People, ideas machines XIII: The origins and evolution of the Cabinet Office, the heart of darkness in the permanent government

Questions for the next PM: How might a *regime complete* government control the government? 'Reform' the Cabinet Office -- or *close* it?

Bureaucracy is a bad European system of government, created by the use of permanent public officials, a system that does not, should not, and cannot exist in England.

Palmerston to Queen Victoria, 1837

He is very depressed about the hopelessness of the present system of governing by 22 gabblers round a table with an old procrastinator in the chair.

Edward Carson, recorded by Leo Amery, 21 July 1915

[W]hen he joined Asquith’s Cabinet [5/15] he was astonished at the lack of method, the absence of any agenda or minutes. He told Asquith this. The latter said that everyone who joined the Cabinet made the same observation but speedily became reconciled to the method of doing business…

Bonar Law

The Cabinet often had the very haziest notion as to what its decisions were.

Curzon, 1918

[The Treasury] had shown neither foresight nor organising capacity, and had plunged the country into something like administrative chaos until the War Cabinet machinery came to the rescue.

Maurice Hankey, the first Cabinet Secretary, 27 October 1922

Unlike Cabinet Ministers who have their fame entombed in rows of bulging biographies, the great Civil Servants often hardly attain to the humble dignity of a footnote to history. A Civil Servant does good by stealth and would blush to find it fame; a Cabinet Minister does good by publicity and would resign if he failed to secure it! It is easy to decide which is the more indispensable to a nation’s welfare. The country easily survives the frequent changes of ministries; it hardly moved a muscle when a Labour Government climbed for a moment to office; but it would receive a staggering blow if the Civil Service suddenly took it into its head to resign tomorrow. Some Governments are in office but not in power; the Civil Service is always in office and always in power.

Baldwin, talk to the Civil Service 1925

The function of the Cabinet Office is essentially one of machinery of Government and not of policy. It cannot be too strongly emphasised that the Cabinet Secretary has no duty of offering advice in any matter of policy or of interfering in any way with the functions of responsible Departments in this respect.

Maurice Hankey, the first Cabinet Secretary

The Secretary of the Cabinet is, in a sense, the ‘prime minister‘s permanent secretary’, to use a phrase of the … Secretary’s predecessor [Sir Burke Trend] on handing me my first brief in 1964; but his loyalty is, no less, to the Cabinet and the doctrine of Cabinet government... He advises also on the practices and conventions about the conduct of ministers in relation to collective Cabinet responsibility, as well as on other matters affecting collective responsibility and loyalty.

PM Wilson reflecting on the role of Cabinet Secretary in retirement (NB. the phrase about loyalty being no less to the Cabinet, this is important as it is used by CO officials to justify not doing what they know the PM wants)

I think it is a good idea to reassert the Civil Service being in the lead in Number 10.

Jeremy Heywood, Cabinet Secretary, 2010.

There are not two departments. I stress there is one department. There is one Cabinet Office of which Number 10 is a subset ... a business unit.’

Gus O’Donnell, Cabinet Secretary

From the decision to elevate the general administrator, the mandarin, and grant to his corporation supreme influence, much of the present discomfiture of the country has followed. Hardly ever has so anachronistic a change occurred in a vital organ of a great empire at a worse moment. The discrepancy between tasks and means is still steadily increasing.

Thomas Balogh, The Apotheosis of the Dilettante, 1959

Bob, this setup is all wrong. The joint ought to be jumping — and it’s not. Your office ought to be quiet as a grave. Instead your office is jumping and the plant looks like a graveyard.

Harriman, The Man Who Sold the Moon. (Analogy: the problem with a permanently ‘jumping’ PM Office while the entities that should be jumping are graveyards.)

This blog concerns the heart of power in the British state — the PM’s office in No10 (PMO) and the Cabinet Office (CO) created in December 1916 out of the crisis of the Somme. The story of the CO’s creation 1916-39 helps enormously if you want to understand how and why core institutions of the British state became pathological, and what could be done to turn the dismal tide.

It’s written not for the political class of SW1 generally, which (unlike the deep state) has demonstrated an intense anti-interest in these things, but for those few aspiring to be the next PM or helping the next PM, taking power amid the wreckage of both the domestic system and the post-1945 international security system and its institutions. In other words, it’s written for a tiny subset of the political class (Insiders) plus a subset of the deep state (Insiders) plus those Outsiders thinking about how to break the grip of Insiders, how to do regime change properly, and what a once-a-century burst of powerful, thoughtful energy looks like in detail.

The most hopeful aspect of our deepening crisis is that crises also push elite talent into rethinking their priorities and this is happening in Britain among Outsiders. Money, talent and ideas are moving fast. The SW1 mainstream is almost entirely unaware, just as they didn’t see the spectral WhatsApp groups signalling the dramatic Silicon Valley shift 2021-3, but NPCs increasingly display blind panic: they can feel their loss of narrative control though their own information ecosystem acts as a multi-billion dollar denial-of-service attack against themselves, so they can’t see straight why their narratives are sinking.

It will be interesting for those working in No10 now but I continue to think it’s extremely unlikely that Starmer will break the pattern of modern PMs: go along with the system, whinge more and more but do nothing more than tinker and shuffle NPCs and spin, then write the usual paragraphs of post-Thatcher memoirs about their ‘frustrating search for the levers of power’, then urge their successor to ‘respect our institutions’, i.e repeat the same pattern of uselessness, moral cowardice and failure. Yes, McSweeney and others would like to do more but the guy who promoted the guy responsible for pandemic preparations and response to be Cabinet Secretary is not the guy to do what’s needed. But 2-4 more years of our present disintegration may well generate a PM and team who return to the Vote Leave premise: tinkering and ‘respect the institutions’ is doomed, our institutions need profound regime change and this is fundamentally entangled with delivering what’s promised to voters, you can’t deliver for voters without regime change because the old Whitehall/Westminster regime will sabotage execution of priorities consciously and unconsciously.

Perhaps the most interesting and deepest lesson from the history of the CO is how a) the players who set up the CO agreed it must *not* do a range of important things — such as control personnel, give policy advice to the PM, interfere in ministerial responsibility etc — yet b) it evolved so the CO now does all of those things they agreed in 1916-17 it would be disastrous for it to do, and c) there has been close to zero focused, knowledgeable, honest Insider discussion about the role of the CO for decades, so much so that MPs aren’t embarrassed to say ‘I don’t really know anything about what the Cabinet Office is and does’, and most ministers leave their pointless career without ever grasping the CO’s role in the pointlessness of their meetings.

Ironically, the main criticism in the CO’s first years was that it would lead to too much power for the PM. But although it started out as the PM’s Office controlling the CO, the PM’s Office gradually dissipated its power to the CO which now controls the PM’s Office. If you watch modern meetings between the PM and Cabinet Secretary, they often involve the PM meekly asking for something and the Cabinet Secretary politely brushing them off. The body language and spoken language has shifted so the PM sounds like the supplicant and the Cabinet Secretary the chief.

Over many decades, the system has evolved an emotional miasma around the PM’s Office to conceal the reality of the PM’s power and on occasion to suggest, subtly, that it just wouldn’t be proper for the PM to do X: well, theoretically PM, but in practice it would be seen as, well, ummm, undermining our institutions. Cue: startled panic — oh oh ok, sorry, I see, yes, well we can’t have accusations like that, I must follow the advice, but … there’ll be the devil of a row, still … nothing for it I suppose… And this pressure is applied with force when the system has to defend itself over its most stupid and indefensible actions, including promoting people for failure, one of its crucial functions and the defence of which is the clearest signal of whether you’re a true Insider or not.

I vividly remember when I got the Trolley to reject various personnel appointments sent for his ‘approval’ — the usual sort of farce where the science guy in HMT dealing with the deep state was suddenly shifted to deal with cows in DEFRA, in the finest traditions of the meritocratic ‘Rolls Royce civil service’. The CO sends appointments in a clever psychological package: bishops etc, which no PM has actually decided in a century, go on top to condition the PM into mindless tick, tick, tick, on a Friday night. Then, after this set up, the operations of the deep state, logical only in its mad closed world, are there to be ticked with the same level of thought and interference as the Bishop of Coventry.

When I suggested to the Trolley that he reject the mad appointments rather than tick them, he was startled: can I do that, Martin [PPS] gives me these as if I have to just tick everything?! Yes Prime Minister! He scribbled ‘No!’ instead of a tick. The PPS was called by the Cabinet Secretary: WTF is going on?! And the PPS said to me, ‘what’s going on you can’t do this?’ And I said, do you mean the PM does not actually control these appointments, it’s just a Potemkin constitutional convention now like bishops, because if that’s what you and the Cab Sec mean you should explain this to the PM…?Long stare… No, I don’t mean that, but, well, it’s highly irregular and although the PM can theoretically reject these appointments, it’s very unwise, and … YOU CAN’T! (See my covid testimony for other examples. My second covid testimony has not been published by the Inquiry because it criticises the Inquiry: logical.)

One of the most interesting things about working in No10, something which really needs an artist’s eye to describe properly — historians cannot do it fully, see blog on Tolstoy and politics — is the subtlety of how this ecosystem has evolved and plays out in meetings, the courtiers of the CO and the modern, all too modern, sensibility of the characters who now become ministers and PM, the former weaving psychological spells and the latter conditioned to want, to need, to believe in them.

Nothing was ready for the war which everyone expected… The longer the Emperor remained at Vilna the less did everybody … do to prepare for the war. Every effort of the men who surrounded the Sovereign seemed directed solely to making his stay as pleasant as possible and enabling him to forget the impending clash of arms. (War and Peace, p723)

Previous PMs would be appalled at the impertinence of Gus O’Donnell’s candid statement about how the PM’s Office is seen as a subset ‘business unit’ of the Cabinet Office (above), and appalled at similar comments from other Cabinet Secretaries (below) — but most appalled, and amazed, at the MPs who let this happen, who proclaim their own castration as a ‘jewel of our constitution’, who rebaptise their castration as ‘defence of our meritocratic institutions against extremists’, when the system is, for example, promoting and honouring officials who should have been fired for killing people in covid.

A PM could reverse this fast. A highly underrated fact about Britain is that, because of our unwritten constitution, a Live Player PM could change things faster and more effectively here than any other western country. And the crisis to motivate such change is deepening every day and will continue to deepen until the election because the old institutions are in a classic systems crisis, the normal historical cycle of regime change, in which they make their own problems worse, including by reinforcing memes among Insiders which stop them seeing reality. It’s a grander version of the Cameron-Osborne OODA-loop-as-denial-of-service-attack-on-their-own-perceptions-of-reality when they responded to ‘Turkey is joining the EU’ by calling their press conference on the roof of the hotel in summer 2016 to denounce ‘lies’ and thus spread like wildfire the meme that ‘Turkey is joining’.

---

The story of the CO’s creation is told in A Man and an Institution: Sir Maurice Hankey, the Cabinet Secretariat and the custody of Cabinet secrecy, by John Naylor.

At the end is a) a summary of crucial points and b) questions for an aspiring PM determined on breaking the cycle of regime failure, to ponder deeply before the next election. I explore the concept of ‘regime completeness’ — i.e the combination of things in a system (Leader, plan, team etc) which is needed — because identifying point solutions for particular problems cannot work in the systems failure we’re experiencing. For non-British readers, you can skip the main part of these Notes and look at the analysis at the end for some things relevant to all modern governments.

Today there is a card-activated connecting door, about 20-30m from the Cabinet room, connecting No10 and 70 Whitehall (70WH) where the CO lives (which most people enter from the main entrance). If you’re standing in Downing Street looking at the famous No10 door, the Cabinet Room is straight ahead behind that door and this internal connecting door to the Cabinet Office is, say, about 50-70m away in a sort of 1 to 2 o’clock direction. No10 staff passes let you through to the CO and some CO passes let you into No10. (It’s also a security risk because if someone smuggled a weapon into 70WH all you’d need to do is take someone’s No10 pass, walk through the unguarded automated door, walk 20-30m and kill the PM: there’s no intervening security. You’d have a decent shot of then getting over the garden wall onto a waiting motorbike. Though the continued lack of drone defences for the PM, five years after I raised the issue and was told ‘there’s nothing’, provides easier methods.) Below 70WH are the COBR rooms and below them is access to tunnels that connect different parts of Whitehall. One tunnel runs from below the CO to the MoD and is used when demonstrations block Whitehall so the PM can walk through the connecting door, down the CO stairs, through the tunnel, then walk from the MoD to the Commons without leaving the secure bubble. Below all this are the nuclear bunkers and a list of people with access to this is known as the ‘Pindar’ list.

The CO was born in the crisis of military failure in World War I and the widespread agreement that both a) the PM, Asquith, couldn’t do the job and b) the ‘machinery of government’ around him had proved deficient. The institutional structure that developed after the Northcote-Trevelyan ‘professionalisation’ of the civil service in the 1850s failed in important ways in the war, see Hankey’s excoriating comments on the Treasury (above) and similar comments about the Foreign Office by many.

This story is connected to the debates in Whitehall from the 1890s through to 1914 about the rising German threat, the need to expand armed forces and the navy, changing technologies, the competing demands of imperial defence far from Europe versus the threat of a European hegemon and invasion, the tensions between defence spending and domestic politics and the reluctance of the rich to pay more taxes, Whitehall arguments over the Belgian guarantee which echoed the arguments of the 1866-7 crisis and 1870 crisis when Bismarck exposed the dilemma over deterrence — threaten, and perhaps deter but perhaps get embroiled in a war, or don’t threaten, and perhaps watch Belgium get snaffled. I can’t go into all that here too but cf. The Weary Titan which explores these themes (part funded by Andy Marshall of the Office of Net Assessment). This period (1848-1914) is a critical case study for everybody in politics as it brings together most fundamental themes of modern politics, including the failure to create institutions able to think through crises of modern speed and scale as technology transformed politics and war and the spiritual crisis of modernity first and best described in Dostoyevsky and Nietzsche (blog).

Deeply frustrated by these debates, Lord Roberts, last to hold the title Commander-in-Chief of the Forces, said to Balfour in July 1904 a decade before the war, ‘At times I despair of any improvement without some national disaster.’ British elites proved unable to create the right institutions to enable better thinking. The institutions which evolved after Northcote-Trevelyan hobbled and disincentivised better thinking. As I’ve pointed out many times, there was only one serious discussion between soldiers and politicians about the connection of military strategy and the Belgian guarantee (CID, 23/8/1911) and it was rather shambolic and definitely inconclusive. Asquith was extremely unsuited to what’s needed for getting to the heart of such a problem: relentless realistic probing about core priorities without wishful thinking and/or prevarication — are we prepared to see Belgium occupied, it yes/no then what does this imply, are we prepared to see France crushed like 1870 again, how do we deter Germany etc? The failure of Asquith and others to face these trade-offs persisted into summer 1914 and caused a fatal paralysis in British statements intended to deter. The result was the worst of all worlds: Germany thought we were making clear we would stay out, so we did not deter, then we felt so entangled we fought, so we fought without even the counter-factual benefit of possibly deterring and therefore avoiding the need to fight.

The institutional failures continued into the management of the war. The crisis of WWI did not produce a political leader close to the abilities of Pitt a century earlier (see blog last year on Pitt and Metternich). In 1915 Asquith complained that the war had produced no great generals and Lt-General Wilson shot back, ‘No, Prime Minister, nor has it produced a statesman.’ Asquith embodied characteristics that have come to typify senior British politicians including a sensibility pathologically prone to push off thinking about the hard questions while avoiding organising things.

Our party system had already evolved to promote people who are adept at schmoozing the party in-group to rise in its ranks but can’t do the job of PM which requires the art of schmoozing and, sometimes, its exact opposite: facing reality, focus on priorities and disciplined execution of complex coordination — all of which becomes more important as the scale of a crisis grows with technology’s development. Asquith had risen through the system to its pinnacle but was out of his depth from summer 1914.

In 1916, the evolving crisis finally spat out Asquith and replaced him with Lloyd George (LG). LG had got to know Maurice Hankey (MH) well and they had much discussed the organisation of government.

NB. The quality of people and institutions — and the underlying seriousness of our political culture — was much higher then than with Ukraine and they failed then. A fortiori Outsiders should not be surprised at the delusions and disasters of SW1 over Ukraine.

---

I won’t repeat here what I’ve said many times about the deep reasons for the Westminster rot. See recent Oxford speech and my Q&A page (including how I actually spent my time in No10 versus the fake stories told about it). It is a systems crisis as I’ve said for twenty years and only systems politics, working on all of these, can fix the rot. Though I think I might have accidentally made ‘systems politics’, the right way of thinking about this, a sort of anti-meme among Insiders— ‘an idea with self-censoring properties, an idea which, by its intrinsic nature, discourages or prevents people from spreading it’ (cf. this interesting sci-fi story).

There are many Labour MPs and spads now in the same insane meetings as 2010-24. Many have quickly grasped that the NPC official story — Rolls Royce machine, only problem was Brexit and Tories, just needs grownups in charge — is junk. And they grasp that if they stick to it and, Sunak-like, try to pretend to voters and themselves that ‘the system works’, their project will cave in as Sunak’s did. Hence Labour ministers and spads, shocked by their crazy meetings, briefing the BBC that it ‘turns out Cummings was right’ about Whitehall.

Someone will have to rebuild the PM’s office and the core of power. And this necessarily involves the critical question: do you try to ‘reform’ the Cabinet Office, or do you close it? I think all Labour attempts to reform the CO 2024-29 will fail, its horrors will continue to metastasise, and ideas such as simply closing the CO and the PM actually controlling the centre of power — a ‘crazy extremist’ idea — will become increasingly ‘mainstream’. (There is an intense power struggle underway now (summer 2025) between Wormald, other parts of the deep state and political forces. Some want rid of Wormald. Others want to reinforce him and extend their control of No10, including by manipulating the appointment of a ‘No10 Permanent Secretary’ — which could easily be one of those things spun as ‘the PM gripping the operation’ which is actually *the operation gripping the PM*. Little of this power struggle has leaked out other than No10 spad unhappiness with Wormald.)

It will greatly help those working on a potential serious regime to understand why and how the current system was built. The current system, obviously, is not what LG and MH built in 1916-18. It’s constantly evolved. It changed a lot under Cameron-May when officials grabbed a lot of power for themselves while pushing ministers and spads away from involvement or even sight of the CO machine, especially with the National Security Secretariat which is now over 500 officials with effectively zero ministerial responsibility.

Our ancestors who built functioning institutions which preserved our civilisation would be amazed to wander around the complex today, through the connecting door, and watch all the meetings of officials only. The officials took over, they run the place, the Cabinet Secretary took the politicians’ power, they use pre-meetings to control the Potemkin meetings with the politicians, they just give scripts to the politicians to read out while decisions are taken elsewhere, ministerial responsibility is now all fake — and the politicians barely seem to notice, never mind care — they would say to each other in amazement. Then they would look at institution after institution outside Westminster and it would all make tragic-comic sense.

To those thinking about 2028/9, remember — our biggest allies in radical change in 2020 was a subset of deep state officials themselves. A serious regime will have unexpected allies as well as bitter enemies.

