Why mandarins matter - keynote speech by the Rt. Hon. Oliver Letwin MP

Monday, September 17, 2012 - 18:00

Oliver Letwin, Minister of State at the Cabinet Office, was invited to the Institute for Government on 17 September to speak about the role of the Civil Service.

Oliver Letwin, Minister of State at the Cabinet Office, speech transcript

A response from Bernard Jenkin MP

Mr Letwin said:

I want to start by telling you a few stories.

They come from a previous life in which – instead of participating in the government of my own country – I was wandering around the world, advising other governments on the restructuring and privatisation of their nationalised industries.

In country A, I looked around for senior civil servants when attending meetings with the minister in charge of the economy. There weren’t any. No one at all, apart from secretaries (in the sense of typists and receptionists). I didn’t find it easy to get anything done there.

In country B, I spent many happy weeks in detailed discussion with a highly intelligent, courteous and urbane group of officials. I had the sense that we were making immense progress. It was only at this point, however, that I discovered that none of these highly intelligent, courteous and urbane people had the slightest connection with the making of decisions. That was done elsewhere – by the monarch and his courtiers.

In country C, by contrast, I found that I was discussing matters with exactly the right group of people. One of the issues which arose was that a certain law needed to be changed. I somewhat tremulously enquired how long it would take to change this law. '24 hours' was the response. It appeared that the processes to which we are used in our democracy were somewhat curtailed in that less than democratic environment.

In country D, I became progressively more unable to understand why what was happening was happening – until I managed to grab the proverbial drink with a senior official who spoke English. He explained to me patiently the personal financial agenda (the very personal, very financial agenda) of each of the principal officials with whom I had been dealing. All then became, and remained, very much clearer.

I can’t end these stories without telling you about country E, in which the government for which I was working found itself displaced – with the aid of a few tanks and other armaments – by the previous occupants: a rather more exciting way of changing administration than we are used to in Britain.

Why am I regaling you with these snippets from my long-past global wanderings?

I am doing so in order to illustrate the fact that our system of government – despite all its current difficulties – is not by any means to be taken for granted.

I want today to set out the way in which I think the administrative civil service ought to function in our liberal democracy. And I want to give you what I regard as two pieces of good news. The first is that I believe the administrative civil service today, at its best, does function in this way. And the second piece of good news is that I believe the reforms being carried through by Francis Maude and Bob Kerslake will hugely improve the chances of it operating in the right way where it is not currently doing so.

But, before I proceed with my main argument, I want to clear a little ground by describing more precisely what I mean when I refer to 'Whitehall administrators'.

To give an idea of scale, I am talking about fewer than 20,000 people at any given time – under 5% of today’s civil service.

What are these people meant to be?

Let me begin to answer that question by explaining what they are not meant to be:

1. although it is certainly useful for them to have had operational experience, when they are acting as administrators, they aren’t meant to be operators or managers (except to a slight extent of one another); they aren’t meant to deliver services;

2. they aren’t meant to project the power of the state – they aren’t the armed forces, or any other form of force;

3. though they need to work closely with economists and accountants and scientists and statisticians, they aren't themselves meant to be experts in any particular technical discipline;

4. though their work is bound up with the making and enforcement of law, they are not meant to be lawyers or judges or police officers;

5. though nearly all that they do depends on finance and has immense effects on business, they are not meant to be financiers or businessmen.

So much for what they aren't meant to be.

What are they meant to be and do?

They are administrators. What they are meant to do is to administer.

And what is this strange activity?

This is a question that I have been gently contemplating for the past thirty years, ever since I made my way from the calm abstractions of philosophical donnery into number 10 Downing Street and found myself surrounded for the first time by the finest exemplars of the administrative civil service.

At that time, I observed Robert Armstrong's Rolls-Royce minutes quietly gliding across from the cabinet office; I listened to Robin Butler and Charles Powell manage the affairs of a great Prime Minister; I witnessed the calm efficiency with which Michael Scholar, Andrew Turnbull, David Norgrove and others despatched the business of government from the private office. These fine officials made me understand a great deal about what an administrative civil service ought to be. Then, as now, the virtues these officials displayed were not universal – but they provided a model of the thing at its best.

At the same time, as I passed under review the broad sweep of political theory from Plato to Rawls, I found to my surprise that, with the honourable exception of Weber, no major theorist has made any appreciable effort to recognise the significance of administration as part of government. Ministers, Parliaments, courts, laws, class relationships, systems, interests - all of these figure of course. But, in the theory of politics, administration is all but absent. It is simply taken for granted that, once arranged in a certain way, the state will conduct its affairs. As if - if only - this were so obvious!

