For the first time, by way of the Post Office (Horizon System) Offences Act (“the Act”), Parliament has legislated to quash criminal convictions. In this post I will argue that, no matter how understandable, indeed laudable, the intention behind this legislation, in its haste to offer a speedy and comprehensive correction to mass injustice, Parliament has crossed a constitutional line and ventured into territory that is properly the preserve of the courts. 

The Act has only one purpose: to quash the convictions of sub-postmasters and others who worked in post office branches and who have suffered as a consequence of the Post Office Horizon IT scandal. I won’t rehearse in detail the circumstances leading to the Act; they are well-known. Beginning in 1996, the Post Office piloted new Horizon software in a number of its branches. By 2000 this was extended across all branches. Due to faults within this software some postmasters’ accounts showed incorrect shortfalls. This led to investigations, and eventually to the suspension, dismissal, and in certain cases the prosecution, of postmasters. Hundreds of postmasters were imprisoned and a number suffered severe financial difficulties including bankruptcy and concomitant family, health and social problems. In total, there were potentially over 900 wrongful convictions across the United Kingdom. The issue is now the subject of a judicial inquiry. 

It has been argued robustly on this blog and elsewhere that the Act gives rise to no significant constitutional objections. In this post I challenge these arguments, well-made though they are. Various points have been made in defence of the Act: that no significant separation of powers issue arises for a constitution that does not strictly separate the branches of government; that one of Parliament’s roles is the redress of grievances; that Parliament is sovereign and can, therefore, legislate legitimately to overturn convictions; and that the circumstances leading to the Act are so unique that it will not become a precedent for future legislation. I take issue with each of these arguments and raise further constitutional objections to the Act. 

It is of course true that the United Kingdom constitution does not embody a formal separation of powers, and that in fact, through the concept of the Crown, the intertwining of executive and legislative roles characterises the very essence of our parliamentary system. But it is mistaken to assume that the inextricable relationship between executive and legislature, embodied in the sharing of functions, extends to the judicial branch. While the courts have historical and symbolic connections to both the executive and Parliament, and while they remain accountable to some extent to the political branches, the independence of the judiciary is a core constitutional principle, guaranteed by section 1 and section 3 of the Constitutional Reform Act 2005 and increasingly recognised through, for example, an independent judicial appointments process. This independence is vital in maintaining the principle which Dicey identified as the necessary counterpoint to Parliament’s legislative supremacy: the rule of law. This principle embodies at least two component elements: the exclusive jurisdiction of the common law courts, and equality before the law. The Act, through which Parliament for the first-time overturns criminal convictions, surely impinges upon the former; later in this post I will argue that it also has potentially problematic implications for the latter.

In historical relief Parliament has played a role in the redress of grievances, but the judicial role of Parliament is in essence now a thing of the past. The growth of the number and range of courts and tribunals, and of their jurisdictional reach, has in effect long removed any real role for Parliament as adjudicator. The final, symbolic transition of the Appellate Committee of the House of Lords to the Supreme Court completed what is to all intents and purposes a separation of the legislative and judicial roles within our constitution. To construe the Act therefore as a revival of this ancient parliamentary role is quite a stretch, and at odds with the growth and consolidation over centuries of the discrete judicial role.

It is said that Parliament is sovereign, and so can pass this Act. As a matter of lawful competence that is certainly true. However, it is not true that Parliament can, with constitutional propriety, assert its legislative supremacy in a way that undermines other constitutional values, in particular the rule of law. While it is impossible under our constitution for Parliament to behave unlawfully since its lawful power is unlimited, it can behave unconstitutionally. The House of Lords Constitution Committee has in the recent past declared law that is inconsistent with the rule of law to be unconstitutional:  see, for example, Constitution Committee, United Kingdom Internal Market Act, HL Paper 151, 17^th report, sess. 2019-21, 16 October 2020 , para 229. Legislating to interfere, in an unprecedented way, with judicial independence is arguably unconstitutional.