Comments [in square brackets] from me, bold is my emphasis…

(I won’t here go into the ravings of SW1’s NPCs in recent weeks as they thrash around in every more absurd Narrative Whiplash. Their cycle will continue to deepen and accelerate as it has since 2016: radicalise to the Left, tell themselves fantasies about public opinion and ‘the centre ground’, blame their problems on a) techbro-driven fascist radicalisation and b) idiot voters fooled by fascist disinformation, and demand louder and louder a return to pre-democratic ideas of censorship of political news ‘to defend democracy from fascism’. It’s inherently comic yet also worrying because the heart of it is the collapse of consensus reality among elites as technology breaks the old centralised information ecosystem, and the emerging political struggle between networks who each think of the other as a mix of delusional and evil is a struggle with a tendency to violence. Cf. The pathological simulacrum and the cycle of narrative whiplash. Also cf. Jon Askonas on modern media, on consensus reality, and on conservatism and technology.)

---

Series on People, Ideas, Machines

XII: Theories of regime change and civil war. Notes on Turchin’s book. And on Timur Kuran, preference falsification/cascades, how sparks start prairie fires.

XI: Leo Strauss, modernity and regime change — and an update 20/5: Notes on: On Classical Political Philosophy

X: Freedom's Forge — the story of American business and industrial production in World War II. Incredible contrast between the America of WWII and now viz building things. Highly relevant to current debates on tariffs, supply chains, AI/drones/robotics etc.

IX: IX: A) Britain's 'Organization of Victory' under Pit 1793-1815 and B) Metternich & European Community. How Whitehall-1795 was more like SpaceX-2025 than Whitehall-2025 is. Real meetings. R&D taken seriously. Procurement and infrastructure taken seriously. Over 230 years Whitehall has gone backwards.

VIII: CIA counterintelligence chief James Angleton, 'a wilderness of mirrors', covert operations, assassinations, moles & double agents, disinformation. A blog on Angleton and the broader history of the CIA and US elites’ attempts to understand the political world. The long-term failures of the CIA on critical geopolitical issues, their security failures and penetration by the KGB, the fundamental problems of building effective intelligence agencies and integrating their work in an overall institutional structure — these deep problems are all extremely relevant to today as Washington increasingly can align on just one thing, hostility to China. Given this history we should not bet on the Washington deep state outperforming the PRC on intelligence and in many areas it seems the PRC has learned lessons from America’s victory over the Soviet Union better than Washington learned them.

VII: On RV Jones, Scientific Intelligence in World War II, how Whitehall vandalised the successful system immediately after the war. Many issues explored in the RVJ blog are relevant to those interested in the future of AI, ‘safety’, and security.

VI: Alanbrooke diaries, incredibly relevant to today’s problems and what military ‘strategy’ really is.

V: Colin Gray and defence planning. What’s the difference between ends, ways, means? What’s the difference between strategy, tactics, operations? Why such confusion? What is defence planning, how does it fit with strategy?

IV: Notes on The Kill Chain — US procurement horrors, new technologies, planning for war with PRC.

III: More on fallacies of nuclear thinking / strategy / deterrence. If you read this and the earlier one you’ll see that almost everything the media says about Putin and nuclear threats is wrong / misguided and, worse, so is much of what is said by international relations/historians/military academics.

II: Thinking about nuclear weapons

I: On innovation in militaries, when does it succeed/fail — e.g why US got ahead on aircraft carriers, RAF defence in 1930s.

Prediction: 1) lessons from UKR will overwhelmingly support the arguments of those who in 2020 argued for radical MoD changes (including taking money from old tank projects that everybody privately admitted were a multi-billion pound disaster) and 2) the correct criticism of the review and connected documents will be seen as a) they did not go nearly far enough, b) the collapse of No10 follow through on defence reform in 2021 was — like the collapse of 2020 plans for planning reform, tax cuts, deregulation, Project Speed, intense focus on R&D and skills etc — a disaster for the country (and a political disaster for the Tory Party). [Me, 3/2022]

On rationalism and politics (2022).

And some other related stuff pre-No10…

On high performance government, ‘cognitive technologies’, ‘Seeing Rooms’, UK crisis management (2019)

On AI, nuclear issues, Project Maven (2019)

On the ARPA/PARC ‘Dream Machine’, science funding, high performance, and UK national strategy (2018)

On ‘systems engineering’ and ‘systems management’ — ideas from the Apollo programme for a ‘systems politics’ (2017)

On China vs US, the ‘Thucydides trap’ book (2017)

And obviously I think that if you’re thinking through AI and geopolitics you should study, or at least skim for a weekend, my chronology of Bismarck. A month of study and you’ll be in the top 0.01% of people who really understand high performance politics, an incredible shortcut! If you take this path, you will have a great advantage over your competitors.

---

Introduction

The government tried to stop Crossman publishing his Diaries in 1975. The case illustrated the inevitable tension between a) ‘government needs secrecy’ and b) ‘in a democracy the voters have a right to know a lot about how government works’. The Cabinet Secretary gave evidence.

The courts allowed the publication. The Labour government responded by strengthening ‘closed government’. Naylor writes of the prevalent attitude among senior politicians and officials that voters have little business in knowing about how they’re governed. One big question is: How and why has secrecy come to envelop the practice of modern Cabinet government?

Ironically Hankey himself was not allowed to publish his memoir of WWI, based on diaries, because of the development of secrecy he drove. For 20 years two of his successors and three PMs refused to allow publication.

Churchill obtained a suspension of the usual canons of Cabinet secrecy to write his memoirs of WWII. But Hankey only got permission in 1961. It was partly convention (e.g not to undermine the confidential relationship between ministers and officials) and partly legal, particularly the Official Secrets Act of 1911 and 1920. The Crossman affair was a denouement of the old system and attitudes. After, more was published.

But Naylor did not start his research with the theme of secrecy, he was looking at the whole story of the creation of the institution, the people who shaped it, and the domestic and political forces that influenced it.

The book considers:

• The pre-history to the crisis, the Committee on Imperial Defence, failures in WWI etc.

• The creation of the CO and Secretariat after the crisis of 1916.

• The Lloyd George regime.

• The crisis when LG fell in 1922.

• The Secretariat post-1922 until Hankey left in 1938.

• The custody of Cabinet secrecy from Hankey to John Hunt.

Cabinet minutes, when they started with the CO and Secretariat, were prepared to record agreement, not promote controversy.

The CO started keeping its own records. Cf. Wilson’s The Cabinet Office to 1945 — which I’ll cover next.

Hankey and his deputy Jones kept diaries available for inspection though their successors did not (or did not admit it).

A large collection of documents Hankey organised, intended for a book he never wrote, are in the PRO.

---

Ch1: The origins of the Cabinet Secretariat

[These notes are chronological, the chapter isn’t. It’s much easier to understand chronologically so I’ve jiggled some stuff between chapters. I’ve added some background to make the story easier to understand.]

The Committee on Imperial Defence (CID)

There was a Cabinet Defence Committee (CDC) set up in 1895. Little is now known about it because, like Cabinet, it kept no records.

There were two main criticisms of the CDC .

1. Poor use of intelligence sources.

2. Failure in defence planning, i.e how to settle the broad principles of national and imperial defence upon the basis of information from all the interested Departments, and set out in principle the size and composition of the military and naval forces necessary for such a defence policy. Those who set up the CDC had assumed a measure of inter-departmental coordination was possible without agreed upon records of planning and in the absence of any permanent staff.

In 1900 there were complaints that a lack of records meant a failure to apprehend decisions. These complaints were fended off by the service departments who argued that records would turn the CDC into a ‘court of revision’. The Admiralty and War Office resisted surveillance by other Cabinet members.

The CID was set up in 1902 in the aftermath of problems in the Boer War.

By 1902 the First Lord of the Admiralty thought that a stronger committee that backed more naval spending would help get the money so the Admiralty changed position. Balfour set up the CID.

• Regular meetings.

• Limited membership. PM, First Lord, SoS for War, four military advisers.

• It circumvented established conventions of Cabinet secrecy which prevented businesslike processes. It had a clerk from the FO to keep records.

The Esher Committee reported in December and recommended a full time staff. Esher wanted to strengthen the CID into ‘a department under the PM’ acting as a General Staff for the Empire. In May 1904 the PM authorised this but called it a Secretariat. It was intended to provide continuity in defence planning regardless of changes to the government: ‘it is not safe to trust matters affecting national security to the chance of a favourable combination of personal characteristics’ (Esher report).

So record keeping and a permanent staff were accepted for defence planning by 1904 but it took more than a decade to establish a similar need for the conduct of Cabinet business.

NB. CID was explicitly the PM’s entity, not a Cabinet (or Cabinet Office) entity. This distinction is very important. Balfour insisted that CID was formally advisory but it inherently eroded departmental autonomy and enhanced the super-departmental authority of the PM.

BUT CID did not become the powerful agency Esher had hoped because Balfour did not insist that CID conclusions should be the basis on which the War Office and Admiralty shape their roles and plans.

• So, for example, the CID conclusion in 1905 that Britain need not take seriously the prospect for invasion did not influence services’ military planning.

• CID made inquiries, wrote documents, made conclusions year after year but the government did not force the military to plan accordingly.

• CID staff played no part in staff discussions with France and the CID did not learn of such discussions until 1911.

• The first senior defence planner, George Clark (later Lord Sydenham), tried to play the role envisaged by Esher but was frustrated under Tories and Liberals.

CID therefore did not solve the fundamental issues of defence planning — the tension between policy and capabilities.

Naylor writes the CID’s failure was determined by the departmentalism inherent in 19th C government. But I would add — Balfour and Asquith could have insisted on CID operating differently but chose not to, they allowed the normal departmentalism to continue and did not exert their authority to enforce change on the critical things.

Hankey became secretary to the CID in 1912, aged 35. He would be at the heart of power from 1912 to 1938. He worked with the system as it had evolved for a decade and accepted its limits.

The processes and structure of CID for a decade provided an institutional model that would be considered and adapted by ministers and Hankey during the war.

Asquith appointed Hankey secretary to the War Council formed in November 1914.

Failure in WW1

In 1915 Asquith complained that the war had produced no great generals and Lt-General Wilson shot back, ‘No, Prime Minister, nor has it produced a statesman.’ Many complained about the ragged discussions in Cabinet, Asquith’s prevarication and lack of energy. Leo Amery recorded Edward Carson’s contempt:

He is very depressed about the hopelessness of the present system of governing by 22 gabblers round a table with an old procrastinator in the chair. (21 July 1915)

When Bonar Law had joined the Cabinet he told Asquith he was ‘astonished’ at the system:

[W]hen he joined Asquith’s Cabinet [5/15] he was astonished at the lack of method, the absence of any agenda or minutes. He told Asquith this. The latter said that everyone who joined the Cabinet made the same observation but speedily became reconciled to the method of doing business and saw its advantage for the special purposes.

Asquith continued to oppose even having minutes.

Years after Hankey’s innovation of Cabinet minutes (below), a historian wrote to him referring to 18th Century secretarial practices. George III was sent a frequent Minute back when the Cabinet’s location was often a private house. Hankey did further research and found traces of minutes going back to medieval years and the Privy Council in the 14th C. He was unaware that records of Cabinet discussions were kept for private purposes by several ministers in the Gladstone/Disraeli era. Salisbury would often gather papers and burn them in the fireplace in the Cabinet room. [I checked and confirmed the fireplace is no longer operational, just as the clock no longer works.] In fact the Minute turned into a ‘PM letter’ in the 19th Century. Hankey thought the last Minute of the old type was Melbourne’s in 1839. The Royal archive’s collection of PM letters starts in 1868. [Naylor implies there is a gap of decades in the records between the old Minute and the new ‘letter’: true?] It seems copies were not kept by PMs.

The most famous expression of frustration with the lack of clear records comes from the 1882 inquiry of Lord Hartington’s secretary directed to Gladstone’s PS:

Harcourt and Chamberlain have both been here this morning and at my Chief about yesterday’s Cabinet proceedings. They cannot agree what occurred. There must have been some decision, as Bright’s resignation shows. My Chief has told me to ask you what the devil was decided, for he be damned if he knows. Will you ask Mr G [Gladstone] in more conventional and less pungent terms?

Some ministers would return to their departments and give a note to officials but this was entirely ad hoc.

The first attempt at change was the War Council established as a supplement to the Cabinet. Then there was the Dardanelles Committee and the War Committee. Both failed. Part of the problem was size. Kitchener and others were understandably reluctant to discuss military secrets ‘with twenty-three gentlemen with whom he was barely acquainted’. Naylor says gossip was rife, security lax and the Times routinely published breaches of Cabinet secrecy. The Cabinet either duplicated the work of its War Committee or engaged in arguments based on limited military appreciations.

Even before the Somme, LG told CP Scott of the Guardian, ‘We are losing the war if we have not already lost it.’ Hankey was appalled at Cabinet dithering over Dardanelles and Gallipoli, recalling in his diary:

The Government are really dreadfully to blame. They put off decisions, squabble, have no plan of operation, and allowed themselves to be dragged into this miserable Solonika affair at the tail of French domestic politics. I can see only one solution — to suspend the constitution and appoint a dictator. (12/15)

[Sounds just like covid and Ukraine.]

And the War Committee was no solution:

The War [Committee] work is hopelessly congested, great questions all urgently awaiting settlement. Yet I could not get a meeting for tomorrow because Runciman was going for a day’s shooting, Lord Curzon for a week-end, and Lord Crawford to address his former constituents. I managed to get a meeting for Monday but the PM said ‘You won’t get anyone.’ Today's meeting had to end soon after 1 pm to enable ministers to attend official luncheons. Thus and thus is the British Empire governed at a critical stage of the war. I’ve done all I can to get meetings; crystallise woolly discussions into clear-cut decisions, and to promote control — but the task is a Herculean one! (11/16)

Even Asquith accepted the War Committee failed. He thought it:

1. Too big.

2. ‘There is delay, evasion, and often obstruction on the part of the Departments in giving effect to its decisions.’ [An often heard phrase among deep state officials is ‘it’s the usual — consent, delay, evade’.]

3. It’s often ‘kept in ignorance by the Departments’ of vital information.

4. It’s ‘over-charged with duties’.

Asquith though did not make any serious attempt to improve the bureaucracy and himself was a block on conveying vital information.

An insight into the evolution of the system is the Dardanelles Committee (May-Nov 1915). It was a Cabinet Committee so supposedly should not have had records but the ministers appointed a secretary and Hankey ‘succeeded in gradually infiltrating it with the Chiefs of Staff and all the procedure of the CID’. Some of these documents were used by LG when he attacked Asquith at the traumatic meeting of the War Committee on 1/12/16.

Crisis, November-December 1916

As the Somme continued in November 1916, allied leaders gathered in northern France. LG, now SoS for War, and MH were there. LG later said that MH took the initiative in discussing a major reorganisation:

We both felt that nothing in the way of a change in the conduct of the war had been accomplished and that in the absence of some dramatic coup things would go on as before until we slide into inevitable catastrophe.

LG wanted to resign. MH opposed and instead suggested that LG insist on a ‘small War Committee … for the day-to-day conduct of the war with full power’, independent of Cabinet and not managed by the PM who has a ‘very heavy job looking after the Cabinet and attending to Parliamentary and Home Affairs’. The Chairman must be ‘a man of unimpaired energy and great driving power’.

By the time of the Somme, influential Tories such as Earl of Derby and former PM Balfour were prepared to withdraw support from Asquith and even support Lloyd George. By the time of Asquith’s fall there was also an appetite for a serious change of institutions.

Curzon said that when the old system collided with the war it ‘crumbled into dust at once’.

I do not think that anyone will deny that the old Cabinet system had irretrievably broken down, both as a war machine and as a peace machine.

There was no agenda, there was no order of business… No record whatever was kept of proceedings, except the private personal letter written by the Prime Minister to the Sovereign, the contents of which, in any case, were never seen by anybody else. The Cabinet often had the very haziest notion as to what its decisions were. (1918)

Lloyd George’s first move was not to remove Asquith but to demand a new War Committee with full powers and a general reform of the constitutional machinery. Asquith would remain PM but direction of the war effort would shift to others. LG, MH and some Tories worked on the proposals and negotiated with Asquith. MH was ambivalent — he wanted Asquith to stay as PM and was nervy about LG’s character and methods (‘brilliant but often unsound’!) but also wanted a fundamental change in the war’s direction. A leak to the Times of Asquith’s imminent neutering scuppered the deal. Asquith resigned, Bonar Law deferred to LG who became PM.

The re-organisation based on Asquith remaining was dropped. From 7 December 1916 the old system was replaced with what LG described as ‘virtually a new system of government in this country’:

1. A War Cabinet of five including the PM. Only Bonar Law had departmental responsibilities. Milner, Curzon and Henderson had no executive tasks. Milner said the group would have to ‘do a lot of Thinking and Deciding’ and have precious little else to do. For MH, it all depended on LG — no one could say these four were ‘the wisest hands to win the war — two are really feather heads’, it’s a ‘mere political expedient of the most transparent kind to tide over a difficult crisis’. Carson said of its first meeting — which he attended but was not a member — that it did more work in 7 hours than all previous meetings together. It often worked from 11ish until the evening.

2. Hankey was Secretary to the War Cabinet. A duty was to write a record of proceedings and decisions. He was the first outsider to attend a Cabinet meeting to record its proceedings (see caveat below).

3. A Cabinet Secretariat to support Hankey and the War Cabinet. It prepared the agenda, organised attendance and papers and records of meetings. It communicated between the War Cabinet and the departments. Hankey had four Assistant Deputies. Young, Tom Jones (a Welsh confidant of LG), Clement Jones and L.C Amery. Staff from the late War Committee were transferred over. Tom Jones became Deputy Secretary until 1930 and was the most important official other than MH. TJ’s role was to act as ‘a fluid person moving about among people who mattered and keeping the PM on the right path’. MH did not try to deny Jones the wide scope the PM had given him. [Such ‘fluid people’ are critical and much depends on their talents and character. TJ was an official but his job was more like a modern special advisor and he did things for PMs which today would not be allowed for an official, such as helping with ‘political’ work.]

4. A network of Cabinet Committees.

MH set out early an important principle:

It is of the upmost importance that the responsibility of the Departments should be in no way weakened or overridden by the development of the Secretariat … [which] is neither an Intelligence Department nor a General Staff, but a machine for the service of the War Cabinet in co-ordinating the action of the responsible Departments.

MH initially envisaged a functional division of responsibilities along lines of machinery (e.g records) and ideas — ideas beyond the scope of ministers constrained by departments to help discussion in the War Cabinet by people without ministerial responsibilities. But this was short-lived and instead the Secretariat was divided along lines of military and civil with no ‘ideas’ function.

Instead ‘ideas’ were to be in a new PM’s Secretariat, an entirely new organisation to be housed in the No10 gardens (and nicknamed ‘the Garden Suburb’) while the Cabinet Secretariat was housed in beautiful houses in the Whitehall Gardens cul-de-sac. This looked broadly at issues other than military. The two Secretariats worked closely but in time MH grew unhappy.

(Whitehall Gardens, built in 1808, had beautiful houses where Disraeli and Peel lived which had gardens leading down to the Thames, until they were cut off by Victoria Embankment later in the century. It was built in the old Privy Garden of the Palace of Westminster, painted by Canaletto in 1747: see below, Montagu House, on the right overlooking the Thames, (Whitehall, looking north), and Whitehall from Richmond House. The CID had been quartered in Disraeli’s old house (No2). Hankey commandeered some more of the houses and set up his offices, separated from the Embankment and Thames by long gardens — ‘no pleasanter spot for an office in the whole of London’. Whitehall Gardens was demolished in 1938, the same year Hankey retired, to make way for new government buildings, a ‘monstrous act of vandalism’.)

