Carol Harlow QC

Editor’s Note: Earlier this week Policy Exchange’s Judicial Power Project published a report by Dr Jason Varuhas on Judicial Capture of Political Accountability. The Project has invited commentators with a range of views to reply to the report. Some might be expected to be more sympathetic to the report and others much less so. The second reply is by Carol Harlow, Emeritus Professor at LSE.


In his paper examining the relationship between courts, the Parliamentary Commissioner for Administration (PCA) and Parliament, Jason Varuhas raises a number of points of constitutional importance; too many to deal with in a short comment. I shall focus on two central and interlinked arguments. First and more specifically, that the office of PCA represents machinery for political accountability and, as such, should be ‘separate and distinct from courts and legal processes’; as a fall-back position, Varuhas argues here that review should be ‘light-touch’. With this I entirely agree. Secondly, and more generally, Varuhas argues that the English courts have in recent years evinced an increased willingness to intervene in the political sphere; that there has been, in his own words, a ‘judicial capture of political accountability’. I have much sympathy with these arguments, though to my mind the picture is more nuanced.

Should the office of PCA be subject to judicial review?

The core of Varuhas’s argument is that the PCA is an officer of the House of Commons responsible to the House through its Public Administration and Constitutional Affairs Select Committee (PACAC, previously PASC). The lawmaker’s intention was, or so he contends, that the PCA should answer only to the House of Commons; similarly, the government is answerable to the Commons over questions of maladministration. The PCA system was, in short, intended as an alternative and autonomous system for dispute resolution running parallel to and independent of courts.

Undeniably the PCA is an officer of the House of Commons but the office is largely independent and autonomous. The PCA is not and never was envisaged as a political actor; section 5 of the Parliamentary Commissioner Act 1967 (the Act) authorises the PCA to ‘investigate any action taken by or on behalf of a government department or other authority to which this Act applies, being action taken in the exercise of administrative functions of that department or authority’ and a Schedule lists the bodies to which the Act applies. It cannot be supposed that the House of Commons could act by resolution to authorise the PCA to act outside this statutory mandate. The target of the PCA has always been ‘maladministration’, a term never precisely defined but initially very narrowly conceived. Crucially too, section 12(3) specifies that nothing in the Act authorises or requires the Commissioner to question the merits of ‘a decision taken without maladministration … in the exercise of a discretion vested in [a] department’. This formulation was at first narrowly construed to exclude both departmental rulemaking and ministerial discretion and only later were these points conceded. Again, the so-called ‘MP filter’, which forms the closest link with the House of Commons, was inserted in the Act as a protection for constituency MPs. Constantly criticised by successive PCAs and Select Committees as a fetter on the office, it is generally regarded today as something of an anachronism; indeed, it is currently the subject of a recommendation from PACA that thisiniquitous prohibition on citizens’ direct and open access’ to the PCA should be abolished and a promise from the (Coalition) Government to ‘consider the role of filter’ in forthcoming legislation.

The office of PCA is that of a complaints-handler, a role recognized in the coupling of the office to that of Health Services Commissioner with the composite title of Parliamentary and Health Services Ombudsman (PHSO). It is recognized also by the two most recent ombudsmen, Ann Abraham and Julie Mellor; by PACAC with its search for a ‘people’s ombudsman’; and by the government, which accepts the ‘vital role’ played by the office in ‘providing individuals with remedies for administrative injustice while at the same time highlighting the positive benefits of effective complaint handling across the public sector, and the importance of learning and sharing good practice and experience’. In short, the PHSO is a member of our oversized family of public ombudsmen and an inherent part of our fragmented administrative justice system. This does not mean, however, that the PCA operates as a parallel and autonomous justice system. The Act is clear about the boundary with courts. Section 5(2) precludes the PCA from investigating situations where the complainant ‘has or had a right of appeal, reference or review to or before a properly constituted tribunal or court of law’ unless the PCA is ‘satisfied’ that in the particular circumstances it is not reasonable to expect the complainant to resort or have resorted to it. There is a hint of hierarchy in this provision.

