Andy Burnham’s Number 10 should have fewer special advisers

Keir Starmer had 42 Special Advisers in Number 10. The expectation is that Andy Burnham will recruit a similar number. In this post, Robert Hazell argues that to avoid the turmoil of his predecessor Burnham should aim to have fewer, more heavyweight Special Advisers, working more closely with the Number 10 civil servants.

The newspapers are full of speculation about who will be Andy Burnham’s Chancellor, and who will be the other senior ministers in his Cabinet. But just as important to the success or failure of a Burnham government will be the quality of the staff appointed to support him in Number 10. One of the reasons why Starmer’s premiership failed was a series of poor appointments, with four chiefs of staff and five directors of communications in less than two years; and Number 10 becoming a byword for turmoil and infighting in place of the discipline and stability which Starmer had originally promised. From Boris Johnson’s premiership onwards, there has been too little continuity of core officials (in particular the Principal Private Secretary, who runs the Prime Minister’s personal office in Downing Street), and too much of the running made by (often junior) Special Advisers (Spads).

Dysfunctional operation in Number 10

Until the arrival of Darren Jones as the PM’s new Chief Secretary last September, Number 10 under Starmer had become notorious for U-turns, indecision, negative briefing and high staff turnover. Part of the problem, as many have observed, was the lack of an overarching vision and clear sense of direction, which could only be provided by the PM leading from the top. But that is not the whole explanation. Number 10 also suffered from a basic lack of organisation, and unclear lines of authority, with senior individuals competing for the PM’s ear.

Whitehall departments found the centre was a black hole, with important questions being submitted for a decision and nothing coming back for months. Meanwhile there was infighting between different teams of Spads, even over small issues, leading to repeated delays on things like public appointments. (This was not new: the Public Appointments Commissioner William Shawcross had strongly criticised Number 10 for delaying appointments under Rishi Sunak). There were also serious lapses of judgement, particularly over the appointment of Peter Mandelson: not just the original mistake, but the way Number 10 has managed to keep the issue repeatedly in the headlines when it should have been working to let the matter drop.

Who are the Number 10 Spads?

Who are the Spads in Number 10?  We know who they are from the annual list of names published by the Cabinet Office; or rather who they were, since the latest data release published last July gives a list of all the Spads in post on 31 March 2025. Of the 130 Spads on the list, 42 are recorded as working in Number 10. Some names are familiar, such as Morgan McSweeney, then Chief of Staff, and Jonathan Powell, National Security Adviser. But the majority are not well known, even to seasoned Whitehall watchers.

Their name and salary is the only information we are given. Crucially, we are not told what they do: there is no indication of who works on policy and who works on comms, who marks which department, who liaises with the party in parliament, which teams they work in, who reports to whom. If Andy Burnham could do one thing to help sort out Number 10, it would be to publish brief job descriptions for all the senior staff.  This is something the Committee on Standards in Public Life (CSPL) recommended over 20 years ago, for all Spads, when it said in its 2003 report Defining the Boundaries:

An annual statement should be made to Parliament setting out: (i) the total number of paid special advisers employed in the year; (ii) their names; (iii) the Ministers for whom they work or have worked; (iv) their particular roles and areas of responsibility; (v) the total salary cost by department; (vi) comparison figures for earlier years.

That recommendation was partially enacted in section 16 of the Constitutional Reform and Governance Act (CRAG) 2010, which requires the Minister for the Civil Service to publish an annual report about the number and cost of Spads; but it leaves out CSPL’s (iv) their roles and responsibilities, and (vi) comparison figures for earlier years. Had comparisons been included for earlier years, the figures would have looked like this:

Average number of Special Advisers recorded per year in Number 10 over the years of each premiership

Source: Cabinet Office annual data releases on numbers and cost of Special Advisers. For a detailed breakdown of the different units in Number 10, see Table 2.1 at pp 21-22 here

The numbers have risen fourfold in the last 30 years, with a big jump under Tony Blair, and a further smaller jump under Johnson and Sunak, leading to the new norm of over 40 Spads in Number 10. Readers can judge for themselves whether the increased numbers have led to an increase in the overall effectiveness of Number 10. But regardless of the numbers, there is a genuine problem in how Spads are recruited. Politicians like to appoint people they know. Burnham will not know policy experts in many fields, and must rely heavily on others. James Purnell, tipped to be his Chief of Staff, has only weeks to assemble a new team: how will he set about doing so?

