It is hard to think of a principle of criminal justice more fundamental than the presumption of innocence. To implement this presumption fully, we should require the prosecution to prove all the elements of the offence with which D has been charged and to prove that any defence or exception to the charge raised on the evidence is unfounded. At common law, the great case of Woolmington [1935] AC 462 endorses full observance of the presumption of innocence. However, there are many exceptions to the presumption created by statutory offences that place an onus of proof on D to disprove the presence of one or more elements of the offence, or to prove a particular defence or exception. This familiar practice is known as imposing a reverse burden of proof. Furthermore, where any excuse or exception is created within a statutory offence, it is generally open to the courts to construe that the burden of proving that excuse or exception falls on D.

Hopes were raised of a fuller observance of the presumption of innocence when, as a result of the Human Rights Act 1998, Article 6(2) of the European Convention became a source of English law. Article 6(2) provides that “Everyone charged with a criminal offence shall be presumed innocent until proved guilty according to law.” The first appellate cases to consider this article, Kebilene [2000] 2 AC 326 and Lambert [2001] UKHL 37, gave rise to our optimism [S&S2 p. 67] that the courts would, employing the interpretive latitude allowed by section 3 of the Human Rights Act 1998, interpret all reverse burdens of proof as mere evidential burdens, at least for offences with an identified culpability and severe sentences.

That optimism was confounded by the decision of the House of Lords in Johnstone [2003] UKHL 37. In the context of a serious trademark offence carrying a substantial term of imprisonment, their Lordships emphasised the need for due deference to the will of Parliament and cautioned against any ready finding that an imposition of a reverse burden was a disproportionate response by Parliament to the social mischief proscribed by the offence. Article 6(2) did not stand alone, observed the House of Lords: it was subsumed within the guarantee of fair trial provided by Article 6 as a whole. A reverse burden did not necessarily preclude a fair trial. It could, as was the case for the trademark offence considered in Johnstone, be a proportionate response to a serious and current social or commercial mischief.

The tension between the approaches in Lambert and Johnstone was palpable. In A-G’s Reference (No. 1 of 2004) [2004] EWCA Crim 1025, the Court of Appeal sought to resolve the uncertainty by favouring the approach taken in Johnstone. Indeed, trial judges and magistrates were instructed not to refer at all to Lambert but to confine their attention to the favoured case. But this would hardly do. The Court of Appeal cannot, by fiat, consign a recent decision of the House of Lords to limbo. Consequently, the decision of the House of Lords in Sheldrake v. DPP; A-G’s Reference (No. 4 of 2002) [2004] UKHL 43 was eagerly awaited. The Lords, at least, did have the authority to lay done a clear pathway for the future.

Unfortunately, clarity has not been achieved by their decision. But the good things first. Lambert is not to be disregarded. It remains important as authority that undue deference must not be paid to the fact that Parliament has found fit to impose a reverse burden. The overarching question is whether the reverse burden is compatible with a fair trial; if it is not, it should be read down as a mere evidential burden by using section 3 of the Human Rights Act 1998. However, the good news runs out at this point, at least for persons who would like to see large effect given to the presumption of innocence. Sheldrake considers that reverse burdens do not necessarily preclude the holding of a fair trial. Whether or not the reverse burden at issue offends against Article 6 requires a proportionality assessment. In conducting that exercise, one should balance, on the one hand, society’s interest in the effective suppression of the social mischief with which the offence is concerned and, on the other hand, D’s right to a fair trial. When balancing these two competing interests, one should take into account the severity of the offence in terms of sentence, ease of proof for one party or the other in relation to the matter covered by the reverse burden, and (from previous authority) whether the matter to be proved or disproved is related to a definitional element of the offence or to a defence. In the light of these factors, the final judgement must be made on whether the reverse burden in question is a fair and proportionate legislative response, in the circumstances of contemporary society, to the social mischief with which the offence is concerned.

The problem with this approach is that, as Professor Andrew Ashworth has remarked ([2005] Crim LR at 219), “it furnishes courts with no clear guidance on how to interpret statutes that impose a burden of proof on the defendant.” Rather like the famous boast of Justice Holmes, armed with the desiderata for decision-making provided in Sheldrake, any given offence could plausibly be interpreted either way. The Lords themselves reviewed much of the previous reverse-burden case law. All were found to be in line with the appropriate criteria, with the exception only of Carass [2002] 1 WLR 1714, a case where the Court of Appeal had disallowed a reverse burden later found to be fair and proportionate by the House of Lords in Johnstone. It is submitted that a balancing exercise of the kind required by the House of Lords in Sheldrake is equally compatible with endorsing the particular reverse burden at issue in Johnstone or with rejecting the same provision, as was done in Carass. It is further submitted that the same is true for all the other cases considered by the House of Lords in Sheldrake, irrespective whether reverse burdens were actually upheld or rejected. It is worth observing that, in the main, reverse burdens have been upheld.

Is there a better way forward? There would be if primacy were given to the line taken by the House of Lords in Lambert. It will be remembered that there was, in that case, a distinct hostility to reverse burdens or, more positively, a bias in favour of the presumption of innocence. This could still be built on, at least for cases where D may be sent to prison if he fails to discharge a burden of proof. But for the moment the approach endorsed in Sheldrake holds sway, together with the prospect of further appeals on whether a particular reverse burden is compatible with a fair trial.