance is justified: Haslett, D. W. “Is Inheritance Justified?” (1986) 15:2 Philosophy & Public Affairs 122.

[63] George Wei, “A Look Back at Public Policy, the Legislature, the Courts and the Development of Copyright Law in Singapore: Twenty-Five Years on” (2012) 24:Special Issue SAcLJ 867 at para 32 [George Wei].

[64] Milton Friedman, Capitalism & Freedom (Chicago: University of Chicago Press, 1962) at 161-162.

[65] Singapore exported S$11.4B worth of local intellectual property and imported S$23.2B of foreign intellectual property in 2020 (see Singapore Department of Statistics, “Overall Exports and Imports Of Services, 2016-2020” <https://www.singstat.gov.sg/modules/infographics/singapore-international-trade> (last accessed 9 Dec 2021)).

[66] Ng-Loy Wee Loon, Law of Intellectual Property of Singapore, 3^rd ed (Sweet & Maxwell, 2021) at para 8.0.4 [Ng-Loy Wee Loon].

[67] [2017] 2 SLR 786.

[68] Ibid at para 49.

[69] Copyright Act 1987, ss 28(3) and 28(6).

[70] Ng-Loy Wee Loon, supra note 66, at para 8.0.4.

[71] Ibid.

[72] Ministry of Law and Intellectual Property Office of Singapore, Singapore Copyright Review Report (17 January 2019) at paras 2.3.5 and 2.3.6 [Singapore Copyright Review Report 2019].

[73] The fixation requirement is a condition precedent for copyright to subsist in any authorial work. This is as copyright, unlike confidential information in breach of confidence, is a property right which requires that the boundaries of the property holder’s rights can be delineated by third parties (See Ng-Loy Wee Loon, supra note 66, at para 6.3.16).

[74] Copyright only protects expressions and not ideas or facts, hence any confidential information that pertains to the latter two cannot be protected at all. This is unlike in breach of confidence where what is termed as ‘confidential information’ does not turn on whether the information is an ‘expression’ or a fact or idea.

[75] Copyright only confers a limited monopoly right. There is a need to prove that there was a substantial taking of the plaintiff’s work by the defendant which is a concept not found in breach of confidence. Moreover, notwithstanding infringement being made out, the defendant may also rely on a copyright defence, such as fair use, to excuse the defendant’s breach. Whether the fair use defences can be made out depends a lot on the policy dimension. Thus, confidential information that enjoys copyright but is of interest to society would be much less protected than those that only concern the individual. This is unlike the breach of confidence protections where the defences are much more narrowly defined to only protect situations where it is against the public interest to enable the plaintiff to succeed.

[76] Ng-Loy Wee Loon, supra note 66, at paras 5.1.2 to 5.1.3.

[77] This problem arises because copyrighted works do not have a registry for the public to view in one compendium to determine which works belong to whom. A voluntary registry was proposed in the public consultations for the Copyright Act 2021 but it was rejected by the committee (See Annex A of the Singapore Copyright Review Report 2019, supra note 72, at 6-8).

[78] Singapore Copyright Review Report 2019, supra note 72, at Annex A pages 29-31.

[79] Ibid.

[80] Ibid at 29: The other proposals made to the law reform committee were: (a) Limitation of remedies to a reasonable fee in a subsequent court case or case brought by the copyright owner before a tribunal; (b) Payment of a government-determined fee to a government body, which will be held on behalf of the copyright owner and paid to the owner upon application to the body; and (c) Payment of a government-determined fee directly to the copyright owner if and when the owner approaches the user.

[81] Singapore Copyright Review Report 2019, supra note 72, at Annex A pages 29-31.

[82] Global Yellow Pages Ltd, supra note 35, at para 24.

[83] Tay Long Kee Impex Pte Ltd v Tan Beng Huwah [2000] 1 SLR(R) 786 at para 44 [Tay Long Kee Impex Pte Ltd].

[84] Ibid.

[85] The Singapore Court of Appeal in Tay Long Kee Impex Pte Ltd did not elaborate on what it would take for a work to pass the threshold requirement, only that the modified warranty in the case did not meet this threshold. However, it has been persuasively argued by Prof David Llewelyn that it should be when the work is ‘banal or commonplace’ (See David Llewelyn, Ng Hui Ming & Nicole Oh, Cases, Materials & Commentary on Singapore Intellectual Property Law (Academy Publishing, 2018) at para 04.025).

[86] Ng-Loy Wee Loon, supra note 66, at para 6.1.12.

[87] Save for section 20(1)(a)(iii) of the Copyright Act 2021, where whether a work is one of ‘artistic craftsmanship’ requires an evaluation of artistic merit by the judge (Ng-Loy Wee Loon, supra note 66, at para 6.1.30), there is no requirement to find that an authorial work has sufficient merit. The Singapore Court of Appeal has affirmed this interpretation in Asia Pacific Publishing Pte Ltd v Pioneers & Leaders (Publishers) Pte Ltd [2011] 4 SLR 381.

[88] Flamelite (S) Pte Ltd v Lam Heng Chung [2001] 3 SLR(R) 610 [Flamelite].

[89] Global Yellow Pages Ltd, supra note 35.

[90] Designers Guild Ltd v Russell Williams (Textiles) Ltd [2000] 1 WLR 2416; endorsed by the Singapore High Court in Virtual Map (Singapore) Pte Ltd v Singapore Land Authority [2008] 3 SLR(R) 86 at para 14. This usually requires that the plaintiff prove that the defendant copied a quantitatively or qualitatively significant portion of the plaintiff’s work (see Creative Technology Ltd v Aztech Systems Pte Ltd [1996] 3 SLR(R) 673).

[91] Flamelite, supra note 88, at para 28.

[92] Ibid at para 31.

[93] Ng-Loy Wee Loon, supra note 66, at paras 10.1.34-10.1.35: The line between what is a ‘general idea’, which is not protected by copyright, and a ‘detailed idea’, which is protected, is not clear. A suggested rough guide is whether the defendant ‘over-borrowed’ the skill, labour and judgement that went into the creation of the work.

[94] [2000] 1 WLR 2416.

[95] Lord Hoffman and Lord Millet found infringement on the basis that the similarities between the plaintiff’s and defendant’s works established that there was copying and those copied parts amounted to a substantial taking of the plaintiff’s work. Lord Scott used a different test where the reference should be how similar, holistically, the plaintiff’s work was to the defendant’s. If there was a substantial similarity between the works, then the infringer incorporated a substantial part of the independent skill and labour contributed by the original author in creating the copyright work and therefore there was infringement. As both works were extensively similar overall, his lordship found infringement. Although the Singapore High Court in Virtual Map (Singapore) Pte Ltd v Singapore Land Authority [2008] 3 SLR(R) 86 did not decide which test is to be preferred, it is likely that Lord Scott’s would be used going forward since the District Court in that case voiced support for his test (Singapore Land Authority v Virtual Map (Singapore) Pte Ltd [2007] SGDC 216 at para 54).

[96] Copyright Act 2021, supra note 1, Part 5.

[97] Copyright Act 2021, supra note 1.

[98] Copyright Act 2021, , supra note 1, s 191; The Statutory Factors are: (a) the purpose and character of the use, including whether the use is of a commercial nature or is for non‑profit educational purposes; (b) the nature of the work or performance; (c) the amount and substantiality of the portion used in relation to the whole work or performance; and (d) the effect of the use upon the potential market for, or value of, the work or performance.

[99] Google v Perfect 10 487 F (3d) 701 (9^th Cir 2007).

[100] See Global Yellow Pages Ltd, supra note 35, at paras 77 to 81: The Singapore Court of Appeal cited the transformative use doctrine but did not voice any opinion on whether it should be applied locally.

[101] George Wei, supra note 63 at paras 23-24.

[102] Copyright Act 2021, supra note 1, Part 7.

[103] Ng-Loy Wee Loon, supra note 66, at para 2.2.3.

[104] Factor (c) in the Copyright Act 2021, supra note 1, s 191.

[105] TRIPS Agreement, supra note 2, art 13: “Members shall confine limitations or exceptions to exclusive rights to certain special cases which do not conflict with a normal exploitation of the work and do not unreasonably prejudice the legitimate interests of the right holder.”

[106] WTO, WTO Analytical Index: TRIPS Agreement – Article 13 (Jurisprudence) at para 5.

[107] Ibid at para 6.

[108] Gerald Dworkin, supra note 54, at 565.

[109] Ibid.

Assessing the Applicability of Issue Estoppel arising from a Foreign Enforcement Judgement in International Commercial Arbitration

A PDF version of the article can be found here

---

ASSESSING THE APPLICABILITY OF ISSUE ESTOPPEL ARISING FROM A FOREIGN ENFORCEMENT JUDGEMENT IN INTERNATIONAL COMMERCIAL ARBITRATION

Desmond Chye*

ABSTRACT

The recent Singapore High Court case of BAZ v BBA provided some clarity as to when and where issue estoppel can apply in seat and enforcement court proceedings. However, the court was not entirely correct. A closer inspection of the Model Law, New York Convention and the International Arbitration Act reveals that issue estoppel can actually only arise from a foreign enforcement judgement in situations where the estoppel would be in favour of enforcement or against setting aside of the award. Notwithstanding, it is better for a future court to avoid applying issue estoppel completely for a host of cogent policy reasons. The case also erred in holding that issue estoppel can never arise under the public policy and arbitrability grounds for refusing enforcement or setting aside. It would be possible (though not recommended) for issue estoppel to apply if the foreign enforcement jurisdiction considered the instant jurisdiction’s public policy.

I.                    INTRODUCTION

Issue estoppel is a doctrine that prevents issues of fact or law that have already been raised and decided on in an earlier proceeding from being submitted again for decision in subsequent proceedings between the same parties.[1] In the international commercial arbitration context, the question is whether this doctrine can apply to bar the instant court (seat or enforcement) from re-evaluating issues raised at a prior seat or enforcement court. Although it is clear that issue estoppel can arise from a seat court judgement in enforcement court proceedings, it is less clear whether an enforcement judgement can have the same effect on another jurisdiction’s enforcement court. It is also unclear whether a foreign enforcement judgement can trigger issue estoppel in the seat court. Such ambiguity festered in Singapore until the recent Singapore High Court (“SGHC”) case of BAZ v BBA[2] which appears to have injected some much needed clarity. Although the SGHC stopped short of unequivocally holding that issue estoppel cannot arise at all from a foreign enforcement judgement in setting aside proceedings, the court strongly indicated that the non-applicability of issue estoppel would at least be the general position going forward. In addition, the SGHC helpfully opined that issue estoppel can arise in enforcement proceedings. As a separate matter, the SGHC also held that foreign enforcement judgements decided on public policy grounds cannot give rise to issue estoppel in the seat and enforcement courts.

A number of issues however plague the court’s reasoning such that BAZ v BBA can only be said to have been partially correct as to whether and when issue estoppel can arise. First, the court failed to realise that issue estoppel can operate either for or against enforcement, and that while the former is permitted under the New York Convention, the latter is not. Notwithstanding this, issue estoppel should not be applied in enforcement proceedings because of strong policy reasons. Second, the court’s holding that issue estoppel cannot arise in setting aside proceedings was only partially right. It can in fact arise, if it operates against setting aside. A future court should however not apply it because of practical and doctrinal concerns with doing so. Third, the court’s finding that issue estoppel cannot apply to the public policy grounds for setting aside and refusing enforcement of the award is only correct as a general proposition. It omitted to consider the situation where the foreign judgement had considered the public policy of the instant jurisdiction. Issue estoppel can potentially arise in such a situation as there would be identity of subject matter but it is advisable for a seat court to avoid doing so for practical reasons.

This article is structured as follows: section II will start off by providing the case details of BAZ v BBA. The subsequent sections will then proceed to analyse the various facets of the court’s reasoning for the various situations that may potentially involve issue estoppel arising from a foreign enforcement judgement. Section III will first examine the applicability of issue estoppel arising in an enforcement court from a foreign enforcement court judgement. Section IV will then consider the applicability of issue estoppel arising from a foreign enforcement court judgement in setting aside proceedings at the seat court. Finally, section V will assess whether issue estoppel can or should arise under the public policy and arbitrability grounds for refusing enforcement or setting aside.

II.                 CASE DETAILS OF BAZ V BBA

In BAZ v BBA, the case revolved around a sale and purchase agreement of shares in an Indian company (“the Company”). The family members of the Company’s founder and several companies controlled by them (“the Sellers”) entered into a share purchase and share subscription agreement (“SPSSA”) with BAZ (“the Buyer”). The Sellers were led by BBA and five of them were minors (“the Minors”). The SPSSA was governed by Indian law and had an arbitration clause providing for disputes arising out of the agreement to be resolved via arbitration held in Singapore as the seat.

A dispute subsequently broke out between the parties regarding the circumstances under which the SPSSA was entered into. The Buyer then commenced arbitration proceedings against the Sellers in Singapore. The arbitral tribunal (“the Tribunal”) ruled in favour of the Buyer and granted an award of damages payable by the Sellers (including the Minors) on a joint and severally liability basis (“the Award”).

Simultaneous proceedings for leave to enforce the Award were then commenced by the Buyer in the New Delhi and Singapore High Court. The New Delhi High Court (“DHC”) was the first to give judgement. It allowed enforcement of the Award against the Sellers, but not the Minors on the ground of public policy (“the DHC Judgement”). In the SGHC, which was the seat court, the Sellers’ contended that the Award should not be enforced because it ought to be set aside for being awarded outside the scope of the Tribunal’s jurisdiction: the Tribunal was not allowed to grant an award against the Minors. The Buyer responded by arguing that the Sellers were precluded from litigating the jurisdictional challenges to the award on the basis of issue estoppel as those same issues were already canvassed before and decided by the DHC in the DHC Judgement. The SGHC ultimately dismissed the Seller’s points on the merits and enforced the award against the Sellers sans the Minors.

The Buyer’s claim of issue estoppel raised two issues. The first is whether a foreign enforcement court judgement can give rise to issue estoppel at the seat court and the second is whether a foreign enforcement judgement ruling on arbitrability or public policy can give rise to issue estoppel. The SGHC answered in the negative for both. As an aside, the court also dealt with an issue unrelated to this case: whether the judgements in one enforcement jurisdiction can create issue estoppel in another enforcement jurisdiction. It was the court’s opinion that issue estoppel can arise in such a situation. The rest of this article shall explore whether the SGHC was correct to make these findings.

