The Westminster Model - Recent Developments

Ministers are frequently accused of being too ready to seek political advantage by announcing impossible or badly thought-through policy objectives, or by allocating insufficient time and resources to otherwise achievable policy objectives.  Jill Rutter’s half jest summarised the underlying issues rather well:

Civil Servants say to Ministers that “We won’t tell you it can’t be done if you won’t sack us when it is not done”.  Maybe it is time we recognised that this constitutional pact has run out of road?

Select Committees and others have responded by recommending that it should be made easier for senior civil servants to challenge Ministers' policy decisions in the same way as they have for many years been able to challenge a Ministerial spending decision.  (Officials can ask for a formal written ‘Direction’ if the spending appears to be irregular, or improper, or to represent poor value for money.)

This webpage summarises five such recommendations:

The last three of these recommendations have been acted upon.

It is important to note the importance of the word ‘encourage’ in the above list.  Nothing in the recommendations would stop Ministers from taking whatever decisions they felt appropriate as long as they were willing to account to Parliament for their decisions.

1 - Policy Directions

The Institute for Government's 2011 report Making Policy Better speaks for itself:

We propose adding a fourth Ministerial Direction:-  poor policy process, where the Accounting Officer (usually the Permanent Secretary) is not satisfied that the fundamentals of policy making have been adequately observed.  This recommendation builds on current practice.  The Treasury currently recommends that AOs should exercise judgement on when they need to “take a principled decision”.

One of the standards they should use to make this judgement is whether “clear, well-reasoned timely and impartial advice” has been provided, and whether the decision is in line with the aims and objectives of their organisation – both of which relate closely to our proposed fundamentals.  Furthermore, this new criterion could be seen as an extension of the current value for money criteria, since there is a good case that a poorly made policy will provide poor value for money.

The Treasury is already considering extending AOs’ responsibilities to ‘feasibility’, but our proposal would also embrace the wider way in which policy is made.

Where the minister wanted to override the objection, they could do so, but would need to give a ‘policy direction’.  Such a change would sharpen the incentives for both parties.  AOs would act in the knowledge that they could be held to account by the departmental select committee for the quality of the policy process, whether or not a direction was issued.

Since the direction would be sent to the relevant select committee and published on the department’s website, the minister would be publicly accountable for taking action despite civil service concerns.  The point of extending the AO remit in this way is not to ensure more directions are issued, but to make clear to officials, and in particular the head of department, that they must take responsibility for good process.  By extension, this will give ministers a stronger incentive to observe good policy process.

I have emphasised the final sentence above because it indirectly explains why there was very little chance that Ministers would accept this constraint on their ability to seek political advantage by making bold promises.  Indeed, I am not aware of any later consideration of the idea.  However, as noted by the IfG, the Treasury were already examining the introduction of Feasibility Directions (see further below).

2 - Procedural Directions

In parallel with all the above, the Public Administration and Constitutional Affairs Committee supported a proposal from the Better Government Initiative informed by Chilcot’s criticisms of Blair government decision-making before the Iraq War.  The suggestion was that

One example, had this mechanism already existed, might have been Tony Blair’s failure to circulate pre-Iraq War legal advice to Cabinet colleagues.  Another might have been a challenge to David Cameron's instruction that officials should not undertake contingency planning for a ‘leave’ vote in advance of the Brexit referendum.

The Constitution Society supported the proposal and suggested that a revised and extended version of the Cabinet Manual should be subject to Parliamentary approval.

This attempt to fetter Ministers' discretion was, of course no more welcome to the Government than had been the IfG's suggested Policy Directions (see above).   Even so, the quality of the Government's two formal responses was disappointing.

Rather depressingly, since the proposal was clearly directed only at departures from established procedures, not the merits of a policy, the initial response completely misrepresented it as being “for a formal Ministerial direction to be given, if Ministers decided to go ahead with a policy against the advice of officials”. Having set up this misrepresentation HMG then (quite rightly) rejected its own foolish proposal.

The second response wasn't much better. The rejection of the case for a procedural direction was based on a distinction between accounting officers’ direct responsibility to Parliament and permanent secretaries’ responsibility to ministers and the prime minister for the conduct of departmental business. But this was a false distinction since permanent secretaries have a duty to Parliament as accounting officers for the efficient conduct of their departments.  

Follow these links to read:

3 - Feasibility Directions

These Directions were introduced in 2011 and allowed officials to require Ministers to direct them to proceed with projects even if officials had put on record their doubts that the project's objectives could be achieved either at all, or within the timescale and resources stipulated by the Minister. 

