I'm going to start with two quotes. One is from a leading civil servant; the other from a
politician. First:

Our system of a permanent politically impartial Civil Service is one of the jewels in Britain’s crown.


On one occasion he believed he had negotiated a compromise between civil servants from two departments
concerning a programme related to both, only to have one tell him that his Minister had rejected the
arrangement. “It never was the Minister,” he said, “but the bloody civil servant winding him up. It was
so annoying you lost the will to live at some points in this process.”

Which was which? The first was me shortly before becoming Minister for the Cabinet Office in
2010 with responsibility for the Civil Service. The second was Sir Bob, now Lord, Kerslake,
shortly after stepping down as Head of the Civil Service in 2014.

I start in this way to make clear that I had no predisposition to be critical of the Civil Service.
Based on my experience as a Minister in the eighties and early nineties my expectations were
high. And the disillusionment was steep and distressing.

I stress that I became disillusioned with the Service, not with civil servants. Any critique of this
type will inevitably be traduced as an assault on decent public servants who supposedly can’t
answer back. It is not an attack on civil servants. It remains my view that we have some of the
very best civil servants in the world, both in Whitehall and on the operational side. But the Civil
Service as an institution is deeply flawed, and in urgent need of radical reform. And it is civil
servants themselves, especially the younger ones, who are most frustrated by the Service and its
culture and practices.

A quick word about what this is not. I do not believe that the Civil Service has become
politicised nor do I wish it to become so. I never knew or wanted to know the political
preferences of civil servants. When I discerned a resistance to reducing the size and reach of
the state I was generally confident that this was for reasons of self-interest rather than ideology.
And by the same token nothing that I propose here is, nor anything I sought to change in office
was, about politicisation, although that was often alleged. The accusation of “politicisation” is a
terrific defence against any attempt by politicians to create a genuinely high-performing civil
service that will actually deliver what a democratically accountable government, of any stripe,
decides to do. Civil servants must be impartial in the sense of being equally able and willing to
serve a government of a different colour. But that is emphatically not the same as being
“neutral” or “independent”, as is sometimes claimed.



The Civil Service suffers from institutional complacency. As the new Minister responsible for
the Civil Service, every draft speech or article presented to me started: “The British Civil Service
is the best in the world.” Yet the complaints by Ministers in all parties about the lack of
institutional capability, inefficiency and failed implementation were legion. When we queried the
evidential basis for this assertion, it turned out that the only relevant assessment was a World
Bank ranking for “government effectiveness”, in which the UK ranked number 16. I remember
a conversation where it was proposed that rather than seek to improve our performance against
these criteria, we should create a different index which might better recognise those qualities in
which the British Civil Service was thought to excel. So I was amused to see this summer a new
index assembled by the Blavatnik School at Oxford, in association with the IFG, and (it coyly
states in the fine print) “supported” by the UK Civil Service. The criteria are clearly selected to
favour Westminster-type politically impartial systems – operational efficiency and effectiveness
barely get a look in – and Singapore, which has a truly impressive bureaucracy, doesn’t even
feature. Yet even in this index we come last among the Westminster-type systems, behind
Canada, New Zealand and Australia.

Is this just me? Oliver Letwin, no enemy of civil servants, indeed married to one, recently
bemoaned the quality of the work produced by policy civil servants outside the centre. He made
the point that the cream tends to float to the centre of government, to No10, HMT and the
Cabinet Office. This has often led PMs and Chancellors to discount their colleagues’ complaints
about the quality of the service. Because the civil servants who surround them are among the
best, deficiencies elsewhere are discounted as the fault of inadequate Ministers.

It is significant that the longer Prime Ministers remain in office, the more jaundiced their view
becomes. Tony Blair recently reflected on this:

If you had a crisis, there is nothing better than that British system...But when it came to how do you do
health service reform or education reform, or the early battles I had in reforming asylum and immigration
policy, I found it frankly just unresponsive.