III.              ENFORCEMENT COURT & FOREIGN ENFORCEMENT COURT JUDGEMENT

The issue here is whether issue estoppel can or should arise from a foreign enforcement judgement in an enforcement proceeding. As a preliminary, it should be pointed out that there are two types of enforcement situations that the Singapore court can face: enforcing a domestic international award (i.e. tribunal’s seat was Singapore) and enforcing a foreign international award (i.e. tribunal’s seat was not in Singapore). The latter is governed by the International Arbitration Act[3] (“IAA”) while the former is governed by the UNCITRAL Model Law[4] (“Model Law”). This is an important distinction as it would impact whether issue estoppel can apply at the enforcement court. We shall now look at each type of award in turn.

A.     Foreign awards

A foreign court judgement can give rise to two different issue estoppel situations: (1) an estoppel operating in favour of enforcement, which is usually when the judgement enforced the award, and (2) an estoppel that operates against enforcement, which is usually when the judgement refused to enforce the award.[5] BAZ v BBA suggests that both kinds of foreign enforcement judgements can give rise to issue estoppel but the IAA and New York Convention[6] (“NYC”) actually only allows situation (1). Notwithstanding this, issue estoppel should not be applied at all for policy reasons.

B.      Current case law position

BAZ v BBA opined that issue estoppel can arise in situations (1) and (2) because it is an implied ground for refusing enforcement under IAA section 31(2). When enforcing international foreign awards, the Singapore court applies IAA sections 31(1) and 31(2) which are in pari materia with NYC Article V.[7] Even though there is no express section for issue estoppel in the IAA and NYC, the court found that it exists as an implied ground for refusal.[8] This is as Article V(1)(e) references a foreign judgement’s decision (the seat court’s) as a ground for denying enforcement which must necessarily imply the use of issue estoppel in enforcement proceedings.[9]

Singapore jurisprudence is silent on BAZ v BBA’s position but English cases support the applicability of issue estoppel in both situations. Chantiers De L’Atlantique S.A v Gaztransport & Technigaz SAS[10] (“Chantiers”) opined that issue estoppel can arise from a foreign court judgement[11] and Yukos Capital Sarl v OJSC Rosneft Oil Co (No 2)[12] (“Yukos”) implicitly accepted that issue estoppel can apply in both situations.[13] Diag Human SE v Czech Republic[14] (“Diag”) then went further and held that issue estoppel can arise from a foreign judgement that refused enforcement.

Despite the apparent English support, BAZ v BBA is still wrong to allow issue estoppel for situation (2) because of two reasons. First, issue estoppel operating against enforcement cannot be implied under the IAA as the grounds against enforcement are exhaustive but not vice-versa. Second, issue estoppel is unlikely to be applicable by virtue of being a procedural doctrine under NYC Article III.

Issue estoppel operating against enforcement is not an additional ground for refusal

As a preliminary, issue estoppel cannot be implied under the existing ground in IAA section 31(2)(f) as it expressly refers to only the seat court’s decision as a ground for refusing enforcement by issue estoppel.[15] The question is therefore whether section 31(1) permits issue estoppel to be implied as an additional ground for refusing or permitting enforcement.

Section 31(1)’s plain reading indicates that the Article V grounds are a maximum and not a minimum rule so only additional grounds against enforcement are precluded under the IAA. Essentially, while enforcement of an award may be refused if the Article V grounds listed in section 31(2) are established, the court is not compelled to do so and has the discretion to allow enforcement.[16]

Although there has yet to be any local cases confirming the court’s residual discretion to not refuse enforcement when a ground for refusal is met, it is likely that a future court find it so. For one, English case law has interpreted ‘may’ as enabling the court to enforce the award anyway despite a ground being met for reasons such as issue estoppel.[17] Second, a future court would prefer to retain discretion over whether to refuse enforcement.[18] Third, even though the SGCA in BloomBerry Resorts and Hotels Inc v Global Gaming Philippines LLC[19] recently interpreted ‘may not’ in Model Law section 34(3) as meaning ‘must not’,[20] the decision would likely be confined to section 34(3) and not extended to other Model Law sections. This is as the holding in that case was in relation to the deadline for commencing setting aside proceedings and not the grounds for refusing enforcement.[21]

As there are no specified limits on the court’s discretion to permit enforcement but not vice-versa, it follows that Article V allows enforcement courts to apply rules of national law that are more favourable to enforcement but not rules that are less favourable to enforcement.[22] Accordingly, issue estoppel that operates against enforcement cannot be implied as a new ground for refusing enforcement because it would make enforcement more difficult. Issue estoppel favouring enforcement would however be allowed as it makes enforcement easier.

Applying these principles to the trio of English cases, we find that only Chantiers and Yukos were correct while Diag was wrong. It is therefore submitted that a future court should not place much weight on Diag. We shall now look at how the estoppel operated in each case to show why the first two cases are more persuasive than the third.

Starting with Chantiers,[23] the award creditor first attempted to enforce an award given by an arbitral tribunal seated in London against the award debtor in France. During which, the award debtor claimed that the award was unenforceable as it was tainted by fraud, but the French court dismissed it and found the award enforceable. The award debtor then sought to set aside the award in the seat court (the English court) which the court then dismissed for lacking merit. As any issue estoppel arising from the French enforcement judgement would have prevented the award debtor from raising the same fraud allegations in subsequent proceedings to oppose enforcement, the estoppel worked in favour of enforcement and thus Flaux J’s obiter that issue estoppel could have arisen in the case was correct.[24]

Turning to Yukos,[25] the case also involved issue estoppel operating in favour of enforcement. The award creditor first attempted to enforce an award accorded by a tribunal seated in Russia against the award debtor in the Netherlands. The award debtor however claimed that the award was unenforceable for having been set aside by the seat court in Russia. Notwithstanding this, the Dutch enforcement court upheld the enforceability of the award. The award creditor then sought to enforce the award in the UK but the award debtor raised the same arguments to prevent enforcement. Here, issue estoppel would have favoured enforcement because, if it was established, it would have deprived the award debtor of possibly relying on the award’s setting aside as a ground for opposing enforcement under Article V.[26] As such, Rix J’s implicit endorsement of issue estoppel operating would be correct – but only if it was limited to the kind of estoppel found in Yukos.

In Diag,[27] the issue estoppel operated against enforcement instead. The case involved an award creditor who was previously denied enforcement in Austria seeking enforcement in the UK. The English Commercial Court denied enforcement on the basis that the judgement by the Austrian Supreme Court gave rise to issue estoppel against the award creditor. This decision is however wrong because the estoppel here worked against enforcement[28] – it prevented the award creditor from raising the same arguments favouring enforcement in the English court. This basically meant that Diag was the exact situation in which a court relies on a foreign non-seat court’s refusal of enforcement as a ground to refuse enforcement, which is technically not allowed under the Model Law. It would thus be advisable for a future Singapore court to follow Chantiers and Yukos instead of Diag.

Issue estoppel is unlikely to be a procedural doctrine under NYC Article III

Some argue that issue estoppel operating against enforcement is permitted under the NYC, not as a separate ground for refusal under Article V, but rather as a part of the procedural grounds in Article III.[29] It is said that issue estoppel is only a procedural rule because the ground for refusal remains the ones in Article V: estoppel only operates to preclude a party from denying the presence of certain facts or legal points when they claim that an Article V ground applied.[30] Accordingly, issue estoppel is argued to be a procedural doctrine that falls under Article III which allows the Contracting state to enforce awards “in accordance with the rules of procedure of the territory where the award is relied upon”.[31]

While it is admittedly possible on the plain words of Article III to treat issue estoppel as a procedural doctrine, a purposive reading of the NYC should discourage this. The fact that Article V(1)(e) only refers to the decision of a foreign seat court setting aside the award as a ground for refusing enforcement must mean that the decision of enforcement courts cannot have the same preclusive effect.[32] If it was otherwise then the NYC would have expressly indicated as so.[33] Also, the existence of Article V(1)(e), which is an application of issue estoppel from the decision of the seat court, as a substantive ground under Article V implies that issue estoppel is to be treated as a substantive ground for the purposes of Article V. Moreover, the NYC’s primary objective is to make the enforcement of foreign awards easier[34] so the grounds for refusing enforcement should be narrowly interpreted whenever there is any ambiguity.[35] Section 31(2) should thus be interpreted as excluding issue estoppel operating against enforcement. As Singapore takes a ‘pro-enforcement’ stance towards foreign awards,[36] this interpretation would also have the added benefit of keeping in line with the contemporary judicial trend.

Policy arguments against allowing issue estoppel operating in favour of enforcement

Although issue estoppel operating in favour of enforcement can potentially arise under the IAA, the enforcement court should be slow to apply it because the benefit of getting greater finality in enforcement proceedings does not outweigh the much greater loss to fairness to the award debtor.

The main argument favouring issue estoppel is that it promotes finality in arbitration proceedings. Finality is seen as highly desirable in international arbitration because it would be wasteful to allow the unsuccessful party to litigate the same points ad nauseam around the world.[37] It is thus argued that enforcement courts from various jurisdictions should mutually recognise and enforce each other’s decisions to achieve this goal.[38]

However, the objective of achieving finality should not be prioritised over the more important objective of ensuring fair enforcement judgements. While speed is good to have, what is absolutely crucial to international arbitration is the ability to earn the trust of the parties involved. This is as the whole process is ultimately optional and consensual and so if the parties do not view arbitration as fair and reliable then chances are they will not agree to use it. Unfairness in arbitration system would thus pose a greater existential threat to international arbitration than a failure to encourage award finality.

Issue estoppel, by creating an unacceptable risk that enforcement judgements would be unfair, should thus be avoided. The doctrine undermines an enforcement judgement’s fairness by precluding an unsuccessful party from raising the same points for re-evaluation, even if the prior enforcement court erroneously dismissed them on politically biased or patently erroneous grounds[39] – a risk not insignificant due to the summary nature of enforcement applications.[40] Moreover, the doctrine would perpetuate any wrongly decided foreign enforcement judgements. This problem would then be compounded by the court’s lack of residual discretion to refuse enforcement where the estoppel favours enforcement, as the estoppel prevents any Article V ground from being established.[41] This is especially problematic since issue estoppel is an equitable doctrine “subject to the overriding consideration that it must work justice and not injustice”.[42]

Issue estoppel also promotes ‘forum shopping’ which penalises the award debtor. As the doctrine makes the first enforcement court’s decision determinative on subsequent courts, the parties would be incentivised to ‘forum shop’ to take advantage of the first judgement’s preclusive effect.[43] The award creditor will thus seek the most lenient enforcement jurisdiction for first enforcement while the award debtor will seek the strictest one.[44] This massively handicaps the award debtor because it is the award creditor who chooses where enforcement proceedings are to be initiated and even if the award debtor can pre-empt the award creditor by first getting a ‘negative declaration’ that the award is unenforceable from an enforcement court of his choice, very few countries allow negative declarations.[45]

Finally, issue estoppel undermines the seat court’s role under NYC Article V(1)(e). To illustrate, if an enforcement court in Country A refuses enforcement, which generates an issue estoppel in other enforcement jurisdictions, then the decision in Country A would have the same practical effect as a setting aside judgement at the seat court.[46] This would amount to an inappropriate intrusion into the seat court’s exclusive power to set aside awards.[47] Accordingly, in the absence of a seat court judgement, each jurisdiction should independently consider whether the award should be enforced under the NYC without giving preclusive effect to enforcement decisions of other jurisdictions.[48]

C.     Domestic awards

The SGHC in BAZ v BBA expressed obiter support for issue estoppel to apply to domestic awards regardless of whether (1) it is in favour of or (2) against enforcement. As IAA section 19 is in alignment with the Model Law grounds for refusing enforcement, which is in turn in pari materia with NYC Article V,[49] situation (1) is possible for the reasons expressed in Part A of this section. Although situation (2) is theoretically possible under section 19’s wide wording, a future court should not take advantage of this pliability to permit issue estoppel operating against enforcement as it would go against the prevailing judicial position that the court’s wide discretion must be exercised in a pro-enforcement way.

Preliminary issue: Can the court refuse enforcement of a domestic award in Singapore?

While it is clear that the courts can refuse enforcement of foreign awards, it was previously unclear whether the same applied to domestic awards. The ambiguity arose from this: while foreign awards can be refused enforcement locally by virtue of IAA sections 31(2) and 31(4), the same arguably did not apply to domestic awards as Model Law Articles 35 and, in particular, 36 lack the force of law in Singapore by virtue of IAA section 3(1). It was thus argued that a domestic award cannot be refused enforcement in Singapore.^[50]

It is now clear that domestic awards can be refused enforcement as the ambiguity was resolved by the Singapore Court of Appeal in PT First Media TBK v Astro Nusantara International BV^[51] (“PT First Media”) which held that domestic awards can be refused enforcement locally via section 19 of the IAA. Basically, refusal of enforcement is possible because section 19 gave the court the discretion to enforce a domestic award which section 3(1) did not override.^[52] This is as the legislative intent behind section 3(1) was just to avoid conflict with the New York Convention’s enforcement of foreign awards and not to remove the court’s right to enforce domestic awards.^[53] Section 3(1)’s exclusion of Articles 35 and 36 thus only prevents the court’s discretion to enforce domestic awards from being limited to the Model Law grounds and does not remove their ability to refuse them.^[54]

 

Can issue estoppel be applied under IAA section 19?

When applying section 19, case law stipulates that the court should “align the exercise of that discretion with the grounds under Art 36”.[55] Whether a court can use issue estoppel to deny enforcement of a domestic award would thus depend on whether issue estoppel is a ground for refusing enforcement under Article 36. This would in turn depend on whether the Model Law grounds are exhaustive as there is no article which provides for the application of issue estoppel.

BAZ v BBA implicitly found the Model Law grounds as non-exhaustive. The SGHC opined that Article 5’s plain words gives room for residual domestic law to apply to enforcement proceedings: the Model Law only blocks the court from intervening in matters governed by the Model Law. This, according to the court, meant that “what was not governed by [the Model Law] must be governed by the other rules of domestic law”.[56] Issue estoppel, a domestic law doctrine, can thus apply notwithstanding its absence as a ground for refusing enforcement under Article 36.[57]

The Article 36 grounds are however exhaustive as per its plain words. The court apparently ignored the phrase “may be refused only [if the Article 36 grounds are met]”[58] – it means that the Model Law omits all grounds, save those listed in Articles 36(1) and 36(2), as the basis for the court to exercise their discretion to deny enforcement of a domestic award. Article 36 therefore encompasses all possible issues arising from the proceedings which would exclude any room for domestic law, such as issue estoppel, to come in as an additional ground for refusal via Article 5.

