TC09657 - [2025] UKFTT 01208 (TC)

Our starting point, of course, must be section 16(6) FA 1994, which very clearly puts the burden of proof on the appellant; it states unequivocally that, subject to certain exceptions which are not in point here, it is for the appellant to show that the grounds on which its appeal has been brought have been established. That approach was, not surprisingly, applied by the Upper Tribunal in Build-A-Bear.

Ms McArdle says that the Appellants need to prove that no criminal act was committed in connection with the customs debt incurred when the bikes were imported to the UK to show that the extended time limit is not available to HMRC and so make out their case that HMRC‘s calculations are wrong. This (she says) is the natural result of the general principle that the person who asserts a case must prove it, and so the burden of proof rests on the Appellants even without section 16(6) FA 1994.

As far as general principles are concerned, I am not with Ms McArdle. What has happened here is that HMRC have communicated a customs debt after the normal three-year period, and the Appellants are saying that HMRC have not shown that there was a lawful basis for their action. HMRC are relying on Article 221(4) to authorise what they have done. The essential part of the Appellants’ case is that more than three years have elapsed since the customs debt was incurred. They do not need to show that a criminal act has been committed, whereas it is an essential part of HMRC’s case that a criminal act has been committed. They are the party who is asserting the affirmative of the issue here, that a criminal act was committed and so their act was authorised under Article 221(4), and on general principles they should bear the burden of proving that. The position on general principles is not particularly important, however, because section 16(6) FA 1994 applies here and, as we have seen, that provides that the Appellants must show that the grounds on which they brought the Appeals are satisfied.

In principle, the UK is entitled under Article 245 of the Code to adopt its own procedural rules for customs duty appeals. Unitrading confirms this, but we can also see from Unitrading at [27] that the latitude Article 245 allocates to Member States is limited by the requirements that the rules they adopt (i) are not less favourable than those governing similar domestic actions (equivalence) and (ii) do not make the exercise of rights conferred by Community law impossible or excessively difficult in practice (effectiveness).

 In addition, the Code is an EU regulation and as such has general application and is binding in its entirety and directly applicable in all Member States; Article 288 of the Treaty on the Functioning of the European Union. A regulation being directly applicable means that it applies immediately as the norm in all Member States, without needing to be transposed into national law, and can create rights and obligations for individuals, who can invoke the regulation directly before national courts and use it as a reference in their relationship with other individuals, Member States or EU authorities.

Pausing here, we can see that in principle the UK has freedom under Article 245 to adopt its own procedural rules (of which section 16(6) FA 1994 is clearly one) in customs duty cases, but those rules must comply with the requirements of effectiveness and equivalence and will automatically yield if they conflict with the Code.

Like Ms McArdle, I have not found the three ECJ/CJEU authorities we looked at in detail to offer explicit authority for a general rule on the burden of proof in customs duty cases. Such general observations as the Advocate General or Court make are, however, consistent with the idea that, where a tax authority wants to base an assertion of liability on a particular fact or provision, it is for the authority to make out its case.

So, we see in Unitrading the Czech government pointing out that, if the customs authorities assert that the statement by the declarant of the country of origin does not correspond to reality, they must bear the burden of proving that. Beyond that, I agree with Miss McArdle that Unitrading is addressing a different point, how to secure a fair and effective trial where the way in which an important piece of evidence is provided makes it difficult to review or challenge it. Beyond the Czech government’s submission, however, I cannot find anything in this case to give support to the idea that there is a general burden of proof on tax authorities on all points they assert or any real debate around the burden of proof at all.

I find Greencarrier to be more helpful. The question in that case was the extent to which customs authorities could sample and then extrapolate from their sample. The court was happy that they could do that if the goods in question were identical with those sampled. We have a broadly similar situation here. HMRC say they have evidence that, where bikes were consigned by City Cycle, an incorrect declaration of origin was given (which gives rise to the criminal act which justifies late notification). The question is can they extrapolate what they say they have learned (that the City Cycle declarations of origin are wrong) to other bikes, where the consignor’s identity is unknown? Based on Greencarrier, they can only extrapolate from City Cycle to other consignments if they are identical. At [31] the Court discussed what identical meant and considered that it might include goods coming from the same manufacturer. This case does not, of course, tell us in terms where the burden of proof lies, rather it tells us when a party which is seeking to establish something can extrapolate from the evidence they already have. Of more general relevance to us is the Advocate General’s ready acceptance “that there is not a shadow of a doubt that the burden of proof of the identity of those goods, a question also debated before the Court, rests on the party which seeks to rely on that identity for the purpose of revision of the declarations, namely, the customs authorities”.