The Treasury defined feasibility as follows:

Feasibility often overlaps with value for money and/or propriety. The judgement to be made is whether government has the ability to carry out the proposed policy effectively and credibly. Precedents, market testing and pilot studies can give confidence that a new policy or proposal will be feasible. Conversely, warning signs include novelty, high administration costs, high error rates and significant compliance costs. Where there is doubt about the quality of administration, the proposed course may well also be inefficient or improper.

The deliverability assessment of a major project is also an aspect of feasibility. Where delivery concerns have been raised (for example, in a gateway review), the full accounting officer assessment would normally be expected to note those concerns, and reflect any mitigating actions taken or planned as a result. Although the accounting officer might expect to be notified of these concerns as soon as they are raised, it is preferable for the written assessment of feasibility to be prepared once any mitigating actions have been taken, so that the accounting officer to can also take those into account.

Whitehall watchers awaited the first feasibility direction with great interest.  Would it be seen as evidence, yet again, of Ministers unrealistic expectations, driven by short term political considerations? Or would it be evidence, yet again, of the need for Ministers to be able to override their cautious, unimaginative and unambitious civil servants?

It was quite telling, therefore, that a 2016 National Audit Office report asserted that Accounting Officers "appear to lack confidence to challenge Ministers where they have concerns about the feasibility or value for money of new policies or decisions, not least because standing up to Ministers is seen as damaging to a civil servant’s career prospects". 

But nothing much appeared to change.  The Times reported, in early 2019, that the retiring head of the National Audit Office , Sir Amyas Morse, was concerned that the balance of power between ministers and senior civil servants had shifted, with officials increasingly unable to challenge bad decisions.

“I still don’t think we’ve sorted out the question of the interaction between the political agenda and delivering good results and value for money,” Sir Amyas said. “There’s pressure to do things too quickly or to announce very high-profile world-beating projects. Allowing ministers to have a say in the appointment of senior officials has led to a position where ministers have a great deal of power over their civil servants. That’s unfortunate. They’re intelligent people. They understand that the consequences of disagreeing with a minister are likely to be pretty ugly.”

The first Feasibility Direction had not appeared until 2018 when a minister took responsibility for the risks associated with accelerated introduction of new 'T Level' exams.  This was a perfectly sensible and uncontentious use of the mechanism. 

A small number of further feasibility directions were issued by the Business Secretary as his officials rushed to support the private sector during the 2020 COVID-19 crisis. 

4- Senior Responsible Officers

The 2013 introduction of SROs looked more promising.  SROs were to be personally accountable, including to Parliament, for the delivery of major projects such as the National Cyber Security Programme.  The key principle had until then been that civil servants who gave evidence to such committees do so “as the representative of the Minister in charge of the Department and subject to the Minister's instructions”.  But MPs could now, for the first time, question civil servants about their delivery of major projects such as the (delayed) introduction of Universal Credit.  The new rules now provided that “Senior Responsible Owners (SROs) for Major Projects” are “expected to account for and explain the decisions and actions they have taken to deliver the projects for which they have personal responsibility”.

It was hoped that newly appointed SROs might be concerned to ensure – before accepting their appointment – that they were not suffering from appraisal optimism, and that their project was properly resourced and had sensible timescales and objectives.  This would reduce the chances of their having to account to their Permanent Secretary and Parliament when things went wrong.  And it would ensure that a senior official – the SRO – was forced to challenge Ministers if a major project were being established without proper resources etc.  But it could work very badly if SROs were to do what officials had done in the past, which was to accept that Ministers are entitled to demand rapid action with limited resources, and so sign up to achieving what they privately believe to be unachievable. 

In practice, little at first appeared to have changed.   SRO appointment letters were little more than that.  They specified neither the programme's objectives nor its resources or timescales.  And most departments at first decided to appoint very senior staff as part-time SROs, rather than nominate those officials who were truly responsible for key projects. The SRO for the National Cyber Security Programme was for instance told that he would need to devote only two days a month to the role:-

“I am writing ... to confirm your appointment as Senior Responsible Owner (SRO) of the National Cyber Security Programme ... This will be a part time role which requires two days per month. As SRO you have personal responsibility for delivery of National Cyber Security Programme and will be held accountable for the delivery of its objectives and policy intent; ..."