Accordingly, case law must depart from their alignment with the Model Law for issue estoppel to apply as a ground for refusal. This is possible because the basis for enforcing domestic awards is IAA section 19 and not Model Law Articles 35 and 36 by virtue of IAA section 3(1).[59] As section 19 does not stipulate how the power should be exercised, it has been interpreted to confer wide discretion on the court to decide the applicable rules for enforcement so long as it adheres to the Model Law’s overarching philosophy for enforcement.[60] Therefore, if a future court interprets that ‘overarching philosophy’ as excluding the exhaustive nature of the Model Law grounds then issue estoppel can be introduced as an additional ground for refusing enforcement. This should however not be done as it would go against the widely accepted view that the Model Law is pro-enforcement, which is to be embodied by having specially identified exceptions to enforcement.[61]

IV.              SEAT COURT & FOREIGN ENFORCEMENT COURT JUDGEMENT

The issue here is whether issue estoppel can or should arise from a judgment of another foreign enforcement court in setting aside proceedings at the seat court. Similar to the enforcement situation, issue estoppel here can either operate (1) in favour of or (2) against setting aside the award. It is likely that issue estoppel can only arise in situation (2) but not (1). Notwithstanding this, a future court should still avoid applying issue estoppel in (2).

A.     Current case law position

The SGHC in BAZ v BBA held that courts should be slow to recognise issue estoppel arising from a prior enforcement judgement as a seat court’s determination should be given primacy over an enforcement court’s determination.[62] The SGCA has yet to decide on this but English law has voiced obiter support for it arising in situations (1) and (2).[63] Both the Singapore and English positions are however flawed for the reasons below.

B.      Issue estoppel operating in favour of setting aside cannot apply under the Model Law

The grounds for a seat court to set aside an award are enshrined in IAA section 24 read with Model Law Article 34(2). Similar to NYC Article V, the grounds in Article 34(2) are exhaustive, but only for the grounds to set aside and not the grounds to refuse setting aside. Regarding the former, Singapore case law clearly states that the IAA provides the only grounds on which an award can be set aside[64] and thus courts have rejected expanding the grounds to include doctrines such as ‘Wednesbury unreasonableness’.[65] Although there is no case authority for the latter, it should be permitted. This is as the grounds in Article 34(2) are aligned with those in NYC Article V, minus Article V(1)(e) which cannot apply in a setting aside proceeding,[66] such that the characterisation of the NYC grounds as a maximum and not a minimum rule[67] can also apply to Article 34(2). Article 34(2) thus prohibits additional grounds for setting aside but not more grounds for refusing setting aside. As such, only the application of issue estoppel operating against setting aside is permitted.

C.     Issue estoppel operating against setting aside should not be applied under the Model Law

Although issue estoppel operating against setting aside can arise under the IAA, a court should not apply it because there are strong doctrinal and practical problems with doing so.

From a doctrinal standpoint, BAZ v BBA was correct to find that the seat court has primacy over the enforcement courts and should thus not defer to prior enforcement decisions.[68] There are three theories as to the role of the seat court which affects its primacy vis-à-vis the enforcement courts: the monolocal, multilocal and transnational theories.[69] Singapore likely follows the monolocal approach which means the seat court enjoys primacy.

Under the transnational theory, an arbitral award’s validity “derives from a distinct arbitral legal order” and not from any national or seat legal order.[70] The seat court therefore has no primacy over the enforcement court. This theory is however unpersuasive because international arbitration cannot be a “truly distinct legal order” when national legal orders retain the ability to review such awards by enforcement or setting aside.[71]

Under the multilocal theory, “the validity of the arbitral award derives not only from the seat, but from all legal orders in which recognition and enforcement of the award are sought” hence a seat court does not have primacy over an enforcement court.[72] Some argue that this approach should be taken to protect parties against the ‘local peculiarities’ of the seat court.[73] This is however unpersuasive because it undermines the more important arbitration principle of respecting the parties’ intention to settle their dispute in a neutral jurisdiction of their choice. The choice of seat is an express intention by the parties to submit their dispute to the seat’s court and law[74] that is often made through careful consideration of the jurisdiction’s neutrality and quality of its arbitration law and judges.[75] As the parties have made an informed choice as to the seat, the law need not protect them from unexpected bias by the seat court because they have made their bed and must lie on it. The alternative argument that the enforcement jurisdiction is more relevant to the dispute and hence should be deferred to by the seat[76] is also unpersuasive. To do so would contravene the parties’ desire for neutrality because an enforcement court is chosen for its close connection to the dispute and may thus unfairly advantage one party over the other.

Under the monolocal theory, the seat court has primacy over determining an award’s validity because “an arbitral award derives its legal validity exclusively from the legal order of the seat”.[77] This approach should be followed for the reasons above and because Singapore case law already supports this proposition implicitly: it has been held that there would no longer be any award to enforce if the award has been set aside at the seat because the seat’s act of setting aside renders the award void ab initio.[78]

From a practical standpoint, the promotion of finality in litigation[79] is not a compelling reason for applying issue estoppel to setting aside proceedings for three reasons. First, applying issue estoppel operating against setting aside denies the court any discretion to refuse setting aside even if the prior enforcement judgement was wrong – a risk not insignificant due to the summary nature of enforcement proceedings.[80] This is as issue estoppel will require the seat court to accept the enforcement’s court finding that no ground under Article V (and thus also Article 34(2)) is established. Second, the pro-enforcement stance of enforcement courts also necessitates the seat court to make their own careful judgement instead of simply following other jurisdictions. Apart from the French courts,[81] major jurisdictions defer strongly to the seat court’s decision.[82] Following a wrongly decided enforcement decision at the seat would thus spread the error down the line to the various enforcement jurisdictions. Third, a seat court would likely have great difficulty applying issue estoppel anyway. This is as it may be challenging to determine what exactly drove the foreign court to decide the way it did, especially if the court used different legal rules, procedures and techniques.[83] BAZ v BBA was thus correct not to apply issue estoppel at all.

V.                 PUBLIC POLICY & ARBITRABILITY

The SGHC’s finding in BAZ v BBA that issue estoppel cannot arise under the public policy and arbitrability grounds for setting aside and refusing enforcement of an award is only accurate as a general proposition. Issue estoppel can arguably arise when the foreign enforcement judgement actually considered the instant jurisdiction’s public policy. It should however not be done as a foreign court would not be well placed to determine the instant jurisdiction’s public policy concerns.

In BAZ v BBA, the SGHC held that issue estoppel cannot arise from a prior enforcement judgement under the public policy and arbitrability grounds in the Model Law (and presumably also the IAA and NYC grounds) because “the public policy and the test for arbitrability are unique to each state” such that “there would be no identity of subject matter even if the ground of arbitrability or public policy has been raised before a different court”.[84] The DHC Judgement therefore created no issue estoppel at the Singapore seat court because although the issue at hand in both cases pertained to the law for minors, the public policy that underpinned the law for each jurisdiction differed.

The SGHC’s finding of no issue estoppel is correct, as a general principle, because the public policy considered cannot be purely ‘international’, which means that identity of subject matter is normally absent. The relevant sections in the Model Law, IAA and NYC[85] expressly reference the application of the setting aside or enforcement jurisdiction’s public policy instead of international public policy so the consideration of ‘purely’ international public policy must be prohibited.[86] A purposive reading supports this interpretation: the public policy ground is supposed to be an escape mechanism to allow national courts to deny effect to awards when it conflicts with the instant state’s public policy and not to enforce international principles.[87] It is therefore only national public policies which mandatorily apply to international matters under national law that can fall within the public policy ground.[88] Since the basis of such public policy is national law, the rationale for which necessarily varying from country to country, it follows that issue estoppel cannot arise from a foreign judgement considering the foreign enforcement jurisdiction’s public policy in a setting aside proceeding which considers the seat’s public policies.

However, if the foreign enforcement judgement considered the instant jurisdiction’s public policy then issue estoppel can potentially arise because the public policies considered in both proceedings would be similar. To use Victrix S.S. Co. v Salen Dry Cargo AB[89] as an illustration: the US enforcement court in that case was obliged, under national law, to consider the effect of enforcing the award on Swedish public policy so arguments relating to Swedish public policy were raised and dealt with in the judgement. If enforcement was subsequently sought in Sweden and the issues pleaded were those pertaining to Swedish public policy, then issue estoppel can arguably arise from the US case as the same Swedish public policy issues would be re-litigated. A setting aside and enforcement court should however not apply issue estoppel in such a situation because a foreign jurisdiction would not be well placed to correctly apply the public policy considerations of the instant jurisdiction. This would be especially so when the jurisdictions use different legal systems or follow divergent legal traditions.

VI.              CONCLUSION

Although Singapore’s arbitration law has the potential to herald an era where issue estoppel plays a more potent effect in enforcement and setting aside proceedings, we should not seize this potential. Understandably, the temptation is great. Issue estoppel would promote greater certainty and finality into enforcement and setting aside proceedings. However, we must resist the temptation as these upsides do not compensate for its massive downsides. An expansive issue estoppel regime would not only go against internationally accepted arbitration principles in law and in spirit, but also unfairly disadvantage the award debtor and aggravate potential errors in the enforcement and setting aside process. It is thus unfortunate that the SGHC in BAZ v BBA succumbed to the temptation, albeit partially, by allowing issue estoppel in enforcement proceedings. As such, it is imperative for a future court to eschew the doctrine of issue estoppel completely by repudiating BAZ v BBA. While this article recognises that doing so may lead to increased transaction costs for the parties involved and further delay the enforcement or setting aside of the award, it submits that these costs would be a necessary evil for there to be fairness. There is after all no point in having a cheap and speedy arbitration if it would create an unjust result.

---

* LLB (Candidate), National University of Singapore, Class of 2023. All errors and views expressed in this article remain my own.

[1] Good Challenger Navegante SA v Metalexportimport SA (The “Good Challenger”), [2004] 1 Lloyd’s Rep 67 [The “Good Challenger”].

[2] [2020] 2 SLR 453 [BAZ v BBA].

[3] (Cap 143A, 2002 Rev Ed) [IAA].

[4] Schedule I of the IAA, supra note 3.

[5] Renato Nazzini, “Enforcement of International Arbitral Awards: Res Judicata, Issues Estoppel, and Abuse of Process in a Transnational Context” (2018) 66:3 Am J Comp L 603 at 630 [Nazzini].

[6] Schedule II of the IAA, supra note 3.

[7] David Joseph QC & David Foxton QC, Singapore International Arbitration: Law and Practice, 2nd ed (Singapore: LexisNexis, 2018) at para 6.10 [Joseph & Foxton].

[8] BAZ v BBA, supra note 2 at paras 35 to 36.

[9] Ibid.

[10] [2014] EWHC 1639 (Comm) [Chantiers]

[11] Ibid at paras 313–315.

[12] [2012] EWCA Civ 855 [Yukos].

[13] As observed by Nazzini, supra note 5 at 26 regarding the passage by Rix LJ in Yukos, supra note 12 at paras 147-149.

[14] [2014] EWHC 1639 (Comm) [Diag].

[15] Nazzini, supra note 5 at 632.

[16] Nazzini, supra note 5 at 608.

[17] Yukos Oil Co v Dardana Ltd, [2002] EWCA Civ 543 at para 8; Dallah Real Estate and Tourism Holding Co v Ministry of Religious Affairs of the Government of Pakistan, [2010] UKSC 46 at para 67 [Dallah].

[18] Joseph & Foxton, supra note 7 at para 6.13.

[19] [2021] SGCA 9.

[20] Ibid.

[21] Ibid.

[22] Nazzini, supra note 5 at 608.

[23] Chantiers, supra note 10.

[24] Nazzini, supra note 5 at 631.

[25] Yukos, supra note 12.

[26] Nazzini, supra note 5 at 631.

[27] Diag, supra note 14.

[28] Nazzini, supra note 5 at 632.

[29] Tom Childs, “Enforcement of International Arbitral Awards Should A Party Be Allowed Multiple Bites at the Apple?” (2015) 26:2 The Am. Rev. of Int’l Arb 269 at 275 [Childs].

[30] Nazzini, supra note 5 at 632.

[31] Childs, supra note 29 at 275.

[32] Nazzini, supra note 5 at 632.

[33] Ibid.

[34] Gary Born, International Commercial Arbitration, 3rd ed (The Hague: Kluwer, 2020) at 3721 [Born]

[35] Nazzini, supra note 5 at 632

[36] Aloe Vera of Am. Inc. v Asianic Food (S) Pte Ltd, [2006] SGHC 78 at para 40

[37] The Honourable Chief Justice Sundaresh Menon, “Patron’s Address” (2015) 81:4 Arbitration: The Int’l J of Arb, Med and Dispute Management 413 at 424.

[38] Lord Jonathan Mance, “Arbitration – a Law unto Itself?” (4 November 2015) 30th Annual Lecture organized by The School of International Arbitration and Freshfields Bruckhaus Deringer 1 at 14.

[39] Maxi Sherer, “Effects of Foreign Judgements Relating to International Arbitral Awards: Is the ‘Judgement Route’ the Wrong Road?” (2013) 4:3 J. Int’l Dispute Resolution 587 at 622 to 623 [Sherer].

[40] Born, supra note 34 at 4164.

[41] Nazzini, supra note 5 at 632.

[42] The “Good Challenger”, supra note 1 at para 54.

[43] Sherer, supra note 39 at 622 to 623.

[44] Ibid.

[45] Childs (2015), supra note 29 at 274 & Herbert Kronke et al., Recognition and Enforcement of Foreign Arbitral Awards: A Global Commentary (The Hague: Kluwer, 2010) at 133-35.

[46] Matthew Barry, “The Role of the Seat in International Arbitration: Theory, Practice, and Implications for Australian Courts” (2015) 32 J. Int'l Arb. 289 [Barry], citing Chief Justice Allsop at 319.

[47] Sebastian Perry & Richard Wolley, “Issue estoppel halts enforcement in London” (29 May 2014) Global Arbitration Review.

[48] Born, supra note 34 at 4163.

[49] Ibid at 3762.

