The Duty to Follow Policies (and its Limits)

If a public body adopts a policy about how it will exercise one of its functions, it must follow it.

This principle has been developed over the last 15 years in a series of cases brought against the Secretary of State for the Home Department. Initially it was regarded as an outworking of the doctrine of legitimate expectation (Saadi at [7]). More recently it has been treated as a freestanding ground of judicial review which is ‘a requirement of good administration‘ (Nadarajah at [68]) and ‘a basic public law right‘ (Lumba at [58]).

The full extent to which the duty has now been cut loose from its traditional moorings in legitimate expectation became evident last year in the Supreme Court case of Mandalia. In that case the claimant successfully relied on the Home Office’s failure to follow an internal policy – a ‘process instruction’ to civil servants – of which he neither had, nor could have had, any knowledge at the time when it should have been applied. Following Mandalia, even an unpublished policy is now binding.

The duty is qualified by another basic public law principle, that policies should not fetter discretion. If the circumstances of an individual case provide a ‘good reason‘ for doing so (Lumba at [26]) a public body may, and sometimes must, depart from its own policy.

Subject to this qualification, the requirement to follow existing policies has developed into an important obligation on public bodies. However, two recent cases expose some of the limits of reliance on policies as a ground of public law challenge.

DHL v Ofcom

In the High Court case of R (DHL) v Ofcom, the question was what should happen where a public body is obliged by statute to adopt a policy about the exercise of a function, but the validity of that policy is called into question when it attempts to use its powers. What does it mean for the policy to be valid, and can its validity be challenged even after it has been in place for some time? 

The Background

The DHL case is a rare example of a challenge by a company against an information notice (a formal requirement to disclose information) served on it by a regulator. The collection of large quantities of data is essential to effective regulation, but it is unusual for any dispute over the validity or enforcement of an information notice to come to court.

The reason for the anomaly of DHL is that in that case the information notice was just the convenient ground on which the company sought to fight a much larger battle.

Ofcom, acting in its capacity as the UK postal services regulator, served the notice (under Schedule 8 to the Postal Services Act 2011) on DHL in the latter’s role as a ‘postal operator‘. DHL, an international courier firm, wanted to resist this classification, which brings with it other regulatory obligations. Most of the judgment of Mr Justice Soole is concerned with the question whether only a traditional postal service company (such as Royal Mail) can be a ‘postal operator’ or whether a wider range of bodies is captured by the term. The heart of the case was a careful interpretation of the undefined statutory term ‘by post‘.

Soole J resolved this question in favour of Ofcom, and so the judgment is important in the postal services sector in establishing that companies like DHL are to be treated as postal operators under UK legislation.

The Policy

Given this context, the other grounds of challenge to the information notice have the distinct feel of being secondary issues to DHL, but that does not diminish their importance to a wider audience.

Ofcom’s statutory duty was ‘to prepare and publish a statement of their general policy’ on the use of information gathering powers, and then to ‘have regard to’ that policy when deciding to serve an information notice.

However, when it first took over responsibility for postal services regulation in 2011 (from former regulator Postcomm), it had little time to put in place arrangements for carrying out its new role. As an ‘interim measure’ it published a document pointing to an existing policy designed to serve a similar purpose under the Communications Act 2003, treating this as the policy that would apply to its Postal Services Act functions.

Parts of that existing policy were generic and could be relevant to information gathering in any context, but most of it was specific to telecoms, and had no possible application to postal services. A replacement policy was supposed to be adopted following a subsequent review, but this presumably dropped to the bottom of Ofcom’s in-tray, because four years later, when the information notice on DHL was served, the ‘interim’ policy was still in place.

DHL challenged the validity of this policy, and therefore the legality of the information notice. Soole J thought that the policy was valid [106], and that even if it was deficient it had to be presumed valid because it had not been challenged within the judicial review time limit after it was originally adopted in 2011 (‘the presumption of regularity’ [108]).


It is is hard to avoid the conclusion that Ofcom dodged a bullet here. The judge’s approach to the issue seems wide of the mark.