It is also argued that the Act is not in itself a precedent because similar legislation has been passed before. Measures which are pointed to as potentially analogous to the Act, such as the War Damage Act 1965, are in fact different. Legislation of this kind –  another recent example is the Litigation Funding Agreements (Enforceability) Bill  – serve retrospectively to change the effect of court rulings, but do not nullify these rulings per se. This Act however does not just change the effects of court decisions, it overturns them. The Act intervenes directly in the judicial process by reversing a decision made by a court: for the first time overturning criminal convictions.

The Government declares that it is not its intention in the Act to create a precedent. With respect, whatever its intention, the Act is a precedent for the legislative over-ride of decisions by the criminal courts. It is also impossible to predict what longer term impact this precedent might have. The Government had mooted the possibility of a ‘purpose section’, analogous to section 1 of the Safety of Rwanda (Asylum and Immigration) Act 2024; seeking to declare the intention not to legislate in this way again. Such a provision would of course not bind a future Parliament; and in any case no purpose section was added to the bill. The Act therefore has the potential to become a benchmark for any comparable situations in the future. 

The Act gives rise to other constitutional issues. One is the potential inconsistency with which it will apply to convictions. The category of convictions which will be quashed is very broadly defined in section 2, potentially encompassing convictions within the relevant period which are not a consequence of errors relating to the Horizon software. This over-inclusion might be thought unavoidable, and that it is better that rightfully convicted people are acquitted than that the wrongfully convicted remain so, but there are problematic anomalies. The Act does not overturn convictions which were considered by the Court of Appeal before the Act took effect. Therefore, those who have not appealed and who would have been unsuccessful were they to do so, now have their convictions overturned by the Act; whereas those whose have appealed unsuccessfully have not. To give an example: those convictions which were referred to the Court of Appeal by the Criminal Cases Review Commission but not overturned are not quashed; those which were refused referral by the CCRC are however quashed. I noted above that equality before the law is a central plank of the rule of law principle. The Act as it stands could have quashed the convictions of the guilty, and possibly, in a small number of cases, not quashed the convictions of the wrongly convicted, leading to markedly unequal outcomes.

The breadth of section 2, and the lack of clarity in its application, gives rise to a further separation of powers issue. The Act gives discretion to the executive. Section 4 vests a power in the Secretary of State to identify the convictions to which section 1 should apply. Section 4(1) places a duty upon the Secretary of State to take all reasonable steps to identify the convictions to which section 1 applies and which are therefore quashed by the provisions in section 1. Under section 4(5): “For the purpose of identifying convictions quashed by section 1(1), the Secretary of State must, in particular, consider any representations made to the Secretary of State which claim that a person has been convicted in England and Wales of a relevant offence, whether or not made by that person.” Similar discretion is vested in the Department of Justice for Northern Ireland. 

The Government argues that the criteria for defining relevant convictions are so unambiguous and “capable of being applied without any element of judgment or discretion” that the minister’s role here is straightforward. This is by no means clear. It may well be that, in the vast majority of cases, there is no discretionary role for the minister since that role is to identify quashed convictions; it is not itself a power to quash them. However, the minister’s role in informing the relevant court is important in practical terms, leading to the adjustment of court records. Furthermore, where a person has been convicted of multiple offences it may not always be clear which are “relevant offences” under section 2 and which are not caught by the Act at all. In these cases, it would seem that the minister’s role could be instrumental in determining whether a person’s conviction or multiple convictions is/are removed from the record or not.

The principal argument in favour of the Act is that legislation is the only feasible way to deliver comprehensive corrective justice quickly. On balance that may be so, but it is unclear to what extent the government fully considered other options. Certainly a Royal Pardon would be an inadequate remedy. A pardon is useful in two circumstances. At the present time, to free from prison or otherwise alleviate the suffering of a wrongfully convicted person. No one sent to prison as a result of a Horizon-related offence is, to my knowledge, still incarcerated. Secondly, a pardon can have historical effects. In retrospective relief a pardon is suited to those offences which the Crown concludes, from today’s perspective, ought not to have been offences, or that the penalties imposed, viewed from today’s perspective, were too severe. However, in such a situation there is usually no question of wrongful conviction according to the law as it stood at the time of conviction. In the Post Office scenario a pardon, applying retrospectively, is an inferior remedy to the overturning of wrongful convictions.