[50] PT First Media TBK v Astro Nusantara International BV, [2014] 1 SLR 372 at para 18 [PT First Media].

[51] Ibid.

[52] Ibid at para 85.

[53] Ibid.

[54] Ibid.

[55] Ibid at paras 86 to 90.

[56] BAZ v BBA, supra note 2 at para 37.

[57] Ibid.

[58] Section 36(1) of the IAA, supra note 3.

[59] PT First Media, supra note 50 at paras 86 to 90.

[60] Ibid at para 50.

[61] Born, supra note 34 at 3164 & 3718.

[62] BAZ v BBA, supra note 2 at para 51.

[63] Dallah, supra note 17 at para 39.

[64] PT Asuransi Jasa Indonesia (Persero) v Dexia Bank SA, [2007] 1 SLR(R) 597 at paras 54-55 and 57 and Soh Beng Tee & Co Pte Ltd v Fairmount Development Pte Ltd, [2007] 3 SLR(R) 86 at paras 60-66.

[65] Sui Southern Gas Co Ltd v Habibullah Coastal Power Co (Pte) Ltd, [2010] 3 SLR 1 at para 19.

[66] Born (2020), supra note 34 at 3438-3439.

[67] Ibid at 3435.

[68] BAZ v BBA, supra note 2 at para 51.

[69] Barry, supra note 46 at 294.

[70] Ibid at 299-301.

[71] Ibid at 300-301.

[72] Ibid at 297.

[73] Jan Paulsson, “Delocalisation of International Commercial Arbitration: When and Why it Matters” (1983) 31:1 Int’l & Comparative L. Q. 53 at 54.

[74] Jean-François Poudret & Sébastien Besson, Comparative Law of International Arbitration, 2^nd ed (London: Sweet & Maxwell, 2007) at para 146 [Poudret & Besson].

[75] Francis Mann, “The UNCITRAL Model Law: Lex Facit Arbitrum” in Martin Domke & Pieter Sanders, eds, International Arbitration: Liber Amicorum for Martin Domke (Leiden: Martinus Nijhoff, 1967) & Noah Rubins, “The Arbitral Seat is No Fiction: A Brief Reply to Tatsuya Nakamura’s Commentary” (2001) 16 Mealey’s Int’l Arb. Report 12.

[76] Emmanuel Gaillard, Legal Theory of International Arbitration (Leiden: Matrinus Nijhoff, 2010) at 32.

[77] Barry, supra note 46 at 294 to 295.

[78] PT First Media, supra note 50 at paras 76 to 77.

[79] BAZ v BBA, supra note 2 at para 47.

[80] Born, supra note 34 at 4164.

[81] CA Paris, 12 February 1993, Unichips Finanziaraia v Gesnouin.

[82] Barry, supra note 46 at 312.

[83] Ibid at 318.

[84] BAZ v BBA, supra note 2 at para 50.

[85] Section 31(4)(b) of the IAA, supra note 3; Article 34(2)(b) of the Model Law; Article V(2) of the NYC.

[86] Born, supra note 34 at 4013 to 4014.

[87] Ibid at 4014.

[88] Ibid.

[89] 825 F.2d 709, 714-15 (2d Cir. 1987).

Falsification in Trust Law: a Critique of the 'But For' Test

A PDF version of the article can be found here.

---

FALSIFICATION IN TRUST LAW: A CRITIQUE OF THE ‘BUT FOR’ TEST

Desmond Chye*

I.                    INTRODUCTION

Under trust law, if a trustee misapplies trust property such as by distributing trust funds to unauthorised persons or by making unauthorised investments, the doctrine of falsification steps in to allow the beneficiaries to falsify the trust account to treat the disbursement as one using the trustee’s own money instead.^[1] The trustee would thus be required to replace the misapplied asset in specie or in money which is, under the orthodox approach, all loss caused to the trust fund even if it was not ‘but for’ the trustee’s breach.^[2] However, the English cases of Target Holdings Ltd v Redferns^[3] (“Target Holdings”) and AIB Group (UK) plc v Redler^[4] (“AIB Group”) renounced the orthodox approach and laid down a new method whereby the ‘value’ of the trustee’s liability is only what was caused ‘but for’ the breach. While there has not been a definitive pronouncement by the Singapore Court of Appeal (“SGCA”) on whether to follow the English direction, it is submitted that it should not be adopted because Target Holdings and AIB Group are of weak authoritative weight given that it was unnecessary for the court to have considered causation in those cases. Moreover, even if the authoritative weaknesses of the English cases can be somehow overlooked, the ‘but for’ test remains fundamentally incongruent with the established nature of the trustee’s duties vis-a-vis the trust fund such that adopting it would engender undesirable conceptual confusion into what is otherwise a straightforward area of law.

II.                 TARGET HOLDINGS WAS DECIDED PER INCURIAM

As a preliminary, it should be noted that the English cases revolved around a very common form of trust used to purchase land: the Quistclose trust. This form of trust is, in essence, a bare trust for the lender with a contractual mandate where the trustee has to apply the trust assets only for the specific purposes stipulated for under the contract.^[5] The trustee would thus commit a breach of trust if he distributed the assets in a manner not provided for in the contract.^[6] In addition, the trustee is also usually an agent of the beneficiary.[7]

In Target Holdings, the client of Target, Crowngate, wanted to purchase a plot of land by way of a mortgage in favour of Target. In order to facilitate the transaction, Target, the mortgage lender, transferred the loan money to their solicitor, Redferns, under a Quistclose trust arrangement where the solicitor-trustee Redferns was to release the money to Crowngate, their associated companies and the vendor only once the requisite documentation had been provided by them. Redferns however committed a breach of trust by releasing the trust money before the documents arrived – paid too early. Two years later, Crowngate defaulted on the mortgage payments and became insolvent. Unfortunately, the property turned out severely over-valued due a decline in the property market. Target then applied for a summary judgement against Redferns for falsification to claim the difference between the mortgage redemption price and the original valuation. The court held that Redferns was not liable for the diminished value of the property as it was not caused ‘but for’ their breach but rather due to the decline in the property market.^[8]

It is however submitted that Target Holdings was wrongly decided as it is underpinned by the dubious premise that there was still an active breach of trust by Redferns for ‘paying early’ to Crowngate at the time of the lawsuit. Lord Browne-Wilkinson, in his hypothetical example given at the start of his analysis, explained that there is a breach of trust should monies in the client account be transferred by “the solicitors to the borrower one day before the mortgage is executed”.^[9] This is however false as such breaches may be ‘self-corrected’ if the conditions for proper payment are subsequently fulfilled.^[10] A ‘self-correction’ arguably occurred in Target Holdings as there was no loss to the trust fund by the time of the trial because the misapplication of funds by ‘paying early’ was properly restored to the correct terms when the required documentation subsequently arrived.^[11] This was possible because “the trustee’s obligation to restore the trust property is not an obligation to restore it in the very form in which he disbursed it, but an obligation to restore it in any form authorised by the trust”.^[12] Lord Browne-Wilkinson thus erred in finding a breach of trust by Redferns – the breach had already been corrected earlier and so ceased to exist by the time of the trial. This leaves his lordship’s ‘but for’ test necessarily obiter which greatly undermines the case’s authoritative weight.

Although there have been objections to the ‘self-correction’ theory, they can be disregarded for the purposes of Target Holdings. The main contention against the ‘self-correction’ theory is that a trustee cannot ‘cure’ his own breach on his own accord as the ‘cure’ must first be adopted by the beneficiary.^[13] This argument is however unconvincing because, as pointed out by Conaglen, a trustee’s trusteeship does not terminate simply because he committed a breach of trust.^[14] Due to the special nature of the trustee’s role in a Quistclose trust where he is also an agent of the beneficiary, the trustee’s authority to act post breach can be analogised with that of an agent’s authority to act under similar circumstances.^[15] Under agency principles, an agent who has breached his mandate can still remain authorised to ‘cure’ his own breach as the agent retains residual authority to bind the principal until the principal takes steps to terminate the agency.^[16] The principal’s act of termination can be implied from context or automatic upon a bad faith breach by the agent.^[17] An inference of residual authority would be strong “[w]here the agent has no reason to believe that the principal would not want the transaction completed”.^[18] Applying the agency principles, Redferns was in fact authorised to ‘cure’ their breach of trust as there was no implied termination of their agency because Target wanted the mortgage to be executed by Redferns once the documents were subsequently furnished. Even if the element of fraud identified by Lord Browne-Wilkinson could have amounted to a bad faith breach of Target’s instructions which extinguished Redfern’s agency authority to ‘self correct’, it could not have changed the case’s decision as the court was procedurally restricted from considering any claims of dishonest fraud in summary judgement.^[19]

Assuming we accept the court’s reasoning that there was a breach of trust, there is still another fatal weakness in the case’s logic supporting the ‘but for’ test: Lord Browne-Wilkinson erroneously concluded that there was case law support for the test. This is as the learned judge relied upon surcharge cases^[20] to find that there must “be some causal connection between the breach of trust and the loss to the trust estate for which compensation is recoverable”, namely, “the fact that the loss would not have occurred but for the breach”.^[21] His Lordship’s conflation of falsification and surcharge cases is problematic because each doctrine serves a fundamentally different purpose. A claim for a surcharge of the trust fund is for situations where the trustee has failed to do something that did not involve a misapplication of trust property. This is often when the trustee has failed to meet his duties of care such that the fund is now worth less than what it would have been if the trustee had complied with his duties.^[22] Implicit in the nature of a surcharge claim is the principle of ‘but for’ causation as the court must determine what the trust fund’s hypothetical value ought to have been without the breach. Falsification claims, on the other hand, are for situations where there was a straightforward misapplication of trust assets which does not need consideration of the future value of the fund.^[23] This is as the nature of the trustee’s duty here is akin to a strict liability to maintain the value of the trust fund as it is.^[24] As such, the learned judge’s finding that there was a ‘but for’ test requirement from case law was likely made per incuriam because the conceptual incongruity between the two types of cases rendered the surcharge cases irrelevant to Target Holdings, a case on falsification.

III.              AIB GROUP IS OF WEAK AUTHORITATIVE WEIGHT

Similar to Target Holdings, AIB Group involved a mortgage transaction that went wrong. Under the loan arrangement, AIB Group loaned £3.3m to a borrower which was secured by a mortgage over the borrower’s property worth around £4.25m. The property was however subject to a prior mortgage from Barclays Bank worth around £1.5m. As such, AIB made the mortgage conditional upon Barclays’ mortgage being redeemed on or before the loan advance by AIB so that AIB would be granted the first legal charge over the land. This was to be done using a Quistclose trust where AIB’s solicitor-trustee, Redler, was instructed by the lender-beneficiary, AIB Group, to redeem the mortgage loan prior to or on the date of securing a first charge in favour of AIB Group so as to use the funds to discharge the Barclays mortgage. Unfortunately, the Barclays mortgage was only partially discharged as Redler’s mistakenly remitted around £300,000 short to Barclays. Things remained this way for about two years until Redler confessed to the error. This triggered AIB to negotiate with Barclays to acquire a second, non-first charge, mortgage over the land to secure the £3.3m loan. The borrower then defaulted on the mortgage payments and the property was sold for around £1.2m which was way below the original valuation due to a property market decline. Out of the sale proceeds, AIB only received £867,697 as £300,000 had to go to the partially discharged Barclays mortgage. AIB then sought a falsification claim against Redler for the difference between the sale amount and the original valuation of £4.25m. The UK Supreme Court rejected AIB’s claim and held that Redler was only liable for the £300,000 paid to clear Barclays’ prior mortgage as even if there had been no breach of trust by Redler in paying the wrong amount, the property would still have sold for only £1.2m due to the subsequent decline in the property market.[25]

It is however submitted that AIB Group should also have been decided on orthodox trust principles as opposed to using the ‘but for’ test. This is as there was arguably acquiescence of Redler’s breach of trust when AIB, with full knowledge of the error, chose to obtain a second mortgage with Barclays to resolve the issue instead of pursuing an action against Redler.^[26] AIB’s acquiescence critically meant that, similar to Target Holdings, there was no longer an active breach of trust by the time of the trial. As such, it was unnecessary for the court to have applied the dubious ‘but for’ test in AIB Group. This leaves the ‘but for’ test once more in obiter dicta purgatory and thus devoid of authoritative weight.

IV.              CONCEPTUAL ARGUMENTS AGAINST THE ‘BUT FOR’ TEST

A.                 Fundamental Incompatibility

Assuming that the technical weaknesses in Target Holdings and AIB Group identified earlier can be somehow surmounted, it still remains undesirable to adopt the ‘but for’ test because it is conceptually incompatible with the principles underlying the doctrine of falsification.

At the heart of the issue is the question of what purpose should the remedy for falsification, ie. ‘equitable compensation’, serve? Here, there are two competing options. The first is for falsification to serve as a ‘substitutive’ remedy. This is one where the obligation to provide compensation for the breach of trust is of strict liability. A trustee cannot escape liability even if the loss to the trust was still going to materialise notwithstanding his breach of trust. In this sense, the trustee’s duty to maintain the trust fund is analogous to the obligation to repay a debt owed, with the ‘debt’ here being the trust fund’s value.

The second option is for falsification to act as a ‘reparative’ remedy where the purpose is to put the beneficiary in the same position he would have been ‘but for’ the breach of trust. In short, falsification’s objective is to achieve restitution. This was the position adopted by Lord Browne-Wilkinson in Target Holdings.^[27] Under this doctrine, the trustee is not obliged to compensate for a breach of trust if the loss would have arisen regardless of the trustee’s breach because, in such situations, the ultimate position of the beneficiaries was unaffected by the breach.^[28] This departure from the orthodox position, where the remedy is ‘substitutive’, would however lead to a number of undesirable implications that would weaken the raison d'etre of falsification.

