Mark Sedwill describes what mid-term transitions teach the centre
Every incoming Prime Minister experiences the same first day, whether they arrive after a general election or a mid-term change of leader: the audience with the Sovereign, the speech in Downing Street, the handshake with the Cabinet Secretary and applause from the No 10 staff, congratulatory calls with foreign leaders and the briefing on the nuclear deterrent, when the burden of the job really sinks in.
The ritual is identical. What follows is not. And it is in that divergence – not the ceremony – that the Civil Service does its most consequential, least visible work.
If Andy Burnham becomes Prime Minister this summer, it will be the fifth mid-term change of occupant in Downing Street in under a decade: David Cameron to Theresa May, Theresa May to Boris Johnson, Boris Johnson to Liz Truss, Liz Truss to Rishi Sunak and now Keir Starmer to Andy Burnham. This might have been expected to produce a settled playbook. It hasn’t. What it has produced, for those of us who have lived through several of these moments, is a clearer sense of where the fixed points lie and where everything turns on judgement.
Continuity of governance
The constitutional starting point is that Britain does not elect a Prime Minister; it elects a House of Commons, which then settles on a government able to command its confidence. A mid-term change of leader within the governing party does not disturb that settlement, so there is a strong presumption of continuity. The opposition invariably calls for a general election; the last Prime Minister to seek a fresh mandate on appointment was Stanley Baldwin, over a century ago. He lost.
That presumption did real work in 2019, when Theresa May resigned in early June and fragile parliamentary arithmetic made the timing of the Conservative leadership contest a genuine constitutional question for me as Cabinet Secretary. I insisted the contest conclude before the summer recess, so Parliament had the opportunity to test confidence in the new Prime Minister on appointment, because a minority government propped up by confidence-and-supply could not be assumed to transfer intact. Where a governing party holds a working majority, as now, that calculation falls away. The underlying discipline does not: the Cabinet Secretary remains the guardian of the constitutional conventions and the judgement calls cannot be reduced to a rulebook.
Policy initiatives by an outgoing government are likewise a matter of judgement. In arguably the most significant decision of her entire administration, Theresa May’s government committed the UK to net zero by 2050 in her final weeks in office. Some were uneasy at so strategic a decision being taken after her resignation announcement, but it went through Cabinet and Parliament with cross-party consensus, so it was clear it would endure. Contrast that with the failure to establish a “burning injustices” commission to reflect the agenda she had set out in her Day One speech. That is the test: not whether an outgoing administration has the votes to force a decision through, but whether it commands consensus durable enough to outlast the government that made it.
The obligation underneath all of this is simple to state and hard to practise: the Government of the day is the Government until it isn’t, and officials serve it in full right up to the last moment, however confidently Westminster is gossiping about the succession. That discipline is what allows the small, discreet planning that does have to happen – within the Cabinet Secretariat, not across the wider Civil Service – to proceed without destabilising the business of government, which continues regardless: hundreds of decisions taken, warrants signed, statutory deadlines met.
That continuity has a harder edge. The nuclear deterrent does not pause for a change of Prime Minister: a new Prime Minister writes fresh letters of last resort within hours of taking office and the outgoing letters are destroyed unread, whether the changeover follows an electoral landslide or a resignation announced on a Wednesday afternoon. If a crisis breaks in the changeover window itself, the machinery for chairing the Cabinet Office Briefing Rooms (COBR) and taking a decision has to work regardless of who currently holds the title, which is why the national security architecture around the Prime Minister is designed to survive a gap in the principal, not just a change of ministers around them. It is the sharpest illustration of the wider principle in this piece: continuity of governance. Some functions cannot tolerate a gap and the officials who run them plan the handover long before there is a name to put on the letter.
Day One
The Day One briefing – usually day zero or day minus several for a mid-term transition – is the Cabinet Secretary’s opportunity with the new Prime Minister, and each Permanent Secretary’s with their new secretary of state, to set the register of the ministerial relationship with the machine that will serve them. The narrow version runs through the Government’s stated priorities and moves on. The better version lays out the scale of the strategic challenges in full – the climate, demographic, technological, geoeconomic and geopolitical forces converging on a fiscally constrained economy – and the hard choices that follow, whether or not they match the priorities with which the Government arrived. Done well, that conversation is completely candid about the challenges of the job and signals complete confidence that the principal can handle being told the truth. Officials who soften it to make the moment comfortable do a new Minister no favours: candour established on day one becomes the working culture; candour introduced later has to overcome the disappointment of rumbling bad news.
Britain is both an island economy and the world’s second-largest service economy – a more exposed combination than it sounds. Two-thirds of national income is traded and no other G7 economy is shaped as much by global trends and shocks. For a middle power in that position, climate, demographic and technological change, layered onto geopolitical and geoeconomic competition and a capricious American administration, amount to a permanent revolution rather than a cyclical condition. It arrives as higher inflation, higher unemployment and higher demands on an already stretched state, on top of productivity that has lagged America, France and Germany for a generation. A ministerial team that hears that arithmetic on day one, rather than in fragments over the following months, starts the job understanding why the response has to match the exposure.
That exposure has a diplomatic dimension, too. A change of Prime Minister is read abroad as a signal, whatever the domestic explanation for it, and allies and adversaries seek to interpret it. Live negotiations do not pause for a leadership contest – a trade round, a defence commitment, a position ahead of a summit – and someone has to be able to tell a foreign counterpart, credibly, whether Britain’s position is going to change or remain the same. The practical discipline is straightforward. The new Prime Minister’s first calls to allied leaders should happen within hours and days, not weeks. Ambassadors need lines to give quizzical counterparts before they have to ask. And officials managing an ongoing negotiation should assume, unless told otherwise, that the UK’s position holds through the transition, rather than wait for fresh instructions that might take weeks to arrive.