Are government failures always the fault of the of civil servants? Of course not. We all know
there are myriad examples of failures caused by Ministers ignoring good advice. I’ll say more
later about the relationship between Ministers and civil servants, and perhaps as important the
relationship between policy and operational civil servants.

But I and others have observed that all too often the first reaction of the Civil Service when
something wrong is discovered is either to cover it up or to find a scapegoat, often someone who
is not a career civil servant and who is considered dispensable. There seems to be an absolute
determination to avoid any evidence that the permanent Civil Service is capable of failure.

Another indicator is that if a Minister decides that a Civil Service leader is not equipped for his
or her task, this has to be dressed up as “a breakdown in the relationship”, with the unspoken
suggestion that this is at least as much the fault of the Minister as of the civil servant. It can
never be admitted that the mandarin was inadequate in any way.

When I suggested that there might be room for improvement, the distinguished former Civil
Service Head, Lord Butler, accused me of a failure of leadership. Actually the leadership failure
is to pretend that all is well when no one, even civil servants themselves, really believes that.

The Gifted Amateur Cult

Fifty years after the Fulton report lambasted the cult of the gifted amateur, there is still some
way to go. While developing our Civil Service Reform Plan, a very senior permanent secretary
blithely stated to me that “the age of the generalist has passed”. I didn’t contest this, as it
seemed probably true. But it isn’t true nor should it be.

Most organisations have and value generalists. In the Civil Service the generalist is someone
skilled at marshalling and analysing evidence, mobilising expertise, and developing solutions. But
a generalist is not the same as a gifted amateur. To be effective they need a serious investment in
their skills. The doctrine “we don’t have generalists any more” has the malign effect that all civil
servants feel that they are meant to be specialists or experts, and therefore feel prohibited from
saying that they don’t know how to do something. Hence the tendency for officials to persist in
trying to execute tasks for which they have inadequate skills or expertise, and feeling unable to be
honest about it.

So what training should generalist administrative civil servants have? I was surprised to discover
how the Fast Stream graduate entry operated. Bright graduates thought they were joining the
Civil Service; and were then surprised to find that they joined a specific ministry where training
took a definite second-place to the job to which they were assigned. My modest reform to make
the Fast Stream programme look and feel more like a typical two year graduate training
programme met with surprising resistance, with four permanent secretaries, including at the
Treasury, showing up to tell me that it was completely impossible. Apparently, if the Civil Service
trained its graduate entry the way high-performing private sector entities do, the government
would fall apart. If I insisted, as I did, that Fast Stream trainees did four six months postings in
different parts of government, then they would be unable to do any useful work. The low
importance attached to hard-edged skills was brought home to me when a graduate trainee said
how surprised a colleague had been that she had opted to take a course in using spreadsheets.
The surprise for me was that it was not a basic requirement for every graduate trainee to be able
not only to use spreadsheets, but to do basic financial modelling. The Civil Service College in
Singapore runs courses in letter drafting, speed reading and touch typing – basic skills in the
digital age, and a big contributor to improved productivity. Repeated requests that the UK Civil
Service should require the same met with blank incomprehension and inaction.

The second eye-opener was when I proposed that senior civil servants headed for very big
responsibilities should be put through top management courses, typically three months, at top
business schools. High performing organisations routinely do this; and I have seen people come
out transformed into a bigger, more confident and capable leaders. So I proposed first that ten
permanent secretaries should go through these courses before the 2015 election. The first
objection was that this would be very expensive and that the Daily Mail would make a fuss. My
response was to say: Bring it on. If the Mail really want to object to us spending £60,000 to train
up someone managing a budget of tens of billions, I’d love to have the argument.

We eventually agreed to do this some 18 months before the election. I was told repeatedly that it
was underway. And yet by the election, instead of ten doing three months at Harvard, Stanford
and Insead, one had had done one week at IMD in Lausanne. Not quite what we had in mind. I
struggled to understand why there was so much resistance to a plan that seemed obviously

beneficial and sensible, but also showed a willingness to make a serious investment in our
people. I concluded that the unconscious explanation is an anxiety about senior civil servants
being put into an environment where they are sharing experiences and learning from their peers
in the private sector. This would be to allow too much daylight into the mystique with which the
world of the mandarin is surrounded. The cost of this protectiveness is clear; and I've always
been reluctant to blame heads of departments for operational failures when they’ve been given
woefully inadequate preparation for the huge responsibilities they are asked to assume.