At the most fundamental level, the ‘but for’ approach misunderstands the nature of falsification – falsification stems from a primary obligation rather than a secondary obligation of the trustee. Under the ‘but for’ approach, the trustee is under a primary obligation to execute the trust terms and a secondary obligation to pay equitable compensation if he does not.^[29] The reality is however that the trustee is under a single primary obligation to account for his stewardship of the beneficiary’s property^[30] because he has to “hold and deal with the trust property in accordance with the trust terms, and to produce it when called upon to do so”.^[31] Liability to pay equitable compensation in a falsification claim is thus a primary obligation as it is the means by which the trustee is forced to restore the trust fund to what the trustee should have kept it at.^[32]

Moreover, involving principles of ‘but for’ causation in falsification also goes against its nature as a remedy for any discrepancies found during an ‘account’ of the trust fund. In trust law, an account is “the first step in a process which enables [the beneficiary] to identify and quantify any deficit in the trust fund and seek the appropriate means by which it may be made good”.[33] Once discrepancies are identified, the beneficiaries can opt to falsify the account by treating those disbursements of trust property as never having been disbursed and thus obliging the trustee to restore the trust fund to this new falsified account. It is implicit in this process that falsification is only a substitutive, and not reparative, remedy. As an ‘account’ of anything is naturally only referable to the point in time at which it was taken, the trustee’s liability to restore the account must necessarily be limited to the discrepancy at the time it was taken and not after or before.^[34] Therefore, introducing any concept of causation would be to go against this nature of an ‘account’ by considering events beyond it. It is thus submitted that the correct purpose of falsification is “to ‘restore’ the claimant to the position he was in before the defendant committed the wrong” instead of “put[ting] the parties into the position they would have been in had no wrong occurred”^[35] – the orthodox interpretation where the “award of “equitable compensation” for misapplication should be based on replacement and not compensation”.^[36]

Although the SGCA in Sim Poh Ping v Winsta Holdings[37] (“Winsta Holdings”) stated that orthodox falsification does have an element of causation, the ‘causation’ they were referring to is more accurately described as a ‘conceptual link’ rather than a ‘causal link’. A link is ‘causal’ when the act results in an outcome that is extrinsic to the act itself while a link is ‘conceptual’ when the outcome is intrinsic.[38] An example is the act of cooling down a room. The act of cooling down the room is tautological with the temperature going down so there is a ‘conceptual’ link between the two. On the other hand, the act of opening the window is not tautological with the cooling down of the room as it might be hotter outside hence the linkage between the two is ‘casual’. Applying this to the SGCA’s description of the ‘causal link’ as “the subject matter of the trust would not have departed from the custody of the trustee to whom it was entrusted but for the trustee’s breach of his custodial stewardship duty”,[39] it would appear that what they were really referring to is a conceptual link because a trustee’s breach of his custodial stewardship duty is tautological with the departure of trust assets.

Aside from the fundamental conceptual problems in the ‘but for’ test, Lord Browne-Wilkinson’s approach also suffers from a tremendous flaw: it allows the wrongdoing trustee to gain a windfall from external events. This is best demonstrated by the following hypothetical example:^[40]

Tom is a trustee of some gemstones in a vault. He however commits a breach of trust by removing a gemstone from the trust collection in the vault and gives it away to an unwitting friend of his. Subsequently, out of pure coincidence, the remaining gems were stolen by a burglar.

Should Lord Browne-Wilkinson’s approach be followed then Tom would not be liable for the gemstone he stole simply because he was lucky that a burglar came along to break the causal link between his theft and the loss to the trust fund. Using luck to determine liability for wrongdoing is unprincipled and unjust to the beneficiary hence it is better to stick to the orthodox approach instead.

Although the orthodox approach can instead lead to a windfall to the beneficiary, as noted by Lord Browne-Wilkinson, it can be justified on the principle that all uncertainty should be resolved against a wrongdoer. This is as the trustee voluntarily assumed control over the beneficiary’s property to act for the beneficiary’s interests which renders the beneficiary a ‘vulnerable’ party that needs protection by stricter rules against the trustee.^[41] While it may be argued that an exception should be carved out for Quistclose trusts as the relationship is possibly more equal due to it being a bare trust where the trustee is to act to the order of the beneficiary, the power differential still tilts decisively towards the trustee. This is as the trustee can exploit the beneficiary’s detached position from the fine details of the trust’s execution to effectively conceal wrongdoing until it is too late for the beneficiary like what happened in Target Holdings.

B.                  Inherent Vagueness

Furthermore, even if we suppose that the need for fairness to the trustee trumps the need to protect the beneficiary, it would still be undesirable to adopt the ‘but for’ test because it is painfully vague as to what sort of trust it should apply to.

In Target Holdings, it was held that the orthodox approach should only apply to ‘traditional’ trusts while the ‘but for’ approach applies to ‘commercial’ trusts.^[42] This was due to the perceived difference between trusts created for commercial purposes and those that are not. In traditional trusts where there is no commercial element, the trust assets are normally held in trust for a number of beneficiaries who have different and normally successive equitable interests to the trust assets (e.g. A for life with the remainder to B). The purpose of the trustee is thus to have the whole fund vested in him so as to be available to satisfy the equitable interest of each beneficiary when it falls due.[43] The trustee’s strict obligation to restore the trust fund thus “reflects the fact that no one beneficiary is entitled to the trust property and the need to compensate all beneficiaries for the breach”.[44] However, this rationale does not apply to commercial trusts because, under the Quistclose structure regularly used in commercial transactions, there is only one beneficiary: the client.[45] Moreover, as observed by the court in AIB Group, the trustee’s duties in Quistclose trusts are much more closely defined and last for a much shorter duration than traditional trusts so the trustee is more akin to an agent than a true trustee.^[46] Due to these special characteristics of commercial trusts, there is no need to view the trustee of a commercial trust as the “guarantor of the integrity of the fund” for all beneficiaries which requires the imposition of onerous duties on.^[47]

However, how do we determine when a trust is commercial in more complex situations? This is not an easy question to answer because the ‘commerciality’ of a trust is not binary but rather lies on a spectrum. ‘Commerciality’ would misleadingly appear binary if one looks only at cases like Target Holdings and AIB Group where the trusts were for purely commercial purposes but such an illusion would be quickly destroyed once a trust consisting of commercial and non-commercial purposes is analysed. An example is whether the following would be considered a ‘commercial’ trust: a family settlement granting a power to invest on a mortgage to the trustee but the trustee negligently parted with the trust money without obtaining the executed mortgage and title deeds in exchange.^[48] In such situations, the ‘commerciality’ of the trust must be assessed on a scale to see if it tips more towards a ‘commercial’ trust than a ‘traditional’ one. Perhaps this could be done based on how many beneficiaries the trust has or how close the trustee was to an agent, like in Target Holdings and AIB Group. However, this would require a subjective assessment on the part of the judge as to where the appropriate tipping point should be which would then have the undesirable effect of greatly increasing the risk of unprincipled and arbitrary judgements. When coupled with the fact that the remedy for each is wildly different, the undesirability of the ‘but for’ test is obvious. It is thus submitted that the local courts should not introduce such uncertainty into an otherwise straightforward area of law.

V.                 A POSSIBLE MIDDLE GROUND?

Some argue that a modified ‘but for’ test would be more palatable for the court to adopt but it is submitted that the best option is still to stick to the orthodox position. The primary issue with any modified approach is in determining what sort of situation would justify the reparative remedy as opposed to the orthodox substitutive remedy. Unfortunately, the two main approaches to this problem are somewhat deficient in one way or another.

A.                 ‘Active’ versus ‘Passive’ Duty

The first approach is the one taken by the English Court of Appeal in Main v Giambrone & Law (a firm)^[49] (“Giambrone”) where the applicable remedy is determined based on whether the breach of trust by the trustee was that of an ‘active’ or ‘passive’ duty. The learned judge in Giambrone held that this principle stemmed from the facts of Target Holdings and AIB Group.^[50] This finding is however somewhat unsatisfactory as the learned judge failed to elaborate on why the solicitors were under an ‘active’ duty when the facts appear to support a contrary finding instead: the solicitors had no duty to do anything other than to hold on to the trust assets if the requisite event for applying trust funds away did not arise.^[51] Another issue is that whether a duty is ‘active’ or ‘passive’ should not factor into what remedy is awarded.^[52] Conceptually, the basis for any remedy in falsification is the duty to “act as custodians of the deposit monies indefinitely” until mandated by the trust instrument to apply the funds away^[53] which is necessarily always a passive duty. Hence, any potential 'active' duty would be additional to the always present passive duty to preserve the trust fund.^[54] This leaves the basis for excluding falsification for ‘active’ duties unprincipled.^[55]

B.                  A ‘Completed’ Trust

The second approach is to determine the appropriate remedy based on whether the trust has been ‘completed’. This would involve an analysis of the scope and purpose of the trust to see if the purpose has been achieved.^[56] If the objectively construed “purpose of the trust is fulfilled and the relevant transaction undergirding the trust is ‘completed’” then a substitutive remedy would be unavailable.^[57] This stems from the idea that the principles of trust law ceases to apply once “there is neither a trust to be reconstituted nor any duties to be performed by trustees”.^[58]

However, the ‘completion’ approach would not actually give a different result from the orthodox approach. Under the former, a trust is ‘completed’ when “the conditions of distribution of trust property are met”.^[59] Based on this definition, it can be said that there are two possible scenarios which may comprise ‘completion’: (1) when the trustee performed his duties impeccably all the way and (2) when the trustee inappropriately disbursed trust property but subsequent intervening events made it such that the conditions for the disbursement were met by the time of the trial. The completion approach appears to have only considered the first scenario as giving rise to a completed trust. But is it really so? The second can also be said to be a ‘completed’ trust as the conditions of the distribution have been met, albeit subsequently. More fundamentally, a trust is also either completed or is not at the time of the trial so it does not matter whether it was incomplete earlier. Therefore, should the trustee’s non-compliance be subsequently corrected by intervening events then the trust would become completed at the point of trial. This then means that there would be no breach of trust for the substitutive remedy to operate on by the time of the trial.^[60] The completion approach would thus only award a substitutive remedy for the scenario where the trustee’s breach was never ‘cured’ (ie. a non-Target Holdings situation) – a scenario that is already well tackled by orthodox principles. The completion approach is hence not strictly necessary when analysed from the earlier self-correction theory. As such, if we have to choose between the two approaches, the self-correction method would be the better option since it does not have the added hassle that the completion approach has of needing to ascertain the trust’s purpose.

VI.              CONCLUSION

All in all, the ‘but for’ test in English law should not be adopted in Singapore because the cases of Target Holdings and AIB Group were not only wrong on the facts but also on conceptual grounds. While it is fortunate that the SGCA in Winsta Holdings appears cognizant of the fact that falsification should not be a reparative remedy but rather a substitutive one,^[61] it remains to be seen whether they would conclusively reject Target Holdings’ reparative approach.[62] It is thus hoped that the SGCA would firmly reject this problematic test as soon as possible as otherwise the conceptual confusion in the doctrine of falsification would continue to stir up a storm of controversy in what is otherwise supposed to be a very calm area of law.

---

* LLB (Candidate), National University of Singapore, Class of 2023. All errors in this article remain my own.

^[1] P.J. Millett, “Equity's place in the law of commerce” (1998) 114 Law Quarterly Review 214 at 226. [Millett].

^[2] Ibid.

^[3] [1996] AC 421 [Target Holdings].

^[4] [2015] AC 1503 [AIB Group].

^[5] J E Penner, The Law of Trusts 11th Ed (Oxford: Oxford University Press, 2019) at paras 7.28-7.32.

^[6] Ibid.

[7] Example: The trustees in Target Holdings and AIB Group.

^[8] Target Holdings, supra note 3 at 440.

^[9] Ibid at 432.

^[10] Millett, supra note 1 at 227.

^[11] Ibid.

^[12] Ibid.

^[13] J. Edelman, “Money awards of the cost of performance” (2010) 4 Journal of Equity 122.

^[14] M. Conaglen, “Explaining Target Holdings v Redferns” (2010) 4 Journal of Equity 288.

^[15] P. Watts, “Some Aspects of the Intersection of the Law of Agency with the Law of Trusts” in P.S. Davies and J. Penner, eds, Equity, Trusts and Commerce (Oxford: Hart Publishing, 2019) at 29.

^[16] Ibid.

^[17] Ibid.

^[18] Ibid.

^[19] This view came from a seminar conducted by Professor James Ernest Penner which the author attended.

^[20] In Target Holdings, Lord Browne-Wilkinson cited cases such as Nestle v National Westminster Bank plc [1993] 1 WLR 1260 and Bartlett v Barclays Bank Trust Co. Ltd (Nos. 1 and 2) [1980] Ch 515 to support his proposition at 434.

^[21] Target Holdings, supra note 3 at 434.

^[22] Paul S. Davies, “Compensatory Remedies for Breach of Trust” in Richard C. Nolan, Kelvin F.K. Low & Tang Hang Wu, eds, Trusts and Modern Wealth Management (Cambridge: Cambridge University Press, 2018) 307 at 317 [Davies].

^[23] Re Dawson (deceased); Union Fidelity Trustee Co Ltd v Perpetual Trustee Co Ltd [1966] NSWR 211 at 216.

^[24] Ibid.

[25] AIB Group, supra note 4, see Lord Toulson’s judgement.

^[26] Peter Watts, “Agents’ Disbursal of Funds in Breach of Instructions” (2016) Lloyd’s Maritime and Commercial Law Quarterly 118 at 121.

^[27] Target Holdings, supra note 3 at 436.

^[28] Ibid.

^[29] Millett, supra note 1 at 225.

^[30] Ibid.

^[31] Sim Poh Ping v Winsta Holdings [2020] SGCA 35 at para 113 [Sim Poh Ping v Winsta Holdings].

^[32] Millett, supra note 1 at 225.

[33] Libertarian Investments Ltd v Hall [2014] 1 HKC 368 at para 168.

^[34] Millett, supra note 1 at 225.

^[35] Davies, supra note 22 at 323.

^[36] Lucina Ho, “An Account of Accounts” (2016) 28 Singapore Academy of Law Journal 849 at para 65.

[37] Sim Poh Ping v Winsta Holdings, supra note 31 at para 114.

[38] Von Wright, Norm and Action (New York: The Humanities Press, 1963) at 40-41

[39] Sim Poh Ping v Winsta Holdings, supra note 31 at para 114.

^[40] The author thanks Professor James Ernest Penner for inspiring this hypothetical scenario.

^[41] Davies, supra note 22 at 319.

^[42] Target Holdings, supra note 3 at 434.

[43] Ibid at 435-36.

[44] Ibid.

[45] Ibid.

^[46] AIB Group, supra note 4 at para 71.

^[47] Richard Nolan, “A Targeted Degree of Liability” (1996) LMCLQ 161 at 162.

^[48] Millett, supra note 1 at 225.

^[49] [2017] EWCA Civ 1193 [Giambrone].

^[50] Ibid at para 61.

