The Spiders of Whitehall

Is Labour prepared to launch an internal coup?

“[…]too many people in Whitehall are comfortable in the tepid bath of managed decline.”

Thus the Prime Minister announced his intention to reshape the fabric of the British state in order to deliver a veritable Christmas tree of targets, missions and commitments. He will be the one to pull the plug on the cooling waters of the Civil Service tub, plunging unwary spiders into the abyssal pipework below.

But how will he do this? On this, the Prime Minister was less clear. What is clear is that the absence of substantial reform to the machinery of the State will actively prevent Labour from delivering on its missions. Mission Control for Clean Power 2030, Labour’s most structured and progressed mission, has very little ability to direct the policy work of the Department. This is analogous to NASA building the Saturn V to get to the moon and having no recourse when Cape Canaveral decides to build a launch gantry for a sub-orbital rocket. This is not mission-led Government, but the creation of a new advisory body.

Labour is discovering, as every Government does, that Whitehall is set up to service the Government of the day. Or, more specifically, the Minister of the day. Officials know that Ministers change reasonably frequently, and that there’s nothing a Minister hates more than being told they can’t make a decision because their predecessor chose the opposite. This means that being able to best serve any given Government means maintaining as much flexibility in policymaking as possible and proceeding through incremental changes about which there is a consensus.

The example par excellence for this is the Treasury, with its so-called Brain. This has an institutional emphasis on resisting long-term commitments and maintaining flexibility as an objective for policy above everything else.

And in many respects – especially democratic ones – this is the right thing to do. Ministers, who are elected, should be able to tell officials to stop doing what they’re doing and do something else. The problem is that in order to enable this flexibility the entire Civil Service is structured as a loose federation of disparate teams each looking at any individual aspect of policy on which a Minister may wish to opine. This is less a Whitehall machine, and more a web. But not a lovely concentric web, more like one of the webs produced by spiders who have been given drugs.

Various Departmental organisational structures

This web hungers, and it hungers for one thing: Steer. This is the peculiar noun Whitehall uses to refer to Ministerial decision-making. Teams obviously can’t act until they know what a Minister wants them to do, and so the first few weeks of a Minister’s term in office is death by a thousand submissions (advice papers) on every intricate aspect of their policy programme.

Understandably, in the face of this blizzard of paper Ministers will prioritise what they decide on first, and teams can find themselves waiting for weeks for even a crumb of Steer. This prompts teams to seek Steer sparingly and to ensure that any advice that goes up has already thought of and answered any questions Ministers may have had. The result is the chaos of the caffeine web above, as officials seek views of stakeholders through consultation, run analysis through internal processes, and even occasionally ask other Departments what they think. As such displacement activity evolves it accrues the badge of Best Practice from the Cabinet Office and another few days is added to the standard policy cycle.

Whitehall is therefore ideally set up to dribble in decisions to Ministers in line with their bandwidth regardless of the scale and salience of that decision. The dozens, hundreds of teams in each Department each have a bilateral relationship with the Minister. Ministers may be asked to decide on matters ranging from the future of the nuclear deterrent to regulations governing soldiers’ boots on the same day.

Occasionally Ministers will despair of being bombarded with requests for minor decisions and ask to do a ‘strategy’. They take the view that if they can give a single strategic steer to the Department they can short-circuit the innumerable decisions they face and get things done faster. But the federal structure of the Civil Service militates almost accidentally against this. The expertise necessary to develop the components of this strategy sits with the teams at the edges of the web, not the centre. And so, a strategy team will send out a commission to all the individual policy teams for what they’d like to do, and prepare a document that lists all the individual items that each team wants to deliver. Occasionally the strategy team will seek to apply additional intellectual rigour and push back on policy teams to be more ambitious, and sometimes they will succeed. But the fundamental structure of the measures within a strategy remains the preserve of each individual team.

To ensure that they are on the right track, strategy teams will keep Ministers up to date with their list of measures, and ultimately seek Ministerial approval for each one. This is right and proper; Ministers bear responsibility for what they put out. But it means that decision-making – and importantly, radical change – is painfully slow regardless of what Ministers do.

Into this federal structure has been thrust Labour’s missions. It should be immediately obvious why Labour are frustrated at the pace of delivery of these, because by themselves they do nothing to meaningfully change how Whitehall organises itself. Perhaps there is a Mission team who sends a commission out to the rest of their Department – in rare cases even beyond – for measures that can contribute to the Mission. Each of those will require Ministerial sign-off and passage through the clutter of Whitehall procedures that have ossified around its policymaking.

As mentioned above, Mission Control for the Clean Power mission has been structured as an advisory body rather than a decision-making body, meaning that choices still rest with the same teams with the same bilateral relationships with Ministers. This is very much not NASA. But it needs to be.

Pulling the Plug

Part of the reason why expertise sits at the edge of the web is that the Civil Service actively militates against any given official accruing enough expertise. A member of the Senior Civil Service is explicitly expected to be able to be deployed into any role across Whitehall regardless of their understanding of the policy area. Whitehall only started nodding to the role of experience in staff hiring in 2018. Anyone involved in engaging Government will tell you of their experience in helping to skill up new officials who, after perhaps two years at the most, will head off to a new policy area where the cycle will start anew.

While improving the expertise and knowledge of officials is part of the solution, the bigger challenge is that Whitehall structures the problem of Government incorrectly.

Government in a democracy is about giving the people what they want. The incentive on politicians is to identify what people want and find ways of offering it to them in exchange for power. The job of civil servants should be to find ways of translating that offer into policy and delivery.

But the precise details of regulation only very rarely form part of that offer. What politicians provide instead is a combination of headline directions of travel and a targeted appeal to particular values that the public hold. Political campaigns are about assembling a coalition of voters based on shared interests and common values, not on secondary legislation¹.

It is worth noting that this applies to Parliament in a very important way. The debate over assisted dying was judged to be Parliament at its best – but this was a topic where values are paramount over regulatory design. This was a topic on which MPs genuinely could mediate the public’s values into law.

The problem of Government, then, is not about ensuring democratic sign-off of every detail of policy but rather about getting democratic sign-off of headline measures and value-based approaches to policy. The Civil Service should look to translate the views of the voter coalition that has won the election into holistic policy rather than rely on Ministers to essentially channel the public zeitgeist into detailed regulatory decisions.

This is a radically different role for the Civil Service, one that would see it make regulatory decisions rather than Ministers. It would require a very different structure to achieve this. Expertise would need to shift to the centre of Departments, where a limited number of officials with a deep understanding of their topic, policymaking and public attitudes could create radically alternative structures that enable voter values to be reflected more rapidly.

Reform has proposed an analogy of this that super-charges SpAds into roles that can no longer just be the preserve of twentysomething former Parliamentary researchers. However, their proposal doesn’t solve the fundamental problem of the Civil Service being the wrong shape. Bilateral relationships between individual teams and Ministers need to end, policy choices and expertise need to be centralised, and entirely new capabilities that provide an understanding of what the public wants need to be developed. Ministers should be able to provide a limited number of strategic steers and have confidence that they will be rapidly translated into policy.

In many ways, Mission Control could be the template for this, if Ministers want to force the pace of change. Ed Miliband has the option of essentially instructing the DESNZ Permanent Secretary that they are now subordinate to Chris Stark and all teams working on topics relevant to Clean Power now report to him. It is, however, unclear whether Ministers actually have the power to do this – it would be equivalent to an internal coup. But if Labour is serious about reforming Whitehall, they may have to consider demoting the spiders.

1. Someone will find an example of when this was in a manifesto and determined an election, to whom I will say: edge cases make bad policy.