Why Spads matter

The appointments are crucial. Spads are a fixture of modern government. In the book I wrote with Ben Yong, Special Advisers: Who they are, what they do, and why they matter (2014), I explained why they matter in the following passage:

Our interviewees were all agreed that special advisers had become indispensable to the working of modern government … This was for many reasons, but the single most important one was ministerial overload. The demands on ministers are so great that they cannot be everywhere at once and they need more people to whom they can delegate. These must be people whom they can trust to know their minds and represent them in their absence. Absences include the growing number of ministerial trips to Europe and further afield. The greatest value of special advisers is to be the minister’s alter ego, giving a steer to officials in the department, briefing the press and meeting with outside interest groups. Although these functions overlap with those of officials, special advisers have a particular authority because they have been personally chosen, and through regular proximity have far more opportunity to get to know the minister’s mind … (pp 194-95).

The Number 10 Policy Unit

I wrote that with Secretaries of State in mind, but a fortiori the same logic applies to Prime Ministers. But how can over 40 Spads in Number 10 claim to know the PM’s mind? In the same chapter of the book, we included a section on the Number 10 Policy Unit. Most of our interviewees agreed that the ideal model was a mix of Spads and civil servants; there was less agreement about the ideal size, with numbers ranging from eight to 20. To match every Whitehall department, the Policy Unit would need at least 15 people. But our interviewees were agreed that the quality of the people was far more important than the quantity, and that the Policy Unit did not need one person per department:

Prime Ministers should only have a few priorities if they want to achieve anything. So there should be really heavyweight figures for each of the four or five priorities that the Prime Minister has. Then you can have younger, less significant figures, or civil servants, covering three or four departments. So if the Prime Minister doesn’t really care about DEFRA, energy, DfID or whatever, you can have people handling a number of those together. But if he’s really interested in health reform, or education, then you need to have big figures doing just that (Former senior adviser, No 10)

The other reason for keeping the Policy Unit small is so that its members can genuinely claim to speak with the voice of the Prime Minister:

Small size was, I think, crucial to effectiveness. The whole team had sufficient contact with the Prime Minister to give them credibility throughout Whitehall and to ensure they understood what was in his mind. There was no need for internal competition for access to him or for power struggles for leadership on particular issues (Former member of John Major’s Policy Unit).

Recruiting the new team

In our book Ben Yong and I recommended that ideally these positions should be advertised, to ensure appointment on merit through open competition, rather than simply through nepotism or party loyalty. But that is impractical when time is of the essence: Burnham may become Prime Minister in less than three weeks’ time, and needs a team in place. The default could be to retain many of the existing Spads in Number 10. The Times reported on 27 June that they ‘have been told to speak now if they wish to be considered for keeping their jobs under a Burnham administration’.

While some continuity would seem desirable, for the knowledge they have gained about the government machine, retaining too many would be a mistake. And James Purnell could take comfort that he does not need to find 40 new recruits: finding 20 would be a lot easier, and quality is more important than quantity. One other way of ensuring quality, to hark back to an earlier recommendation, would be to publish brief CVs together with their job descriptions, so that people can see how well qualified they are for their new role. In our book we suggested this was far more important than publication of individual salaries, which gives rise to unnecessary friction and bickering.

We proposed putting publication of the pay of Spads on the same basis as their equivalents in the Senior Civil Service – namely, to publish their Pay Bands (PB 1, 2, 3 or 4), but no more than that. This would give sufficient indication of their seniority and experience. But alongside that small reduction in transparency, we proposed an increase by publishing details of special advisers’ roles. That is what the Public Administration Select Committee had called for in 2001, and the government gave no convincing reason why their roles should not be disclosed. Spads complained to us about some ministers’ lack of clarity in what it was they wanted. Having job descriptions would help to sharpen up ministers’ thinking, increase their accountability, and demystify what spads do. It is another respect in which they are unnecessarily depicted as people who live in the dark.

Conclusion

One further reason why Purnell might want to publish details of Spads’ roles is Burnham’s plan to have a Number 10 in the North, headed by Caroline Simpson as his new Deputy Chief of Staff. If part of the Number 10 operation is based elsewhere, it will be important for Whitehall as well as outsiders to know exactly how functions are divided between Downing Street and Manchester. The best way to do that will be a clear organogram of each office, together with brief job descriptions for all the Special Advisers and the senior officials.

Burnham should also think hard before getting rid of Darren Jones or Jonathan Powell. Between them they have an invaluable wealth of experience. Powell’s goes all the way back to his time as a highly effective Chief of Staff to Tony Blair, and more recently as Starmer’s National Security Adviser. Darren Jones as Chief Secretary to the PM has done a lot to steady the ship in Number 10; he knows which of the current Spads are effective; he would provide continuity, and a bridge to the Parliamentary Labour Party and to Starmer loyalists who may still harbour doubts about the new regime.

About the author

Professor Robert Hazell is the founder of the Constitution Unit and served as Director from 1995 to 2015.

Image attribution: Contains public sector information licensed under the Open Parliament Licence v3.0.