I think the same explanation holds for the well-documented reluctance to welcome and make the
most of recruits from other sectors. The Baxendale Report into external hires showed how little
interest the Civil Service has had in learning from those who come in from other organisations;
that they became, as one of them put it, at best country members of the club. There was little
sense that the organisation adapted to the incomers or learned from them; all the adaptation was
expected from those coming in. Too often the incomers get spat out. Yes, sometimes this tissue
rejection is the fault of the external recruit. But too often it is the system at fault. As one
distinguished business leader who came into government said to one of his successors:

“You will find that mandarins operate behind a wall. Eventually you will find a door. But it only opens
from the inside.”

Mandarins come out on top

Part of maintaining the mandarin mystique is that they pretty much always get the top jobs.
Policy nearly always trumps operational and technical skills for the leadership roles. It feels like a
class divide: there are the white-collar policy mandarins, and the blue-collar technicians who do
operations, finance, procurement, IT and digital, project management, HR, and so on. All the
attempts to create genuine parity of esteem have failed. This has to change in the future. Many
government failures could have been prevented if operational and technical teams had the same
access to Ministers as do policy officials. But there is far too much tendency to keep them at
arms-length and below the salt, as the passive takers of policy ordained from on high which they
are expected unquestioningly to implement. On the notorious Universal Credit programme,
policy was developed in Whitehall; ‘implementation’ was in Sheffield; and IT development in
Warrington. Not surprising that hundreds of millions were written off in wasted costs, whereas
implementers brought into the policy development process early could have pushed back on
policy changes that inevitably complicated the project. It was interesting that the PAC Report
into UC stated that “the problems came to the attention of the Department as a result of a
review commissioned by the Secretary of State”. In other words it was the minister calling the
attention of officials to the implementation car crash, which is not the way round it is meant to


As in most bureaucracies, the culture is hostile to innovation. While pursuing Civil Service
reform, I was often asked: when will this end? The answer of course is never. Reform and
improvement is always a work in progress. There is no steady-state management any more in
progressive organisations. While the words encourage innovation the behaviour and incentives
do the reverse. Jonathan Powell, himself a civil servant before becoming Tony Blair’s chief of
staff, summed it up well:
The system is stacked against civil servants who might want to get things done. There is very little upside
gain for an official who succeeds in resolving a problem and a huge downside risk for permitting
something to go wrong.

The best organisations learn more from the things that are tried that don’t work than from those
that do. I introduced what I wasn’t quite brave enough to call the Francis Maude award for
failure: to recognise a team that tried something new that didn’t work, stopped doing it, and
ensured that the organisation learned from it. There were 80 nominations; 79 for projects that
had not failed, thus simultaneously missing and illustrating my point. The Singapore Civil
Service consider this willingness to innovate as so important that they elevate it to an obligation.
On the wall of the Civil Service College is this quotation:

What we want is for all officers to see it as part of their job to question the assumptions and past ways
of doing things and suggest ways to improve and innovate.

Surely nobody could argue with that proposition. Yet it is the inverse of the behaviours that are
rewarded and encouraged in our system. Government is today one of the few remaining
environments in which disruption is a dirty word. There is an bias to inertia. It needs to be
replaced by a bias to action. One aspect of this is that too often process trumps outcome, a
failing identified by many of the external hires interviewed for the Baxendale Report.