^[51] Ding Hang Seah, “Custodial Stewardship Duties of a Trustee: The Search for a Principled Approach” (2019-2020) 37 Sing L Rev 119 at 125-26 [Seah].

^[52] Ibid.

^[53] Giambrone at para 62.

^[54] Seah, supra note 51 at 125-26.

^[55] Ibid.

^[56] Youyang Pty Ltd v Minter (2003) 196 ALR 482 at paras 48-49.

^[57] Seah, supra note 51 at 126.

^[58] Ibid at 127.

^[59] Ibid.

^[60] See the ‘self-correction’ theory in Part 2 of this article.

^[61] Sim Poh Ping v Winsta Holdings, supra note 31 at para 125: “[In the] situation involving custodial breaches, the monetary award is substitutive – it seeks to restore the trust fund or the fund of the principal either in specie or by a monetary sum in lieu. The usage of the term “equitable compensation”, with its reparative origins, in cases of custodial breaches may have well led academics and courts to view the monetary awards ordered for custodial breaches similarly – and in our view, wrongly – in a reparative light.”

[62] Ibid at para 122.

Is Singapore's Constitutional Supremacy An Illusion?

A PDF version of this article is available here.

---

IS SINGAPORE’S CONSTITUTIONAL SUPREMACY AN Illusion?

Desmond Chye*

I.         Introduction

The conventional wisdom is that Singapore operates under the principle of constitutional supremacy by virtue of Article 4 of the Constitution of the Republic of Singapore.[1] Unfortunately, the reality is much murkier than what the Constitution would like one to believe for there has been a steady stream of academic debate, since at least the 1980s, over whether constitutional supremacy actually exists in form and in substance locally. In this article, I shall explore the arguments for and against constitutional supremacy and conclude that while it does exist in theory, it is unfortunately weak in practice. This is as the capabilities of and the actions by the bodies supposedly co-equal under the Constitution are not really synonymous with that of a state practising constitutional supremacy.

II.         when does constitutional supremacy arise?

The core difference between parliamentary and constitutional supremacy lies in the scope of legislative power accorded to parliament. Under the former, the legislature has the unfettered right to make or undo any law whereas under the latter, the legislature’s ability is fettered by the Constitution such that any acts of parliament inconsistent with the Constitution would be void.[2] A parliament is ‘unfettered’ if it meets Dicey’s criteria for a ‘sovereign parliament’: (1) when parliament is free to amend all laws, fundamental or otherwise in the same manner as other laws; (2) when there is no legal distinction between constitutional and other laws; and (3) when there are no other bodies which can nullify an act of parliament.[3] As such, Dicey observed that constitutional supremacy requires the following characteristics: (1) the constitution must be written; (2) it must be rigid; and (3) acts by the legislature must be capable of invalidation if they amount to overstepping their constitutionally prescribed authority.[4]

In addition to the Diceyean factors, the constitution must also be a ‘grundnorm’[5] for it to be supreme. Similar to how a creature cannot be superior to its maker, an “instrument cannot validate itself by declaring itself valid” hence a constitution can only attain supremacy if it is authorised by a higher power.[6] For constitutions which are already at the apex of the legal order, this ‘higher power’ must be external to the organs created under the constitution.[7] Examples include popular acceptance via a referendum, a gift by a departing colonial overlord to its former subjects, and even a revolutionary upheaval that imposes a new legal order illegitimately.[8] As a corollary to this principle, a constitution created by a legislative act cannot give rise to constitutional supremacy for it would mean that parliament, not the constitution, is the grundnorm.

III.         SINGAPORE’s theoretical basis for constitutional supremacy

There is a theoretical basis for constitutional supremacy in Singapore. On the Diceyean front, the Constitution clearly meets the required criteria as amending the Constitution is legally fettered. Article 4 of the Constitution declares that “any law enacted by the Legislature after the commencement of this Constitution which is inconsistent with this Constitution shall, to the extent of the inconsistency, be void”, which is enforced by the Judiciary under Article 93.[9] Article 5 of the Constitution further stipulates that bills seeking to amend the Constitution must obtain the support of “not less than two-thirds of the total number of Members of Parliament (excluding nominated Members)”.[10] On the matter of the grundnorm, some difficulties have been voiced but those views can be safely ignored for the reasons elaborated below.

Relying on the grundnorm theory, Andrew Harding opined that the Constitution cannot be supreme because it was created by an Act of Parliament, the Republic of Singapore Independence Act, 1965.[11] Harding theorised that when Singapore became independent, the Malaysian Parliament failed to transfer its plenary powers to the Singapore Parliament via the Constitution and Malaysia (Singapore Amendment) Act, 1965[12] due to drafting errors in section 5 of the Singapore Amendment.[13] This was as the offending section stated that the “executive authority and legislative powers of the Parliament of Malaysia” [emphasis added] would be “transferred so as to vest in the Government of Singapore” [emphasis added].[14] The problems with the phraseology are twofold: (1) Malaysia’s Parliament did not have executive authority; and (2) it could not have been intended to vest executive and legislative powers in the “Government” as that customarily refers to the Executive.[15] Harding further posited that “Government” could not have referred to the legislature as otherwise the phrase “Legislature of Singapore” would have been used instead, as evidenced by the usage of the latter phrase in section 7 of the same Act.[16] According to Harding, these errors then deprived Singapore’s Parliament of its required plenary powers to subsequently enact Singapore’s post-independence constitution via the RSIA.[17] Harding thus opined that the only way the Constitution could have come into being via the RSIA is by rationalising the whole affair as a ‘revolutionary situation’.[18] Under this theory, Parliament assumed plenary powers on its own accord in the ensuing legal vacuum which it then used to pass the RSIA.[19] As a result of this, Harding concluded that the Constitution is not the grundnorm but rather “a manifestation of the grundnorm” that rests in parliamentary supremacy.[20]

However, Harding’s theory is unsustainable as it relies on a flawed assumption that the Constitution of the State of Singapore, 1963[21] was no longer valid at the time the RSIA was passed.[22] The 1963 Constitution was actually still valid post independence as section 7 of the Singapore Amendment enabled it to apply when it stated that “[a]ll present laws in force in Singapore immediately before Singapore Day shall continue to have effect”[23] such that the 1963 Constitution “evolved into or succeeded as the constitution of the 1965 State of Singapore on Singapore Day”.[24] This means that the RSIA was not an act of a supreme parliament creating a new constitution but rather that of an inferior parliament amending the existing constitution.[25]

Moreover, Harding’s reasoning with regards to the errors in section 5 is dubious. As analysed by former Chief Justice Chan Sek Keong, the apparent incongruity of the phrases used can be easily avoided by taking a purposive interpretation of them. The former Chief Justice pointed out that the phrase “executive authority” is likely to be free standing as it is not qualified by the words “of Parliament of Malaysia to make laws” such that it should not be read as so qualified.[26] For the issue with the phrase “Government of Singapore”, the former Chief Justice perceptively observed that any ambiguity would be removed should it be read purposively—it would refer to Singapore’s Parliament and the Head of State respectively.[27]

Lastly, even if Harding’s interpretation of section 5 of the Singapore Amendment is correct, Parliament would have still acquired the necessary plenary powers to enact the RSIA without the need for the ‘revolutionary situation’ envisaged by him. As held by the Singapore Court of Appeal in Public Prosecutor v Taw Cheng Kong,[28] the mere fact that section 5 of the Singapore Amendment failed to transfer plenary powers from the Malaysian Parliament did not leave the Singapore Parliament entirely without recourse.[29] This was as Singapore’s Parliament was able to separately obtain the required plenary powers from the “political fact of Singapore’s independence and sovereignty” which had “the consequences of vesting the Legislative Assembly of Singapore with plenary powers on Singapore Day”.[30] Harding’s proposition is thus not convincing in proving that Singapore lacks a theoretical basis to find constitutional supremacy.

IV.         weak constitutional supremacy

Notwithstanding the theoretical basis for finding constitutional supremacy in Singapore, whether it actually exists in practice is another matter. The Diceyean factors must not only be met in form but also in substance. Unfortunately, it does not appear that constitutional supremacy is strongly adhered to locally. This is as the legal fetters on what laws the legislature can pass are too lenient and, the Judiciary’s fear of overstepping its constitutional role has inadvertently given too much leeway to the executive and legislature to enact laws which weaken constitutional supremacy.

A.             Weak Legal Fetters

Enforcing the text of the Constitution alone is insufficient to achieve constitutional supremacy in practice as the Constitution lacks clauses, such as an eternity clause,[31] that impose legal immutability on its core provisions. In fact, if one looks at the text of the Constitution, there are no limitations whatsoever against the permissible range of amendments the legislature can make.[32] This means that in theory, any constitutional bills, including those that would ‘constitutionalise’ ordinary bills establishing a dictatorship, can be passed so long as the requirement for two-thirds majority support in Parliament under Article 5 is met. Such an outcome would render Article 4’s supremacy clause toothless against a rogue legislature bent on effacing the Constitution—a threat not insignificant in Singapore where the ruling party has held on to their parliamentary supermajority since independence.

In light of the lenient constitutional text, a common law check in the form of the basic structure doctrine to limit the range of permissible amendments is thus necessary. The courts are however unwilling to adopt anything other than an emaciated form of the doctrine. At its core, the basic structure doctrine involves the “judiciary deduc[ing] certain core characteristics of the constitutional order from the text, underlying philosophy, and history, and declar[ing] these to be unamendable”.[33] There are two flavours to the doctrine based on how rigorously it is applied, the ‘thick’ version in Kesavananda Bharati v State of Kerala[34] which allows courts to invalidate amendments should any constitutional values[35] be undermined, and the ‘thin’ version espoused in Mohammad Faizal bin Sabtu v Public Prosecutor[36] which only allows for invalidation if the amendment derogates from Westminster-style separation of powers.[37] The former has yet to be endorsed locally and it is nebulous whether the latter applies as all mentions of thin doctrine have only been in obiter and were draped in an abundance of caution.[38]

Even if the thin version applied, it would be of questionable efficacy. Firstly, it does not police the relationship between the government and its citizens as it only deals with “how political power is organised and divided between the organs of State in a particular society”.[39] This is problematic because a constitution at its core is a ‘social contract’ between the governors and the governed that sets out the rights and liabilities each party owes to the other. If this ‘contractual’ relationship is incapable of enforcement, then the constitution’s raison d’être is undermined.

Secondly, the thin version does not appear very effective at preserving ‘Westminster-style’ separation of powers. A core tenet of the Westminster system is that every citizen must have a right to representation in parliament—a constituency cannot be denied an elected representative.[40] However, the thin version allows a denial of representation, as demonstrated by the court’s obiter dicta in Wong Souk Yee which dispensed with the need to hold a by-election to fill up any post-election vacancies arising in a Group Representation Constituency (“GRC”):

Even if the right to representation forms part of the basic structure of the Constitution, it does not follow that there is a particular form of representation that is “fundamental and essential” to the Westminster model of government such that it cannot be departed from. In our judgment, there is nothing in principle that would prevent Parliament from devising the GRC scheme in such a way that a GRC could be left to be represented by less than its full complement of Members where one or more of them has vacated his or her seat. It is therefore unnecessary to consider the existence and scope of the basic structure doctrine for the purpose of disposing of this appeal.[41] [emphasis added]

Although the court explained it in terms of the ‘form of representation’ which suggests that having complete or incomplete representation is just a variation on the Westminster system, this is fallacious thinking. Fundamentally, the right to representation is a binary question: one either has representation or he does not. As such, only variations as to the quality of representation can count as variations to the Westminster system. In Wong Souk Yee however, the issue was not with the quality of representation, which the learned judge seemed to have implied by using the ‘complete or incomplete representation’ comparison, but rather the total absence of representation vis-à-vis the fifth parliamentary seat. In short, the voters were denied a voice separate from the other members due to the vacant seat. This analysis is possible because GRC members do not all act as one Member of Parliament serving the constituency. If it was converse, then Wong Souk Yee’s reasoning would have been sustainable for the seat would have remained completely filled until all members vacated. Seen in this light, the Westminster system was in fact derogated when the voters were denied a by-election to replace their missing Member of Parliament. The thin doctrine thus failed its purpose of protecting the Westminster system.

B.             Excessive Judicial Deference to the Legislature and Executive

The prevailing philosophy of the courts, born out of the fear of judicial overreach, is unfortunately too deferential to the legislature and executive for them to serve as the enforcer of constitutional supremacy. According to the Diceyean characterisation of constitutional supremacy, the legislative, executive and judicial branches are co-equal under the constitution with the duty entrusted to the judicial branch to ensure that all organs of state adhere to their constitutionally demarcated roles.[42] In discharging this sacred role, it is regrettable that the local courts adopt a ‘green light’ approach.[43] Under this philosophy, the judiciary is not “the first line of defence against administrative abuses of power: instead, control can and should come internally from Parliament and the Executive itself”.[44] In short, the judiciary prefers the government to practice ‘ownself check ownself’ and would seek to trust the government’s judgment as much as possible.[45] While there may be perfectly legitimate, compelling socio-political reasons for the court to adopt this lenient approach, it comes at the expense of constitutional supremacy by emaciating the judiciary’s checking power.

Perhaps the most vivid example of the damage wrought to constitutional supremacy is the fact that clauses which oust judicial review of legislative or executive acts can be valid. This occurred in the watershed case of Teo Soh Lung v Minister for Home Affairs[46] which affirmed the constitutionality of the amendments to Article 149 of the Constitution that ousted the court’s judicial review powers over cases relating to subversion or national security under the Internal Security Act.[47] In coming to their decision, the court displayed their ‘green light’ preferences by deferring to the decision of Parliament to leave it “to the Cabinet and the President acting in accordance with the advice of the Cabinet to determine whether it is necessary in the interests of national security to detain a person”[48] under the ISA. Such an approach threatens the foundations of constitutional supremacy by allowing a supposedly co-equal body to be ‘more equal’ than the rest.