The first thing to note is that if there had been no (or no valid) policy in place, Ofcom could not have exercised its information gathering function, because it would have been unable to meet the requirement to have regard to the policy that it was supposed to have. If a statutory duty to have a policy goes unsatisfied, it undermines the ability to exercise the function to which that policy relates. Ofcom did not attempt to argue the contrary position.

For this reason, Ofcom needed its ‘interim’ policy still to be valid. But was it? The judge thought that it was a pragmatic solution to the lack of time and resources in 2011, and this may be right. But what justification could there have been for still having the same policy in place and relying on it four years later?

Statutes often require regulators to adopt policies about the exercise of various functions. What Parliament is requiring the regulator to do in each case is to address its mind to the powers it has been given, think about how they should be used in the context of the sector being regulated, and make a public statement of how it will use them so that companies and consumers know what to expect. Plainly Ofcom did no such thing. When it ‘read across’ [98] an existing telecoms policy, what it did in relation to its postal services powers was to adopt the form, but not the substance, of a policy.

If this was just about acceptable as a holding position in 2011, it must have long since ceased to be so by 2015. To take any other approach is to reduce the statutory duty to a box-ticking exercise, which cannot have been what Parliament intended. Ofcom should have turned its mind to how to exercise its information gathering powers in the specific context of postal services, but did not.

As for the ‘presumption of regularity’, it is difficult to see how this operates to regularise an ongoing failure to do something that needs to be done. On the contrary, such a failure is always continuing and therefore always open to challenge. If Ofcom’s interim policy was ever ‘regular’ as such, it must have ceased to be so when by the passage of time it lost any meaningful claim to be interim. What remained was a published document which failed to satisfy the basic statutory requirement – the absence, in fact, of a policy.

In summary, the judge in DHL did not treat the statutory duty on Ofcom to adopt and apply a policy with sufficient seriousness, leading to a conclusion which is hard to justify.


In passing, it is worth noting that there was also a proportionality challenge to Ofcom’s information notice in DHL. Soole J summarised the test as requiring ‘an assessment to be made as to whether the interference or inconvenience caused to the recipient [of the information notice] is proportionate to the reasons for gathering the information’ [124]. Being satisfied that Ofcom had regulatory reasons for wanting the information, he rightly had little difficulty in dismissing the claim that the requirement to produce it was too burdensome. It would require an extreme case for a challenge of this sort to work.

Lee-Hirons v Secretary of State for Justice

In the Supreme Court decision in Lee-Hirons v Secretary of State for Justice, the question concerned the effect of failing to follow a policy to take certain steps after the exercise of a function. Does this mean that the decision to exercise the function can be quashed?

The Background

On its facts, Lee-Hirons was a case as different from DHL as can be imagined.

Mr Lee-Hirons had a serious criminal history, with 61 convictions to his name. He also suffered from persistent mental health problems. In the light of these, following his latest conviction, a court ordered his detention in hospital, with a restriction order limiting the power of the hospital to discharge him.

Eventually he was discharged by the direction of a tribunal. However, faced with evidence that his mental health was deteriorating again, the Secretary of State for Justice exercised a power under section 42 of the Mental Health Act 1983 to order his renewed detention.

The government had a longstanding policy, adopted for the purposes of the Act, of how such recalls to hospital should be dealt with. This imposed three requirements relating to the reasons the patient should be given for the renewed detention.

First, at the time of the recall, someone should explain to him in simple terms what was taking place. Second, as soon as possible afterwards, and in any event within 72 hours, the hospital should verbally explain the reasons for his readmission. Third, by no later than 72 hours after the recall, he should be given that explanation in writing.

In this case, there were entirely sound reasons for the renewed detention, but only the first of these steps was followed. A verbal explanation of the reasons was not given until 15 days after readmission to hospital, and a written statement followed only several months later [20].

The Judgment

The policy in this case was unusually important. As the Supreme Court noted, deprivation of liberty has been subject to strict control since Magna Carta [23]. Providing reasons for detention is an element of those controls and a requirement of the European Convention on Human Rights (ECHR). The policy was designed to ensure this was done properly.

On the other hand, there were in fact good reasons for the renewed detention of Mr Lee-Hirons, and they were communicated to him – certainly sufficient for compliance with the ECHR, albeit later than government policy required and therefore in breach of public law.

