I. The Problem of Unauthorised Dispositions
What should our commercial law do about the problem of unauthorised transfers of chattels? More formally, if someone grants you a proprietary interest in a thing, how far should your interest be affected by the fact that the grantor neither owned the thing nor was authorised by the owner to dispose of it?
The question is a chestnut, but none the worse for a new look. For one thing, a lot of what has been said about it in England is either platitudinousFootnote 1 or aimed at a specific problem arising out of unauthorised dispositions rather than the issue as a whole.Footnote 2 For another, even though this is an area where civilian lawyers think very differently, surprisingly little is said about how far their experience might inform our own.Footnote 3 Third, even though the present system can be politely described as an incoherent mess, there has been only one reform in the last 50 years – namely, the unlamented disappearance in 1995 of market overt,Footnote 4 a comparatively small change precipitated not by reforming zeal or academic pressure, but by an adventitious scandal in the art world.Footnote 5 And lastly, the point matters. Even though an increasing proportion of this field is covered by specific regimes – on the ranking of security interests, on priorities in insolvency or for that matter the background rules on equitable and legal interests – we still need logical and defensible background principles against which these special rules can operate.
Hence the present article. It aims to look at the subject in the round, where necessary drawing from how things are done elsewhere. To keep it within bounds, it has a few limitations. It will not cover specific self-contained systems, for example those applicable to company charges or registered ships (though it will discuss how such schemes should be accommodated in any general regime); nor will it cover complexities over unauthorised dispositions of part of a bulk.Footnote 6 Coverage will be angled towards non-consumer cases, that is to cases where owner and ultimate transferee are businesses. This is because, even though the present law rarely in fact treats consumers differently,Footnote 7 the need to provide special protection to entirely private parties may well raise separate difficulties. So too, discussion will be limited to wrongful dispositions: cases where a possessor is authorised by law to override an owner's title, such as execution sales or disposals of seized goods,Footnote 8 do not raise the same issues and will be ignored.
These matters aside, however, it will aim at as much generality as possible. To anticipate briefly, it will make the case for enacting a general principle of entrustment – broadly, a rule that anyone voluntarily entrusting possession of goods to another ought to take the risk of subsequent malversation – and then to go on to work out the implications of such a rule (which are less straightforward than one might think). For ease of reference, it will use three stock characters: O, a chattel owner (or sometimes holder of some lesser interest); P, an intermediary possessor of the chattel purporting to convey a proprietary interest in it; and R, the supposed recipient of that interest. Admittedly this is an over-simplification, since the chattel may well pass through the hands of two or more parties between leaving O's hands and arriving in R's (an important point, on which more below). But we will employ this scheme as a general template.
II. The Present Position: A Summary
To begin with, a short summaryFootnote 9 of the present English position and attempts to reform it may be helpful.
The starting point is that R gets no better right than P had: nemo plus iuris transferre ad alium potest quam ipse habet.Footnote 10 This is so at common law,Footnote 11 as regards not only ownership but all proprietary interestsFootnote 12 ; in one limited case it is preserved by statute in the form of s. 21 of the Sale of Goods Act 1979.Footnote 13 Hence if R wishes to prevail over O it must invoke a specific exception. Of these there are about half-a-dozen, depending on how you count them.
First, O may be estopped from asserting its right against RFootnote 14 if it expressly or impliedly represents to RFootnote 15 that P owns the goodsFootnote 16 or that it has authorised P to dispose of them.Footnote 17 Common-law in origin,Footnote 18 though statutorily acknowledged,Footnote 19 this principle can in principle validate any kind of disposition.Footnote 20
Second, the Sale of Goods Act 1979 deals with the repercussions of the rule allowing a buyer to become owner without delivery, or conversely take delivery without yet being owner.Footnote 21 Under ss. 24Footnote 22 and 25(1)Footnote 23 the seller in the first case, and the buyer in the second, can pass title to a second good-faith buyer R, provided the latter takes delivery from P.Footnote 24 R must be a buyer, pledgee or someone in an analogous positionFootnote 25 ; in other cases R, however faultless, remains unprotected.Footnote 26
Third, in very limited circumstances R is protected against a subsequent claim by O on the basis of a simple entrustment by O to P. This arises under s. 2 of the Factors Act 1889, extending a narrower common-law protection derived from the agency doctrine of ostensible authority.Footnote 27 P must have been in the business of dealing in goods of that sort; O must have entrusted them to P for sale or pledgeFootnote 28 ; and P must have passed (though not necessarily delivered) them to R in the ordinary course of business.