Whitehall and the Privy Garden looking north from Richmond House, 1747

Residence of Peel, 4 Whitehall Gardens (and where he died after falling off his horse)

Secrecy and distribution. MH did not use shorthand but did make detailed notes. There was wrangling over the years as ministers adapted to some records being kept and there were arguments over the extent to which individual comments should be recorded or only decisions etc. MH opposed a full record of comments because, e.g a function of the minutes was to give to other officials and they should not see that minister X said Y in opposition to the agreed decision (or collective responsibility was undermined).

You can see the power of the notetaker in reflections upon how MH did the job — he could elaborate a conclusion ‘which had not been expressed in so many words by anyone at the meeting but which was accepted afterwards as representing the outcome’, and people would joke that they couldn’t tell what decision had been reached but must await MH’s minute. There was new formality around papers submitted to Cabinet and inevitable problems of balancing the need for security with the need to share with senior people responsible. And MH was inevitably caught between the PM not trusting people and those people badgering him for more information.

In 1889 HMG passed legislation (Official Secrets Act) to protect official secrecy, prompted by a 1878 incident in which a FO clerk memorised the text of an Anglo-Russian treaty and sold it to a London newspaper for £40. Over 20 years only one prosecution brought under the OSA was for a matter other than military/naval affairs. Complaints about leaks by Cabinet ministers was regular, including by Asquith during the war.

MH’s general approach was to err on the side of sharing too much rather than too little with ministers. But he also guarded Cabinet secrecy. The addition of notes and better communication did not mean the end or dilution of Cabinet secrecy. In 1916 there had been a formal investigation into the Dardanelles disaster. The inquiry requested notes of War Council discussions. MH got Asquith to refuse handing them over, though in the end a compromise was agreed after some threats of imprisoning ‘in the Tower’ from the Inquiry commissioners — the lead of the inquiry was allowed to peruse the documents to satisfy himself that nothing substantial had been hidden. MH’s argument was that release of the Notes from a sub-Cabinet body would compromise collective agreement — secrecy must be preserved so people could continue to talk freely unafraid their comments would be used as evidence against them. MH also kept his own detailed diary which LG was aware of. (There were odd occasions in the past when outsiders had been let into Cabinet, e.g 26/11/1882 when Sir Edward Hamilton attended.) [See below for reflection on my experience of this.]

Even good institutional mechanisms cannot solve fundamental political and strategic issues. The business of No10 and Whitehall improved. But the high level strategy of the war did not change much, attrition continued in Flanders. LG did not feel able until spring 1918 to replace the Chief of the Imperial General Staff, Robertson, and even then he did not dare replace the commander of the BEF, Douglas Haig. The friction around Cabinet and meetings was largely fixed but the friction between politicians and generals continued and relations deteriorated in 1917. The War Cabinet did not include representatives of the service departments and could not bridge differences between politicians and generals/services.

The military was frustrated that despite its advice the politicians kept questioning the ‘Westerner’ approach of Robertson and Haig. MH summed it up in December 1916 when he wrote that the War Cabinet was ‘really up against it as they don’t believe in Robertson’s ‘Western Front’ policy, but they will never find soldiers to carry out their “Salonica” policy’. The difference was not institutional, it was political-strategic, and could not be solved by ‘better process’. Naylor writes that LG did not have the political support from Tories to reject the military preference. The soldiers therefore continued with their ideas until LG demanded a change in spring 1918.

Hankey’s verdict on the change with LG and the new War Cabinet:

Gone were the scramble of Ministers to get their pet subject discussed at Cabinet meetings. Gone were the endless rambling discussions with no one to give a decision. Gone was the exasperating waste of time while the affairs of a department were discussed by people who knew little of the matter and had received no Memoranda on the subject. Gone were the humiliating and dangerous doubts of what the decision was, or whether there had been a decision at all. (Supreme Command)

---

Ch2: The Lloyd George regime

Asquith did not make much trouble for the new regime.

• The new Secretariats started work.

• LG spoke to Parliament less.

• There were more reports to the nation rather than reports to Parliament.

• One minister was not in Parliament (John Maclay).

The shift was shocking enough that AJP Taylor said ‘parliamentary government, as it had been known for the last century or so, ceased to exist’.

LG stressed that ministers remained responsible for departments to Parliament in ‘exactly the same way’.

Media criticism often confused responsibilities between the PM and Cabinet Secretariats.

Naylor writes that the ‘Whig theory of British history’ was a belief that the pattern of Cabinet government which had operated in the Victorian period represented the culmination of 1,500 years of progress towards representative government. The LG changes seemed to run contrary to this. And by spring 1918 when the war reached another crisis, many argued the changes had not been vindicated by improved performance. And, per above, the improved machinery had not solved the core political-strategic problems.

There were still often problems stemming from the inherent difficulty of problems, the characteristics of Ministers, and the disagreements over military strategy. On 18 March 1918 amid the military crisis, MH wrote of how the ministers kept avoiding the questions on the agenda, getting side-tracked, LG not initialling conclusions nor allowing MH to act without them, such that everything was getting clogged up and No10 a bottleneck.

Hankey personally was seen as contributing both to the development of the convoy system and the tank, two of of the most important innovations and both of which required overcoming strong Whitehall inertia. At the end of the war the Cabinet granted him £25,000 for his services and it was widely seen as reasonable. MH himself recorded that ‘I did as much to win the war as any of the Admirals and Generals’.

He went with LG to the peace talks where he was also widely praised for bringing order to the chaos of political meetings.

Now there came the question of how the wartime system would change with peace.

MH favoured maintaining a small Cabinet over a reversion to the old system. He thought the small size ‘coupled with the elasticity which enabled the arbiters to hear all sides of the case’ was crucial to the wartime success. He wanted to call the new variation ‘the Reconstruction Cabinet’. The Milnerites wanted to keep the war system and have an Imperial Cabinet. MH thought a Cabinet of 22 divided into imperial and domestic was a possible compromise but preferred the small option.

After the Versailles conference was over the Commons insisted on a change to the wartime system and in November 1919 normal Cabinet was restored quickly with the big departments returning to the Cabinet room.

Once the war system of a handful was politically impossible, MH preferred to revert to ~20 on the basis that 10-12 would just cause endless rows and whinging.

The Machinery of Government Committee chaired by Haldane recommended the continuation of the Cabinet Secretariat. MH’s own role continued. Suggestions about a new staff for the PM merging the two secretariats went nowhere for the moment and MH was opposed. (There were ideas about such a staff being an Intelligence organ. Naylor says in the long run this concept emerged again with the Central Policy Review Staff under Heath, although that was for Cabinet not the PM alone and it sat in the Cabinet Office. Cf. Stacey, 1975.)

Fascinating — Hankey was not made a Privy Councillor nor asked to swear its oath and he recorded in his diary that this meant although he would maintain secrecy ‘for so long as necessary’, he did not regard himself and his heirs ‘bound by an oath that I have never been asked to take’, ‘my memoirs will be the more interesting for my not having taken the Oath’, and if he had taken it ‘I am not sure that I should feel free to keep this diary’!

Minutes now were called Conclusions and were supposed to be shorter — MH assured Churchill that he would destroy his ‘pencil notes’ of discussion once the formal Conclusions were drafted. His assistant was banned from the room then readmitted. [The beginning of the creep towards ~10 non-ministers who now sit ‘against the wall’.] Specific conclusions were sent to Departments for action or information. Secret documents were circulated to named individuals. MH had tried to get the Cabinet to agree to return papers to the CO for him to keep in the event of resignation or death, but they refused and instead said that ministers could not make public use of documents without the permission of the King. The ‘immemorial custom’ (MH) of Cabinet had been that papers remained the personal property of Ministers on vacating office. In wartime it had shifted because of the number of secret documents. But now ministers reverted to the old rule. (CID in 1908 stipulated that keeping private files of CID papers was not allowed. CID continued this approach and ministers complied.) Remarkably, when LG left office he took with him the Irish Treaty of 1921 and kept it until the CO recovered it in 1945 after his death — and nobody had ever asked to see it!?! (And Naylor implies it was the only copy!?)

The OSA presumed guilt until innocence was proved and could be used in extremis. It was widened in 1911 and 1920 but it was a grey area the extent to which the OSA applied to ministers (see below).

Also in 1919 the Treasury’s role was strengthened and the Permanent Secretary of HMT was made Head of the Civil Service. This proved MH’s ‘most daunting problem in the immediate post-war years’ (Naylor). HMT officials claimed a precedent going back to 1867 but this was a formal specified statement. Both the creation of the CO and the HMT takeover of the civil service reduced the powers of departments and strengthened the powers of the PM over and concerning them.

HMT seconded an official to sit in the CO to resume ‘the old established arrangement’ of departments thrashing out with HMT ideas with spending implications. MH agreed.

In 1921 HMT proposed shifting budgetary provision for the CO from within Treasury to a separate status. MH opposed as he wanted the CO to stay hidden from Parliamentary scrutiny but he lost the argument. MH viewed the PM as Head of the CO as well as First Lord of the Treasury and said he should have only one department, accepting a view of HMT as a foster parent of the CO. His fear of scrutiny was clearly substantial given he objected to more theoretical autonomy. (Naylor does not explain why Fisher wanted this, given Fisher’s desire to take over the CO, see Ch3.)

The post-war Cabinet Office consisted of:

• Cabinet Secretariat

• a much smaller Secretariat for CID

• an office responsible for publishing Official War Histories (the CO kept the documents so it was natural)

• an office responsible for communication with the League of Nations [moved to the FO shortly but an early sign of the post-1918 tendency for the PM to encroach on the old duties of the Foreign Secretary].

Functionally it was divided into home (led by Tom Jones) and external affairs covering imperial, foreign, and defence affairs. MH tried hard to get the CID restored and it met in June 1920 [edited from 2020, thanks to commenter] for the first time in 5 years.

There were 123 staff in 1922, a score less than 1919. [The CO cannot tell you today how many people work there but they had a communications office about as big as this recently.]

In autumn 1922 the LG government fell. There were rows over foreign policy and the PM’s style. The SoS at the FO, Balfour, had been excluded from the War Cabinet. The FO had been diminished. There was FO resentment at MH’s role and influence. (MH was also a student of phrenology which influenced him particularly on the characteristics of the Poles!) LG had resented FO secrecy pre-war and now often left diplomats in the dark about his activities. Curzon was not much of a check as he wanted to stay in office and deferred to LG on many issues but their relationship deteriorated.

A crisis blew up over Turkey in what came to be known as the Chanak crisis. LG was pro-Greek. Baldwin was appalled at LG’s Cabinet and watching its meetings, ‘I felt I was in a thieves’ kitchen’, nobody seemed to have any principles and there was ‘the most awful cynicism’; LG’s methods were ‘a profound threat’ because of ‘the wide appeal which apparently easy, authoritarian solutions had over the slow processes of democracy’. Cabinet was divided. The PM’s Secretariat was in chaos with no effective head. The public was indifferent. The press was hostile. Much of Parliament was critical. Public statements from No10 were seen as reckless. Handling of the Dominions was poor.

The Cabinet Secretariat got some blame for some for the debacle. Naylor says this was unfair: Curzon lost the argument in Cabinet, there was no secret cabal in the CO. The fall of LG was for deeper political reasons. Curzon summed up the situation as there being ‘two Foreign Offices’ and the No10 one did not communicate with the other. Underlying everything was that LG’s conduct of affairs was personal, secret, arbitrary. People disliked his handling of the press and personal corruption including sale of honours. While much could be swallowed during the war crisis now people were fed up of it. LG joked that while he could ‘conceive of circumstances arising in which I might be compelled to act on principle’, nobody could say that of FE Smith or Churchill! (LG said it was better to sell honours than policies like Tories and Liberals.)

LG recognised the shifting political sands and could have resigned and made it easier to make a comeback but Beaverbrook said ‘the glitter of his supreme office held him in chains and he could not bear to give up power and patronage’, it was a weakness ‘and he knew it was weakness’. Although LG maintained his grip on Cabinet, the lower ranks of Tories had had enough.

My experience is that all Prime Ministers suffer by suppression. Their friends do not tell them the truth; they tell them what they want to hear. It was so with Asquith… George was also misled. People are always apt to think that what had been will be. (Bonar Law)

The fate of the Cabinet Secretariat now hinged on Bonar Law’s attitude to whether it had contributed to this ‘suppression’.

---

CH3: Crisis 1922

Parts of the press had started criticising the Cabinet Secretariat — ironically some criticism for supposedly encroaching on policy but some for failing to coordinate policy and allowing a shambles! MPs gave speeches in 1922 about the dangers of the Secretariat and the growth of the PM’s power, the dangers of ‘the Presidential system’, which threatened Parliament and the constitution. It was defended by Austen Chamberlain as the mechanism of Cabinet control — it ‘has no executive function, has no administrative function, displaces no other Department’. LG described it as ‘a communicating Department, they are a means of transmitting to Departments the decisions’, it is ‘purely a recording machine’ and has ‘nothing whatsoever to do with any question of policy’. And LG explicitly linked the ‘old dignified ways’ to ‘the cataract of 1914’ and said ‘the world wants a change from the old methods’.

[Obviously the development of the CO has proved critics right in the sense that it has usurped ministerial powers, it very much is involved in policy, it has gone far beyond secretarial work etc.]

The PM’s Secretariat also caused trouble with Cabinet and press. Beaverbrook ran a campaign against MH and deliberately confused it with the Cabinet Secretariat. It was not popular with departments because LG used it to fend off ministers who wanted to talk to him.

MH had a strict policy of the Cabinet Secretariat having no contact with the press. Ministers still sometimes blabbed and the departments had bigger press offices so more leaked out there. And there were suspicions and worries that a new government might use records from the previous government to discredit them.

Overall, the new institution was associated with LG’s leadership and habits therefore there were questions about whether it would survive his departure.

Balfour dismissed much of the criticism of LG’s ‘personal rule’ — PMs are always attacked either for being ‘figure-heads run by abler men or tyrants. There has never been a middle position.’

After the Carlton Club revolt and the collapse of LG’s position, MH thought a reduction in the Secretariat’s staff inevitable. MH quickly divested himself of things like the League of Nations to the FO. (He retained some loyalty to LG. Tom Jones said of LG that one could say fifty things about LG, all true and contradictory.)

There was a lot of press speculation on the end of the LG ‘system’. And MH wrote that ‘the whole hierarchy of the civil service’ is trying to ‘down me’. MH had to persuade Bonar Law to keep the Secretariat and defeat Warren Fisher, Permanent Secretary at HMT who wanted to absorb the Secretariat. [No surprise, a Treasury power grab!]

Bonar Law told him that he wanted the recording of minutes to continue but the expense of the Secretariat to be reduced. MH agreed. Fisher argued that HMT was ‘the central Department of the Government’ therefore the Cabinet Office must be an integral part of HMT. MH argued that the Cabinet Secretary often had to deal with disputes between HMT and departments and had to be seen as independent, not a Treasury official. They came to a deal. But Bonar Law gave a speech, in the election campaign, in which he said the Secretariat ‘must come to an end’ — but we must retain an agenda and ‘a definite record of our decisions’. The Cabinet Secretary and ‘whatever help he needs’ should be part of the Treasury which is ‘the central department of government’. This confirmed exactly what MH had wanted to avoid — the Secretariat being absorbed into HMT.

Then Fisher proposed that MH continue as secretary to the Cabinet and CID and become clerk of the Privy Council (this role being suddenly open as the previous Clerk had been convicted of ‘bothering women’ in Hyde Park and ‘even if he wins his appeal will probably have to retire’!).

The proposal eases everything. The Cabinet is constitutionally a Committee of the Privy Council, as the CID is of the Cabinet. The whole arrangement is symmetrical and logical but one that I have often thought of. Fisher then astonished me by saying it had been in his mind throughout…

[HMT] had shown neither foresight nor organising capacity, and had plunged the country into something like administrative chaos until the War Cabinet machinery came to the rescue. This he [Fisher] did not dispute and even strengthened what I said.

This arrangement also meant that MH would brief the king (as he had about CID pre-LG) and it would solve the issue of someone being in Cabinet who had not sworn the Privy Council oath of secrecy.

MH spoke to Bonar Law with a prepared resignation statement. Bonar Law read it, said he was ‘too sensitive’, and agreed MH’s requests:

• to defend him in Parliament against the many attacks,

• the Secretariat would not ‘be swallowed by the Treasury’ and

• he should have his own staff which would be reduced to a minimum.

Fisher largely surrendered, asking only that MH move office from Whitehall Gardens which MH agreed to but then did not do. Fisher officially noted that he regarded the logical step to be the absorption of the CO into HMT but he would not push it while MH was in post.

Fisher consistently held that he had deferred to MH’s personal position and he did not accept the Secretariat’s status long-term. But the matter was essentially closed until MH stood down in 1938 and the fight resumed. The two worked together in important ways on rearmament in the 1930s.

HMT maintained that a Blue Note in 1887 had made clear the role of the HMT Permanent Secretary as also head of the civil service but Baldwin had to admit in Parliament (24/2/1926) that while the association of the two functions dated to 1867, the file containing the Minute and associated papers had been lost for fifty years! [Has this Blue Note ever been found?] But the 1919-20 reorganisation and confirmation of Fisher’s dual role helped put things on an official footing. Bridges, a later Permanent Secretary of HMT who also became Cabinet Secretary, took a different view to Fisher and thought the Secretariat should be independent of HMT because of its role in handling disputes with HMT.

Fisher saw the civil service along with the Army, Navy and Air Force as one Service of the Crown with four Divisions. He also abolished the Home/Foreign Civil Service distinction in the pursuit of a unified Civil Service —

… a single efficient entity, a unified machine in which advancement came by merit, whose standards of conduct were of the highest and whose administrative capabilities would not be impeded by jealousies or red tape.

Many saw in Fisher’s general bureaucratic victory and the departure of LG a win for the Insiders against the Outsiders. Thomas Balogh, in The Apotheosis of the Dilettante (1969), said of Fisher’s vision that when during the war something hard had to be done to avert disaster, outsiders invariably had to be brought in!

Bonar Law drew an analogy with Rome which had two different modes for government — peace mode and war mode. During the war, they had had to centralise with the War Cabinet etc but now it’s peace we should ‘do as the Romans did’ and go back to the old peacetime ways. But after the election Bonar Law tempered his criticisms and the Secretariat carried on much as before though a bit smaller (102 to 63 staff). Bonar Law also described his view of the PM job as different to LG who had decided most things himself — he would act more like the head of a large business where most things are done by others and he would provide ‘general supervision’.

MH would have a run-in with Curzon who attacked the Secretariat publicly then asked for their help in providing support because the FO couldn’t do it promptly — ‘You abuse us like that then sponge on our efficiency because the rotten Foreign Office cannot provide an efficient stenographer’.

[NB. Naylor barely mentions it but another important report was the Haldane report on government machinery handed down in December 1918.]

---

CH4: The Secretariat in the 1920s: policies and procedures

Hankey was seen as impartial by key players. E.g he could be asked for records to resolve disputes and people trusted him to be straight.

MH had to deal with a slew of memoirs of the war and disputes. Ministers and soldiers wanted to rebut charges. There was no formal system for reviewing ministerial memoirs. Churchill discussed his account with Asquith and this approach became institutionalised. MH pointed out to people that he had no formal role and could exempt nobody from either the Official Secrets Act nor their Privy Council oath. He suggested that only the serving PM could absolutely OK revelations about Cabinet discussions.