Beemsterboer deals with the detailed provisions in Article 220(2)(b). We are not concerned with these, and we need to remember that this provision is very detailed, nuanced was Ms McArdle’s description, and very different in structure from the provision we are concerned with. We do, however, see the Advocate General making a very general statement (at [47]) that “Under the generally accepted rules of procedural law, the party relying on the conditions laid down in a provision must as a general rule prove that they have been satisfied.” He does not express any doubt about that being a rule of general application or seemingly entertain any doubt about it being the starting point in the Code. We see that approach being followed in paragraphs [48]-[51], but not slavishly; it is departed from where there are good policy reasons for doing so. He takes the same approach at [54]-[57].

We see the Court observing (at [39]) that “… in accordance with generally accepted rules on the allocation of the burden of proof, it is the responsibility of the customs authorities which wish to rely on [a particular provision] to carry out post-clearance recovery to adduce, in support of their claim, evidence that the issue of incorrect certificates was due to an inaccurate account of the facts provided by the exporter.” Although this case is concerned with the detailed provisions of Article 220(2)(b), both the Advocate General and the Court acknowledge the generally accepted rule that a person who wishes to rely on a provision must prove their entitlement. That is not the inevitable end point at every turn of Article 220(2)(b), but it is the unquestioned starting point.

If we turn to the language of the provision we are concerned with, Article 221(4), we see that it provides an exception to the general rule (that communication of a custom debt must take place within three years of the date on which the debt was incurred) where the customs debt is the result of a criminal act. Read naturally, it restricts communication of a customs debt after three years to circumstances where the debt is the result of a criminal act.

If we accept HMRC‘s assertion about where the burden of proof lies, however, it changes the way Article 221(4) works quite radically. The practical effect of holding that the burden of proof lies on the debtor would be that the only limitation on late communication would be if the debtor could show that the debt did not result from a criminal act. Effectively, what (read naturally) is a precondition for late communication (that the customs debt results from a criminal act) would have turned into a limited defence.

Whilst Article 245 gives Members freedom to introduce their own procedures for dealing with appeals, they should not affect the fundamental rights and obligations the Code, as a directly applicable EU regulation, gives rise to. Allowing section 16(6) FA 1994 to shift the burden of proof in the way HMRC suggest would turn the way Article 221(4) works on its head, from what appears to be its straightforward reading (that there can only be late notification if the customs debt results from a criminal act) to a strained, unnatural reading (that there can always be late notification unless the debtor can show that the customs debt did not result from a criminal act).

I was not addressed on how this issue is approached in other Member States or at EU level, and of course the procedure for referring this type of issue to the CJEU for a preliminary ruling is no longer available. Whilst I cannot be sure that Article 221(4) does not operate across the EU in the way HMRC claim for it in the UK, I am wary of adopting an analysis which appears to allow a domestic procedural rule to have such a radical and unexpected effect on the natural reading of the Article, particularly when the natural reading of the Article and what seems to be the generally accepted approach to how the burden of proof operates in scenarios like this (see [137] below) point in the same, but entirely opposite, direction.

As well as completely reframing the provision, putting the burden of proof on the debtor would put a high burden on the debtor (to prove that the debt did not result from any criminal act at all). It would require the debtor to scour the criminal code of the relevant jurisdictions (certainly the jurisdiction of import and possibly the jurisdiction of export too) and try to find out whether it, or anyone else involved in the importation, might have committed a crime which impacted on the customs debt and only if they could prove a complete negative could they resist late communication of the customs debt. In that connection, we should bear in mind that the criminal act HMRC rely on here is a strict liability offence. For obvious reasons, it may be difficult for a debtor to identify every criminal act (particularly those where there is no mens rea requirement) which might have been committed and then establish that it was not.

I was not addressed in detail on the operation of the principle of effectiveness, but it seems to me that placing the burden of proof on the debtor and taking the burden seriously (putting the debtor to proof of the total negative, which it seems to me we must do – we cannot proceed on the basis that the burden of proof is on the debtor but that this requirement will generally not be taken very seriously) must come very close to breaching, if it does not breach, the principle of effectiveness.

Ms McArdle says that determining whether it is impossible or excessively difficult to protect one’s rights in an appeal is a question involving consideration of evidence and there is no evidence before me on this point. In the absence of persuasive evidence, there can be no finding that it is excessively difficult or impossible for the Appellants to argue their position on appeal.

I do not consider that evidence is needed to establish that it is harder to prove a negative than a positive. The authors of Phipson on Evidence (20^th ed at 6-06) explain that the reason why the burden of proof rests on the person who asserts the affirmative of the issue is because, in the nature of things, a negative is more difficult to establish than an affirmative. Putting the burden of proof on the debtor would be a very onerous and potentially very expensive thing to do, which could very likely result in debtors not challenging, or unjustifiably failing to challenge successfully, the late notification of customs debts which do not result from any criminal act at all, which is the exact opposite of the natural reading of Article 221(4), that late notification should only take place in circumstances where the customs debt is the result of a criminal act.

