Politics

No Minister: What “ministerial directions” tell us about how government works

A recent prime ministerial decision to overrule a senior civil servant sheds light on an instrument that deserves to be better understood

January 20, 2021
Boris Johnson issued a ministerial direction not to settle with former adviser Sonia Khan after she brought an unfair dismissal claim. Photo: Aaron Chown/PA Archive/PA Images
Boris Johnson issued a ministerial direction not to settle with former adviser Sonia Khan after she brought an unfair dismissal claim. Photo: Aaron Chown/PA Archive/PA Images

In August 2019 Sonia Khan, special adviser to the then-chancellor Savid Javid, was marched out of Downing Street with a police escort following a clash with Dominic Cummings. This incident was revealing about the government’s impatience with the rules of good conduct that normally apply. The case has also brought to light the little-known area of “ministerial directions.” We now know that Boris Johnson specifically instructed the former chief executive of the civil service John Manzoni not to settle Khan’s unfair dismissal claim, even though Manzoni—and in all likelihood, his legal advisers—thought doing so would better ensure value for money than contesting it further. The recent story raises important questions about how Whitehall operates.

What are ministerial directions?

A ministerial direction is a form of protection for civil servants. If permanent secretaries (the most senior civil servants) believe that what the minister tells them to do goes against standards of public spending, the official can put their objection in the public domain (since 2011) and require a letter from the minister, which gives them cover. Ministerial directions are thus formal instructions from ministers telling their department to proceed with a proposal.

This structure has been around since 1990, and is based on the proposition that each permanent secretary is also the “accounting officer” for their department (a practice developed in 1872). He or she is the person whom parliament (through the Public Accounts Committee) would call to account for how the department spends its money. In the event of a direction, the formal accountability for the decision is pushed up to the minister.

Civil servants have a duty to seek a ministerial direction if they think a spending proposal breaches any of the following criteria set out by HM Treasury:

  • Regularity—if the proposal is beyond the department’s legal powers, or agreed spending budgets;
  • Propriety—if it does not meet “high standards of public conduct,” such as appropriate governance or parliamentary expectations;
  • Value for money—if something else, or doing nothing, would be cheaper and better;
  • Feasibility—if there is doubt about the proposal being “implemented accurately, sustainably or to the intended timetable.”

Many disagreements are sorted out in informal discussions prior to seeking a direction, as both sides may want to avoid a spat becoming pubic. But it is thought that the National Audit Office rather welcomes permanent secretaries seeking a direction and thinks it should happen more often. What we do not know is how many directions have been avoided through skilful persuasion by accounting officers! To ask for one may be career limiting for the permanent secretary, although ministers rarely seem to suffer—even if it turns out their direction should not have been given.

Use during the pandemic

The vast majority of the 89 ministerial directions issued since 1990 (39 of them since the 2010 election but none between 2011 and 2014) have been on value for money grounds. They are usually given at the rate of two or three a year. One of the first was the famous incident involving Tim Lankester and the Pergau Dam.  Two of them were issued on notoriously wasteful projects, Kids Company and the London Garden Bridge (one of Johnson’s Big Ideas). The imbroglio of Kids Company is still being worked through and was wider than a value for money issue; the Garden Bridge never materialised.

There have, however, been no less than 14 issued in relation to coronavirus. The speed required in responding to the epidemic inevitably meant that there would be more, but this number is still unexpectedly high.

Benoit Guerin, a senior researcher at the Institute for Government, notes that “ministerial directions—once seen as a ‘nuclear option’—have become much more common in the last year, as government worked to support businesses during the coronavirus pandemic.”

The Khan direction

A further worrying feature about the Khan letter was the delay in the matter being made public (ministerial directions on further funding for the Garden Bridge and British Steel indemnity were also slow to come to light). The Khan direction was the longest to reach public attention, taking nine months—the direction having been given to Manzoni on 3rd March 2020.

Manzoni wanted to continue negotiations with a view to settling the case on the grounds of the cost of fighting the claim. The instructions from the Prime Minister however were clear they should not seek to settle.

The direction included this curious sentence: “The legal position is clear that the Prime Minister can withdraw consent for the appointment of any special adviser.” That is indeed clear, but it is equally apparent that this does not override the legal protections Khan had in the law of the land. The case was eventually settled in December 2020 for an undisclosed “five figure sum,” its timing roughly coinciding with Dominic Cummings being asked to leave his position. The likelihood is that it would have been a lower settlement if such a pig-headed response had not been given earlier.

OneWeb and the Towns Fund

“The weaknesses of the accounting officer framework are most obvious when policies or projects that are clearly flawed go ahead without a direction being requested,” notes Guerin.

The system can however work effectively. For instance, when the government joined a consortium bid of $500m in respect of OneWeb, a satellite company that was in Chapter 11 bankruptcy in the US, the then-acting permanent secretary of the Department for Business, Energy and Industrial Strategy wrote to the then-secretary of state Alok Sharma:

“The purchase—both in its scale and the fact that the company is early in its journey towards a first-of-a-kind satellite constellation and generating revenue—is unusual for government… Having reflected carefully on the information provided, I have concluded that whilst there may be a commercial case for investing alongside other commercial investors if you accept advisors’ assessment of OneWeb’s business plan projections, as a standalone high-risk investment with a possibility that the entirety of the investment is lost and no wider benefits accrued, I cannot satisfy myself that this investment meets the requirements of Value for Money as set out in Managing Public Money.” Sharma duly gave the direction to go ahead with the successful bid. It will be instructive to follow the fate of that company and see whether Sharma is held to account.

However, the National Audit Office and the Institute for Government have pointed to several surprising examples where large-scale government projects, about which civil servants had clear concerns, did not lead to a request for direction. Those include the National Programme for IT in the NHS (£10bn); the FiReControl project (£469m); and the rollout of Universal Credit.

One such recent example is the Towns Fund—a scheme designed to boost economic growth in certain areas of the country—where two ministers put forward each other’s constituencies for payments from the fund when they were not anywhere close to being priorities on the list which had been drawn up by civil servants.

Vigilance is needed

“The Ministerial Direction system is good at ensuring that public spending is legal and within budgets, but it doesn’t do so well at areas that are not so clear cut, like the value for money or feasibility of policies or projects,” argues Guerin. Like so many things, it can be used inappropriately, and we must be vigilant. In the Khan case, the public would have been better served by an early settlement, which would also have avoided a public servant wastefully having to fight the government through the courts.