Fourth, good-faith acquisition is partially protected where P defrauds O of goods which it then transfers for value to a good-faith receiver R. Essentially, R's right becomes indefeasible if two requirements are satisfied: namely, that P obtained a voidable title from O,Footnote 29 normally under a sale contract voidable for fraud,Footnote 30 and that there has been no effective avoidance of P's title before the disposition to R. This was always the rule at common lawFootnote 31 ; when the transaction between P and R is one of outright sale, it is statutory.Footnote 32
Fifth, there are miscellaneous other cases of protection, one of the best-known being that afforded by statute to buyers, other than motor dealers, of vehicles let on hire purchase.Footnote 33
Lastly, all the above exceptions exist against the background of other regimes, which serve to complicate the picture further. First, the rules of equity always hover in the background, shielding good-faith purchasers against equitable interestsFootnote 34 – a matter of some significance in the case of non-possessory security interests short of ownership, because such interests are nearly all equitableFootnote 35 and there has since 2013 been no requirement for non-UK companies to register them, even when they affect property in England.Footnote 36 Second, there are a number of regimes on priorities existing more or less independently of the general law: for instance the rules in Part 25 of the Companies Act 2006 on charges created by UK companies, Sch. 1 to the Merchant Shipping Act 1995 on dealings with registered ships, and the Cape Town Convention regime for interests in aircraft.Footnote 37 Further, it should not be forgotten that insolvency law special rules apply to protect good-faith purchasers where property is disposed of by an insolvent whose power of disposition has otherwise been curtailed.Footnote 38
III. Reform Efforts
Such a ramshackle structure might be expected to spur reform. It has not: the only exception is certain proposals primarily connected with security law which might have had incidental effects on unauthorised dispositions as a whole.Footnote 39 In 1966 the now-defunct Law Reform Committee suggested minor changes,Footnote 40 inspired mainly by a clutch of cases where fraudsters had sold cars to innocent dupes for cash.Footnote 41 These would have tinkered with the rules on buyers in possession, extended the voidable title rule to cover almost all cases of fraud, and extended the then rule of market overt to cover all goods bought in retail premises or at public auction. Nothing happened. In 1989 the Diamond Report,Footnote 42 a document largely concerned with security interests, suggested in passing that anyone entrusted with goods under a contract of sale, lease or hire purchase might be empowered to pass title to an innocent buyer. Again nothing happened. Five years later, the Department of Trade and Industry tentatively floated a proposal under which in essence any innocent purchaser from a person in possession of goods with the owner's consent would get good title.Footnote 43 Unfortunately this suggestion, which would hardly have raised an eyebrow on the other side of the English Channel,Footnote 44 was very superficially supported and inadequately argued. It was intemperately attacked,Footnote 45 and quickly forgotten. In 2005 the Law Commission expressed an intention to investigate the whole subject of nemo dat in non-theft cases,Footnote 46 but by 2011 admitted that it had lost interest.Footnote 47 Meanwhile, the only reform that did take place was the one mentioned at the beginning of this article, namely the abolition of market overt.
IV. The Present Law: An Appraisal
In the light of all this, it is difficult to regard the present English position as anything other than an arbitrary and unpredictable mess. There is no overall plan for deciding when a good-faith receiver ought or ought not to be protected. Furthermore, the protections provided to third-party acquirers are spectacularly capricious. For example, a receiver R can frequently sleep easy if it buys goods in P's possession, less so if it lends against them, and hardly at all if it takes those same goods on hire, uses them or asserts a lien over them.Footnote 48 Again, a financier providing a van under a finance lease can repossess it if the lessee dishonestly sells it; but if it lets the same vehicle out on hire purchase or conditional sale it loses out unless the buyer is a motor dealer.Footnote 49 In fraud cases the distinction between a voidable title in P, which can benefit R provided it takes in good faith, and a void title, which cannot, is at best hard to draw and at worst incomprehensible.Footnote 50
Again, there is no unified concept of a good-faith receiver, with the result that the hurdles to be cleared by R in order to receive protection vary curiously according to the nemo dat exception in question. If buying from a seller or buyer in possession R must not only be in good faith but take delivery; if from a mercantile agent or voidable title-holder, a paper transaction is enoughFootnote 51 (though in the latter case the sale must be in the ordinary course of businessFootnote 52 ). Yet again, the burden of proof is apt to swing like a weathercock. When sued by O, if R relies on a voidable title in P it can sit back and tell O to establish its caseFootnote 53 ; to profit from any other exception to nemo dat it must it seems plead and prove it.Footnote 54
Indeed, the strongest indication that the status quo is unsatisfactory is that, while many writers have described it, few if any have seriously tried to support it. Furthermore, such efforts have been fairly consistently unconvincing. For instance, the objection raised against the Department of Trade's 1994 proposal essentially comprised a plea that any substantial downgrading of existing property rights must be bad; that it would be unthinkable if the depositee of a car or a ring was empowered unilaterally to expropriate the owner by selling it; and that the proposal was an open invitation to fraud.Footnote 55 Of these, the first begs the question; the second is highly controversial (assuming nothing to chose between O and R in terms of fault, it is perfectly arguable that O, as the person choosing whether and where to entrust goods, should take the risk of subsequent malversation); and the third is simply unsupported (indeed, it presumably entails a belief that most of Europe, where person in possession with the owner's consent can generally pass title, must compared with England be a hotbed of dishonesty).
V. The Way Ahead
A. An Entrusting Rule
If English commercial law is to reflect live up to its reputation of being as straightforward and uncomplicated as possible, we clearly need to put matters on a rational footing. What will be suggested in this section is the introduction of a general “entrusting rule”.