The truth is - as my examples from other countries are intended to illustrate - administration is anything but obvious. Wollheim once said of artistic style that it is an achievement of an artist to have a style - any style. And in the same way, it is an achievement of a state to have an administration - any administration.

The historians have a much firmer grip on this than the political theorists. They chart the development of administration -- in Rome, in China, in Byzantium, under the Angevins, under Napoleon and in the modern state. They recognise that it is an achievement to construct and maintain an administration. But historians are historians. They write history. They do not, on the whole, deal in abstractions -- so they inevitably leave us with the question unanswered: what is it exactly that administrators in a modern liberal democracy do?

My answer, after a rather prolonged period of contemplation, is that the administrative civil service in a modern liberal democracy characteristically needs to engage in four distinct but related activities:

• accumulation of knowledge;

• transmission of decisions;

• advice; and

• guardianship.

The trick that needs to be pulled off is to engage simultaneously and successfully in each of these four types of activity. Where the administrative civil service succeeds in pulling off that trick – which, at its best, it does – it brings something of inestimable value to Britain. And, by the same token, it is hugely in Britain’s interest that we should have a civil service reform of the kind now being promoted by Francis and Bob – to ensure that, so far as possible, these four critical activities are carried out successfully and simultaneously in all parts of the civil service.

Accumulation is the aspect of administration that most interested Weber. As he pointed out – and as is, once pointed out, obvious – the administrative civil service in the modern state needs to perform the role of ensuring that someone knows the answer to the question 'how does the system work?'.

We too often forget that the functioning of a modern, liberal state depends not only on law and law courts but also on the maintenance of settled process. When Maitland remarked that liberty resides in the interstices of the law, he was highlighting the significance of due process – which is the only safeguard of fairness and stable expectations, whether in court or in dealings with power. In the absence of due process, every trial becomes a trial out of Kafka and every dealing with government becomes something out of Darkness at Noon.

But however great the temptation to forget the fundamental importance of due process, commentators are even more inclined to forget that the maintenance of due process depends on having administrators to keep track of what the processes are. And this accumulation of knowledge of process is the first task of the administrative civil service in a modern liberal state.

Of course, there are ever-present dangers. A fixation with process can become absurdly bureaucratic. Process can become a substitute for achieving effects. And, at worst, administrators can hide behind process as a reason for not even attempting to achieve the effect desired by Ministers. I have no doubt that we currently suffer in the UK from too much process – some of it, alas, introduced by the previous government. One of the purposes of the present coalition government, both in our Red Tape Challenge and in our civil service reform is to prune back process where it has become too inhibiting and too nearly an end in itself. I am delighted to say that there is no more enthusiastic a proponent of such pruning than the current Cabinet Secretary.

But, as we seek to restore proportionality where it is lacking and to reduce the burdens of process where they have become excessive, we should remind ourselves that we do require an administrative civil service that understands, respects and operates due process where it is needed and to the extent it is needed.

The second aspect of administration in the modern British state – transmission – is less obvious but no less important.

Read a textbook account and you will find the following, charming but illusory description of our constitution: a liberal democracy like ours elects a legislature and an executive; the legislature makes law, which the systems of civil and criminal justice enforce; the ministers who constitute the executive make decisions which are either in the form of laws proposed to the legislature or in the form of actions sanctioned by law.

The reason why this charming description of our constitution is illusory is that it entirely ignores the vital question, 'how are the decisions of ministers transmitted?'.

A minister sitting in an office (even supposing that a minister acting solo were able to provide himself or herself with an office) is in principle capable of making any number of decisions. But without someone to transmit those decisions, they would remain poetical aspirations rather than actions.

The activity of transmission is very little remarked. But it is both difficult and complex. It consists not only of recording, but also of translating, enlarging, clarifying, encoding, promulgating, authorising and, often enough, paying and accounting.

Of course, just as the administrators' understanding of due process can become a disproportionate process-fetish, so the administrators' ability to follow process in transmitting ministerial decisions can become labyrinthine. Simple objectives can be turned into items of such great complexity that the original aim is either ludicrously delayed or altogether lost in the morass of refinements. Ministers need constantly to be on their guard against these tendencies – and experiments such as contestable policy formation, as with the National Planning Policy Framework, are well worth using as a corrective. Fresh eyes belonging to practitioners from outside the administrative civil service can sometimes see clearly through bizarre and unnecessary tangles that those engaged in administering processes have come to regard as normal and inevitable.

