Is the civil service reformable?

A bi-partisan approach to reform is currently adopted, but can it work?

Ed Straw
11 February 2015

Finally, some Westminster politicians have concluded that real reform of the Civil Service will not be realized through party politics. A bi-partisan approach is being adopted. But, will it work? Can it work without the wider transformation of the whole system of government? These notes argue, no.

1) The Civil Service exercises considerable power but nowhere in any definition of democracy is a quasi-independent self-appointed unelected government administration legitimised. The Whitehall Civil Service is amongst the most centralising organisations in the developed world.

2) The existing Northcote-Trevelyan civil service model of 1854 was designed for the parliamentary, political and legislative roles of government, and not for the delivery of the vast array of public services costing nearly half the country’s GDP, services that largely did not exist in 1854, nor in their current scale or complexity. Neither was the NT model designed for the modern government roles of dealing with global finance, corporate power, media power, intergenerational conflicts of interest, or supra national governments.

3) Who said this: ‘it is against the system, and not against individuals that we raise our objection’. Answer: Sir Stafford Northcote and Sir Charles Trevelyan to the Lords of HM Treasury in 1855, when introducing their reform of the Civil Service; it is not a people problem but a system problem.

4) I kicked off the modern round of attempts at civil service reform in 2004 with The Dead Generalist for Demos. Jonathan Powell took this up, Tony Blair spoke at the launch of the Cabinet Office plan for The Professionalisation and Specialisation of the Civil Service. This was being taken seriously and backed politically from the highest level. The humbly titled First Division Association, the lobbying organization for Whitehall civil servants, had its director – every bit a spin doctor as any ‘politicised’ appointee – brief against me. But the plan went into action. And what happened? Deckchairs were rearranged alongside much fiddling, further reform reports published and ten years later the generalist culture still rules. After being laid upon for a while, like memory foam the civil service returns to its original shape.

5) A former minister explaining why politicians avoid real reform of the Civil Service, said “they’ll find a way to get back at you.” Indeed they do: by stirring up yet another SPAD spat – should ministers really be allowed their ‘politicised’ special advisers – with the news media ever ready to use any stick to poke the politicians; and by leaking against any ‘disloyal’ ministers.

6) My realization why I was wasting my time back in 2004 occurred recently when discussing civil service reform with a senior peer. His answer: the existing Northcote-Trevelyan model dating from 1854. Gosh. Why do constitutionalists stick with Northcote-Trevelyan I wondered? Because civil service reform in isolation results in handing more power to the executive – the government of the day. It has far too much already. So the lesser bad is retained. Everyone knows the civil service is a major part of the problem, but so too is the political establishment, the Lords, the public service regime, centralisation, the absence of real local government, receding democracy, and state institutions. This is a problem of the whole system, not one part alone, and its solution will only be found through analysing it as one.

7) For example, the Public Accounts Committee is widely applauded for ‘nutting’ the top person for another weapons procurement failure. Surely this is working? But go back and read NAO reports from the 60s and experience déjà vu, except not only do you think you’ve seen this all before, you have, many times – over-budget, late, and well below specified performance. Nothing has changed. The ‘nutted’ permament secretaries or agency heads return to their departments wearing their scars with pride, heroes for defending them and the status quo. In any well-functioning system of government, a PAC is essential. How can its entirely necessary and well-aimed reports be turned into lasting change in its targets? The answer is that the government has to have control of all the other motivators of behaviour – appointments, recruitment, pay, promotion, systems, etc; and over just who in a democracy is responsible for delivery and who has the levers to make it happen. Without these, public bodies have no imperative to change radically. Nothing is making them change. They just absorb, and continue much as before. In organizational theory this is termed alignment where all of the motivators and levers are pointing and working in the same direction. It’s key. Nearly all of government and public service bodies in the UK are significantly out of alignment.

8) What is political and what not? Briefly delivery is political – why else do we elect governments other than to produce beneficial change on the ground? Feedback of the results of all policies, programmes, legislation is not. But the system we have runs arcane debates about a tiny population of ‘politicised’ SPADs and the sanctity of civil service responsibility for delivery, whilst allowing most results to be politicised by either being left under the carpet or spun in any direction. Maths is not political.

9) Feedback is essential – slowly the Office for National Statistics has acquired more independence and more scorekeeping powers, but for a fraction of what government does. NICE is a scorekeeper but not independent. The NAO has its merits but has developed few/no new approaches to value for money in government in the last 30 years. Why are there only three separations of powers, when so clearly more are needed to make modern government work? Surely feedback is a prime candidate for a fourth separation? I’ve termed it the Resulture.

10) With effective feedback, we can see how power could be reallocated and wider public checks and balances installed such that government can take on responsibility for delivery, and appoint executive ministers, without becoming yet more dominant. The civil service can then be split in two: the independent NT model for the parliamentary, political and legislative roles and for a part in policy vetting, and the rest under the control of the executive.

11) Government is an organisation. Its problems and solutions are to be found in organisational theory not traditional political and constitutional analysis. But the influencers in and around government – academics, think tanks, political activists, journalists – are either dependent on the system for work (I was) or newsflow, or have not the education or experience in this field nor appreciate what it can do for governments of every form, including the EU. It is as if the analysis has not gone beyond Newton’s Laws. A presumption exists in every article, speech, discussion, and book that the body of constitutional knowledge essentially cast in ancient Greece is whole. Plenty of debate within this body but none outside. We need the concept of constitutional extension, built on organization and systems theory.

12) Ten years after my first attempt at reform, I’ve published a book, which includes substantive sections on the transformation of the civil service. I think you’ll find it informative:

Who is bankrolling Britain's democracy? Which groups shape the stories we see in the press; which voices are silenced, and why? Sign up here to find out.


We encourage anyone to comment, please consult the oD commenting guidelines if you have any questions.
Audio available Bookmark Check Language Close Comments Download Facebook Link Email Newsletter Newsletter Play Print Share Twitter Youtube Search Instagram WhatsApp yourData