He also had many problems with the Official Histories of WWI.

In 1889 HMG passed the Official Secrets Act (above). Over 20 years only one prosecution brought under the OSA was for a matter other than a British citizen disclosing secrets. It was widened in 1911. The first section concerned espionage. Section 2 construed as an offence the action of any person who, having any information obtained ‘or to which he had access owing to his position’ in office, ‘communicates the information to any person other than a person to whom he is authorised to communicate it’. In the cursory debate of 1911, S2 was never discussed. In 1919 the courts ruled that it applied to any information, not just secret data, obtained through an official position. The 1920 OSA amendment was largely about espionage. Naylor writes that the Attorney General made a hash of explaining it in Parliament and this reverberated for decades. In 1972 the Franks Committee claimed that it had always been the intention for the OSA to operate against leaks of all kinds. [This wide interpretation clearly extends extremely far from a measure to stop espionage or the disclosure of secret and important information.]

It was a grey area the extent to which the OSA applied to ministers. Many ministers and PMs (including Asquith and LG) acted as if it clearly did not apply to them. The OSA was not invoked to stem the flood of wartime memoirs. And it has not stopped the flow of ministerial memoirs since.

From 1934 ministers have supposedly not retained comprehensive collections of Cabinet papers, as was normal pre-war.

Bonar Law dispensed with the PM’s Secretariat and reverted to the Private Secretary system. BL got throat cancer and was replaced by Baldwin. There was no warm personal relationship between Baldwin and MH but there was mutual respect and the Secretariat survived another change of PM.

MH was not consulted about the snap 1923 election over Protection and Baldwin consulted with no officials on the matter before launching the campaign. Tom Jones stayed on as a speechwriter for BL and Baldwin. He opposed the policy — was a fierce liberal and a free trader professor of economics — yet wrote SB’s campaign speeches on Protection, a partnership MH thought ‘the most extraordinary feature of the election’.

In June 1920 CID met for the first time in 5 years. A Standing Defence Sub-committee was charged with surveying defence commitments and this committee in effect carried on CID’s business. The Standing Defence Sub-committee was tasked in March 1923 with a review of the co-ordination of the three Services. This committee spun off another sub-committee chaired by Balfour on whether the Navy would gain complete control of the Fleet Air Arm or whether it would remain part of RAF. (It was given to the RAF but the Navy fought on.) The Sub-committee recommended that the Chiefs of Staff of the three Services constitute a committee of the CID to discuss questions affecting their joint responsibilities. The COS Sub-Committee thereby acquired functions which had been intended but never secured for its parent, the CID. It was a victory for Air Marshal Trenchard who had pushed for this since 1919. MH became Secretary to the COS committee too. Chiefs of Staff came and went — MH remained!

MacDonald continued the work of the Secretariat and CID in 1924. This brought MH into contact with the Webbs. Beatrice Webb recorded her impression:

An attractive personality, trusted and liked by all Cabinets in succession for the good reason that he likes them and is absolutely loyal and amazingly appreciative of the different statesmen he serves. A simple-minded soldier of the conventional type, devout Christian, a puritan in habits, a perfect gentleman in manners. He assumes that the men he serves are public-spirited however they may differ in opinion and capacity. He has plenty of shrewd intelligence, but no intellect; abundance of good temper but no wit; irony or sarcasm would, I think, be inconceivable on his own part and somewhat unintelligible in others. Hankey, like other simple-minded persons, mistakes power over other people for real distinction of thought and feeling. But this lack of censoriousness – this slightness of critical facility, combined with absolute integrity, kindness and loyalty and quickwitidness, make him an ideal secretary to Cabinets. If ever a man was perfectly suited to his job, it is Maurice Hankey.

Little changed. MacDonald asked the CO to inform the media of the topics discussed at Cabinet, which MH didn’t like and after Labour fell the practice was stopped.

The Campbell affair brought Labour down. The details are tortuous and not relevant (pp141ff). Nutshell:

• A journalist published a piece in the Communist Party paper Workers’ Weekly calling on the armed forces to organise to smash capitalism. The DPP and Attorney General agreed it was a crime. The DPP secured a warrant for Campbell’s arrest, then after publicity the AG thought he’d made a mistake and wanted to u-turn.

• Now he brought in the PM and Cabinet. There was a Cabinet discussion. Cabinet decided that ‘political’ prosecutions should not proceed without Cabinet agreement. And it agreed with the u-turn on the basis of AG comments on a letter from Campbell (which turned out not to exist!). The entire discussion sounds shambolic and just like a modern Cabinet discussion of legal advice!

• There was then a political furore in which MacDonald claimed to MPs he had never been involved and the entire matter had been at the discretion of the Law Officers.

• But many people knew it had been discussed in Cabinet. MacDonald challenged the note of Cabinet by Tom Jones. But Jones also had original verbatim notes which he had not destroyed (interesting too!). MH thought the PM’s statement to Parliament ‘a bloody lie’! So a classic political problem was entangled with the duties of the new Secretariat to record Cabinet discussions and the records might lead to the fall of a government.

• In the tortuous argument, MH went into great detail and tried to persuade the PM his memory was mistaken and he had indeed approved the Minute he now said was wrong, and MH could prove that the Minute had been circulated to Cabinet and nobody had disputed the record.

• The PM would not admit error. He lost a vote. The government fell.

• MacDonald maintained the whole affair was an ‘extraordinary series of muddles’ which was true! But he also never admitted some established facts and he blamed the Secretariat to others after his fall.

Interestingly, Jones suggested to MH that MH try to influence the King NOT to grant an election — an example of how already the Cabinet Secretary’s position was informally accumulating power in ways unintended by LG or other PMs.

In the election that followed MacDonald was undone by the infamous Zinoviev letter. This is also too complex to go into in detail. Nutshell:

• On 10 October a document arrived at the FO purporting to be from Zinoviev, head of the COMINTERN, to the head of the Communist Party of Great Britain outlining a plan to use a new Labour Government for a Bolshevik revolution.

• It took days to wind through the machinery to the PM who asked for views on its authenticity. The Permanent Secretary, Crowe, said it was genuine, others were unsure.

• As Whitehall deliberated and planned an official response, a copy was leaked to the Daily Mail which threatened to publish it. Crowe decided on his own authority to publish immediately along with the official government protest without the PM’s agreement. It was published four days before the election leading to a great scandal. MacDonald later claimed he could have been reached by phone.

• The Soviet government said the document was a forgery.

• MacDonald told his colleagues he felt like he’d been ‘sewn in a sack and thrown in the sea’.

• MacDonald’s campaign sank. He had no clear answer to the accusations.

• Labour ministers felt with justification they were the victims of a very dirty trick.

• The new Tory government assured the Commons it was genuine.

• Naylor writes that a) Hankey accepted the letter as genuine all his life, b) now the authenticity of the letter remains in dispute but c) there is no doubt the letter was planted in the press by the intelligence community and there was a ‘sordid’ intrigue between that community and the Tories.

• In 1999 a report was published with supposedly full access to MI5, MI6 and other secret files. It concluded it was probably a forgery written by White Russians and probably leaked to the Mail by MI6. There have also been suggestions it was an operation conducted by George Ball, MI5 agent and Conservative supporter.

A legacy of the red scares was the Chancellor of the Duchy drawing up secret plans to deal with events like a General Strike. MH played a role drafting Emergency Regulations 1926 (under the Emergency Powers Act 1920) issued by Royal proclamation at the end of April 1926 when negotiations over coal miners’ pay and conditions collapsed. However the courts declared the General Strike illegal — only the coal dispute was covered by the 1906 Trade Disputes Act — and the TUC folded. [The unions involved in the General Strike became liable by common law for incitement to breach of contract and faced asset seizures. The Trade Disputes and Trade Unions Act 1927 banned sympathy strikes, general strikes, and mass picketing, creating a system whereby trade union members had to ‘opt-in’ to paying the political levy to the Labour Party.]

The Baldwin Cabinet renounced the precedent that prosecutions of a political character required Cabinet sanction. Baldwin described it as ‘unconstitutional, subversive of the administration of justice and derogatory to the office of the Attorney-General’. The parliamentary debate on this created the modern understanding of the constitutional position of the Attorney-General. In exercising the enforcement of the criminal law the AG is wholly independent of the Government and responsible only to Parliament.

MH drafted departmental staff to help with Cabinet Committees. This cut down the need for more central people and ensured departmental experts were involved.

At this time the CO had three main functions:

1. The Secretariat, Cabinet records etc.

2. The CID.

3. Cabinet Committees.

MH repeated a critical point:

The function of the Cabinet Office is essentially one of machinery of Government and not of policy. It cannot be too strongly emphasized that the Cabinet Secretary has no duty of offering advice in any matter of policy or of interfering in any way with the functions of responsible Departments in this respect. (CAB 63/37, The British Cabinet Office.)

A fourth function was assigned to the CO in 1925 — the Committee of Civil Research (CCR), an effort intended to mirror the CID, operating neither as a Cabinet Committee nor a department. It could only investigate and recommend and had no fixed numbers or procedures, no executive office was under its control, no department was in obedience, nobody sat on it of right except the PM. Its function was to give ‘connected forethought from a central standpoint to the development of economic, scientific and statistical research in relation to civil policy and administration and it will define new areas in which enquiry will be valuable’. It had been foreshadowed in the Haldane report. It was set up under Labour but then Baldwin gave it to Balfour who understood CID and was also a patron of science and a pioneer in the coordination of government scientific research. Industrial research should not be seen as ‘a desirable luxury’ but essential said Balfour. 1925-30 it investigated many topics including the steel industry and electricity and nuclear research. It was the first attempt to create a research department at the highest level of government to recruit economic and scientific specialists outside departments.

It was closed in 1930 as it became caught up in the creation of the new Economic Advisory Council (EAC) by the Labour government in 1930 and power struggles with the Treasury.

NB. CID and CCR were not strictly speaking Cabinet Committees though they functioned largely as if they were and with the support of the CO Secretariat.

---

CH5: Twin Institutions

MH said in 1927 that he gave more time to the affairs of CID than to Cabinet.

The Admiralty had been angry when Labour did not authorise the construction of the Singapore naval base. Cabinet prioritised ‘international cooperation’ hoping restraint would lead to arms control. MacDonald implied to MPs that CID was not working properly and the Singapore decision was an example. MH was rightly angry. He told MacDonald that Singapore was ‘the worst case that can be taken’ to illustrate CID problems as ‘no question has been more exhaustively studied in the last twenty years’. MacDonald withdrew his criticism.

MacDonald also queried the influence of CID but MH reminded him that it did not take ‘decisions’, it resolved technical issues and provided ‘recommendations’ to Cabinet.

MH also recorded the proceedings of the COS Committee and spent a lot of time mediating disputes. The Tories renewed construction of Singapore but then there were great disputes about fortifications and between all and HMT.

MH and Jones disagreed in their advice to Baldwin on defence expenditure. MH argued that it was far better to invest in defence and keep people off the dole which ‘rots the morale of the people’. Jones preferred to spend on other things.

MH could influence discussion of things like the Locarno pact but he could not get his way if Cabinet ministers flatly disagreed. He also opposed discussions about a Channel tunnel. He fought to maintain Britain’s ‘Belligerent Rights’, i.e our longstanding claimed right to starve continental enemies in the event of war by closing naval traffic. He was enraged the Tories entertained what he regarded as stupid schemes for ‘freedom of the seas’ and babble about international law etc.

MH strongly opposed creating a Ministry of Defence and argued CID could do the job better and cheaper. He argued defence was too comprehensive to relegate to a ministry. CID could coordinate not just the Services but every Department and ‘in the widest sense of the term [CID] may be said to fulfil many of the functions of a Ministry of Defence’. MH was so committed to a revival of CID he said several times he would resign if it were replaced or changed much. Cf. a 1931 note he wrote arguing that such a MoD would go beyond the limit of ‘wise rationalisation’ to ‘over-centralisation’ which would ‘defeat its own object’.

He argued that the COS Committee determined priorities and could work with HMT to establish an overall figure for defence spending.

[In some sense, the NSC is a re-creation of aspects of CID in that it tries to coordinate across the entire system and can pull in any part of the state for discussion and coordination, including police. See below.]

MH thought that British governments had been too appeasing of America, giving way on issue after issue hoping it would improve relations. Instead we should stop making concessions. The American and Belligerent Rights issue came together when MacDonald visited Washington in 1929 to discuss the Naval Treaty. MH rallied the deep state to scupper America’s ambush of MacDonald and much-feared paragraphs were not announced! MH even broke his own rule on the press to brief The Times (off the record) on Belligerent Rights. Naylor concludes that ‘by present day standards, Hankey exceeded the bounds for civil servants’ but successive PMs did not see him as behaving improperly.

In 1929 MacDonald was PM again. Business continued mostly as usual though MacDonald wanted the Minutes to be sparser following his disaster years earlier (above)! (MH kept a separate series of notes with more details in his own files.) There’s only one recorded request 1929-35 from a minister to change the Minutes.

MH was pressed into service at the Hague Conference and was again appalled by the FO’s uselessness: they’d ‘done nothing: no organisation, no interpreters, … no accommodations booked, no nothing’.

The CCR could not develop synoptic planning because of HMT objections. But there remained a feeling among ministers something different was needed. MacDonald badgered for something like an ‘Economic General Staff’ operating like the CID but MH thought the analogy poor and likely to be a ‘feckless duplication of work already being done quite efficiently by departments’ which he wanted the CO to stay out of, and any ‘expert body’ independent of departments would just mean it was boycotted by departments including HMT. Instead there emerged by 1930 a committee comprised of officials plus outside experts, the Economic Advisory Council, with Jones as secretary. [These arguments have been repeated cyclically since, see Thoughts below.]

Labour returned to the Singapore base and ordered a go-slow again, enraging MH. Part of the justification was the morass of disarmament conferences taking place which Labour hoped would work.

At the London Naval Conference, Britain agreed to reduce our naval strength viz US and Japan. Churchill and others were very critical. The Admiralty stated bluntly that we were ‘reducing our strength as compared with the others’ and that full Imperial Defence was not compatible with the restriction from 70 to 50 cruisers. MH’s biographer concluded of the detailed discussion that MH fell victim to the diplomatic professional deformity of wanting a deal at almost any price and therefore agreed things he regretted.

Vansittart (Permanent Secretary at the FO) warned in May 1930 that Germany would rearm and the disarmament conferences would not stop it. MH agreed and started trying to undo the Ten Year Rule agreed in 1919 and confirmed officially again in 1928 and June 1930 by CID. In Jan 1931 MH sent a memo to MacDonald on military developments. The PM did not disagree with the analysis but felt that no action should be taken pending work of the Disarmament Conference. The COS concluded that developments in Asia should cancel the Rule and on 22 March 1932 CID accepted the COS view. But Cabinet insisted that the acceptance of the COS report could not justify increased spending on defence given the economic situation and they mandated further exploration of disarmament prospects.

20 months passed. After Hitler took power and dwindling hopes for disarmament, the COS view was put into effect formally by Cabinet in November 1933 and a Defence Requirements Sub-Committee (DRC) of CID was created, with MH as Chair, to plan a program for repairing deficiencies in armed forces: in this 20 months the Rule was theoretically dropped by the CID but in practice it continued.

By the end of 1930 the Depression had caused an economic and political crisis. Unemployment was over 2 million. The pound was under pressure viz the gold standard. Labour was stuck between demands for public spending cuts and their politics. The May report of May 1931 recommended spending cuts, benefits cuts and wage cuts.

The Cabinet fell apart. MacDonald said he’d resign but then formed a National Government. Labour felt betrayed. MH was sympathetic. He believed the crisis needed a cross-party/bigger than party solution. An interesting procedural anomaly preceded the new Cabinet: a Conference of Ministers met first and decided the new Cabinet would follow established procedures, i.e men NOT in Cabinet decided how the Cabinet would function and that it would be small (10) but with more getting the papers.

In a break with precedent the new government after the election (which decimated Labour) decided Ministers could publicly break with collective agreement on some economic issues.

MH was mostly an orthodox voice on the economy: Labour was spending too much, pensions and benefits were too high, we’re spending beyond our means etc. He wanted belt tightening and more for defence. He worried foreigners thought Britain ‘exhausted’ and unwilling to compete for vital interests. Interestingly a) he’d made a point of not voting since 1918, b) he voted for the National Government in 1931, c) we know this because the KING asked him about it and told him he should vote for the National Government!

The Secretariat continued as normal with the new government. Papers considered ‘exceptionally secret’ bore a special label in red print directing it be kept under lock and key, opened only by the named recipient, and copying was forbidden. This applied to ~2-3% of Cabinet papers in the estimate of MH’s deputy who organised this. Interestingly, budgets were discussed in Cabinet before they were announced but the Cabinet Secretary took no notes and no discussion was recorded.

LG threatened to publish Cabinet documents. MH assembled a legal case and informed LG’s secretary he would be prosecuted for breaching the OSA. LG backed off. MH also had to read volumes of LG’s Memoirs and suggest changes. LG was told he had to seek release from his Privy Council oath and guard against violation of the OSA. Constitutionally the King still had control over what Cabinet papers a former minister could quote from.

Ironically, MH himself would be denied permission to publish his own war memoir until 1961.

---

CH6: Hankey’s Last Years

There was a general principle that criminal law should not be adapted to ends not originally intended. This was ditched with the OSA. The second Labour government started using it for ends other than espionage and terrorism.

In 1925 the government has used the OSA to block publication of a book about Casement, executed 1916, based partly on diaries which the government suppressed for decades.

In 1932 it was used against Compton Mackenzie, a former British intelligence agent. He’d published two volumes of memoirs. He assumed there’d be no problems with the third. Then he was hit with OSA and advised by lawyers to plead guilty. He said he was told the government was using his case to send a message to LG and other ministers.

In 1934 a row over a book by Lansbury’s son using Cabinet documents provoked MH to action. One of the documents had been marked ‘most secret and confidential’. Hitherto there had been no formal rule or legal opinion on publication of Cabinet documents, it was all convention that publication required PM consent. (Baldwin had told MPs in 1927 that ministers were obliged to consult the government.) MH spoke to the PM and AG immediately. The issues in the papers were still very alive and contentious. MH argued it was a clear breach of the OSA. Legal wheels rolled. The publishers were told to recall all copies. Lansbury’s son was convicted of two misdemeanour counts under OSA. He was fined. But Lansbury himself was not prosecuted! (Perhaps it was felt that the principle only required the son prosecuted and the sight of a prosecution of a Cabinet minister by political opponents would look bad.) The prosecution worked: ministers did not publish Cabinet documents without permission and a formal rule had been established. (In 1944 the Cabinet Office confiscated papers from Lansbury’s family, the only case of outright confiscation according to Naylor and one which went far beyond Cabinet documents and cannot be defended.)

MH then used the case to argue that his Secretariat should acquire control of all Cabinet papers. CID papers were always collected from ministers and not left in their personal possession. Cabinet practice had been haphazard post-1918. MH argued that papers with ministers could easily be stolen by agents etc [obviously true]. And the assumption then was that Cabinet records would remain permanently closed, which held until 1966. They began to ask ministers to return papers but they could not insist. The Law Officers ruled that they held them lawfully but this right did not extend to heirs etc so papers were pursued when the minister died. MH deployed officials to request the papers from ministers and did not inform them of the Law Officers’ opinion that they could refuse!

