Identifying the rules and practices that should properly be regarded as part of our flexible and uncodified constitution is nowhere more difficult than in relation to the internal workings of the executive.  At one time this difficulty was normally attributed to a culture of secrecy within government that kept information about its internal organisation out of the public domain.  But as Alan Page and I document in our forthcoming Executive Self-Government and the Constitution (OUP, expected April 2025), which shows how constitutionally significant rules and practices within the executive have evolved over the course of this century, a far greater volume of material about such rules and practices is now published – a response, one might surmise, not so much to the formal requirements of the Freedom of Information Act as to the informal pressures and expectations fuelled by the internet, social media and the 24-hour news cycle. 

Despite all this information, ignorance and misunderstanding of key principles of executive organisation and control remain substantial.  The sheer volume of material, and the fact that little of it is framed in the formal shapes familiar to lawyers – statute and regulation – with most appearing as ‘guidance’, ‘standards’, ‘declarations’ or ‘letters’, discourages brief exposition and hinders its admission to accessible sources like basic constitutional law or politics texts.  One unfortunate effect of the low level of public and professional knowledge which results is the ease with which well-established principle and practice can be misrepresented to, and in, the media.  2025 has already yielded a powerful example, in the shape of the campaign of criticism emanating from ministers and directed at the Attorney General, Lord Hermer.  Superficially, the issue here is the reputation, and survival in office, of a minister.  More fundamentally, the campaign calls into question the attitude of the government to respect for legal standards, both in the formulation of policy through legislation and in the implementation of policy through administrative decisions.  At the same time, it diverts attention from changes in the position of the Attorney General which are of considerable long-term constitutional significance.  

Legal risk and administrative decisions

Labour ministers are complaining, it appears, that Lord Hermer is imposing a ‘freeze on government’ by reason of his ‘finickity’ attitudes to the legality of proposed administrative action and the framing of new legislative proposals.  They are reported as saying that he is causing delays to policies even if there is a relatively marginal risk of illegality, and is inserting himself into discussions on legislation at the policy stage.  And his insistence on respect for international law norms is regarded as excessive in the face of policy challenges like those presented by illegal immigration.  There is a nice irony in the fact that Labour politicians who complained loudly about the Johnson government’s readiness to pursue policies of dubious legality – and in some cases, involving international legal obligations, of clear illegality – now complain that the Attorney is making it more difficult for them, as ministers, to do the same thing.   

One may ask how it is that the Attorney who, no matter how close he may be to the Prime Minister, is still only the government’s legal adviser, can single-handedly impose a freeze on the business of government.  It is for the relevant minister to decide, in relation to administrative action, whether to run the risk that it might be held illegal by the courts.  It is for the government’s Parliamentary Business and Legislation Committee (PBL), bringing together a number of senior ministers with the Lord President of the Council (and Leader of the House of Commons) in the chair, to decide how departmental proposals for legislation advance through the pre-Parliamentary stages.  In each case the advice of the Attorney may be expected to be highly influential; in neither is it determinative.  Executive Self-Government and the Constitution, analysing the way legal advice, among other executive functions, has developed over the course of the current century, offers a number of examples of individual and collective ministerial pushback against internal legal advice.  Indeed, our suggestion is that this tendency has increased in recent years, notably in the period of the Johnson and Sunak governments.  This has been the case in relation to advice on both administrative legality and legislative standards.

Administrative legality 

On administrative legality ministers, and public commentators on the issue, misrepresent the extent to which Lord Hermer is departing from the approach of his predecessors.  According to The Times, ‘Late last year Hermer strengthened the guidance on legal risk [replacing this], changing it to say government lawyers must advise ministers that policies are unlawful if they believe an argument could not be properly put forward in court. This was not previously the case. The new guidance states that they should only as a last resort put forward legal arguments that are ­legitimate but which have a high risk of failing.’

This is said to be a ‘reversal’ of changes introduced by Suella Braverman, the former Attorney General, ‘who relaxed the guidance amid concerns that lawyers had taken a “computer says no” approach and “needlessly” hampered policies.’ Hermer’s so-called ‘reversal’ consisted in substituting, as the criterion for advising that a policy was unlawful, the absence of any ‘tenable’ legal argument to support it, in place of any ‘respectable’ legal argument.  This is a distinction without a difference: ‘tenable’ argument and ‘respectable’ argument are each defined as one that ‘a lawyer representing the government could properly advance . . . before a court or other tribunal.’ Nor did Braverman ‘relax the guidance.’  The test her guidance adopted was the same as the one in the previously existing guidance, developed not by the Attorney General but by the Government Legal Department (GLD) (which is led by the Treasury Solicitor) and first published in 2015.  Like Hermer’s, it made clear to government lawyers that ‘If . . . you conclude that there is no respectable legal argument that can be put to a court, then you will need to advise that the proposed action is unlawful.’

This blog has already addressed the overheated response of authors writing under the banner of the Policy Exchange think tank to the provisions of Hermer’s guidance emphasising the importance of consideration of international as well as domestic law.   Those authors have also expressed alarm about provisions of the guidance which, they say, signal greater personal involvement of the Attorney in the legal scrutiny of government policies, leading to the ‘unwarranted and unconstitutional development’ of his ‘wielding a de facto veto over a wide range of policy decisions.’  But it was Braverman’s guidance that introduced the suggestion that senior GLD lawyers might ‘need to alert’ the Attorney General’s Office (AGO) before advice was given that proposed action could not be supported by respectable legal argument.  Hermer has added guidance about other precautions to be taken when advising on action supported by no more than a ‘tenable’ legal argument (para 20), and has extended the scope of such upward reference to cases where the legal risk is assessed as being otherwise high or medium-high.  