And this is part of a broader problem. Behaviours, especially among the most senior, are
frequently at odds with stated policies. We commissioned independent evidence-based reviews
into the treatment of women, BAME, LGBT and people with disabilities, so that the Civil
Service Diversity Plan would be hard-edged and practical rather than the tokenistic and generic
draft I saw first. They all concluded that while the Civil Service had model progressive policies,
the reality too often was profoundly different. In the Women in Whitehall report, one woman
said: “I was told that the reason I wasn’t being interviewed [for a promotion] was that I would
have outperformed the man whose turn it was”. It remains the case that too often promotions
are made on the basis of personal patronage and time-serving rather than on merit and talent.
And while on diversity, there is I believe less personality and intellectual diversity than I recall
from before. There seems a premium on blandness. High-performing organisations relish
having a decent quotient of the quirky, difficult and maverick, and know how to manage and use
them. One of the very best officials I worked with was told that the only question mark over his
suitability to be a permanent secretary was that he was “too colourful”.

Honesty and integrity are two of the Civil Service’s statutory values. Yet it is surprising how
often Ministers are told things that are simply not true. On two specific occasions I was told
that the cost of implementing a change, in each case to civil servants’ own employment terms
and conditions, was literally 100 times what turned out to be the actual cost. Quite often I would
be told that the law precluded a particular course. More often than not it was not true. Civil
servants should “speak truth unto power”. But it must be the truth. No sane Minister wants to
embark on a course without the best and most candid advice on all its implications. But it must
be advice based on true facts and the best evidence there is. Oliver Letwin summed this up:

On probing the causes of the unclear, jargon-ridden and ill-evidenced papers that too frequently came my
way, I often found that the problem was not just a stylistic inelegance but rather an inability to think
clearly about whether a proposition being put forward actually corresponded with the facts…sometimes –
indeed, distressingly often – officials had put together a ‘view’ or ‘recommendation’ without knowing the
essential facts.

And when, after receiving candid and well-evidenced advice, Ministers make a decision, it is the
duty of officials to execute it. Again, surprisingly often, this simply doesn’t happen. On one
occasion I asked a cross-departmental group of officials why a Cabinet Committee’s very clear
decision had simply been ignored. The answer? “We didn’t think it was a very strong mandate”.
What? What on earth do you need? A Papal Bull?

In 2004, Tony Blair as Prime Minister, no less, announced in a speech that in future all Senior
Civil Service appointments would be for a four year fixed term. It didn’t happen. It wasn’t until
2013, many years after Canada, Australia and New Zealand, that we finally secured a much lesser
change: that permanent secretaries would in future be on a five year fixed term.

I thought this was just part of the unwritten sub-culture until discovering what we came to know
as “the document of shame”. This set out the criteria for selecting potential permanent
secretaries, and was drawn up by consultants at the behest of the then leadership of the Civil
Service in I think 2009. It included such precious jewels as

• knowing when to “serve” the political agenda and manage Ministers’ expectations versus leading
their Department

• Balances Ministers’…immediate needs or priorities with the long-term aims of the Department,
being shrewd about what needs to be sacrificed

• Tolerates high levels of ambiguity and uncertainty …and at times irrational political demands.

I had always assumed that sentiments of this nature were simply passed from one leather
armchair in the Athenaeum to another. But that a willingness to ignore the decisions of
democratically accountable Ministers was regarded as such an essential quality in a permanent
secretary that it needed to be recorded in writing was astonishing.


So what is needed for the future?


First is a change in culture. This is both the most important and the most difficult. You change
an organisation’s culture not by trying to change the culture but by changing people’s behaviours.
The Civil Service of the future must have a culture with these qualities:

• Honesty about itself – especially about its own performance

• Truthfulness and accuracy in its advice to Ministers

• A genuine openness to learning and influence from outside, including external hires

• A genuine commitment to training and hard skills

There are proper experts in organisational change who will have views on what is needed to
achieve all this. All I would say is that it has to start at the top of the organisation. There is no
substitute for leadership by example.

Innovation and “Up or out”

I referred earlier to the innovation-hostile culture, and the bias to inertia. One of the reasons for
this is that bureaucracies, unlike most high-performing organisations, and indeed most militaries,
have no “up or out” expectation. This sounds brutal but shouldn’t be. There comes a point in
any demanding work environment where some people on a career path have reached a point
from which they will not rise any further. It doesn’t mean they all should all be pushed out;
every organisation needs continuity, experience and an institutional memory. But if you have too
many people who have exhausted their ability to progress AND their ability to make a positive
contribution, they tend to justify their continued presence in the organisation by questioning,
delaying, or obstructing action. They become part of the bias to inertia. There’s an old saying:
for every one person trying to make something happen there are four trying to stop it. When I
recounted this to some Israeli civil servants, they said: “Only four?”