It was also complicated by the fact that in 1931 the Law Officers thought the statement printed on Cabinet papers — ‘This document is the property of His Britannic Majesty’s Government. Secret, Cabinet’ — was legally dubious, and ‘property of His Majesty’ would have been more accurate! No change had been made. There was discussion about this but MH opined — ‘better to leave sleeping dogs lie’. The old heading continued despite doubts as to its legality. Many continued to confuse the meaning and significance of the two different phrases. It did become accepted that Cabinet papers were not the ‘property’ of ministers. And most ministers complied with the request, only half a dozen held out but were not pursued. LG and Churchill refused the ‘invitation’ and kept their papers. The Secretariat did not try to seize them from their heirs. Only the Irish treaty was recovered from LG because there was no other copy!

There was a tricky question of biographers’ access to papers of those dead — denying all access seemed unfair on those who happened to die (versus those alive and in possession of papers) yet creating rules seemed very hard given the constitutional and legal problems, so in very English fashion the Secretariat decided to make no rules but review cases individually. The Cabinet ended up discussing the subject and agreeing that relatives of the dead could be given access in some circumstances.

For MH, the system of Cabinet records was to help the efficient conduct of public affairs, not to help history or settle endless disputes between individuals.

[There’s a lot of historical detail in this chapter I’ll ignore here.]

A massive exemption was created again for war documents 1939-45. WSC wrote many ‘Personal Minutes’ which were seen at the time as intended for use in a post-war memoir. At the end of the wartime government, Churchill had Bridges (then Cabinet Secretary) draft an approach for those ministers: ministers could keep telegrams, minutes, documents circulated to Cabinet which they wrote and signed themselves and should be seen as ‘their personal property except that they will be bound by the rules governing the use of official papers which are well established’. This novelty regarding ‘personal property’ held for him and the wartime Cabinet. Attlee and the Secretariat then firmly closed the door again.

For a short time the CO tried to maintain that records of Cabinet proceedings should never be disclosed and were exempt from the Public Records Act 1958 which stipulated delivery of public records to the Record Office after 50 years. This could not be sustained and it became established that Cabinet records would be made accessible after 50 years [later 30].

Defence in 1930s

MH prioritised defence of the Empire in the East. This meant naval projects. The RAF was suspicious of him. Contrary to the view of MH and service chiefs, Treasury and other officials and ministers shifted some resources to air defence from the navy and army in 1934. In 1935 he worked on a new defence White Paper. Germany was rearming and introduced conscription in March 1935. MH thought the only way to square the circle was more money therefore public support and debate. There was serious tension between MH, Warren Fisher at HMT and Vansittart at FO. MH was deeply critical of Fisher’s papers though he did partly defend him in later years from criticism over his obstruction of parts of rearmament.

By the mid-1930s parts of the system started to revolt against MH’s dominance of CID. Liddell Hart told people MH had been there too long. Trenchard (resigned 1930) publicly stated that the secretaryship of Cabinet and CID should be separated. MH ‘deplored’ the suggestion. But criticism mounted in the mid-1930s of the defence system. He had warned of a Minister of Defence without a Ministry — the original idea floating around (see above). Now he stressed that it would pose a dilemma of dual control not just between the PM and Minister but between the Minister and the Fighting Service Ministries. He accepted the system for planning a war was deficient but blamed the now-abandoned Ten Year Rule.

In March 1936 Baldwin overruled MH on appointing Thomas Inskip as a defence deputy — Minister for Coordination of Defence. [This was famously but anonymously described as ‘the most cynical appointment since Caligula made his horse a consul’.] Inskip was not a military expert. MH simultaneously appointed a senior officer as deputy secretary in the CID — ‘Pug’ Ismay who would serve Churchill with distinction. He also strengthened the Joint Planning Committee. MH’s role was unchanged by the changes. Inskip sided with those who argued that the RAF must build up its fighter force.

MH also believed that Britain should prioritise her own defence over plans to help allies on the Continent. He did not believe in the League’s ability to enforce sanctions against Italy and he thought trying to use it against Italy would fail. Generally the military leadership agreed with MH that ‘collective security’ was a dangerous delusion. The COS thought that our position could not withstand an attack from Germany in Europe and Japan in Asia at the same time and pleaded for time.

Naylor argues that the push by MH and others in the system against the League and ‘collective security’ plus the outcome of the Hoare-Laval Pact contributed to Hitler’s belief that Britain wouldn’t fight for anything but herself. [In Britain the Pact led to an outcry in media and Parliament. The Chief Whip told the PM the MPs wouldn’t stand for it.]

In April 1936 MH summed up his attitude to rearmament:

The essential thing is to stave off war and to strengthen ourselves in case one day we can stave it off no longer — which God forbid.

There was a fundamental difference between the MH view — appease to buy time for rearmament — and the Chamberlain view — appease because it can bring a permanent peace. Naylor says MH ‘does not stand in the ranks of the appeasers’ because the policy was for him ‘a diplomatic mask designed to disguise the inability of British arms yet to sustain any other foreign policy’; he conciliated from a ‘conviction of weakness rather than a sense of mission’.

There was criticism of proliferating committees under the CID system. In early 1937 a standing ministerial committee was created, designated a CID Sub-Committee on Defence Plans (Policy). There were three committees — the CID, the DRC and the DP(P) — all of which overlapped with no clear demarcation.

MH worked closely with Chamberlain who took over as PM in 1937 but he also agreed with Balfour that Chamberlain’s judgement could not be trusted. As Chancellor, Chamberlain had not been interested in MH views on rearmament and defence priorities.

MH, the military and Vansittart agreed on the need to try to avoid fighting with Italy while they rearmed. Eden disagreed thinking appeasement would encourage more claims. [Of course, both perspectives could be right: we should avoid fighting Italy now AND/BUT this makes a fight more likely later. Of course, a core issue was that Chamberlain hoped that appeasement might lead to a permanent settlement.] In a footnote Naylor says that notes from a MH trip to Italy in 1937 lend themselves to the interpretation that MH had emotional sympathies with Italian fascism but he does not explore this further.

In the last major review of defence preparations before the war in winter 1937-8, HMT/Fisher again asserted caution because ‘we are rapidly drifting into financial chaos and are in danger of undermining ourselves before the Boche feels it desirable to move’. In December 1937 MH asserted that the RAF should pursue a larger proportion of light and medium bombers viz heavy bombers but Inskip opposed him and ignored it. The claim by some that MH pushed for fighters over bombers is not supported by the COS and HMT records (p249).

In 1937 the Economic Affairs Council was overhauled — officials recommended a return to the structure of the Committee of Civil Research of 1925! The Council had fallen out of use, was generally ineffective and this was a fitting end.

MH was pleased Eden resigned in Feb 1938.

I woke this morning with a strange feeling of relief. I am sorry to say that generally I wake on how we are to provide for some horror in the next war. Today I felt there was just a possibility of peace. I only hope I am right.

In August 1938 MH resigned as Cabinet Secretary and was replaced by Edward Bridges. He’d scouted out some directorships such as on the Suez Canal Company board (under the PM’s control). He recommended to the PM that his successor continue to combine the Secretariat and CID. He said it brought great advantages including of speed for the jobs to be unified in one office. But it was tricky to find someone with the skills needed for both. Fisher pushed for Edward Bridges. MH pushed Ismay saying a service officer would be better but both candidates would be great. Chamberlain decided to break up MH’s job: Bridges for the Secretariat and Ismay for CID.

Naylor says it’s hard to assess MH’s contributions on rearmament because his voice ‘blends into a chorus’.

MH left with bad blood with Churchill over the latter’s access to and use of secret information especially re the RAF to criticise policy. MH ended up writing to WSC complaining that his cultivation of information from officers was ‘subversive to discipline’ and that he strongly opposed such officers sharing such information. Relations never recovered. Naylor says the letter was unusual, perhaps unique, was a blunder, and is probably explained by the cumulative pressure MH felt over many years. (An interesting aside is Ismay’s comment that he found the three years 1936-9 ‘more difficult and anxious’ for him than the constant disasters of 1939-42.)

Coinciding with his leaving was the CO leaving Whitehall Gardens which was destroyed in August to make way for the Main Building of the MoD. [It’s interesting MH did not try to use his power to scupper this destruction!] MH handed over the Bridges the War Book he had laboured over.

[I have not gone into the huge questions on appeasement and defence planning but see blog on RV Jones for insight into the air aspect.]

---

CH7: The custody of Cabinet secrecy

Bridges made no big changes between August 1938 and the start of the war. According to Naylor, Churchill wanted to make some big changes when he took over but Ismay persuaded him not to.

The CO moved first to private houses in Richmond Terrace then after bombing to government buildings in Great George Street.

CID was closed at the outset of war and essentially merged into the new War Cabinet system with its Secretariat merging into the Military Section of the Cabinet Secretariat. Bridges remarked that the abolition of some two dozen subcommittees of CID was the biggest blow the Germans had endured. The Secretariat was split between the civil and military.

Minutes became shorter: the pros and cons of a problem were stated plus the conclusion or decision. Individual views were largely ignored.

The Secretariat grew from ~200 1939 to ~600 1945 (681 in 1976 but half were doing stats). Bridges expanded on MH’s idea of seconding departmental officials for 6-12 months to the CO so departments trusted the work.

MH accepted a War Cabinet post under Chamberlain and continued under WSC until removed in 1942. He then worked on his own memoir of WWI. By September 1943 he’d written to Bridges to arrange a vetting of his book. Bridges was concerned about the effect on the confidential relationship between ministers and officials. MH made changes and reminded Bridges that WWI had been treated as sui generis from the start. Bridges was unmoved. He objected to the depiction of the influence of officials on ministers’ decisions and to MH’s use of his diary to recover ‘the inner histories’ of government: ministers would become ‘much more chary of speaking freely’ to officials. For Bridges it was not another memoir about WWI but the first memoir about the influence of officials! MH appealed to the PM who sided with Bridges.

In 1957 MH tried again with Macmillan but was again told he could not publish. After correspondence Macmillan replied enclosing a copy of OSA and wrote that MH would not need a lawyer to realise that OSA ‘makes it an offence for any person who has held an official position to publish any information which he has acquired by virtue of holding that position’. So MH had pushed a broad interpretation of OSA to curtail ministerial behaviour he disliked and now found OSA used against him. One can feel some sympathy that MH watched memoirs of the recent war be published while he was still prevented from publishing his memoir of the previous war. Macmillan softened his stance when MH modified the book and a modified version was allowed in 1961. This itself was an exception to the post-1945 insistence that officials not publish memoirs.

Cadogan had also started a diary, known to Halifax, as had other officials. But Bridges did not yet know about it in dealing with MH. MH claimed his diary had been ‘absolutely essential’ to his work in WWI given the lack of official records. MH was then told that he’d have to remove all quotations from diaries.

Bridges was succeeded by his deputy, Norman Brook, in 1946. Brook served as Joint Permanent Secretary at the Treasury and as Head of the Civil Service. After that the jobs were split up again.

The OSA was used in 1938 against a journalist warning a criminal of arrest but this led to worries about its application to humdrum criminal matters. It was then used against WSC’s son-in-law Duncan Sandys who had implicitly threatened to reveal the poor state of London’s air defences. It was suggested that he could be imprisoned for two years if he did not reveal his official sources. But Sandys invoked his rights as an MP and the Speaker and Clerk of the Commons said that a breach of the privilege of rights of a MP had occurred. The AG backed off and said there was no question of trying to exercise powers of interrogation under OSA.

WSC articulated the core questions:

1. Does OSA enable HMG to bring criminal proceedings against MPs and will HMG do this?

2. Should HMG use OSA to prevent ‘exposure of Ministers who have neglected their duty’?

3. Is Parliamentary privilege in the larger sense involved in attempts by HMG to intimidate MPs?

In 1939 the government pursued an amendment to the Act while the Commons pursued the breach of privilege. The PM blamed an ‘extraordinary catalogue of accidents’ for mistakes over the Sandys case! Parliament made clear that it protected its proceedings from the 1911 and 1920 legislation and the government was forced to limit its use of OSA to espionage. Section 6 was modified, section 2 was intact.

[NB. This has been an issue recently. Some Tories claimed they could not have spoken in the Commons about the secret Afghan flights and super-injunction (revealed July 2025) ‘because of OSA’ but this is clearly false. OSA is relevant to what they say outside the Commons but does not limit what they say in the House.]

The restrictions on ministerial memoirs established in 1934 remained. Officials were not allowed to use official documents to settle public disputes. Vansittart complained when he was attacked post-war but the system did not budge.

Wartime continued to be treated as sui generis. E.g Alanbrooke’s diaries were allowed — though the original publication of them involved considerable modification of passages criticising Churchill which only were published long after he was dead.

In 1957 Bridges issued new regulations stating that Crown Servants were liable to prosecution under OSA for publication of ‘any official documentation which has not already been made public’ — i.e all official information, not just secret information.

The 1958 Public Records Act changed the bias from A) no publication by officials to B) an acceptance of publication after some decades.

At the end of the Victorian period the Home Office enforced a prohibition on access to records post-1800. Even post-1945 the HO retained its records post-1800 and transferred few of those to the Public Record Office. The FO decided in 1924 to open records through to the end of 1878. After 1954 some system was brought to the process.

Even aspects of intelligence work have been allowed, e.g Masterman’s The Double Cross System. Some documents were sealed for a century, e.g documents relating to the Abdication crisis of 1936.

Eden’s memoirs used Cabinet documents. Macmillan kept a diary he drew on for his own memoirs.

The Franks Committee 1971 described Section 2 as ‘an ancient blunderbuss’ in need of replacement.

The Wilson government failed to sustain a legal objection to the 1975 publication of Crossman’s Diaries — extracts in The Times 26/1/75. Crossman’s executors and the Times were determined on publication without following the Cabinet Secretary’s diktats. The government did not use OSA in their court case, they made a common law argument, which was the ‘coup de grace’ (Naylor) against Section 2. The Lord Chief Justice ruled the public interest would not be damaged by publication of a record of Cabinet at a decade’s remove. The judgment was not carte blanche for publication but the judge articulated the concept of balancing restrictions ‘in the public interest’ against ‘another public interest, such as freedom of speech’.

Crossman’s diaries are the first sign of ministers complaining about a ‘rival administration’ of officials, suspicions that officials were meeting without ministers to shape critical decisions, a growing feeling there is a wiring of real power hidden from the Cabinet, suspicion of the real power of the Cabinet Secretary exceeding that of ministers, a feeling that the drafting of Minutes suddenly revealed that a minister had lost a battle which was not evident in the meeting (!). [Such thoughts were novel and it would have shocked Palmerston’s generation to think ministers might feel like this. Such feelings now are widespread.]

In parallel a committee of Privy Counsellors (Radcliffe committee) inquired into the subject of ministerial memoirs and advised the proscription for 15 years, half that of the Lord Chief Justice in the Crossman case. It stressed the need to protect officials — there must be a ‘scrupulous reticence with regard to the attitudes and personalities’ of officials when writing memoirs. The Cabinet Secretary’s role in vetting memoirs was strengthened. And it was stated that ministers ought not to discuss officials still in service.

There’s a lot of detailed history on thinking about memoirs in the 1960s and 70s I’ve left out here.

Naylor says that there are documents relating to MH’s ministerial work in intelligence and bacteriological warfare that may never be released. [Is this true? Have they been published, destroyed, kept secret? Please leave links if you know.]

NB. the old practice was to keep the existence and membership of Cabinet Committees secret. Now there is a list published.

---

Some thoughts

NB1. As always when I describe Whitehall, remember that I also know and am friends with many excellent officials. The media stories of 2019-20 about widespread hatred and fighting were false — my biggest supporters were in the deep state, not among political people. My comments are not universal, when I say ‘officials’ I don’t mean ‘all officials’, I mean ‘the dominant subset which pushes the system’. They are directed at the dominant trends/characters which the pathological system now promotes at the expense of good public servants who are increasingly driven out.

NB2. These thoughts are intended to help develop a network thinking about how to do regime change properly. This needs many things discussed openly over the next 1-3 years to develop as deep a consensus as possible about critical issues (given constraints) but also, obviously, not everything should be discussed openly! Some ideas not discussed here will occur to readers. Feel free to suggest anything but my responses may not always be entirely open. An obvious example of something which should not be discussed openly before it happens is the precise list of people who should be removed and ideas for who replaces certain roles.

1/ The post-1850s civil service over time has destroyed the old responsibility to Parliament. The Northcote-Trevelyan report came partly from the model of the East India Company which had applied ideas from the Chinese civil service about competitive exams, ‘meritocracy’ etc. Gladstone wanted to replace the old system. The Northcote-Trevelyan report of 1854 was not implemented but came as the shambles of the Crimea War showed change was needed. Change and reform was in the air so it was implemented.

As Northcote-Trevelyan worked its way through after the 1850s, permanent officials acquired more power, ministers lost power. Individual responsibility for ministers and officials gradually declined. For Pitt and Palmerston, ministerial responsibility to Parliament was real: they were really in charge and really responsible to Parliament. After Northcote-Trevelyan, this constitutional-theoretical responsibility to Parliament gradually became more and more fake. Ministers now pretend to ‘run my department’ and ‘take responsibility’ but they know little of what goes on in the departments, they can hire and fire nobody, and they are rarely responsible in any real way. When they say in Parliament ‘I’m responsible’, everyone knows it’s something of a charade and the minister saying it is not responsible in any normal sense, and may not even have been aware of the thing they’re taking responsibility for.

As I’ve pointed out many times, the brilliant book Now It Can Be Told by General Groves illustrates timeless core ‘unrecognised simplicities’ of high performance. One of his principles for running the Manhattan Project was:

Authority was invariably delegated with responsibility, and this delegation was absolute and without reservation.

Whitehall has rendered this crucial simple principle effectively unlawful/impossible. If you try to enact its reinstatement and say ‘I want named directly responsible individuals in charge of XYZ so we all know the score’, as I did in 2020, it’s seen as a fundamentally hostile act to modern Whitehall. As long as this continues, it’s pointless hoping for serious change in performance.

2/ The CO’s creation was in response to the failure of the newly ‘professionalised’ system. Whitehall including the Treasury and FO failed to think about deterrence and war properly pre-1914 — see Hankey’s and others’ comments. And Whitehall proved worse at adapting in a crisis than the pre-Northcote Trevelyan system had been. The Cabinet Secretariat/Office were created partly because the newly ‘professionalised’ system could not solve its problems and in panic they essentially threw power at Lloyd George to centralise and decide much himself. After the 1916-18 crisis passed, they were left with the machinery created in this panic, in a similar way to how much of Washington today is what FDR and Hopkins created 1933-45 in the panic over the Great Depression and threat of Communism.

It’s fascinating how both Asquith and Lloyd George lacked confidence in the professional military leadership’s strategy and execution (as Pitt sometimes did) yet felt politically unable to fire generals (unlike Pitt who moved people and promoted people). Hence the carnage continued with offensive after offensive despite many ministers having grave doubts about the strategy. The rapid promotion of talent and demotion of duffers that characterised the Pitt story has gone. The supposed ‘meritocracy’ and ‘professionalisation’ of Whitehall with Northcote-Trevelyan yielded Haig and others at the pinnacle of the bureaucracies, strong bureaucratic resistance to innovation, and the Somme.

The story of Pitt, Wellington and Nelson is messier, ‘noisier’, with much more dramatic public controversy and Parliamentary debate — and amid many failures, phenomenal successes including two rare geniuses (Wellington and Nelson) discovered and empowered who built great institutions to defeat one of the great military geniuses in history.