Ms McArdle also says that the Appellants are large and sophisticated businesses profiting from importation, and so they ought to know that the possibility of being liable for debts arising over three years ago is a very real one, and so they should retain evidence beyond the three-year time limit to be able to support their position if subsequently challenged. She also says that the Appellants have produced a lot of evidence (not before me in this preliminary issues hearing) on the question of whether “the customs debt the result of an act which, at the time it was committed, was liable to give rise to criminal court proceedings” and this necessarily undermines any argument of procedural unfairness if the burden of proof is applied following section16(6) FA 1994. Finally, she says that the Appellants had plenty of time to make any arguments or produce any evidence before and during an extensive Right to be Heard Process, but no dispute as to the consignments in question was raised by them until 2023.

I do not agree that the position of a particular debtor in a particular case (its size/sophistication and whether it has been able to produce evidence) is relevant to the abstract question where the burden of proof lies generally. Similarly, I cannot see why (or how) the Appellants not raising this issue during the Right to be Heard Process affects where the burden of proof lies on the issue once it has been raised.

The Code imposes a minimum period of 3 years for the retention of records, and this dovetails neatly with the basic period for notifying customs debts in Article 221. In the UK the retention period is 4 years (see regulation 9 of the Customs Traders (Accounts and Records) Regulations 1995), but I do not consider that anything turns on this when we are looking at the way the Code functions. There is an obvious symmetry between the minimum period for preserving records and the ordinary three-year period for notifying customs debts. Once that period has expired, as a matter of EU law, there is no expectation that a trader has retained any records. It would seem odd (to put it mildly) against such a background for customs authorities to be able to notify a debt to a trader without any evidential basis for their action and for the trader, who by then is entirely free to dispose of their records, to bear the burden of proving that no debt had arisen. I do not think that it is controversial to say that legal certainty is an important principle of EU law. I cannot see how any ability on the part of customs authorities to act in the way I have just described could comply with the requirement of legal certainty.

Turning to the concept of equivalence, although in UK direct tax cases the burden of proof is generally on the taxpayer to displace an assessment, the burden of proof on the validity of assessments (issues such as whether there has been a discovery, whether the conditions that assessments may only be brought if the loss of tax was brought about carelessly or deliberately or there was insufficient disclosure, and the extended time limits for careless or deliberate behaviour) generally lies on HMRC; the position here is usefully summarised in Pump Court Tax Chambers: Tax Litigation Handbook at 12.10 et seq. I was not addressed in detail on all forms of UK tax, particularly indirect tax, litigation. Interestingly, although this position is not free from doubt, the burden of proof on the validity of VAT assessments (whether the assessment is made to best judgment and whether HMRC have failed to meet the test that allows assessments to be made up to a year after facts justifying the assessment come to their knowledge) tends to sit on the taxpayer; Tax Litigation Handbook at 4.49-4.52, 4.57-4.61 and 12.33-12.34. Given the doubt expressed in relation to the VAT position and that the authorities discourage focus on the “best judgment” (as opposed to quantum) aspects of an appeal against a VAT assessment, it seems to me that a fair (if high level) summary of the effective UK approach to the burden of proof on matters such as extended time limits and other factors which go to the validity of an assessment is that the burden of proof is generally on HMRC rather than the taxpayer.

We have seen in the three ECJ/CJEU authorities we have reviewed that there seems to be a ready acceptance that the ordinary approach to the burden of proof (put colloquially, If you want to rely on a provision, it’s your job to prove your entitlement to do so) should apply in the Code unless there is a good reason for it not to. Taking such an approach to the burden of proof here would also be consistent with the natural reading of Article 221(4) (that late notification is only allowed if the customs debt resulted from a criminal act) and its interaction with Article 16, and my initial conclusions on effectiveness and equivalence suggest that there is every reason to keep with that approach rather than reverse the burden of proof.

For these reasons, I consider that, where reliance needs to be placed on Article 221(4) before a customs debt can be notified, the legal burden of proof is on the customs authorities to establish that they are entitled to rely on it.

I am not sure that there really is any conflict between section 16(6) FA 1994, which can be construed (as Ms Sloane suggests) as saying no more than that, if it is a part of the Appellants’ case that the burden of proof on the extended time limit issue is on HMRC, it is for them to make good that argument. However, to the extent there is any conflict between Article 221(4) construed in the way I have just described and the way section 16(6) FA 1994 would otherwise operate, the directly applicable provisions of the Code must prevail. There was no suggestion in Build-A-Bear that section 16(6) FA 1994 conflicted with any relevant directly applicable provision of the Code, and I do not consider that my conclusion is in any way in conflict with (or not consistent with) the Upper Tribunal’s decision on the burden of proof in that case.