But SROs were strengthened by the introduction of Accounting Office Assessments - see further below.  The Universal Credit SRO appointment letter, for instance, requires the SRO to prepare an Accounting Officer Assessment (see further below) 'if the programme might depart from the four standards (regularity, propriety, value for money and feasibility), or from the agreed plan – including any contingency – in terms of costs, benefits, timescales, or level of risk'. It is also firmly linked to the Business Case, so the SRO is personally accountable for delivering the intended economic and net present values.

Indeed, the various publicly available Universal Credit appointment letters show that the relevant SRO was able to renegotiate the programmes timescales and be clear to Parliament what the reasons are.  This will have followed private negotiations with his Ministers.  So this particular Minister/official dynamic appears to be working very well. 

5 - Accounting Officer Assessments

As from 2017, and following a Public Accounts Committee recommendation, the Treasury announced that ‘Accounting Officers should personally approve, in advance, all significant initiatives, policies, programmes and project’ and so be able to provide assurance to Parliament that those activities provide value for money and are feasible etc.  The guidance went on to say (emphasis added):

The analysis should consider the issue in the round. A ministerial policy decision cannot be sufficient justification alone for proceeding. The accounting officer’s job is to try to reconcile ministers’ policy objectives with the standards for use of public funds.

The full accounting officer assessment should provide a frank examination of the key issues including any sensitive issues. It should address the essence of the policy which is being delivered, its purposes and its prospect of successful delivery or implementation. It is therefore not usually published in full, but is shared with the Treasury.   A summary of the key points from an accounting officer assessment of a major project should however be prepared and published. 

Will It Make a Difference?

it will be interesting to see whether other SROs are able to develop the positive relationship with their Ministers that has been achieved in DWP.  Ministerial Directions were once regarded as nuclear weapons - more effective in the silo rather than launched.  But they have come to be seen as a grown-up way of allowing Ministers to account for political decisions to override strict value for money criteria.  SROs’ ability to prepare Accounting Officer Assessments are similarly unlikely to be used very often, but they should, over time, help curb Ministers' desire to order officials to achieve challenging objectives be achieved within impossible timescales and with inadequate resources. 

The open question is whether the threat, so to speak, of SRO appointments and Accounting Office assessments will discourage ministers from announcing badly thought through policy decisions and projects such as Prime Minister Cameron's 'Big Society' or Prime Minster Theresa May's social mobility agenda.  Both aspirations were sincere but there was no organisation or institutional weight behind them. 

(A senior minister was asked to summarise the 'Big Society' concept in a single sentence.  He couldn't - and his questioner, Peter Hennessy, recalled that he had been told that it was more "a state if mind" than a specific idea!)

The signs so far are not promising, given what has happened so often during the initial response to the COVID-19.  Officials are clearly not obliged to prepare an AO Assessment immediately a minister indulges in some blue sky thinking.  And the Treasury’s guidance allows a bit of wriggle room later on:

Often, big intricate decisions have long lead times. In such cases, it is good practice to make the accounting officer assessment in principle at an early point, firming it up at suitable strategic points as the policy or proposal is developed. This makes for orderly evaluation of the key features of the policy, with no surprises at the final decision point. Apart from providing time to redesign a policy or proposal, early assessment may flag up how the proposal can be better designed to meet both ministers’ and parliament’s requirements, or whether there is a for a ministerial (or board) direction, particularly when proposed spending is imminent or an existing spending stream no longer complies with the four accounting officer standards.


The Achilles heel may be the private nature of the policy assessments. 

Parliament has been given a role in monitoring ‘major projects’: 

Accounting officers who have considered an assessment for a project in the Government’s Major Projects Portfolio (GMPP), in line with this guidance, and approved it, should provide to Parliament a summary of the key points from the assessment which informed their judgement. Accounting officers may choose to publish similar information from assessments made in other circumstances at their discretion, but there is no requirement to do so.

‘Policy’ AO Assessments, however, will not automatically be shared with Parliament, in which case they would not appear to be different in substance from the traditional Mandarin warnings that ministers’ policy proposals are ‘brave’: – ‘courageous’ even!  So it could all come down to the energy and inquisitiveness of MPs and Select Committees.  They could – perhaps supported by the National Audit Office and the media – start insisting on seeing AO Assessments.  If so, we could see a significant improvement in the way this country is governed.


Martin Stanley

Spotted something wrong?
Please do drop me an email if you spot anything that is out-of-date, or any other errors, typos or faulty links.