The story of WWI is politicians sitting on doubts, failing to act, more publicly ‘unified’, less ‘noisy’, less Parliamentary controversy — and much less success. It was a catastrophe from which we’ve never recovered, a catastrophe which Alanbrooke felt keenly in the darkest moments of 1942 when he wrote in his Diaries that the real reason for the WW2 disasters was the the best of the younger generation had been slaughtered in WWI’s trenches.

Just as the system failed to generate realistic thinking about Germany and preparations pre-1914, so it also mostly failed in similar ways with Hitler. But this failure was a mix of institutional failure to focus thinking (like pre-1914), the personal characteristics of senior ministers (e.g Chamberlain’s fundamental misreading of Hitler), and (real/imagined) political constraints (e.g ministers had to care far more about voters’ desires for spending on non-defence than Pitt did).

Many stories of bureaucratic inertia in WW1 and WW2 feel very familiar to me having watched covid in ways that the stories of 1790s-1800s are less familiar to my time in SW1 (and more similar to other times and places). Modern bureaucratic madness has its own particular kind of stories and characters. For sure, every war involves disasters and stupidities. But there was a dynamism you see in our response to the Napoleonic Wars which was already diminished by 1914 and had diminished further in Alanbrooke’s tales of Whitehall in his Diaries, in RV Jones’ memoir (links above), in the battles to develop Special Forces against extreme opposition and in other tales.

We have continued in the same direction for decades since. This has been generally pathological and this can be seen int he way the successful things were attacked by ‘the system’. E.g The successful Alanbrooke/CoS Committee model: ended. The successful RV Jones model for scientific intelligence: ended immediately after 1945. The SAS: closed immediately after 1945. It’s to be expected that our response to Ukraine was extremely delusional, extremely incompetent, and generated more lies and delusions rather than improvements.

3/ Secrecy is used to hide failure. When you create a permanent class that promotes almost entirely from within, they inevitably prioritise protecting the permanent system. Secrecy and ‘national security’ is therefore used to stop not just voters but also the politicians nominally in charge understanding the system’s failures and insisting on individual responsibility.

I saw it in covid when the CO tried to stop us publishing SAGE minutes, claiming it was technically part of the Cabinet Committee system therefore must not publish minutes (we won on this). I saw it in how the CO classified disasters in the nuclear enterprise and extraordinary failures viz PRC penetration of Whitehall — secrecy aimed mainly at keeping secret almost unbelievable failure (much greater than is realised by/disclosed to 99% of MPs) and limiting criticism of officials. Similarly child abuse and the grooming gangs were covered up by officials protecting the system which they tried to justify inside the system as ‘we’re stopping racists exploiting news’.

Official secrecy generally — the Official Secrets Act, FOI regime, vetting systems, RIPA, Investigatory Powers Act, the leak/briefing problem of ministers/spads/officials, powers to investigate and prosecute etc — is another complex dysfunctional mess.

A very few things need much more serious secrecy but most things should be more transparent.

4/ Institutional mechanisms cannot be relied on to solve fundamental political and strategic issues. The lack of clarity about decisions in the pre-CO Cabinet was bad and having proper records — or clarity there was no decision — was good. But the most important disagreements and questions are almost inevitably so deep and uncertain that better bureaucratic process cannot provide an obvious answer, it can only help clarify the true problems.

This does not imply fatalism about improving bureaucracies. But it’s important to be realistic that lifer-bureaucrats tend to reduce all problems to ‘needs better process’ and it isn’t true. The failure of the post-CO Cabinets to solve critical problems with Germany in 1917-18 and in the 1930s show this very clearly.

Getting good answers to the deepest problems of war and peace depends in the end on the talents, disposition, sensibility etc of key individuals as well as the quality of ‘process’.

5/ The CCR. It’s very interesting that they created the Committee of Civil Research (CCR) in 1925 to provide No10 with serious analytical skills, a sort of partial forerunner to the No10 data science team we created 2020. Its function was to give ‘connected forethought from a central standpoint to the development of economic, scientific and statistical research in relation to civil policy and administration and it will define new areas in which enquiry will be valuable’.

It’s also interesting that it did not survive and HMT helped kibosh it. Similarly the CO has repeatedly tried to kill ‘10ds’ (saved in 2024 by an alliance of some officials and McSweeney).

I disagree with Heseltine on most important issues, and I think he’s one of many who thinks I should be in jail for Brexit, but it’s interesting that he is unique (in my knowledge of ministers of the last 50 years) in taking information systems very seriously. He created MINIS, a system to track responsibility and cost. He himself has explained that it was a feature of having been a successful businessman and I’m sure this is true. And he also explained officials always shut it down and his MP colleagues let them.

Every time I left a department the officials closed the system down and I have to say, with respect to my much admired colleagues, so did they, because it was an extremely boring thing. It was tedious to the degree to go through this infinite detail of money being spent and making decisions about it.

6/ CID/national security/defence capabilities.

The CID was set up to improve coordination across Whitehall on defence.

Note the interesting ideas from Esher after the Boer War about the PM’s office taking charge of the CID and how the CID — a pre-CO entity —was an entity responsible to the PM, not to Cabinet or the CO.

CID lacked power because Balfour did not insist that CID conclusions should be the basis on which the War Office and Admiralty shape their roles and plans. CID played no part in shaping discussions with France over the Belgium guarantee. There was only one meeting between ministers and officials to discuss this (in 1911). CID therefore did not solve the fundamental issues of defence planning — the tension between policy and capabilities.

In WW2 this was handled much better via the CoS Committee which could think, decide and enforce decisions. Many things which worked better in the war, including CoS and JIC and scientific intelligence, were changed for the worse after 1945.

Today the NSC is supposed to be the place for coordination of defence and security across Whitehall. It does not work well, partly because of how it works and partly because nothing can work well when it’s trying to coordinate multiple entities which are themselves dysfunctional, including the MoD and CO. The NSC is run by the CO and the CO is one of the entities theoretically supposed to be coordinated by the NSC — a relationship which tends to the dysfunctional, as the NSC can hardly be expected to be tough on the failures of the CO when it is run in the operational sense by the CO.

E.g NSC is carefully steered away from scrutiny of the MoD. MoD, HMT and CO prefer to keep these issues to their own pathological processes away from prying political eyes. NSC under the Tories was often Potemkin made worse by the fact that the leaking culture got so bad that the SoS for Defence, Wallace, and other ministers including Truss were repeatedly investigated by the CO for leaks from it and these leaks inevitably meant senior people such as ‘C’ did not share some sensitive things.

The issues of defence planning which led to the creation of CID are even more pressing today. The role of the CO confuses the shambles further because a) CO officials now have institutional incentives/motives/ideas and b) the CO makes the processes more opaque to the PM and PM staff, often using secrecy and ‘spads aren’t cleared to see X’ as an excuse. Everything is clouded by fog and the MoD then often rightly blames the CO for problems. Another example: J Powell is the PM’s National Security Adviser, but because he is ‘political’ officials have insisted he cannot be actually in charge of budgets, which have been put in the hands of an official who is lower in the pecking order and therefore given the brush off by the MoD and agencies.

7/ HMT has too much power while also generating very bad budget and SR processes and totally failing on cost control (which it isn’t interested in and can’t do).

There’s a cyclical discussion about HMT and the need for coordination from No10 and outside expertise. We saw it in 1929-30 with the Economic Advisory Council. Similar discussions played out after 2008 and in 2020.

• Ministers and advisers sense HMT is not doing what’s needed, because of some mix of a lack of skills and resistance to outside ideas (e.g I suggested 50 or 100 year bonds in 2020 (suggested to me by a very smart hedge funder, not my idea), which would have been a dramatic triumph — HMT officials very visibly resisted even discussing such ‘outlandish’ ideas).

• HMT won’t let anybody else develop anything powerful.

• The PM also gets the feeling that HMT isn’t doing what’s needed but won’t insist on big changes.

• So ‘advisory’ bodies of ‘outside experts’ are created which HMT ensures are talking shops with little/no influence. E.g Any reviews have written into their terms of reference that HMT processes are not part of the review etc. (E.g see how HMT neuters every science review from considering HMT even though HMT vandalism is one of the main problems. In 2020, HMT tried to argue that it was not subject to decisions of the XO Cabinet Committee.)

• After a while the new outside experts thingy proves at best of minimal value so is wound up: ‘didn’t work out old boy, duplicates other things too, too many meetings, time to tidy it up’.

• The cycle repeats with each crisis.

HMT can be fixed without legislation if a PM wants to.

See below.

8/ A core task of the centre — coordinating between permanent large bureaucratic institutions — is inherently tricky and can never be ‘solved’.

Departments are locked into a somewhat zero-sum game for money and power with other departments and resist providing information. And they resist coordination by any central entity which might point out failures, savings, alternative ideas etc. The services resist coordination and all processes such as Cabinet Committees and CO structures will have similar problems. The core focus and the core people will determine the extent to which the structures work. E.g Alanbrooke made the chiefs of staff committee work. There is no substitute for leadership.

But improving the way budgets are made and priorities conceived and executed — i.e the dysfunction of No10-HMT-CO — would radically diminish inevitable tensions. Cf. below on Jobs and Apple’s highly unusual organisation.

9/ ‘Consent, delay, evade’ is a constant problem. It can only be defeated by priorities, relentless focus, determination to impose your will over months and years, and politicians being prepared to remove officials who subvert via consent, delay, evade. If you’re clearly prepared to remove people, you don’t have to remove many people. If you’re not, the entire system operates on consent, delay, evade.

Newspapers now routinely carry comments to the effect that Starmer is ‘furious’ that things don’t happen faster. But officials know he won’t fire people so they can ‘consent, delay, evade’ with impunity. Briefing, complaining and shuffling structures cannot affect anything important. Officials have seen it all before many times. When Starmer’s next reorganisation is announced, it will immediately be crystal clear to officials how they can continue consent, delay, evade.

10/ It’s crucial to consider the contrast between a) the clear agreement on ‘what the CO must not do’ when it was set up and b) how it now does all of these things.

The Cabinet Secretariat was created in the 1916 crisis to:

• write an agenda for Cabinet,

• record decisions and communicate them to relevant people,

• help ensure sound information flow on cross-departmental matters for the war effort.

It was explicitly not for providing policy advice.

It was explicitly not for controlling personnel decisions.

It was explicitly stated that it must not interfere with the responsibility of departments to the PM/Cabinet.

It was explicitly stated that it must not interfere with the responsibility of ministers to Parliament.

It was explicitly stated that it must not get big, it must remain very small and not sprawl.

Hankey:

It is of the upmost importance that the responsibility of the Departments should be in no way weakened or overridden by the development of the Secretariat…

The function of the Cabinet Office is essentially one of machinery of Government and not of policy. It cannot be too strongly emphasised that the Cabinet Secretary has no duty of offering advice in any matter of policy or of interfering in any way with the functions of responsible Departments.

Lloyd George: it’s ‘purely a recording machine’ and has ‘nothing whatsoever to do with any question of policy’.

It was not even stated that it would not interfere in ministerial appointments. The idea would have seemed ludicrous. The idea of the Cabinet Office maintaining permanent records on Ministers’ personal lives, leaked when deemed appropriate to dish a career, would have seemed inconceivable. But of course if a PM said today they are abolishing PET (the CO’s HR STASI, holder of the personnel files) it would be denounced by most Insiders as ‘dangerous Orbanism’, ‘a power grab’, ‘undermining our institutions’ etc.

Now it does exactly those things which its creators said it must not do. And further, over time, and particularly post-2010, the CO loosened the PM’s scrutiny and control so it became an empire with close to zero effective visibility, never mind responsibility, to any minister including the PM.

The Cabinet Office has become a vast, powerful, independent agent with its own policy machine and communications team, its own bureaucratic interests (not the same as the PM’s office), its own HR system which controls careers for the rest of the civil service, its own security and intelligence assets and networks. Close to 100% of its meetings are ‘officials only’.

And the Cabinet Secretary has become much more powerful than any minister except the PM.

• When a bomb goes off or a minister is suspected of having unsuitable girlfriends (happened a bit in the last few years), it is to the Cabinet Secretary that very sensitive reports come first — and he decides what ministers are told and when, if anything.

• The Cabinet Secretary can call the police and start or stop criminal investigations.

• The Cabinet Secretary job has evolved from ‘must not interject his own policy advice’ to ‘the PM’s main policy adviser’.

• The Cabinet Secretary job has evolved from ‘must not interfere in appointments’ to ‘in charge of all appointments’.

• The Cabinet Secretary job has evolved from ‘must not interfere with responsibility of ministers’ to ‘the Cabinet Secretary chairs meetings holding ministers to account and advises the PM on removing ministers’.

• The Cabinet Secretary job has morphed from the Hankey role to combining both a much bigger role as head of the Cabinet Office AND head of the civil service to make the job ‘the spider at the centre of the web’ (Lord Armstrong), including power over all appointments.

• The Cabinet Office and morphed role of the Cabinet Secretary has contributed to the collapse of ministerial power and responsibility. As failure has intensified and individual responsibility has been designed out of the system, power has become more intensely centralised with officials in the CO.

The CO today tries to present itself to the PM’s staff as effectively working for the PM and the PM’s priorities: you don’t need to hire any more people PM, we all work for you! But it is not its true constitutional position, nor how it sees itself, nor how it works in practice.

Per Robin Butler, it sees its official role as ‘serving the Cabinet’ not the PM.

I think none of us [Cabinet Secretaries] saw the Cabinet Office as the instrument to deliver what Tony [Blair] wants. We did not see it [the Cabinet Office] as an executive body that was to deliver the will of the Prime Minister… The Cabinet Office supports the ministers collectively and acts as a broker between them in cases of disagreement.

And of course it’s the Cabinet Secretary who writes the minutes defining what the ‘ministers collectively’ decide.

And O’Donnell made clear that the CO now sees the PM’s office as a subset of the CO and run by the CO:

Let me be absolutely clear… There is one Cabinet Office and Number 10 is a subset of the Cabinet Office… Number 10 is part of the Cabinet Office… There are not two departments. I stress there is one department.

And many of its officials have increasingly come to see themselves as serving the interests of the CO itself — a sensibility they justify often by saying to themselves ‘look at the rubbish MPs, it’s our duty to keep the show on the road’ etc.

Again, Butler said out loud in 2010 to the Lords Committee that he wants the Cabinet Secretary and head of the civil service jobs to stay merged because it means the former then controls appointments, the real power:

The second [reason], I am afraid, is a more Machiavellian one, which is that to have some influence and control of the senior appointments gives the Cabinet Secretary … leverage over government departments.

You bet, Robin, thanks for saying it out loud, though it’s a shame the MPs never noticed! Control appointments and you have real power, while the Potemkin ministers are patted on the head and sent off to ‘take full responsibility’.

The power in the Cabinet Office has grown in proportion to the growing frustration, sometimes rage, of successive PMs about how ‘the centre’ works. All ‘reforms’ (except what we started in 2020 which were abolished/reversed) have left it stronger. Starmer, like Sunak, tried some tweaks on arriving. Now he whines that he says things but nothing much seems to happen — just like every recent PM. But they never use their power as PM to change it.

Potemkin Cabinet suits officials. Officials today do not want voters to understand how much power they now have and how little power the elected politicians now exercise. Officials rightly think that were voters to understand there would be strong demands for the politicians to take back control. They therefore encourage the continuation of the Potemkin ‘Cabinet Government’ charade each week I’ve written about many times, a charade in which Cabinet is theatre — theatre for the Ministers and MPs, helping them pretend to themselves they are ‘running the country’, and theatre for the media, so they keep reporting the Official Story. Meanwhile more things are run from, and most important meetings including pre-meetings occur in, the Cabinet Office than the PM’s office.

CO officials like to describe themselves to No10 spads as ‘the gearbox’ for the PM’s Office but they are not, they are the sand in the gearbox.

It’s crucial to grasp that as the Cabinet Office/Secretary have hoovered up power, they have made greater and greater efforts at narrative control, to manage appearances, to make it look as if the PM and ministers are in charge. Look at this quote from the Official History of the Cabinet Office:

[T]he central contention of this history is that the first duty of the Cabinet Secretary is to make ministers look in control of events, farsighted and wise, governing in the interests of the nation as a whole… [T]he processes of government must appear rational and smooth, no matter what is going on beneath the surface. The more the processes can be kept secret, the better the chance of success.

11/ No10: much more is now demanded of it while the CO stops it acting.

Ironically, as Whitehall has become more pathological and made ministerial responsibility largely fake and direct responsibility practically impossible, this has thrown more weight on the character and sensibility of the PM, being the only one with the constitutional power to deal with many pathologies.

We now have a ‘centre of power’ which is a) swamped by demands from other broken institutions to solve problems nobody else has the authority/skills to solve and b) habituated to the role of a sort of Media Entertainment Service in which it perpetuates fake stories about how government works while being unable to execute competently itself. It’s both 100X more expected to ‘grip’ problems than 50 years ago and, paralysed by the CO, unable to do so other than in very rare and usually very limited/focused cases.

The need for some institutional mechanism to perform the function of focused execution of complex coordination, given the growing demands on the centre and the diminishing personal skills of our politicians on average, is a big part of the history of the CO since 1916. To a large extent, CO mechanisms today are supposed to generate for the government crucial skills which our MPs and spads largely lack, and processes they cannot create for themselves, while maintaining various constitutional fictions about ‘ministers deciding’ and ‘Cabinet government’.

---

Some Questions for a new PM wanting to be ‘a government that controls the government’

Questions such as the role of Cabinet, how to restore real responsibility, how to recruit and incentivise elite talent, how to structure No10 and change powers of HMT and Cabinet Office are all entangled. To stress again: fixing government is a systems problem.

The below does not touch on the core political (and, deeper, the spiritual) issues a serious new PM must consider, it focuses on some institutional questions raised by the above. Obviously I have my own detailed ideas on these questions and a preferred model for a new PM but won’t go into much of this here.

Many of these questions are interdependent. What you decide about closing the CO obviously changes what you do with other aspects of No10. What you decide about Cabinet changes other things. What you decide on how to restore direct personal responsibility changes your plans on elite talent, given you can only recruit elite talent if you also create jobs with authority and responsibility together, per Groves…

How to recruit elite talent? Making jobs non-fake will be very controversial as it hasn’t happened, or even been conceivable, in living memory in Whitehall with very few exceptions. (Even in May 2020, with tens of thousands of recent deaths, creating non-fake jobs leading projects like vaccine research or testing was bitterly resisted by much of Whitehall.) They won’t stay without leadership. 99% of the great people we brought in 2020, such as technology specialists who could have made millions in Silicon Valley instead (and have after leaving!), left in the months after I left No10 (and all of them I think by 2022) because of the combination of a) clearly the PM had abandoned the plan and b) the dominant CO officials made clear they would be thwarted, not allowed in meetings, if necessary persecuted etc.

Parties and the Commons gradually shifted over a couple of centuries from some elite talent to almost zero elite talent and active hostility to it. We went from Pitt (Napoleonic Wars) to Asquith (World War I) to Truss/Starmer. This is arguably impossible to fix in a representative democracy and libraries are full of comments to this effect. If so, then crucial jobs must be given to non-MPs with responsibility operating in un-traditional ways. But even if it can be partly reversed, it will take many years and is clearly not something that can or will be fixed in a couple of years.

