Speech by BIS Permanent Secretary Martin Donnelly

Speech by BIS Permanent Secretary Martin Donnelly at the Institute for Government, Monday 30 June 2014: Positive Neutrality and Trust – the policy role of a permanent civil service.

"It may safely be asserted that, as matters now stand, the Government of the country could not be carried on without the aid of an efficient body of permanent officers, occupying a position duly subordinate to that of the Ministers who are directly responsible to the Crown and to Parliament, yet possessing sufficient independence, character, ability, and experience to be able to advise, assist, and, to some extent, influence, those who are from time to time set over them."

These resounding words date from November 1854, when the Northcote Trevelyan Report on civil service reform was presented to Parliament. They remain true today. I want to examine the justification for the provision of policy advice and delivery to Ministers by permanent civil servants. That requires some prior definition of the issue.

I then wish to argue that permanence and what I call positive neutrality are both more efficient, and more capable of building and maintaining trust in the process of government than any likely alternative.

It is important to stress that a non-political policy civil service is of course only a small part of the central UK civil service; and a tiny part of the wider public sector. There has been a welcome focus in recent years on the importance of efficient and responsive public service delivery, from health to tax to welfare. These are all vital and challenging tasks requiring specific expertise. I however wish to focus on the central policy civil service function which is often known for shorthand as Whitehall, though that today is increasingly a virtual rather than a physical place, which works directly for and with Ministers, and is concerned with the development and implementation of government policy.

Within this system civil servants recruited by competitive examination and promoted within a system designed to reward merit and be independent of political or other external influence- principles set out by Northcote and Trevelyan 160 years ago - directly advise Ministers. They expect to retain their roles, including at the most senior levels, when Ministers and indeed Governments change. They do not take part in political debate, although they clearly influence its outcome by their advice and effectiveness. For policy civil servants the notion of working for Her Majesty's Government, in support of the public interest as defined by Ministers responsible to an elected Parliament, is a real one.

Northcote and Trevelyan changed the political climate which had allowed rampant patronage across the public service. Their work led to the unified civil service structure we can still recognise today. Much has changed since the 1850s when Dickens was in full flight (and gave us the Circumlocution Office in Little Dorrit as a reminder that critiques of bureaucratic self-serving have a long pedigree) and most British adults did not have the right to vote.

This system has with some adaptations served Britain continuously over a turbulent 150 years, through wars that saw the civil service increase massively then reduce, the introduction of full democracy, the welfare state, the end of Empire and European Union membership, the rise of a knowledge based, service economy and changes in the structure of the United Kingdom from Irish independence through devolution to today's constitutional debates.

But it is right to look at whether this model remains the best option for policy governance in the second decade of the twenty first century. It is certainly not sufficient in itself. Advice can be independent but unhelpful; implementation can be well-meaning but of limited competence; decisions can be technically correct but fail to convince the public.  And the experience of those in the system can be too narrow or unimaginative to ensure an effective joined up response to new social or economic challenges. This requires the permanent civil service to know how to engage with external expertise in policy formulation and execution; and also to encourage technical experts to join, for short periods or a substantial career, flexible structures which make full use of their skillset.


Nonetheless independence does offer a promising starting point. When combined with a career structure able to reward professional commitment it limits the attractions of telling Ministers what they might like to hear rather than offering a more objective assessment of options, their likely success and affordability. It allows experience – collective as well as personal – of effective policy-making to be built up and drawn upon. It also provides a structure allowing frank discussion of options and priorities within a safe policy space across government and beyond.

A precondition of impartial advice and critical assessment of delivery options is that there is sufficient trust among Ministers to take official advice seriously. And the wider public have to trust that the process of government is being carried out to their benefit.

Trust cannot be assumed and has to be earned, both internally and externally.  Baroness Onora O'Neill, whose seminal 2002 Reith Lectures and subsequent work remain critical to any understanding of how to earn and maintain trust in a modern society, has commented on the extent of challenge civil servants and all those in positions of authority face in building and maintaining trust.

For Whitehall officials the challenge begins at home, within government. It is not a given that an incoming Minister, perhaps of a different political party, probably of a different temperament and outlook from her or his predecessor, should immediately trust official advice.  A key challenge for policy officials is to achieve and maintain that trust.  To add value for Ministers across government requires the foundation of a relationship of mutual trust, built on professional respect, and evidence of competence, and able to handle the pressure of events.  

The risk to good policy advice can stem not only from some private agenda but from the temptation to become too close to the wishes of a particular Minister. Civil servants must be politically aware and politically sensitive. Working in proximity to Ministers and to Parliamentarians leaves most officials with considerable admiration for the work politicians do in and out of government, for limited reward and often much unfair criticism. Their task has become harder as the media have become more pervasive, more immediate and, with less time or space to explain complex arguments.  Parliamentary speeches and debate no longer receive the public attention they once did.