The Supreme Court was faced with the question whether the failure to comply with the policy infected the decision to detain and rendered it unlawful. It was clearly reluctant to find that a post hoc administrative error led to this rather large conclusion.

Giving the judgment of the court, Lord Wilson sought to disconnect the failure to comply with the policy from the decision to detain. Reaching back to a concept used by Lord Kerr in Lumba, he said that the detention was unlawful only if the public law breach caused by failing to follow the policy had a ‘direct’ bearing on the decision to detain [34].

In Lord Wilson’s view it did not – ‘In my opinion there is no link, let alone a direct link, between, on the one hand, the Minister’s wrongful failure for 12 days to provide to the appellant an adequate explanation for his recall and, on the other, the lawfulness of his detention’ [39]. Failure to follow the policy was unlawful in itself, but it did not generate the further illegality of a wrongful decision to detain.


The Supreme Court’s judgment in Lee-Hirons is both understandable and surely correct.

However, it is important to be clear why, when non-compliance with a policy will usually make a decision unlawful, the failure of the Secretary of State to comply with his policy did not have that effect on the decision in this case.

Relying on the adjective ‘direct’, to establish a test of the degree of connection required between a policy and a related decision, lacks that clarity. When the courts take refuge in abstract concepts – in this case ‘directness’ – those concepts often become a repository for whatever meaning a judge finds it convenient to project onto them in an individual case. They have little predictive capacity, and do nothing to contribute to legal certainty.

Perhaps for this reason, while (like the rest of the court) agreeing with Lord Wilson, Lord Reed made a further attempt to articulate an explanation for the outcome in Lee-Hirons. He gave three reasons, but it is really unnecessary to look beyond the first of them, which relates to timing – ‘…the duty is to provide reasons ex post facto. It remains capable of meaningful performance even after the 72-hour period has expired. Such performance can if necessary be enforced. Delayed performance does not, in this situation call into question the validity of the antecedent decision’ [52].

In other words, the failure to follow the policy related to, and rendered unlawful, only the things done after the decision to detain. It did not have the effect of reaching back in time to infect a decision which was otherwise valid at the time it was made.

The chronology matters, because it goes to causation. It is not necessary to say that there was no ‘direct’ link between non-compliance with the policy and the impugned decision; in reality there was as Lord Wilson says no link at all.

Summary and Conclusion

Over recent years, the courts have taken an increasingly strong line on the need for public bodies to comply with their own policies. Culminating in Mandalia, case law establishes the requirement to comply with a policy, whether it is published or not, as a free-standing public law duty.

This has potentially large consequences, creating a ground of challenge against any public body that is so lax as to forget to do what it said it would, or so administratively inefficient as to fail to achieve it in practice. But the very significance of these consequences is also a reason why the courts might seek to limit them in individual cases.

Lee-Hirons does so robustly, though with something of a struggle to articulate exactly why. Its key lesson can in fact be simply stated – a decision which would otherwise be lawful is not compromised just because of a failure to comply with a policy that requires a series of post hoc steps to be taken. That failure will itself be unlawful, but does not retrospectively invalidate the original decision.

DHL is a harder case to justify. Where a public body adopts a policy of its own volition – and the force of the duty to comply with policies lies in the fact that a body which fails to do this is hoist with its own petard – there may be no particular requirement as to its content. But where there is a statutory duty to have a policy, that policy ought to be meaningful and the result of proper consideration of the power being exercised.

If merely having a published document in place is sufficient to discharge a statutory duty to have a policy, the public law duty to follow that policy is also seriously undermined.

It is unrealistic to imagine that policies will usually be challenged in their own right (even though they can be) – most often they will become the focus of serious legal scrutiny only when the powers to which they refer are being exercised. In the majority of cases, the presumption of regularity ought to rescue a policy which has been in place for some time. But where the failings of a purported policy are sufficient to undermine it entirely, so that it constitutes no kind of satisfaction of the statutory duty, the mere passage of time should really not be sufficient to redeem it.

However, for the moment, DHL constitutes an unhelpful precedent to the opposite effect.