I think a serious Leader should make recruiting elite talent a core priority over 2025-9 with a clear story about why being an MP is not doomed to being a largely fake job and how they will restore responsibility/authority to non-elected executive roles in Whitehall (and elsewhere). This hasn’t happened in a serious way in living memory but if done seriously, month after month with good communication, it would be a very compelling story. It is only credible if the Leader is credible.

What is Cabinet for?

A group of 25-30 is unarguably incapable of acting as a serious decision making group or even a coherent discussion forum. Cabinet’s size means that even before 1914, it had an inherent tendency towards being ‘22 gabblers round a table with an old procrastinator in the chair’ (Carson). It has grown since then to over 25 and there’s a been a tendency to ‘keep ministers quiet’ by adding them to a ‘attend Cabinet’ list. And at least half a dozen officials and spads sit against the wall, mostly pointlessly other than as a sign of their supposed prestige in the system. There was incredible resistance in the CO to all my attempts to stop this practice (similar to my requests for an end to the absurd email chains with 100 people by which the centre tries to organise itself). A small (handful) ‘War Cabinet’ worked better than normal Cabinet in WW1 and WW2 for obvious reasons.*

• Option A: Potemkin. Leave it as a Potemkin forum while real discussions are elsewhere. Pros: keeps lots of colleagues happy because the Downing Street camera parade continues. Cons: a lot of time is wasted on managing the Potemkin theatre and continuing a fake system is inherently bad.

• Option B: much smaller. Make it much smaller (<10), one non-minister taking notes. Its role is hashing out critical questions. Pros: Cabinet government becomes real instead of fake. Cons: only works with serious ministers. Howls of anguish as the TV parade cast list is cut, the MPs go crazy (and you’re ramping up the ‘elite overproduction’ competition).

• Option C: an inner and ‘normal’ Cabinet. Pros: gains of a smaller group and avoid some howling. Cons: needs complex handling and it admits the normal Cabinet is fake.

• Option D. I won’t put my preferred option here for now but there are interesting alternatives to the obvious connected to how you use Cabinet Committees. Whatever the formal story there is always an inner group exercising power and influence. One can formalise this or leave it informal.

• Should there be ministers without portfolio solely to help the PM in ‘Thinking and Deciding’ (above) without having either to spend their own time on departmental affairs or having inevitable personal incentives generated by them having a departmental job?

* Cabinet confidentiality. Cabinet confidentiality collapsed from 2017 with May’s political authority and it accelerated under Boris. Ministers got in the habit of simply calling hacks immediately and giving full readouts. I suggested to the PM after our 2019 victory that it was the moment to return to confidentiality and enforce it by firing someone for leaking. He declined. The problem worsened and therefore also made it more Potemkin, as we couldn’t discuss sensitive issues in a place which leaked in minutes. A group of ~30 will tend to leak even if the culture is better than it is.

Will you restore ministerial responsibility and give ministers the power to execute priorities?

The evolved system whereby Permanent Secretaries work for the Cabinet Secretary while the PM and PM Office have nothing to do with setting priorities, monitoring, intervening, and nothing to do with the talent pipeline is connected to many of the problems including a) consent, delay, evade, and b) fake responsibility.

One way or another, the PM must form priorities and make individuals responsible for executing them.

You can only do this if you can recruit elite talent and fundamentally change powers over the civil service including ending the current system in which only the PM can insist on a civil servant being replaced.

Option A: leave Ministers as fake and appoint others with proper power and responsibility.

Option B: make (at least some) Ministers truly responsible and therefore able to replace officials (with some sort of light involvement from No10).

I favour (B) partly because I also favour trying to restore the Commons as a serious place and a necessary (not sufficient) condition is restoring real ministerial responsibility to the Commons.

How to change, formally/explicitly and informally/implicitly, the balance of power over money and budgets between a) No10 and the PM’s Office, b) No11 and HMT?

There are many problems with how the No10/HMT relationship has evolved, inter alia:

• The way HMT works makes it impossible for a government to think through then execute genuine priorities.

• Long-term budgets now are almost sure to be a disaster, with funny money and dodgy accounting par for the course — ‘savings’ which are obviously ‘costs’ in every sense except HMT definitions etc. HMT ‘baselines’ are used to create funny money ‘savings’. HMT often insists on a department writing to everybody notifying that ‘there is no funding guaranteed after date X’, so then people get laid off etc, and ‘savings’ are registered, then after date Y the program is restarted, people are rehired etc, so millions are lost in reality but HMT accounts suggest ‘savings’. The worst example is the MoD’s budgets since 2015, after the Osborne-Heywood deal known as ‘the Heywood wedge’, which created structural black holes and a rolling series of lies which has ended up with vast lies and vast deception of Parliament.

• HMT is rubbish at cost control which is the pretext for its control of budgets, constant micromanagement and undermining of serious planning. This was very clearly illustrated in covid but ignored.

• Spending Reviews are a terrible way to do long-term budgets.

• ‘Golden Rules’ have become farcical.

• The ‘Treasury business case process’ involves a bunch of PPE 27 year olds opining in detail on everything from whether a Fields Medallist should get money for a research project to how SF uses drones. Everything has to get ‘business case sign off’. This often shocks ministers who discover months after thinking ‘we decided to do X’ that ‘Treasury hasn’t signed off the business case’ so X is not actually happening. HMT officials will use this to scupper even PM priorities or agreed action by the PM and Chancellor if officials dislike them.

• HMT withholds data from No10 and the PM.

• HMT is terrible at data, computers, and analysis. ‘10ds’ terrified HMT officials because they realised their grip of power via having the best information was threatened. (C.f GOD’s memory of the ERM fiasco: ‘Getting hold of a Reuters screen and finding out what was happening to the exchange rate was difficult.’ After the 2008 crisis, they put the National Economic Council inside COBR ‘to be able to get [data] on the screens’.)

• The CO uses the HMT-No10 tension for its own purposes, adding another layer of bureaucratic knife-fighting, very hard for ministers to see, never mind interpret or act against.

There was much punditry about my changes in 2020 but they were modest and obviously sensible. For example, the ‘Orbanism’ (Jonathan Powell) amounted to:

A. I insisted on live sharing of data between No11 and No10 instead of the PM Office being denied information by HMT officials.

B. I insisted on one joint spad team instead of officials’ preferred option of two warring groups which officials can play off against each other.

C. I insisted we develop budgets and spending reviews jointly, no black boxes with the PMO informed in a huge rush at the end (HMT’s normal mode to maximise their power).

This wasn’t ‘Cummings trying to control the Treasury’. It was simply reducing absurd friction and improving planning.

Sunak and his main adviser agreed and stuck to them. These three simple, obvious changes greatly improved the quality of thinking and planning (vis our 2019 experience) especially during the chaos of covid.

As soon as I left Sunak’s team stopped this arrangement which contributed to No10 losing grip over Whitehall and a return to No10 vs No11 battles.

Changing HMT is connected to procurement reform, creating proper long-term budgets, creating a team dedicated to and expert in real cost control etc.

HMT can be sorted out by a mix of PM determination to use the PM’s full powers, other changes below, and judicious use of Cabinet Committees. It does not need primary legislation.

Should the Cabinet Secretary continue as head of the civil service, or someone else?

• Option A: The Permanent Secretary of HMT — the Warren Fisher model 1919-39.

• Option B: A separate head of the civil service — mostly the model from 1968 to 1981.

• Option C: The Cabinet Secretary — mostly the model since 1981.

• Option D: There is no head of the civil service — the Pitt-Palmerston era model. This would particularly make sense if you intend to do ‘less but better’ in No10 and empower other entities — including over how they do hire-fire-incentivise-train etc. There’s something aesthetically pleasing about the simplification of guillotining the disastrous bureaucratic rationalism of a ‘centre’ for the civil service — with endlessly proliferating directives, divisions, strategies etc — and a return to the organic English institutional evolution, bit-by-bit in response to circumstances.

See Robin Butler’s view (above) on the advantages for the Cabinet Secretary and CO in merging the two roles, which is exactly why they must be split up.

To what extent should the concept of permanent civil servants remain, should we replace ‘Permanent Secretaries’ with ‘directly responsible individuals’?

One can sum up many thousands of pages and millions of words written about Whitehall’s failures over decades — and why each ‘Inquiry’ embeds the next failures, and why each PM’s memoirs say the same things about their failed search for elusive ‘levers of power’ — in two short sentences.

A/ DRIs are a feature of all high performance organisations.

B/ DRIs scarcely exist in Whitehall now and the concept is deeply hostile to the dominant Whitehall culture.

A PM could quickly change this.

In Whitehall, the alternative to responsibility is well summarised by Gus O’Donnell — whether something is failing or not, ‘we need more resource’. If your project seems to work ‘we need more resource to reinforce it’. If it’s failing, ‘we need more resource to fix it’. If it’s a disaster, ‘we need much more resource to save it’. Nothing ever needs less resource. And stopping things so it needs no resources provokes the second largest degree of resistance you see in Whitehall. Officials will go to almost any lengths to make ministers accept the principle that X must continue, even if in much reduced form — they know the minister will soon be gone and X can spring back to life ‘with more resource’.

How to handle unavoidable tension between a) departments/agencies, b) the centre? How do you balance ensuring execution of priorities and avoiding bad micromanagement?

Departments are organised for specific functions with specific legal powers. They are mostly historical (e.g HMT, FO, HO). Many of the most important problems do not fit neatly into a single department. Departments resist coordination, information sharing and oversight.

Sometimes this is reasonable: rubbish individuals and institutions in the centre interfere ignorantly, cause duplication and confusion etc, so departments resist ‘coordination’. It’s often unreasonable: it’s a desire to hide money, hide failure, avoid responsibility. All of which applies to the modern No10/CO setup. Sometimes it’s a mix — ‘everyone’s right and everyone’s unhappy’ as they say in Russia, such as with budgets and SRs where HMT is right to think everyone is lying and cheating AND HMT’s process is dishonest and stupid and rewards lying and cheating.

Big companies have similar problems. It is very interesting to look at how Steve Jobs reorganised Apple in a very different way to almost all big companies to try to deal with some of these problems. I wrote a little about this a few years ago here. Nutshell: business history and theory make the case that as entrepreneurial firms grow large and complex, they must shift from a functional to a multidivisional structure to align accountability and control and prevent the congestion that occurs when countless decisions flow up the org chart to the very top. Jobs rejected this, fired all general managers, replaced divisional P&Ls with just one P&L, aligned incentives to company profits rather than any more local goals etc. I’ll go into how Whitehall could learn from this another time.

NB. ~100% of those in SW1 who attack ‘micromanagement’ are the most low agency people in Britain with zero history of successful management and zero idea how exceptional leaders often ‘micromanage’. Yet these clowns write columns about ‘reorganising No10’ — in between ‘how to beat Putin’ and ‘how to deliver a two state solution’ and ‘what Elon doesn’t understand about managing tech companies’.

WhatsApp groups are an inexhaustible supply of hilarious updates from the Bluesky NPC front. Here are a pundit and a professor often quoted by other NPCs on the subject of Whitehall. They bleat to their Bluesky friends how they couldn’t control their own diaries. ‘I always did everything I could’ but ‘unstoppable’… Yes, I’m sure you did do everything YOU could…! Top tip for MPs and others thinking about to do regime change: do not take management advice from characters who can’t exert their will over their diary secretary yet think they can give Elon management advice…

How to balance transparency and secrecy? Cf. the evolution of SF operations in Iraq: they shifted to radical transparency with intelligence, and accepting higher chances of leaks, in order to speed up the operational cycle a lot. This worked brilliantly at the operational and tactical levels (it did not and could not solve the fundamental political problems at the root of the Iraq fiasco, but that’s not the important point). When does it not work, which parts of government can’t operate safely like this?

How to fix the current mess with the services, MoD, agencies, JIO/JIC, NSS etc? What lessons from CID and the chiefs of staff committee? It’s interesting that Asquith chose not to enforce CID’s conclusions on the military pre-1914. It was not involved in planning with France. Per the Esher Committee 1904, it’s possible for our flexible constitution to allow the creation of entities that report to the PM directly with wide powers.

This is a vast topic I’ll return to.

How to create a place which brings together all thinking about war across all its different domains (conventional, SF, space, intelligence, propaganda, diplomacy etc)? Most of you will be thinking — wait, this place must already exist, right?! NO! It does not! I found this out in 2019 when I asked the Permanent Secretary to MoD where this place is situated in the system and he gave a hollow laugh and said, ‘It doesn’t exist, unless you count Cabinet.’ They’ve gone through the Ukraine tragi-comedy without such an entity and I’ve pointed out the total lack of coherent ends-ways-means from the start.

How to use Cabinet Committees? Recently close to 100% have been too big even though they were invented to cut the size and improve focus. But they can work well in the British system. Part of the reason is that constitutionally they have the same force as full Cabinet (something not grasped by most MPs).

The most effective conventional things 2019-20 were the XS/XO Cabinet Committee system which was sort of cloned to the Covid Taskforce. Crucial was limiting the number of ministers. XS was half a dozen. XO only had one permanent minister chairing. In covid it was critical to avoid a large number of ministers always there.

This connects to changes in HMT and long term budgets (above).

How much of the CRAG Act 2010 to amend/repeal?

E.g Butler et al watched the Blair 1997 Order in Council and were determined to ensure it wasn’t repeated.

I have come to the view that a Civil Service Bill is right because I was responsible for the Order in Council that enabled up to three special advisers from Number 10 to give instructions to civil servants, and I recommended Mr Blair to do that because in fact that was what was happening in the case of Alastair Campbell and Jonathan Powell, so I thought we had better be legal. But it was so easily done, it was done the first weekend by an Order in Council and it rather shook me to realise how easily the fundamental structure of our civil service could be changed, and once that Rubicon was crossed you could never go back. Therefore I thought it would be right to entrench this through an Act of Parliament [CRAG]. (Butler)

How much of CRAG needs to be repealed given other things one can do?

What are the most important things that can’t be done until it is changed?

What’s the worst that Commissioners can do to delay until they are removed?

What’s the scope for the usual activists to bring JRs under CRAG to try to delay a new government?

How to mitigate these delays?

E.g the Civil Service Commissioners must go. This job is the PMO’s, see below. But their existence can be finessed for a few months while legislation goes through.

How can Parliament be restored as the place capable of controlling the power of the executive and courts?

E.g Select Committees are particularly poor on defence and security. Anonymous X accounts provide more useful information on the MoD’s uselessness than the Defence Select Committee which is told much less than junior hacks.

How to restore the reputation of MPs?

Part of the answer is pay. MP pay should reflect productivity growth in the private sector which is also at the core of what money is available for the public sector. I would tie it to some metric like median private sector wages. If they go up, MPs get a raise. If they fall, the MPs get a cut. Incentives are aligned in a healthy way. MPs can’t game this. And it makes them think about everything they’re voting on: does this improve private sector productivity?

I wanted to do this in No10. I suggested to Sunak in December 2022 that he do this.

Also stop all taxpayer funding for political parties. This would be popular and it would immediately dish the left. Correct in principle, bad for the left, popular.

How to do medium term planning much better? This is extremely hard. It inevitably involves constant adjustment of plans because reality proves assumptions wrong. There is no magic process to make it work. Governments everywhere struggle with the same problems. Senior people inevitably get sucked into one crisis after another so longer term planning is pushed to less senior people therefore it has less traction. The intelligence services do not do ‘own side’ intelligence so that machinery is not applied to many crucial problems. (When I asked JIC to consider the likelihood of chaos after the 2020 US election, the JIC chairman said it was the first time they’d done something on America.)

Another huge problem I’ll return to.

What’s the role of spads given other decisions? Spads have grown in importance and number in proportion to the increasing powers of officials vis ministers and the decreasing performance of Whitehall. Many now say ‘a serious government needs far more, and better, spads’. This is true if you are trying to make the existing system work better. But it is *not* necessarily true if you are trying to change the system profoundly. Spads now are partly ‘a person who answers to a minister unlike ~100% of the rest’ but if the minister is truly responsible for the department and can hire/fire people to senior jobs, then the nature of ‘spads’ inevitably is transformed.

How to deal with the growth of judicial review and the quiet revolution whereby officials and government lawyers, working with a network outside government, have redefined ‘respect for the rule of law’ to mean ‘doing what lawyers claim is in keeping with international law’ which is very elastically defined to include many things which are not incorporated in domestic law?

Another vast subject I’ve touched on before including in covid evidence. Officials and lawyers have been nudging the system year by year towards a state where officials refuse to do what ministers decide on the basis that ‘it’s incompatible with the rule of law’ by which they increasingly mean ‘lawyers’ interpretation of international law’.

E.g the appalling head of the government legal service in the CO resigned in 2020 because he opposed the government’s actions with the Internal Market Bill on the basis of an ‘attack on the rule of law’ by which he meant ‘international law’.

The constitutional position in Britain for many centuries has been that international law is *not* binding on the government, nor does it direct the conduct of officials, unless implemented in domestic law. The new position is extremely radical but purports to be conservative. The next phase of this campaign is to insinuate that passing domestic legislation to restate the traditional position is itself ‘an attack on the rule of law’ which should be rejected by officials ‘to defend the rule of law’ — an Alice in Wonderland subversion.

This shape-shifting has been possible because of the degradation of quality of MPs who often do not know the basics of the constitution. The rot of the Tory Party is well summarised by how they went along with this for 14 years instead of stamping on it. NPC ministers like Cleverley trotted out on TV lines articulating the far left position over and over without even realising what they’re doing (horrors like Grieve were different, they know what they’re doing).

How to create a ‘war mode’? Bonar Law’s point about Rome having a peace mode and a war mode for the constitution is of great importance. Covid exemplified this. A pandemic or war forces government to do things in ways which should not be normal in peace time. War demands that the balance shift between speed, clear authority etc over established freedoms, normal processes etc. Ideally we would have a more Roman approach and clarify peace/war mode. Though there will also also be a tendency for some to invoke war mode, just as governments are drawn to using terrorist legislation for non-terrorism ends.

How to structure No10?

Some thumbnail historical sketches:

• Pitt. Phenomenally able PM, great wisdom with youthful energy. Hired whoever he wanted to work with him in No10, no real rules in any modern sense. Ministers truly responsible to Parliament. No permanent civil service in the modern sense though there were officials of various kinds appointed for long stretches. Talent acquisition taken seriously, many brilliant young people brought in. Infrastructure and procurement taken very seriously. Responsibility and authority often co-located. Nepotism was normal, sometimes bad, sometimes a shortcut to great talent. None of the modern ‘anti-corruption’ machinery but vastly less waste and corruption than now, cf. the long-term relationship of BAE and MoD. And Westminster jumped on corruption while now it authorises it and spins for it.

• Churchill. A patchily phenomenally able PM of great spiritual depth drawing on the residues of an aristocratic culture, with self-awareness of some flaws, hence after many disasters appointing and keeping Alanbrooke whose chiefs of staff committee ran the war competently and forced focus on the critical questions. Some rules in the modern sense on hiring in No10 but not a critical barrier in the war, though and other Whitehall pathologies were damaging (e.g the tendency to endless committees). Ministers more responsible to Parliament than now but already responsibility diluted by modern bureaucracy. Cabinet not great, huge time wasted in meandering discussions of Cabinet and Cabinet committees. Whitehall showing modern signs of sloth and decrepitude but able people still able to tunnel through or force change and speed.