Crudely summarised, this means that in so far as R can prove that it received chattels in good faith from a possessor P, and that O did not lose possession against its will, R should succeed as against O. This should, however, be subject to three conditions. First, R must have believed when it received the goods that P either owned them, or otherwise had authority to deal with them. Second, R must not have had knowledge of matters indicating to a reasonable person in its position that P did not have the power of disposition. Third, R must have given value (subject to one qualification, referred to below).
In addition, it will be suggested that any scheme should have four further features. First, it should not be limited to ownership proper but should cover all proprietary interests, whether it is a question of O losing them or R gaining them. Second, it should be capable of defeating the interests of third parties other than O, provided that those entitled were not themselves dispossessed without their consent. Third, it should (subject to a few exceptions) supplement existing exceptions to nemo dat. And fourth, it should acknowledge that it existed against a background of specific schemes and ideally allow the latter to be slotted in as easily as possible. All these matters are covered in detail below.
B. Why an Entrusting Rule?
Why choose an entrusting rule? It is suggested that it can be justified on four bases.
1. The point of principle
The main argument advanced here is as follows. Any discussion of nemo dat is at bottom a discussion of how to balance the interests of an original owner O with those of a good-faith acquirer R once P is out of the picture. There are, of course, any number of ways of doing this (some of which will be mentioned below). But the entrusting rule, or something like it, is it is suggested the only one that, in reaching the balance, gives adequate weight to the essential nature of ownershipFootnote 56 itself.
According to one's point of view, one can regard ownership as an institution resting on a number of possible justifications: for example, autonomy,Footnote 57 efficiencyFootnote 58 or the practical needs of commerce.Footnote 59 But whichever view one takes, it is suggested that two features stand out as distinguishing characteristics, without which an institution would not be ownership as we know it. One is the idea that ownership exists as the irremovable residual or background right to dictate how a thing is to be used or exploited, which continues to subsist whatever other lesser rights may come and go.Footnote 60 The other is a degree of permanence. Interests in assets can be difficult to characterise convincingly as ownership if they are precarious or readily defeasible without any action on the part of the person entitled.Footnote 61
These two features, suggestive as they are of relatively permanent interests existing in the background whether or not consciously exercised, suggest (it is submitted) that if we are to pay proper respect to the institution of ownership in deciding nemo dat conflicts, we need a system that makes ownership rights presumptively indefeasible, unless and until the owner chooses to do something consistent with consent to being divested.Footnote 62 And this is precisely the essence of an entrusting rule. A principle of this kind respects these features of ownership by insisting that before O can be expropriated there must be an affirmative act on its part consistent, at least outwardly,Footnote 63 with an intent to alienate it.Footnote 64 Conversely, it refuses to deprive O of ownership where the property was taken without such a consent, this being inconsistent with the status that ought to be accorded to property rights.Footnote 65
Of course an entrusting rule is not the only possible approach. A large number of other possible criteria have been suggested for deciding who ought to win a contest of priorities between owner and acquirer: for instance relative fault,Footnote 66 some more general form of equitable apportionment,Footnote 67 allocative efficiency of scarce goods,Footnote 68 relative needFootnote 69 or some sophisticated combination of these.Footnote 70
Such approaches nevertheless all face one common difficulty: namely, that unlike the entrustment principle they treat title conflicts similarly to other issues, such as contract or accident law, and as a result fail to give sufficient weight to the special nature of ownership.Footnote 71 But they raise other problems too. Notably, the adoption of any approach based on the actual situation of the parties and applied on a case-by-case basis makes the law more unpredictable and the task of settling title disputes quickly and efficiently correspondingly harder. Determining entitlement (or dividing loss) according to fault, for instance, potentially turns every property dispute into a complex evidential dispute about relative blame. Furthermore, it raises formidable problems with repeated dispositions and multiple parties. Imagine O entrusts a thing to P; P sells it to R1, which on-sells to R2. Whose fault would be relevant, and in what proportions, to determine R2's rights against O: R1's, R2's or both?Footnote 72 To be fair, it has been suggested that we can overcome such complaints by applying fault criteria on a “typical situations” basis: that is, by grouping cases into typical situations, determining the rule to be applied to each, and then applying it without reference to the facts of the actual dispute.Footnote 73 But this raises its own further difficulties. Identifying typical situations is not straightforward (how widely or narrowly should they be drawn?). Furthermore, any move in this direction is effectively regarding the chosen criterion not so much as a test, as merely a guide to be taken into account. It is no doubt for reasons such as this, that criteria such as relative fault have (it is suggested rightly) never attracted much following, either in EnglandFootnote 74 or for that matter in any major common or civil law jurisdiction.
2. The relation between entrusting and the present law
Apart from the argument of principle, it is suggested that an entrusting rule would not only simplify matters, but do so with a welcome lack of drastic change. This is because the existing position under English law is already nearer to an entrusting rule than one might believe.Footnote 75 This point may seem surprising, but a moment's thought will confirm it. Of the specific exceptions to nemo dat described in Section II above, three explicitly depend on voluntary entrustment. Sections 24 and 25 of the 1979 Act both demand that the seller or buyer, as the case may be, should be or remain in possession with the consent of the owner; and the same goes for the factor's possession in s. 2 of the Factors Act 1889. Equally two more exceptions, s. 23 on voidable title and s. 27 of the Hire Purchase Act 1964 dealing with vehicles on hire purchase, largely assume it: one cannot readily have a voidable title or a hire-purchase arrangement without an underlying entrustment of possession. Turning to the more general law, it is also worth remembering that other rules incidentally raising issues of unauthorised disposition, such as those on equitable title, unregistered charges and insolvency, are also concerned overwhelmingly with situations where O, the entity entitled, acquiesces in a middleman P possessing the property later alienated to R.