The key question is whether there was an alternative way of overturning these convictions that could have been achieved through the courts. One option was mass appeals to the Crown Court from the Magistrates’ Court and, where applicable, to the Court of Appeal Criminal Division from the Crown Court. The Crown could, in the former cases, offer no evidence and the cases would fall. In the latter cases the Crown could in the vast majority of cases, or indeed every case, not oppose the appeal. If thought necessary, although surely it would not be needed, legislation could quickly be passed to reverse the burden of proof for these appeals. This would compel the Crown to prove beyond reasonable doubt why any element of a conviction should stand.

One objection to the appellate route is that many people wrongfully convicted are now so disillusioned or afraid that they would not partake in appeals. But in such cases the Attorney General could act in the public interest, leading appeals for those who have died or lack capacity, or who simply do not want to be involved. 

This approach would have had practical as well as constitutional advantages. In practice it would have allowed the Crown, if it so wished, to resist appeals in particular cases, to ensure that appeals relating to convictions that have nothing to do with Horizon and where there is overwhelming evidence of actual wrongdoing, were resisted. This would have helped avoid some of the anomalies referred to above in relation to inequality of outcomes. 

Another advantage is that each conviction would have been publicly declared. The Act quashes “relevant convictions” but in some cases people who have been exonerated may not know. There is no master list kept by the courts of all those convicted of offences as defined by section 2. The appellate route would, by contrast, require the specific identification of each conviction. This would mean the public exoneration by a court of every person wrongfully convicted. There may be some added vindication for convicted parties in a court overturning their wrongful convictions – a day in court as it were – and a degree of satisfaction when the same justice system that convicted them now acknowledges and corrects its mistakes. This is lost by a legislative route. 

On the other hand if any relevant convictions were not identified by such an appeal process, convictions would have stood until they had been appealed. In that respect the current Act seems to offer a more comprehensive remedy, albeit one that gives rise to inconsistencies, and to constitutional concerns. 

A final point worth noting is that the bill’s passage was very hasty. It was introduced in the Commons on 13 March and moved swiftly, receiving its third reading on 29 April. The government cited the urgency of the issue, but the Horizon scandal has been known about for years. It not at all clear why more time could not have been allotted to full parliamentary consideration of alternative approaches. The announcement of a general election by the Prime Minister meant also that its passage through the Lords was significantly truncated. Second reading took place on 13 May, with Committee stage scheduled for 3 June. But with Parliament proroguing on 24 May, all remaining stages were taken on 23 May. The announcement of an election is of course an exceptional circumstance, but it is very regrettable that a bill of such constitutional significance should pass through the upper house before the Constitution Committee had an opportunity to report and before the House could take full account of both its constitutional implications and any alternative path that might have been taken.

The bill as it went through Parliament enjoyed cross-party support and passed with overwhelming majorities in both Houses. The government has had a difficult job in unravelling this problem and the Act is a laudable attempt to correct an appalling catalogue of injustices ascribable largely to private prosecutions brought by the Post Office. But it seems that in its haste to find a solution, Parliament has too readily dismissed the feasibility of the appellate route that could have provided more nuanced outcomes and would have better respected the institutional integrity of the courts within our constitution.

Stephen Tierney is Professor of Constitutional Theory, University of Edinburgh. He serves as Legal Adviser to the Constitution Committee of the House of Lords. This blog post is written in a personal capacity.

(Suggested citation: S. Tierney, ‘The Post Office (Horizon System) Offences Act 2024: Necessary remedy or unwarranted interference with judicial independence?’, U.K. Const. L. Blog (3rd June 2024) (available at https://ukconstitutionallaw.org/)