There is also the danger of sheer inaction. Either through torpor or through positive reluctance, administrative civil servants can, at their worst, defeat Ministerial objectives, just by ensuring that when the Minister has decided to act nothing actually happens. Such failures of transmission are enemies of democracy – and one of the things that our civil service reform programme is designed to do is to eliminate such failures.

But the fact that transmission can become over-elaborate and under-effective should not blind us to the fact that our administrative civil service, at its best, is fully capable of translating ministerial decisions into action. We need improvement through reform. But we have real strength on which we can build.

Such transmission of decisions is, however, a very different thing from achieving the effects and outcomes that ministers desire. And this is where we come to the third characteristic activity of the administrative civil service – the provision of advice. It is extraordinarily important to distinguish between what such advice should be and what it should not be.

To begin with what it should not be, the civil service is not called upon to formulate political programmes or to determine national objectives. A state in which the civil service did so, would be something other than a democracy – since democracy consists in the ability of the electorate to make a choice between programmes and objectives put forward by competing political parties, and to hold elected politicians to account for their performance; pace the calls for an apolitical national strategy from some, including the present select committee on administration, any attempt by the administrative civil service to formulate such a strategy independent of the political programme of the elected government would be a subversion of democracy.

But a programme or an objective is a different thing from a fully specified policy or a fully specified decision. And it is, I think, precisely into this gap between programme and policy, or between objective and decision that the activity of civil service advice properly fits.

The administrative civil servant is called upon to perform the extraordinarily difficult task of discerning the nature of the programme or objective sufficiently clearly – and of gauging the effects of both government action and citizen reaction sufficiently certainly – to be able to advise the minister accurately on which specific policy or decision will be most likely to achieve the objective. Often enough, this will involve a creative act – identifying subordinate, specific objectives that flow from higher level, general objectives, or identifying issues with which a Minister (given their higher level objectives) would want to concern themselves if they were aware of it.

We are dealing, here, with something that demands an intellect which is both imaginative and subtle – because how you do something may have an effect not only on the result but also on the political character of the action, and it is therefore extraordinarily difficult to know where objectives end and implementation begins. What may appear from a crude perspective to be merely instrumental and accidental may well in fact be essential.

Lest this should seem to be a series of gnomic utterances, let me give a concrete – though, I believe, fictional – example.

Suppose that a minister has stated, in his or her manifesto, the objective of improving community and village halls. There are, of course, numerous ways in which this objective can be met. The government could dole out taxpayers' money and specify precisely the manner in which it is to be spent on community and neighbourhood halls. Or the government could encourage locals to raise money, do the designs and contribute labour.

Manifestly, each of these policies may – if defined in sufficient detail and carried through properly – fulfil the narrow objective. But the choice between them is not neutral or merely practical. As will be rapidly evident, one of the two proposed means of implementation is much more in tune with the wider objectives of the present coalition government; and there could well be some other government with whose wider objectives the other means of achieving the narrow and specific objective would be more in tune.

So the administrative civil servant, in giving advice on how to achieve ministerial objective A is bound also constantly to bear in mind ministerial objectives B to Z. And this is not a straightforward matter. It requires an understanding of the relative priority attached to differing objectives by differing ministers, as well as an understanding of how to balance short term effects against long term effects. And, if the advice is to be useful, it needs also to be based on a clear line of sight from decision to action: the civil servant needs to be able to envisage how the policy on which he or she is advising can be implemented.

But there are even more exacting requirements that we place on our Ministers and on our administrative civil servants.

So far as Ministers are concerned there is of course – as is well rehearsed – a dual requirement. They need to seek and then listen to advice from their officials. But they need also to be sufficiently self-confident to exercise their own judgement and reject the advice if they are not persuaded by it.

And the administrative civil service – when providing policy advice – is required to perform a corresponding double act. We need civil servants who will give well informed, fearless advice – speaking truth to power. But, if the Minister rejects the advice, then we need those very same civil servants to transmit and implement the Ministerial decision with which they disagree, as energetically and effectively as they would have transmitted and implemented the decision that they recommended. This is by no means an easy task for a human being to perform. Perhaps unsurprisingly, it is not always performed as well as it should be – and, it is an aim of our civil service reform programme to ensure that this becomes (despite its inherent difficulty) absolutely and everywhere the norm. But, once again, we have models on which to build because this is precisely the double act that the best of our administrative civil servants already perform as a matter of course.

This brings me to the fourth characteristic activity of the administrative civil service – guardianship.

Of all the roles of the administrative civil service, this is the one that is most problematic – and the one that our constitutional arrangements make it especially important for our administrators to play.