Like any other statutory body, the PCA is bound by the terms of the mandate. She possesses, however, a wide measure of autonomy and statutory discretion in the exercise of her duties. Section 5(2) provides that she ‘may conduct an investigation’ where a judicial remedy is potentially available ‘if satisfied that in the particular circumstances it is not reasonable to expect [the complainant] to resort or have resorted to it’, while section 5(5) permits the PCA in determining whether to initiate, continue or discontinue an investigation to ‘act in accordance with his own discretion’. It also provides that ‘any question whether a complaint is duly made under this Act shall be determined by the Commissioner’ – a formulation which, since the seminal Anisminic decision, cannot be regarded as unassailable. And rightly so. Ombudsmen, like statutory tribunals, make mistakes. In Cavanagh and Others v Health Services Commissioner, for example, the PHSO sought to extend a complaint made on behalf of a patient into inefficiency in the operation of a hospital into an inquiry into the clinical competence of medical practitioners. Similarly, in the recent JR 55 case, a Northern Ireland ombudsman recommended financial compensation by a medical practitioner in a claim that would normally be handled by civil courts. The courts in both cases assumed their traditional role of protecting the citizen against abuse of power.

Should ombudsman investigations have binding force?

Both the Occupational Pensions and Equitable Life affairs involved the reverse question: review of ministerial refusal to abide by an ombudsman’s recommendations for redress, a rare but not unknown occurrence. This is a matter for which the Act makes no explicit provision (other than the section 10 provision for reports). In common with other ombudsmen, however, the PCA customarily makes recommendations for redress and compensation; indeed, the Principles for Remedy issued by the office advise that it may, if it is decided that the organization has ‘got things wrong’, make recommendations including explanations, apologies and recommendations for the organization to put them right. Though warning that there are no automatic or routine remedies for injustice or hardship resulting from maladministration or poor service, the Principles acknowledge the possibility of financial compensation and advise the administrator how it should be calculated. As the Principles suggest, ombudsman recommendations are normally regarded as advisory and non-binding.

This understanding was reflected in the carefully worded recommendations. Each of the five recommendations contained the phrase that ‘the Government should consider whether’ to act on her Report. Very unusually, the Government declined to accept the findings of maladministration as well as the recommendations for financial redress. We might bear in mind here that the PCA’s jurisdiction to investigate had been contested. The PCA has no jurisdiction to conduct and initiate its own investigation. Section 5(1)(A) of the Act provides for investigation only when a ‘written complaint is duly made to a member of the House of Commons by a member of the public who claims that a person has failed to perform a relevant duty owed by him to the member of the public’ and this is referred on to the PCA by an MP. By blocking up groups of complaints and publishing a Special Report to PASC, successive PCAs have side-stepped this limitation. Notably both the Occupational Pensions and Equitable Life episode involved just such ‘Big Inquiries’. Ann Abraham, then PCA, told MPs in a letter that:

I had been shown indications that maladministration might have caused injustice to those who had complained to me – and to those in a similar position as those complainants. I also believed that my ability to access evidence which was not available to complainants meant that an investigation by me would achieve a worthwhile outcome, whatever its result. I therefore decided to conduct an investigation.

In Law and Administration, Richard Rawlings and I commented that ‘[a]n unkind critic might call this a “fishing expedition”,’ perhaps hinting that a judicial review might have been justified. In fact, the Government cooperated with the investigation albeit that the outcome was a ministerial decision not to accept the findings of maladministration. This concerned PASC, which (like Varuhas) argued for a full immunity from legal process:

To ask a court to review the Ombudsman’s findings would effectively make matters which are currently not justiciable subject to judicial decision. In these circumstances Parliament’s role would be diminished to that of an interested bystander. We believe that when there are disputes between Government and the Ombudsman, Parliament is the proper place for them to be debated.

If governments were ‘routinely’ to consider rejecting findings of maladministration, however, the ‘broad common understanding’ would break down and, failing new legislation, judicial review might become appropriate (HC 1081[75-8]).

In Law and Administration, we also described judicial review as an entirely logical step from the standpoint of the judiciary; either the ombudsmen were carrying out an administrative function as investigators, in which case their decisions were clearly reviewable; or if the ombudsmen were adjudicators, they could be classified with subordinate jurisdictions. The PCA is a public official with a strong measure of discretionary power from whose potentially incorrect decisions there is no clear avenue of appeal. The ‘Big Inquiry’ involves the PCA in a new and more political role. There is a very clear argument here for judicial review, if only to the limited extent recognised in the case of tribunals – which are incidentally subject to appeal – in Cart and Eba (R(Cart) v The Upper Tribunal; (MR(Pakistan)) v The Upper Tribunal [2011] UKSC 28; Eba v Advocate General for Scotland [2011] UKSC 29).

