Tortoise with a Chance of Transmogrification? Good Faith in English Contract Law

Whether contract is a fundamentally individualistic institution, or should necessarily entail some attitude of respect between counterparties, is a debate which rumbles on. In a masterly recent contribution, ‘Good Faith in English Contract Law: A Humble ‘3 by 4’ Approach’, Mindy Chen-Wishart and Victoria Dixon advance the argument for recognising the principle of good faith in English contract law.[1]

Good faith, they argue, should be like the humble ‘measured tortoise’ – an organising principle the explicit recognition of which would help illuminate and taxonomize the law, reveal its ethical content and allow for incremental development in line with the character of the common law.[2] Good faith, after all, already exists in the subterrane underlying ‘many existing and apparently disparate rules’.[3] And though we have been speaking prose all along without knowing it, they argue for making explicit what has thus far been implicit. ‘The humble ‘measured tortoise’ can go far’, they say, ‘[b]ut it need not do so, and certainly it will be in no hurry.’[4]

My objective in this blog post is not to attempt a frontal assault, but more humbly, to highlight what I think is a significant shortcoming in an otherwise compelling argument. I suggest that Chen-Wishart and Dixon neglect a crucial comparison on which their argument depends – namely, the balance between the relative risks and advantages of their proposal. Two distinct but interrelated Questions are left open:

Question 1: What advantages does their proposal promise (assuming faithful implementation) compared to continuing under the status quo?

Question 2: How do the advantages identified in Question 1 outweigh the risks of the proposal inadvertently transmogrifying into something else?

Given that Chen-Wishart and Dixon’s proposal involves (at least to some extent) a leap of faith, one might rightly wonder why – if the measured tortoise need not go far at all – it should be allowed to go anywhere at all?

Of course, as Chen-Wishart and Dixon themselves recognise, their argument, as it currently stands, ‘can only be a starting point […] much further work can and should be done.’[5] My hope is that, by raising these tentative issues, I might in a small way contribute to the broader debate.

The ‘3 by 4’ Approach

According to Chen-Wishart and Dixon, ‘good faith’ can be broken down into three subsidiary concepts: (1) honesty, (2) fair-dealing, and (3) fidelity to contractual purpose. The stringency with which these apply is to depend on where along a spectrum the nature of a contractual relationship falls: (1) arms-length (baseline), (2) symbiotic, (3) involving a party with a recognised vulnerability, or (4) involving a fiduciary relationship (the strictest).[6] The three concepts, when combined with the four levels of stringency of application, constitute their ‘3 by 4’ approach.    

While it is a marked improvement from ‘good faith’ as an amorphous monolith, Chen-Wishart and Dixon’s ‘proposed taxonomy […] is relatively coarse-grained [and] provides no algorithmic certainty’.[7] Instead, as they recognise, the acceptability of their proposal would turn heavily on how it is applied in practice. To succeed, their model of good faith must behave like a measured tortoise – not a ‘relentless woodpecker […] forcefully adding to the substantive content of the parties’ obligations’ and ‘peck[ing] away at non-conforming parts of the law, rapidly transforming it’.[8] Yet, as they concede, the difference between these two creatures is only one of degree.[9]

These preliminary observations demonstrate the precarity of Chen-Wishart and Dixon’s position. At its heart, their argument depends on striking a fine balance correctly. Accordingly, justifying their proposal requires a comparison of three counterfactual worlds. The first is one in which we continue as per the status quo – where, as Moore-Bick LJ put it in MSC Mediterranean Shipping Company SA v Cottonex Anstalt, the law develops along established lines rather than under a general organising principle of good faith.[10] The second is Chen-Wishart and Dixon’s ideal scenario – where their proposal is adopted and faithfully implemented. The third is a deviation from the second – where having set off on its journey, the measured tortoise inadvertently transmogrifies into quite something else.

For Chen-Wishart and Dixon’s proposal to succeed, it cannot stop at simply showing why the second counterfactual is preferable to the first. It must also take into account the risks and dangers of the third counterfactual eventuating. In other words, it must be able to satisfactorily answer the two Questions I set out at the beginning.

Question 1
What advantages does their proposal promise (assuming faithful implementation) compared to continuing under the status quo?

Firstly, Chen-Wishart and Dixon must demonstrate why their ideal scenario would be preferable to the status quo. The question can be put, obversely, as ‘what do we stand to lose by persisting under the status quo?’ Indeed, the more compelling the answer to Question 1 is, the less of a hurdle Question 2 will be – the opportunities stemming from the principle’s recognition would outweigh the risks posed by it.