Against this background it is important for officials to separate strong professional support for Ministers from crossing the line to become uncritical personal commitment, which ultimately is unhealthy for both politicians and officials. It can also make cross-government coordination more difficult.

Officials have to separate honestly held personal views from policy advice. Ministers appreciate clear advice and recommendations but they have a right to know that it is based on their own political judgements, not those of an official. Ultimately if an official is uncomfortable with the fundamental direction of policy in the area she or he works in they may need to move, to ensure that the system can deliver the quality of service required by Ministers.

Building Trust between Officials and Ministers

To work effectively and achieve and maintain that mutual trust, we need to be explicit about the differences between a politician and a policy civil servant and clear on the appropriate behaviour of the latter. I identify three broad themes to guide official behaviour.

Firstly not to do for one Minister what would not be done for another of a different party or outlook in the same situation. This Kantian approach allows whole-hearted commitment but stops short of indulging personal preferences or crossing the line into political polemic. Providing a convincing defence of government policy should be a core Whitehall skill; rubbishing the Opposition is not the function of permanent officials.

Secondly to avoid having to answer the question 'why wasn't I told that this might happen?' by erring on the side of disclosure to Ministers of everything that might take place. This must include challenging optimism bias, without allowing that challenge to become an excuse for inaction. We know for example that the timescale of policy delivery is a key component of a successful policy and needs as much critical scrutiny as the policy itself. It also includes ensuring a thorough evidence base for policy options, and being transparent about the inevitable limitations of the evidence when time or other constraints require rapid decision-making.

Third to offer some advice that is not accepted. If Ministers are never challenged they are unlikely to be getting the best advice around an issue. If the boundaries of their views are not subject to scrutiny they may miss the opportunity to change their views as the evidence changes. The desire to please ministers is rightly a strong motivator of policy civil servants; but it needs to be tempered by challenge – which in my experience good Ministers expect and have a right to. Ultimately Ministers have the last word, so mutual trust requires the ability to offer conflicting views in the confidence that it will not erode that trust.

This also applies within official structures. It is not always easy for policy civil servants working within a hierarchy to offer different and even opposing views to Ministers, where the same facts lead to differing judgements. But it is part of our professionalism to ensure that rigorous analysis of evidence and judgements about likely outcomes do not level down diversity of opinion. Culturally Ministers can be more at home with internal disagreement than civil servants; ensuring a genuine fair hearing for different options is a key role for the senior civil service in their quality control of advice to Ministers.

Clearly any policy challenge must be within the framework of the government's overall political direction, and therefore provided with political sensitivity. And after decisions are made they need to be implemented with professionalism,  rigour and as much pace as is consistent with effective delivery – the gap between the press release and the policy impact on the ground is inevitably longer than would be politically ideal. A policy civil servant may need to draw on the achieved trust levels to argue for slower safer delivery on occasion, despite its frustrations. Equally the challenge to deliver more, faster, with less resource must also be accepted and worked on professionally.

Officials and Ministers - differences between the two

What then are the differences between officials and Ministers? Fundamentally  Ministers have democratic legitimacy in their own right; and officials do not. They can expect their work to be taken seriously, based on their recruitment on merit and subsequent performance, but Ministers must have the last word. Officials are there to support them, within the law and respecting financial proprieties laid down by Parliament. Those who are not comfortable with this reality do not have a place as permanent policy civil servants. Nor should wider society expect officials to be a brake on the delivery of contentious political decisions supported by the democratic process.

Second, politicians generally have a different, shorter time horizon. So there is a tension between the optimal time needed to prepare or revise a policy and the political need to announce and implement it, which has to be recognised. To caricature, politicians want what Churchill called action this day; officials welcome more time to weigh the evidence and assess delivery options. Both are valid starting points. The value added comes in how they are brought together. Mutual trust between Ministers and officials, accepting the legitimacy of these different approaches, makes it easier to find the best available outcome; recognising the political constraints that exist as well as the dangers that go with too rapid decision-making.

To risk a further caricature, officials are trained to see complexity and understand ambiguity – policy delivery is difficult, with costs, downsides and uncertainties, and often with wider consequences elsewhere in government.  Without political leadership and explicit willingness to take risks not many decisions could be taken. The opposite caricature is a desire to take immediate decisions and announce them without having thought through all the consequences, resource constraints and the scale of risks, particularly in medium term projects. Or to assume that these stubborn facts can be overcome through greater zeal and Leninist commitment by those within the system.