His supposed appetite for increased involvement is however belied by unpublished advice from AGO, which has not been modified, asking lawyers to think carefully before they request Law Officer advice, and to consider other possible sources like outside counsel or specialist resources within GLD.  And the general criterion for involvement of the Law Officers in the legal advice function has not changed from the principle consistently stated in the Ministerial Code since 1997, and repeated in Hermer’s guidance (para 22), that ‘the Law Officers must be consulted in good time before the Government is committed to critical decisions involving legal considerations.’

Legal standards and preparation of legislation

It is hardly surprising that ministers new to office and impatient to make their mark should fret when they find that the process of turning their policy ideas into legislative proposals involves a certain amount of quality control operated by agencies within executive government.  In periods when governments command large House of Commons majorities, such controls may be more important than those exercisable by Parliament.  There is an elaborate instruction book for Departments on preparation and passage of legislation, the Cabinet Office’s Guide to Making Legislation, which is operated by its Parliamentary Business and Legislation Secretariat.  Traditionally, it has been those drafting legislation – Parliamentary Counsel – who have been seen as the custodians of values regarded as integral to the legal order, such as non-retrospection, proper use of delegation, and respect for the liberties of the subject.  If they felt that the proposals they were dealing with threatened such values, they could in the face of departmental insistence seek the intervention of the Attorney and the other Law Officers to achieve settlement of the issue at ministerial level.    

Since Labour ministers were last in government, however, the Attorney, along with the Advocate General for Scotland, has taken a much more proactive stance, issuing guidance to departments – Working with the Law Officers on Legislation– which requires submission to them, early in the drafting process, of a memorandum identifying key legislative policy issues, including those mentioned above.  It also indicates the position that the Law Officers might take on certain kinds of provision, and identifies provisions for which their consent is needed: essentially retrospection and early commencement.  This document appeared in 2019 – under a Conservative government and during the tenure of Geoffrey Cox.  

Labour ministers’ complaint about Lord Hermer appears to be that he is actually applying this guidance: “He was willing to swallow — to a certain degree — not necessarily following the right process in the early days because they had to work at pace, he understood things had to be done quickly and he wasn’t quite as obstructive,” they said. “But as we bed down, I think it’s become obvious that he does want things early, and he does want an appropriate interval for him to consider things, . . . ” Relaxation of legislative standards in periods following elections has been a regular occurrence in recent years, with the PBL Committee tolerating rushing bills out quickly in the early days of a new Parliament, to show fidelity to manifesto promises, if necessary on the basis of half-baked policy.  Lord Hermer’s critics seem to want to perpetuate such an approach to legislating, prioritising political impact over clarity, coherence and other legal values: an approach  that led, during the Johnson Government, to unprecedentedly vigorous criticism by House of Lords Committees, including the  Delegated Powers and Regulatory Reform Committee and the Secondary Legislation Scrutiny Committee.

Conclusion

The personal campaign against Lord Hermer also distracts attention from a more significant constitutional development within the executive.  The years since 2000 have seen both a massive growth in the size of the government legal service and the almost complete centralisation of legal advice and civil litigation in the Government Legal Department.   GLD has absorbed the legal advice teams of large departments, whose chief legal adviser formerly answered, through the Permanent Secretary, to the minister.  Those lawyers are now under the line management of the Treasury Solicitor, the head of GLD – a non-ministerial department – with the rank of Permanent Secretary.  It is hardly surprising that this remarkable piece of centralisation – not achieved in any other function of government – should have caused a re-examination of the relationship between the Attorney General and other government lawyers.  That relationship has traditionally been defined by the guidelines indicating when legal issues should be referred to the Law Officers (formerly set out in detail in the Ministerial Code, now contained in the Cabinet Manual).  This recognises the primacy and finality of the Attorney’s legal advice; but it does not give the Attorney any powers over GLD beyond their traditional function of ‘superintendence and sponsorship’.  

Since 2019 Attorneys General have been seeking to enlarge their powers.  One signal was the replacement of the GLD’s 2015 Guidance Note on Legal Risk by the Attorney General’s Guidance on Legal Risk issued during Suella Braverman’s tenure as Attorney in 2020-22.  Her predecessor, Geoffrey Cox, was dissatisfied with the limitations imposed by his role of ‘superintendence and sponsorship’ of GLD, and floated the idea that GLD should be absorbed into an Attorney General’s Department so as to be subject to ministerial control – a proposal originally made by Sir Robert Andrew’s inquiry into government legal services in 1989.  While this idea seems to have received as cool a welcome within government in 2019 as it did in 1989, Law Officer involvement in the affairs of GLD has been reinforced by the creation of a ‘ministerial strategy board’; Braverman, as Attorney, commissioned a review of its activities; and attempts continue to create a framework agreement for GLD of the kind that regulates the relationships of bodies like executive agencies and non-ministerial departments with their sponsoring departments.  The Crown Prosecution Service and the Serious Fraud Office each have such an agreement with the Attorney General.  The fact that discussion of such an agreement has been going on between GLD and the Attorney General’s Office since 2019 suggests that there are significant issues to be resolved.  The general direction of movement – towards stronger centralised ministerial control of the legal advice and litigation function – is, however, apparent.  Is it the right one?

Terence Daintith is Professorial Fellow at the Institute of Advanced Legal Studies, University of London, and the author, with Professor Alan Page, of Executive Self-Government and the Constitution (OUP, forthcoming, April 2025).   

(Suggested citation: T. Daintith, ‘Attacking the Attorney General: Some constitutional and administrative context’, U.K. Const. L. Blog (6th March 2025) (available at https://ukconstitutionallaw.org/))