However, it is hard to know precisely what to make of Chen-Wishart and Dixon’s equivocal argumentation. Respectfully, they seem to want to have their cake and eat it too. A key argument that they advance to support their proposal, and at the same time neutralise objections, is that ‘we already have it’.[11] Good faith already permeates the law, even if we have not realised it yet. Indeed, their proposal relies on judges using the selfsame skills they have used all along – in ensuring the law walks in step with the unarticulated requirements of good faith – to continue developing the law after the organising principle is explicitly recognised.

How much of a difference, then, will their proposal make? They answer this with a classic lawyer’s evasion – the measured tortoise may but need not go far.[12] Unfortunately, this type of reasoning cuts both ways. If the law has successfully developed to satisfy the substantive requirements of good faith through disparate doctrines, then why is their proposal needed? If, on the other hand, the status quo is unsatisfactory, then Chen-Wishart and Dixon should explain which legal fixtures are at issue and in what concrete ways they hope to see the law change.

Critically, answering Question 1 requires more than just broad-brush arguments about transparency or ethical illumination.[13] It requires explaining why developments in line with good faith are currently being impeded by the lack of the principle’s explicit recognition. Only by engaging at a sufficient level of specificity can we substantively evaluate the value of Chen-Wishart and Dixon’s proposed taxonomy to the development of English contract law.

Question 2
How do the advantages identified in Question 1 outweigh the risks of the proposal inadvertently transmogrifying into something else?

The importance of Question 2 rests on recognising that it is one thing to say how the law should develop, and another thing entirely to predict how it will or is likely to. Addressing the latter is especially pertinent, given that one of the major objections to recognising good faith in English contract law is the fear that it is likely to introduce uncertainty into the law. Unfortunately, however, Chen-Wishart and Dixon’s argument confines itself to only addressing the former point.

This weakness is well-illustrated by the analogy they draw between their proposal and the neighbour principle laid down by Lord Atkin in Donoghue v Stevenson.[14]  The neighbour principle, they argue, ‘has facilitated cautious incremental development through reasoning by analogy with existing authorities’[15] and ‘has not imposed negligence liability whenever carelessness results in foreseeable harm.’[16] Recognising an organising principle of good faith ‘would thus not be foreign to the modus operandi of the common law.’[17]

However, by glossing over the tortured history of the neighbour principle, their analogy tells only half the story. In fact, the other half is a cautionary tale of how protean organising principles do not sit easily with incremental reasoning.

The familiar cases of Dorset Yacht,[18] Anns[19] and Caparo[20] (to name but the few most prominent ones) remind us of how what began modestly transmogrified into a persistent source of instability in the law, notwithstanding attempts to reign it in.[21] The House of Lords and Supreme Court, at various turns, vacillated and split on fundamental questions of the principle’s characterisation.[22] The lower courts struggled to apply the principle consistently and accurately.[23] Whatever meaning Lord Atkin originally intended his words to bear, the neighbour principle, once recognised, took on a life of its own.

No doubt, Chen-Wishart and Dixon are aware of this history. Indeed, they clarify that the same doctrinal reasons for rejecting the application of the neighbour principle as a general duty of care formula ‘appl[y] equally to the notion of good faith in contracting’ – good faith cannot be an ‘independent and unmediated cause of action’.[24] Yet, as I have argued, this clarification is not quite enough—two crucial questions remain unanswered. Should we expect the neighbour principle’s history of uncertainty and vacillation to repeat itself if we explicitly recognise good faith as an organising principle? And to the extent that we should, is it worth it?

Concluding Thoughts

Any comparison of counterfactual worlds – of the sort required to ground Chen-Wishart and Dixon’s argument – would undoubtedly benefit from lessons gleaned from actual experiences in other jurisdictions. Canada is a jurisdiction whose recent recognition of good faith – as a ‘general organizing principle […] that underlies many facets of contract law’[25] – most closely resembles Chen-Wishart and Dixon’s proposal. At this time, it would be unwise for me to venture into any substantive comment, except to refer the reader to literature evaluating the illuminating developments there,[26] and echo Paul Davies’ scepticism towards optimistic early predictions that the transparency of that approach would only serve to enhance commercial certainty.[27]

For now, I conclude by admitting that it is always easier to criticise than it is to build and defend. As such, this blog post should in no way be seen as a deprecation of Chen-Wishart and Dixon’s impressive contribution to this important debate. I merely point out that, as their argument currently stands, one is still left wondering whether – given that the measured tortoise need not go far at all – it should be allowed to go anywhere at all? I suggest that finding answers to this requires a keen focus on, and robust engagement with, the two Questions I set out above.