In fact both ministers and officials share a strong interest, and desire, to get things right first time, to communicate policy clearly, to explain the constraints of real world delivery and to get stakeholders on board. Policies often need improvement as they are being implemented; external circumstances change with implications even for successful policies; resources can become more constrained so that further hard choices have to be made; and things go wrong unexpectedly. Policy success needs to be judged against the evidence available at the time decisions are made, and that in turn requires transparency and mutual support between officials and Ministers.

A safe policy space

If this trust exists the next step can be taken – to justify a safe policy space for discussion within ministries and across government. Effective policy making must be evidence based. That requires engagement with experts and even more importantly with those likely to be affected, directly or indirectly. Most policy has financial implications and involves tradeoffs with other desirable outcomes, meaning that there are likely to be losers as well as winners. Policy involves value-driven decisions, including on willingness to allow adults to run some risks, on levels of redistribution of income through the tax or welfare systems, on legal protections or definition of crime – issues on which views differ and where debate can be helpful.

There also has to be acceptance that no policy exists in isolation - other governmental goals are legitimate and must be taken into account. In Isaiah Berlin's terminology we have to be foxes rather than hedgehogs, not concentrating on one Big Thing to the exclusion of all else.  Different Ministries inevitably have different priorities for resource use and policy delivery. That is one reason they exist.  Government collectively must produce a sustainable range of policies which balance competing claims for limited money, legislative time or delivery expertise.  So policy officials must ensure that discussions or even arguments about policy tradeoffs take place across government on the basis of shared evidence and analysis, within a structure that allows for private discussion to determine priorities, which are then accepted and implemented across all departments.

Northcote and Trevelyan were perceptive in identifying that this type of conflict resolution requires a united cadre of officials with mutual trust and a shared organisational structure. One key criterion for assessment of policy officials is how they manage policy divergences between departments, work with the coordinating role of the Cabinet Office, and ultimately support Cabinet decisions engaging collective Ministerial responsibility. A shared Whitehall culture of transparent, unified decision-making is assisted by secondment of officials between departments and to and from the centre.  A habit of shared working and acceptance of its longer-term benefits is an important benefit of a relatively stable senior team of officials, who also provide the framework for working effectively with those who bring complementary external experience to the policy-making process.

Pre-emptive publicity about potential decisions can be used to bias those decisions in favour or one or another group, can make evidence gathering more difficult in a highly politicised environment, or weaken the political position of a Minister or part of Government seeking to take a difficult but necessary decision as one element of a wider programme. So transparency within the system should not be equated with public openness of the decision-making process. A private space for political debate within a government is a necessary part of executive cohesion which in turn ensures that government decisions are coherent across departments and over time, and which produces better outcomes.

Clearly the trust officials need to deliver their value during internal policy or resource debate must be matched by complete discretion about how these processes reach a political conclusion. That discretion should in my view extend to subsequent memoirs or public commentary and go beyond strict legal requirements. It applies both to domestic and to diplomatic officials. It is an investment in maintaining that relationship of trust for future generations of officials and Ministers.

Avoiding a monopoly on advice

The relationship between permanent officials and Ministers cannot and should not be a monopoly. Political advisers have a key role which cannot be played by civil servants. Officials are not chosen for their ability to give political advice.  Unsurprisingly the evidence is that they are not as good at it as politicians.  Political advisers add value through additional challenge to official advice, including suggesting new approaches, and their own advice on the politics and presentation of a decision. In my experience they do this well, because Ministers know the difference between good and bad political input. And they function as part of the wider unified civil service structure.

The crucial specificity of the British system is that political or personal advisers are not a separate layer of administration. In France or other countries with a cabinet system advice will go to the member of the Minister's cabinet for approval and, if necessary change, before being sent to the Minister. Cabinet officials often have considerable decision-making powers in their own right and many issues never make it to the Minister. One implication of this system is that it requires many fewer junior Ministers.  Another is that inter-Ministerial coordination tends to function on two separate levels – political and official - leading to a higher risk of policy incoherence and conflict over resources.

The Whitehall model ensures that official advice is seen directly by the Minister. Additional comments can be provided by special advisers and by the Minister's private office, but they do not change the advice itself. It is this direct access, together with career progression which does not depend on Ministerial patronage, which allows honest and occasionally unwelcome advice to be provided. It is up to officials to make use of this privilege wisely, and to politicians to support it as the best available way to maximise their own real decision-making power.

A further consequence of this model is that policy officials have to work wholeheartedly for the government of the day. They should not be looked to as a brake on politically divisive policy, nor expected to brief those outside government about policy disagreements or alternatives. If voters do not like policy outcomes the answer lies in the ballot box, not in expecting permanent officials to take on a political role, however unpopular a government may become at some points in its mandate.