The point also works in the converse way. Hardly any of the current exceptions work where O has been deprived of possession against its will: indeed, the one that regularly had this effect, market overt, has been abolished. It is true that estoppel remains capable of protecting the third party in cases of theftFootnote 76 : but in practice virtually all estoppel cases also turn on entrustment.
In effect, therefore, it is suggested that the effect of introducing a general entrustment rule should be seen as embodying not so much a change in the present law, as a confirmation of the existing pattern inherent in it by the removal of a number of exceptions. In large part, moreover, these are exceptions that are hard to justify. Few would seriously support, for example, the Helby v Matthews Footnote 77 rule excluding hire-purchase transactions from the protection given to R in respect of goods sold under reservation of title because they contain a technical option in the hirer not to buy,Footnote 78 or the bar on mercantile agents passing title to R if (unknown to R) they were entrusted with goods for some purpose other than sale or pledge.Footnote 79
3. The partial parallel with the UCC
We have so far not mentioned American developments. Until the coming of the UCC, most American states essentially applied the English system of a number of discrete exceptions to nemo dat.Footnote 80 Article 2 of the UCC, however, has now greatly widened the third party's protection. Article 2-403 in particular extends the ability of an entrustee to transfer goods in two vital ways: by saying that P obtains a voidable title in almost every case of fraudulent purchase,Footnote 81 and even more importantly by saying that any entrustment of goods to a dealer in those goods allows the latter validly to transfer them to a good-faith receiver.Footnote 82 This solution goes a good deal of the way towards a general entrustment rule, though admittedly not the whole way: thus it does not allow non-dealers to pass title on the basis of entrustment,Footnote 83 and affects only the interests of the immediate entruster O (thus leaving other entrusters’ interests to trip up unwary buyersFootnote 84 ). Furthermore, it is largely based on the same ideas: namely, that in at least certain circumstances a voluntary entrusting ought to serve to justify expropriating O.Footnote 85 It is also worth adding that, although this article is not specifically concerned with the specialist topic of security interests, Article 9 of the UCC also goes some way towards protecting a buyer in the ordinary course of business against pre-existing security interests, even if the latter are otherwise perfected and thus on principle effective against third parties.Footnote 86
4. The European dimension
If the UCC in the American context has gone a large way towards adopting a de facto theory of entrustment, European civil lawyers have gone the whole course. By and large the theory of entrustment is now regarded as entirely orthodox,Footnote 87 so that when faced with unauthorised dispositions of the kind dealt with in this article civil law regimes subordinate O's rights to R as a matter of course in any case where O was not dispossessed without its consent. These systems are worth a look, not only because they may provide inspiration for reform here but also since (as will appear below) they point up problems that might go unnoticed in an entirely common-law treatment.
The most carefully modulated system in this respect is German law. The civil code explicitly sets the scene, providing that anyone receiving a chattel in good faith from a person he believes to be the owner gets good title, free of all prior interests, save where he is grossly negligent or the original owner was the victim of loss, theft or some other form of involuntary dispossession.Footnote 88 Furthermore, where P is a businessperson it suffices that R merely believed P to have authority to dispose of it even if he did not believe P to be the owner.Footnote 89 The codal principle, moreover, protects not only buyers but those taking pledges or claiming a number of lesser interests in goods.Footnote 90 Switzerland is very similar in approachFootnote 91 ; indeed, in one respect its civil code is even more generous to third-party acquirers than Germany's.Footnote 92
France is more impressionistic. It admittedly has no provision explicitly applying an entrusting principle. It has nevertheless reached much the same result through inventive interpretation of the famous provision “en fait de meubles, la possession vaut titre”,Footnote 93 a provision originally aimed at something rather different but increasingly pressed into service as a protection for buyers.Footnote 94 The main difference be between France and Germany is that protection is less widely extended beyond good-faith buyers, and that a lack of good faith in R is more readily found: but these are details.
The support provided by commentators in civilian jurisdictions is moreover similar to that advanced at the beginning of this section. That is, they argue that if we need to draw a line between protection of property and worry-free trade between honest merchants, consent by O to dispossession is the least unconvincing place to draw it, since then there is at least some conscious consent by O to events that might lead to its dispossession.Footnote 95 Although, as already touched on, the details vary,Footnote 96 the general principle seems to work fairly well. It is noteworthy that it was adopted without serious question by the drafters of the Draft Common Frame of ReferenceFootnote 97 as a blueprint for a future European law of things, and for much the same reasons.Footnote 98
VI. Working Out the Entrusting Rule: When Should R Be Protected?
It is one thing to advocate an entrusting rule; quite another matter, and in practice rather more important, to work out in more detail what form it should take. That is the aim of the remainder of this article.