Machinery of government is policy
The sharpest lesson from the last decade of transitions is also the least understood outside the centre: the machinery-of-government decision that shapes a new government is not administrative housekeeping – and it is not really a first-day decision, either. It has to be settled in the days before, because appointing the great offices of state is itself a Day One act, and a Prime Minister cannot appoint ministers to jobs that do not yet exist or that might be abolished or reconfigured. It is the first substantive policy choice of the incoming Prime Minister, taken before the government exists, and it hardens into fact faster than almost anyone entering office expects.
Theresa May’s decision to create the Department for International Trade and the Department for Exiting the European Union illustrates the point. It was settled in the days before she became Prime Minister. Both departments made obvious political sense for a leader who had campaigned to remain in the European Union and needed to demonstrate she would deliver Brexit. However, putting trade policy in the hands of a department tasked from day one with striking deals with third countries effectively closed off the European Economic Area, the customs union and single market as viable end-states for the UK before a single negotiating session had taken place with Brussels. Government policy had not then been set. The negotiation later moved back toward something closer to a customs arrangement. Parliament rejected it and a harder Brexit followed. The point is not which approach was right. It is that the policy was effectively set by the organisational chart agreed before Cabinet was appointed and, once set, unwinding it would have cost more political capital than any government – let alone one that became a minority government months later – had to spare.
The lesson for officials advising an incoming Prime Minister is clear: press the machinery question in the days beforehand and be clear about the policy impact, because the first Cabinet appointments cannot be announced until it is settled. Once departments exist and ministers acquire pet initiatives and mini-funds, institutional behaviour reasserts itself fast. If Theresa May’s machinery decision effectively set policy before her government had considered it, Tony Blair’s experience shows the impact of not taking decisions on machinery early enough. By his own account, he regrets not restructuring government machinery in his first term as he did in his second, despite a landslide majority, and wishes he had spent the political capital of 1997 sooner. A Prime Minister who cannot be confident of a decade in office – which, in recent form, describes most of them – has even less time to spare.
The Prime Minister’s team
Good transitions share a structural feature: the person who plans the transition should not be the person running the leadership campaign. Boris Johnson’s appointment of Ed Lister – kept entirely off the campaign trail, free to focus on what a functioning No 10 would need – is close to a template. It reflects a wider truth about candidates for the premiership: most are superstitious about being seen to plan for a job they haven’t won, in the same way a government can be reluctant to permit contingency planning for an outcome it doesn’t want, as before the 2016 referendum. Officials should expect that reluctance and work with it rather than wait for explicit instruction.
The second judgement the centre has to make quickly is about the incoming Prime Minister’s authentic operating style – there is no single template for the job and imitating the wrong one fails fast. Some prime ministers govern like President Reagan: setting direction, communicating, taking the big decisions, delegating implementation. Others govern like Mrs Thatcher: immersed in the detail of every department. Both can work. What doesn’t work is a big-picture leader forced into a details-first mode, or vice versa.
The effective No 10s of the last three decades – notably Tony Blair’s and David Cameron’s – had small, stable core teams with high trust, who kept internal friction private and remembered their job was to serve the boss’s agenda, not their own. The less effective ones fractured. The chiefs of staff who make the stable model work understand a discipline the American military captures well: don’t wear the stars. The job is to be the Prime Minister’s voice, not a public figure in their own right; every chief of staff who has sought their own profile has weakened the operation around them – and weakened their Prime Minister. Jonathan Powell, Ed Llewellyn and Gavin Barwell are good models to follow.
The Burnham transition
Andy Burnham’s big idea is devolution. If he takes the ambitious approach and decides England’s devolved functions – health, education, transport, culture – are to move north with his “No 10 North”, that has to be settled in the days before he takes office, not worked out afterwards, so that the ministers he appoints on Day One are being appointed to a structure that already exists. Left until ministers are in post, it risks becoming a political gimmick disconnected from the power centres of Whitehall. Devolution is not like most policy choices, where a government can correct course after a slow start; departmental geography sets the incentives that shape delivery for years, because officials build their careers, their networks and their measures of success around whichever structure exists. Unwinding the wrong model several years in costs far more than getting it right from the outset. That is what makes the machinery decision more than administrative housekeeping for this particular agenda: it is the mechanism, not merely the message. Officials supporting that transition should press the structural question first, resist the pull toward a quiet, politics-free August and be candid that the machinery choice is not preparatory to the policy – it is the policy, and it will not be easy to revisit.
Two further points apply beyond any one government. The scale of economic transformation now required – across labour and capital markets, energy security, welfare and pensions, skills, defence, public-service productivity – has to move at the pace the underlying trends are moving, leaving no room for scarce political bandwidth on second-order policies that please activists without shifting the trajectory. Relentless focus on a handful of priorities is close to the only viable strategy. And an agenda at that scale should draw on centre-right as well as centre-left thinking – job-market liberalisation alongside welfare-to-work, regulatory as well as planning reform, a defence-anchored industrial strategy. Officials should expect to serve a blend rather than assume an administration’s lineage predicts its programme.
None of this is a rulebook. Civil servants, who like precedent and structure, should resist the temptation to treat it as one. What the last decade of mid-term transitions offers instead is a discipline: serve the outgoing government fully until the moment it ends; treat the machinery-of-government decision as the first and most consequential policy choice a new Prime Minister will make; and give that advice while there is still time to matter, because political capital, once spent, does not come back.
The audience with the Sovereign, the speech in Downing Street, the applause from the No 10 staff – none of that tells a new Prime Minister anything about how to govern. What happens in the hours after is the whole job. And it starts with the org chart.
Baron Sedwill of Sherborne is a cross-bench peer. He was Cabinet Secretary and Head of the Civil Service 2018–2020.