The careful and cautious judgment of the Court of Appeal in R. (Bradley) v Work and Pensions Secretary had three main stages. First, it ruled that the ministerial decision was reviewable. Since there was no clear statutory protection, this is surely unexceptional by the standards of every contemporary judicial review system and indeed justified by a line of cases stretching back to the celebrated decision in Padfield v Minister of Agriculture, Fisheries and Food [1968] AC 997, which requires ministers to give reasons for decisions. That opened up several possibilities. Either the court could focus on the government’s decision not to grant compensation or it could focus on the findings of maladministration. Rightly, it chose the second, less intrusive possibility. This left open the standard or intensity of review. The Wednesbury decision again offers two possibilities: either (i) to annul a decision that is so unreasonable that no reasonable decision-maker would have made it or (ii) to require the decision-maker to act reasonably in the sense of rationally; to hold, in other words, that the findings of fact in an ombudsman investigation are presumptively binding in the sense that they can only be rejected for good reason. In taking this second line (which as the Court of Appeal acknowledged, was controversial) the Court relied on a line of cases that allow a decision-maker to reject findings of fact made by an adjudicative body on the basis of a ‘contrary (albeit rational) view which the decision maker preferred’. Again the reasoning seems to reflect Padfield. What is the point of setting up a fact-finding body be it an ombudsman investigation, a tribunal, a public inquiry or parliamentary committee, if its findings are simply ignored? There is no statutory avenue of appeal and Parliament is not the place for a parallel fact-finding inquiry. And, as Wall LJ stressed, the decision is ‘procedural rather than substantial. The decision is quashed as unlawful, and the Minister must think again’ (Bradley at [138]). This does not seem unduly onerous.

Are Courts supplanting the role of political and democratic institutions by interventions into political accountability processes?

Where does this leave Varuhas’s wider contention concerning ‘judicial capture of political accountability’? I agree that we have for some time seen a definite trend to ratchet up the intensity of review of political decision-making. The trend is once again nuanced; some judges and courts in some situations take what Varuhas calls an ‘aggressive’ stance; other judges in different situations take an opposite view. He cites the Evans case, where the Supreme Court saw fit to overturn the decision of a Cabinet Minister accountable to both the Cabinet and Parliament, and the indefensible case of the Litvinenko public inquiry, where a well-reasoned letter drafted no doubt by a conscientious civil servant was ruled by the High Court to be irrational. This to me is to substitute judicial prejudice for the reasoned opinion of a scrupulous Home Secretary, surely better placed in the instant case to weigh the public interest in open justice against the necessities of foreign affairs.

I would add the first instance decision in Reilly and Hewstone where Lang J drilled deeply into parliamentary proceedings, impugning a ministerial statement made in support of (admittedly retrospective) legislation to rule the Jobseekers (Back to Work Schemes) Act 2013 incompatible with Art 6(1) of the ECHR.  The Court of Appeal later acknowledged that the 2013 Act had been effective in retrospectively validating the non-payment of Job Seekers Allowances, but held that the Act was nevertheless incompatible with Art 6(1), leaving in place the declaration of incompatibility that assigns the final decision (subject to any further appeal) to the Government. (R (Reilly (no.2) and Hewstone v Home Secretary, Bevan and others v Home Secretary [2016] EWCA Civ 413).

What is significant and welcome in the present context is that the Court of Appeal drew back from the intrusive thrust of the High Court judgment, seemingly accepting the admission that Article 9 of the Bill of Rights (‘proceedings in Parliament ought not to be impeached or questioned in any court’) had been breached. Stressing that their own reasoning did not involve any ‘questioning’ of the kind, the Court of Appeal added a more general procedural caveat:

We do, however, wish to observe that Lang J was put in a difficult position by the Secretary of State’s reliance in his evidence on the justifications for the legislation advanced by Lord Freud in his speech in the House of Lords and by the extent of the Parliamentary materials placed before her in the voluminous bundles produced by the parties. That made the line between criticising the Secretary of State’s case and questioning proceedings in Parliament hard to identify or observe. It has become relatively commonplace in public law proceedings for every last word spoken or written in Parliament to be placed before the court. In particular, debates are relied upon extensively when they should not be and, furthermore, the conclusions of Select Committees are prayed in aid with the court being asked to “approve” them. [T]hat should not happen.

Do the Bradley requirements breach these guidelines? I do not think so.

Carol Harlow
Emeritus Professor, London School of Economics

Share This