A. PossessionFootnote 99
The first vital requirement for an entrusting rule, as mentioned above, is that O should have consented to put P in possession. Normally this will be simply a matter of fact: was P placed in de facto control of the goods and did O consent to lose that control?
As for consent, it is suggested that there is no need to define this in detail,Footnote 100 and that in practice the concept would be generously interpreted it would be unlikely to cause difficulties,. Thus existing statutory provisions in England explicitly requiring P to be in possession with O's consent are already construed widely, as looking to immediate consent only and virtually ignoring complicating factors such as deceit or trickeryFootnote 101 ; this also reflects practice in many civil law jurisdictions,Footnote 102 and would no doubt continue. Nevertheless, a number of points could do with clarification.
First, what must be placed in P's possession? Apart from the goods themselves, it is submitted that symbols of them, for example a key or swipe-cardFootnote 103 giving access to a warehouse or container, should clearly also suffice. This not only fits neatly with the treatment of possession elsewhere in English law,Footnote 104 but also reflects the fact that means of access of this kind are regularly used to transfer goods especially where physical transfer would be impracticable. The same also goes for possession where there is evidence of attornment on the part of a third party.Footnote 105
What about documents, whether traditional documents of title such as bills of lading or others which cannot be used to transfer possession? The answer is that if forms of possession such as attornment or symbolic possession through a swipe-card are admitted, a fortiori the same must go for documents: if an owner O entrusts them to P, rather than having them taken out of its hands, R should be in no worse position than it would have been in respect of the goods. This solution has the advantage of maintaining continuity with many of the existing exceptions to nemo dat in England, where dealings with documents of title are treated as dealings with the goods themselvesFootnote 106 ; it would also in practice largely approximate the situation in England to that under Article 7 of the UCC in the US, which deals specifically with documents of title – though the latter is, to say the least, complex.Footnote 107
B. The Good Faith of R
The second issue is much more important: to defeat O, R should have to be in good faith and without notice of P's lack of power to deal with the goods. On principle this goes without saying: but on the detail some important issues arise.
One, already touched on in connection with German law, concerns the precise belief to be required of R. Should R have to show that it believed P actually owned the goods? Or ought it to suffice that R thought that P had the authority from the owner to make the transfer, even if R knew the goods were, or might be, owned by someone else? The point looks narrow, but it may matter. For example, it is quite plausible that O may sell to P on retention of title terms; that P, not having paid O, sells to R; that R knows this fact, but believes that P is authorised to sell on the goods even before payment (possibly on the basis that it assigns any right to payment to O). If a belief in ownership is necessary, R loses: if a belief in authority suffices, it wins.Footnote 108 This is not something English lawyers readily discuss, and under the present law the answer seems to vary.Footnote 109 Civil law jurisdictions also reach different answers; but at least in commercial cases most accept that a belief in P's authority is enough.Footnote 110 It is submitted that this latter solution is the better one. If we regard a voluntary surrender of possession as justifying imposing on O the risk of subsequent misdealing by P, there is no reason to complicate the matter by imposing an artificial requirement that R's good faith relate to the ownership of, rather than the empowerment to deal with, the thing concerned.Footnote 111
Another crucial issue concerns what degree of knowledge (or fault) should disable R from relying on the entrusting rule. No one could seriously argue that R should win if it actually knew of P's lack of authority, or suspected it and chose to turn a blind eye to an inconvenient possibility. But what of the more difficult case where R, even if not knowingly dishonest, was to some extent at fault: whether by failing to appreciate facts obvious to a reasonable person, missing a more or less obvious inference from facts known to it, or failing to make proper enquiries (all of which are subtly different)?
Under the present English law of nemo dat, this issue remains unsettled, save that there is no doubt that at least some conduct short of positive dishonesty will do. So much is clear because several statutory provisionsFootnote 112 demand both good faith (which by legislative fiat means honesty and no moreFootnote 113 ) and a lack of notice of P's lack of authority to deal, which presumably means something different. But what counts as notice for these purposes remains spectacularly unclear. One can cite suggestions that the doctrine of constructive notice (i.e. liability for anything short of actual knowledge of facts) should be driven out of commercial lawFootnote 114 ; that constructive notice is potentially relevant, at least in some casesFootnote 115 ; that the existence of an unusual background to a sale itself constitutes noticeFootnote 116 ; that notice is actual knowledge of facts yielding a reasonable inference that a disposition is unauthorised, even if that inference remains undrawnFootnote 117 ; that there is no positive duty in a buyer to investigate underlying factsFootnote 118 ; or that it all depends on what sleuth-work is usual in the circumstances.Footnote 119
For once, civil law practice is not much help either. In France the requirement of good faithFootnote 120 has been held fairly consistently to bar the buyer's claim to protection where the sale was in circumstances suggestive of skullduggery.Footnote 121 In Germany gross negligence explicitly excludes protection: a point on which there is a great deal of detailed law, not always consistent.Footnote 122 The Draft Common Frame of Reference (DCFR) adopts a strict standard under which anything other than slight negligence bars R.Footnote 123
In short, we need a clean slate. It is suggested that the best approach is as follows. First, there should be no general duty in R to enquire as to P's position. Businesspeople do not generally owe a duty of care to safeguard the integrity of others’ property rights,Footnote 124 and to demand that any purchase of goods be accompanied by due diligence would disproportionately increase delay and cost with no guarantee of comparable gain. Indeed, there is much to be said for resisting any introduction of such a duty even where enquiry is usual practice: although failure to make enquiries might be evidence of knowledge, it would (it is suggested) be going too far to make the protection of innocent third parties strictly dependent on their acting in the ordinary course of business. On the other hand, it is suggested that, in line with developments in the law generally,Footnote 125 actual knowledge of facts should defeat protection if those facts would indicate to a reasonable receiver that something was wrong. To this extent, it is suggested that the law ought to continue to demand both honesty in fact and, separately, lack of notice in this sense.
