Is it now time to formalise the processes for scientific advice in emergencies?

PDF Version

In mid 2020, at the height of the pandemic, the House of Lords Science and Technology Committee conducted an enquiry into the Science of COVID-19. I submitted written evidence suggesting that the relationship between SAGE and the Government should be put on a more formal footing. In doing so I was encouraged by a number of senior figures including members of SAGE and of the Lords Committee itself, but in the event my suggestion wasn’t taken up.

The question of whether the scientific advisers “cosied up” excessively with Government, or whether close working was essential in the circumstances, will run and run. My own view is that a more formal relationship or protocol could help in the future, and that if we are currently in a period of relative calm, this may be the time to consider it.

How could a formal protocol help? I first drew up such a protocol when I came into the post of Chief Scientific Adviser to the Home Office in the aftermath of the resignation of most of the members of the Advisory Council on the Misuse of Drugs. Certainly during my own time in the Department, the protocol functioned very well in delineating the “scientific” and the “political” aspects. I remember very well the late Andrew Miller MP, then chair of the Commons Science and Technology Committee, saying to me:

“The way scientific advice should work is that the Government should be free to reject the advice, as long as they give written reasons for doing so.”

I gently pointed out to Andrew that this was exactly how the ACMD protocol worked, and that precisely this had happened on a particular issue to do with the drug khat. Andrew asked me to point him to the Home Secretary’s published letter on the matter and the next time I saw him he said:

“I saw Theresa recently and I told her that I thought that was a rather good letter”!

The details of a written protocol would be the subject of discussion and negotiation, which would itself be a good opportunity to discuss publicly the way that the system should work, but here are some possible elements; more details of issues that could be considered are given in my original evidence document.

  1. Membership of the committee. This could include not just a list of members but (perhaps more importantly) some indication of the range of expertise covered. This would both justify the authority of the committee and also make it clear what were the limits of its specialist competence.
  2. The publication of advice and supporting evidence. In my view it would almost always be appropriate for the advice and for all the as yet unpublished supporting evidence to be put into the public domain simultaneously with the advice being proferred to Government. This would enable wider evidence-based scrutiny of any particular proposal or advice.
  3. The way the advice will be handled. I would go with Andrew Miller’s suggestion that the Government should undertake to consider the advice and to publish reasons if they choose not to accept it. This would mean that the actual decision-makers (the Government) would not be constrained by advice they had received, but on the other hand would enhance transparency and accountability if they decided to use other considerations.
  4. The form of the advice. Obviously the advice would be based on cited or appended evidence that the committee had considered, or on the acknowledged, preferably documented, expertise of committee members. It could be advisable to have a pre-specified range of ways of describing the strength of evidence, perhaps along the lines of that used by the Intergovernmental Panel on Climate Change. Another important issue is the way that minority or dissenting views will be represented.

“No military plan will survive contact with the enemy.” Similarly, there’s little doubt that in a serious emergency any agreed protocol for SAGE might need to be modified in the light of circumstances. Nevertheless, setting out principles and processes in the way I’ve set out could be a very good starting point.

About the Author

Sir Bernard Silverman FRS

I am a statistician whose research has ranged widely across theoretical and practical aspects of statistics. The focus has been on computational statistics, researching the ways that computing power has changed our ability to collect, analyse, understand and utilise data. I have collaborated in many scientific fields, and with various areas of industry and government. After an academic career, I spent seven years as Chief Scientific Adviser to the Home Office. I now work freelance, with roles including research, policy, consultancy, and expert advice.


Twitter: @silverman_b