The other three roles, accumulation, transmission and advice, work naturally together. It is only if our administrative civil servants have a great accumulated knowledge of the due process of government that they can be expected to transmit ministerial decisions effectively; and it is only if they have a full understanding of process and of the transmission of decisions that they will be able to advise usefully on the development and implementation of democratically determined policy programmes. But, in all three of these roles, our administrative civil servants are called upon to be servants of whichever set of ministers our democracy has placed in government. Whereas, in their capacity as guardians, our administrative civil servants are called upon to play an altogether different role – as servants, not of ministers but of the crown, accountable to Parliament.

I say 'the crown' because, in our sadly unwritten constitution, the crown is the metaphor for the persisting state which rises (in the person of Her Majesty) above the process of party politics and above whatever is at any given time the present ministerial incumbency.

In their role as guardians, administrative civil servants act on behalf of the crown to ensure that the government as a whole acts with propriety and in conformity with the law.

Why, one might ask, is this necessary when there are courts to ensure conformity with the law?

It is of course true that, over the past half century or so, the judges have developed administrative law to a degree that was unimagined a century ago. And our governments today are governed by that judge-made UK law – as well as by European law, human rights law and international law – to a degree that would equally have been unimagined in 1912.

Nevertheless, it remains true that UK governments (like any government, but even more than governments subject to the clear rules of a written constitution) have wide discretion about how to act -- and it is one of the unspoken roles of the civil service to ensure that this wide area of discretion is not abused.

The importance of this civil service role can hardly be over-stated. It is one of the great bulwarks against tyranny. The administrative civil service provides a continuing safeguard that ministers of any persuasion will not be able to use the machinery of the state to personal or party political advantage.

What makes the role particularly difficult to perform is that it needs to be performed in a way that does not turn the civil servants into being the civil masters. There is a constant danger that the administrative civil servant will use his or her position as guardian of propriety to seek to prevent Ministers from doing things that it is in fact permissible and proper (but, not from the civil servant’s point of view, convenient) that they should do. And, on the other side, there is the danger that – fearing this reversal of roles – civil servants will not act as guardians sufficiently to prevent genuine impropriety. It is no easy matter for administrative civil servants to steer between the Scylla of unjustified constraint and the Charybdis of insufficient constraint. Once again, it is an aim of our civil service reform programme to ensure that – so far as possible – we do constantly steer between the Scylla and this Charybdis. And, once again, we can build upon the fact that the best of our civil servants achieve such skilful piloting every day.

So much for my analysis of the roles of the administrative civil service and for my qualified but enthusiastic endorsement of the quality of our administrative civil servants in carrying out those roles.

I want just to end with a plea.

It is addressed to all those leading the service and to all those commentators who have an influence over the service.

My plea is this: let us not make the crude mistake of attempting to liken the administrative civil service – the fewer than 20,000 people involved in accumulation, transmission advice and guardianship – with any other entity in the land.

Most, indeed almost all of the activity of the modern state is in some sense business-like. Of the 400,000+ people in the wider civil service and the millions employed in the wider public services, the overwhelming majority are engaged in performing tasks with clear objects, arranged in units, led and managed by leaders and managers in the way that any private or social enterprise has to be led and managed.

As I have said, it is immensely useful, for administrators to have had some real experience of this operational activity. But the work of the administrative civil service is not the same sort of thing as operational activity. It is intrinsically governmental. It exists to promote and enable what Michael Oakeshott called a civil association, not an enterprise association. It is a profession in its own right, no less demanding and no less valuable than other professions. Accordingly, while we can and must ensure that the business-like aspects of its activities -- its accounting and control of money, its use of physical assets, its procurement techniques and the like – are performed in a business-like way, (as Francis Maude and Danny Alexander have been ensuring through the Efficiency and Reform Group and the Major Projects Authority), we must never allow ourselves to be gulled by the crude falsehood that all would be for the best in the best of all possible worlds if only all administrative civil servants were to be trained in some other profession or were to spend more time reading books written by management consultants.

The special tasks required of the administrative civil service to enable Ministers to operate a liberal democracy under the rule of law are very special indeed.

The skill involved in understanding the processes of government, in transmitting ministerial decisions effectively, in advising ministers wisely and faithfully on the translation of political programmes into actions, and in acting appropriately as guardians against the abuse of state power is very great.

The possession of these skills on the part of the best of our administrative civil servants is very precious; and the need for them to be possessed by those administrative civil servants who do not currently possess them is equally great. Through our civil service reform programme, building on the huge prowess of the best, we must ensure that each generation of administrative civil servants has the self-confidence to hand those special skills to its successors.