This leaves a third issue: the burden of proof. Under the present English law it is generally on the receiver RFootnote 126 (subject to one anomalous exceptionFootnote 127 ). This seems right: the vital right of an owner to follow its property into whoever's hands it may come should not be taken away by presumption,Footnote 128 from which it is arguable that the initial burden should be on the recipient to prove that it has taken honestly. On the other hand, it is suggested that if honesty is shown by R, it should then be up to O to show that some other bar applies, such as that O was involuntarily dispossessed or that R had knowledge of some fact indicating irregularity: any other result would be in effect an unfair demand on R to prove a wide-ranging negative.
C. Direct and Indirect Entrusting
A hidden problem in the English treatment of unauthorised dispositions is that discussion generally takes the easy way out and assumes that only three parties are involved: O dealing with P, and P with R. This assumption, which spills over into legislation,Footnote 129 can cause problems where, as may well happen, the chain is longer. Imagine, for instance, that P is a buyer in possession from O; P sells to X, who takes in bad faith; R then buys in good faith from X. Any effect of s. 25(1) of the 1979 Act is exhausted by the sale to X: it follows that R loses out, despite being a buyer in good faith in competition with a seller who quite voluntarily delivered the goods before ownership had passed.Footnote 130 This limitation also has the converse result that the receiver R's interests trump only those of O and no one else. If a third party Y itself entrusted the goods to O, its interests remain unaffected. This latter point matters commercially. It means, for instance, that if Y sells to O and O to P, both sales being on reservation of title terms with no payment made,Footnote 131 an innocent buyer R who pays cash to P takes free of O's interest but not Y's.Footnote 132 And so too where shipping documents are returned by a bank Y to an importer O under a trust receipt,Footnote 133 delivered but not transferred to a buyer P and sold by P to R. In this case too, and for the same reason, R can defeat O's rights but not Y's.
These results are unacceptable. In so far as a credulous owner O deserves to bear the risk of misdealing when pitted against a good-faith buyer R, it should not matter how many possessors intervene or who precisely was guilty of misappropriation: all that should matter is that the owner is not deemed worthy of protection and the buyer is. Put conversely, a person in the position of R should be guaranteed, not simply protection from P's immediate predecessor but clear title, subject only to the exclusion of loss or theft.Footnote 134 It is therefore submitted that any entrusting rule should protect the good-faith receiver R in all cases, the only exception being where the original owner O was dispossessed against its will. This, it is worth noting, is the result in FranceFootnote 135 and in Germany.Footnote 136
D. A Need for Delivery?
For an entrusting rule to apply, there is no doubt that O must lose possession to P, since otherwise there can be no entrustment at all. But should it be equally necessary that R gain it?Footnote 137 The present English approach is ambivalent. To succeed under ss. 24 and 25 of the Sale of Goods Act 1979 (sellers and buyers in possession) R must take delivery: elsewhere, by contrast – estoppel, the Factors Act and voidable title – it need not.
The difference is difficult to justify, and it is suggested that the latter is the better solution. It is difficult to see why the presence or absence of delivery should make any difference to the equities as between O and R,Footnote 138 assuming the transaction between P and R is capable even without it of creating a proprietary interest in the latter.Footnote 139 The whole doctrine of entrusting depends on possession by P, and on the impression of entitlement created by O in letting P have possession. But why should anyone care whether R takes possession? What matters is reliance by R, for example by paying money: delivery is beside the point.Footnote 140 Of course, if R fails to take delivery it may itself lose its rights to some other third party: but whether R chooses to take that risk is R's business.
Two further points reinforce this conclusion. One is that in large numbers of contemporary sales buyers do not necessarily take possession at all: R might, for example, have bought the goods in P's possession with a view to resale to a third party which would then take delivery direct from P. There is no reason to penalise R in such circumstances. Another point is that, even where there is a requirement of delivery under the present law, it can be satisfied by a pretty meaningless rigmarole.Footnote 141 Furthermore, the legal nature of delivery can give rise to distinctions of striking complexity,Footnote 142 which do no credit to anyone; suppressing the requirement of delivery at a stroke removes this sorry complication entirely from this area of the law of property.
