In this post, Mark Summers KC and James Stansfeld of Matrix Chambers comment on the Supreme Court’s decision in Public Prosecutors Office of the Athens Court of Appeal v O’Connor (Northern Ireland) [2022] UKSC 4.


The problems associated with appellate time limits under the Extradition Act are long-standing and notorious. As part of the then Government’s drive to rid the system of delays that had previously characterised extradition, the 2003 Act introduced rigid, and peremptory, time periods in which appeals were required to be lodged.

The problems with such a system were obvious and immediately felt. Unrepresented defendants in custody, often without access to even a fax machine, were the prime, but by no means only, group of persons dealt serious and predictable injustice by this inflexible regime.

What immediately followed was, ironically, an equally predictable stream of additional High Court litigation testing the limits of the statutory regime; what constituted a valid ‘notice’ of appeal (would a mere letter do?); what ‘grounds’ did the notice require (would a letter which contained none do?); what did ‘giving’ notice mean (did it include service, and if so on which of the many extradition actors?); did the CPR service deeming provisions apply (could notice be given outside of court hours?); could an irregular notice be cured (what if only the front page of a notice had been faxed by the prison?). Etc, etc. The Court’s answers to most of these questions were invariably coloured heavily by a desire to alleviate the injustices caused by an unnecessarily harsh statutory scheme. Liberty is, after all, at stake. But difficult facts and a desire to bend normal principles to accommodate them, make for, at best, inconsistent and superficial decisions.

Matters came to a head in the Supreme Court in 2012 in Pomiechowski v Poland [2012] 1 WLR 1604, when this Court decried the unfairness of the system and strove to find a more sustainable way around it. For some of the appellants in that case, an expansive and ‘generous’ interpretation of an appeal ‘notice’ was alone sufficient to remove injustice. For one, however, not even that device could work (at least not without reversing prior Supreme Court authority holding that ‘giving’ notice means serving as well as lodging: Mucelli v Albania [2009] 1 WLR 276 ). The solution devised in Pomiechowski? To venture article 6 ECHR and the right of access to justice into pastures it was never intended to reach. ECtHR case law had consistently refused to apply article 6 to the substantive conduct of extradition proceedings. This Court however, imaginatively reasoned that, because UK citizens possessed a common law right to enter and remain within the UK and the extradition proceedings under the 2003 Act could affect that right, such extradition proceedings fell within article 6 because such a defendant was entitled to a fair determination of his common law right to remain within the jurisdiction. Reasoning generally then that the absolute and inflexible time limits for appeal in the 2003 Act did not meet the required article 6 standard of access to justice, the HRA served therefore, in the case of a UK citizen at least, to require the statutory provisions to be read as being subject to judicial discretion to extend time.

The limits to what this Court could achieve this way were plain. The new implied discretion could not apply to non-UK citizens, nor foreign states. For the former at least, the system remained ‘discriminatory’ and ‘unsatisfactory’.

Parliament eventually answered this Court’s call to act. The Anti-social Behaviour, Crime and Policing Act 2014 did so by inserting into the various defence appellate provisions of the 2003 Act (ss.26(5), 103(10), 108(7A)) a discretion to extend time ‘if the person did everything reasonably possible to ensure that the notice was given as soon as it could be given’.

This was and is no general discretion to extend time where it is in the interests of justice to do so, nor is it one which enables consideration of the underlying merits of a putative appeal. The sole focus of the power is instead on the conduct of ‘the person’. But the power nonetheless worked reasonably well to remove the injustices caused by the system and matters quietened down. At least they did until Szegfu v Hungary [2016] 1 WLR 322; in which the High Court afforded a broad interpretation to the meaning of ‘the person’ and applied to the Act the ‘surrogacy’ principle, by which the faults of lawyers are attributed to the party personally. In this context, that meant that a defendant (in custody etc.) who had done all s/he could do to initiate an appeal (by instructing lawyers etc.) was shut out (and surrendered abroad) because their instructed lawyer had failed to act with appropriate diligence. Szegfu undoubtedly marked a sharp change of judicial direction. Previously striving to bend an unfair system towards fairness, the High Court apparently now regarded the system as sufficiently fair, and indulgence to defendants no longer needed.  According to the High Court, the sole mischief to which the new discretion was aimed was unrepresented defendants in custody; judicial latitude in its interpretation was no longer required for defendants (such as Mr Szegfu) who were represented.

The High Court in Northern Ireland promptly disagreed (O’Connor v Greece [2017] NIQB 77) and the matter came before the Supreme Court.

On 2 February 2022, the Supreme Court gave judgment in Public Prosecutors Office of the Athens Court of Appeal v O’Connor (Northern Ireland) [2022] UKSC 4; [2022] 1 WLR 903. The Court unanimously held that Szegfu was wrong.

The decision is neither controversial nor surprising. The mischief to which the new discretion was aimed was assuredly not confined to unrepresented defendants in custody. Rather, some of the key injustices which had led to the introduction of the discretion had involved failures by lawyers (see, e.g. R (Mann) v Westminster Magistrates’ Court [2010] EWHC 48 (Admin)). Attributing their faults to their clients was neither intended by the scheme introduced by the 2014 amendments, nor remotely warranted in principle. Especially where liberty of the individual is at stake. Recall, for example, that Mr O’Connor himself was (once his appeal was admitted) discharged on Article 3 ECHR grounds as a result of the treatment that awaited him in the notorious Greek prison system.

Context is everything. It was never likely that, given the pivotal role it had played in its creation, the Supreme Court was realistically going to accept neutering of its hard-won judicial discretion in this way. The ‘surrogacy’ principle is not, after all, of universal application. It was doubly unlikely to do so in an appeal carrying a proven Article 3 violation. In fact, the Supreme Court had only given Greece permission to appeal at all on condition that it undertook not to arrest Mr O’Connor in the UK even if it won the appeal.

One is left wondering why the Szegfu neutering attempt ever happened in the first place? No doubt the lack of underlying merit in Mr Szegfu’s appeal did not incline the High Court to entertain his out-of-time appeal. Facts undoubtedly drive decisions. But it bears observing that the whole discussion in Szegfu was arid (obiter) in any event; the faults in late service were Mr Szegfu’s own, not his lawyers. It required no ‘surrogacy’ principle to exclude his attempt to appeal.