Even though there has been some rhetorical flourish of late decrying ouster clauses, the fact remains that the courts are still very hesitant to invalidate them. In Nagaenthran a/l K Dharmalingam v Public Prosecutor,[49] the defendant claimed that section 33B(4) of the Misuse of Drugs Act (reproduced below) is an unconstitutional ouster clause because it blocks the court from reviewing whether the Public Prosecutor’s determination was valid.[50]

Section 33B(4) of the MDA: “The determination of whether or not any person has substantively assisted the Central Narcotics Bureau in disrupting drug trafficking activities shall be at the sole discretion of the Public Prosecutor and no action or proceeding shall lie against the Public Prosecutor in relation to any such determination unless it is proved to the court that the determination was done in bad faith or with malice.” [emphasis added]

The court, in a radical departure from Teo Soh Lung, opined that an ouster clause would be “constitutionally suspect for being in violation of Article 93 of the Singapore Constitution as well as the principle of the separation of powers”.[51] However, the actual impact of that statement remains highly unclear as the court continued to approach the case in a conservative, ‘green light’ way by taking a very narrow view of what counts as an ouster clause. This was as the court, after incanting the famous phrase from Chng Suan Tze v Minister for Home Affairs[52] that “[a]ll power has legal limits and the rule of law demands that the courts should be able to examine the exercise of discretionary power”,[53] found that section 33B(4) of the MDA was not intended by Parliament to be an ouster clause but rather only an immunity clause, which is constitutional.[54] Chief Justice Sundaresh Menon explained that the clause’s purpose is not to “exclude the jurisdiction of the courts to supervise the legality of the [Public Prosecutor]’s determination under [section] 33B(2)(b) of the MDA” but rather only to “immunise the [Public Prosecutor] from suit save on the stated grounds”.[55]

With due respect to the learned Chief Justice, it is surprising that the clause in Nagaenthran is not an ouster clause because section 33B(4) of the MDA appears to be on all fours with the ouster clause in section 8B(2) of the ISA. Section 33B(4) of the MDA also prevents the court from reviewing the validity of the decision maker’s determination save on some exceptions (bad faith and malice in this case). While this was something the Chief Justice noticed when he observed that “Parliament intended the inquiry under [section] 33B(2)(b) to be determined solely by the [Public Prosecutor] and not by the court”,[56] he ultimately found that both clauses were distinguishable. The core thrust of the Chief Justice’s argument appears to be that there was no ouster of the court’s jurisdiction over how the Public Prosecutor made his determination because the court would not have adjudicated on it anyway.[57] This was as the Chief Justice, concurring with Parliament, treated the determination’s validity as non-justiciable for being something the judiciary could not have competently adjudicated on due to a lack of specialist knowledge on the matter.[58] There is however a logical deficiency in this argument: finding that one would not have adjudicated on a case, even if given a choice, is not equivalent to not stripping away one’s right to adjudicate on the case. The former is a voluntary choice by the Judiciary to abstain from making a judgment while the latter is an involuntary ouster of the court’s discretion to judge by Parliament. The clause in Nagaenthran should thus be interpreted similarly as Teo Soh Lung’s—one that ousts judicial review. Alas, it would appear that the ‘green light’ approach still holds much sway over the judiciary.

V.         CONCLUSION

While constitutional supremacy has firm theoretical roots in the soil of our legal landscape, it unfortunately does not have much of a trunk on the surface. I express no judgment on whether this legal deficiency is a positive or negative one for our nation. It is a complex issue with strong arguments on both sides. Hardnosed pragmatism should trump rigid legal dogma, but at the same time, loosening the guardrails might invite trouble should the winds of authoritarianism blow in our direction. The purpose of this article is not to opine on what we should strive for but rather just to raise awareness of how weak constitutional supremacy is in Singapore.

---

* LLB (Candidate), National University of Singapore, Class of 2023. All errors and views expressed in this article remain my own.

[1] (1999 Rev Ed) [Constitution].

[2] A V Dicey, Introduction to the Study of the Law of the Constitution (Great Britain: Liberty Classics, 1982) at 3–4 [Dicey].

[3] Ibid at 37–39.

[4] Ibid at 78–79.

[5] A J Harding, “Parliament and the Grundnorm in Singapore” (1983) 25:2 Malaya L Rev 351 at 357 [Harding].

[6] Ibid.

[7] H Kelsen, General Theory of Law and State, translated by A Wedberg (Cambridge: Harvard University Press, 1945) at 114–115.

[8] Ibid.

[9] Constitution, supra note 1.

[10] Ibid.

[11] (Act 9 of 1965) [RSIA].

[12] (Act 53 of 1965) [Singapore Amendment].

[13] Harding, supra note 5 at 364–366.

[14] Singapore Amendment, supra note 12.

[15] Harding, supra note 5 at 364–366.

[16] Ibid.

[17] Ibid.

[18] Ibid.

[19] Ibid.

[20] Ibid.

[21] [1963 Constitution].

[22] Chan Sek Keong, “Basic Structure and Supremacy of the Singapore Constitution” (2017) 29: Special Issue Sing Ac LJ 619 at 658 [Chan, “Basic Structure”].

[23] Singapore Amendment, supra note 12.

[24] Chan, “Basic Structure”, supra note 22 at 655.

[25] Ibid at 658.

[26] Ibid at 650.

[27] Ibid.

[28] [1998] 2 SLR(R) 489 (CA).

[29] Ibid at paras 29–33.

[30] Ibid.

[31] See Article 79(3) of the Basic Law for the Federal Republic of Germany in the revised version published in the Federal Law Gazette Part III, classification number 100–1, as last amended by Article 1 of the Act of 29 September 2020 (Federal Law Gazette I p 2048): “Amendments to this Basic Law affecting the division of the Federation into Länder, their participation on principle in the legislative process, or the principles laid down in Arts 1 and 20 shall be inadmissible”.

[32] The requirements for a valid amendment to the Constitution are found in Article 5.

[33] Jaclyn L Neo, “Towards a “Thin” Basic Structure Doctrine in Singapore (I-CONnect Column)” (17 January 2018), I-CONnect: Blog of the International Journal of Constitutional Law, online: <http://www.iconnectblog.com/2018/01/towards-a-thin-basic-structure-doctrine-in-singapore-i-connect-column/>.

[34] AIR 1973 SC 1461 per Sikri CJ’s judgment.

[35] These values are normally abstract ideals drawn from a constitution’s preamble.

[36] [2012] SGHC 163.

[37] Ibid at para 11.

[38] See Ravi s/o Madasamy v Attorney-General [2017] 5 SLR 489 (HC) at paras 65–66 and Wong Souk Yee v Attorney-General [2019] 1 SLR 1223 (CA) at paras 77–78 [Wong Souk Yee].

[39] Yong Vui Kong v Public Prosecutor [2015] 2 SLR 1129 (CA) at para 71.

[40] Vellama d/o Marie Muthu v Attorney-General [2013] 4 SLR 1 (CA) at para 79.

[41] Ibid at para 78.

[42] Dicey, supra note 2 at 40.

[43] Chan Sek Keong, “Judicial Review—From Angst To Empathy, A Lecture to Singapore Management University Second Year Law Students” (2010) 22: Sing Ac LJ 469 at para 29.

[44] Ibid.

[45] Ibid at para 37.

[46] [1989] 1 SLR(R) 461 (CA) at paras 15–16 [Teo Soh Lung].

[47] (Cap 143, 1985 Rev Ed) [ISA].

[48] Teo Soh Lung, supra note 46 at para 19.

[49] [2019] SGCA 37 [Nagaenthran].

[50] (Cap 185, 2008 Rev Ed) [MDA].

[51] Nagaenthran, supra note 49 at para 74.

[52] [1988] 2 SLR(R) 525 (CA).

[53] Nagaenthran, supra note 49 at para 73.

[54] Ibid at para 74.

[55] Ibid at para 51.

[56] Ibid at para 67.

[57] Ibid at paras 67–74.

[58] Ibid at paras 66–67.

Director's Duties: Re-Examining the Bona Fide Test

A PDF version of the article can be found here.

---

Director’s duties: Re-examining the bona fide test

Desmond Chye & Russell Vaz*

I.                    Introduction

It is well-established that directors are fiduciaries of the company they serve. As fiduciaries, they owe a host of duties, including the duty to act bona fide in the company’s best interests. Unfortunately, recent developments have created uncertainty over how the test is to be applied. In particular, whether the test has a substantive objective component in addition to a subjective one. The dominant interpretation is that both components are part of the test. However, closer inspection exposes some flaws in this interpretation. This article will explore the arguments for the contrary position: that the test for the duty to act bona fide in the company’s interests is purely subjective.

II.                 The Original Test

Originally, the Singapore courts’ test for assessing bona fides was purely subjective. Directors need only act in “what they consider – not what a court may consider – is in the interests of the company” to satisfy the duty.[1] Courts were motivated by “strong policy considerations” to avoid “coerc[ing directors] into exercising defensive commercial judgment” that “will dampen, if not stifle, the appetite for commercial risk and entrepreneurship”.[2] As such, the duty would not be breached if the directors acted in the “honest and reasonable belief that they were for the best interest of the company, even if those decisions turned out subsequently to be money-losing ones”.[3]

It is apposite to note that the test may occasionally dip into the realm of objectivity. This is also known as the evidential standard version of the objective test. In the Singapore High Court case of Cheam Tat Pang v PP[4] the Learned Judge made the following remark:[5]

“It is settled law that if directors take risks which no director could honestly believe to be taken in the interests of the company, such actions could well support allegations that the directors in question had acted in breach of their fiduciary duties to the company”.

The Learned Judge’s remarks appear limited to establishing the evidential proof of the subjective mind of the director in question to see if he did in fact act reasonably. As such, the evidential objectivity did not detract from the overall subjectivity of the test.

III.              The Dominant Interpretation of the Current Test

The locus classicus for the new test is Ho Kang Peng v Scintronix.[6] The facts of the case are simple: the defendant director effectively paid bribes to advance the company’s overseas interests.[7] In applying the bona fide test, the courts stated:[8]

“However, this does not mean that the court should refrain from exercising any supervision over directors as long as they claim to be genuinely acting to promote the company’s interests. First, “where the transaction is not objectively in the company’s interests, a judge may very well draw an inference that the directors were not acting honestly”... The test in Charterbridge Corporation Ltd v Lloyds Bank Ltd [1970] Ch 62 (at 74) of “whether an intelligent and honest man in the position of a director of the company concerned, could, in the whole of the existing circumstances, have reasonably believed that the transactions were for the benefit of the company”, has been accepted and applied by this court in Intraco (CA) (at [28]). On the other hand, it will be difficult to find that a director has acted bona fide in the interests of the company if he “take[s] risks which no director could honestly believe to be taken in the interests of the company”… Secondly, it seems that the requirement of bona fide or honesty will not be satisfied if the director acted dishonestly even if for the purported aim of maximising profits for the company.” (emphasis added)

By advocating for the standard of an “intelligent and honest man”, and stating that acting for the company’s best interests would be insufficient, the courts seemed to be introducing a substantive objective component.

This interpretation of Scintronix has been largely accepted as the orthodoxy. Subsequent cases, such as the Singapore Court of Appeal case of Goh Chan Peng v Beyonics Technology Ltd[9] appear to support this view, stating that the bona fide test has “both subjective and objective elements”.[10] Eminent local academics such as Professors Dan Puchniak and Tan Cheng Han SC have also adopted the position that the test has a substantive objective component.[11]

IV.              The Alternative Interpretation of the Current Test

A.      Case Authority supporting a Purely Subjective Standard

A closer inspection of Scintronix reveals that it may not have laid down a discrete objective component at all. Phrases like “intelligent and honest man” which hinted at an objective standard were tempered with statements that they would only be used to “draw an inference” or when directors only “purport” to act in the company’s best interest.[12] This suggests that the courts were only advocating the use of an objective evidentiary tool to determine the director’s subjective state of mind, keeping in line with the traditional test. Furthermore, the court never explicitly stated that there would be an objective component. This is an odd oversight considering the uninterrupted pedigree of the purely subjective test.

The apparent support of Beyonics is also to be doubted. The court commented that “where the transaction is not objectively in the company’s interests, a judge may very well draw an inference that the directors were not acting honestly” (emphasis added).[13] This statement suggests that the courts were using an objective evidentiary tool as explicated by Professor Walter Woon. This possibility was noted by the Singapore High Court in Ong Bee Chew v Ong Shu Lin,[14] acknowledging that Beyonics could have merely used an objective evidentiary tool. We note that while Ong Bee Chew did ultimately support the two-part test, this was for procedural fairness rather than doctrinal accuracy as Beyonics was not available to the parties at the time of their submission.[15]

Furthermore, as suggested by Professor Hans Tjio, Scintronix may have simply applied the original Charterbridge test instead.[16] The court in Charterbridge Corporation Ltd v Lloyds Bank Ltd[17] held that the objective standard of an “an intelligent and honest man” would only be applied where the director exercised no discretion at all.[18] The rationale behind it is simple – failing to engage in any subjective consideration whatsoever, an objective assessment remained the only way to determine if he had acted in the company’s interests.[19] In Scintronix, the court found that “[t]he wrong committed by the Appellant in the present case cannot be regarded as an error of judgment – it arose because he failed to exercise any judgment at all.” (emphasis added).[20] As determined by Professor Tjio, this indicates Scintronix was merely a use of the original Charterbridge test in Singapore, applying the objective standard only when no discretion was exercised.[21] This would leave the traditional subjective test largely intact.

B.      Policy Arguments supporting a Purely Subjective Standard

The main public policy rationale for a substantive objective component is to prevent directors from carrying out immoral acts they believe were in the company’s best interests. Under a purely subjective test, he would be exonerated, lowering the standards of commercial morality.  However, the need to maintain a minimum standard of commercial morality must be balanced against the need to avoid stifling entrepreneurship with excessive judicial interference. 

Unfortunately, the two-part test risks stifling entrepreneurship. Its vagueness instils undue fear in directors who would be unsure of what standard to act on. Courts have elucidated that the objective component depends on whether, “objectively, the transactions were not in the company’s interests”.[22] However, in an attempt to mitigate judicial interference, they have softened the standard, stating that is only serves to “[hold] directors to minimum standards of commercial morality” and that the court will thus apply “a very low baseline in order to avoid unnecessary interference”.[23] Unfortunately, ‘commercial morality’ is a nebulously defined criterion that does little to resolve the uncertainty.

Furthermore, the test is phrased very widely as it takes the perspective of an ‘honest and intelligent director’ such that even negligence can potentially fall under the objective limb. This problem was evident in Scintronix where the court made the following remark:[24]

“He simply continued a highly irregular and improper practice which he understood to have been initiated by the previous management under a different form without so much as inquiring why it was made, whether it would implicate the Company, and whether proper sanction had been obtained. He had failed to exercise reasonable care.” (emphasis added)

While the subsequent cases of Beyonics and Ong Bee Chew all stressed a ‘minimum standard’, they never repudiated Scintronix’s requirement for ‘reasonable care’ such that it is possible for the broad standard to persist since that would comport with what is expected of an ‘honest and intelligent’ director. 