E. Must R Take for Value?
The English law of nemo dat regards it as obvious beyond argument that a gratuitous transferee can never be protected as an innocent receiver. Indeed, even though two provisions of the Sale of Goods Act 1979, ss. 24 and 25, and also s. 2 of the Factors Act 1889, make no mention whatever of any need for R to give value, they are said on high authority to require it implicitly.Footnote 143
Strictly speaking, there is no necessity about this. There would be nothing incoherent were we to say that an owner entrusting goods to a possessor took the risk of unauthorised gifts as well as alienations for valueFootnote 144 : and indeed while in Germany the gratuitous transferee R essentially loses out to the original owner,Footnote 145 some civil law systems do protect donees in the same way as other alienees.Footnote 146 Nevertheless, on balance it is submitted that the English view is the sounder on principle. With a gratuitous transferee who cannot point to any element of detrimental reliance, the equities seem pretty strongly weighted in favour of the original owner.Footnote 147
On the other hand, the question of gratuitous transfer cannot necessarily be dismissed as simply as this. Although an innocent donee R ought not to become owner of the thing at the expense of O, it ought nevertheless to be protected in another way, namely by being insulated from liability beyond an obligation to return the goods. If at the time of any demand it has in good faith disposed of them (or lost them), or the goods have deteriorated, its liability should be limited accordingly. After all, one can hardly blame a possessor for alienating, or failing to take care of, goods it believes with good reason to be its own. The point matters in England, because if the common-law position were left untouched, the law of conversion would in most casesFootnote 148 make R liable for the full value of the goods at the time it got them, even if when demanded back they had been devalued or lost; furthermore, if R had disposed of them, however innocently, it would be fixed with a liability for their value at the time of disposal.Footnote 149 Any reform would thus have to provide for this (which, incidentally, is the solution of German lawFootnote 150 ).
F. Proprietary Interests Other Than Ownership
Discussions of nemo dat and unlawful dispositions regularly assume that the problem is essentially one of deciding about ownership. In fact the issue is much more nuanced. O might be not an owner but a pledgee, lienholder or mortgagee. Conversely P may have purported to grant R something less than ownership; not only a pledge (a possibility at least recognised in the Factors Act 1889 and in ss. 24 and 25 of the 1979 Act), but, for example, a lease, a charge or a possessory lien. It is suggested that in any rational overall scheme, there must be a common rule for all proprietary interests, whether we are talking about the interest which O stands to lose, or that which R stands to gain.
As regards O's interests it is not difficult to see why this must be. If an entrustment by O to P can justifiably cause O to be stripped of full ownership, a fortiori there can be no objection to its defeating some lesser interest in O. This point is accepted as obvious by many civil lawyersFootnote 151 ; moreover, there is evidence of at least a dim appreciation of it in England, where courts have on occasion strained to interpret some legislative reference to an “owner” in a nemo dat context as referring equally to someone in the position of O but with a lesser interest.Footnote 152
Moreover, as a matter of principle a similar a fortiori argument ought to apply to R. If we are happy in an entrustment case to grant R absolute ownership and thus to eclipse O's interest entirely, there is no good reason not to do the same where the transaction between P and R creates some lesser interest like a lease or a pledge, which merely burdens O's right with some lesser interest of R's. This is indeed partly recognised in many European jurisdictions by the extension of protection ad hoc to such interests, generously in GermanyFootnote 153 though less so in France.Footnote 154 It is also grudgingly admitted in England, in that a few of the exceptions to nemo dat protect pledgees and analogous receivers.
All this is, of course, subject to a major constraint: we are talking only of proprietary interests, as against mere personal claims vested in R.Footnote 155 But what ought to count as a proprietary interest? In the present English context, unless pre-empted by statuteFootnote 156 R's protection seems to embrace any traditional legal or equitable interest, including that of an equitable chargee or lienholder.Footnote 157 But what about possessory interests? Imagine R takes a lease of goods, or buys them subject to reservation of title: does this yield a mere contractual right against P, or a property interest? English lawyers instinctively say the former, on the basis that a lessee is never, and a buyer under a mere conditional agreement to sell not yet, an owner.Footnote 158 But, at least where R is in possession, this seems perverse. Possession carries within it its own (proprietary) rights. Thus title to sue for conversion inheres in a lessee,Footnote 159 or a purchaser in possession subject to reservation of titleFootnote 160 ; both too, it seems, can also cite their possession to resist claims to surrender of the goods , whether by the contractual counterpartyFootnote 161 or anyone else.Footnote 162 If so, it is suggested that such rights should be regarded as possessing sufficiently proprietary characteristics to be brought within the protection of any legislation so as to be exercisable against O in addition.Footnote 163
G. Protection for Other Parties
The main thrust of this article concerns the protection of R and the proprietary interest transferred to it. But there is one case where it needs to go further and provide some sort of a shield for third parties. Imagine that P, entrusted by O with goods, dishonestly sells them to an innocent buyer R using the services of X, a dealing agent or auctioneer; and assume further that X acts without fault. The problem is that even if R is protected, without more the intermediary X, who claims no proprietary interest, is not.Footnote 164 Instead X would face strict liability to O in the tort of conversion,Footnote 165 at least in so far it exercised any physical control over the goods in the course of arranging the sale.Footnote 166 This is perverse: in so far as a receiver R is protected from liability to O, the same must go a fortiori for those whose part in the transaction is merely ancillary. More formally, therefore, it is suggested that in any entrustment scheme it would need to be provided that those acting innocently on the instructions of either P or R in a transaction where R's interest would be protected should themselves receive a statutory shield against liability in conversion.