So a lean effective organisation with a bias to action will guard against this. That doesn’t mean
rejecting large numbers as worthless. It means supporting, financially and otherwise, those who
have run out of road to help them find the next thing they can do. As I say, this is a very normal
practice in high-performing organisations. The British Civil Service must embrace it. At the
same time, it must rid itself of the tendency to solve the problem of underperforming officials
by promoting them or moving them sideways, a practice known in America as “turkey farming”.

Parity of Esteem

Second is the need to value operational, commercial, financial and technical skills as highly as
“administrative” skills. I think there is a simple solution to this. Every line department should
have a twin leadership: a policy leader and an operational leader. In the half or so of
departments that have heavy operational responsibilities the permanent secretary would always
be the operational leader; and in the other half it would be the policy leader. At the top of the
Civil Service there would always be a full-time Civil Service Chief Executive, which we
introduced in 2014, and a Cabinet Secretary. The Chief Executive would be the leader for the
whole Service for its operational, financial, commercial and technical functions (which of course
comprise the overwhelming majority of civil servants); and the Cabinet Secretary would
continue to be the leader of the mandarin stream. The all-important role of Head of the Civil
Service would alternate between the Cabinet Secretary and the Chief Executive. Without
something like this, which would be revolutionary in Whitehall terms, I see no prospect of parity
of esteem ever happening. And without parity of esteem there is no prospect that the technical
experts, whose advice on the implementation of policy is essential, will ever be in a position to
make their advice count before, as it is so often the case, it is too late. When it comes to
appointing permanent secretaries, the pattern today is that the Head of the Service and other
mandarins have excessive influence, leading inevitably to appointments being made in the image
of the incumbent mandarin cadre. We really struggled to deliver what every Prime Minister has
wanted, which is to be offered a choice of candidates for permanent secretary roles. This was
eventually and grudgingly agreed by a Civil Service Commission led then, as now, by a former
permanent secretary.

Functional Leadership

Third, the Civil Service needs strong functional leadership at the centre of government. This
sounds boring and technocratic but it is really important. Let me explain what I mean. A number
of functions are common across the whole of government. Every department claims that what
it does is completely unique and distinctive, and of course much is genuinely unique. But most
of their requirements for property, IT and digital, procurement, HR, finance and project
management are common to the whole of government. And even when they are not, you still
need one place where there’s a critical mass of technical expertise. Between 2010 and 2015 we
led from the Cabinet Office an efficiency and reform programme that saved the taxpayer £19
billion in 2014-5 compared with the 2009-2010 baseline, which with previous years’ savings
added up to a cumulative £50 billion. In addition to downsizing the Civil Service by over 20%,
we renegotiated contracts with the government’s biggest suppliers, reformed procurement to
open government contracts to smaller UK-based suppliers, improved dramatically the success
rate of the government’s major projects, exited numerous properties and upgraded much of the
rest, and Britain went from being a country which was a byword for expensive government IT
car crashes to last year being ranked by the UN top in the world for digital government.