This approach has implications for Opposition politicians, and again raises the issue of mutual respect between the different political and official cultures. Our system requires politicians not in government to distinguish between the policies they may abhor and the officials responsible for advising the Government on how to develop and deliver them. It also requires officials shortly before general elections to prepare actively to support and implement policies which may be diametrically opposite to the ones they have worked on and defended in recent years.

In my experience policy officials achieve this partly by developing a professional culture of political detachment, and occasionally ironic humour, recognising that elections do not change external reality, but also by enjoying the intellectual challenge of taking a new approach to issues. Personal familiarity between senior officials and politicians across the Westminster spectrum also helps to increase understanding of complementary roles.

Achieving full mutual trust between officials and Ministers must always be work in progress. There is a case for more joint training and team building between the political and official leadership of departments, and perhaps between Permanent Secretaries and the Cabinet as a whole. 

Building trust with wider society

But assuming that trust is substantially achieved, it is still only part of the solution. How do officials convince a sceptical wider world, as well as commentators, business, the third sector, and individual citizens that Whitehall really is good for them, or at least better than any plausible alternative model?

Onora O'Neill has identified honesty, competence, and reliability as key requirements for trust. How do these apply to the relationship between a permanent policy civil service and the wider public? And how do we best express them?

Honesty is probably the most straightforward. One hundred and fifty years of public service tradition allied to a rigorous system of audit and control through the National Audit Office and Public Accounts Committee together with the personal responsibility of the Accounting Officer, traditionally also the Permanent Secretary, for funds voted by Parliament give the United Kingdom a system which is the envy of the world in the honest management of public money.

Criticism has focussed more on issues of competence and reliability. Here there is no room for complacency. Continued civil service reform remains crucial. As turnover of companies in the FTSE 100 shows, it is challenging to have to renew constantly the skills needed to engage effectively with a rapidly changing economy and society. There is also a much greater level of transparency and media engagement now with civil service processes and decisions.  O'Neill's Reith Lectures make the argument that transparency has done little to build or restore public trust, and notes that ' the press are skilled at making material accessible but erratic about making it assessable'.

From a policy civil service perspective we need to be more assertive in helping citizens to assess available information about complex issues and programmes that are under scrutiny. Managing risk means accepting that things will go wrong, or certainly less than fully right, being quick to acknowledge problems and any failings that have contributed to them, setting out the remedies being put in place and learning lessons for the future. It also means celebrating success and communicating clearly what can be achieved with available resources, to avoid unrealistic expectations leading to disappointment. And it requires more proactive public communication setting out why officials are capable of doing their jobs to a high standard of professionalism.

It is increasingly clear that much trust-related dialogue takes place across social media. Digital literacy is becoming a core skill for policy officials seeking evidence about policy impact and also seeking to engage with sometimes sceptical citizens directly. Ensuring that there is a recognised area for officials to engage in this way while respecting both political boundaries and their own private space is a key challenge for the coming years.

So competence requires continued improvement, learning from best practice across the private and third sectors, and reviewing successful policy-making internationally. It also requires effective communication of risk management; explanation suitable to modern media of how and why decisions were made on the evidence available; and willingness to admit, correct and learn from mistakes rapidly.

Trust should come from public evidence of competence which allows officials to operate without their ability to do so being a contentious issue on every occasion. When the necessary communication to allow outsiders to assess that officials are working effectively has taken place the wider challenge of communicating policy is for Ministers. It is after all their policy, and they are entitled to the credit for it.

A public profile in itself is not helpful for Whitehall officials, any more than for the large backstage crew of a successful play. Ministers know that policy making is a team effort, and it is important that officials do not come onstage except in specific and limited circumstances such as Parliamentary hearings. Otherwise the critical distinction between political and administrative roles risks becoming blurred, to the detriment of both. Trust is often best expressed implicitly through a lack of public concern or interest when everything is proceeding smoothly.

To conclude, for policy-neutral officials to add value in government needs a specific culture as much as a set of clear rules. It requires officials to deliver a strong professional commitment to Ministers within mutually understood boundaries which separate them from political activity; Ministers to accept that the occasional inconveniences of this system deliver them a better outcome for individual policies and overall government effectiveness over time than would come from a more fluid, less professional structure; both to work together to deliver the best outcomes available in the political space; and wider society to trust that it is getting a good deal.

I believe that this system is worth the effort to adapt and improve while keeping the basic distinction between Ministers and officials simply because it offers most potential for good policy-making and delivery. To maintain it we need to be assertive in explaining why this is so. The above reflections are my personal contribution to that debate.

Martin Donnelly
30 June 2014