If the objective standard is truly, as a plain reading of the judgement indicates, that of an ‘honest and intelligent director’, the substantive objective test would impose too harsh a burden on directors. This is as 99% of all domestic companies are Small Medium Enterprises.[25] It is unrealistic to expect small businesses to have the funds to hire professionals to function as directors. Nor is it realistic to expect all business owners, many of whom are uneducated, to perform the role of an “honest and intelligent director”. Imposing such an exacting standard would “dampen, if not stifle, the appetite for commercial risk and entrepreneurship”.[26]

C.     Applying the Purely Subjective Test to existing Case Law 

The existing case law has dealt solely with the issue of bribery. Bribery satisfies the targeted fact matrix of being both immoral and prima facie in the company’s interest. However, such situations could have been easily prevented by a purely subjective test. As noted in Scintronix, bribery does not help the company’s long-term interests, only its short-term interests.[27] Ergo, a director could subjectively believe giving a bribe to be in the company’s short-term interest but still fail to subjectively believe it is in the company’s interests as a whole, after taking its long-term interests into account. The situation of receiving a bribe, as in Beyonics, is much more straightforward – the director accepted a bribe to do something he otherwise would not have done. It is unlikely any director could subjectively believe that accepting a bribe would further the company’s interests.

V.                 Conclusion

The bona fide test plays a critical role in regulating director’s duties. Uncertainty over its components could prompt unwarranted defensive decision making, curtailing the economic potential of Singapore’s businesses. A clearer test is therefore needed. While we hope that the courts will take the flaws of the two-part test into account, we recognise that the current orthodox interpretation of the bona fide test will likely remain the law for the foreseeable future. Our article merely aims to explore the possibility of an alternative rather than to overturn the status quo.

---

* LLB (Candidate) (NUS), Class of 2023. All errors and views expressed in this article remain our own.

[1] Cheong Kim Hock v Lin Securities [1992] 2 SLR 349 [Cheong Kim Hock] at para 26.

[2] Vita Health Laboratories Pte Ltd v Pang Seng Meng [2004] 4 SLR 162 [Vita] at para 17.

[3] ECRC Land Pte Ltd v Ho Wing On Christopher [2004] 1 SLR(R) 105 at para 49.

[4] [1996] 1 SLR 541.

[5] Ibid at para 80.

[6] [2014] 3 SLR 0329 [Scintronix].

[7] Ibid at paras 32-34.

[8] Ibid at paras 38-39.

[9] [2017] 2 SLR 592 [Beyonics].

[10] Ibid at para 35.

[11] D Puchniak, CH Tan & SS Tang, “Company Law” (2017) 18 SAL Ann Rev 247 at paras 9.7-9.8.

[12] Scintronix, supra note 6 at paras 32-34.

[13] Beyonics, supra note 9 at para 36.

[14] [2017] SGHC 285 [Ong Bee Chew] at para 78.

[15] Ibid at para 78.

[16] Hans Tjio, P Koh & PW Lee, Corporate Law (Academy Publishing, 2015) [Tjio, Koh & Lee (2015)] at para 09.043.

[17] [1970] Ch 62 [Charterbridge].

[18] Ibid at 74.

[19] Tjio, Koh & Lee (2015) supra note 16 at para 09.043.

[20] Scintronix, supra note 6 at para 40.

[21] Tjio, Koh & Lee (2015) supra note 16 at para 09.043.

[22] Beyonics, supra note 9 at para 35.

[23] Ong Bee Chew, supra note 14 at para 84.

[24] Scintronix, supra note 6 at para 40.

[25] Singapore Department of Statistics, “Topline Estimates For All Enterprises And SMEs, Annual” (accessed 2 January 2021)

<https://www.tablebuilder.singstat.gov.sg/publicfacing/createDataTable.action?refId=15808>

[26] Vita, supra note 2 at para 17.

[27] Scintronix, supra note 6 at para 37.

Counting Double Counting

The PDF version of this article can be found here.

---

counting double counting

Desmond Chye & Russell Vaz*

I.               INTRODUCTION

The law against double counting is seldom taught but remains an important part of the criminal law. It ensures an offender is not punished twice for the same crime. This is especially vital in this day and age where the preponderance of similar offences makes double counting far more likely. In this article, we will attempt to explain the law on double counting; in particular, the dispute over the test for ‘same offences’ and the effect that Tan Khee Koon v Public Prosecutor[1] has had on this dispute.

II.            WHAT IS DOUBLE COUNTING?

Double counting is not to be confused with its more famous relative, double jeopardy. As helpfully explained by the Singapore High Court in Chong Kum Heng v Public Prosecutor, “[t]he rule against double jeopardy is that a person cannot be made to face more than one trial for the same offence”.[2] In contrast, the rule against double counting prevents double punishment for the same offence, where the same set of facts gives rise to liability under more than one written law.

The statutory basis for the prohibition on double counting derives from section 40 of the Interpretation Act:[3]

“Where any act or omission constitutes an offence under 2 or more written laws, the offender shall, unless the contrary intention appears, be liable to be prosecuted and punished under any one of those written laws but shall not be liable to be punished twice for the same offence.” [emphasis added]

A plain reading makes it clear that the determining factor in finding double counting is whether the offences charged constitute the ‘same offence’. However, as this was not defined in the IA, case law is instructive on what it means.

III.           THE MEANING OF ‘the saME OFFENCE’ IN CASE LAW

The starting point would be the Singapore High Court case of Tan Khee Koon,[4] where the court applied the test in the Malaysian case of Jamali Bin Adnan v PP.[5] However, the endorsement of Jamali, which is a double jeopardy case, indicates that the definition of the ‘same offence’ is shared with double counting, thus introducing some complications.

In the area of double jeopardy, the locus classicus on what constitutes the ‘same offence’ is the English House of Lords case of Connelly v Director of Public Prosecutions.[6] In Connelly, there were two proposed approaches to finding similarity. Lord Morris adopted a more generous approach where the similarity need only be substantial.[7] On the other hand, Lord Devlin took a stricter view that the offence must be exactly the same in law because “legal characteristics are precise things and are either the same or not”.[8] However, it is unclear from the remaining Law Lords’ judgements which approach is to be preferred.

The most authoritative court to have weighed in on the conflicting approaches is the Privy Council (on appeal from Singapore). In Wee Harry Lee v Law Society of Singapore,[9] a solicitor was subject to two disciplinary proceedings for violations under section 84(2) of the Legal Profession Act.[10] The provisions violated were:

“[The solicitor] (a) has been convicted of a criminal offence, implying a defect of character which makes him unfit for his profession; or (b) has been guilty of … grossly improper conduct in the discharge of his professional duty …”[11]

The first proceeding arose from (b), which resulted in a two-year suspension, while the second proceeding founded on (a) resulted in a further two year suspension.

Their Lordships ultimately allowed the solicitor’s appeal against the further suspension, holding that the second proceeding was “an abuse of the disciplinary process”.[12] In coming to their conclusion, the court endorsed both approaches proposed in Connelly, but stopped short of deciding which approach was to be preferred as the result would be the same regardless. Even if the case fell outside the narrow scope of Lord Devlin’s test, it would still be covered “by the alternative form of relief which [Lord Devlin] favoured as mitigating the rigour of his strict test”.[13] This is likely the rule in Reg v Elrington,[14] where Cockburn CJ held that “[w]here a person has been charged with an offence (whether he be acquitted or convicted), he cannot be again tried “on the same facts in a more aggravated form””.[15] This leaves only two possible conclusions: the offences were the same using either Lord Morris’ or Lord Devlin’s test, or that the Elrington approach was used instead.

Assuming it was the former, it would appear impossible for their Lordships to have used Lord Devlin’s approach in Harry Lee if we were to examine the offences charged closely. This is because the emphasis in (a) is on the lawyer’s character (in other words, propensity for future professional misconduct) for which his criminal conviction is evidence thereof whereas (b)’s emphasis is on the lawyer’s actual act of wrongdoing. These are two conceptually distinct requirements that would certainly fail the exact similarity test. Therefore, in reaching the conclusion that they did, their Lordships must have considered the offences substantially—but not exactly—similar.

It is also interesting to note that the subsequent Singapore High Court case of Lim Keng Chia v Public Prosecutor[16] interpreted Harry Lee to stand for the former, as evidenced by the learned judge’s description of the case:[17]

“In Wee Harry Lee’s case, the Privy Council was asked to determine whether the two successive sets of disciplinary proceedings brought against the appellant were based on the same instance of misconduct; and if so, whether the second set of such proceedings amounted to a violation of the doctrine of autrefois convict and acquit or at the very least, an abuse of the disciplinary process provided for in the Legal Profession Act (Cap 217). The Privy Council answered these questions in the affirmative. It must be pointed out, however, that the Privy Council in Wee Harry Lee’s case was faced with two sets of the same sort of proceedings.”

The use of the phrase “same sort of proceedings” implied the learned judge’s belief that it was the similarity between the offences that was critical to the outcome. It would thus be reasonable to believe that in the future, the courts will not treat Harry Lee as a case applying the rule in Elrington.

We return to the decision in Tan Khee Koon, which began the controversy by citing Jamali. Ironically, this citation is the closest thing we have to a resolution. The test in Jamali involves an inquiry into whether essential ingredients of the offences are the same. In Jamali, the offences were held to be different because the essential ingredients of armed robbery differed from that of an offence under the Malaysian Internal Security Act 1960.[18] “[A]rmed robbery can be regarded as an aggravated form of theft causing fear of instant death or hurt on the intended victim voluntarily by means of a deadly weapon” while “the essential ingredients under the Internal Security Act 1960 are simply control of firearms and/or ammunitions without lawful authority”.[19] This bears a striking similarity to Lord Morris’ test, which the court in Jamali treated as the majority judgement without making any mention of Lord Devlin’s test. This signals an implicit endorsement of Lord Morris’ test over Lord Devlin’s by the court in Tan Khee Koon, albeit in obiter.

However, as the citation in Tan Khee Koon was merely obiter and no court has yet to make a definitive pronouncement, there regrettably remains some element of uncertainty as to whether Lord Morris’ substantial similarity approach is truly preferred.

IV.           TAN KHEE KOON’S SIGNIFICANT ADDITION TO DOUBLE COUNTING

Double counting arose as an issue in Tan Khee Koon. The case involved an accused who tried to steal $20,000 but was only able to obtain $4,500. He was charged with two separate offences: the first for obtaining $4,500 (“1st Offence”) and the second for attempting to obtain $20,000 (“2nd Offence”), with the former sum naturally being a component of the latter sum.[20] Although one concerned an attempt and the other concerned the actual commission of an act, statutory provisions prevent one from being punished with both.[21] The court therefore had to decide the following: can obtaining $4,500 be considered the same offence as attempting to obtain $20,000 if the $4,500 was a part of the $20,000?

The answer was simple: the 2nd Offence did not consist of an attempt of the 1st Offence alone. It consisted of a series of attempts; only one of which was committed. The learned judge approached the problem in an ingenious manner. His Honour broke down the 2nd Offence into 2 ‘sub-offences’, ie, (1) obtaining $4,500, and (2) attempting to obtain $14,740. In the result:

(a)   1^st Offence: Obtaining $4,500

(b)  2^nd Offence: Obtaining $20,000 = Obtaining $4,500 (‘Sub-offence 1’) + Attempting to obtain $15,500 (‘Sub-offence 2’)[22]

It becomes clear that Sub-offence 1 is identical to the 1^st Offence.[23] In order to avoid double counting, it became necessary to separate the offending ‘sub-offence’ from the remaining portions.[24] However, the nature of the 2nd Offence prevented that, causing it to be struck out entirely.[25]

Tan Khee Koon represents a significant addition to our legal firmament. A single offence can now be considered an amalgamation of several smaller ‘sub-offences’. Double counting is now far easier to successfully plead, especially with infinitely divisible elements like money.

Unfortunately, this rule introduces uncertainties into the already fractious law on double counting. Chiefly, how far can an offence be broken down? Such ramifications can only be answered by the courts.

V.             CONCLUSION

Given the importance of double counting as a statutory safeguard against punitive punishment, the considerable uncertainty over what is required to find it is highly undesirable. The endorsement of Jamali’s essential ingredient test in Tan Khee Koon, albeit in obiter, suggests it will be used going forward. Being the practical embodiment of Lord Morris’ substantial similarity test, it would appear that the debate is somewhat provisionally resolved. However, while resolving one uncertainty, Tan Khee Koon has created another: the court’s method of breaking down offences bypasses the need for offences to be holistically similar. As such, it would be in the interest of the courts to demarcate the boundaries of Tan Khee Koon’s rule to avoid stretching the concept of similarity too far and conclusively decide which test for similarity they prefer.

---

^* LLB (Candidate) (NUS), Class of 2023. All errors and views expressed in this article remain our own.

[1] [1995] 3 SLR(R) 404 [Tan Khee Koon].

[2] [2020] SGHC 21 at para 50.

[3] (Cap 1, 2002 Rev Ed) [IA].

[4] Tan Khee Koon, supra note 1 at para 105.

[5] [1986] 1 MLJ 162 [Jamali]

[6] [1964] AC 1254 [Connelly].

[7] Ibid at 1306.

[8] Ibid at 1340.

[9] [1983–1984] SLR(R) 768 (PC) [Harry Lee].

[10] (Cap 217, 1970 Rev Ed); that provision has since been redesignated as Section 83(2) of the Legal Profession Act (Cap 161, 2009 Rev Ed Sing), where it remains otherwise unchanged.

[11] Harry Lee, supra note 9 at para 2.

[12] Ibid at para 27.

[13] Ibid.

[14] 1 B. & S. 688 [Elrington]; cited in Connelly, supra note 6, at 1202.

[15] Ibid at 696.

[16] [1998] 1 SLR(R) 1.

[17] Ibid at para 8.

[18] Jamali, supra note 5 at 166.

[19] Ibid at 167.

[20] Tan Khee Koon, supra note 1 at para 100.

[21] Ibid at para 122.

[22] Ibid at paras 114-117.

[23] Ibid at para 115.

[24] Ibid at para 116.

[25] Ibid at para 119.