• Thatcher. Strong leader, many virtues. Poor at handling her team. Some able ministers (relative to today’s). Ministerial responsibility to Parliament greater than today but closer to today than to 1850. (Carrington’s resignation over the Falklands was praised partly because he was NOT seen as actually responsible for the FO’s failures.) Civil Service very powerful. She could push it on her priorities but Yes, Minister was real and she didn’t change its evolution, contra advice from John Hoskyns that this was fundamental to achieving her policy goals. Much of Whitehall continued in a Left direction uncurbed by ministers. Public services, infrastructure etc in many ways went backwards (e.g exam reform and National Curriculum a huge cultural vandalism-disaster). Many ideas from Tory think tanks about improving government actually created the nightmare of quangos and agencies we’re cursed with now. She exemplified the modern situation: a very determined PM, sleeping a few hours per day, can make the system bend on a few big things but outside the ‘lighthouse beam’ of direct PM focused energy, the system drifts left and decays.

• Blair. After first term failure it evolved to: Policy Unit, Strategy Unit, Delivery Unit, hugely powerful HMT/Chancellor. Huge focus on talking to the old media. Able to push through a few priorities connected to communication with the old media, e.g waiting lists. Main domestic legacy: constitutional legislation like HRA, Equality Act etc. Blair admits he only really understood the problems of Whitehall and No10 after his first term then he was stuck in Iraq and never got the chance to change it properly. More than any PM since Thatcher he’s publicly stated obvious truths about how Whitehall evolved but his successors ignored him. Treasury and Cabinet Office both grew in power. Cabinet shrank in power and influence. No10 seemed to become more powerful because it developed the No10 as Media Entertainment Service model, based on ideas from Gould and Mandelson observing the 1980s and early 1990s in America. Blair’s personal dominance of Westminster camouflaged the deeper trends in Whitehall. By 2010, Whitehall was much further left and much more powerful viz ministers and the country.

• Tories 2010-24. With the exception of 2019-20, no clear plan and no challenge to business-as-usual in Whitehall. Ministers as NPCs. Ironically, Blair’s successors preferred to copy the Media Entertainment Service model he started with rather than his post-No10 advice — and were less politically successful. They couldn’t see the Media Entertainment Service model was intrinsically bound up in Blair’s period being the last phase of the old centralised media’s dominance, the crackup of the phase beginning in the 19th Century with mass newspapers and telegraph etc. As Blair ruled, the internet was breaking the foundations of the model. I saw the Cameron project prepare for 2010. They were clueless about Whitehall. Maude, Letwin et al trusted the system and longed to sink into its embrace. Over and over that set of Tories said to me, before 2010 and long after, ‘There’s nothing wrong with the civil service, Dom, other than Blair being in charge for a decade, once the natural party of government is back in power, you’ll see…’ Our team in the DfE was more successful than others. Why? Partly because we prepared legislation and plans in advance unlike the rest of the Cameron project. Partly because 2011-12 we purged 15 of the top 20 people and built a Private Office supportive of plans rather than sabotaging them, something never mentioned, obviously, in any analysis of our success. Tory pundit/think tank world prefers to push fairy tales about ‘a strong minister with a clear idea and good communication’.

One way for an aspiring PM to think about how to approach the CO:

Option A: Establishmenty-’reform’.

• ‘The PM Office must be strengthened!’ ‘More spads, better spads’. A Policy Unit as influential as Thatcher’s (NB. it was <3). A Strategy Unit and Delivery Unit like Blair.

• ‘The Cabinet Office must be reformed, smaller, more focused!’

• Reheated Francis Maude / Michael Barber rhetoric about ‘civil service reform but this time we mean it’ in line with dozens of attempts since the 1960s.

• Outcome? The usual. Attempts to strengthen both the CO and PMO together simply intensify the core dysfunctions and the inherent power struggle between the political staff and officials, because most power in No10 is zero sum between No10/CO/HMT. ‘Reform’ rhetoric turns into committees and powerpoints for White papers. Officials have seen it all before. After 18 months it’s obviously dead. Government in mid-terms. Another term down the toilet.

Repeated attempts at reforming the Cabinet Office and civil service generally have, whatever the story, a) never led to the PM thinking ‘it now works’ and b) never solved the recurrent deep problems of the civil service repeated constantly across the political spectrum since the 1968 Fulton Report. The standard pattern applied to everything — consent, delay, evade — is particularly potent when applied to preserving their own powers.

Conventional ‘reform’ doesn’t work because it doesn’t touch the core issues of talent, responsibility, incentives, culture and power. It does not challenge the core of the Northcote-Trevelyan system. And it’s done by people a) with no experience of creating a high performance organisation who b) have been trained to believe fairy tales about the ‘meritocratic world-renowned’ system.

Option B: solve the core problems repeated for decades and best explained in Yes, Minister.

A possible path:

1. Close the Cabinet Office. Things in it either a) stop (e.g the second separate vast communication empire), b) return to departments/other entities (e.g much of NSS), or c) shift to the control/responsibility of the PM’s Office (e.g things supposed to be for ‘coordinating’ and ‘delivery). NB. The PM is also minister for the civil service (unknown to many in SW1!).

2. The Cabinet Secretariat returns to being responsible to the PM’s Office (as in 1916) to do the original job: agenda, minutes for Cabinet, secretarial service for Cabinet Committees etc. NB. The Secretariat provides secretarial services for Cabinet committees, it *stops* controlling them, writing its own policy papers for them etc. And it is not in charge of ‘policy’ or ‘delivery’ or ‘strategy’. The Secretariat becomes an office of <~20 (given the changes to technology it can be smaller than under MH) and, like under MH, it should draw staff from departments so they are familiar with material and are not seen, as they are now, as interfering.

3. Attached to the PM’s Office, working with the Secretariat, is a Library and a Librarian of integrity and scholarship. The Library maintains records and has a classified reading room for people to read papers that should not be circulated/retained.* The Secretariat hands its papers to the Library. Rules on access to previous government papers change so they’re optimised for ‘enable the system to learn from mistakes’, not ‘hide failure’.

4. The Cabinet Secretary runs the Secretariat only. The Cabinet Secretary is not responsible for policy advice (per 1916), is not responsible for Whitehall personnel decisions beyond the Secretariat (per 1916), and is not head of the civil service (per 1916). It’s an interesting job but with about 1,000X less power than now and much less than ministers instead of being much more powerful than ministers.

5. The PM appoints someone to run the PMO who is in charge of the entire No10/central entity. The division of official and ‘political’ ends, the inbuilt friction and warring between tribes ends. This does not, obviously, eliminate all the normal problems of the centre of power. It clarifies responsibility for handling these problems: the PM and the ‘CEO’ of the PMO.

6. There is no ‘head of the civil service’ in charge of promotions and appointments across the system. The only central entity with power over appointments across the system is the PMO — in the sense that it can trump anybody else and insist on hires/fires — but PMO engagement is limited to action needed to enforce PM priorities and intervene when it sees failure, when it must interfere much faster and more effectively than now. There is no general need for the PMO or any central entity to interfere in personnel elsewhere unless it is failing. The point of making DRIs central to departments/agencies is precisely to restore direct personal responsibility and to ensure power is co-located with responsibility, per General Groves.

7. Civil service appointments shift from ‘closed by default’ to ‘open by default’, i.e open to applicants from anywhere. The culture of the civil service being a closed caste appointing and promoting almost entirely from within — with some Potemkin processes whereby senior people go off for a few years so they can claim ‘private sector experience’ — ends. Hiring becomes actually meritocratic instead of fake meritocratic.

8. A National Security Adviser and team, responsible to the PMO, runs those things in the PM’s office which need central coordination including the NSC. E.g One of these things is defining priorities for the intelligence services, a function now conducted (badly) away from all ministerial eyes in the bowels of NSS.

9. ‘10ds’ (data science/AI) strengthened in the PMO. Also in charge of ensuring all data relevant to the PMO, including all data held by HMT, is available live and in the right format. Responsible for quality assuring all data that goes to the PM so we don’t have repeats of the Cabinet Office putting up obviously fake graphs to the PM showing the entire country travelling on HS2 to justify continuing the obviously catastrophic project. (This was exposed as fake and the PM went ahead in January 2020 knowing it was fake but that’s a separate point.)

10. A small Systems Team, combining experts in policy and project management, in the PMO for monitoring and enforcing priorities and watching how things interact. This includes managing what replaces ‘Permanent Secretaries’, i.e DRIs. A model for this ‘systems management’ is George Muller’s approach to Apollo.

11. Ministers have power to hire/fire and the ‘official’ head of their department is a DRI with a very clear definition of priorities for the term of the government. The PMO can intervene but the bias should be: state priorities, make people truly responsible, remove fast if failing, do not set up duplicating processes in the centre to compensate for department/agency failure.

12. Many other things I won’t go into now. E.g A crucial question is: there is no entity in the British state whose job it is to think about and prepare for war/conflict across all domains then coordinate it when it happens. Its absence is obvious in the utter shambles of Ukraine ‘strategy’, the total inability to define any even vaguely coherent ends-ways-means, the way MoD was allowed to get worse and worse during a war etc. Such an entity must exist. But how and where depends on many other questions.

This path means the PM exercising powers more fiercely than has been seen since 1945 but in narrower fields and the most basic shift is to a) restore the PM in charge of the centre, b) clarify PM priorities and direct responsibility, and c) shift to real meritocracy — all of which replaces the fake responsibility and fake meritocracy we’re suffering under today.

Until the quality of Ministers improves a lot some will try to make stupid appointments but it will all be out in the open — they will actually be responsible for stupid appointments instead of evading responsibility like now. Errors will always happen but this new system means they’ll be fixed faster and a core job of the PMO will be thinking about talent, just as the core job of a CEO is thinking about talent. Ultimate responsibility will be with the elected PM, not the unelected Cabinet Secretary.

* There’s a better way of recording decisions than the 1916 Minutes system. The Secretariat was originally justified by the lack of clarity over Cabinet discussions. An easy way to handle it now is what I did 2019-20 for various entities including XS, XO and the Covid Taskforce — I said to the Cabinet Secretary that agreed action should be typed on a screen where all can see and at the end of the meeting these actions would be agreed then executed to agreed timelines. The Cabinet Room did not have, or allow, screens in 2019. XS and XO were conducted (mainly) in COBR which has screens. In covid I got the meetings moved out of COBR into Cabinet — to escape the strict security protocols such as no open internet — and put screens in the Cabinet Room. This low tech upgrade helped clarify real problems a lot and speeded things up because action was clear, agreed, and then instantly communicated without delays over detailed minutes (rarely of use anyway). And the twice weekly rhythm of XO meant it operated in a completely different way to normal Whitehall. For the purposes of good government, this system is much faster and more effective than the old minutes system. The advantages of minutes to historians is not the point.

---

‘Regime completeness’

A last thought…

If you, dear next PM, think ‘surely I can make Establishment-y reform work’ — you will have to explain exactly why your scheme, leaving the CO intact, will succeed where all previous attempts have failed. If Thatcher presided over the spread of Yes, Minister, why will you succeed? What is your plan to cope with ‘consent, delay, evade’?

If you do not close the CO immediately and create a system where the PM is again in charge of ‘the centre’, then you and your spads will find yourselves in the same cycle of meetings as everyone else for decades, with the same results.

Further, it’s foolish to worry about the ‘noise’. If you aspire to be a PM who turns the tide, then you will inevitably be attacked as ‘fascist’ for core policies on immigration, boats, asylum etc. And you will have to deliver on those things. So the important thing is to deliver and win the argument with the voters and crush the London Insider class as soon as possible. And this means: close the CO and control what must be controlled. The screeching from former Permanent Secretaries will be lost amid the louder screeching and their are huge advantages in compressing the row over Whitehall into the row over other things in the first year. And the time to strike on such things is when you are at the height of your prestige, the height of democratic legitimacy, after winning the election. That is the time to do the hardest things, when the system feels least able to resist. I tried and failed to persuade the Trolley of this in January 2020. Let his abject failure — his contemptible waste of our efforts in 2019 — be a warning to you.

Yes, being the PM who decides that ‘the government controls the government’ will be ‘noisy’.

But there is no escaping ‘noise. You can have the ‘noise’ of Cameron and May and Sunak and Starmer — relentless contempt and loathing. Or you can have the noise of a once-a-century leader inevitably attacked by a subset of Insiders-hacks-academics but with the upside of winning public support and the self-respect of being the Live Player who finally faces reality. You can be attacked for being a ‘puppet’ controlled by others or for being a ‘fascist/tyrant’. As Balfour remarked about Lloyd George above, PMs are always attacked for one or the other.

This raises the issue of ‘regime completeness’.

What do I mean by this?

Many ideas about change now seem sensible in their own terms but if you understand the system you know idea X has almost no practical chance of happening because other parts of the system will scupper it. X cannot happen in isolation because of the systems crisis.

For example, genuinely first rate change of defence procurement. One can write schemes embodying the right approach and we should do this. But it is impossible to implement them without also having many other things in place including a supportive PM, control of HMT and CO etc. Otherwise your scheme for defence procurement goes into the HMT and CO processes and emerges as … not the new scheme but continuing with the old system with some spin, quite possibly even worse than the old system. Defence procurement is regime complete.

Similarly, taking AI and robotics seriously. One can write policies and plans but we’ve seen repeatedly the sabotage of such plans. Making technology a true priority is regime complete.

Regime completeness means:

1. A Leader with certain skills and temperament/sensibility who grasps the scale of the crisis and would prefer to fail unconventionally than fail like every PM for decades.

2. A network of elite talent.

3. A detailed Plan — for policy, for law, for bureaucratic changes — and the Plan To Do the Plan, i.e for truly grasping power and ensuring it’s a government that controls the government, the plan for making the plan happen amid the power system which exists or will exist. Such plans are not just ‘policy’ plans. They are also political, bureaucratic-Jedi, and dynamic.

4. A political strategy which can win an election and sustain a government trying to do the Plan.

5. A communication machine which can figure out how to tell the big picture story of a radically different government and handle the tactics. The collapse of the old media and the old audiences means this needs new ideas and skills but also means, I think, there is a lucky coincidence. If you can build the right machine at this moment, you will be the only entity in Britain which can do mass political communication with elite performance, providing a great strategic advantage and a tail wind for the endeavour. If you’re worried about the ‘noise’ then build the Meme Machine to overwhelm it.

6. Creating a positive flywheel of execution and story-telling. Success in reality and with voters — beyond SW1’s ephemeral hysterical news cycles — builds supportive coalitions which allows further progress and crushes opponents. I’m often described as a cynic but I’ve always thought that bringing the culture of elite high performance to Westminster would be mostly very popular outside Westminster (and universities) if you can sustain it for a few years so voters see their lives improve. In contrast, the conventional wisdom of Insiders has been that voters just follow the news and don’t care about the quality of government.

And regime completeness brings us back to the character and sensibility of the Leader.

A Leader will have to a) reassure a lot of Insiders while b) showing the country they are serious about real change and c) recruiting elite talent needed to execute real change.

These things are inherently in tension. And any Leader aspiring to turn the tide — who takes over the Tories, or creates a Startup Party, or Farage — who inches close to power will come under extreme pressure from a vast network of powerful/influential people to ‘play the game’, i.e not disrupt how power and money work in Westminster:

Of course you must win, winning means saying all sorts, we understand, but… All this talk of closing the Cabinet Office, of repealing the HRA, of firing tens of thousands and ripping up how Whitehall does everything. All this will just provoke a resistance you can’t beat. It creates too many enemies. It’s not what Thatcher did. I’ll always be on your side, of course, but so many won’t be if you go down this path. It would be such a tragedy for you to sabotage yourself at the start. There’s another way. With just a few appointments and a few subtle changes, you can get almost everything you want. And you can focus on what the voters care about and you’ll go down in history. The British system is subtle. The PM’s powers are vast. You can get what you want. Just don’t go to war with the system. Don’t start purging the Permanent Secretaries. Don’t provoke a Whitehall strike at the start of your government. Don’t become captured by revolutionaries. We’ll help you get what you want while staying friends with the system. Now, just appoint X to run your office, it will send a reassuring signal. Next…

This will be all the more powerful because it will contain some truth. A PM is powerful but can’t make too many enemies. Battles must be sequenced. Much of London will go berserk if things like appointments and procurement truly change. The old system has large networks of people and money dependent on the old ways continuing.

But the core of the tension must be faced. A PM who continues broadly in the post-Churchill tradition can make friends with the system, be treated by it with respect, and after they’ve left to write memoirs about their search for the elusive ‘levers of power’, they’ll all be friends again and get made Chairman of the British Museum or whatever — like Major, Blair, Cameron, Sunak. But they’ll leave the pathological CO in charge, feed Whitehall’s pathological institutions with more money, they won’t be a government that controls the government, so they won’t do what they’ll say they’ll do to win the 2029 election. They’ll drive the system deeper into crisis, maybe over the brink.

The Leader who can lead a regime complete government, therefore, is necessarily one who is prepared to gamble on both failure and being enemies of the Insider class in order to grasp at true greatness. They’ll need the very odd mix of skills and sensibility and endurance for an exhausting decade of operating within masks, behind each mask another mask. They’ll be hated by a section of the country in a way Cameron and Sunak — or even Thatcher — are not hated. But they also might be a true historic character whose name will echo for generations, like Pitt or Churchill.

Such people are few and far between. It takes luck to get into position to make such choices. The intersection of the characters and the luck is beyond calculation. Regime completeness is super, super hard and there’s a logic to why it tends to be wars and revolutions that throw such characters up in the moment.

But our alternative to seeking regime completeness is continued rot, ruin and our most dangerous crisis since 1939-42 — to be faced with a lot less inherited strength than 1939-42.

Finally, I must admit that my ideas above are based on another crucial difference between me and Insider conventional wisdom. I think that many of Insiders’ worst errors are generated by false ideas they have about psychology and communication. But I must admit that despite winning campaigns I’ve had close to zero success in persuading Insiders of why we won and why their ideas are false. The Insider perspective is that I won campaigns because of luck, not the reasons I think, therefore the above ideas are nonsense. But if you listen carefully to the Insiders, you will see that almost everything they say about ‘Whitehall’ and ‘delivery’ is really based on their assumptions about the psychology of voters and communication — which they’ve repeatedly demonstrated they do not understand…

---

Please post sources below for people and the AI(!) to study on how to do regime change after Starmer.

There’s a book, The Private Office (Henderson). Someone should use LLMs to pull together modern evidence from different sources (oral histories, memoirs etc) of how the PM’s Private Office works. If someone does this to a standard that’s useful to me, along with a short explanation of how they used LLMs to do it, I’ll send them £5,000.

Another challenge. There is an epic Pitt biography by John Ehrman in three volumes and over 2,000 pages. At some point I’ll read it myself. But I have a lot on my plate. Can someone feed the 3 volumes into LLMs and extract notes that mirror my interests as displayed in my blog?

• Examples of the unrecognised simplicities of high performance, similar to the last chapter of Groves’ book on responsibility/authority, speed, talent etc.

• Examples of how Whitehall/No10 was organised in particularly bad or good ways relevant to our debates. E.g How did Pitt deal with HMT? How did Cabinet evolve under Pitt? How did Whitehall evolve?

• Write a 5,000 word max briefing on this book for the next PM.

If someone does this to a standard that’s useful to me, along with a short explanation of how they used LLMs to do it, then I’ll send them £5,000.

Thanks for subscribing…

September to December will see more stuff on the Actual Plan for 2028/9 and the possibilities for the Startup Party project — either a pure startup or an existing entity effectively closing and re-opening as TSP…