None of this could have been achieved without strong leaders of these functions at the centre
of government, themselves technical experts, with strong Ministerial support, backed by spend
and other controls to prevent departments from doing the wrong things. On property the
wrong thing might be taking a lease that looked right for that department but left another with
costly unfilled space. Functional leadership coupled with spend controls would enable the space
to be used optimally with the best result for the taxpayer. On procurement the wrong thing was
every department buying its own commodity consumables. We found that some departments
were paying seven times as much for their printer cartridges as others. The wrong thing was
every department dealing with the major suppliers to government separately so that each one
was taken for a ride separately. The wrong thing was every department running its own website,
or more usually multiple websites, expensively, inconsistently and impenetrably. The award-
winning and world-leading gov.uk, replacing well over 1000 websites, would not have been
possible without the strong functional leadership model. The wrong thing was every part of
government marking its own homework on the management of major projects, so that they were
all said to be doing splendidly despite two thirds running well over budget and timetable.
Establishing the Major Projects Authority to provide consistent oversight, assurance and support
nearly halved the failure rate. The wrong thing was every department and agency having its own
internal audit function that was internal to itself rather than a single service giving the Treasury
genuine insight into how departments were spending taxpayers’ money. The wrong thing was –
and is – every departmental finance director being accountable only to the permanent secretary,
with no serious line of accountability to the centre. This has left the quality and reliability of
basic management information pitifully inadequate, with an inevitable lack of accountability for
departments. Senior business leaders we brought into government as non-executives on
departmental boards were shocked by the poor quality and impenetrability of government
financial and management data.

To the extent that we could we created strong functional leadership. It wasn’t perfect. And we
were heavily resisted by the Treasury which in the main with the exception of Danny Alexander
was at best uninterested in and sometimes actively hostile to our entire programme of efficiency
savings. We couldn’t even persuade them of the need for a powerful financial management
function led by a senior CFO figure, although the Conservative Party had committed to this
before the election. We should have been more insistent on building new functions, albeit using
many of the existing people, rather than trying to adapt the existing organisations. In a
makeshift way we made it work. But we couldn’t make it permanent, and of course many of the
departmental barons hated it. Not Ministers, interestingly, most of whom were perfectly happy
with what we were doing, and wanted to concentrate on what was genuinely the core business of
their departments. But for much of the mandarinate this was an assault on their autonomy and
empires. And what we know about empires is that they fight back.

And boy, are they fighting back! The mantra tends to be: “we definitely want to continue with
the reforms. But they’re now embedded in the departments, and it’s definitely now safe to relax
the central controls.” When you hear those words you know that what they really mean is that
the reforms are embedded six feet under, and that the departments are cheerfully going back to
their old ways. So GDS, which became a model for other governments to follow, including in
the USA and Australia, is becoming side-lined and underpowered. The powerful and
revolutionary idea of “government as a platform” is dead. The financial management function,
for which the Treasury briefly and reluctantly created a dedicated full-time leader, has now
reverted to being “led” from a big spending department by the departmental finance director –
and excellent though he is, it means functional leadership in that arena is dead. Imperceptibly,
inch by inch, with a control dropped here or not enforced there, the old silos and departmental
baronies are re-emerging, with nothing to restrain the old unreconstructed behaviours from
taking hold once more. There will be a substantial cost to the taxpayer in wasted spending, and
to the citizen in service improvement foregone. I have no doubt that a future Civil Service, like
all big complex organisations, needs a vigorous model of stronger, not weaker, central functional

Are there other things needed? Yes but probably beyond the scope of this lecture. British
Ministers’ offices are woefully under-powered compared with those in Australia, Canada or New
Zealand. That’s not only the fault of the Civil Service; Prime Ministers tend to be strangely
negative about empowering their colleagues. David Cameron did support our modest move in
this direction, overcoming strong resistance from the leadership of the Civil Service when we
proposed permitting EMOs, or Extended Ministerial Offices. This model went as far as was
possible within current statutory constraints, but was only a pale reflection of what Ministers in
other Westminster systems enjoy. And in any event this tiny modest advance has now been
reversed. The empire struck back.

One final reflection. I took no part in the Brexit referendum debate, thinking both sides’ cases
wildly exaggerated, and for me the arguments were quite finely balanced. I thought then that a
vote to leave meant some certain short-term downside, with longer term upside opportunity. I
still think that. But if we are to realise those upside opportunities, then as a country we need to
maximise our competitive advantages. One of those can be and must be a Civil Service that is
genuinely the best in the world. One that is lean, proactive, innovative, that makes the most of
the amazing people with a passion for public service who join it, with a bias to action and a

relentlessly truthful and open culture. What I propose here are a few modest steps in that