H. The Effect of an Entrusting Rule on Existing Statutory Exceptions
We suggest below that a general entrusting rule should supplement, and not replace, the existing protections available to a good-faith purchaser (see “The place of an entrustment regime in the scheme of things”). However, it is suggested that this should be subject to one qualification concerning the current s. 2 of the Factors Act 1889 and ss. 24 and 25 of the Sale of Goods Act 1979.Footnote 167 As was pointed out above, these provisions contain more than their fair share of anomaly, narrow distinctions and obscure draftsmanship. Furthermore, virtually all the situations they deal with would in any case be covered – and covered in a more extensive and logical way – by the scheme proposed here. There seems no reason to complicate matters by keeping these antique and partial provisions, and they should go.
One might say the same about s. 23 of the 1979 Act, dealing with voidable title. Theoretically no harm would be done by getting rid of it, since all it does is partially re-enact the common-law position for one particular special case,Footnote 168 and most instances will be covered by the new scheme anyway. On the other hand, in one matter it goes further than any entrustment rule: it does not require P to be in possession of the goods at all.Footnote 169 Thus if O sells to P under a voidable contract but remains in possession, and P then sells on to R, R's interests are protected. This result seems entirely unobjectionable, and since there is no reason to throw it into doubt we might as well preserve s. 23 as a harmless confirmation of it.
I. Exceptions
For simplicity's sake, we have argued the case for a general entrusting rule on a largely commercial basis. It must be recognised that with goods not traded or used commercially, there may be a case for requiring different treatment. There is something to be said, for example, for allowing non-commercial purchasers of consumer goods to succeed in the absence of actual knowledge that the disposition to them is unauthorised (as is presently the case with s. 27 of the Hire Purchase Act 1964Footnote 170 ). There might even be an argument in favour of an analogous protection for owners of such goods, preventing any buyer acting in the course of a business from invoking the entrustment rule as against them.Footnote 171 And there may be other examples of goods that exceptionally ought to be protected. An example might be artefacts on loan for display, on the basis that the social interest in preserving their public availability requires some of the normal proprietary rules to be qualified.Footnote 172
J. The Place of an Entrustment Regime in the Scheme of Things
As mentioned above, what is being suggested here is not a universal but a fall-back principle. No rule can accommodate all participants or events; nor for that matter should it try to, especially when it comes to detailed regimes based on statutory registers of security interests, or – even more importantly – systems based on international conventions. Indeed, this residuary feature has been a feature of civil law systems that recognise an entrustment principle, which have never had much difficulty with carving out exceptions to accommodate specific regimes.Footnote 173 So too any replacement English system would take effect as a default rule, subject to exceptions.
What actual or possible exceptions are we talking about?
First, there are principles relevant to unauthorised dispositions arising out of the general law (as opposed to legislation): most obviously the rule protecting the good-faith purchaser of goods subject to any equitable interest, and the principles of agency and the rules of estoppel. They are well-established, and it would be a recipe for confusion to try to cut them back; instead it should be accepted that, while they will be partly overlaid by any entrusting rule, they may well continue to have a field of application outside it.
Second, specific statutory schemes providing for the adjustment of existing ownership or security rights. Examples of these are the rules relating to securities over chattels registrable under Part 25 of the Companies Act 2006, those covering interests in aircraft under the Cape Town Convention,Footnote 174 and more recondite matters such as the Agricultural Credits Act 1928.Footnote 175 These schemes are concerned with security interests of one sort or another, such as the rules on disposals by bankrupts,Footnote 176 or the rules of Part II of the Merchant Shipping Act 1995 on transfer of title to registered ships.Footnote 177 Moreover, this list may well grow, if (for example) it is thought desirable to add a workable scheme of secured lending to replace the Bills of Sale Acts,Footnote 178 to make title retention agreements or finance leases registrable securities,Footnote 179 to replace the present privileged status of possessory security in commercial contexts with a requirement of registration,Footnote 180 or for that matter to create an all-embracing registration scheme on the lines of Article 9 of the UCC.Footnote 181
VII. Conclusion
The import of this paper can be briefly summarised. It is difficult to deny that the current position as to nemo dat in England is an arbitrary and unpredictable mess. No one, given a clean sheet of paper and a brief to design a new system, would come up with the one we have. It follows that we need a substantial recasting, with a view to producing a default rule on unauthorised dispositions of chattels that is easy to understand, rational and logical. Such a rule should follow three principles:
-
(1) The background rule should be one of entrustment, under which a proprietor putting or leaving another in possession of goods prima facie takes the risk of subsequent misdealing.
-
(2) The principle should be universal and not limited to ownership. It should be capable, where it applies, of defeating or protecting any proprietary interest.
-
(3) However, a general principle of this kind should emphatically be only a prima facie rule. It should be open to exceptions where there is good reason to admit them, for example where it is necessary to have a specific scheme covering particular types of security interest, or where particular actors are regarded as in need of special protection. What is necessary is a simple and workable underlying scheme that meshes as sweetly as possible with exceptions of this kind.
